José Luís de Almeida - Universidade do Porto

Transcription

José Luís de Almeida - Universidade do Porto
Caracterização Psico-social de Voluntários Saudáveis
em Ensaios Clínicos de Farmacologia Humana e sua Correlação
com os Resultados de Tolerabilidade e Farmacocinética
José Luís de Almeida
Caracterização Psico-social de Voluntários Saudáveis
em Ensaios Clínicos de Farmacologia Humana e sua
Correlação com os Resultados de Tolerabilidade e
Farmacocinética
José Luís de Almeida
Título:
Autor:
Edição:
Produção gráfica:
Depósito-legal:
ISBN:
Ano:
Capa:
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
José Luís de Almeida
De autor
Tipografia Nunes, Maia
#########
#########
2008
Nuno Almeida, sobre imagem de Jared Mechaber (reprodução autorizada)
II
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Dissertação de candidatura ao grau
de Doutor apresentada à Faculdade
de Medicina da Universidade do Porto
III
Artigo 48º, § 3º: «A Faculdade não responde pelas doutrinas expendidas na dissertação».
(Regulamento da Faculdade de Medicina do Porto. Decreto-Lei nº 19337, de 29 de Janeiro de 1931)
IV
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
CORPO CATEDRÁTICO DA FACULDADE DE MEDICINA DO PORTO
Professores Catedráticos
Doutor Manuel Maria Paula Barbosa
Doutor Manuel Alberto Coimbra Sobrinho Simões
Doutor Jorge Manuel Mergulhão Castro Tavares
Doutora Maria Isabel Amorim de Azevedo
Doutora Maria Amélia Duarte Ferreira
Doutor José Agostinho Marques Lopes
Doutor Patrício Manuel Vieira Araújo Soares da Silva
Doutor Daniel Filipe Lima Moura
Doutor Belmiro dos Santos Patrício
Doutor Alberto Manuel Barros da Silva
Doutor José Manuel Lopes Teixeira Amarante
Doutor José Henrique Dias Pinto de Barros
Doutora Maria Fátima Machado Henriques Carneiro
Doutora Isabel Maria Amorim Pereira Ramos
Doutora Deolinda Maria Valente Alves Lima Teixeira
Doutora Maria Dulce Cordeiro Madeira
Doutor Cassiano Pena de Abreu e Lima
Doutor Altamiro Manuel Rodrigues Costa Pereira
Doutor Rui Manuel Almeida Mota Cardoso
Doutor António Carlos Freitas Ribeiro Saraiva
Doutor Álvaro Jerónimo Leal Machado de Aguiar
Doutor António José Pacheco Palha
Doutor José Luís Medina Vieira
Doutor José Carlos Neves da Cunha Areias
Doutor Manuel Jesus Falcão Pestana Vasconcelos
Doutor João Francisco Montenegro Andrade Lima Bernardes
Doutora Maria Leonor Martins Soares David
Doutor Fernando Manuel Mendes Falcão dos Reis
Doutor Rui Manuel Lopes Nunes
Doutor Amadeu Pinto de Araújo Pimenta
Doutor António Albino Coelho Marques Abrantes Teixeira
Doutor José Eduardo Torres Eckenroth Guimarães
Doutor Francisco Fernando Rocha Gonçalves
V
Professores Jubilados
Doutor António Augusto Lopes Vaz
Doutor António Carvalho Almeida Coimbra
Doutor António Fernandes da Fonseca
Doutor António Fernandes Oliveira Barbosa Ribeiro Braga
Doutor António Germano Pina Silva Leal
Doutor António Luís Tomé da Rocha Ribeiro
Doutor António Manuel Sampaio de Araújo Teixeira
Doutor Cândido Alves Hipólito Reis
Doutor Daniel Santos Pinto Serrão
Doutor Eduardo Jorge Cunha Rodrigues Pereira
Doutor Fernando de Carvalho Cerqueira Magro Ferreira
Doutor Francisco Zarco Carneiro Chaves
Doutor Henrique José Ferreira Gonçalves Lecour de Menezes
Doutor João Silva Carvalho
Doutor José Augusto Fleming Torrinha
Doutor José Carvalho de Oliveira
Doutor José Fernando Barros Castro Correia
Doutor Levi Eugénio Ribeiro Guerra
Doutor Luís Alberto Martins Gomes de Almeida
Doutor Manuel Teixeira Amarante Júnior
Doutor Maria da Conceição Fernandes Marques Magalhães
Doutor Mário José Cerqueira Gomes Braga
Doutor Serafim Correia Pinto Magalhães
Doutor Valdemar Miguel Botelho dos Santos Cardoso
Professores Aposentados
Doutor Abel José Sampaio da Costa Tavares
Doutor Alexandre Alberto Guerra Sousa Pinto
Doutor Amândio Gomes Sampaio Tavares
Doutor Artur Manuel Giesteira de Almeida
Doutor Carlos Rodrigo Magalhães Ramalhão
Doutor Fernando Tavarela Veloso
Doutor Francisco de Sousa Lé
Doutor Joaquim Germano Pinto Machado Correia da Silva
Doutor José Manuel Costa Mesquita Guimarães
Doutor Manuel Augusto Cardoso de Oliveira
Doutor Manuel Machado Rodrigues Gomes
Doutor Walter Friedrich Alfred Osswald
VI
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Aos meus afectos
VII
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
AGRADECIMENTOS
Expresso o meu reconhecimento e gratidão
… a quem contribuiu para esta dissertação:
– Professor Doutor Patrício Soares da Silva, meu Orientador. Não fora o seu apoio e
incentivo e provavelmente nunca teria eu encontrado a motivação necessária para
levar por diante os trabalhos desta dissertação.
– Professor Doutor António Albino Teixeira, meu Co-Orientador, pelo apoio e ânimo,
sempre disponível e sempre tolerante com a minha impaciência.
– Professor Doutor Rui Coelho, pela amiga disponibilidade para comigo discutir aspectos relacionados com os instrumentos psicométricos utilizados nos trabalhos desta
dissertação.
– Professor Doutor Todd B. Kashdan, da Universidade George Mason, pela permanente
atitude de desafio e questionamento sobre a forma de relatar os resultados.
– Professor Doutor Amílcar Falcão, por generosamente ter-me aconselhado nas análises estatísticas e, particularmente, pelo desenvolvimento do modelo estatístico
utilizado na análise farmacocinética.
– Equipa clínica da Unidade de Farmacologia Humana de BIAL, nomeadamente o
Professor Doutor Manuel Vaz da Silva (Director Clínico e Investigador Principal),
a Drª Teresa Nunes (Coordenadora das Operações de Ensaio) e, em especial, a
Drª Benedita Azevedo (Assistente de Ensaio Clínico), pela colaboração amável e
competente.
… a quem permitiu esta dissertação:
– Dr. Luís Portela, Presidente do Conselho de Administração de BIAL – Portela & Cª
SA, por ter generosamente autorizado a realização dos trabalhos desta dissertação
com os participantes em ensaios clínicos da Unidade de Farmacologia Humana de
BIAL.
IX
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
«As we know, there are known knowns.
There are things we know we know.
We also know there are known unknowns.
That is to say we know
there are some things we do not know.
But there are also unknown unknowns,
the ones we don’t know we don’t know.»
XI
Donald Rumsfeld,
em conferência de imprensa no
Departamento de Defesa dos
Estados Unidos da América,
em 12 de Fevereiro de 2002
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
ÍNDICE
Introdução
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Objectivos
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Capítulo I Caracterização sócio-demográfica e das motivações e percepções dos
participantes em ensaios clínicos de fase I, bem como da influência
das características de personalidade nas mesmas.
«Why healthy subjects volunteer for phase I studies and how they .
perceive their participation?»
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Eur J Clin Pharmacol 63, 1085-1094, 2007
Capítulo II Caracterização da importância dos valores interpessoais na auto-selecção
de voluntários para ensaios clínicos de fase I
21
«Interpersonal values of healthy research volunteers participating
in phase I drug studies»
. . . . . . . . . . 33
. . . . . . . . . . . . . . . . . . . . . 35
Enviado para publicação
Capítulo III Os traços de personalidade como filtro de auto-selecção de participantes . . . . . . . . . . 41
em ensaios clínicos de fase I e sua influência nas avaliações de tolerabilidade
a) «Personality characteristics of volunteers in Phase 1 studies and likelihood . . . . . . . . . . . . . 43
of reporting adverse events»
b) «Healthy subjects volunteering for phase I studies: Influence of curiosity, . . . . . . . . . . . . . . 53
exploratory tendencies and perceived self-efficacy»
c) «Who volunteers for phase I clinical trials? Influences of anxiety, social . . . . . . . . . .
anxiety and depressive symptoms on self-selection and reporting of adverse events»
d) «The role of socio-economic conditions and psychological factors on
the willingness to volunteer for phase I studies»
Int J Clin Pharmacol Ther 46, 340-348, 2008
Int J Clin Pharmacol Ther 46, 109-118, 2008
. . . . . 63
. . . . . . . . . . . . . . . . . 71
Eur J Clin Pharmacol 64, 575-582, 2008
Pharm Med, 22, 2008 (aprovado para publicação; em impressão)
XIII
Capítulo IV Prevalência de queixas adversas de origem não medicamentosa e sua .
correlação com as características de personalidade
. . . . . . . . . . . . . 79
«Adverse nondrug-related complaints by healthy volunteers in Phase I studies . . . . . . . . . . . . 81
compared to healthy general population»
Int J Clin Pharmacol Ther, 46, 2008 (aprovado para publicação; em impressão)
Capítulo V Análise exploratória da influência das variáveis psicométricas relevantes . . . . . . . . . . . . . 91
no processo de auto-selecção de voluntários para participação em ensaios
de fase I sobre a farmacocinética dos fármacos estudados nos mesmos ensaios
Resultados não submetidos a publicação
Discussão e Conclusões .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Resumo, Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Bibliografia
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XIV
129
Introdução
Introdução
INVESTIGAÇÃO DE NOVOS MEDICAMENTOS
A cadeia de desenvolvimento de novos medicamentos pode sistematizar-se do seguinte
modo: Investigação exploratória → Estudos pré-clínicos → Ensaios clínicos de Fase I (farmacologia humana) → Ensaios clínicos de Fase II (terapêuticos exploratórios) → Ensaios clínicos
de Fase III (terapêuticos confirmatórios) → Registo (autorização de introdução no mercado)
→ Comercialização.
Contudo, parecem necessárias alterações significativas ao modelo vigente de investigação
e desenvolvimento de novos medicamentos. Por um lado, assiste-se a uma redução progressiva do número de novos medicamentos introduzidos em cada ano, apesar de um aumento
crescente do investimento em investigação [Woodcock, 2007; Woosley & Cossman, 2007].
Por outro lado, os custos de investigação em função de cada novo medicamento introduzido em comercialização aumentaram marcadamente [Woodcock, 2007], particularmente os custos dos ensaios clínicos [FDA, 2004], atingindo valores próximos dos 900 milhões
de dólares [Kola & Landis, 2004], cujo retorno financeiro obriga a preços de comercialização insustentáveis para muitas sociedades, mesmo para as mais desenvolvidas [Woodcock,
2007]. Além disso, o tempo médio de desenvolvimento de um novo medicamento é muito longo, rondando os 12 anos entre a data de patente e a data de introdução no mercado [EFPIA, 2008], o que se tornou um impedimento e desincentivo à inovação [Woosley &
Cossman, 2007]. Por outro lado, ainda, o paradigma vigente de desenvolvimento de novos
medicamentos é de difícil adaptação a certas áreas terapêuticas, como é o caso das doenças raras ou o das doenças degenerativas, onde existem aliás claras necessidades por suprir.
Assim, por várias razões, nomeadamente os custos excessivos, a baixa taxa de produtividade e a estagnação da inovação, considera-se que o actual modelo de investigação e
desenvolvimento de novos medicamentos está em crise e necessita de evoluir [FDA, 2004],
de modo a melhorar a sua eficiência e a capacidade de satisfazer as necessidades terapêuticas presentes e futuras [Buckman et al., 2007]. É esse, pelo menos em parte, o ponto de
partida de programas recentemente lançados tanto nos Estados Unidos da América (FDA
Critical Path Initiative) [FDA, 2004] como na União Europeia (Innovative Medicines Initiative, IMI) [EFPIA/EU, 2007], que visam melhorar as condições para um desenvolvimento de
soluções terapêuticas inovadoras, de um modo mais rápido e com custos mais aceitáveis.
3
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
A maior debilidade do actual modelo de investigação de novos medicamentos deve-se à necessidade de vários milhares de fármacos iniciarem o desenvolvimento para que um novo medicamento atinja a fase de comercialização [EFPIA, 2008]. A principal razão para tão grande taxa
de insucesso é o facto das diversas fases do desenvolvimento apresentarem baixa capacidade
de previsão da utilidade clínica final [Kola & Landis, 2004]. Como a investigação exploratória
e pré-clínica é pouco preditiva da eficácia e da segurança dos fármacos no Homem, o seu real
valor terapêutico é determinado sobretudo em ensaios clínicos, o que faz com que grande parte
da taxa de descontinuação se situe na fase clínica. A percentagem dos fármacos que iniciaram
desenvolvimento clínico e atingem a fase de registo não ultrapassa os 11% e a dos que chegam
efectivamente a ser comercializados não atinge 8% [Kola & Landis, 2004]. Mas, mesmo assim,
entre os que iniciam comercialização, vários são descontinuados devido a problemas de segurança só detectados, por vezes, apenas após anos ou mesmo décadas de uso [Woosley, 2004].
É, portanto, urgente o aperfeiçoamento dos processos de desenvolvimento de novos medicamentos [Woosley & Cossman, 2007]. Um dos mais prementes desafios é a melhoria da capacidade de previsão do insucesso/sucesso em cada fase do desenvolvimento e a redução dos
custos associados à investigação de novas terapêuticas farmacológicas. Tal passa, sobretudo,
por melhorar a capacidade das fases mais iniciais de desenvolvimento predizerem a eficácia e
a segurança de um fármaco, de modo a que as decisões de descontinuação possam ser mais
precoces e informadas (“kill early and kill smart”).
O advento da medicina personalizada
O paradigma de desenvolvimento de novos medicamentos tem-se apoiado numa abordagem que se pode considerar de base “populacional”. A efectividade (isto é, a utilidade) clínica
de um novo medicamento é considerada como demonstrada após a realização de pelo menos
dois ensaios clínicos aleatorizados e controlados na população-alvo de doentes [Woodcock,
2007]. No caso dos ensaios serem “positivos”, isto é, se demonstrarem eficácia significativamente superior do fármaco-teste face à do placebo, ou eficácia pelo menos idêntica à do
comparador activo (se for esse o controlo), e se a tolerabilidade e segurança foram aceitáveis,
considera-se que a relação risco/benefício é favorável e, então, o novo medicamento pode iniciar a sua comercialização. Contudo, apesar de até ao momento ainda não se ter desenvolvido
melhor alternativa, o ensaio clínico aleatorizado apresenta certas limitações. A interpretação
usual de um ensaio aleatorizado e controlado, comparando duas intervenções numa população e em que uma intervenção se revela “superior” à outra, é a de que a intervenção “superior”
será a melhor escolha para qualquer indivíduo dessa população [Woodcock, 2007]. Isso tem
como consequência que, apesar da diversidade das características individuais, os doentes sejam tratados geralmente de modo idêntico. Ao adoptar-se a “resposta média” (que é a que,
no fundo, é traduzida pelos resultados desses ensaios) como recomendação para o global da
população-alvo, desvaloriza-se a importância da variabilidade individual na resposta ao fármaco, quer em termos de eficácia quer em termos de segurança. Por isso, não se podem considerar completamente surpreendentes as estimativas que apontam para que apenas 20-60%
dos doentes recebam o tratamento mais efectivo [Foernzler, 2008] e que, apenas nos Estados
Unidos da América, mais de 100.000 pessoas morram anualmente devido a reacções adversas
4
Introdução
a medicamentos, constituindo uma das primeiras cinco causas de morte [Lazarou et al., 1998].
Ou seja, o actual paradigma de desenvolvimento de medicamentos não tem capacidade
discriminativa satisfatória das diferenças na resposta em função das características individuais, que, conhecendo-se, permitiriam a identificação a priori dos doentes não respondedores
ou dos que não toleram adequadamente o medicamento, evitando-se, assim, que o medicamento fosse prescrito a um doente com características preditivas de “não respondedor” ou
de problemas de segurança. Para resolver este problema, assiste-se actualmente à emergência
de um novo paradigma, a que se convencionou chamar “medicina personalizada” [Allison,
2008], que tem como objectivo proporcionar o fármaco certo, ao doente certo, na dose certa, no
tempo certo (isto é, à primeira vez), conforme definido pela FDA (“right drug, right patient, right
dose, right time”) [Buckman et al., 2007].
O conceito de medicina personalizada pressupõe a existência de novas modalidades diagnósticas (“Good therapeutics follows good diagnostics”) que possam ser usadas para gerir a variabilidade em numerosos cenários durante o desenvolvimento e posterior utilização clínica dos novos
medicamentos [Lesko, 2007]. Serão marcadores biológicos capazes de predizerem o metabolismo, a eficácia, a tolerabilidade e a segurança de uma dada intervenção farmacológica num
determinado indivíduo [Foernzler, 2008]. A descodificação do genoma humano, o desenvolvimento da farmacogenética e da farmacogenómica, e a evolução tecnológica crescente ao nível
de metodologias analíticas (genómica, proteómica, metabolómica, imagiologia não invasiva,
etc.) e da bioinformática, são factores facilitadores de um desenvolvimento da medicina personalizada e do seu estabelecimento como paradigma prevalente, tanto ao nível de desenvolvimento de novos medicamentos como da prática terapêutica de rotina, num futuro não longínquo
[Lesko, 2007]. Considera-se que a medicina personalizada será o futuro [FDA, 2004], restando apenas a dúvida relativamente ao tempo que demorará a implementar-se [Lesko, 2007].
A matéria-prima da medicina personalizada é a informação. Assim, tanto no decurso
dos ensaios clínicos como de estudos observacionais, crescentemente emergirá a necessidade
de recolher informação maciça de cada indivíduo, usando as novas ferramentas diagnósticas.
Essa informação, analisada em conjunto com a de outros indivíduos, constituirá uma base
para a definição de marcadores de eficácia e de marcadores de segurança, através da comparação das características dos doentes respondedores versus os não respondedores e das características dos doentes com boa tolerabilidade versus os com problemas de tolerabilidade ou
segurança. Nisso assentará a possibilidade de um tratamento “personalizado” dos doentes,
seja em contexto de ensaio clínico seja em contexto de rotina clínica, devidamente “estratificados” de acordo com os marcadores biológicos apropriados.
Deste novo paradigma esperam-se benefícios para o doente e para os sistemas de saúde,
nomeadamente a prescrição de apenas medicações eficazes, evitamento/controlo dos efeitos
adversos, dosagem mais segura e rigorosa durante a terapêutica, maior número de opções
terapêuticas e redução dos custos de tratamento devido a um melhor diagnóstico. O desenvolvimento de marcadores de eficácia e segurança permitirá, também, tempos de desenvolvimento mais curtos e redução do custo médio do desenvolvimento de cada novo medicamento.
5
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Ensaios em voluntários saudáveis no contexto do desenvolvimento
de novos medicamentos
À excepção de fármacos que pela sua toxicidade (como é o caso, por exemplo, de agentes citotóxicos destinados ao tratamento de doenças oncológicas) são primeiro investigados
em voluntários com a doença-alvo, o desenvolvimento de novos medicamentos engloba uma
série de ensaios de farmacologia humana realizados em voluntários saudáveis (ditos ensaios
de fase I). A bateria típica de ensaios de fase I de um novo medicamento engloba os ensaios
iniciais (“first-in-man”) de segurança, tolerabilidade e farmacocinética, ensaios de avaliação do
impacto de factores intrínsecos (idade, género, função renal, função hepática…) e do impacto
de factores extrínsecos (interacções medicamentosas, dieta…) na farmacocinética, farmacodinamia e tolerabilidade, e ensaios biofarmacêuticos (bioequivalência relativa de formulações,
proporcionalidade de dose, presença de alimentos…).
Os ensaios de fase I apresentam um elevado potencial para a melhoria da eficiência do
modelo de desenvolvimento de novos medicamentos. Por um lado, o processo de autorização
de introdução no mercado (AIM) de um novo medicamento engloba habitualmente ensaios
de fase I realizados num total de mais de meio milhar de voluntários, o que pode representar
cerca de 20% da população submetida a ensaio clínico até à altura do pedido de AIM. Assim,
existe o potencial de desenvolvimento de marcadores de resposta e de segurança mais sofisticados que os actualmente utilizados e que, quando aplicados nesta população, podem permitir a tomada de decisões de descontinuação mais precoces, evitando-se avançar indevidamente para as fases de ensaio em doentes. Enquanto que o custo do conjunto de ensaios de fase I
de um novo medicamento não ultrapassa, em média, 10 milhões de euros, os ensaios de fase
II custam 75 milhões de euros e os de fase III custam 200 milhões de euros [Gaw, 2008]. Por
outro lado, entre os fármacos que iniciam ensaios terapêuticos, a taxa de insucesso é de 40%
na fase II [Kola & Landis, 2004] e, entre os fármacos que mostraram evidência de efectividade
na fase II, a taxa de insucesso é de 50% na fase III [Woosley & Cossman, 2007]. Esta altíssima
taxa de insucesso nas fases II e III torna, pois, evidente que os ensaios de fase I, tal como são
actualmente realizados, apresentam um poder de previsão da utilidade clínica extremamente
baixo. Urge, pois, robustecer o seu papel como fonte de informação que possa contribuir para
decisões mais precoces de descontinuação do desenvolvimento e, assim, reduzir a taxa de insucesso nos ensaios de fase II e III e os enormes custos associados.
Por outro lado, os ensaios de fase I podem também desempenhar um papel importante
na melhoria da eficiência das fases II e III de desenvolvimento de novos medicamentos, ao fornecerem informação importante para a estratificação dos doentes nos ensaios terapêuticos,
reduzindo assim o risco de ensaios negativos e proporcionando a “personalização” das recomendações de utilização em rotina clínica. Para além disso, a informação obtida em ensaios de
fase I tem potencial uso no estabelecimento de relações farmacocinéticas-farmacodinâmicas
como estratégias de ponte (“bridging”) entre diferentes populações de doentes (por exemplo,
diferentes grupos étnicos, novas populações com a doença-alvo, doentes pediátricos, geriátricos ou grávidas, doentes com insuficiência hepática ou renal) ou diferentes formulações de um
determinado medicamento/fármaco (por exemplo, novas doses ou intervalos de dose, novas
formas de libertação e novas vias de administração) [Rolan et al., 2003].
6
Introdução
RECRUTAMENTO E SELECÇÃO DE VOLUNTÁRIOS SAUDÁVEIS
PARA ENSAIOS DE FASE I
Não existem dados estatísticos sobre o número de ensaios clínicos de fase I e/ou o número de voluntários saudáveis estudados nos mesmos, a nível mundial. Contudo, é razoável
presumir-se que se realizam anualmente milhares de ensaios clínicos de fase I, num total de
centenas de milhar de voluntários saudáveis.
Porque depende do consentimento informado e livremente expresso, a amostra de participantes num ensaio clínico de fase I é uma “amostra selectiva”, potencialmente não representativa da população geral como um todo. Assim, parece importante conhecer em que medida
os voluntários saudáveis que participam em ensaios clínicos de fase I diferem da população
geral e qual o potencial impacto das eventuais diferenças nos resultados dos ensaios clínicos
[Tishler et al., 2005]. Pela não aleatoridade da selecção dos participantes nos ensaios, o processo de recrutamento e selecção de voluntários para participação nos ensaios de fase I pode
introduzir factores de confusão limitantes da validade externa, isto é, da generalização dos
resultados à população geral, caso ocorram relações sistemáticas entre as características individuais que favoreceram a participação e as variáveis do estudo [Marcus & Schutz, 2005].
A “amostra selectiva” de participantes é o resultado da acção de vários selectores que,
para efeitos de sistematização, postulamos situarem-se em três domínios: a) selecção situacional ou de contexto; b) auto-selecção; e, c) heterosselecção (Figura 1.1).
Validade externa?
População geral
A
B
C
População do estudo
Figura 1.1. Domínios dos factores de selecção de participantes em ensaios de fase I:
A) selecção situacional ou de contexto; B) auto-selecção; e, C) heterosselecção
No domínio situacional ou de contexto, podemos considerar múltiplos selectores que
são essencialmente externos às características do indivíduo ou às limitações impostas pela
investigação propriamente dita, mas que potencialmente limitam a possibilidade de um indivíduo da população geral constituir-se como virtual voluntário (Figura 1.2).
7
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Selecção situacional
Indivíduo da população geral
Local de residência
Pressão social/familiar
Situação económica
Informação
Transportes
Disponibilidade de tempo
Potencial voluntário
Figura 1.2. Possíveis factores situacionais ou de contexto selectivos da participação em ensaios de fase I.
O primeiro selector situacional é o acesso à informação que torna um qualquer indivíduo
da população geral consciente da possibilidade de poder participar. São múltiplos os canais
que hoje em dia se podem utilizar para divulgar ensaios clínicos, desde a internet, publicidade
na rádio ou em meios de comunicação escritos, referência de ex-participante ou membro da
equipa (“boca-a-boca”), etc. As estratégias de recrutamento de voluntários saudáveis variam
largamente entre centros clínicos. Nalguns casos, utiliza-se publicitação junto do público em
geral, através dos meios de comunicação de massas. Noutros casos, o recrutamento é dirigido
a grupos específicos (por exemplo, estudantes universitários) e através de meios mais ou menos selectivos. A estratégia de recrutamento tem uma influência óbvia no tipo de indivíduos seleccionados e, consequentemente, na sua representatividade da população geral. Este selector
é absoluto, dado que ninguém pode ponderar voluntariar-se se não souber que o pode fazer.
Outros factores de selecção situacional são, por exemplo, a situação sócio-económica,
a distância do local de residência e a facilidade de transportes até à unidade onde decorrerá
o ensaio, e as percepções sociais e familiares positivas ou negativas sobre a participação, que
podem condicionar a predisposição à participação. Estes selectores condicionam significativamente a participação do indivíduo, mas não são absolutos, porque aquele mantém alguma
capacidade de arbítrio.
No domínio da auto-selecção, podemos considerar múltiplos factores selectores que são intrínsecos ao indivíduo potencialmente candidato à participação num ensaio de fase I (Figura 1.3).
A auto-selecção deriva do facto da participação num ensaio clínico pressupor prévio
consentimento informado. Ou seja, a amostra de participantes é uma população auto-seleccionada de sujeitos que consentiram em participar. A auto-selecção condiciona um potencial
viés, também designado por “viés de voluntário” ou “viés de não resposta” [Marcus & Schutz,
2005]. Constitui-se como um factor de confusão que pode inclusive atingir a própria definição
8
Introdução
Auto-selecção
Potencial voluntário
Valores
Motivações
Personalidade
Voluntário
Figura 1.3. Factores auto-selectivos de participação em ensaios de fase I.
da “norma”: por exemplo, os valores normativos das escalas psicométricas, incluindo as que
usámos no presente trabalho, são definidos com base na aplicação na população “normal”.
Mas restará sempre a questão sobre se efectivamente representam “toda” a população, porque as escalas só são aplicadas em quem aceita e não está assegurado que as características
das pessoas que rejeitaram a participação se distribuam de um modo idêntico às das que
aceitaram colaborar na validação da escala.
A auto-selecção pode assumir várias formas e pode ocorrer em diferentes alturas. Postulamos que o factor de confusão de auto-selecção num ensaio de fase I conduzido em voluntários saudáveis pode ocorrer principalmente em três momentos:
–– ­No momento de um indivíduo decidir responder ou não a um anúncio (por exemplo,
um anúncio no jornal ou na internet, ou um cartaz) ou outros meios pelos quais tomou
conhecimento sobre a possibilidade de se voluntariar para um ensaio de fase I (por exemplo, um contacto pessoal).
–– No momento de finalmente decidir participar, após tomar conhecimento dos detalhes
específicos do estudo, durante os procedimentos do consentimento informado.
–– Em qualquer altura durante o estudo, porque um participante mantém o direito de abandonar o ensaio sempre que o quiser fazer. Uma eventual perda diferencial na continuação
do estudo pode englobar-se no conceito de “viés de não resposta”.
Pode ainda teorizar-se um factor de confusão de selecção adicional: a participação repetida. Para além das questões metodológicas e éticas que a participação repetida levanta
— nomeadamente o risco de “profissionalização”, quando a repetição é excessiva [Tishler &
Bartholomae, 2003] —, a “experiência” prévia pode condicionar os resultados dos ensaios.
Porque são saudáveis, não é de presumir qualquer efeito directo benéfico para a saúde
dos participantes em ensaios de fase I decorrente da administração do produto investigacional. Por outro lado, usualmente, a participação obriga a internamento por vários dias, proce-
9
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
dimentos desconfortáveis ou mesmo dolorosos, e risco de acontecimentos adversos, por vezes
graves (embora muito raramente), como o caso recente ocorrido no Reino Unido, em que voluntários saudáveis desenvolveram falência multiorgânica de instalação rápida, por libertação
maciça de citocinas, após administração intravenosa do imunomodulador TGN1412 [Nada &
Somberg, 2007]. Por outro lado, a compensação financeira não é normalmente elevada, devido às restrições éticas e legais impostas de forma a evitar que o apelo financeiro possa, pelo
seu excesso, constituir-se como factor limitante a um consentimento isento de toda a espécie
de coacção. Assim, pode perguntar-se quais são então as razões porque indivíduos normais se
voluntariam para ensaios de fase I?
A atenção dada às motivações e atitudes dos participantes em ensaios de fase I tem sido
muito limitada e a bibliografia disponível é muito escassa [Tishler & Bartholomae, 2002].
Numa revisão sistemática recente da literatura relativa ao tema, dois autores [Tishler & Bartholomae, 2002] encontraram apenas sete publicações que cumpriam os critérios para revisão
[Bigorra & Baños, 1990; Esecover et al., 1961; Hassar et al., 1977; Martin et al., 1968; Novak et
al., 1977; Van Gelderen et al., 1993; Vrlovac et al., 1990]. A mais recente tinha sido publicada
em 1993 [Van Gelderen et al., 1993] e quatro tinham sido publicadas nos anos 60 ou 70 do
século XX [Esecover et al., 1961; Hassar et al., 1977; Martin et al., 1968; Novak et al., 1977],
numa altura em que as condições de realização de ensaios de fase I eram bem distintas das
actuais, nomeadamente no que respeita ao recrutamento de prisioneiros [Martin et al., 1968;
Novak et al., 1977] e colaboradores das empresas farmacêuticas [Hassar et al., 1977].
O motivo primário auto-relatado para participação num ensaio clínico de fase I tem sido
descrito como sendo o incentivo financeiro, o que aliás corresponde ao expectável. As amostras de voluntários saudáveis compreendem habitualmente pessoas com baixo rendimento
[Ayd & Calabresi, 1972; Bigorra & Baños, 1990; Tishler & Bartholomae, 2003; Van Gelderen et
al., 1993], entre as quais se incluem os estudantes universitários. Contudo, empiricamente conhecemos que há muitas pessoas economicamente desfavorecidas que rejeitam liminarmente
a ideia de participação; ou seja, embora a compensação financeira seja importante para muitos indivíduos, não é suficiente para outros se voluntariarem para a participação em ensaios
de fase I. E há indivíduos que invocam razões de natureza não financeira, nomeadamente altruísticas, como motivo de participação. Assim sendo, pode teorizar-se que os valores e a personalidade do indivíduo são factores importantes na decisão de participar num ensaio clínico.
Os traços de personalidade e os valores individuais são características que indicam uma
preferência para agir ou perceber o mundo de uma determinada maneira. Os valores determinam preponderantemente o comportamento humano e são o alicerce no qual assenta a
personalidade do indivíduo, ditando-lhe o que é justo, desejável, aceitável ou certo. Um grupo
de valores pessoais, o dos valores interpessoais, pode ser particularmente relevante para a
decisão de participar em ensaios de fase I. Os valores interpessoais são aqueles que influenciam o comportamento e as relações sociais, incluindo a motivação para tomar ou não tomar
atitudes em favor de outrem. Um traço de personalidade é descrito como uma característica
individual, em muitos aspectos heredo-constitucional e razoavelmente estável, que é responsável pela consistência do comportamento. Os valores servem de sistema de “orientação”
utilizado por uma personalidade quando confrontada com a escolha entre alternativas [Beu-
10
Introdução
kman, 2005]. Nos trabalhos desta dissertação pretendemos, pois, avaliar a influência dos
valores interpessoais, da personalidade e de factores psicológicos como selectores potenciais
de participação em ensaios de fase I.
No domínio da heterosselecção, os selectores postulados são os critérios de inclusão/exclusão dos sujeitos, definidos no protocolo de ensaio clínico, e o livre arbítrio do investigador,
que aliás é muitas vezes assumido na lista de critérios de selecção protocolados (Figura 1.4).
A participação efectiva num ensaio de fase I é restringida a indivíduos que, para além de
consentirem na participação, cumprem os critérios de selecção definidos no protocolo, que
na maioria dos casos se baseiam na demografia (por exemplo, idade, género, índice de massa
corporal ou grupo étnico), estilo de vida (por exemplo, não fumadores) e a demonstração da
condição de “saudável”. Um critério arbitrário dá habitualmente ao investigador o poder de
decidir a exclusão de um determinado indivíduo apenas com base na sua percepção subjectiva
sobre se o indivíduo cumpriria ou não as exigências do estudo.
Os critérios de inclusão/exclusão definidos nos protocolos de fase I são usualmente muito estritos e restritivos (por exemplo, ao limitarem comummente a participação a “jovens”).
Isso conduz a amostras mais homogéneas, o que pode reforçar a validade interna do ensaio,
mas representa uma ameaça à sua validade externa, isto é, ao potencial de generalização dos
resultados.
Heterosselecção
Voluntário
Critérios de selecção
Arbítrio do investigador
Participante
Figura 1.4. Factores selectivos de participantes em ensaios de fase I do domínio da heterosselecção.
11
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
AUTO-SELECTORES DE PARTICIPAÇÃO EM ENSAIOS CLÍNICOS DE FASE I
E IMPACTO NOS RESULTADOS DOS MESMOS
Os trabalhos desta dissertação focaram sobretudo na auto-selecção, embora também
tenham sido explorados alguns aspectos da selecção situacional.
Como referido, o consentimento informado é um agente de auto-selecção e teorizou-se
que é condicionado pelos valores interpessoais e pela personalidade e, eventualmente, por
alguns estados psicológicos específicos.
No que respeita aos valores interpessoais, estes foram avaliados através do Inventário de
Valores Interpessoais, versão portuguesa do Survey of Interpersonal Values (SIV) [Gordon, 1993].
Os resultados obtidos numa amostra de participantes em ensaios clínicos foram comparados
contra os valores normativos da população portuguesa, utilizando uma versão validada para
Portugal por A. Menezes-Rocha e colaboradores [Gordon, 2001].
As características de personalidade associam-se a diferentes marcadores biológicos. Por
exemplo, níveis elevados de ansiedade e inibição comportamental têm sido associados a níveis
elevados de 5-hidroxitriptamina, enquanto comportamentos elevados em tendências exploratórias (por exemplo, curiosidade e procura de sensações) têm sido relacionados à dopamina e,
em menor extensão, à noradrenalina [Zuckerman, 1991]. Ou seja, diferentes características de
personalidade podem associar-se a diferentes ambientes biológicos e, também por essa via, a
diferentes respostas aos fármacos. O estudo desta matéria tem sido negligenciado, mas pode
ser importante no paradigma da medicina personalizada. No que respeita às características
de personalidade, a caracterização foi feita em dois níveis, com objectivos distintos, se bem
que complementares:
–– Comparação dos traços de personalidade obtidos pela aplicação do NEO Personality Inventory Reviewed (NEO PI-R) [Costa & McCrae, 1995] numa amostra de participantes
em ensaios clínicos versus os valores normativos da população portuguesa, utilizando
uma versão validada para Portugal (Inventário de Personalidade NEO Revisto) por M.
Pedroso-Lima e A. Simões [Costa & McCrae, 2000]. Se bem que este inventário permita
uma avaliação muito abrangente da personalidade de um indivíduo, caracterizando 5
factores (domínios) e seis facetas (traços) dentro de cada domínio, o NEO-PI-R é de
difícil implementação prática, pelo seu custo, tempo de aplicação e dificuldade de interpretação, o que constituem obstáculos à sua utilização generalizada e rotineira.
–– Por isso, com base na nossa experiência empírica e sobretudo apoiados na informação
publicada, tentámos identificar, a priori, dimensões de personalidade que pudessem, concomitantemente: i) estar relevantemente implicadas na auto-selecção; ii) ter impacto
hipotético sobre os resultados (“outcomes”) dos ensaios; iii) dispusessem de escalas de
avaliação de aplicação rápida e não dispendiosa e de interpretação fácil, viabilizando a
sua implementação na caracterização de voluntários para ensaios de fase I, caso os resultados suportassem tal recomendação.
12
Introdução
São escassos os estudos publicados que tenham comparado os traços de personalidade
dos participantes em ensaios de fase I com a população normativa [Tishler et al., 2005]. Em
estudos em que foi utilizado o Eysenck Personality Questionnaire [Ball et al., 1993; Cami et al.,
1989], os voluntários de ensaios de fase I apresentaram graus mais elevados de extroversão
e níveis mais baixos de neuroticismo em relação à população geral [Ball et al., 1993] e a um
grupo de controlo [Ball et al., 1993]. Noutro estudo [Pieters et al., 1992], a aplicação do Dutch
Personality Inventory mostrou também que os voluntários de ensaios de fase I eram mais extrovertidos que a norma, mas num estudo em que se utilizou o Freiburg Personality Inventory não
se encontraram diferenças significativas, ao nível de extroversão e de neuroticismo [Meyer et
al., 1995]. A aplicação do State-Trait Anxiety Inventory (STAI) mostrou também que voluntários
de ensaios de fase I apresentavam níveis de ansiedade (estado e traço) mais baixos do que os
valores normais [Pieters et al., 1992]. Em dois estudos, verificou-se que os voluntários de ensaios de fase I apresentavam uma tendência mais elevada para a procura de sensações e maior
necessidade de aventura e risco do que a norma [Farre et al., 1995; Pieters et al., 1992], se bem
que apenas num dos estudos as pontuações dos participantes em ensaios de fase I tenham
sido mais elevadas do que as pontuações da norma na subescala “experience seeking” [Farre et
al., 1995].
Um traço de personalidade relacionado com uma ampla faixa de parâmetros relacionados com a saúde é a auto-eficácia [Bandura, 1977]. A auto-eficácia percepcionada é definida
como a crença de um indivíduo na própria capacidade de autogerar comportamentos para
obter os resultados desejados quando confrontado com barreiras e obstáculos [Kashdan &
Roberts, 2004]. Indivíduos com elevada auto-eficácia preferem tarefas mais desafiantes, investem mais esforço e são mais persistentes do que os indivíduos com baixa auto-eficácia percepcionada, que duvidam das suas capacidades, evitando-as e considerando-as mais frequentemente como ameaças pessoais [Bandura, 1994], como pode acontecer num ensaio clínico.
Um parâmetro de avaliação sempre presente nos estudos de fase I é a monitorização
dos efeitos adversos. Meyer e colaboradores [Meyer et al., 1996] verificaram a existência de
correlações positivas entre o neuroticismo e o número de queixas de origem não medicamentosa numa população de estudantes de medicina. Alguns estudos demonstraram que certos
estados psicológicos podem influenciar a resposta subjectiva aos fármacos, nomeadamente o relato de queixas [Ball et al., 1993; Meyer et al., 1995]. Num estudo em doentes [Ong
et al., 2004], aqueles com maiores níveis de ansiedade apresentaram mais efeitos adversos
após administração medicamentosa do que os indivíduos com baixa ansiedade. Os autores
concluíram que tal se devia ao facto dos indivíduos mais ansiosos apresentarem vigilância
somática aumentada e uma tendência para dramatizar a interpretação de sensações físicas
ligeiras, conduzindo assim a um sobre-relato de queixas. Há também evidência da influência
da personalidade no efeito nocebo (ocorrência de efeitos adversos em consequência da toma
de placebo) [Drici et al., 1995].
Existem também dados, embora escassos e por vezes inconsistentes, que sugerem que a
farmacocinética pode ser afectada pelos traços de personalidade [Meyer et al., 1996]. Num
ensaio em voluntários saudáveis, 12 com neuroticismo elevado e outros 12 com neuroticismo
13
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
baixo, os autores [Nakano et al., 1980] verificaram que os indivíduos com níveis mais baixos
de neuroticismo absorveram uma dose de diazepam de forma mais lenta e atingiram menores
concentrações plasmáticas do que indivíduos com níveis mais elevados de neuroticismo. Os
autores repetiram a experiência com administração intraduodenal e não encontraram diferenças nas taxas de absorção, tendo concluído que o esvaziamento gástrico ocorre mais rapidamente nos indivíduos com neuroticismo mais elevado. Inversamente, outros autores [Jackson
et al., 1985] não encontraram diferenças na farmacocinética da teofilina causadas por níveis
diferentes de neuroticismo em voluntários saudáveis e em doentes asmáticos.
Pessoas com nível elevado de tendências exploratórias e de sentimento de auto-eficácia
e que gostam de actividades de risco poderiam estar mais disponíveis a voluntariarem-se para
ensaios clínicos. Nos trabalhos desta dissertação colocámos a hipótese de que os indivíduos
com traço de curiosidade mais elevado, que são mais exploratórios por definição, são mais
susceptíveis de se voluntariarem para ensaios clínicos do que indivíduos com traço de curiosidade baixo, os quais têm maior probabilidade de vivenciar estas situações como indesejáveis.
Colocámos, também, a hipótese de que indivíduos saudáveis possuindo as qualidades inerentes à auto-eficácia seriam mais predispostos a voluntariarem-se para ensaios clínicos.
Embora a ansiedade se possa manifestar de diversas formas, a participação num ensaio
de fase I parece justificar a investigação, em particular, da ansiedade-traço e os reflexos que
esta poderá ter em determinadas situações sociais (ansiedade social). Por exemplo, efeitos
adversos ou procedimentos dolorosos podem ser considerados especialmente ameaçadores
para os indivíduos ansiosos. Por outro lado, a interacção estreita com pessoas previamente
desconhecidas, como é o caso da equipa clínica e dos demais participantes, e o confinamento
a uma enfermaria aberta com poucas condições de privacidade podem ser entendidos como
especialmente ameaçadores para pessoas com ansiedade social elevada. Consequentemente,
estabelecemos a hipótese de que pessoas com ansiedade-traço e ansiedade social elevadas
seriam menos receptivas a voluntariarem-se para ensaios de fase I.
Tendo em atenção esta informação bibliográfica e a nossa experiência empírica, tentámos então identificar instrumentos psicométricos que pudessem ser de fácil aplicação prática
e capazes de mensurar as características de personalidade (e alguns estados psicológicos)
presumivelmente implicadas na auto-selecção e, eventualmente, com influência nos resultados
dos ensaios de fase I.
No pólo dos factores psicológicos “positivos”, seleccionámos os instrumentos Self-Efficacy
Scale [Sherer et al., 1982], na versão Escala de Avaliação da Auto-Eficácia Geral preparada por
J. L. P. Ribeiro [Ribeiro, 1995], e Curiosity and Exploration Inventory-Trait (CEI-T) [Kashdan, 2004],
de Todd B. Kashdan, director do Laboratory for the Study of Social Anxiety, Character Strengths, and
Related Phenomena, da Universidade George Mason, Fairfax, Virgínia, EUA, o qual acedeu em
colaborar e nos ajudou a validar uma versão portuguesa.
Para a mensuração de factores psicológicos “negativos” – nomeadamente, a ansiedade,
a ansiedade social e a sintomatologia depressiva – utilizámos as escalas State-Trait Anxiety Inventory (STAI) [Spielberger et al., 1970], Social Avoidance and Distress (SAD) e Fear of Negative Evaluation (FNE) [Watson & Friend, 1969], e Beck Depression Inventory-II (BDI-II) [Beck et al., 1996].
Estas escalas foram aplicadas a um grupo de voluntários de ensaios de fase I e a um grupo de
14
controlo composto por indivíduos que nunca tinham participado em ensaios de fase I e que
apresentavam características sócio-demográficas idênticas ao grupo de voluntários. O grupo
de controlo foi dividido em subgrupos, de acordo com a resposta ao convite para participação
num ensaio clínico concreto. Os resultados nas diferentes escalas foram comparados entre os
subgrupos do grupo de controlo e entre o grupo de controlo e o grupo de participantes. As
pontuações no grupo de participantes foram correlacionadas com o relato de acontecimentos
adversos durante a participação em ensaios clínicos e com a cinética dos fármacos testados
nos mesmos ensaios.
15
Objectivos
17
Objectivos
Os principais objectivos desta dissertação foram os seguintes:
1. Investigar aspectos relacionados com a selecção situacional ou de contexto de
participantes em ensaios clínicos de fase I (Capítulo I e Capítulo III.d).
2. Caracterizar sócio-demograficamente os voluntários para ensaios clínicos de fase
I (Capítulo I e Capítulo III.b e III.c).
3. Caracterizar as motivações e percepções relativas à participação em ensaios clínicos de fase I (Capítulo I e Capítulo III.d).
4. Caracterizar a relação entre as características da personalidade e as motivações e
percepções relativas à participação em ensaios clínicos de fase I (Capítulo I).
5. Caracterizar o processo de decisão que leva um indivíduo a voluntariar-se para um
ensaio clínico de fase I (Capítulo I e Capítulo III.d).
6. Caracterizar a importância dos valores interpessoais na auto-selecção de voluntários para ensaios clínicos de fase I (Capítulo II).
7. Caracterizar a importância das características de personalidade na auto-selecção
de voluntários para ensaios clínicos de fase I (Capítulo III.a, III.b, III.c e III.d).
8. Determinar a relação entre as características sócio-demográficas, de personalidade
e factores psicológicos e a incidência de efeitos adversos durante a participação
em ensaios clínicos de fase I (Capítulo III.a e III.c).
9. Caracterizar a prevalência de queixas adversas de origem não medicamentosa em
voluntários de ensaios clínicos de fase I e em não voluntários e determinar a sua
relação com as características de personalidade e factores psicológicos (Capítulo IV).
10.Determinar a relação entre os factores psicológicos e os parâmetros farmacocinéticos obtidos durante a participação em ensaios clínicos de fase I (Capítulo V).
19
CAPÍTULO I
Caracterização sócio-demográfica e das motivações e percepções dos
participantes em ensaios clínicos de fase I, bem como da influência das
características de personalidade nas mesmas
«Why healthy subjects volunteer for phase I studies and how they perceive their participation?»
Eur J Clin Pharmacol 63, 1085-1094, 2007
Caracterização sócio-demográfica e das motivações e percepções dos participantes em ensaios
clínicos de fase I, bem como da influência das características de personalidade nas mesmas
Eur J Clin Pharmacol (2007) 63:1085–1094
DOI 10.1007/s00228-007-0368-3
PHARMACOEPIDEMIOLOGY AND PRESCRIPTION
Why healthy subjects volunteer for phase I studies
and how they perceive their participation?
Luis Almeida & Benedita Azevedo & Teresa Nunes &
Manuel Vaz-da-Silva & Patrício Soares-da-Silva
Received: 15 May 2007 / Accepted: 7 August 2007 / Published online: 20 September 2007
# Springer-Verlag 2007
Abstract
Objective To characterize the motivations and attitudes of
healthy subjects who volunteer for phase I studies as well
as their perception of the informed consent procedure and
participation in the study.
Methods Subjects (n = 136, 48.5/51.5% males/females,
26.9±5.5 years; 51.9% students, 42.2% employed, 5.9%
unemployed) anonymously answered questionnaires regarding study participation and personality (Revised NEO
Personality Inventory).
Results Financial reward was the most important motivation
and was most valued by subjects with a lower monthly
income (p<0.01) and lower education (p<0.05). Personality
traits correlated differently with the various motivators for
participation. “Word-of-mouth” (mainly by ex-participants)
was the source of information on the phase I study for 94.9%
of the participants. Most (88.2%) subjects consulted other
people (family member/partner, 53.3% of participants;
friends, 40.0%; physician, 25.0%) before deciding to
participate. In 80% of the cases, the people consulted
recommended that the subject not participate, with the risk
of serious complications being the main objection in 75.7%
cases. The information provided was generally assessed
favorably; 34.6% of subjects considered the discomfort
caused by participation to be less than that originally
reported during the consent procedure, 58.8% considered it
to be identical and 6.6% considered it to be higher. Most of
the subjects (81.6%) declared that they never felt at risk of a
serious complication during their participation in the study,
16.2% occasionally felt that they were at risk and 2.2% felt
that they were at risk for a significant period of time.
Personality traits correlated with the decision to ask someone
else’s opinion, the manner in which information provided
was rated and the perception of the risk taken during
participation.
Conclusion Financial reward as the main motivator for
participation in phase I studies was less valued by healthy
volunteers with a higher income and education. A subject’s
motivations and perceptions toward participation were
found to be influenced by individual characteristics,
including some personality traits.
L. Almeida (*) : B. Azevedo : T. Nunes : M. Vaz-da-Silva :
P. Soares-da-Silva
Department of Research & Development, BIAL,
A Av. da Siderurgia Nacional,
4745-457 S. Mamede do Coronado, Portugal
e-mail: [email protected]
L. Almeida : P. Soares-da-Silva
Institute of Pharmacology and Therapeutics, Faculty of Medicine,
University of Porto,
Alameda Professor Hernâni Monteiro,
Porto, Portugal
Keywords Clinical pharmacology . Healthy subjects .
Phase I studies . Recruitment . Volunteering
M. Vaz-da-Silva
Department of Medicine, Faculty of Medicine,
University of Porto,
Alameda Professor Hernâni Monteiro,
Porto, Portugal
Introduction
M. Vaz-da-Silva
Hospital de S. João,
Porto, Portugal
Phase I studies in healthy subjects are of fundamental
importance in the early drug development process as an
26
23
CAPÍTULO I
1086
Eur J Clin Pharmacol (2007) 63:1085–1094
and Bigorra and Baños [2] in their studies published in the
European Journal of Clinical Pharmacology in 1993 and
1990, respectively. The subjects were also requested to
complete the Revised NEO Personality Inventory (NEOPI-R) [9], a highly regarded assessment of normal personality based on the personality Five-Factor model [9–11]. It
consists of 240 statements to which the respondents rate
their level of agreement using a five-point scale. Items are
organized into the five broad domain scales, each of which
is subdivided into six eight-item subscales, with each of
these corresponding to a specific personality trait or facet.
The five broad domains or factors and their six specific
personality traits or facets of personality measured by NEOPI-R are the following: (1) Neuroticism (N), predisposition
to experience psychological stress, irrational ideas, poor
impulse control [facets: anxiety (N1), angry-hostility (N2),
depression (N3), self-consciousness (N4), impulsiveness
(N5), and vulnerability (N6)]; (2) Extroversion (E), measures a dimension of sociability, liveliness and cheerfulness
[facets: warmth (E1), gregariousness (E2), assertiveness
(E3), activity (E4), excitement seeking (E5), and positive
emotions (E6)]; (3) Openness to Experience (O), imagination, intellect, intellectual curiosity, need for variety, nondogmatic attitudes and willingness to entertain novel and
unconventional values [Facets: fantasy (O1), aesthetics
(O2), feelings (O3), actions (O4), ideas (O5), and values
(O6)]; (4) Agreeableness (A), measures a dimension from
sympathetic, cooperative, and altruistic to egocentric,
skeptical and competitive [Facets: trust (A1), straightforwardness (A2), altruism (A3), compliance (A4), modesty
(A5) and tender-mindedness (A6)]; (5) Conscientiousness
(C), disciplined striving after goals and a strict adherence to
principles, control of impulses, planning, organizing, ability
to carry out a task [Facets: competence (C1), order (C2),
dutifulness (C3), achievement striving (C4), self-discipline
(C5) and deliberation (C6)].
In this study we used the NEO-PI-R form S (selfadministered version) that has been validated for the
Portuguese population and reported to have good psychometric properties [12]. With the support of adequate
proprietary software [13], the individual results were
interpreted by reference to the Portuguese population norms
prepared for each scale and subscale, taking the gender and
age of the individual into consideration. These were
presented as direct scores for each subscale and converted
into the corresponding percentile of the norm.
This study was approved by an Independent Ethics
Committee (IEC) before any of the subjects were
approached. The phase I studies in which the subjects had
participated had been previously approved by the national
central Ethics Committee for Clinical Research [Comissão
de Ética para a Investigação Clínica (CEIC), Lisbon,
Portugal] and by the national drug regulatory authority
intermediate phase between in vitro and animal experiments
and clinical trials in the target patient population. However,
as the subjects are healthy, it is not expected that they will
have a direct health benefit from participating in such
studies. In addition, study participation usually requires
confinement for several days, frequent blood drawing and
other inconvenient procedures, and the drug being tested
may on occasion cause adverse effects. Therefore, one may
ask why normal subjects volunteer for such studies?
Limited attention has been given to this aspect of phase I
studies, and the published literature contains only a few
reports of studies investigating the motivations and attitudes
of healthy subjects who volunteer to participate in drug
clinical trials [1]. In a recent systematic review of the
literature on the use of financial incentives in clinical
research, Tishler and Bartholomae [1] found only seven
publications meeting the criteria for review [2–8]. The most
recent of these had been published in 1993 [7], and four out
of seven had been published in the 1960s and 1070s [3–6]
when phase I study conditions were very different from
current ones, particularly in terms of the recruitment of
prisoners [5, 6] and employees of pharmaceutical companies [4]. Therefore, there is a clear need for up-to-date
studies that assess the motivations, attitudes and perceptions of potential participants in phase I studies.
As a contribution towards a further clarification of such
issues, we report here our assessment study of a group of
participants in phase I drug studies in terms of their
motivations for participating, whether they asked (or not)
someone else’s opinion before deciding to participate and
how they have perceived the informed consent procedure
and study participation. We also explored the influence of a
subject’s demographic, socio-economic and personality
characteristics on his/her motivations and perceptions
toward participation.
Materials and methods
One hundred and thirty-six subjects who had participated in
phase I drug studies conducted at the Human Pharmacology
Unit (HPU) of BIAL (Portela & Co. SA,S Mamede do
Coronado, Portugal) between January 2006 and July 2006
involving the use of orally administered new chemical
entities currently in clinical development as new putative
medicines for the treatment of CNS-related diseases were
invited to anonymously complete questionnaires on their
demographic, socio-economic and life-style characteristics
as well as personal habits. The subjects were also invited to
complete a questionnaire regarding their motivations and
experiences as study participants. This questionnaire consisted of 14 questions (summarized in Table 2) and was
derived from the questionnaires used by Van Gelderen [7]
27
24
Caracterização sócio-demográfica e das motivações e percepções dos participantes em ensaios
clínicos de fase I, bem como da influência das características de personalidade nas mesmas
Eur J Clin Pharmacol (2007) 63:1085–1094
1087
pation is limited to one study per each drug being
investigated. Since the HPU only performs studies for its
owner, which has a relatively small number of drug entities
in clinical development, repeat participation is relatively
low: most subjects (69.9%) completing this question had
participated in only one study.
An inverse correlation was found between the number of
times a volunteer participated and monthly income: subjects
earning less than 50% of the average national salary in the
industry and services sector had lower repeat participation
rates than those subjects who have a monthly net income of
at least 50% of the average national salary (p<0.01); this
apparently paradoxical finding is due to the fact that
subjects with the lowest monthly income are usually
students, which is a transitory situation (some repeat
participation in young professionals includes an early
participation when the subjects were still students).
[Instituto da Farmácia e do Medicamento (INFARMED)
Lisbon, Portugal]. These studies had involved confinement
for up to 14 days and the frequent drawing of blood for
pharmacokinetic analysis and safety procedures. All participants were selected according to the selection criteria
defined in the study protocols, which were relatively similar
between studies. Volunteers who admitted to smoking more
than ten cigarettes per day or to drinking more that 14
alcohol units per week were excluded. Before any of the
phase I studies started, the participants attended one or
more information meetings in which they received information, both verbally and in writing, on the proposed study,
including the study design, procedures, inconvenience or
discomfort during the study, precautions, possible adverse
events and financial reward. The subjects had the opportunity to ask someone else’s opinion before making their final
decision.
The statistical analysis was performed with the STATISTICAL PACKAGE FOR SOCIAL SCIENCES (SPSS) ver. 11.5 (SPSS,
Chicago, IL). Spearman’s rho correlations between different
variables were performed; statistical significance was
established at an alpha error of 5% (p<0.05), two-sided.
Why do subjects volunteer to participate?
Question 2 (Q2) One of the questions of the questionnaire
was aimed at characterizing the subject’s motivation for
volunteering to participate in such clinical studies. The
participants had to rate, using a scale from 0 (no
importance) to 5 (extreme importance), the following
reasons: curiosity, progress of science/medicine, a free
health check-up, financial reward and other. Financial
reward was the most important single motivation (Fig. 1),
receiving a score of “5” (extreme importance) and “4” by
58.1 and 30.1% of participants, respectively; 9.6% and
2.2% of the subjects gave this motivational aspect a “3”
and “2”, respectively. No subject considered financial
reward to be unimportance. A desire to contribute to the
progress of science/medicine and the possibility of a free
medical check-up were rated similarly by participants:
approximately 20% (19.1% for progress of science/medicine and 20.6% for free medical check-up) rated these
motivations with a “5” and approximately 30% rated it
with a “4” (27.9% for progress of science/medicine and
30.9% for free medical check-up). On average, curiosity
was considered to be a less powerful motivator for
participation, but 11 (8.1%) participants gave it a “5” in
terms of importance and a further 27.2% gave it a score of
“4”. Other motivations reported by the subjects to be
important were to “get a new experience” and “to break
routine” (n=4), to be with a friend who also participated
(n=4), to contribute to the development of new medicines
for the treatment of a disease that affected a close friend or
relative (n=3), to meet unknown people (n=3), and to get a
rest (n=3).
No significant correlations were found between the
different motivators and gender, age or principal occu-
Results
All participants agreed to complete the questionnaires, but
some did not answer all of the questions. The main
demographic, social and life-style characteristics of the
study cohort are presented in Table 1, and the results of the
14-item participation questionnaire are presented in Table 2.
In terms of age, occupation, habitation, education level,
monthly income, smoking, drinking coffee and physical
exercising, there were no major differences between the
male and female groups of participants. The women
differed from the men primarily with respect to civil status
(fewer were married and more were divorced), lower
consumption of alcoholic beverages and a higher use of
medications (mainly oral contraceptives). Approximately
50% of the subjects of both gender groups were university
students of a wide range of teaching fields: engineering,
26.6%; health-related professions,16.7%; literature,15.0%;
management/economy,10.0%; education,10.0%; sciences,
8.3%; law,6.7%; other,6.7%. The monthly income of the
students depended mainly from parents/family allowance
and/or scholarship.
History of participation in phase I studies
Question 1 (Q1) Due to ethics concerns (“professionalization” of volunteers), the number of studies in which a
subject is allowed to participate is limited to a maximum of
two studies per year; for methodological reasons, partici-
28
25
CAPÍTULO I
1088
Eur J Clin Pharmacol (2007) 63:1085–1094
Table 1 Socio-demographic and life style characteristics of the study population
Variable
Parameter
Male
Female
Total
Sample
Age (years)
n (%)
Mean ± SD
Median
Range
Caucasian
Other
Student
Employed
Unemployed
<25%
25%–50%
51%–100%
101%–150%
>150%
Parents/family home
Subject’s own home
Rental house/flat
Rental room
University facility
Master’s degree
Licensed
Bachelor’s degree
12 years
9 years
6 years
Single
Married/living together
Divorced
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
66 (48.5%)
26.6±5.2
26
18–43
97.0%
3.0%
53.0%
43.9%
3.0%
24.1%
27.6%
29.3%
12.1%
6.9%
43.9%
24.2%
18.2%
10.6%
3.0%
3.0%
19.7%
3.0%
62.1%
10.6%
1.5%
69.7%
28.8%
1.5%
75.4%
24.6%
33.3%
66.7%
65.2%
34.8%
69.7%
30.3%
18.2%
81.8%
72.7%
27.3%
70 (51.5%)
27.2±5.8
26
19–45
100.0%
0.0%
50.7%
40.6%
8.7%
22.8%
29.8%
28.1%
12.3%
7.0%
37.7%
29.0%
17.4%
8.7%
7.2%
2.0%
23.2%
10.1%
58.0%
4.3%
1.4%
75.4%
14.5%
10.1%
62.3%
37.7%
30.0%
70.0%
71.0%
29.0%
32.9%
67.1%
47.1%b
52.9%
58.8%
41.2%
136 (100.0%)
26.9±5.5
26
18–45
98.5%
1.5%
51.9%
42.2%
5.9%
23.5%
28.7%
28.7%
12.2%
7.0%
40.7%
26.7%
17.8%
9.6%
5.2%
2.9%
21.5%
6.7%
60.0%
7.4%
1.5%
72.6%
21.5%
5.9%
68.6%
31.4%
31.6%
68.4%
68.1%
31.9%
50.7%
49.3%
32.8%
67.2%
65.7%
34.3%
Ethnic group (%)
Occupation (%)
Monthly net income in relation to the mean national
net salary in the industry and servicesa
Habitation
Level of education
Civil status
Keep affective relationship?
Smoking
Drinks coffee
Drinks alcoholic beverages
Use of medicines
Regular physical exercising
a
b
Approximately €1000/month (net)
Mostly oral contraceptives
pation. An inverse correlation was observed between the
motivators financial reward and monthly income (Spearman’s coefficient of correlation: R =−0.217, p<0.05):
subjects with a higher monthly income tended to consider
the financial reward to be less important than those with a
lower income. Significant positive linear correlations were
found between the manner subjects rated the importance
of the different motivators and a number of personality
traits: between curiosity and trust (R=0.246, p<0.01),
straightforwardness (R = 0.191, p < 0.05) and tendermindedness (R = 0.187, p< 0.05); between progress of
medicine/science and competence (R = 0.208, p < 0.05)
and self-discipline (R=0.196, p<0.05); between financial
reward and modesty (R=0.244, p<0.01). The following
significant negative linear correlations were found:
between free medical check-up and assertiveness (R=−
0.209, p<0.05); between curiosity and values (R=−0.187,
p<0.05); between progress of medicine/science and values
(R=− 0.248, p<0.01).
Question 3 (Q3) Surprisingly, 8.8% of participants declared
that they would agree to participate even if there was no
29
26
Caracterização sócio-demográfica e das motivações e percepções dos participantes em ensaios
clínicos de fase I, bem como da influência das características de personalidade nas mesmas
Eur J Clin Pharmacol (2007) 63:1085–1094
1089
Table 2 Results of the 14-item questionnaire (n=136)
Question
Answer
n (%)
Q1. Participation: history, including current study
One study
Two studies
Three studies
Four or more studies
Q2. Motivations for study participation (See Fig. 1)
Q3. Would you have accepted to participate if there was no financial reward?
No
Maybe
Yes
Q4. How did you hear about the possibility of participation?
Poster
Flyer
HPU staff
Ex-participant
Someone who never participated
Other way
Q5. Would you accept to recommend participation to a friend/relative?
No
Maybe
Yes
Q6. Before participating, did you ask someone else’s opinion?
No
Yes
Yes, family/partner
Yes, friend
Yes, physician
Yes, unspecified
Yes, other
Q7. How did your family or close friends reacted regarding participation?
Almost all said yes
Some said yes, some said no
Almost all said no
All said no
Q8. If they recommended you should not participate, what reasons did they give?
Risk
Guinea pig role
Missing work/school
Away from home
Other
Q9. How do you consider the financial reward?
Good
Reasonable
Bad
Q10. How do you consider the information provided before the study?
More than enough
Enough
Not enough
Q11. How do you consider the inconvenience/ discomfort in relation to the information provided?
Lower than informed
Similar to that informed
Higher than informed
Q12. How did you feel about the possibility of an occurrence of a serious complication?
Never fear that
Occasionally
Significant period of time
30
27
95 (69.9)
24 (17.6)
13 (9.6)
4 (2.9)
52 (38.2)
72 (52.9)
12 (8.8)
7 (5.1)
10 (7.4)
22 (16.2)
94 (69.1)
18 (13.2)
3 (2.2)
1 (0.7)
14 (10.3)
121 (89.0)
16 (11.8)
120 (88.2)
64 (53.3)
48 (40.0)
30 (25.0)
25 (20.8)
5 (4.2)
23 (20.3)
53 (46.8)
37 (32.6)
2 (1.9)
94 (75.7)
52 (40.0)
11 (8.7)
9 (7.8)
2 (1.7)
46 (33.8)
85 (62.5)
5 (3.7)
60 (44.4)
73 (54.1)
2 (1.5)
47 (34.6)
80 (58.8)
9 (6.6)
111 (81.6)
22 (16.2)
3 (2.2)
CAPÍTULO I
1090
Eur J Clin Pharmacol (2007) 63:1085–1094
Table 2 (continued)
Question
Answer
n (%)
Q13. How confident were you that you could trust the skills of the clinical staff?
Always sure
Occasional doubts
Doubted significant period of time
Q14. Would you agree to participate in a future study?
No
Maybe
Yes
financial reward; 52.9% declared that “maybe” they would
participate.
No correlation was found between declared willingness
to participate without financial reward and age, gender,
occupation or income. However, a significant positive
linear correlation was found between education and
probability of agreeing to participate without financial
reward (R=0.177, p<0.05): a larger number of more highly
educated subjects were less motivated by financial reward.
In terms of correlations between personality traits and
willingness to participate without financial reward, subjects
who declared that they would participate or perhaps
participate without financial reward showed: (1) higher
levels of neuroticism (p=0.001) and its traits anxiety (p<
0.05), depression (p<0.01), self-consciousness (p<0.05),
1 (0.7)
19 (14.0)
116 (85.3)
impulsiveness (p<0.05) and vulnerability (p<0.05) and a
higher level of values (p < 0.05; (2) lower levels of
conscientiousness (p<0.05 and its traits dutifulness (p<
0.05, self-discipline (p<0.05 and deliberation (p<0.05 and
a lower level of trust (p<0.05).
How did the subjects hear about the possibility
of participation?
Question O4 (Q4) Subject recruitment followed strict
standard procedures approved by an IEC. No radio, TV or
newspaper advertising was allowed, and the only public
steps taken to recruit subjects was through posters affixed
in public areas of the universities of Oporto’s metropolitan
Progress of medicine
60
60
50
50
% of subjects
% of subjects
Curiosity
40
30
20
10
40
30
20
10
0
0
0
1
2
3
4
Importance
Min
5
0
Max
1
2
3
4
Importance
Min
For-free medical check-up
5
Max
Financial rewarding
60
60
50
50
% of subjects
% of subjects
Fig. 1 How do you rate, on a
scale from 0 (not important at
all) to 5 (of extreme importance), the following reasons for
agreeing to participate in the
clinical trial?
130 (95.6)
6 (4.4)
0 (0.0)
40
30
20
10
40
30
20
10
0
0
0
Min
1
2
3
Importance
31
28
4
5
Max
0
Min
1
2
3
Importance
4
5
Max
Caracterização sócio-demográfica e das motivações e percepções dos participantes em ensaios
clínicos de fase I, bem como da influência das características de personalidade nas mesmas
Eur J Clin Pharmacol (2007) 63:1085–1094
1091
area and flyers (leaflets) distributed near the universities.
However, this type of advertising was largely inefficient:
while 12.5% of the participants declared that they had
learned about the possibility of participating in the clinical
studies through reading the poster or leaflets, only four
subjects (2.9%) learned of the study exclusively by reading
the posters and/or leaflets. “Word-of-mouth” by someone
who had already participated in a clinical study was the
source of information for 69.1% of the participants; an
additional 13.2% participants were exclusively informed by
“word-of-mouth” from someone who had never participated. In total, “word-of-mouth” was the means of information
for 94.9% of the participants; in 13.2% of the cases,
information was obtained by more than one source.
No correlations were found between the different ways
information on the clinical studies was obtained and gender,
age or monthly income. Differences were found with
respect to occupation: all subjects who stated they had
heard of the study through the flyer were students, while all
of the unemployed people who volunteered had heard of
the study by “word-of-mouth”, mainly by an ex-participant
(75%). Information provided by the posters was particularly
effective to university graduates.
asked someone else’s opinion less often than those with a
lower income (R = −0.202, p< 0.05). Differences in a
number of personality traits were found between subjects
who had asked someone’s else opinion before participating
and those who did not: subjects who had asked someone
else’s opinion tended to be higher in neuroticism (p<0.05),
self-consciousness (p<0.05), impulsiveness (p<0.05) and
vulnerability (p <0.05), and lower in order (p<0.05),
achievement striving (p<0.05), self-discipline (p<0.05)
and conscientiousness (p<0.05). When subjects who asked
a family member/partner’s opinion were compared with
those who did not, they were found to be lower in anger–
hostility (p<0.05) and fantasy (p<0.05). When subjects
who asked a physician’s opinion were compared with those
who did not, they were found to be higher in achievement
striving (p<0.05) and gregariousness (p<0.05).
Question 5 (Q5) After participating in the clinical study,
most subjects declared that they would recommend study
participation to a friend or relative. No correlation was
found with any of the socio-demographic characteristics
assessed. In terms of personality traits, subjects who
answered they were not sure whether they would recommend that someone participate tended to show higher
scores on depression (p<0.05) and fantasy (p<0.05) and
lower scores on excitement seeking (p<0.05), conscientiousness (p<0.05), dutifulness (p<0.05), achievement
striving (p<0.05) and self-discipline (p<0.05).
Question 8 (Q8) The main objection presented by those
consulted was the risk of serious complications: this was
referred by 76% of participants, and in 41% of cases
represented the single objection; to play the role of a
“human guinea pig” was referred to by 40% of the
participants, and was the sole objection in 11% of the
cases. To miss school or work was an objection presented
to 9% participants, and the inconvenience of being away
from home for several days was referred to as a reason for
none participation by 8% of the subjects. All subjects (n=9)
to whom being away from home was presented as a reason
for not participating were female (R=0.284, p <0.01).
Significant negative correlations were found between: risk
of serious complications and age (R=−0.237, p <0.05);
missing work/school and age (R=−0.244, p<0.01); missing
work/school and monthly income (R=−0.249, p <0.01).
Question 7 (Q7) Among those subjects who consulted
other people (88%), in approximately 80% cases at least
some of those consulted recommended that the subject not
participate. Despite such a high rate of recommendation not
to participate, all of those receiving this advice ultimately
did decide to participate.
Decision-making process of volunteering
Question 6 (Q6) Before agreeing to participate, 88% of the
participants asked someone else’s opinion regarding participation. Of these, a family member/partner was consulted
by 53% of the participants, friends by 40% and a physician
by 25%. Some subjects asked the opinion of several people:
20% asked the opinion of both a family member/partner
and friends; 5% asked the opinion of a family member/
partner, friends and a physician; 14% asked the opinion of
both a family member/partner and a physician; 10% asked
the opinion of both friends and a physician.
No correlations were found between the decision of
asking someone else’s opinion and any of the demographic
characteristics. A significant negative correlation was found
with monthly income: subjects with higher monthly income
Subjects’ opinion regarding financial reward
and information provided
Question 9 (Q9) Financial reward for participation in the
study had been defined by the IEC and took into account
the number of hours the subject should spend participating in a given trial. The net financial reward the subject
received per hour of involvement corresponds approximately to 80% of the hourly net salary earned by an
employee in the industry and services sector. One third
(34%) of the subjects considered the financial reward to
32
29
CAPÍTULO I
1092
Eur J Clin Pharmacol (2007) 63:1085–1094
Subjects’ willingness to repeat participation
be good, 63% considered it sufficient and 4% considered
it low. No correlation was found between the evaluation
of the appropriateness of the financial reward and age,
occupation or income; however, a correlation with
gender was found: female subjects tended to appreciate
the amount paid more highly than their male counterparts
(R=0.244, p <0.01).
Question 14 (Q14) Of the subjects who agreed to participate in the study, 85% stated that they would agree to
participate again in another clinical trial, and a further 14%
declared that “maybe” they would agree. The lone subject
(1%) who declared an unwillingness to repeat participation
mentioned that blood sampling had been very difficult to
tolerate.
Question 10 (Q10) In terms of the amount of information
that the subjects’ received before signing the informed
consent form, 44% considered it “more than enough”, 54%
as “enough” and 2% as “not enough”. No correlation was
found with age, gender, occupation, monthly income or
education level. Subjects who considered the information
provided “more than enough” were found to be relatively
higher in the following personality characteristics: warmth
(p<0.05), positive emotions (p<0.05), feelings (p<0.01),
tender-mindedness (p<0.05) and competence (p<0.05).
Discussion
The results of this study reveal that the decision to
volunteer for participation in a phase I study and the
attitudes and perceptions of subjects while participating are
influenced by a number of socio-demographic characteristics and personality traits. They also show that the
importance of the financial reward as motivator for study
participation is inversely correlated with the subject’s
education and income levels.
All invited subjects agreed to participate in the survey.
This very high participation rate is probably due to the fact
that the subjects were able to fill in the questionnaires
before their discharge from the phase I study, while they
were required to remain confined to the research facility.
To our knowledge this is the only study published in the
last decade with the aim of systematically characterizing the
perceptions and motivations of healthy subjects towards
participation in phase I studies. One of the pitfalls of the
few studies reported to date that have examined the
association between motivation for participation and
the characteristics of volunteers to phase I studies is the
lack of information on their socio-economic status [1]. In
our study, such data were collected and correlated with the
participant’s motivations for participation. More than 90%
of the participants have at least 12 years of scholarship and
a remarkable one third of the subjects have a university
degree; these figures are higher than those found in the
normal population and are the result of a recruitment
strategy that has been followed since the beginning of such
studies by Bial’s HPU in 2000 up to the present. Bial’s
HPU was the first clinical pharmacology unit licensed by
the Portuguese Health Ministry for performing phase I
studies with new drugs. Public perception of phase I studies
in Portugal is traditionally negative and, consequently,
many restrictions to such studies were self-imposed in
order to facilitate their licensing and continuation as well as
to minimize the risk that any one specific aspect could be
negatively perceived and attract the attention of the mass
media or authorities. The participation of unemployed
people or subjects with low education could be considered
to be exploitation. It is intuitively assumed that a higher
Question 11 (Q11) With respect to the risk and discomfort
caused by the study in relation to the information provided
before participation, 56% of subjects considered the
discomfort lower than that earlier stated in the informed
consent procedure and 59% considered it identical. No
correlation was found with any of the socio-demographic
characteristics assessed. In terms of personality traits,
subjects who expressed a less favorable opinion were
found to be higher in terms of values (p<0.05) and order
(p<0.05) personality traits.
Subjects’ perception of risk and staff skills
Question 12 (Q12) To the question “During the study, how
did you feel regarding the possibility of occurrence of a
serious complication”?, 82% answered that they never felt
to be at risk at all, 16% reported that they occasionally felt
at risk and 2% felt that they were at risk for a significant
period of time. No correlation was found with any of the
socio-demographic characteristics assessed. Subjects who
perceived themselves to be at risk at any time during the
study showed a higher anxiety trait (p<0.05).
Question 13 (Q13) The HPU is required to have at least
one experienced physician present in the clinical research
facility at all times (i.e. during the entire 24-h period)
throughout the whole study period (i.e. from admission
until discharge of healthy subjects). This was a requisite
required by the Ministry of Health at the time of granting
the license to operate. In terms of confidence with the
clinical staff’s skills to treat any serious clinical complication that eventually could occur, 96% of the subjects were
always convinced that staff was competent enough.
33
30
Caracterização sócio-demográfica e das motivações e percepções dos participantes em ensaios
clínicos de fase I, bem como da influência das características de personalidade nas mesmas
Eur J Clin Pharmacol (2007) 63:1085–1094
1093
of personality characteristics correlate with the declared
willingness to participate without financial reward: subjects
that rate higher in neuroticism, anxiety, depression, selfconsciousness, impulsiveness and vulnerability are more
likely to consider volunteering under such conditions.
Published data on healthy volunteer’s recruitment strategies are missing. However, it is known that such strategies
vary greatly from site to site and country to country. As
mentioned, the unfavorable environment in Portugal on the
participation in phase I studies has lead to self-imposed
restrictions. No mass advertising is allowed, and priority
has been given to recruiting university students living in the
metropolitan area where such research facilities are located.
The study population surveyed here reflects just such a
strategy: approximately half of participants are students and
a further one third have a university degree (most of them
having participated for the first time while students), and
“worth-of-mouth” was the most efficient approach for
finding new volunteers. Interestingly, the likelihood of a
participant to recommend to a friend that he/she participate
in a phase I study was shown to be influenced by the
former’s personality characteristics.
Little is known about the decision-making process of
volunteering for participation in phase I studies. In a series
of 100 phase I study participants, Rabin and Tabak [29]
reported that only 42% of the subjects sought out
information or checked alternatives before making a
decision. In our population, the proportion of subjects
who consulted other people (family, friends, a doctor,
among others) before agreeing to participate was high
(88%), and the decision to ask that for that external opinion
appeared to be influenced by the individual’s personality
traits. Opinions were predominantly negative (mainly fear
of serious complications) – i.e. advice was mostly opposed
to participation in the study – but were not taken into
account by the subjects because ultimately they did
volunteer to participate. However, we were unable to
determine the proportion of subjects who asked someone
else’s opinion and actually followed their recommendation
of no participation; studies are currently being planned to
specifically answer this question.
According to the Declaration of Helsinki and the ICH
Good Clinical Practice recommendations, participation in
human research must fulfill one fundamental prerequisite:
informed consent. Informed consent requires providing
enough information, both in terms of quantity and quality
(information must be understood by the subject) to allow
the subject to take an informed decision. However, it is
always difficult to determine the optimum amount and type
of information to give to the volunteers, and the opinion of
the study participants can be one approach to improve the
informed consent procedure [7]. In our study, volunteers
expressed a positive opinion in terms of the information
education assures both a better understanding of the
informed consent form and a lower susceptibility to accept
unfavorable situations or those perceived to be highly risky.
The few unemployed subjects enrolled as study participants
in the phase I study surveyed here were young
professionals who had recently completed their university
graduation and not yet found a job. In terms of net monthly
income, approximately 25% of the subjects earned less than
25% of the average national salary in the industry and
services professions (approximately €1000/month, after
taxes), and an additional 25% of subjects earn between
25% and 50%. The income is apparently low, but it should
be taken into account that most subjects were students and
singles, and a significant proportion of them lived with their
parents/family and, presumably, had no household costs.
In agreement to findings reported in the literature [1, 2,
7, 14–26], financial reward was the single, most important
motivation for volunteering as a participant in our population. However, other motivators also played a role, as we
found [27] and as has also been suggested by other authors
[28]. Among these other motivators, some can be considered to be pragmatic (such as the opportunity for a free
clinical check-up) and others may be altruistic in nature
(such as contributing to the development of new medicines). Although the relatively small sample size does not
allow us to draw definitive conclusions, we found no
correlation between the different motivators and gender,
age or occupation in our population. Van Gelderen et al. [7]
reported a difference in the reason for participation and the
age of volunteer (with the younger volunteers being more
motivated by money and the older volunteers being
especially motivated by the idea of a free medical
examination), but their population included aged people,
which is in sharp contrast to our population. The significant
inverse correlation between financial reward as a motivator
and monthly income was not unexpected, and can lead to
ethical concerns because economically disadvantaged volunteers may be particularly vulnerable [1, 24]. On the other
hand, we also found an inverse correlation between
financial reward and education, indicating that highly
educated people are less susceptible to monetary inducement. The study also shows that the reason(s) why a subject
participates in a study may be influenced by the subject’s
personality.
An unexpected finding was the significant proportion of
subjects who stated that they would volunteer for participation without the promise of a financial reward. The
subjects answered the questionnaire in an anonymous and
confidential way and, therefore, they had no reason for
saying anything but the truth. However, it is most likely
that at the time such participation would translate into
positive action, many of these subjects would actually
refuse to participate. The results do suggest that a number
34
31
CAPÍTULO I
1094
Eur J Clin Pharmacol (2007) 63:1085–1094
provided regarding the inconvenience or discomfort caused
by study participation, but it appears that a significant
proportion of subjects (44%) considered the amount of
information excessive (“more than enough”).
Subjects with higher anxiety trait tended to become more
anxious in relation to the possibility of an occurrence of a
serious complication during study participation.
In conclusion, financial reward was less valued as a
motivator for participation in phase I studies by healthy
volunteers with a higher income and education. A subject’s
motivations and perceptions toward participation would
appear to be influenced by the characteristics of the
individual, including some personality traits.
9. Costa PT, McCrae RR (1992) Revised NEO Personality Inventory
(NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI). Professional manual. Psychological Assessment Resources, Odessa, FL
10. Digman JM (1990) Emergence of the five factor. Annu Rev
Psychol 41:417–440
11. Goldberg LR (1990) An alternative “Description of Personality”:
the big-five factor structure. J Pers Soc Psychol 59:1216–1229
12. Pedroso-de-Lima M, Simões A (2000) NEO PI-R: Inventário de
Personalidade NEO Revisto. Manual profissional. CEGOC–TEA,
Ltd, Lisbon
13. Rocha D (2000) NEO PI-R: CEGOC system, software para
cotação informatizada do teste NEO-PI-R. CEGOC-TEA, Lisbon
14. Anderson JA, Weijer C (2001) The research subject as entrepeneur. Am J Bioethics 1:67–69
15. Ayd FJ, Calabresi P (1972) Motivations and rewards for
volunteering to be an experimental subject. Clin Pharmacol Ther
13:771–781
16. Dickert N, Grady C (1999) What’s the price of a research subject?
Approaches to payment for research participation. N Engl J Med
341:198–203
17. Erb K, Staks T, Su C-A (1994) Trends in the remuneration of
volunteers. Appl Clin Trials 3:54–58
18. Gribbin K, Schaie KW (1976) Monetary incentive, age, and
cognition. Exp Aging Res 2:461–468
19. Hyun I (2006) Fair payment or undue inducement? Nature
442:629–630
20. McGee C (1997) A piece of my mind: subject to payment? JAMA
278:199–200
21. Russel ML, Moralejo DG, Burgess ED (2000) Paying research
subjects: participant’s perspectives. J Med Ethics 26:126–130
22. Tishler CL, Bartholomae S (2003) Repeat participation among
normal healthy research volunteers: professional guinea pigs in
clinical trials? Perspect Biol Med 46:508–520
23. Vere DW (1991) Payments to healthy volunteers: ethical problems. Br J Clin Pharmacol 32:141–142
24. Viens AM (2001) Socio-economic status and inducement to
participate. Am J Bioethics 1:69
25. Weijer C (1999) What’s the price of a research subject? N Engl J
Med 341:206
26. Zink S (2001) Maybe we should pay them more. Am J Bioethics
1:88–89
27. Almeida L, Azevedo B, Nunes T, Vaz-da-Silva M, AlbinoTeixeira A, Falcao A, Coelho R, Soares-da-Silva P (2006)
Healthy subjects’ motivations to volunteer for participation in
human pharmacology studies (Abstr 1071). J Clin Pharmacol
46:1077
28. Newton L (1982) Inducement, due and otherwise. IRB 4:4–6
29. Rabin C, Tabak N (2006) Healthy participants in phase I clinical
trials: the quality of their decision to take part. J Clin Nurs
15:971–979
Acknowledgements We express our deep gratitude to the volunteers
for agreeing to participate, and to Grazina Binkauskiene, PhD, and
Vitalija Avdejenkova, PhD, for their assistance in the statistical
analysis.
References
1. Tishler CL, Bartholomae S (2002) The recruitment of normal
healthy volunteers: a review of the literature on the use of
financial incentives. J Clin Pharmacol 42:365–375
2. Bigorra J, Baños JE (1990) Weight of financial reward in the
decision by medical students and experienced healthy volunteers
to participate in clinical trials. Eur J Clin Pharmacol 38:443–446
3. Esecover H, Malitz S, Wilkens B (1961) Clinical profiles of paid
normal subjects volunteering for hallucinogen drug studies. Am J
Psychiatry 117:910–915
4. Hassar M, Pocelinko R, Weintraub M, Nelson D, Thomas G, Lasagna
L (1977) Free-living volunteer’s motivations and attitudes toward
pharmacologic studies in man. Clin Pharmacol Ther 21:515–519
5. Martin DC, Arnold JD, Zimmerman TF, Richart RH (1968)
Human subjects in clinical research: a report of three studies. N
Engl J Med 279:1426–1431
6. Novak E, Seckman CE, Stewart RD (1977) Motivations for
volunteering as research subjects. J Clin Pharmacol 365–371
7. Van Gelderen CEM, Savelkoul TJF, Van Dokkum W, Meulenbelt
J (1993) Motives and perception of healthy volunteers who
participate in experiments. Eur J Clin Pharmacol 45:15–21
8. Vrlovac R, Francetic I, Rotim K (1990) Drug trials on healthy
volunteers in Yugoslavia. Int J Clin Pharmacol Ther Toxicol
28:375–379
35
32
CAPÍTULO II
Caracterização da importância dos valores interpessoais na auto-selecção de
voluntários para ensaios clínicos de fase I
«Interpersonal values of healthy research volunteers participating in phase I drug studies»
Enviado para publicação
Caracterização da importância dos valores interpessoais na auto-selecção
de voluntários para ensaios clínicos de fase I
)NTERPERSONALVALUESOFHEALTHYRESEARCH
VOLUNTEERSPARTICIPATINGINPHASE)DRUGSTUDIES
,UIS!LMEIDA4ODD"+ASHDAN0ATTY:ORBAS2UI#OELHO!NTvNIO!LBINO4EIXEIRA0ATRqCIO3OARESDA3ILVA
$EPARTMENTOF2ESEARCHAND$EVELOPMENT")!,0ORTELA#O3!3-AMEDEDO#ORONADO0ORTUGAL
)NSTITUTEOF0HARMACOLOGYAND4HERAPEUTICS&ACULTYOF-EDICINE5NIVERSITYOF0ORTO!LAMEDA0ROFESSOR(ERNhNI-ONTEIRO0ORTO0ORTUGAL
$EPARTMENTOF0SYCHOLOGY'EORGE-ASON5NIVERSITY&AIRFAX6!53!
3ERVICEOF0SYCHIATRYAND-ENTAL(EALTH&ACULTYOF-EDICINE5NIVERSITYOF0ORTO0ORTUGAL
)NSTITUTEFOR-OLECULARAND#ELL"IOLOGY)"-#5NIVERSITYOF0ORTO0ORTUGAL
#ORRESPONDENCETO
$R,UIS!LMEIDA$EPARTMENTOF2ESEARCHAND$EVELOPMENT")!,!!VDA3IDERURGIA.ACIONAL3-AMEDEDO#ORONADO0ORTUGAL
0HONE&AX%MAILLUISALMEIDA BIALCOM
!"342!#4
0ARTICIPATIONINPHASE)STUDIESWITHDRUGSDEPENDSONINFORMEDCONSENT7EREPORTASTUDYASSESS
ING WHETHER THE±INTERPERSONAL VALUES²OFHEALTHYVOLUNTEERSFORPHASE)PARTICIPATIONAREDIFFERENT
FROM THENORMALPOPULATION/URSAMPLECONSISTEDOFCONSECUTIVEPHASE)STUDYVOLUNTEERS)N
TERPERSONALVALUESWEREASSESSEDUSINGTHE3URVEYOF)NTERPERSONAL6ALUES3)6ANDCOMPAREDTOTHE
SCALEVALIDATIONSAMPLEOFINDIVIDUALS3)6CONTAINSSIXSCALESWHICHCHARACTERISE THEFOLLOWING
VALUES3UPPORTBEINGTREATEDWITHCOMPASSION)NDEPENDENCEFREEDOMTOMAKEONE´SOWNCHOICES
"ENEVOLENCEGENEROSITYANDCAREFOROTHERS#ONFORMITYADHERENCETOSOCIALNORMSANDRULES,EADER
SHIPBEINGINCHARGEOFOTHERPEOPLEAND2ECOGNITIONBEINGADMIREDBEINGCONSIDEREDIMPORTANT
AND ACHIEVING RECOGNITION #OMPARED WITH THE SCALE VALIDATION SAMPLE PHASE ) STUDY PARTICIPANTS
REPORTEDGREATERVALUINGOF3UPPORT)NDEPENDENCEAND"ENEVOLENCEANDLOWERLEVELSOF #ONFORMITY
AND,EADERSHIPP´S/URRESULTSSUGGESTTHATTHEINTERPERSONALVALUESOFPARTICIPANTSINPHASE
)STUDIESDIFFERFROMTHEGENERALPOPULATIONWITHIMPLICATIONSFORINTERPRETINGCLINICALTRIALS
+%97/2$3
#LINICALTRIALSHEALTHYVOLUNTEERSINTERPERSONALVALUESPHASE)DRUGSTUDIES
).42/$5#4)/.
The development of new medicines is based on
non-clinical laboratory studies and clinical trials performed in humans aiming to characterize the drug
safety and efficacy before starting commercialization.
Before performing therapeutic Phase II and Phase III
studies with a new drug in patients with the target disease, a series of Phase I trials is typically conducted in
healthy research volunteers. These studies with healthy
volunteers aim to characterize the safety margin and
the adverse reactions profile, and to obtain pharmacokinetic (e.g., drug absorption, distribution, metabolism
and excretion, and drug-drug interactions) and pharmacodynamic information.
Ideally, samples of healthy volunteers selected for
phase I studies should be representative of the popula-
35
tion for whom the drug is intended. However, pragmatic issues of convenience, practicability and cost
prevent the recruitment of such an ideal population
(Tishler et al., 2005). The concerning effect of including healthy research volunteers with unrepresentative
characteristics is that they may be physiologically or
metabolically different from the target population
(Meyer, 2001). Variability in pharmacodynamic and
pharmacokinetic responses among healthy volunteers
during clinical trials may be due to several factors,
such as genetic background and differing physiological or environmental conditions (Morselli, 1990). Variability due to psychological conditions exists (Xu,
2007), but is largely overlooked (Tishler et al., 2005).
Although thousands of healthy volunteers participate in Phase I trials with drugs each year, relatively
few studies have examined their motivations and char-
CAPÍTULO II
acteristics (Tishler & Bartholomae, 2002). Phase I studies often involve inconvenient research facilities stays,
invasive medical procedures, restricted diets, and risks
of adverse side effects from the medication. Given these
risks and discomforts, one can question the motives for
participation in these studies.
Phase I trial samples are usually comprised of students and other people with less income. While monetary reward is often reported as a primary motivation
(Almeida et al., 2007; Ayd & Calabresi, 1972; Bigorra
& Baños, 1990; Tishler & Bartholomae, 2003; Van
Gelderen et al., 1993), many economically disadvantaged individuals reject participation in Phase I studies, suggesting that volunteering for participation is
a complex decision and influenced by factors beyond
financial incentive. There is some evidence suggesting
that personality traits have an influence on the decision
to participate in Phase I clinical trials (Almeida et al.,
2008a; Almeida et al., 2008b; Almeida et al., 2008c).
Traits that have been found to be elevated in Phase I
healthy volunteers in comparison to their peers in normal population include altruism (Bigorra & Baños,
1990), curiosity and exploratory behaviour (Almeida et
al., 2008c), sensation seeking and disinhibition (Farre et
al., 1995; Pieters et al., 1992). Individuals high in trait
anxiety and social avoidance are more likely to refuse
participation in Phase I studies (Almeida et al., 2008b).
Additional areas of individual difference which
may influence the decision to volunteer or participate
in Phase I studies are personal value orientations. Personal values represent beliefs about what is important
to an individual, such as opinions on what is fair, desirable, acceptable, or just (Beukman, 2005). These beliefs serve a guiding function, influencing life decisions
and how to invest time and energy when confronted
with competing options (Rohan, 2000). One group of
personal values – the interpersonal values –, may be
particularly relevant in the decision to participate in
Phase I studies. Interpersonal values are those influencing social behaviour and social relationships; this includes the motivation for doing or not doing things for
others, and the preferred pattern of relating to other
people (e.g., dimensions of dominance-submission and
hostile-friendly). The current study explores the importance of interpersonal values in healthy volunteers
participating in Phase I studies. It is hypothesized that
healthy research volunteers will differ from a normative sample in terms of the interpersonal values deemed
important.
0/05,!4)/.!.$-%4(/$3
Participants were 138 consecutive healthy volunteers enrolled in a Phase I drug trials with either a new
anti-epileptic drug or anti-parkinsonian medication
at the Human Pharmacology Unit of BIAL - Portela
& Cª SA (S Mamede do Coronado, Portugal). The current study was approved by the governing Independent
Ethics Committee. All aspects of subject recruitment,
screening, evaluation, and compensation for the Phase
I trials were approved by the National Ethics Committee for Clinical Research (CEIC - Comissão de Ética de
Investigação Clínica, Lisbon, Portugal) and the National Drug Regulatory Authority (INFARMED, Lisbon,
Portugal).
Prior to giving written consent for the Phase
I study, volunteers attended at least one information
meeting where they received study information in oral
and written forms. Putative participants were informed
of the phase I study design and procedures, potential
discomforts, known and possible unknown adverse
side effects and the approved financial compensation.
Known side effects ranged from headaches and nausea to
hair loss, skin rash, and risk of haemorrhage. The study
required residence in the research facilities totalling
up to two weeks, a restricted diet, and frequent blood
drawings for pharmacokinetic analysis. Volunteers were
made aware that there were no direct benefits for participation in the study outside of financial compensation.
After informed consent was obtained, participants
completed a version of the Survey of Interpersonal Values (SIV) (Gordon, 1993) that has been validated for
the Portuguese population (Gordon, 2001). The SIV is
a well-established measure in the area of personality assessment which provides measures of six interpersonal
value orientations. “Support” assesses the value of being
treated with understanding, kindness, and consideration and receiving encouragement from others. “Conformity” measures the value of adherence to regulations
and behaving in ways that are socially correct, accepted
and proper. “Recognition” measures the value of being admired, being considered important, attracting
favourable notice, and achieving recognition. “Independence” assesses the value of having freedom to make
one’s own decisions and the right to do what one wants
to do. “Benevolence” measures the value of being generous, sharing with others, helping the unfortunate, and
doing things for others. “Leadership” assesses the value
of being in charge of and having authority over others.
The SIV is a 30-item self-administered questionnaire. Each item is composed of three short phrases
describing three different situations. Participants were
instructed to read each of the statements and mark
one statement that was most important to them and one
statement that was least important to them, leaving the
third item unmarked. Each domain contains five items
which have shown good internal consistency — Support
(α=.67), Conformity (α=.81), Recognition (α=.69),
Independence (α=.78), Benevolence (α=.78), and
Leadership (α=.87) — and high test-retest reliability.
36
Caracterização da importância dos valores interpessoais na auto-selecção
de voluntários para ensaios clínicos de fase I
Scores from the SIV are computed and summarized along a percentile distribution, grouping scores
into 5-point percentile intervals. The authors (Gordon,
2001) developed a categorical scale in which scores
were further grouped into the following percentile categories: Very High (P95-99), High (P70-90), Average
(P35-65), Low (P10-30), and Very Low (P1-5). Using
the SIV Technical Manual software (Rocha, 2001), SIV
scores of the Phase I sample were transformed into percentile categories by referencing the scores to the Portuguese normative sample while accounting for gender
and age. Statistical differences on the percentile categories between the current sample and the normative
group for each of the interpersonal value domains were
then computed using the Chi-Square Goodness-of-Fit
Test. Statistical analyses were performed with the Statistical Package for Social Sciences version 11.5 (SPSS
Inc, Chicago, Ill).
Of the final sample of 138 healthy research volunteers, 46.4% were male (n=64) and 53.6% female
(n=74), and they were primarily Caucasian (97%), ranging in age from 18 to 43 years (mean±SD = 25.8±4.9
years). Students comprised the majority of the healthy
research volunteers (60.8%), 32.3% were employed,
and 6.9% were unemployed. Most subjects reported
monthly earnings of í500 or less (60.4%), with 86.1%
reporting a monthly income under í1000.
Our healthy research volunteers were compared to
the 807 individuals comprising the normative sample
when the SIV was developed (Gordon, 2001). Similar
to our healthy research volunteers, the majority of the
comparison sample were Caucasian (>95%) and 50.4%
female with a mean±SD age of 30.0±5.8 years (Gordon,
2001). The mean±SD scores in the normative sample
were 13.56±4.71 for Support, 15.43±5.86 for Conformity, 9.58±4.39 for Recognition, 16.67±6.08 for Inde-
pendence, 16.67±5.07 for Benevolence and 18.07±7.33
for Leadership (Gordon, 2001). In the normative sample, significant negative correlations were found between Support and Conformity (r=-.18, p<0.01), Support and Independence (r=-.20, p<0.01), Support and
Leadership (r=-.54, p<0.01), Recognition and Conformity (r=-.17, p<0.01), Independence and Conformity (r=-.40, p<0.01), Independence and Recognition
(r=-.44, p<0.01), Benevolence and Recognition (r=.35, p<0.01), Benevolence and Independence (r=-.20,
p<0.01), Leadership and Conformity (r=-.37, p<0.01),
and Leadership and Benevolence (r=-.46, p<0.01). A
significant positive correlation was found between Recognition and Support (r=.34, p<0.01) and Conformity and Benevolence (r=.16, p<0.01) (Gordon, 2001).
2%35,43
We initially examined correlations across the
six interpersonal values in the Phase I study healthy
volunteers sample. Significant negative correlations
were found between Support and Conformity (r=-.39,
p<0.001), Support and Leadership (r=-.32, p<0.01),
Recognition and Conformity (r=-.24, p<0.01), Independence and Conformity (r=-.34, p<0.001), Independence and Recognition (r=-.40, p<0.01), Benevolence
and Recognition (r=-.25, p<0.01), Benevolence and Independence (r=-.39, p<0.01), Leadership and Conformity (r=-.29, p<0.01), and Leadership and Benevolence
(r=-.41, p<0.001). A significant positive correlation
was found between Recognition and Support (r=.32,
p<0.01). Overall, these results are in high agreement
with the correlations described for the normative sample in the SIV technical manual (Gordon, 2001).
Figure 1 displays the frequency distributions of
SIV subscale scores for the Phase I study sample. Table 1
100
Percentile
75
50
25
0
Support
Conformity
Recognition
Independence
Benevolence
Leadership
&IGURE&REQUENCYDISTRIBUTIONSOF3)6SUBSCALESCORESFORTHE0HASE)STUDYSAMPLEOFHEALTHYRESEARCHVOLUNTEERSN"OXANDWHISKERSPLOTSINDICATEMEDIAN
QUARTILEANDRANGEVALUES
37
CAPÍTULO II
4!",%-EANSTANDARDDEVIATIONANDPERCENTILECATEGORIESFOREACHOFTHE3URVEYOF)NTERPERSONAL6ALUE3)6SUBSCALESINTHE
0HASE)STUDYSAMPLEOFHEALTHYRESEARCHVOLUNTEERSN
6!,5%3
-EAN3$
0%2#%.4),%#!4%'/29
PVALUE
6ERY(IGH
0
(IGH
0
!VERAGE
0
,OW
0
6ERY,OW
0
3UPPORT

#ONFORMITY

2ECOGNITION

)NDEPENDENCE 
"ENEVOLENCE

,EADERSHIP

#HISQUAREGOODNESSOF½TTESTCOMPARINGTHE0HASE)STUDYSAMPLEWITHTHENORMATIVESAMPLE
presents the descriptive statistics and percentile category distribution for the SIV subscales. Results from the
Chi-Square Goodness-of-Fit Test comparing the interpersonal value distributions of the Phase I study sample
with the normative sample are also presented. Phase
I study participants reported higher levels of importance on the values of Support (p<0.001), Independence
(p<0.001) and Benevolence (p<0.001) and significantly
lower levels of importance on the values of Conformity
(p<0.001) and Leadership (p<0.001) than the normative sample. While no differences were found on the
Recognition subscale (p=0.059), a trend was detected
with participants reporting more importance for the
value of Recognition than the normative group.
$)3#533)/.
Values represent an individual’s belief of what is
important in one’s life (Beukman, 2005) and these beliefs influence perceptions of self, others, and the world.
Values also guide people toward certain behaviours
and not others. The goal of this study was to examine
whether healthy volunteers enrolled in a Phase I clinical trial differed from a normative sample in terms of
their interpersonal value orientations. We found that
healthy volunteers participating in a Phase I study
placed higher importance on the values of Support, Independence and Benevolence, and lower importance on
the values of Conformity and Leadership than did their
peers from the normative group.
Results are consistent with the defining features
of these value orientations. Phase I study designs and
procedures are associated with potential risks. As individuals with non-conformist and independent value
orientations show a higher tendency to pursue novel
and risky experiences (Zuckerman, 1991), it is not surprising that these values were more important in the
Phase I study sample than the normative sample.
People who possess benevolent values often engage in prosocial behaviour aimed at the preservation
of human welfare (Bardi & Schwartz, 2003). Thus they
may be more inclined to put themselves at risk for the
potential to better the lives of others via clinical drug
trials. In addition, there was a negative relationship between the values of Benevolence and Independence and
no relationship was found between Benevolence and
Conformity. These findings suggest that Phase I volunteers may be comprised of unique subgroups, each with
varying personality characteristics.
Recent research on the personality traits of individuals volunteering for Phase I studies have found that
individuals low in trait anxiety and social avoidance
(Almeida et al., 2008b), and high in curiosity/exploration and self-efficacy (Almeida et al., 2008c) are more
willing to volunteer for phase I studies. Taken together,
financial motives, interpersonal value orientations, and
personality traits constitute self-selection biases which
could potentially impact Phase I study findings. There
is some research suggesting that personality characteristics may influence pharmacokinetics and pharmacodynamics (Meyer, 1992; Tishler et al., 2005; Xu,
2007) and the probability of presenting or reporting
adverse side effects (Almeida et al., 2008a; Almeida et
al., 2008b). Several personality characteristics are associated with monoamine functions. Although no oneto-one relationship exists between a single neurotransmitter and a personality trait, some neurotransmitters
play a key role in influencing behavioural tendencies.
Elevated levels of 5-hydroxytriptamine are associated
with increased anxiety levels and behavioural inhibition whereas behaviours high in approach tendencies
(e.g., curiosity and sensation seeking) have been linked
to dopamine and to a lesser extent norepinephrine
(Zuckerman, 1991). Currently, the research on the biological characteristics of individuals volunteering for
Phase I studies is limited.
38
Caracterização da importância dos valores interpessoais na auto-selecção
de voluntários para ensaios clínicos de fase I
Questions should be raised as to how these personality differences distinguish volunteer samples from
general population. One way to accomplish this would
be to implement the assessment of relevant personality
characteristics into study screening procedures so that
researchers can better understand their sample and any
limitations to the inferences they can make. Currently,
the research on personality traits unique to healthy research volunteers is relatively scarce and more studies
are needed to assess how personality dimensions specific
to healthy volunteers influence Phase I study outcomes.
!#+./7,%$'%-%.43
We thank the healthy research volunteers and the clinical
staff of the clinical pharmacology unit of BIAL – Portela
& Co, SA (especially M. Vaz-da-Silva, MD, PhD, Teresa
Nunes, MD, and Benedita Azevedo, BSc Psychol) for their
cooperation. We thank Dr Grazina Binkauskiene and Professor Amílcar Falcão for their assistance in the statistical
analysis.
2%&%2%.#%3
!LMEIDA , !ZEVEDO " .UNES 4 6AZDA
3ILVA - 3OARESDA3ILVA 0 7HY
HEALTHY SUBJECTS VOLUNTEER FOR PHASE )
STUDIESANDHOWTHEYPERCEIVETHEIRPAR
TICIPATION%UR*#LIN0HARMACOL
!LMEIDA , &ALCAO ! 6AZDA3ILVA -
#OELHO 2 !LBINO4EIXEIRA ! 3OARESDA
3ILVA 0 A 0ERSONALITY CHARACTERIS
TICSOFVOLUNTEERSTO0HASESTUDIESAND
LIKELIHOODOFREPORTINGADVERSEEVENTS)NT
*#LIN0HARMACOL4HER
!LMEIDA , +ASHDAN 4 .UNES 4 #OELHO
2 !LBINO4EIXEIRA ! 3OARESDA3ILVA 0
B7HOVOLUNTEERSFORPHASE)CLINI
CAL TRIALS )N¾UENCES OF ANXIETY SOCIAL
ANXIETY AND DEPRESSIVE SYMPTOMS ON
SELFSELECTION AND REPORTING OF ADVERSE
EVENTS%UR*#LIN0HARMACOL¯
!LMEIDA,+ASHDAN4"#OELHO2!LBINO
4EIXEIRA ! 3OARESDA3ILVA 0 C
(EALTHYSUBJECTSVOLUNTEERINGFOR0HASE)
STUDIESIN¾UENCEOFCURIOSITYEXPLORATO
RYTENDENCIESANDPERCEIVEDSELFEF½CACY
)NT*#LIN0HARMACOL4HER
!YD&*#ALABRESI0-OTIVATIONSAND
REWARDSFORVOLUNTEERINGTOBEANEXPERI
MENTAL SUBJECT #LIN 0HARMACOL 4HER "ARDI ! 3CHWARTZ 3( 6ALUES AND
BEHAVIOR3TRENGTHANDSTRUCTUREOFRELA
TIONS0ERS3OC0SYCHOL"ULL
"EUKMAN 4, #ULTURE VALUES AND
WORKRELATED VALUES ! THEORETICAL OVER
VIEW )N 4HE EFFECT OF SELECTED VARIABLES
ON LEADERSHIP BEHAVIOUR WITHIN THE
FRAMEWORKOFATRANSFORMATIONALORGANI
SATIONPARADIGM5NIVERSITYOF0RETORIA
"IGORRA*"AuOS*%7EIGHTOF½NAN
CIAL REWARD IN THE DECISION BY MEDICAL
STUDENTSANDEXPERIENCEDHEALTHYVOLUN
TEERSTOPARTICIPATEINCLINICALTRIALS%UR*
#LIN0HARMACOL
&ARRE-,AMAS8#AMI*3ENSATION
SEEKINGAMONGSTHEALTHYVOLUNTEERSPAR
TICIPATINGINPHASE)CLINICALTRIALS"R*#LIN
0HARMACOL
'ORDON ,6 3URVEY OF )NTERPERSONAL
6ALUES 4ECHNICAL -ANUAL 4HIRD %DITION
#HICAGO 3CIENCE 2ESEARCH !SSOCI
ATES
'ORDON ,6 )NVENTfRIO DE6ALORES )N
TERPESSOAIS3)6-ANUAL4mCNICO,ISBON
#%'/#4%!,TD
-EYER&0!SPEKTEEINERDIFFERENTIEL
LEN(UMANPSYCHOPHARMAKOLOGIE)N/L
DIGS+ERBER*,EONARD*0EDS0HARMAKO
PSYCHOLOGIE%XPERIMENTELLEUNDKLINISCHE
!SPEKTE&ISCHER*ENAPP
-EYER&00SYCHOANDIMMUNOPHAR
MACOLOGICALFACTORSRELEVANTTOSELECTION
OFVOLUNTEERSINCLINICALSTUDIES)NT*#LIN
0HARMACOL4HER
-ORSELLI0,!REPHASE)STUDIESWITH
OUTDRUGLEVELDETERMINATIONACCEPTABLE
&UNDAM#LIN0HARMACOLS¯S
0IETERS -3 *ENNEKENS3CHINKEL ! 3CHOE
MAKER (# #OHEN !& 3ELFSE
LECTION FOR PERSONALITY VARIABLES AMONG
HEALTHY VOLUNTEERS "R * #LIN 0HARMACOL
2OCHA $ 3)6 #%'/# 3YSTEM SOFT
WARE PARA COTAljO INFORMATIZADA DO
TESTE3)6#%'/#4%!,TD,ISBON
39
2OHAN-*!ROSEBYANYNAME4HE
VALUES CONSTRUCT 0ERS 3OC 0SYCHOL 2EV 4ISHLER #, "ARTHOLOMAE 3 4HE RE
CRUITMENTOFNORMALHEALTHYVOLUNTEERS
! REVIEW OF THE LITERATURE ON THE USE OF
½NANCIALINCENTIVES*#LIN0HARMACOL
4ISHLER #, "ARTHOLOMAE 3 2EPEAT
PARTICIPATION AMONG NORMAL HEALTHY RE
SEARCH VOLUNTEERS PROFESSIONAL GUINEA
PIGS IN CLINICAL TRIALS 0ERSPECT "IOL -ED
4ISHLER #, "ARTHOLOMAE 3 2HODES !2
0ERSONALITY PRO½LES OF NORMAL
HEALTHY RESEARCH VOLUNTEERS A POTENTIAL
CONCERN FOR CLINICAL DRUG TRIAL INVESTIGA
TORS-ED(YPOTHESES
6AN 'ELDEREN #%- 3AVELKOUL 4*& 6AN
$OKKUM 7 -EULENBELT * -O
TIVES AND PERCEPTION OF HEALTHY VOLUN
TEERSWHOPARTICIPATEINEXPERIMENTS%UR
*#LIN0HARMACOL
8U & %FFECT OF PERSONALITY TYPE ON
PHARMACODYNAMICS THROUGH CHANGING
PHARMACOKINETICS -ED (YPOTHESES :UCKERMAN - 0SYCHOBIOLOGY OF PER
SONALITY #AMBRIDGE 5NIVERSITY 0RESS
.EW9ORK
CAPÍTULO III
Os traços de personalidade como filtro de auto-selecção de participantes em
ensaios clínicos de fase I e sua influência nas avaliações de tolerabilidade
a) «Personality characteristics of volunteers in Phase 1 studies and likelihood of reporting
adverse events»
Int J Clin Pharmacol Ther 46, 340-348, 2008
b) «Healthy subjects volunteering for phase I studies: Influence of curiosity, exploratory
tendencies and perceived self-efficacy»
Int J Clin Pharmacol Ther 46, 109-118, 2008
c) «Who volunteers for phase I clinical trials? Influences of anxiety, social anxiety and
depressive symptoms on self-selection and reporting of adverse events»
Eur J Clin Pharmacol 64, 575-582, 2008
d) «The role of socio-economic conditions and psychological factors on the willingness to
volunteer for phase I studies»
Pharm Med 22, 2008 (aprovado para publicação; em impressão)
CAPÍTULO III
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
43
CAPÍTULO III
44
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
45
CAPÍTULO III
46
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
47
CAPÍTULO III
48
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
49
CAPÍTULO III
50
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
51
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
International Journal of Clinical Pharmacology and Therapeutics, Vol. 46 – No. 3/2008 (109-118)
Healthy subjects volunteering for Phase I
studies: Influence of curiosity, exploratory
tendencies and perceived self-efficacy
Original
©2008 Dustri-Verlag Dr. K. Feistle
ISSN 0946-1965
L. Almeida1,2 , T.B. Kashdan3, R. Coelho4, A. Albino-Teixeira2,5
and P. Soares-da-Silva1,2
1Department of Research and Development, BIAL (Portela & Co., SA), S Mamede
do Coronado, 2Institute of Pharmacology and Therapeutics, Faculty of Medicine,
University of Oporto, Porto, Portugal, 3Department of Psychology, George Mason
University, Fairfax, VA, USA, 4Service of Psychiatry and Mental Health, Faculty of
Medicine, and 5Institute for Molecular and Cell Biology (IBMC), University of Oporto,
Porto, Portugal
Curiosity and self-efficacy of Phase I volunteers
Key words
Phase I studies –
healthy volunteers –
personality – curiosity –
self-efficacy
Received
July 24, 2007;
accepted
November 6, 2007
Correspondence to
Dr. L. Almeida
Department of Research
and Development, BIAL
(Portela & Co., SA),
S Mamede do
Coronado, Portugal
[email protected]
Abstract. Objective: To test the hypothesis that trait-curiosity and perceived self-efficacy influence the willingness of healthy subjects to volunteer for participation in Phase I
studies. Materials and methods: A group of
healthy subjects who had never participated
in clinical studies (“index group”) were invited to participate in a Phase I study. They
were assessed with regard to trait curiosity
(Curiosity and Exploration Inventory; CEI-T)
and perceived self-efficacy (Self-Efficacy
Scale; SES) and subjects who accepted the invitation to participate were compared with
those who refused and with a group of healthy
subjects who had previously participated in
clinical studies (“validation group”). Results:
A significant positive correlation was found
between the willingness to participate and the
CEI-T total score (R = 0.28; p < 0.01), exploratory tendencies (R = 0.34; p < 0.001), SES total score (R = 0.30, p < 0.01), initiative and persistence (R = 0.29, p < 0.01), planning/goal setting (R = 0.19, p < 0.05) and social self-efficacy (R = 0.29; p < 0.01). The “index group”
subjects who accepted the invitation to participate showed significantly greater CEI-T exploratory tendencies (Z = –3.334, p = 0.001,
Mann-Whitney test) and total scores (Z =
–2.703, p < 0.01) and greater SES total score
(Z = –3.131, p < 0.01), initiative and persistence (Z = –3.065, p < 0.01), planning/goal
setting (Z = –2.173, p < 0.05) and social
self-efficacy (Z = –2.954, p < 0.01) than subjects who refused. No differences were found
between the subjects in the “index group”
who accepted the invitation and subjects in
the “validation group”. Using a logistic regression model, both CEI-T exploratory tendencies and SES initiative/persistence were
significant predictors of participation. Conclusion: Subjects higher in curiosity/explora64
53
tion and in perceived initiative/persistence
are more willing to volunteer for Phase I studies. The impact of these self-selection biases
on Phase I study results is unknown but deserves further evaluation.
Introduction
There are several reasons against participation in a Phase I clinical trial: participation
often requires frequent blood drawings and
other inconvenient procedures, implies confinement for several days, and sometimes
causes adverse effects; on the other hand,
subjects are healthy and it is not expected that
they will directly benefit from participating in
such studies. Therefore, one could ask why a
healthy subject is willing to volunteer.
The population samples in Phase I studies
can be considered self-selected because participation depends on informed consent.
Self-selection may complicate the interpretation of the results and represent a threat to
their generalizability. However, limited attention has been given to potential selfselection biases in volunteer enrolment and
few studies have examined whether the characteristics of healthy subjects who volunteer
to participate in clinical trials are different
from their peers who decline participation
[Tishler and Bartholomae 2002].
Prior investigations of volunteer motivation concluded that the financial incentive is
usually the main motive of whether or not to
participate [Almeida et al. 2007a, Bigorra and
CAPÍTULO III
110
Almeida, Kashdan, Coelho et al.
Baños 1990, Hassar et al. 1977, Van Gelderen
et al. 1993, Vrlovac et al. 1990]. Other commonly self-reported motivations for volunteering for Phase I studies include opportunities for free clinical check-ups, and contributing to scientific advances to help others, and
“curiosity” [Almeida et al. 2007a, Bigorra and
Baños 1990, Farre et al. 1995, Van Gelderen
et al. 1993]. However, it was not clear
whether curiosity corresponded just to a “curiosity-state” (which is largely based on individual interests, expectations and prior
knowledge) or would represent a “curiositytrait” (personality characteristic of whom has
the propensity to more readily enter in novel
and challenging experiences).
It has been reported that personality traits
such as altruism [Newton 1982], self-control,
self-confidence, emotional stability [Meyer et
al. 1995], and extraversion [Pieters et al. 1992]
may increase subject’s likelihood of participation. In our study, we were interested in two
particular personality traits that might differentiate people who volunteer for clinical trials
and those who refuse: curiosity and self-efficacy. Since both of these traits are related to
psychological, physical, and social well-being
[Kashdan 2004, Kashdan and Steger 2007,
Maddux 1995, Silvia 2006], if volunteers for
clinical trials differ from the rest of the population on these traits, potential problems might
arise in the interpretation of findings including
limits to their generalizability.
Curiosity can be defined as the recognition,
pursuit, and self-regulation of novel and challenging opportunities, often reflecting intrinsic
values and interests [Kashdan et al. 2004]. Trait
curiosity refers to a predisposition toward feelings of interest and curiosity and the resulting
exploration of novel, puzzling, and uncertain
situations [Kashdan and Fincham 2004]. A person high in “trait curiosity” prefers novelty,
complexity, uncertainty and conflict and shows
a preference for situations that are a match or
slightly exceed their skills and derive satisfaction from the process of learning and discovery.
Task or “state curiosity” is a transitory psychological state evoked by a specific interesting object or activity. We hypothesized that normal
healthy subjects with greater trait curiosity, who
are more exploratory by definition, are more
likely to volunteer for clinical trials and subjects
with less trait curiosity are more likely to view
these situations as undesirable.
65
54
Another personality trait related to a wide
range of health-related outcomes is self-efficacy [Bandura 1977]. Perceived self-efficacy
is defined as beliefs about one’s capacity to
self-generate behaviors to obtain desired outcomes even when confronted with barriers
and obstacles. People with high perceived
self-efficacy choose to perform more challenging tasks, invest more effort and persist
longer than those who are low in self-efficacy; people with low self-efficacy, doubting
their capabilities, avoid or abort difficult
tasks and often view them as personal threats
(as opposed to challenges) [Bandura 1994].
In terms of the willingness to perform difficult tasks, self-efficacy has been consistently
shown to be a more important determinant
than anticipatory anxiety or the amount of
prior task-relevant experience [Arch 1992,
Bandura 1997]. We hypothesized that normal
healthy subjects possessing the qualities inherent to self-efficacy would be more likely
to volunteer for clinical trials whereas the relative absence of self-efficacy would lead to
avoidance responses even in the face of
valued contingencies such as financial
incentives.
The primary objective of this study was to
evaluate the influence of trait curiosity (assessed by the Curiosity and Exploration Inventory – Trait, CEI-T [Kashdan et al. 2004,
http://mason.gmu.edu/�tkashdan/curiostywork.
html/]) and perceived self-efficacy (assessed by
the Self-Efficacy Scale, SES [Sherer et al.
1982]) on subjects’willingness to volunteer in a
group of normal healthy subjects already enrolled in Phase I drug studies (“validation
group”) and in a group of healthy subjects who
had never participated (“index group”) and
were invited to participate in a Phase I study.
Since there is a possibility of overlap between
curiosity and perceived self-efficacy, we also
examined the relative independent contribution
of each of these personality characteristics (and
their lower-facet dimensions) on the prediction
of participation.
Methods
Study populations
At the time of their participation in Phase I
clinical trials with new chemical entities cur-
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
111
Curiosity and self-efficacy of Phase I volunteers
Table 1.
Main demographic and socio-economic characteristics of the study populations.
Index group
(IG)
Variable
Gender
Accepted
(n = 51)
Refused
(n = 59)
(n = 216)
47.1%/52.9%
42.4%/57.6%
45.8%/54.2%
26.3 ± 6.6
26.6 ± 7.3
26.1 ± 5.3
Median (years)
25
24
25
Range (years)
18 – 45
18 – 45
18 – 45
Male/Female
Mean ± SD (years)
Age
Ethnicity
Occupation
Caucasian/Other
92.2%/7.8%
98.3%/1.7%
95.8%/4.2%
Student
56.0%
55.9%
57.3%
Employed
38.0%
32.2%
36.7%
Unemployed
6.0%
11.9%
6.0%
< 25%
30.0%
28.6%
27.3%
35.0%
24.5%
28.5%
25.0%
28.6%
29.7%
10.0%
12.2%
9.9%
0.0%
4.1%
4.7%
Single
72.5%
67.2%
77.9%
Married/living together
23.5%
27.6%
17.1%
Divorced
3.9%
5.2%
5.0%
4 years
2.0%
6.8%
0.0%
6 years
3.9%
3.4%
1.0%
9 years
9.8%
5.1%
7.0%
12 years
60.8%
55.9%
63.8%
Monthly net income
25 – 50%
in relation to the
51 – 100%
mean national net
salary in the industry
101 – 150%
*
and services sector
> 150%
Civil status
School degree
completed
Validation
group
Bachelor
5.9%
6.8%
7.5%
Licensed
15.7%
20.3%
17.6%
Mastership
2.0%
1.7%
3.0%
*approximately € 1,000/month, net.
rently in clinical development as new putative
medicines for the treatment of neurological
diseases conducted at the Human Pharmacology Unit of BIAL (Portela & Co. SA,
S Mamede do Coronado, Portugal), a group
of healthy volunteers (n = 216) were invited
to participate in the current study. In accordance with the current regulatory requirements, their participation in the phase I studies had been preceded by standard informed
consent procedures, including full information regarding the risks and inconveniences
of study participation. All phase I participants
who were invited to participate in the current
study accepted the invitation.
A group of normal healthy subjects
(n = 119) who had never participated in a clin-
66
55
ical trial was constituted (“index group”).
Since it is known that healthy subject’s
socio-economic level [Viens 2001], income
[Tishler and Bartholomae 2002], educational
level and age [Van Gelderen et al. 1993] may
influence the willingness to volunteer for
Phase I participation, efforts were made to assure that the “index group” would match the
Phase I participants sample in terms of main
demographic (age, gender, ethnicity and occupation) and socio-economic (monthly income, civil status and school degree completed) characteristics.
Subjects of the “index group” were also
provided with the typical informed consent
information prepared for a real Phase I study
and invited to participate. Among the 110
CAPÍTULO III
112
Almeida, Kashdan, Coelho et al.
study, and two subgroups (“accepted” and
“refused”) were constituted accordingly.
Main demographic and socio-economic characteristics of the study populations are
presented in Table 1.
Self-report measures
Figure 1. CEI-T scores in the “index group” (IG)
subsets of subjects who “refused” and who “accepted” the invitation for participation in a Phase I
study, and in the “validation group” (previous participants in Phase I studies): Box-and-whiskers plot
showing median, quartiles and range values.
subjects who answered the invitation, 51
(46.4%) subjects declared that they would be
willing to participate and 59 (53.6%) individuals refused participation in such Phase I
67
56
Subjects of both groups were invited to
complete Portuguese adaptations of the CEI-T
[Kashdan et al. 2004], to operationalize trait
curiosity and exploratory tendencies, and the
SES [Ribeiro 1995], to operationalize trait
self-efficacy.
The 7-item CEI-T assesses two relatively
independent dimensions of curiosity: (1) Exploration – general strivings to seek out novelty, challenge, uncertainty, and intriguing
experiences (4 items; e.g., “Everywhere I go,
I am out looking for new things or experiences”) and (2) Absorption or Flow – the propensity to be deeply absorbed in the interesting target of one’s attention (3 items; e.g.,
“When I am participating in an activity, I tend
to get so involved that I lose track of time”)
[Kashdan 2002]. The 2-dimensional approach of the CEI-T focuses on the defining
features of curiosity rather than different objects that induce curiosity [Kashdan 2002].
Respondents rate the items according to a
7-point Likert-type scale that indicate agreement with the statements provided from
“Strongly Disagree” to “Strongly Agree”.
The original English version of the CEI-T
has good psychometric properties, is relatively unaffected by socially desirable responding, and is relatively independent from
positive affect [Kashdan et al. 2004]. The
Portuguese CEI-T version was submitted to
confirmatory factor analysis and also showed
good psychometric properties. A model comparison procedure introduced by Bollen
[1980] was used to test whether the 2-factor
model with Exploration and Absorption as
separate, but correlated components of curiosity provided a better fit to the data than a
one-factor model. The hypothesized 2-factor
model was compared to a model where the
zero order association between the two dimensions of curiosity was constrained. The
models could be directly compared by interpreting the change in the �2 (per change in df).
The 2-factor model with Exploration and Ab-
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
113
Curiosity and self-efficacy of Phase I volunteers
Figure 2. SES scores in the “index group” (IG) subsets of subjects who “refused” and who “accepted” the
invitation for participation in a Phase I study, and in the “validation group” (previous participants in Phase I
studies): Box-and-whiskers plot showing median, quartiles and range values.
sorption as separate, but correlated components of curiosity, fit the data well, �2 = 24.56,
p = 0.03, �2/df = 1.89, TLI = 0.97, CFI = 0.98,
RMSEA = 0.05, SRMR = 0.03. Allowing Exploration and Absorption to covary freely in a
2-factor model resulted in a significant improvement in model fit, �2 (1) = 11.12, p =
0.001. Loadings on the Exploration factor
ranged from 0.41 – 0.82, loadings on the
Absorption factor ranged from 0.57 – 0.85,
and the two factors correlated at 0.74. Thus,
the 2-dimensional CEI-T is synonymous with
the English version.
The 23-item SES [Sherer et al. 1982] focuses on the willingness to initiate behavior,
willingness to expend effort in completing
behavior, and persistence in the face of adversity. 6, 5 and 4 items compose the Initiative/Persistence, Planning/Goal Setting, and
Social Self-Efficacy subscales, respectively.
It has shown to be a valid predictor of aca-
68
57
demic and professional behavior [Ferrari et
al. 1992, Sherer et al. 1982], self-esteem
[Woodruff and Cashman 1993], and general
adjustment [Martin et al. 1996, Sherer and
Adams 1983]. In our study, we used a 15-item
Portuguese validated adaptation of the original scale [Sherer et al. 1982], performed by
Ribeiro [1995]. The Portuguese adapted scale
shows good psychometric properties [Ribeiro
1995]. The items are rated on a 7-point
Likert-type scale, ranging from 1 (“Fully Disagree”) to 7 (“Fully Agree”).
Statistical considerations
Statistical analyses were performed with
the Statistical Package for Social Sciences
version 11.5 (SPSS Inc., Chicago, IL, USA).
The criterion for statistical significance was
set at an � error of 5% (p < 0.05), 2-sided.
CAPÍTULO III
114
Almeida, Kashdan, Coelho et al.
Table 2. Spearman’s correlation between CEI-T and SES scores and interest
into participation in a Phase I study.
Scales
Spearman’s rank correlation
CEI-T
Total
R = 0.28
p = 0.004
Exploration
R = 0.34
p = 0.000
Absorption
R = 0.14
n.s.
Total
R = 0.30
p = 0.002
Initiative and Persistence
R = 0.29
p = 0.002
Planning/Goal Setting
R = 0.19
p = 0.048
Social Self-Efficacy
R = 0.29
p = 0.002
SES
coxon-Mann-Whitney rank sum test was
used for the comparisons. In the “index
group”, Spearman’s correlations between the
CEI-T and SES scores and the answer to the
invitation to participate were performed.
In addition to univariate analyses, the independent contribution of each CEI-T and
SES sub-scale on the prediction of study participation in the “index group” was tested by
using logistic regression.
Ethics
n.s. = not statistically significant.
This study was approved by an Independent Ethics Committee (Comissão de Ética
Independente da UFH, Porto, Portugal).
Table 3. Statistical comparison (Wilcoxon-Mann-Whitney rank sum test) of
CEI-T and SES scores between subjects of the “index group” (IC) who “accepted” the invitation to participate in a Phase I study versus those who “refused”
(A), and versus the “validation group” (B).
Results
(A)
IC - Accepted vs.
IC - Refused
Scales
(B)
IC - Accepted
vs. Validation
group
CEI-T
Total
Z = –2.703
p = 0.007
Z = –1.247
n.s.
Exploration
Z = –3.334
p = 0.001
Z = –0.454
n.s.
Absorption
Z = –1.366
n.s.
Z = –1.839
n.s.
Total
Z = –3.131
p = 0.002
Z = –0.456
n.s.
Initiative and Persistence
Z = –3.065
p = 0.002
Z = –0.621
n.s.
Planning/Goal Setting
Z = –2.173
p = 0.030
Z = –0.164
n.s.
Social Self-Efficacy
Z = –2.954
p = 0.003
Z = –0.140
n.s.
SES
CEI-T and SES univariate
analysis
CEI-T and SES scores are displayed in
Figures 1 and 2. The results of the Spearman’s
correlation between CEI-T and SES scores
and interest into participation in the “index
group” are presented in Table 2. Statistical
comparison of CEI-T and SES scores between “index group” subjects who accepted
the invitation for participation in a Phase I
study versus those who refused participation,
and versus those subjects who had previously
participated (“validation group”) are presented
in Table 3.
n.s. = not statistically significant.
Trait curiosity
The first set of analyses consisted of comparing CEI-T and SES scores to examine if
there were differences between those subjects
in the “index group” who declared they
would accept to participate and those who refused. An internal validation was performed
by comparing the “index group” subset of
subjects who declared they would accept to
participate with the “validation group”. Tests
for normality were performed using the
Shapiro-Wilk test. Since the dependent variables were not normally distributed, the Wil-
69
58
Within the “index group”, respondents
who accepted the invitation for participating
in a Phase I study showed significantly greater
Exploration score (Z = –3.334, p = 0.001) and
Total Score (Z = –2.703, p < 0.01) than those
who refused; no significant difference was
found in Absorption. No significant differences were found between those “index
group” subjects who declared they would accept to participate and the “validation group”.
In the “index group”, a significant positive
linear correlation was found between the interest in participation and the CEI-T Total
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
115
Curiosity and self-efficacy of Phase I volunteers
Table 4.
Summary of logistic regression analysis of CEI-T and SES scores as predictors of participation in a Phase I study.
Scales
�
SE
Exp(B)
95%CI for Exp(B)
Exploration
0.174
0.088
1.190
1.002, 1.413
Absorption
–0.059
0.087
0.943
0.795, 1.119
Initiative and Persistence
0.098
0.049
1.103
1.003, 1.214
Planning/Goal Setting
–0.040
0.064
0.961
0.849, 1.089
Social Self-Efficacy
0.081
0.073
1.085
0.940, 1.251
–6.629
1.928
0.001
CEI-T
SES
Constant
p < 0.05
n.s.
p < 0.05
n.s.
n.s.
p = 0.001
� = regression coefficient; SE = standard error of the mean; Exp(B) = odds ratios; CI = confidence interval; n.s. = not statistically significant.
Score (R = 0.28, p < 0.01) and CEI Exploration (R = 0.34, p < 0.001).
ble 4. Both CEI-T Exploration (p < 0.05) and
SES Initiative and Persistence (p < 0.05)
were found to be statistically significant
predictors of participation.
Perceived self-efficacy
Within the “index group”, respondents
who accepted the invitation for participating
in a Phase I study also showed significantly
greater scores on all variables: Total Score (Z
= –3.131, p < 0.01), Initiative and Persistence
(Z = –3.065, p < 0.01), Planning/Goal Setting
(Z = –2.173, p < 0.05) and Social Self-Efficacy (Z = –2.954, p < 0.01). No significant
differences were found between those “index
group” subjects who declared they would accept to participate and the “validation group”.
In the “index group”, a significant positive
linear correlation was found between the interest in participation and the SES Total Score
(R = 0.30, p < 0.01), Initiative and Persistence (R = 0.29, p < 0.01), Planning/Goal Setting (R = 0.19, p < 0.05) and Social Self-Efficacy (R = 0.29, p < 0.01).
Multivariate analysis on the
prediction of study participation
To explore the contributions of each
CEI-T and SES component to the prediction
of participation, a logistic regression model
was composed in the “index group” using
CEI-T Exploration and Absorption, and SES
Initiative and Persistence, Planning/Goal
Setting and Social Self-Efficacy as covariates.
The results of the model are presented in Ta-
70
59
Discussion
This study examined whether normal
healthy subjects participating and willing to
volunteer for Phase I clinical trials differ from
those who do not take part, in terms of tendencies to be curious, exploratory, and self-efficacious. Supporting our hypotheses, subjects
who were more curious and self-efficacious
were more likely to volunteer for participation in Phase I studies. There is potential
overlap between the personality traits of curiosity and self-efficacy, as people possessing
these qualities are prone to view difficult situations as challenges rather than threats and
they are more likely to engage in approach behavior even in the face of avoidance motives
to prevent negative outcomes. Upon controlling for shared variance among the CEI-T and
SES sub-scales, we found that the unique predictors of participation were CEI-T Exploration and SES Initiative and Persistence. Thus,
these traits appear to be of particular importance in understanding the willingness (and
reluctance) to volunteer for Phase I clinical
trials.
The results are consistent with the theoretical framework. Participation in Phase I studies with drugs in clinical development may be
perceived as a challenging situation because
CAPÍTULO III
116
Almeida, Kashdan, Coelho et al.
it involves the discomfort, performance of
unaccustomed procedures, close social interactions with the clinical staff and other participants, the risk of occurrence of expected and
unexpected adverse events, and, according to
current Good Clinical Practices and regulatory requirements, the volunteers must be
made aware about all such inconveniences
and risks before giving their informed consent. Since highly curious and self-efficacious subjects show a higher tendency to embrace and even thrive in challenging situations
[Kashdan and Silvia (in press), Maddux
1995], it seems plausible that they are more
likely to view a Phase I study as an appealing
and satisfying situation.
During the clinical development process
of a new medicine, normal healthy volunteers
are utilized in Phase I clinical trials to provide
pharmacokinetic, pharmacodynamic and tolerability information to help to determine the
doses to be used in future Phase II and III clinical trials and, afterwards, in the target patient population. Ideally, samples of normal
healthy volunteers selected for Phase I studies
should be representative of the population for
whom the drug is intended. However, pragmatic issues of convenience, practicability
and cost prevent the recruitment of such an
ideal population [Tishler et al. 2005]. Although it is accepted that healthy volunteers
do not necessarily represent the target population, it is assumed that they should represent
at least their age group [Pieters et al. 1992].
However, our study corroborates data from
other studies [Meyer 2001] suggesting that
participants in Phase I studies differ in important ways from the population from which
they are drawn. For example, healthy people
who volunteer for Phase I studies appear to
show a greater tendency toward sensationseeking and to be more extraverted and selfconfident than normative samples [Ball et al.
1993, Cami et al. 1989, Farre et al. 1995, Pieters
et al. 1992], and volunteers were also shown
to be less socially anxious [Almeida et al.
2007b] and depressed [Almeida et al. 2007c]
than controls.
The overarching aim of this line of research is to determine whether normal healthy
subjects differ from the population of interest
in a meaningful way [Tishler et al. 2005]. The
self-selection bias will be important only if it
interferes on the study outcomes or upon the
71
60
interpretation of the studies. There is evidence that some personality characteristics
may influence the pharmacodynamic response [Meyer 1992], objective pharmacokinetic data [Meyer 2001] and subjective
tolerability information gathered from normal healthy volunteers [Meyer et al. 2000].
For instance, there are suggestions that extroverted personality characteristics influence
the subjective tolerability and physiological
response of normal healthy subjects to drug
administration [Meyer 2001]. Also, there is
some evidence that volunteer’s type of personality may influence the subjective response to placebos, which are commonly
used in Phase I clinical trials with new drugs,
and therefore impair the evaluation of new
drugs in such studies [Drici et al. 1995]. However, the variability in drug response and/or
pharmacokinetics due to personality or psychological features is an overlooked topic,
systematic research in this field is still missing, and the impact of the self-selection bias
on the study results is largely unknown
[Tishler et al. 2005]. Once known, the characterization of the volunteers’ personality and
psychological characteristics that actually
may affect their subjective response and
pharmacokinetic parameters could be addressed in the screening process for controlled Phase I studies, through the formation
of balanced groups [Meyer et al. 2000], or in
the statistical analysis [Tishler et al. 2005].
Conclusion
Results from our study, which to our
knowledge was the first to use a control
group, suggested that participants in Phase I
studies show a preference to seek novel, uncertain, and complex situations (as shown by
greater CEI-T Exploration scores) and are
more willing to initiate behavior and to persist
(as shown by greater SES Initiative and Persistence scores), compared with their control
peers. Results were consistent across comparisons between subjects who had never participated in a Phase I study and who declared that
they would consider participating versus
those who declined invitation to participate
(behavioral intentions), and between subjects
who accepted the invitation versus a group of
subjects who had previously participated.
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
Curiosity and self-efficacy of Phase I volunteers
These findings may have implications for interpreting threats to the generalizability of the
results of Phase I studies. The impact of these
self-selection biases on the Phase I study results is unknown but deserves further evaluation.
Acknowledgments
We thank the healthy subjects and the
clinical staff of the clinical pharmacology
unit of BIAL – Portela & Co., SA (especially
Professor Manuel Vaz-da-Silva, Dr. Teresa
Nunes and Benedita Azevedo) for their cooperation. We thank Professor Amílcar Falcão
and Dr. Grazina Binkauskiene for their assistance in the statistical analysis.
References
Almeida L, Azevedo B, Nunes T, Vaz-da-Silva M, Soaresda-Silva P. Why healthy subjects volunteer for Phase I
studies and how they perceive their participation?. Eur
J Clin Pharmacol. 2007a; 63: 1085-1094.
Almeida L, Azevedo B, Rocha J, Nunes T, Falcao A,
Vaz-da-Silva M, Coelho R, Albino-Teixeira A, Soaresda-Silva P. Fear of negative evaluation and social
avoidance and distress of subjects who volunteer to
Phase I studies versus a control group. Clin Pharmacol
Ther. 2007b; 81: S29-30 (abstract).
Almeida L, Azevedo B, Nunes T, Falcao A, Vaz-da-Silva
M, Coelho R, Albino-Teixeira A, Soares-da-Silva P.
Mood of subjects who volunteer to Phase I studies versus a control group. Clin Pharmacol Ther. 2007c; 81:
S29 (abstract).
Arch E. Affective control efficacy as a factor in willingness to participate in a public performance situation.
Psychology Reports. 1992; 70: 3-9.
Ball CJ, McLaren PM, Morrison PJ. The personality
structure of “normal” volunteers. Br J Clin Pharmacol.
1993; 36: 369-371.
Bandura A. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev. 1977; 84: 191-215.
Bandura A. Self-efficacy. In: Ramachaudran VS (ed).
Encyclopedia of human behavior. Vol 4. New York:
Academic Press; 1994. p. 71-81.
Bandura A. Self-efficacy: The exercise of control. New
York: Freeman; 1997.
Bigorra J, Baños JE. Weight of financial reward in the
decision by medical students and experienced healthy
volunteers to participate in clinical trials. Eur J Clin
Pharmacol. 1990; 38: 443-446.
Bollen KA. Issues in the comparative measurement of political democracy. American Sociological Review.
1980; 45: 370-390.
Cami J, Llorente M, Farre M, Badenas J-M, Ugena B.
Personality of healthy volunteers participating in
Phase I clinical trials. Person Individ Diff. 1989; 10:
1199-1200.
72
61
117
Drici MD, Raybaud F, De Lunardo C, Iacono P, Gustovic
P. Influence of the behaviour pattern on the nocebo response of healthy volunteers. Br J Clin Pharmacol.
1995; 39: 204-206.
Farre M, Lamas X, Cami J. Sensation seeking amongst
healthy volunteers participating in Phase I clinical trials. Br J Clin Pharmacol. 1995; 39: 405-409.
Ferrari JR, Parker JT, Ware CB. Academic procrastination: Personality correlates with Myers Briggs types,
self-efficacy, and academic locus of control. Journal
of Social Behavior and Personality. 1992; 7: 495-502.
Hassar M, Pocelinko R, Weintraub M, Nelson D, Thomas
G, Lasagna L. Free-living volunteer’s motivations and
attitudes toward pharmacologic studies in man. Clin
Pharmacol Ther. 1977; 21: 515-519.
Kashdan TB. Social anxiety dimensions, neuroticism,
and the contours of positive psychological functioning. Cognitive Therapy and Research. 2002; 26:
789-910.
Kashdan TB. Curiosity. In: Peterson C, Seligman MEP
(eds). Character strengths and virtues: A handbook
and classification. Washington, DC: American Psychological Association and Oxford University Press;
2004. p. 125-141.
Kashdan TB, Fincham FD. Facilitating curiosity: A social and self-regulatory perspective for scientifically
based interventions. In: Linley PA, Joseph S (eds).
Positive psychology in practice. Hoboken, NJ: John
Wiley & Sons; 2004. Chapter 30, p. 483-503.
Kashdan TB, Rose P, Fincham FD. Curiosity and exploration: Facilitating positive subjective experiences
and personal growth opportunities. J Person Assess.
2004; 82: 291-305.
Kashdan TB, Steger MF. Curiosity and pathways to
well-being and meaning in life: Traits, states, and everyday behaviors. Motivation and Emotion. 2007; 31:
159-173.
Maddux JE. Self-efficacy, adaptation, and adjustment:
Theory, research and application. New York: Magnum; 1995.
Martin TR, Flett GL, Hewitt PL, Krames L, Szanto G. Personality correlates of depression and health symptoms: A test of a self-regulation model. Journal of Research in Personality. 1996; 31: 264-277.
Meyer FP. Aspekte einer differentiellen Human-Psychopharmakologie. In: Oldigs-Kerber J, Leonard JP (eds).
Pharmakopsychologie. Experimentelle und klinische
Aspekte. Jena: Fischer; 1992. p. 187-200.
Meyer FP. Psycho- and immunopharmacological factors
relevant to selection of volunteers in clinical studies.
Int J Clin Pharmacol Ther. 2001; 39: 300-310.
Meyer FP, Haschke R, Rohl FW. The healthy volunteer in
clinical pharmacology: personality and motivation.
Eur J Clin Pharmacol. 1995; 48: 91-96.
Meyer FP, Troger U, Rohl FW. Reliability of symptom reports by healthy volunteers treated with placebo over
several time periods. Arzneimittelforschung. 2000;
50: 768-771.
Newton L. Inducement, due and otherwise. IRB: A Review of Human Subjects Research. 1982; 4: 4-6.
Pieters MS, Jennekens-Schinkel A, Schoemaker HC, Cohen AF. Self-selection for personality variables among
healthy volunteers. Br J Clin Pharmacol. 1992; 33:
101-106.
Ribeiro JLP. Adaptação de uma escala de avaliação da
auto-eficácia geral. In: Almeida L, Ribeiro I (eds).
Avaliação psicológica: formas e contextos. Braga:
Apport. 1995; p. 163-176.
CAPÍTULO III
118
Almeida, Kashdan, Coelho et al.
Sherer M, Adams CH. Construct validation of the
Self-Efficacy Scale. Psychological Reports. 1983; 53:
899-902.
Sherer M, Maddux JE, Mercandante B, Prentice-Dunn S,
Jacobs B, Rogers RW. The self-efficacy scale: Construction and validation. Psychological Reports. 1982;
51: 663-671.
Silvia PJ. Exploring the Psychology of Interest. New
York: Oxford University Press; 2006.
Tishler CL, Bartholomae S. The recruitment of normal
healthy volunteers: A review of the literature on the
use of financial incentives. J Clin Pharmacol. 2002;
42: 365-375.
Tishler CL, Bartholomae S, Rhodes AR. Personality profiles of normal healthy research volunteers: a potential
concern for clinical drug trial investigators? Med Hypotheses. 2005; 65: 1-7.
Van Gelderen CEM, Savelkoul TJF, Van Dokkum W,
Meulenbelt J. Motives and perception of healthy volunteers who participate in experiments. Eur J Clin
Pharmacol. 1993; 45: 15-21.
Viens AM. Socio-economic status and inducement to participate. Am J Bioethics. 2001; 1: 1F-2F.
Vrlovac R, Francetic I, Rotim K. Drug trials on healthy
volunteers in Yugoslavia. Int J Clin Pharmacol Ther
Toxicol. 1990; 28: 375-379.
Woodruff SL, Cashman JF. Task, domain, and general efficacy: A reexamination of the Self-Efficacy Scale.
Psychological Reports. 1993; 72: 423-432.
73
62
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
Eur J Clin Pharmacol (2008) 64:575–582
DOI 10.1007/s00228-008-0468-8
CLINICAL TRIAL
Who volunteers for phase I clinical trials? Influences
of anxiety, social anxiety and depressive symptoms
on self-selection and the reporting of adverse events
Luis Almeida & Todd B. Kashdan & Teresa Nunes &
Rui Coelho & António Albino-Teixeira &
Patrício Soares-da-Silva
Received: 6 November 2007 / Accepted: 31 January 2008 / Published online: 5 March 2008
# Springer-Verlag 2008
Abstract
Objective To investigate the influence of anxiety, social
anxiety and depressive symptoms on the willingness of
healthy subjects to volunteer for phase I studies and to
report adverse events.
Materials and methods A group of healthy subjects who had
never participated in a clinical trial (“Naïve Subjects”) were
invited to participate in a phase I study. All subjects were
assessed for trait anxiety (State-Trait Anxiety Inventory, STAIT), social anxiety (Social Avoidance and Distress, SAD, and
Fear of Negative Evaluation, FNE) and depressive symptom-
atology (Beck Depression Inventory, BDI-II). Subjects who
accepted the invitation to participate were compared with those
who refused. The personality traits of a group of “Actual
Participants” were examined, and the relation of these traits to
adverse events reported during participation was evaluated.
Results A significant inverse correlation was found between
the STAI-T (R=−0.203, p<0.05) and SAD (R=−0.204, p<
0.05) scores and the willingness to participate. Naïve
Subjects who refused the invitation to participate showed
higher scores on STAI-T (Z=−2.600, p<0.01) and SAD
(Z=−2.524, p<0.05) inventories. Logistic regression using
BDI-II, STAI-T, SAD and FNE as covariates also showed
that the only unique predictors of participation were the
STAI-T ( p<0.05) and SAD ( p<0.01) scores. Significant
positive correlations were found between trait anxiety and
reporting of adverse events.
Conclusion Participants in phase I studies are a selfselected sample defined by low trait-anxiety and social
avoidance behaviors. This self-selection bias may affect the
study results because less anxious subjects tend to report
fewer adverse events. The characterization of a participant’s
personality traits may be important in phase I studies.
L. Almeida (*) : T. Nunes : P. Soares-da-Silva
Department of Research and Development,
BIAL (Portela & Co, SA),
A Av. da Siderurgia Nacional,
4745–457 S Mamede do Coronado, Portugal
e-mail: [email protected]
L. Almeida : A. Albino-Teixeira : P. Soares-da-Silva
Institute of Pharmacology and Therapeutics, Faculty of Medicine,
University of Porto,
Alameda Professor Hernâni Monteiro,
Porto, Portugal
T. B. Kashdan
Department of Psychology, George Mason University,
Fairfax, VA, USA
Keywords Adverse events . Anxiety . Depressive
symptoms . Healthy volunteers . Phase I studies .
Social anxiety
R. Coelho
Service of Psychiatry and Mental Health, Faculty of Medicine,
University of Porto,
Rua do Campo Alegre,
Porto, Portugal
Introduction
A. Albino-Teixeira
Institute for Molecular and Cell Biology (IBMC),
University of Porto,
Alameda Professor Hernâni Monteiro,
Porto, Portugal
Participation in phase I studies implies (1) subjection to
unaccustomed procedures, some of which are relatively
invasive and painful, such as venous catheterization, (2) the
risk of occurrence of expected or unexpected drug adverse
74
63
CAPÍTULO III
576
Eur J Clin Pharmacol (2008) 64:575–582
ipants”) and a group of healthy subjects who had never
participated in a phase I clinical trial (“Naïve Subjects”) and
who matched the actual participants in terms of demographic and socio-economic characteristics. A total of 200
subjects in the Actual Participants group and 100 subjects
in the Naïve Subjects group were estimated to provide a
power of 80% to detect a difference of 2.5 in mean scales
scores, at a significance level of 0.05. The subjects did not
receive any financial compensation for participating in this
study.
events and (3) close social interactions with the research
staff and often confinement for several days in research
facilities with other participants. Consequently, participation in phase I studies can be perceived as a challenging
event and that a number of personality characteristics will
most likely have an effect on the subject’s willingness to
volunteer.
Although anxiety can manifest in a number of different
ways, participation in a phase I study appears to support the
investigation of two types of anxiety—trait anxiety and
social anxiety. For example, potential adverse effects or
painful procedures may be considered especially threatening by anxious people. Additionally, the close interaction
with previously unknown people, such as clinical staff and
other study participants, and confinement to an open-space
ward with minimal privacy are conditions that can be
perceived as threatening by people with elevated social
anxiety. Preliminary results by our group [1] also indicated
that subjects with elevated depressive symptoms tend to
self-exclude from phase I participation. Based on these
results, we hypothesized that people reporting elevated trait
anxiety, social anxiety and depressive symptoms would be
less willing to expose themselves to the risks, discomfort and
interpersonal interactions required by participation in a phase
I study. The primary objective of this study was, therefore, to
evaluate the influence of these personality dimensions on the
willingness of subjects to volunteer for participation in phase
I studies. Since there is conceptual overlap between these
personality constructs [2–5], we examined the independent
contribution of each of them on the subject’s willingness to
volunteer in phase I studies.
Differences in the personality characteristics of healthy
volunteers in comparison with those of their counterparts in
the normal population does not necessarily mean that they
are clinically meaningful. There are suggestions, however,
that some personality traits or psychological states may
influence the pharmacokinetics[6–9], pharmacodynamics
[10–12] and probability of presenting clinical complaints
[10]. However, results from systematic research in this field
are still missing, and the impact of the self-selection bias on
study results is largely unknown[7]. The secondary aim of
our study was, therefore, to investigate correlations between
the reported personality traits of a group of participants and
the reporting of adverse events, a common endpoint in
phase I clinical trials.
Actual Participants This group consisted of 198 consecutive normal healthy volunteers who participated in phase I
clinical trials conducted at the Human Pharmacology Unit
of BIAL (S Mamede do Coronado, Portugal). These trials
were pharmacokinetic studies of new oral anti-epileptic and
antiparkinsonian drugs and involved confinement for 3–14
days and frequent blood drawings. Before consenting to
participate, participants received oral and written information on the study design, procedures, inconvenience,
discomfort, precautions, possible adverse events and financial compensation. The financial compensation policy for
participation in the phase I studies was similar in all studies.
Naïve Subjects This group consisted of normal healthy
subjects who never had participated in a clinical trial.
Members of the Naïve Subjects group were selected using
the same method of recruitment as that for phase I
participants to maximize similarities with the Actual
Participants. Naïve Subjects (n=117) were provided with
written information relative to a phase I clinical trial and
invited to participate. That clinical trial implied procedures,
confinement and residence at the clinical research facility
that were similar to those required of participants in the
Actual Participants group; financial compensation was also
similar. Two subgroups were created among those who
responded to the invitation to participate (n=110): “Accepted” (n=51) and “Refused” (n=59).
Procedures and self-report measures
Trait anxiety, social anxiety and depressive symptoms were
measured by Portuguese adaptations of the trait anxiety
portion of the State-Trait Anxiety Inventory (STAI)[13], the
Social Avoidance and Distress (SAD) and Fear of Negative
Evaluation (FNE) scales[14] and the Beck Depression
Inventory II (BDI-II) [15].
Material and methods
Populations
Trait anxiety The STAI is the most widely used selfadministered instrument for measuring anxiety in adults
[16]. It consists of two separate scales: “state anxiety”
(STAI-S), which refers to a transitory emotional state
The subjects of this study comprised a group of actual
participants in phase I drug clinical trials (“Actual Partic-
75
64
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
Eur J Clin Pharmacol (2008) 64:575–582
577
in phase I clinical trials. In accordance with current
regulatory standards, an adverse event was defined as any
undesirable event occurring to a subject during the study,
following the administration of an investigational product,
irregardless of whether or not the event was considered to
be drug-related. The possibility of a causal relationship
between the adverse event and study drug was assessed by
the investigator, following pre-specified rules similar to those
issued by the World Health Organization (www.who-umc.
org). Adverse events were coded according to the Medical
Dictionary for Regulatory Activities (MedDRA). For crossover studies, each study period was considered to be one
“test”. The following variables were derived: percentage of
tests with adverse events; percentage of tests with drugrelated adverse events; number of adverse events/test;
number of drug-related adverse events/test; number of
adverse events/day of test; number of drug-related adverse
events/day of test; number of central nervous system (CNS)related adverse events/test; number of CNS-related adverse
events/day of test.
characterized by subjective feelings of tension that vary in
intensity over time, and “trait anxiety” (STAI-T), which
refers to a relatively stable and long-standing disposition to
respond to stress with elevated anxiety and a tendency to
perceive a wide range of situations as personally threatening
[13]. Each scale consists of 20 statements, and scores vary
from a minimum of 20 to a maximum of 80. We used the
STAI-T sub-scale in our study. High scores reflect greater
trait anxiety. Prior research has shown that the STAI-T has
strong internal reliability and construct validity [17].
Social anxiety Social anxiety is defined as “a marked and
persistent fear of one or more social or performance
situations in which the person is exposed to unfamiliar
people or to possible scrutiny by others” [18]. Commonly
endorsed fears include public speaking, performance,
speaking to strangers and meeting new people. Two scales
were used here to measure the behavioral and cognitive
components of social anxiety. The SAD consists of 28 true–
false items assessing social anxiety and fear-motivated
avoidance derived from potential or actual social interactions [19]. High SAD scores reflect a greater likelihood of
avoiding potentially stressful social interactions. The FNE
consists of 30 true–false items assessing the expectation
and fear of being evaluated negatively by other people and
catastrophic responses to mildly negative social events [20,
21]. Scores can vary from a minimum of 0 to a maximum
of 28 (SAD) and 30 (FNE). Prior research has shown that
both the SAD and FNE possess good psychometric
properties in clinical and non-clinical populations [22].
Application of the SAD and FNE to the Portuguese
population showed results similar to those found in studies
performed in other countries [23]. Hofman [24] recently
found problems in the original scoring instructions of the
SAD; consequently, we used his corrected scoring system
in our study.
Analyses
The first set of analyses compared the results of each scale
between the “Accepted” and “Refused” sub-groups of the
Naïve Subjects group; as internal validation, the Naïve
Subjects “Accepted” subgroup was compared with the
Actual Participants group. Tests for normality were performed using the Shapiro–Wilk test. When the dependent
variable showed a non-normal distribution, the Wilcoxon–
Mann–Whitney rank sum test was used for comparison;
when it showed a normal distribution, t tests were used. For
the BDI-II, the following cutoffs for depressive symptoms
severity were considered, as suggested by the authors of the
BDI-II [25]: 0–13=minimal; 14–19=mild; 20–28=moderate; 29–63=severe. Groups were compared on the distribution of subjects by severity degree with chi-square tests
or Fisher’s exact tests. In addition to these univariate
analyses, the independent contribution of the STAI-T, SAD,
FNE and BDI-II on the prediction of study participation
was tested using multivariate logistic regression. Spearman’s rho correlations were calculated between scales
scores and (1) willingness to participate and (2) variables
of adverse events.
Statistical analyses were performed with the Statistical
Package for Social Sciences ver. 11.5 (SPSS, Chicago, IL).
Depressive symptoms The BDI is one of the most widely
used instruments for assessing depressive symptoms. The
most recent version (BDI-II) [25] corresponds to diagnostic
criteria for major depressive disorder in the fourth edition of
the Diagnostic and Statistical Manual of Mental Disorders
(DSM-IV) [18]. The BDI-II is a 21-item self-report scale
for assessing cognitive, affective and somatic symptoms of
depression during the previous 2 weeks. Each of the 21
items is summed into a single score. Scores can vary from a
minimum of 0 to a maximum of 63. Higher total scores
indicate more severe depressive symptoms. A meta-analysis
[26] suggested evidence of excellent construct validity in
clinical and non-clinical samples.
Ethics
The current study was approved by an Independent Ethics
Committee (Comissão de Ética Independente da UFH, S
Mamede do Coronado, Portugal).
Adverse events Adverse events reported by the Actual
Participants group were recorded during their participation
76
65
CAPÍTULO III
578
Eur J Clin Pharmacol (2008) 64:575–582
Results
showed significantly lower scores on the STAI-T (Z=
−2.600, p<0.01; Wilcoxon–Mann–Whitney test) compared
with Naïve Subjects who denied participation (“Refused”
subgroup). No significant differences were found between
Naïve “Accepted” respondents and the Participants Group
(Z=−0.985, p>0.05). Significant differences were found
when the full Naïve Subjects group was compared with
Actual Participants (Z=−3.718, p<0.001). Using Spearman’s rho, we found a statistically significant inverse relation
between STAI-T scores and the willingness to participate in a
phase I clinical trial (R=−0.203; p<0.05). Taken together,
the results showed that subjects high in trait anxiety tend to
self-exclude from participation in phase I studies.
Table 1 presents the demographic and socio-economic
characteristics of study populations. The Naïve Subject
and Actual Participant groups were similar in terms of
relevant demographic and socio-economic characteristics
such as age, gender, ethnic origin, monthly income and
education. However, a few differences were apparent when
the Naïve Subjects Accepted and Refused groups were
compared: subjects who accepted the invitation to participate in a clinical trial showed a higher preference for
exercising/sports than those who refused, and, surprisingly,
the rate of refusal to participate was higher among
unemployed than among employed subjects.
Social anxiety Statistically significant lower SAD scores
(Z=−2.524, p<0.05) were found in the “Accepted” subgroup
compared with the “Refused” subgroup of the Naïve Subjects.
Significantly higher SAD scores were found in the “Accepted” subgroup of Naïve Subjects compared with the Actual
Participants group (Z=−2.435, p<0.05) and when the entire
Naïve Subjects group was compared with Actual Participants
(Z=−5.038, p<0.001). Using Spearman’s rho, we found an
STAI-T, SAD, FNE and BDI-II univariate analysis
Figure 1 shows the median, quartiles and range scores on
the STAI-T, SAD, FNE and BDI-II scales.
Trait anxiety Naïve Subjects expressing a willingness to
participate in a phase I clinical trial (“Accepted” subgroup)
Table 1 Main demographic
and socio-economic characteristics of the study cohorts
Parameter
Variable
Gender (%)
Age (years)
Ethnic origin (%)
Occupation (%)
Monthly net income in relation
to the mean national net salary
in the industry and services
sectora (%)
Civil status (%)
School degree completed (%)
All values for all parameters,
with the exception of “Age”,
are given as a percentage of
that study cohort
a
Approximately €1000/month,
net
Smoking (%)
Coffee drinking (%)
Exercising/sports (%)
Alcohol consumption (%)
Male/female
Mean ± SD
Median
Range
Caucasian/other
Student
Employed
Unemployed
<25%
25-50%
51-100%
101-150%
>150%
Single
Married/living together
Divorced
4 years
6 years
9 years
12 years
Bachelor
Licensed
Masters
≥1 cigarette/day
≥1 cup/day
≥1 session/week
≥1 occasion/week
77
66
Actual Participants
(n=198)
Naïve Subjects
Accepted
(n=51)
Refused
(n=59)
47.5/52.5
26.4±5.5
25
18–45
95.5/4.5
55.7
38.9
5.4
25.2
27.7
31.4
10.7
5.0
76.2
17.8
5.9
0.0
1.0
7.0
63.8
7.5
17.6
3.0
34.9
67.6
62.5
49.5
47.1/52.9
26.3±6.6
25
18–45
92.2/7.8
56.0
38.0
6.0
30.0
35.0
25.0
10.0
0.0
72.5
23.5
3.9
2.0
3.9
9.8
60.8
5.9
15.7
2.0
21.6
58.8
72.5
52.9
42.4/57.6
26.6±7.3
24
18–45
98.3/1.7
55.9
32.2
11.9
28.6
24.5
28.6
12.2
4.1
67.2
27.6
5.2
6.8
3.4
5.1
55.9
6.8
20.3
1.7
27.1
66.1
47.5
44.1
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
Eur J Clin Pharmacol (2008) 64:575–582
40
30
50
25
ep
lP
A
ct
ua
aï
N
aï
ve
ve
lP
ua
ct
aï
te
d
se
ar
tic
-R
ef
u
ip
cc
ep
te
d
-A
ve
N
aï
ve
N
A
63
p<0.05
n.s.
tic
0
an
ts
5
0
-R
ef
us
ed
10
10
ip
an
ts
15
20
40
56
35
49
30
n.s.
n.s.
BDI-II Score
25
20
15
10
42
35
28
21
nt
ci
ct
ua
lP
ar
ti
ve
aï
N
aï
N
pa
te
ep
cc
-A
lP
ua
ct
A
d
ed
ar
-R
tic
ef
ip
cc
ep
-A
ve
aï
N
us
d
te
ed
us
ef
-R
ve
aï
N
inverse relation between the SAD score and willingness to
participate in a phase I clinical trial (R= −0.204, p<0.05).
Taken together, the results indicate that people to whom
social interactions cause distress tend to self-exclude from
participation in phase I studies.
Significantly higher FNE scores were found in the
“Accepted” subgroup of the Naïve Subjects than among
the Actual Participants (Z=−2.439, p<0.05) or when the
whole Naïve Subjects group was compared with the Actual
Participants (Z=−4.298, p<0.001). Using Spearman’s rho,
we found no significant correlation between FNE scores
and willingness to participate in a phase I clinical trial (R=
−0.108, p>0.05). Overall, results suggest that the fear of
negative evaluation does not substantially contribute to selfexclusion from study participation.
A
0
an
ts
7
0
s
14
5
ve
FNE Score
20
cc
30
-A
40
p<0.05
p<0.05
ar
SAD Score
35
60
d
n.s.
p<0.01
70
N
80
STAI-T Score
Fig. 1 STAI-T, SAD, FNE and
BDI-II scores in the subsets of
the Naïve Subjects who “Refused” and who “Accepted” the
invitation to participate in a
phase I clinical trial, and in the
Actual Participants group. Boxand-whiskers plot showing median, quartiles and range values.
For a description of the selfreport measures, see the section
Procedures and self-report
measures
579
p>0.05) or between the “Accepted” subgroup of the Naïve
Subjects and Actual Participants. Using Spearman’s rho, we
failed to find a significant relation between BDI-II scores
and willingness to participate (R=−0.067, p>0.05). However, groups differed in the distribution of subjects into
depressive severity categories: in Actual Participants, 5.1,
1.9 and 0.0% of the subjects presented mild, moderate and
severe symptoms, respectively; in Naïve Subjects, the results
were 7.8, 6.9 and 2.6%, respectively (chi-square=2.69, df =
3, p<0.01). Overall, the results suggest that subjects
presenting moderate or severe depressive symptoms tend to
self-exclude from participation in phase I studies.
STAI-T, SAD, FNE and BDI-II multivariate analysis
on the prediction of study participation
Depressive symptoms No statistically significant differences were found in BDI-II scores between the “Accepted”
and “Refused” subsets of the Naïve Subjects (Z=−1.006,
To explore the unique contributions of trait anxiety, social
avoidance and distress, fear of negative evaluation and
78
67
CAPÍTULO III
580
Eur J Clin Pharmacol (2008) 64:575–582
Table 2 Summary of logistic regression analysis of STAI-T, SAD, FNE and BDI-II scores as predictors of participation in a phase I clinical trial
Self-report measuresa
β
Standard error
Exp(B)
95%CI for Exp(B)
STAI-T
SAD
FNE
BDI-II
Constant
-0.072
-0.096
-0.030
0.043
4.956
0.029
0.031
0.029
0.032
0.943
0.930
0.908
0.970
1.044
142.094
0.88,
0.86,
0.92,
0.98,
0.99
0.97
1.03
1.11
Significance
p<0.05
p<0.01
n.s.
n.s.
p<0.001
β , Regression coefficient; Exp(B), Odds ratios; CI, confidence interval; n.s., not statistically significant
a
STAI, State-Trait Anxiety Inventory; SAD, the Social Avoidance and Distress scale ; FNE, Fear of Negative Evaluation (FNE) scale; BDI-II,
Beck Depression Inventory II
depressive symptoms in predicting participation, we developed a logistic regression model. The results of the model
are presented in Table 2. The only statistically significant
predictors of participation were the STAI-T (p<0.05) and
SAD (p<0.01) scores.
with lower trait anxiety scores reported fewer adverse
events.
Correlation between STAI-T, SAD, FNE and BDI-II
and the reporting of adverse events
Participation in phase I clinical trials is based on free
informed consent. Therefore, participants are a self-selected
population of volunteers, and it is reasonable to assume that
factors affecting the willingness to volunteer induce a selfselection bias that has a potential impact on study outcomes. The results of the current study show that more
anxious, socially avoidant and depressed people are less
Discussion
Information regarding adverse events was available for 470
study periods (“tests”). Mean number of days per test was
10.0±9.0 (range1–47). Adverse events (a total of 382) were
reported in 57.2% of tests. Most adverse events were
nervous system disorders or gastrointestinal disorders,
according to the System Organ Class (SOC) classification
of the MedDRA dictionary. Table 3 presents the most
frequently reported adverse events, classified according to
the MedDRA Lowest Level Term. The mean number of
adverse events per test was 0.88±1.3 (range 0–7), of drugrelated adverse events per test, 0.64±1.2 (range 0–7), of
adverse events per day of test, 0.18±0.22 (range 0.0–1.0),
of drug-related adverse events per day of test, 0.12±0.18
(range 0.0–0.9), of CNS-related adverse events per test,
0.42±0.82 (range0.0–5.0) and of CNS-related adverse
events per day of test, 0.08±0.14 (range 0.0–0.8).
Post-hoc evaluation showed a power of approximately
70% for the correlations between scales’ scores and
adverse event variables. No significant correlations were
found between the BDI-II, SAD and FNE and adverse
events. Significant positive correlations were found
between STAI-T scores and each adverse event variable:
percentage of tests with adverse events (R=0.163, p<
0.05), percentage of tests with drug-related adverse events
(R=154, p<0.05), number of adverse events per test (R=
0.193, p<0.05), number of drug-related adverse events per
test (R=0.158, p<0.01), number of adverse events per day
of test (R=0.188, p<0.01), number of drug-related adverse
events per day of test (R=0.166, p<0.05), number of CNSrelated adverse events per test (R=0.175, p<0.05) and
number of CNS-related adverse events per day of test (R=
0.173, p<0.01). Overall, the results suggest that subjects
Table 3 Adverse events coded by the MedDRA Lowest Level Term
(LLT) with a frequency of at least 1.0%
79
68
Adverse event LLT
Frequency (%)
Dizziness
Somnolence
Nausea
Headache
Dysmenorrhoea
Pharyngitis
Myalgia of lower extremities
Increased levels of creatine phosphokinase
Rhinitis
Tremor of hands
Vomiting
Epistaxis
Frontal headache
Rhinorrhoea
Toothache
Ecchymosis
Heartburn
Lumbar pain
Catheter site phlebitis
Diarrhoea
Loose stools
Nasopharyngitis
Paraesthesia tongue
Sore throat
10.0
6.8
4.2
3.9
2.6
2.6
2.1
2.1
2.1
1.8
1.8
1.8
1.6
1.6
1.6
1.3
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
Eur J Clin Pharmacol (2008) 64:575–582
581
together with strangers in a ward with minimal privacy
for several days. Therefore, it is reasonable to expect
subjects with higher social avoidance to self-exclude from
participation in such studies.
Although it is accepted that phase I study participants do
not necessarily represent the general population, it is
assumed that they should represent at least their age group
[30]. However, our study corroborates data suggesting that
participants in phase I clinical trials may differ from the
population from which they are drawn [30]. That said, the
self-selection bias will be important only if it interferes with
study outcomes or conclusions. Overall, our results suggest
that subjects lower in trait anxiety report fewer adverse
events. Since study participants represent a self-selected
population of less anxious subjects than the full population
[31], it may be concluded that: (1) globally, the selfselection may cause a decrease in the likelihood of
reporting adverse events during phase I clinical trials; (2)
in case of parallel-group clinical trials in which tolerability
is the main endpoint, between-group imbalances in trait
anxiety may lead to biased results. As a prevention strategy,
the assessment of the individual’s anxiety level could be
included in the volunteer’s screening procedures and
adequate stratification could be carried out during subject
assignment to different study groups.
In conclusion, our study found that less anxious and less
socially avoidant subjects are self-selected for participation
in phase I clinical trials. The impact of these self-selection
biases on the phase I clinical trial results has not been
completely characterized, but there is a strong suggestion
that subjects low in anxiety tend to report fewer adverse
events. Therefore, the characterization of participant’s
anxiety levels may be important in phase I studies.
likely to volunteer for participation in phase I clinical trials.
The results also suggest that phase I study participants with
lower trait anxiety scores report relatively fewer adverse
events when participating in phase I studies.
The Naïve Subject and Actual Participants groups
showed similar demographic and socio-economic characteristics, and the Actual Participants group is representative of
our historical database of participants in phase I studies
[27]. Anxiety-trait, social anxiety and depressive symptoms scores in the Naïve Subjects group are relatively
similar to those of healthy samples with similar demographic characteristics [15, 23, 29]. However, since reports
of phase I studies usually do not include information on the
socio-economic status of study participants [28], it is
unclear whether our sample is representative of subject
populations volunteering for phase I studies in other
clinical pharmacology units. The median STAI-T score
was 35 and 42 in the “Accepted” and “Refused” subgroups
of the Naïve Subjects, respectively; this difference is
statistically significant (p<0.01) and clinically meaningful
[17]. In terms of social anxiety (SAD scores), the median
score in the “Refused” subgroup was twofold higher than
that in the “Accepted” subgroup (8 versus 4, respectively),
which is also statistically significant (p<0.05) and clinically meaningful [14].
It is noteworthy that the scores for all scales were
slightly higher for the Naïve Subjects “Accepted” subgroup
than for the Actual Participants group. This may be due to
the fact that Naïve Subjects “Accepted” subjects only
expressed an “intention” regarding participation, and it
may be speculated that at the decision-making moment,
some subjects (most likely among those with high scores)
would reverse their consent to participate.
There is conceptual overlap among the personality traits
under study. Upon controlling for shared variance among
the STAI-T, SAD, FNE and BDI-II, we found that the only
unique predictors of participation were trait anxiety and
social avoidance. These traits appear to be of particular
importance in terms of understanding the willingness (and
reluctance) to volunteer for phase I clinical trials. Participation in phase I clinical trials with drugs in clinical
development involves discomfort and the risk of experiencing expected and unexpected adverse events. According
to current good clinical practice and regulatory requirements, volunteers must be made aware of all such
inconveniences and risks before consenting to participate.
Since subjects with greater anxiety are more likely to
perceive a wider range of situations as threatening[13], it
seems reasonable that they are more likely to view a phase I
clinical trial as an unappealing and stressful situation.
Participation in a phase I clinical trial also requires
relationships with unknown people and the ability to
embrace socially challenging situations, such as living
Acknowledgements We thank the healthy subjects and the clinical
staff of the clinical pharmacology unit of BIAL – Portela & Co, SA
(especially Prof. Manuel Vaz-da-Silva, and Benedita Azevedo, BSc
Psychol) for their cooperation. We also thank Prof. Amílcar Falcão
and Dr. Grazina Binkauskiene for their assistance with the statistical
analysis.
References
1. Almeida L, Azevedo B, Vaz-da-Silva M, Nunes T, Falcao A,
Coelho R, Albino-Teixeira A, Soares-da-Silva P (2006) Presence
of depressive symptoms in a group of subjects who accepted to
participate in Phase I studies in comparison with a control group
(abstract 1067). J Clin Pharmacol 46:1076
2. Gorenstein C, Andrade L (1996) Validation of a Portuguese
version of the Beck Depression Inventory and the State-Trait
Anxiety Inventory in Brazilian subjects. Braz J Med Biol Res
34:367–374
3. Kashdan TB, Roberts JE (2004) Social anxiety's impact on affect,
curiosity, and social self-efficacy during a high self-focus social
threat situation. Cogn Ther Res 28:119–141
80
69
CAPÍTULO III
582
Eur J Clin Pharmacol (2008) 64:575–582
19. Herbert JD, Rheingold AA, Brandsma L (2001) Assessment of
social phobia and social anxiety. In: Hoffmann SG, DiBartolo PM,
Juster HR (eds) Social anxiety and social phobia. Allen & Bacon,
Needham, p 20
20. Monfries MM, Kafer NF (1994) Private self-consciousness and
fear of negative evaluation. J Psychol 128:447–454
21. Stopa L, Clark DM (2001) Social phobia: Comments on the
viability and validity of an analogue research strategy and British
norms for the fear of negative evaluation questionnaire. Behav
Cogn Psychother 29:423–430
22. Trower P, Sherling G, Beech J, Harrop C, Gilbert P (1998)
The socially anxious perspective in face-to-face interaction:
An experimental comparison. Clin Psychol Psychother 5:155–
166
23. Pinto-Gouveia J, Fonseca L, Robalo M, Allen A, Matos A, Gil E
(1986) Ansiedade social: Utilização dos questionários de autoresposta SAD, FNE e SISST numa população portuguesa.
Psiquiatr Clín 7:43–48
24. Hofmann SG, DiBartolo PM, Holaway RM, Heimberg RG (2004)
Scoring error of social avoidance and distress scale and its
psychometric implications. Depress Anxiety 19:197–198
25. Beck AT, Steer RA, Brown GK (1996) BDI-II: Beck Depression
Inventory, 2nd edn. The Psychological Corporation, San Antonio
26. Beck AT, Steer RA, Garbin MG (1988) Psychometric properties
of the Beck depression inventory: twenty-five years of evaluation.
Clin Psychol Rev 8:77–100
27. Almeida L, Azevedo B, Nunes T, Vaz-da-Silva M, Soares-daSilva P (2007) Why healthy subjects volunteer for phase I studies
and how they perceive their participation? Eur J Clin Pharmacol
63:1085–1094
28. Tishler CL, Bartholomae S (2002) The recruitment of normal
healthy volunteers: A review of the literature on the use of
financial incentives. J Clin Pharmacol 42:365–375
29. Andrade L, Gorenstein C, Vieira-Filho AH, Tung TC, Artes R
(2001) Psychometric properties of the Portuguese version of the
State-Trait Anxiety Inventory applied to college students: factor
analysis and relation to the Beck Depression Inventory. Braz J
Med Biol Res 34:367–374
30. Pieters MS, Jennekens-Schinkel A, Schoemaker HC, Cohen AF
(1992) Self-selection for personality variables among healthy
volunteers. Br J Clin Pharmacol 33:101–106
31. Van Gelderen CEM, Savelkoul TJF, Van Dokkum W, Meulenbelt
J (1993) Motives and perception of healthy volunteers who
participate in experiments. Eur J Clin Pharmacol 45:15–21
4. Kashdan TB, Roberts JE (2007) Social anxiety, depressive
symptoms, and post-event rumination: affective consequences
and social contextual influences. J Anxiety Disord 21:284–301
5. Kowacs F, Socal MP, Ziomkowski SC, Borges-Neto VF, Toniolo
DP, Francesconi CR, Chaves ML (2003) Symptoms of depression
and anxiety, and screening for mental disorders in migrainous
patients. Cephalalgia 23:79–89
6. Nakano S, Ogawa N, Kawazu Y (1980) Influence of neuroticism
on oral absorption of diazepam. Clin Pharmacol Ther 27:370–374
7. Tishler CL, Bartholomae S, Rhodes AR (2005) Personality
profiles of normal healthy research volunteers: a potential concern
for clinical drug trial investigators? Med Hypotheses 65:1–7
8. Vessel ES (1982) On the significance of host factors that affect
drug disposition. Clin Pharmacol Ther 31:1–7
9. Meyer FP (1992) Aspekte einer differentiellen Human-psychopharmakologie. In: Oldigs-Kerber J, Leonard JP (eds) Pharmakopsychologie. Experimentelle und klinische Aspekte. Fischer,
Jena, pp 187–200
10. Meyer FP (2001) Psycho- and immunopharmacological factors
relevant to selection of volunteers in clinical studies. Int J Clin
Pharmacol Ther 39:300–310
11. Mizuki Y, Suetsugi M, Ushijima I, Yamada M (1997) Differential
effects of dopaminergic drugs on anxiety and arousal in healthy
volunteers with high and low anxiety. Prog Neuropsychopharmacol Biol Psychiatry 21:573–590
12. Mizuki Y, Suetsugi M, Ushijima I, Yamada M (1996) Differential
effects of noradrenergic drugs on anxiety and arousal in healthy
volunteers with high and low anxiety. Prog Neuropsychopharmacol Biol Psychiatry 20:1353–1367
13. Spielberger CD, Gorsuch RL, Lushene RE (1970) STAI manual
for the state-trait anxiety inventory. Consulting Psychologists
Press, Palo Alto
14. Watson D, Friend R (1969) Measurement of social-evaluative
anxiety. J Consult Clin Psychol 33:448–457
15. Coelho R, Martins A, Barros H (2002) Clinical profiles relating
gender and depressive symptoms among adolescents ascertained
by the Beck Depression Inventory II. Eur Psychiatry 17:222–226
16. Keedwell P, Snaith RP (1996) What do anxiety scales measure?
Acta Psychiatr Scand 93:177–180
17. Spielberger CD, Gorsuch RL, Lushene RE (1983) Stait-Trait Anxiety
Inventory (Form Y). Consulting Psychologists Press, Palo Alto
18. American Psychiatric Association (1994) DSM-IV. Diagnostic
and statistical manual of mental disorders, 4th edn. American
Psychiatric Association, Washington D.C.
81
70
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
71
CAPÍTULO III
72
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
73
CAPÍTULO III
74
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
75
CAPÍTULO III
76
Os traços de personalidade como filtro de auto-selecção de participantes em ensaios
clínicos de fase I e sua influência nas avaliações de tolerabilidade
77
CAPÍTULO III
78
CAPÍTULO IV
Prevalência de queixas adversas de origem não medicamentosa e sua correlação
com as características de personalidade
«Adverse non-drug related complaints by healthy volunteers in phase I studies
compared to healthy general population»
Int J Clin Pharmacol Ther 46, 2008 (aprovado para publicação; em impressão)
Prevalência de queixas adversas de origem não medicamentosa e sua correlação com
as características de personalidade
International Journal of Clinical Pharmacology and Therapeutics, Vol. 46 – No. /2008 (-)
©2008 Dustri-Verlag Dr. K. Feistle
ISSN 0946-1965
Original
Adverse nondrug-related complaints by
healthy volunteers in Phase I studies
compared to healthy general population
L. Almeida1,2, R. Coelho3, A. Albino-Teixeira2,4 and P. Soares-da-Silva1,2
1Department
of Research and Development, BIAL-Portela, SA,
S. Mamede do Coronado, 2Institute of Pharmacology and Therapeutics,
Faculty of Medicine, 3Service of Psychiatry and Mental Health, Faculty of Medicine,
4Institute for Molecular and Cell Biology (IBMC), University of Oporto, Porto, Portugal
Key words
Phase I studies – healthy
subjects – personality –
nondrug-related adverse
complaints – adverse
events
Received
March 31, 2008;
accepted in revised form
June 30, 2008
Correspondence to
Dr. L. Almeida
Department of Research
and Development
BIAL, A Av. da
Siderurgia Nacional,
4745-457, S. Mamede
do Coronado, Portugal
[email protected]
Adverse nondrug-related complaints
Abstract. Objective: Some complaints
that are reported as adverse drug reactions by
healthy subjects during participation in Phase
I studies are common complaints in healthy
individuals from the normal population. The
objective of this study was to compare the incidence of complaints in a group of 192
healthy volunteers in Phase I studies with a
control group of 112 healthy subjects who
matched the Phase I group participants in
terms of demographic and socioeconomic
characteristics, and to investigate the relationship between some psychological factors
and the incidence of complaints. Methods:
Both groups completed a questionnaire on the
incidence of complaints during the previous 2
– 4 weeks. Trait anxiety was assessed by the
trait scale of the State-Trait Anxiety Inventory (STAI-T), depressive mood by the
Beck’s Depression Inventory-II (BDI-II) and
perceived self-efficacy by the Self-Efficacy
Scale (SES). Results: Compared to the control group, Phase I volunteers presented a significantly lower incidence of stomach pain,
back pain, limb or joint pain, headaches, fainting spells, palpitations, shortness of breath,
constipation, loose stools or diarrhea, nausea,
gas or indigestion, feeling nervous or anxious, feeling restless, getting tired very easily,
muscle tension, aches, or soreness, and concentration difficulties. Significant positive
correlations were found between the STAI-T
and BDI-II scores and the incidence of several complaints; inversely, the SES score correlated negatively with several complaints.
Conclusion: The incidence of complaints in
healthy subjects is not of a random character
and depends on psychological characteristics. Volunteers in Phase I studies are a
self-selected sample with a lower tendency to
report nondrug-related adverse events than
their peers from the general population. The
impact of this self-selection bias on the assessment of tolerability during Phase I studies
deserves further evaluation.
81
Introduction
One of the primary objectives of Phase I
studies in healthy subjects is to investigate the
safety and tolerability of new drugs before
proceeding with studies in the target population. How ever, the fact that the general
“healthy” population has a high incidence of
complaints that are also typical side effects of
drugs (e.g. fatigue, headache, flu-like symptoms, abdominal pain, etc.) is an essential but
overlooked aspect that may complicate the
evaluation of adverse events in Phase I studies. In 1968, Reidenberg and Lowenthal
[1968] reported that only 16% of subjects
from a sample of 293 healthy medical students and university staff presented no complaints in the preceding 3 days, after being
asked to complete a questionnaire based on
side effects of drugs. In 1996, Meyer et al.
[1996] repeated the experience on 130 healthy
medical students using a similar questionnaire
and found that only 11% of the participants
were free of any complaint. Furthermore, these
authors found that subjects who presented less
complaints were less nervous, more emotionally stable, more highly motivated and less
anxious than subjects who presented more
complaints [Meyer et al. 1996].
In previous studies, we have shown that
subjects who volunteer for Phase I studies
tend to have lower levels of anxiety, social
anxiety and depressive mood [Almeida et al.
2008], and higher levels of curiosity, exploratory tendencies and perceived self-efficacy
[Almeida et al. 2008] than their peers in the
general population. Consequently, because
participation is based on informed consent,
CAPÍTULO IV
2
Almeida, Coelho, Albino-Teixeira and Soares-da-Silva
participants in Phase I studies are a self-selected population of subjects with personality
characteristics not representative of the population as a whole. Admitting an influence of
personality characteristics on the incidence
of nondrug-related symptoms, as shown by
Meyer et al. [1996], it may be hypothesized
that those personality-related self-selection
biases may lead to a self-selected population
of participants in Phase I studies with lower
levels of complaints than their peers in the
general population. To contribute to the clarification of this issue, we conducted the current study aiming:
– to compare the incidence of complaints
between a group of healthy subjects who
volunteered for Phase I studies and a control group of healthy subjects who had
never participated in a clinical trial,
– to investigate the relationship between
trait anxiety, depressive mood and perceived self-efficacy and the incidence of
nondrug-related complaints.
Material and methods
Populations
The study population comprised of a group
of healthy subjects who volunteered for participation in Phase I drug studies conducted at
the BIAL’s Human Pharmacology Unit, S.
Mamede do Coronado, Portugal (Phase I
volunteers, n = 192) and a control group of
healthy subjects (n = 112) who were self-reported as healthy and were not taking medicines on a regular basis other than oral contraceptives or vitamins. The initial control group
comprised of 120 subjects, but 4 subjects
were excluded from the analysis because they
declared they had previously participated in
Phase I studies and further 4 subjects were excluded because they were found to be regularly taking an antidepressant (n = 2), a
benzodiazepine (n = 1) and an unidentified
thyroid disorder drug (n = 1).
In both groups, subjects were invited to
anonymously and confidentially answer a set
of questionnaires aiming to characterize their
demographic, socioeconomic and psychological characteristics, without drug administration. In the Phase I volunteers group, the
questionnaires were completed at the time of
82
volunteering for Phase I study participation.
In order to get a control group reasonably
matching the demographic and socioeconomic characteristics of the Phase I participants,
the control group was selected mimicking the
usual recruitment procedure for Phase I studies, which is described elsewhere [Almeida et
al. 2007]. Most participants in Phase I studies
conducted in our clinical pharmacology unit
are university students, or recent graduates
who are starting their professional career
[Almeida et al. 2007]. For constituting the
control group, subjects were contacted in the
universities and work places from which
participants usually come and were asked to
participate in this investigation. It was emphasized that this investigation did not constitute a clinical trial, only consisting of filling
out a set of questionnaires, in an anonymous
and confidential way. Neither Phase I volunteers nor control group subjects received any
financial compensation for their participation
in this specific study. The study was approved
by an Independent Ethics Committee (Comissao de Ética Independente da UFH, Porto,
Portugal).
Self-report measures
Subjects complaints in the recent
past
The assessment of recent complaints was
performed through a self-report questionnaire (Table 1) derived from some sections of
the Patient Health Questionnaire (PHQ),
which is a self-report version of the PRIMEMD Patient Problem Questionnaire (PPQ)
[Spitzer et al. 1999]. The questionnaire was
answered in an anonymous and confidential
way. Subjects were asked how much they
have been bothered during the last 4 weeks by
any of a list of somatic complaints that are often found as adverse events in Phase I studies
(Questions #1). Questions #2 asked about the
frequency in the previous 2 weeks of complaints mainly related with depressive mood,
and Questions #3 asked about the frequency
of anxiety-related complaints in the previous
4 weeks. Questions #4 asked how much subjects have been bothered during the last 4
weeks about personal circumstances, such as
negative life-events or interpersonal, social or
Prevalência de queixas adversas de origem não medicamentosa e sua correlação com
as características de personalidade
3
Adverse nondrug-related complaints
Table 1.
Self-reported complaints questionnaire.
1. During the last 4 weeks, how much have you been bothered by any of the following problems?
a. Stomach pain
b. Back pain
c. Pain in your arms, legs, or joints (knees, hips, etc.)
d. Menstrual cramps or other problems with your periods*
e. Pain or problems during sexual intercourse*
f. Headaches
g. Chest pain
h. Dizziness
i. Fainting spells
j. Feeling your heart pound or race
k. Shortness of breath
l. Constipation, loose bowels, or diarrhoea
m. Nausea, gas, or indigestion
2. Over the last 2 weeks, how have you been bothered by any of the following problems?
a. Little interest or pleasure in doing things
b. Felling down, depressed, or hopeless
c. Trouble falling or staying asleep, or sleeping too much
d. Feeling tired or having little energy
e. Poor appetite or overeating
f. Feeling bad about yourself, or that you are a failure, or have let yourself or your family down
g. Trouble concentrating on things, such as reading the newspaper or watching television
h. Moving or speaking so slowly that other people could have noticed? Or the opposite – being so
fidgety or restless that you have been moving around a lot more than usual
i. Thoughts that you would be better off dead or of hurting yourself in some way
3. Over the last 4 weeks, how often have you been bothered by any of the following problems?
a. Feeling nervous, anxious, on edge, or worrying a lot about different things
b. Feeling restless so that it is hard to sit still
c. Getting tired very easily
d. Muscle tension, aches, or soreness
e. Trouble falling asleep or staying asleep
f. Trouble concentrating on things, such as reading a book or watching TV
g. Becoming easily annoyed or irritable
4. In the last 4 weeks, how much have you been bothered by any of the following problems?
a. Worrying about your health
b. Your weight or how you look
c. Little or no sexual desire or pleasure during sex
d. Difficulties with husband/wife, partner/lover or boyfriend/girlfriend
e. The stress of taking care of children, parents, or other family members
f. Stress at work outside of the home or at school
g. Financials problems or worries
h. Having no one to turn to when you have a problem
i. Something bad that happened recently
j. Thinking or dreaming about something terrible that happened to you in the past – like your house
being destroyed, a severe accident, being hit or assaulted, or being forced to commit a sexual act
*Only in females
83
CAPÍTULO IV
4
Almeida, Coelho, Albino-Teixeira and Soares-da-Silva
Table 2. Main demographic and socio-economic characteristics of the study groups.
Characteristic
Gender
Age
Volunteers group
(n = 192)
Control group
(n = 112)
Male
46.4%
44.6%
Female
53.6%
55.4%
26.3 ± 5.4
25.9 ± 6.4
Mean ± SD (years)
Median (years)
25
24
Range (years)
18 – 45
18 – 45
Caucasian
95.8%
95.5%
Other
4.2%
4.5%
University student
56.0%
54.4%
Employed
38.2%
34.8%
Unemployed
5.8%
8.0%
Missing
0.0%
1.8%
Monthly net income in relation
< 25%
26.7%
28.9%
Ethnic origin
Occupation
to the mean national net salary
25%-50%
27.9%
28.9%
in the industry and services
51% – 100%
30.3%
26.7%
sector#
101% – 150%
10.3%
11.8%
> 150%
4.8%
3.3%
Single
77.0%
72.3%
Married/living together
17.3%
24.3%
Divorced
5.8%
3.6%
Missing
0.0%
0.9%
4 years
0.0%
2.7%
3.6%
Civil status
School degree completed
6 years
1.0%
9 years
7.3%
8.0%
12 years
62.8%
58.9%
Bachelor
7.3%
6.3%
University graduation
18.3%
18.8%
Masters
3.1%
1.8%
Smoking
 1 cigarette/day
35.4%
24.1%
None
64.6%
75.9%
Exercising/sports
 1 session/week
61.5%
60.7%
None
37.5%
39.3%
Missing
1.0%
0.0%
 1 drink/week
49.0%
50.0%
None
51.0%
50.0%
None
66.7%
59.8%
Irregularly
17.7%
18.8%
Daily
15.6%
17.0%
Missing
0.0%
4.5%
Alcohol drinking
Use of medicines
#approximately
€1,000/month, net.
84
Prevalência de queixas adversas de origem não medicamentosa e sua correlação com
as características de personalidade
5
Adverse nondrug-related complaints
financial difficulties. In Phase I volunteers,
the questionnaire was answered at the time of
screening procedures for Phase I study participation, after getting the subject’s written informed consent, but the questionnaire was not
considered to be part of the screening.
Trait-anxiety, depressive mood and
perceived self-efficacy
Trait anxiety, depressive mood and perceived self-efficacy were measured by Portuguese adaptations of self-report inventories.
The trait anxiety was assessed by the trait
anxiety section of the Spielberger’s StateTrait Anxiety Scale (STAI-T) [Spielberger et
al. 1970]. The trait anxiety refers to a relatively stable and long-standing disposition to
respond to stress with anxiety and a tendency
to perceive a wider range of situations as
threatening [Spielberger et al. 1970]. STAI-T
scores vary from a minimum of 20 to a maximum of 80 and they have a direct interpretation: high scores mean more trait anxiety and
low scores indicate less. Depressive mood
was assessed by the Beck Depression Inventory-II (BDI-II) [Beck et al. 1996]. BDI-II is a
21-item self-report scale for assessing cognitive, affective, and somatic symptoms of depression during the previous 2 weeks. Scores
can vary from a minimum of 0 to a maximum
of 63. Higher total scores indicate more severe depressive symptoms [Coelho et al.
2002]. Perceived self-efficacy is defined as
beliefs about one’s capacity to self-generate
behaviors to obtain desired outcomes even
when confronted with barriers and obstacles
[Kashdan and Roberts 2004], i.e., it refers to
one’s impression of what one is capable of doing. The perceived self-efficacy was assessed
through the Self-Efficacy Scale (SES)
[Sherer et al. 1982]. It is composed of three
subscales: Initiative/Persistence, Planning/
Goal Setting, and Social Self-Efficacy. In our
study, we used a 15-item Portuguese validated
adaptation [Ribeiro 1995] of the original SES
scale [Sherer et al. 1982]. In Phase I volunteers, the scales were applied at the time of
their admission to Phase I study participation.
version 11.5 (SPSS Inc, Chicago, Ill). The criterion for statistical significance was set at an
 error of 5% (p < 0.05). Each self reported
complaint was compared between the Phase I
volunteers and the control group using the
Pearson’s 2 test. STAI-T, BDI-II and SES
scores were not normally distributed; consequently, the Mann-Whitney rank sum test was
used for between-group comparisons. Correlations between the STAI-T, BDI-II and SES
scores and the self-reported complaints were
investigated using Spearman’s rho.
Results
Groups did not significantly differ in terms
of demographic, socioeconomic and life-style
characteristics (Table 2). The proportion of
patients using medicines was 33.3% in the
Phase I volunteers group and 35.8% in the
control group. Approximately half of the subjects were using medicine on a daily basis
(mostly oral contraceptives by females, in
both groups). The remaining took medicine
sporadically/irregularly, mostly for the alleviation of menstrual pain, occasional headache or
rhinitis.
Incidence of complaints
Table 3 presents the statistics of the selfreported complaints in both groups. For
Questions #1, #2 and #3, Phase I volunteers
showed significantly lower levels of complaints than the control group for all questions
asked except 2.f (“Feeling bad about yourself, or that you are a failure, or have let yourself or your family down”) and 3.e (trouble
falling asleep or staying asleep). For Questions #4, the two groups were similar for all
but three questions: Phase I volunteers were
significantly less worried about their health
than the control subjects (4.a), had less difficulties with their romantic partner (4.d) and
less stress caused by their interaction with
their relatives (4.e).
Analyses
Scales scores in the Phase I
volunteers and control groups
Statistical analyses were performed with
the Statistical Package for Social Sciences
Figure 1 displays box-and-whiskers
plots of the STAI-T, BDI-II and SES Total
85
CAPÍTULO IV
6
Almeida, Coelho, Albino-Teixeira and Soares-da-Silva
Table 3.
Complaints comparison between the Phase I volunteers group and the control group (Pearson’s Chi-square test). Part I.
Questions
#1
Phase I Volunteers (n = 192)
2 test
Control Group (n = 112)
How much bothered
How much bothered
A lot
p
value
Not
A little
A lot
Not
A little
a
86.5%
10.9%
2.6%
69.6%
25.9%
4.5%
< 0.01
b
58.6%
34.0%
7.3%
48.2%
33.9%
17.9%
< 0.01
< 0.01
c
71.4%
26.6%
2.1%
60.7%
27.7%
11.6%
d*
61.2%
27.2%
11.7%
54.8%
32.3%
12.9%
n.s.
e*
89.3%
7.8%
2.9%
85.5%
14.5%
0.0%
n.s.
< 0.05
f
55.2%
39.1%
5.7%
45.5%
42.0%
12.5%
g
93.2%
5.2%
1.6%
87.5%
8.9%
3.6%
n.s.
h
89.6%
8.9%
1.6%
84.8%
11.6%
3.6%
n.s.
i
96.4%
1.6%
2.1%
89.3%
8.9%
1.8%
< 0.05
j
84.4%
13.0%
2.6%
71.4%
24.1%
4.5%
< 0.05
k
92.2%
5.2%
2.6%
81.3%
16.1%
2.7%
< 0.01
l
80.7%
15.6%
3.6%
68.8%
22.3%
8.9%
< 0.05
m
75.0%
24.0%
1.0%
68.8%
24.1%
7.1%
< 0.05
Questions
#2
How often bothered
How often bothered
Not at all
Several
days
> 50%
days
Nearly every
day
Not at all
Several
days
> 50%
days
Nearly every
day
p value
a
63.5%
30.2%
5.2%
1.0%
47.3%
43.8%
7.1%
1.8%
< 0.05
b
65.6%
30.2%
4.2%
0.0%
50.9%
40.2%
5.4%
3.6%
< 0.01
c
59.9%
29.7%
8.9%
1.6%
50.0%
29.5%
12.5%
8.0%
< 0.05
d
45.0%
48.7%
6.3%
0.0%
33.0%
45.5%
13.4%
8.0%
< 0.001
e
67.2%
29.2%
3.1%
0.5%
54.5%
32.1%
8.0%
4.5%
< 0.05
f
81.2%
16.8%
2.1%
0.0%
72.3%
24.1%
1.8%
1.8%
n.s.
g
75.5%
21.4%
2.1%
1.0%
58.9%
35.7%
0.9%
4.5%
< 0.05
h
87.5%
10.4%
2.1%
0.0%
75.9%
21.4%
1.8%
0.9%
< 0.05
i
97.4%
2.6%
0.0%
0.0%
91.1%
7.1%
1.8%
0.0%
< 0.05
Questions
#3
Not at all
Several
days
> 50%
days
Nearly every
day
Not at all
Several
days
> 50%
days
Nearly every
day
p value
a
52.6%
41.1%
6.3%
0.0%
43.8%
48.2%
8.0%
0.0%
n.s.
b
78.6%
21.4%
0.0%
0.0%
68.8%
27.7%
3.6%
0.0%
< 0.05
How often bothered
How often bothered
c
79.2%
20.3%
0.5%
0.0%
62.5%
31.3%
6.3%
0.0%
< 0.001
d
82.2%
16.8%
1.0%
0.0%
68.8%
25.9%
5.4%
0.0%
< 0.01
e
73.4%
21.9%
4.7%
0.0%
64.3%
28.6%
7.1%
0.0%
n.s.
f
82.3%
16.7%
1.0%
0.0%
68.8%
26.8%
4.5%
0.0%
< 0.01
g
68.6%
28.3%
3.1%
0.0%
58.0%
34.8%
7.1%
0.0%
< 0.05
*Only in females; n.s. = not significant.
86
Prevalência de queixas adversas de origem não medicamentosa e sua correlação com
as características de personalidade
7
Adverse nondrug-related complaints
Table 3.
Complaints comparison between the Phase I volunteers group and the control group (Pearson’s Chi-square test). Part II.
Questions
#4
Phase I Volunteers (n = 192)
2 test
Control Group (n = 192)
How much bothered
How much bothered
A lot
Not
A lot
p
value
63.4%
8.9%
< 0.001
54.5%
19.6%
n.s.
Not
A little
A little
a
52.6%
44.8%
2.6%
27.7%
b
28.6%
58.9%
12.5%
25.0%
c
83.9%
11.5%
4.7%
73.2%
20.5%
5.4%
n.s.
d
79.5%
15.3%
5.3%
67.0%
26.8%
4.5%
< 0.05
e
87.5%
12.5%
0.0%
78.6%
17.0%
3.6%
p < 0.01
f
44.8%
45.3%
9.9%
39.3%
47.3%
13.4%
n.s.
g
38.0%
54.2%
7.8%
42.0%
43.8%
13.4%
n.s.
h
80.7%
17.2%
2.1%
76.8%
18.8%
4.5%
n.s.
i
63.5%
28.1%
8.3%
63.4%
28.6%
8.0%
n.s.
j
87.5%
12.0%
0.5%
85.7%
10.7%
3.6%
n.s.
*Only in females; n.s. = not significant.
scores in the Phase I volunteers and the control group.
Significantly lower STAI-T scores (Z =
–3.750, p < 0.001) were found in Phase I volunteers compared to control group subjects. A
STAI-T score < 40 was reported by 74.9% of
Phase I volunteers compared to 55.4% of the
control group while 25.1% of Phase I volunteers vs. 44.6% of the control group had a
STAI-T score of  40 (p < 0.001).
Although mean BDI-II scores were lower in
Phase I volunteers than in the control group, the
difference was not statistically significant (Z =
–1.459). However, the groups differed in terms
of the distribution of subjects by complaint severity: 5.2% of Phase I volunteers (n = 192) vs.
8.0% of the control group (n = 112) had mild
complaints, 1.6% vs. 6.3% had moderate complaints and 0 vs. 2.7% had severe complaints
(2 = 2.70, df = 3, p < 0.01).
Significantly higher SES Total (Z = –2.878,
p < 0.01), Initiative and Persistence (Z =
–2.959, p < 0.01), Planning/Goal Setting (Z =
–2.079, p < 0.05) and Social Self-Efficacy (Z =
–2.236, p < 0.05) scores were found in Phase I
volunteers compared to the control group.
Figure 1. Box-and-whiskers plot showing median,
quartiles and range values of STAI-T, BDI-II and
SES total scores in the Phase I volunteers (n = 192)
and the control group (n = 112).
87
Correlation between scales
scores and complaints
Table 4 presents the results of the correlation analysis between the reported complaints
CAPÍTULO IV
8
Almeida, Coelho, Albino-Teixeira and Soares-da-Silva
Table 4. Correlation coefficients between the STAI-T, BDI-II and SES scores and the self-reported complaints in the phase I volunteers and the control group of healthy subjects (Spearman’s rho).
Questions
#
Phase I Volunteers (n = 192)
STAI-T
BDI-II
SES Total
Control Group (n = 112)
STAI-T
BDI-II
1.a
0.189*
0.271**
1.b
0.221*
0.331***
1.c
0.207*
0.333***
0.193*
0.246**
SES Total
1.d
1.e
1.f
0.233**
0.397***
0.513***
1.g
0.222*
0.200*
1.h
0.269**
0.240*
1.i
0.365***
0.299**
-0.184*
0.376***
-0.306**
1.j
0.232**
0.259**
1.k
0.150*
0.203*
1.l
0.197**
1.m
0.154*
-0.196**
0.213*
-0.203*
0.316***
0.353***
0.343***
2.a
0.244**
0.418***
-0.279***
0.486***
0.521***
-0.197*
2.b
0.432***
0.499***
-0.325***
0.585***
0.619***
-0.235*
2.c
0.259***
0.349***
0.443***
0.543***
2.d
0.178*
0.354***
-0.227**
0.458***
0.582***
2.e
0.263***
0.402***
-0.220**
0.486***
0.462***
2.f
0.311***
0.370***
-0.356***
0.512***
0.495***
2.g
0.200**
0.297***
-0.330***
0.435***
0.402***
2.h
0.194**
0.389***
0.409***
2.i
0.210**
0.413***
0.447***
0.563***
3.a
0.323***
0.494***
-0.191**
0.477***
3.b
0.337***
0.347***
-0.184*
0.0.400***
0.444***
3.c
0.269***
0.388**
-0.199**
0.413***
0.493***
3.d
0.154*
0.375***
0.439***
0.449***
3.e
0.230**
0.294***
0.493***
0.483***
3.f
0.281***
0.372***
-0.245**
0.481***
0.450***
3.g
0.308***
0.421***
-0.155*
0.465***
0.463***
4.a
4.b
0.214**
0.434***
0.341***
0.235*
0.209**
0.274***
-0.177*
362***
0.335***
0.296***
-0.205**
0.260**
4.f
0.243**
0.208**
4.g
0.312***
0.377***
-0.206**
4.h
0.274***
0.317***
-0.154*
0.277**
0.300**
4.i
0.268***
0.416***
-0.273***
0.422***
0.454***
4.j
0.193*
0.246**
0.328***
0.285**
4.c
4.d
4.e
0.302**
0.307**
0.295**
0.407***
0.332***
0.347***
*p < 0.05, **p < 0.01, ***p < 0.001, empty cells = not significant correlation coefficients.
88
-0.323***
Prevalência de queixas adversas de origem não medicamentosa e sua correlação com
as características de personalidade
9
Adverse nondrug-related complaints
and the STAI-T, BDI-II and SES scores. Significant direct correlations were found between the STAI-T and BDI-II scores and
many complaints. The magnitude of the correlation was usually lower in the Phase I volunteers than in the control group, which may
be due to the fact that Phase I volunteers are a
self-selected population with lower scores of
STAI-T and BDI-II and lower levels of complaints. Significant inverse correlations were
found between SES subscales and several
complaints; here the correlation was higher in
the Phase I volunteers than the control group,
due to the fact that Phase I volunteers have
higher scores of SES.
Discussion
To our knowledge, this is the first study in
which the tendency to report nondrug-related
complaints was compared between a group of
healthy subjects who volunteered for Phase I
participation and a control group of healthy
subjects from general population.
Many of the complaints asked about in the
questionnaire are those frequently reported as
adverse events in Phase I studies. Compared
with the control subjects, Phase I volunteers
had a significantly lower incidence of stomach pain, back pain, limbs or joints pain,
headaches, fainting spells, palpitations, shortness of breath, constipation, loose stools, or
diarrhoea, nausea, gas or indigestion, feeling
nervous or anxious, feeling restless, getting
tired very easily, muscle tension, aches, or
soreness, and concentration difficulties, which
are common adverse events during Phase I
studies. Our study clearly shows that the incidence of complaints in the recent past (2 – 4
weeks) in Phase I study volunteers was lower
than that of a sample of healthy subjects from
normal healthy population with the same demographic, socioeconomic and life-style
characteristics, and suggests that the incidence of many complaints may be influenced
by the individual psychological characteristics. This view was reinforced by the finding
of a significant direct relationship between
trait anxiety or depressive mood and the incidence of many complaints, and a significant
inverse correlation between perceived
self-efficacy and the incidence of several
complaints.
89
As previously reported by our group [Almeida et al. 2008] and other authors [Ball et
al. 1993, Cami et al. 1989, Farre et al. 1995,
Meyer et al. 1995, Pieters et al. 1992], healthy
subjects who volunteer for Phase I studies
may significantly differ from their peers in the
general population in many aspects of their
personality characteristics. Healthy subjects
with higher anxiety and social avoidance
[Almeida et al. 2008] and those with less perceived self-efficacy [Almeida et al. 2008]
were found to exclude themselves from volunteering for Phase I studies. The current
study shows that this self-selection bias led
to a sample of Phase I volunteers with a
lower tendency to report nondrug-related
complaints than their peers from the general
population.
Phase I data form the basis for go/no go
decisions in drug development, and therefore
the evaluation of the adverse events in Phase I
studies is of paramount importance. However, the interpretation of adverse events of
new drugs is very difficult, because at this
stage there is no or only little clinical experience and, on the other hand, adverse events
may occur due to study conditions [Hermann
et al. 1997] or the nocebo effect [Lutfullin et
al. 2005]. The results of this study indicate
that an additional bias or confounding factor
can be the subject’s personality, because it
influences the probability of reporting a
nondrug-related complaint that, when occurring during a study, has to be considered as an
adverse event. The self-selection of participants with a lower tendency to report nondrug-related complaints may potentially decrease the likelihood of reporting adverse
events during Phase I clinical trials and, in
case of multiple group studies, may lead to biased results due to between-group differences. It has been suggested that the assessment of the individual’s personality could be
of significant importance in all Phase I studies
[Meyer 2001], because it would allow for a
stratification of volunteers according to these
traits in order to obtain homogeneous study
groups and/or to control for the personality
influence on the Phase I data at the time of the
statistical analysis [Tishler et al. 2005].
In conclusion, the incidence of complaints in healthy subjects is not of a random
character and depends on pschological characteristics. Volunteers in Phase I studies are a
CAPÍTULO IV
Almeida, Coelho, Albino-Teixeira and Soares-da-Silva
self-selected sample with a lower tendency to
report nondrug-related adverse events than
their peers from the general population. The
impact of this self-selection bias on the assessment of tolerability during Phase I studies
deserves further evaluation.
Acknowledgments
We thank the study subjects and Professor
Manuel Vaz-da-Silva, Dr Teresa Nunes and
Benedita Azevedo for their cooperation. We
thank Professor Amílcar Falcao and Dr Grazina Binkauskiene for their assistance in the
statistical analysis.
References
Almeida L, Azevedo B, Nunes T, Vaz-da-Silva M, Soaresda-Silva P. Why healthy subjects volunteer for Phase I
studies and how they perceive their participation? Eur
J Clin Pharmacol. 2007; 63: 1085-1094.
Almeida L, Kashdan T, Coelho R, Albino-Teixeira A,
Soares-da-Silva P. Healthy subjects volunteering for
Phase I studies: the influence of curiosity, exploratory
tendencies, and perceived self-efficacy. Int J Clin
Pharmacol Ther. 2008; 46: 109-118.
Almeida L, Kashdan T, Nunes T, Coelho R, AlbinoTeixeira A, Soares-da-Silva P. Who volunteers for
Phase I clinical trials? Influences of anxiety, social
anxiety and depressive symptoms on self-selection
and reporting of adverse events. Eur J Clin Pharmacol.
2008; 64: 575-582.
Ball CJ, McLaren PM, Morrison PJ. The personality
structure of “normal” volunteers. Br J Clin Pharmacol.
1993; 36: 369-371.
Beck AT, Steer RA, Brown GK. BDI-II: Beck Depression
Inventory – second edition manual. San Antonio, TX;
1996. The Psychological Corporation.
Cami J, Llorente M, Farre M, Badenas J-M, Ugena B.
Personality of healthy volunteers participating in
Phase I clinical trials. Person Individ Diff. 1989; 10:
1199-1200.
Coelho R, Martins A, Barros H. Clinical profiles relating
gender and depressive symptoms among adolescents
ascertained by the Beck Depression Inventory II. Eur
Psychiatry. 2002; 17: 222-226.
Farre M, Lamas X, Cami J. Sensation seeking amongst
healthy volunteers participating in Phase I clinical trials. Br J Clin Pharmacol. 1995; 39: 405-409.
Hermann R, Heger-Mahn D, Mahler M, Seibert-Grafe
M, Klipping C, Breithaupt-Grogler K, De Mey C. Adverse events and discomfort in studies on healthy subjects: the volunteer’s perspective. Eur J Clin Pharmacol. 1997; 53: 207-214.
Kashdan TB, Roberts JE. Social anxiety’s impact on affect, curiosity, and social self-efficacy during a high
self-focus social threat situation. Cognit Ther Res.
2004; 28: 119-141.
Lutfullin A, Kuhlmann J, Wensing G. Adverse events in
volunteers participating in Phase I clinical trials: a sin-
90
10
gle-center 5-year survey in 1,559 subjects. Int J Clin
Pharmacol Ther. 2005; 43: 217-226.
Meyer FP, Haschke R, Rohl FW. The healthy volunteer in
clinical pharmacology: personality and motivation.
Eur J Clin Pharmacol. 1995; 48: 91-96.
Meyer FP, Troger U, Rohl FW. Adverse nondrug reactions:
an update. Clin Pharmacol Ther. 1996; 60: 347-352.
Meyer FP. Psycho- and immunopharmacological factors
relevant to selection of volunteers in clinical studies.
Int J Clin Pharmacol Ther. 2001; 39: 300-310.
Pieters MS, Jennekens-Schinkel A, Schoemaker HC, Cohen
AF. Self-selection for personality variables among
healthy volunteers. Br J Clin Pharmacol. 1992; 33:
101-106.
Reidenberg MM, Lowenthal DT. Adverse nondrug reactions. N Engl J Med. 1968; 279: 678-679.
Ribeiro JLP. Adaptaçao de uma escala de avaliaçao da
auto-eficácia geral. In: Almeida L, Ribeiro I (eds.) Braga:
APPORT. 1995. Avaliaçao Psicológica: formas e contextos p. 163-176.
Sherer M, Maddux JE, Mercandante B, Prentice-Dunn S,
Jacobs B, Rogers RW. The Self-Efficacy Scale: Construction and validation. Psychol Rep. 1982; 51:
663-671.
Spielberger CD, Gorsuch RL, Lushene RE. STAI manual
for the state-trait anxiety inventory. 1970: Palo Alto:
Consulting Psychologists Press; 1970.
Spitzer RL, Kroenke K, Williams JB. Validation and utility
of a self-report version of PRIME-MD: the PHQ primary
care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA. 1999; 282:
1737-1744.
Tishler CL, Bartholomae S, Rhodes AR. Personality profiles of normal healthy research volunteers: a potential
concern for clinical drug trial investigators? Med
Hypothes. 2005; 65: 1-7.
CAPÍTULO V
Análise exploratória da influência das variáveis psicométricas relevantes no processo
de auto-selecção de voluntários para participação em ensaios de fase I sobre a
farmacocinética dos fármacos estudados nos mesmos ensaios
Resultados não submetidos a publicação
Análise exploratória da influência das variáveis psicométricas relevantes no processo de auto-selecção de voluntários
para participação em ensaios de fase I sobre a farmacocinética dos fármacos estudados nos mesmos ensaios
Análise exploratória da influência das variáveis psicométricas
relevantes no processo de auto-selecção de voluntários para
participação em ensaios de fase I sobre a farmacocinética dos
fármacos estudados nos mesmos ensaios
de um ensaio com a mesma entidade química. Ou seja,
nenhum indivíduo da população estudada participou em
mais do que um ensaio com um dado fármaco, embora possa ter participado em ensaios de mais do que um fármaco.
Durante a preparação dos resultados para análise foi
considerada a necessidade de fazer as seguintes exclusões:
–– O nebicapone é um inibidor da enzima catecol-Ometiltransferase (COMT), da qual é simultaneamente
substrato. Nos ensaios usaram-se diferentes doses de
nebicapone. Assim, obtiveram-se diferentes graus de
inibição da COMT e, consequentemente, diferentes
taxas de conversão de nebicapone em metil-nebicapone sob acção da COMT. Ou seja, o efeito farmacodinâmico do nebicapone afecta de modo significativo
a cinética do fármaco. Considerou-se, pois, que a farmacocinética do nebicapone não era um modelo adequado para o tipo de análise pretendida e os ensaios
com nebicapone foram retirados do estudo.
–– O acetato de eslicarbazepina é um pró-fármaco
de eslicarbazepina (BIA 2-194). A R-licarbazepina
(BIA 2-195) é um metabolito menor. Quando se usa
um método aquiral de doseamento, o método não
distingue entre eslicarbazepina e R-licarbazepina e
a mistura enantiomérica é relatada como BIA 2-005
(R-S-licarbazepina). Na maior parte dos ensaios com
acetato de eslicarbazepina realizados na UFH foi usado um método aquiral. Contudo, em dois ensaios utilizou-se um método quiral, impedindo que os dados
1. Objectivo
O objectivo desta análise foi o de explorar a relação
entre os factores psicológicos que mostraram influência
significativa na auto-selecção dos indivíduos como voluntários para participação em ensaios de fase I – nomeadamente as tendências exploratórias (avaliadas pelo
Curiosity and Exploration Inventory Trait, CEI-T), a iniciativa e a persistência (avaliadas pela Self-Efficacy Scale,
SES) e a ansiedade (avaliada pelo State-Trait Anxiety Inventory, STAI) – e a cinética dos fármacos estudados nos
ensaios de fase I em que esses indivíduos participaram.
A hipótese de partida foi a de que os factores psicológicos estudados podem, potencialmente, afectar a cinética dos fármacos, particularmente a sua absorção.
2. Material
Foram incluídos, inicialmente, todos os ensaios de fase
I realizados na Unidade de Farmacologia Humana (UFH)
de BIAL com a população estudada nos trabalhos desta dissertação e que englobaram os seguintes fármacos:
nebicapone (BIA 3-202), acetato de eslicarbazepina (BIA
2-093) e trans-resveratrol (BIA 6-512). Em todos estes ensaios fizeram-se doseamentos das concentrações plasmáticas do fármaco administrado e/ou dos seus metabolitos.
De acordo com as regras de selecção de voluntários em
vigor na UFH, nenhum indivíduo pode participar em mais
93
CAPÍTULO V
cinéticos desses ensaios pudessem ser integrados com
os dos demais ensaios com acetato de eslicarbazepina.
Foram, assim, retirados da análise.
Após estas restrições, e as óbvias exclusões dos indivíduos que foram aleatorizados à toma de placebo nos ensaios controlados com placebo e dos indivíduos para os
quais não se dispunha de avaliação psicométrica, o número de indivíduos disponível para análise (isto é, aqueles para os quais simultaneamente se dispôs de dados farmacocinéticos e de dados psicométricos) foi o seguinte:
– Acetato de eslicarbazepina: n = 23
– Trans-resveratrol: n = 37
Dado que as amostras finais podem considerar-se relativamente reduzidas, decidiu-se adoptar uma abordagem
populacional à análise dos dados cinéticos, antes do estudo da sua correlação com os resultados psicométricos.
senvolvimento de um modelo de efeitos mistos baseado numa aproximação paramétrica. Recorreu-se para o
efeito ao programa estatístico informático mais universalmente utilizado para este tipo de análise – o NONMEM (Nonlinear Mixed Effects Model) (NONMEM Project
Group, University of California at San Francisco) [Sheiner et al., 1977; Sheiner & Beal, 1980; Sheiner & Beal,
1981; Sheiner & Beal, 1983]. Utilizou-se a versão VI,
mais recente.
Utilizando o NONMEM para a determinação de parâmetros farmacocinéticos populacionais pode, mediante
algumas assunções, obter-se estimadas finais tanto para
os parâmetros de efeitos fixos (θ) como para os parâmetros de efeitos aleatórios (η, ε). O procedimento através
do qual o NONMEM procura essas estimadas finais baseia-se num processo iterativo utilizando, para o efeito,
um algoritmo de gradiente que permite minimizar uma
função-objectivo conhecida pela designação de mínimos
quadrados expandidos (extended least squares - ELS). A
estrutura geral da função-objectivo utilizada pelo NONMEM é a seguinte:
3. Técnica analítica de doseamento
dos fármacos
As concentrações plasmáticas de BIA 2-005 e de
trans-resveratrol foram determinadas sob condições de
Boas Práticas de Laboratório, usando métodos previamente validados. As amostras de plasma para a determinação das concentrações plasmáticas de BIA 2-005 foram
preparadas a partir de amostras de sangue venoso colhidas em tubos Vacutainer® com heparina-lítio como anticoagulante e centrifugadas a aproximadamente 1500 g
durante 10 minutos, a 4 ºC. As doses de acetato de eslicarbazepina foram 800 mg ou 1200 mg. Tipicamente,
as amostras de sangue foram colhidas 0.5, 1, 2, 3, 4, 6,
8, 12, 18 e 24 h após a dose. O método de doseamento
consistiu em cromatografia líquida de alto desempenho
(HPLC) com detecção espectrométrica de massa tandem
(MS-MS) e encontra-se descrito em artigo submetido
para publicação [Vaz-da-Silva et al., 2008]. O limite inferior de quantificação era de 10 ng/mL.
As amostras para a determinação das concentrações
plasmáticas de trans-resveratrol foram preparadas a partir de amostras de sangue venoso tomadas em tubos Vacutainer® com EDTA K3 como anticoagulante e centrifugadas a aproximadamente 1500 g durante 10 minutos, a
4 ºC. O método de doseamento consistiu em HPLC com
detecção espectrométrica de massa e encontra-se descrito
em artigo em publicação [Almeida et al., 2008]. O limite
inferior de quantificação era de 0,5 ng/mL.
OBJ(θ,ξ,y) = Σ [(yj - f(θ,xj)).v(θ,ξ,xj)-1.(yj - f(θ,ξ,xj))
t + ln(det v(θ,ξ,xj))]
onde xj e yj correspondem, respectivamente, ao vector dos valores da variável independente e ao vector dos
valores da variável dependente para o indivíduo j. O
símbolo f refere-se, igualmente, a um vector, de tal forma que f(θ,xj) corresponde ao valor médio populacional
de yj, enquanto que v representa a matriz de variânciacovariância populacional relativa a yj que, por sua vez, é
função dos parâmetros de efeitos fixos (θ), dos parâmetros de efeitos aleatórios (ξ) e dos valores que toma a variável independente. Porque se trata de uma população,
y representa o conjunto dos vectores yj, razão pela qual a
função-objectivo é constituída por um somatório cuja dimensão é igual ao número de indivíduos que compõem
a população. Finalmente, em linguagem matricial, det A,
At e A-1 significam, respectivamente, determinante de A,
matriz transposta de A e matriz inversa de A.
Para efeitos práticos, v corresponde ao valor que diferencia uma regressão não-linear ordinária de uma regressão não-linear ponderada, mas que, na função-objectivo
utilizada pelo NONMEM, se encontra um pouco mais
desenvolvida correspondendo à seguinte expressão:
v (θ,ξ,xj) = G(θ,xj).Ω.G(θ,xj)t + σ2 Inj
4. Programa estatístico
onde Ω corresponde à matriz de variância-covariância
relativa aos ηj (parâmetros de efeitos aleatórios inter-individuais), σ2 representa a variância de ε (variabilidade
residual) e Inj refere-se à matriz identidade nj-dimensio-
A caracterização do comportamento cinético do BIA
2-005 e do trans-resveratrol foi efectuada através do de-
94
Análise exploratória da influência das variáveis psicométricas relevantes no processo de auto-selecção de voluntários
para participação em ensaios de fase I sobre a farmacocinética dos fármacos estudados nos mesmos ensaios
a modelização fármacoestatística pretendida procedeu-se
à selecção do modelo monocompartimental aberto, assumindo a existência de uma cinética de incorporação e de
eliminação de primeira ordem. A implementação deste
modelo farmacocinético no NONMEM permite o ajuste
dos dados e cuja descrição na sua forma recursiva apresenta o seguinte aspecto:
nal. A função G dá-nos a matriz Jacobiana das derivadas parciais de f(θ,xj) com respeito aos elementos de ηj,
quando ηj = 0 [Sheiner & Beal, 1980; Metzler, 1987].
A probabilidade de se obterem os valores de concentração observados como uma função dos parâmetros do
modelo é o que se denomina "função de probabilidade".
Segundo o princípio de máxima probabilidade, os parâmetros farmacocinéticos mais prováveis para uma determinada população (os que melhor ajustam os dados
de concentração-tempo observados) serão aqueles que
maximizam essa função de probabilidade (Lmáx.) [Murphy, 1982].
Quando os erros (parâmetros de efeitos aleatórios)
possuem uma distribuição estatística normal (tal como
é assumido na presente situação pelo NONMEM), o
método dos mínimos quadrados expandidos adquire a
designação de método de máxima probabilidade, de tal
forma que se pode relacionar a função-objectivo acima
descrita com a função de máxima probabilidade (Lmáx.)
da seguinte forma:
A02 = A01 exp (-Ke Δt) + A11 Ka [exp (-Ke Δt) - exp
(-Ka Δt)] / (Ka - Ke)
A12 = A11 exp (-Ka Δt) + F.d
C = A02 / Vd
Ke = CL / Vd
onde A01 e A02 representam, respectivamente, as
quantidades de fármaco no organismo aos tempos 1 e 2;
A11 e A12 representam, respectivamente, as quantidades
de fármaco no local de absorção aos tempos 1 e 2; C é
a concentração prevista; F é a fracção de dose disponível; Ka é a constante de absorção; Ke é a constante de
eliminação; d é a dose introduzida no local de absorção
a tempo 2; Vd é o volume de distribuição determinado
pelo modelo de regressão; CL é a depuração (“clearance”)
determinada pelo modelo de regressão; e Δt é o intervalo
de tempo entre t1 e t2 (t2 - t1).
OBJ (ELS) = -2 ln (Lmáx.)
Ou seja, a minimização da OBJ (ELS) corresponde à
maximização de Lmáx., pelo que o valor da função-objectivo funciona como um índice da bondade do ajuste
conseguido pelo modelo fármaco-estatístico considerado
[Yamaoka et al., 1986; Yamaoka et al., 1987].
6. Modelo de regressão múltipla
Os modelos de regressão múltipla pretendem dotar os
parâmetros farmacocinéticos populacionais de um carácter dinâmico. Da conjugação entre o modelo farmacocinético adoptado e um modelo de regressão múltipla no qual
intervêm um conjunto de descritores clínicos (covariáveis)
com significância estatística, resulta a elaboração de um
modelo estrutural capaz de retirar rigidez aos parâmetros
farmacocinéticos tal qual são descritos tendo em consideração apenas o modelo determinístico seleccionado.
A construção do modelo estrutural pode revelar-se
um processo extremamente complicado e moroso, especialmente quando o modelo determinístico envolve bastantes parâmetros e/ou existe uma quantidade apreciável
de covariáveis capazes de, a priori, influenciar o comportamento cinético do fármaco. Quando isso acontece, o
número de combinações possíveis capazes de definir o
modelo estrutural aumenta drasticamente, embora nesses casos a tendência normal seja a de se começar por experimentar aquelas variáveis cuja transcendência clínica
é mais evidente.
Com o objectivo de tentar limitar a construção do
modelo estrutural apenas às covariáveis cuja probabilidade real de entrada no modelo de regressão seja claramente superior, alguns autores têm sugerido, utilizando
5. Modelo farmacocinético
A selecção do modelo farmacocinético corresponde à
primeira fase da modelização fármacoestatística. Na realidade, o modelo determinístico é, na sua forma mais
simples, o primeiro responsável quer pelo ajuste das concentrações observadas, gerando-se dessa forma os parâmetros farmacocinéticos populacionais básicos, quer pela
posterior capacidade preditiva atribuível ao modelo fármacoestatístico no seu todo [Grasela & Sheiner, 1991].
Por este motivo, a selecção do modelo farmacocinético
corresponde a uma etapa determinante na modelização
fármacoestatística, estando aconselhada a comparação
prévia de modelos determinísticos distintos, em função das concentrações disponíveis, sem a introdução de
quaisquer covariáveis, devendo a escolha recair no mais
simples de entre aqueles que demonstrem capacidade
para ajustar de forma satisfatória os dados disponíveis
[Wade et al., 1991].
Tendo em atenção a experiência prévia e a informação
disponível, não parece ser controversa a selecção do modelo monocompartimental aberto para a realização do
presente estudo. Resumidamente pode dizer-se que para
95
CAPÍTULO V
distintas filosofias, um trabalho prévio de triagem que
permita aquilatar da possibilidade que uma determinada covariável tem de entrar no modelo, relativamente
às restantes covariáveis [Maitre et al., 1991; Bennet &
Wakefield, 1994].
Na presente situação, a construção de um modelo
fármacoestatístico que explique o comportamento cinético dos fármacos em estudo centrou-se essencialmente
na eventual influência das componentes psicométricas
que se demonstraram auto-selectoras de participação em
ensaios clínicos de fase I [Almeida et al., 2008a, 2008b
e 2008c] – nomeadamente as tendências exploratórias
(avaliadas pelo CEI-T), iniciativa e persistência (avaliadas pela SES) e ansiedade (avaliada pelo STAI) – sobre
o processo de incorporação dos fármacos. Assim, no
presente processo de modelização fármacoestatística, o
modelo estrutural básico corresponde ao estado inicial
sem a introdução de nenhuma covariável de acordo com
a seguinte estrutura:
7. Modelo estatístico
Para além do modelo estrutural que permite a determinação da tendência central apresentada pelos parâmetros
farmacocinéticos populacionais, interessa também possuir informação acerca do grau probabilístico associado
à dispersão de cada parâmetro, garantindo um carácter
estocástico ao modelo final. Por este motivo, a utilização
conjunta de um modelo estatístico permite determinar o
intervalo de valores que, previsivelmente, os parâmetros
farmacocinéticos individuais poderão tomar tendo como
base de trabalho o valor médio populacional proporcionado pelo modelo estrutural e a estatística inferencial.
As fontes de erro que podem fazer com que os parâmetros farmacocinéticos de um elemento da população
se afastem dos valores médios populacionais encontramse representados pelos parâmetros de efeitos aleatórios
interindividuais e intra-individuais. Se se conseguir desenvolver um modelo estatístico que permita calcular o
conjunto de valores que estes parâmetros de efeitos aleatórios pode tomar, então o problema fica resolvido, na
medida em que, ao valor médio populacional, pode então associar-se as correspondentes medidas de dispersão
e com isso aumentar a capacidade preditiva.
Utilizando equações gerais, pode dizer-se que os parâmetros farmacocinéticos populacionais (φ) serão função (g) de um modelo estrutural dependente da presença
de covariáveis (z) e de parâmetros de efeitos fixos (θ):
Ka (h-1) = θ1
CL (L/h) = θ2
Vd (L) = θ3
onde θ1, θ2 e θ3 correspondem aos parâmetros de
efeitos fixos que, neste caso concreto, coincidem com os
parâmetros farmacocinéticos básicos.
A evolução do modelo estrutural básico para um modelo estrutural intermédio está relacionada com a tentativa de introdução no modelo das covariáveis psicométricas
contínuas em estudo de acordo com o seguinte esquema:
φ = g (z,θ)
Ka (h-1) = Σ (θi * VCOθi); com i = 1,2,3...n
CL (L/h) = θ2
Vd (L) = θ3
Por outro lado, as concentrações previstas (yi), que
funcionam como variável dependente, serão função (f)
dos parâmetros farmacocinéticos populacionais (φi) e
das variáveis independentes (xi), como o tempo ou as
doses administradas:
No sentido de serem exploradas todas as opções possíveis, também a categorização das variáveis psicométricas contínuas foi alvo de análise. No processo de categorização optou-se por determinar a mediana das VCO e
criar assim as VCA (variáveis categóricas) cuja influência
se exercerá de uma forma multiplicativa, podendo aumentar (θVCA > 1) ou diminuir (θVCA < 1) o valor dos
parâmetros farmacocinéticos básicos entretanto determinados.
yi = f (φ, xi)
Com estas equações gerais estamos a trabalhar apenas
com a tendência central relativa aos parâmetros populacionais, ou seja, utilizando apenas o modelo estrutural,
pelo que, até ao momento, a única fonte de variabilidade
é-nos dada pela letra "i". Esta refere-se a uma qualquer
observação capaz de condicionar o valor de y pela simples razão de podermos considerar, por exemplo, tempos
ou doses distintas. No entanto, tudo se passa como se
todos os elementos pertencentes à população apresentassem exactamente o mesmo comportamento, não existindo variabilidade interindividual ou intra-individual.
Através da associação de um modelo de erro, um
indivíduo "j", pertencente à população anteriormente
considerada, vai poder diferir dos valores médios popu-
Ka (h-1) = [Σ (θi * VCOθi)] * [Π (θi )VCA]; com i =
1,2,3...n
CL (L/h) = θ2
Vd (L) = θ3
96
Análise exploratória da influência das variáveis psicométricas relevantes no processo de auto-selecção de voluntários
para participação em ensaios de fase I sobre a farmacocinética dos fármacos estudados nos mesmos ensaios
computacional, mas fundamental como ponto de partida
para uma análise correcta dos dados. Normalmente, a
sua execução é levada a cabo utilizando, para o efeito,
o modelo fármacoestatístico mais simples possível, correspondendo, geralmente, ao modelo estrutural básico já
anteriormente descrito e ao modelo de erro proporcional, tanto para a variabilidade interindividual como para
a variabilidade residual.
Analisando as tabelas e os gráficos assim gerados,
torna-se possível detectar, à partida, eventuais erros na
introdução dos dados (erros fixos), tais como concentrações incorrectas devidas, por exemplo, a um engano nas
casas decimais que, não sendo corrigidas, irão adulterar
toda a análise populacional.
lacionais, passando as equações gerais a incluir, então,
os parâmetros de efeitos aleatórios interindividuais (ηj) e
intra-individuais (εij):
φj = g (zj,θ,ηj)
yij = f (φj, xi,εij)
Quando se utiliza o NONMEM, devido à linearização
que este programa adopta para a introdução dos parâmetros de efeitos aleatórios no modelo fármacoestatístico,
os modelos de erro resumem-se, basicamente, ao modelo
aditivo e ao modelo proporcional. Voltando às equações
gerais, tem-se portanto:
φj = g (zj,θ) + ηj (modelo aditivo)
φj = g (zj,θ) x (1+ηj) (modelo proporcional)
onde ηj corresponde à diferença entre o valor típico populacional e o valor correspondente ao indivíduo
"j", sendo que η é uma variável aleatória, normalmente
distribuída, de média zero e variância ω2, cuja caracterização representa a variabilidade interindividual. Para a
variabilidade residual (expressão mais rigorosa do que
variabilidade intra-individual) as equações gerais tomarão a seguinte forma:
8.2. Construção do modelo estrutural
Já se referiu qual a estratégia seguida para, a partir de
um modelo estrutural básico, se atingir, mediante a incorporação progressiva de parâmetros e/ou covariáveis, o
modelo estrutural completo. No entanto, o crescimento
do modelo básico encontra-se associado à utilização de
meios de diagnóstico que, de acordo com critérios estatísticos apropriados, vão permitindo a incorporação daquelas variáveis cuja introdução no modelo se encontre
plenamente justificada, do mesmo modo que, pelo caminho, vão ficando variáveis que, por não beneficiarem o
modelo, vão sendo preteridas.
Não entrando em grandes detalhes, dir-se-á apenas que a entrada de parâmetros e/ou covariáveis num
modelo, tornando-o mais complexo, apenas se justifica
caso essa adição origine melhorias estatisticamente significativas na bondade do ajuste. Regra geral, não deve
atender-se apenas a critérios isolados, mas sim a um conjunto de informação que pode resumir-se no seguinte:
a) Alterações na função objectivo (FObj);
b) Diminuição nas fontes de variabilidade (ηj e εij);
c) Análise gráfica de confirmação.
yij = f (φj, xij) + εij (modelo aditivo)
yij = f (φj, xij) x (1+εij) (modelo proporcional)
onde εij corresponde à diferença entre a concentração
prevista pelo modelo populacional e a observada no indivíduo "j" a tempo "i", sendo que ε é uma variável aleatória,
normalmente distribuída, de média zero e variância σ2,
cuja caracterização representa a variabilidade residual.
Para o presente trabalho, optou-se inicialmente por
utilizar o modelo proporcional, tanto para a variabilidade
interindividual como para a variabilidade residual, pela
maior facilidade que este tipo de modelo de erro permite no que diz respeito à selecção das estimadas iniciais.
Na realidade, de acordo com o modelo proporcional, à
raiz quadrada de ω2 e σ2 correspondem directamente os
coeficientes de variação de φj e yij. No entanto, durante
a construção do modelo final, o modelo aditivo foi igualmente experimentado, incluindo o conjunto de combinações possíveis entre estes dois modelos.
a) Alterações na função objectivo (FObj)
A FObj funciona como uma medida global da bondade do ajuste conseguido já que, baseando-se no resultado final dos parâmetros estimados, corresponde a -2
ln (Lmáx.), onde (Lmáx.) representa a função de máxima
probabilidade. Quando um modelo actual difere de um
outro anterior pela sua maior complexidade, ou seja, por
possuir mais parâmetros na sua estrutura, a diferença entre as funções-objectivo (DFObj) segue aproximadamente uma distribuição qui-quadrado (χ2), cujos graus de
liberdade equivalem à diferença de parâmetros existente
entre os dois modelos em comparação.
Assumindo-se a existência de uma distribuição es-
8. Construção do modelo
fármacoestatístico
8.1. Revisão da informação
A primeira fase da construção do modelo fármacoestatístico consiste na verificação da informação introduzida nos arquivos de entrada de dados para o NONMEM.
Trata-se de uma etapa bastante simples do ponto de vista
97
CAPÍTULO V
que, por outro lado e idealmente, a representação PRED
vs WRES não deverá apresentar nenhum padrão de distribuição, significando com isso que a variabilidade aleatória existente não segue nenhuma tendência explicável
através do modelo estrutural, ficando a dever-se exclusivamente a factores aleatórios.
tatística conhecida e pela qual se orientam as DFObj,
podemos, mediante a atribuição de uma significância
estatística, testar a bondade de ajuste para modelos que
constituam, entre si, um par complexo/reduzido. O modelo reduzido será idêntico ao modelo complexo com a
excepção de que um ou mais parâmetros tomarão valor
nulo ("0" para coeficientes e "1" para potências).
Quando se pretende contrastar hipóteses, o modelo
reduzido corresponderá à hipótese nula (H0), enquanto
que o modelo complexo representará a hipótese alternativa (HA). Nesta fase de incorporação progressiva, parâmetro a parâmetro, no modelo estrutural, utiliza-se uma
DFObj ≥ 6,6 para a aceitação de HA relativamente à H0,
encontrando-se este valor associado a um valor de p ≤ 0,01
[Wiest et al., 1991; Bruno et al., 1992; Yukawa et al., 1992].
Obviamente que, quando não está presente um par
complexo/reduzido – como acontece, por exemplo, com
a introdução de uma covariável no modelo sem que, com
isso, se altere o número de parâmetros (zero graus de liberdade) – este teste estatístico já não tem aplicabilidade,
embora mande a boa regra que se opte por aquele modelo que apresentar menor FObj, decisão que deverá ser
complementada por outros meios de diagnóstico.
8.3. Construção do modelo estatístico
Basicamente, as regras que se aplicam para a construção do modelo estrutural são transponíveis para o desenvolvimento do modelo estatístico. Sucede que, pelas suas
características intrínsecas, torna-se mais difícil desenvolver com precisão o modelo de erro mais adequado, não
só pelo facto de o NONMEM apresentar algumas limitações neste campo, mas também porque, para o mesmo
nível de precisão, a quantidade de informação necessária
para se obterem resultados similares aos dos parâmetros
de efeitos fixos teria de ser muito superior.
Existe, no entanto, um sinal inequívoco de que o
modelo estatístico possui demasiados ηj, que se revela
quando algum ω (SD para o modelo aditivo e CV para
o modelo proporcional) toma valores próximos de zero.
Quando isso sucede, não significa que o parâmetro farmacocinético em questão não apresenta variabilidade interindividual (o que não é realisticamente plausível), mas
sim que o modelo, no seu todo, consegue encontrar uma
explicação indirecta para essa mesma variabilidade.
b) Diminuição nas fontes de variabilidade (ηj e εij)
A ideia subjacente ao desenvolvimento de um modelo
estrutural reside na tentativa de se procurar explicação
para as diferenças cinéticas existentes entre os indivíduos
pertencentes a uma determinada população. Assim, a introdução de novos parâmetros e/ou covariáveis no modelo só fará sentido se for acompanhada de uma diminuição das fontes de variabilidade aleatória (interindividual
e/ou residual). Por este motivo, a construção do modelo
estrutural deverá ser, normalmente, acompanhada por
uma diminuição nos valores de ηj e εij, significando esse
facto que houve uma transferência de variabilidade aleatória para o modelo estrutural que é, justamente, o que
se pretende com a sua elaboração.
8.4. Refinamento do modelo
O objectivo desta etapa reside na pesquisa de parâmetros excedentários cuja saída do modelo possa melhorar
a precisão dos que nele permanecerem. A técnica básica nesta fase corresponde à determinação da matriz de
variância-covariância para o melhor modelo conseguido
até ao momento, por forma a que se calculem os intervalos de confiança de todos os parâmetros. Depois, resta
procurar os parâmetros cujos valores nulos ("0" para os
coeficientes e "1" para as potências) estejam incluídos
nos intervalos de confiança, procedendo-se, de seguida,
à sua eliminação do modelo. As ferramentas a utilizar
para o refinamento do modelo incluem:
i) erro padrão (SE) da média dos parâmetros;
ii) intervalos de confiança (IC) de 95%;
iii) magnitude relativa (%) dos SEs.
c) Análise gráfica de confirmação
Uma cuidada análise gráfica pode fornecer informações muito importantes, tanto como ferramenta de diagnóstico para a adição de parâmetros e/ou covariáveis,
como também para a confirmação de que um determinado desenvolvimento no modelo corresponde a uma
melhoria no ajuste dos dados.
Normalmente os gráficos mais sugestivos são aqueles
que confrontam as concentrações previstas (PRED) com
as concentrações observadas (DV) ou com os residuais
pesados (WRES). Na prática, quanto melhor for o modelo, maior será a correlação entre PRED vs DV, enquanto
8.5. Teste formal do modelo
Trata-se de um processo oposto ao seguido durante
a construção do modelo fármacoestatístico. Durante essas etapas procede-se a uma incorporação progressiva de
parâmetros e/ou covariáveis, enquanto que, agora, o que
se pretende é inverter o sentido eliminando-se progres-
98
Análise exploratória da influência das variáveis psicométricas relevantes no processo de auto-selecção de voluntários
para participação em ensaios de fase I sobre a farmacocinética dos fármacos estudados nos mesmos ensaios
WRES VS. PRED
WRES
PRED
sivamente parâmetros e/ou covariáveis por forma a que,
no modelo final, apenas se venham a manter aqueles que
são realmente importantes.
Devido aos fundamentos estatísticos que estão na base
da incorporação progressiva de parâmetros e/ou covariáveis e também pelas múltiplas comparações que se têm
de efectuar até que o modelo chegue a esta fase, durante
a eliminação progressiva, apesar de se utilizar o mesmo
teste estatístico baseado na distribuição χ2 das DFObj,
adopta-se um valor de p ≤ 0,005, o que significa uma
DFObj ≥ 7,8 por cada parâmetro eliminado, seguindo-se
o mesmo raciocínio para o contraste de hipóteses (H0 e
HA) [Driscoll et al., 1989; Moore et al., 1989; Bruno et
al., 1992].
9. Resultados
Figura 2 - Residuais pesados (WRES) vs concentrações previstas (PRED)
de BIA 2-005.
9.1. BIA 2-005
9.1.1. Modelo estrutural básico
A Figura 1 apresenta o gráfico das concentrações de BIA
2-005 previstas (PRED) versus as concentrações observadas (DV) e a Figura 2 apresenta o gráfico dos residuais pesados (WRES) versus as concentrações previstas (PRED).
A Figura 3 apresenta a distribuição dos valores de Ka
de BIA 2-005 na população estudada (IPRED).
12
10
Frequência (n)
8
6
4
PRED VS. DV
PRED
2
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Ka(h-1)
DV
Figura 3 - Distribuição dos valores de Ka de BIA 2-005 na população
estudada (IPRED).
O modelo estrutural básico pode ser descrito do seguinte modo:
FObj = 8288,335
Ka (h-1) = θ1
ωKa= 63,4%
CL (L/h) = θ2
ωCL = 19,4%
Vd (L) = θ3
ωVd = 23,6%
σ = 31,5%
Figura 1 - Concentrações previstas (PRED) vs concentrações observadas
(DV) de BIA 2-005.
99
CAPÍTULO V
9.1.2. Modelo estrutural final
A Figura 4 apresenta o gráfico das concentrações previstas (PRED) versus as concentrações observadas (DV) e
a Figura 5 apresenta os residuais pesados (WRES) versus
as concentrações previstas (PRED) no modelo final.
É visível pela análise gráfica, e especialmente pelo
decréscimo observado na função-objectivo, que houve
uma evolução entre o modelo básico e o modelo final,
sendo de destacar a introdução da variável psicométrica “ansiedade-estado” (STAI-S) como factor preditivo
discriminante relativamente à velocidade a que ocorre a
absorção (Ka), sem influência na CL ou Vd. A introdução
PRED VS. DV
PRED
das outras variáveis psicométricas não mostrou valor
preditivo discriminante de qualquer dos parâmetros farmacocinéticos (Ka, CL e Vd).
O modelo estrutural final pode ser descrito do seguinte modo:
FObj = 8280,355
ωKa = 64,3%
Ka (h-1) = θ1
CL (L/h) = θ2 ωCL = 19,5%
Vd (L) = θ3
ωVd = 25,0%
σ = 30,6%
A Figura 6 apresenta a distribuição da população estudada (IPRED), dicotomizada em função da mediana do
STAI-S, por valores de Ka.
6
STAI-S = < 33
STAI-S > 33
5
DV
Frequência (n)
4
3
2
1
0
0.2
Figura 4 - Concentrações previstas (PRED) vs concentrações observadas
(DV).
WRES
PRED
WRES. VS PRED
Figura 5 - Residuais pesados (WRES) vs concentrações previstas
(PRED).
0.4
0.6
0.8
1.0
1.2
1.4
Ka(h-1)
Figura 6 - Valores de Ka na população estudada (IPRED) utilizando
“ansiedade-estado” (STAI-S) como variável categórica.
A Tabela 1 apresenta as estimadas dos parâmetros do
modelo fármacoestatístico final. Para a variável contínua
“ansiedade-estado” (STAI-S) não se cumpriram critérios
estatísticos que permitissem a sua introdução no modelo final, mas a sua categorização através da dicotomização em função da mediana dos valores observados permitiu que tanto a “ansiedade-estado” (STAI-S) como a
“ansiedade-traço” (STAI-T) contribuíssem para a explicação de parte da variabilidade observada ao nível da absorção (Ka). A permanência da variável “ansiedade-estado”
(STAI-S) em detrimento da variável “ansiedade-traço”
(STAI-T) no modelo final fica apenas a dever-se ao facto
da primeira se apresentar mais influente tendo em consideração as ferramentas estatísticas associadas ao processo de modelização. No entanto, também a variável
“ansiedade-traço” (STAI-T) mostrou ser capaz de discriminar a velocidade de absorção consoante os indivíduos
100
Análise exploratória da influência das variáveis psicométricas relevantes no processo de auto-selecção de voluntários
para participação em ensaios de fase I sobre a farmacocinética dos fármacos estudados nos mesmos ensaios
PRED VS. DV
Tabela 1 - Estimadas dos parâmetros do modelo fármacoestatístico final.
Parâmetros
Estimadas
Precisãoa
θSTAI -S
ωKa
ωCL
ωVd
σ
1,38
2,83
68,4
0,60b
64,3%
19,5%
25,0%
30,6%
26,9%
4,8%
6,8%
24,0%
63,8%
34,9%
49,8%
19,0%
DV
θ1
θ2
θ3
PRED
a Magnitude do erro padrão da média expresso em percentagem.
b Corresponde a uma diminuição de 40% na Ka para os indivíduos cuja pontuação
para “ansiedade-estado” (STAI-S) se encontra abaixo da mediana (pontuação ≤
33) em relação aos valores totais observados.
Figura 7 - Concentrações previstas (PRED) vs concentrações observadas
(DV) de trans-resveratrol.
WRES VS PRED
WRES
PRED
se situassem acima ou abaixo da mediana dos valores
observados na sua globalidade. Enquanto que os indivíduos com valores de “ansiedade-estado” (STAI-S) abaixo da mediana em relação ao valores totais observados
(pontuação ≤ 33) apresentam um factor multiplicativo
de 0,60 relativamente à Ka (correspondente a uma diminuição de 40% na Ka para os indivíduos cuja pontuação
para “ansiedade-estado” se encontra abaixo da mediana),
no caso da variável “ansiedade-traço” (STAI-T) dicotomizada (pontuação ≤ 34), o factor multiplicativo é de 0,75
relativamente à Ka (correspondente a uma diminuição
de 25% na Ka para os indivíduos cuja pontuação para
“ansiedade-traço” se encontre abaixo da mediana).
Para além desta observação relativamente à influência
das variáveis psicométricas na velocidade de absorção,
merece igualmente destaque o facto do modelo final
apresentar estimadas extremamente precisas e robustas
para a generalidade dos parâmetros de efeitos fixos (Ka,
CL e Vd) e de efeitos aleatórios (ωKa, ωCL, ωVd e σ).
Este facto revela que o modelo determinístico se encontra estabilizado e, por esse motivo, o efeito sugerido pela
“ansiedade-estado” (STAI-S) na velocidade de absorção
(Ka) não resulta de um artefacto matemáticoestatístico,
mas sim de uma evolução consistente do modelo que
explica os dados observacionais.
Em conclusão, a aplicação do modelo mostrou que
indivíduos mais ansiosos apresentaram maior rapidez de
absorção de BIA 2-005, em relação aos menos ansiosos.
Figura 8 - Residuais pesados (WRES) vs concentrações previstas (PRED)
de trans-resveratrol.
A Figura 9 apresenta a distribuição dos valores de Ka
na população estudada (IPRED).
O modelo final corresponde ao modelo básico e pode
ser descrito do seguinte modo:
9.2. Trans-resveratrol
9.2.1. Modelo estrutural
A Figura 7 apresenta o gráfico das concentrações de
trans-resveratrol previstas (PRED) versus as concentrações observadas (DV) e a Figura 8 apresenta o gráfico
dos residuais pesados (WRES) versus as concentrações
previstas (PRED).
101
FObj = 1282,305
ωKa = 25,2%
Ka (h-1) = θ1
CL (L/h) = θ2
ωCL = 22,5%
Vd (L) = θ3
ωVd = 77,3%
σ = 128,5%
CAPÍTULO V
14
10. Bibliografia
12
Almeida L, Falcao A, Vaz-da-Silva M, Soares E, Costa R,
Loureiro AI, Fernandes-Lopes C, Rocha JF, Nunes T,
Wright L, Soares-da-Silva P (2008). Pharmacokinetic and safety profile of trans-resveratrol in a rising
multiple-dose study in healthy volunteers. Mol Nutr
Food Res. (In press).
Almeida L, Falcao A, Coelho R, Albino-Teixeira A, Soares-da-Silva P (2008a). The role of socio-economic
conditions and psychological factors on the willingness to volunteer for phase I studies. Pharm Med (In
press).
Almeida L, Kashdan T, Nunes T, Coelho R, Albino-Teixeira A, Soares-da-Silva P (2008b). Who volunteers
for phase I clinical trials? Influences of anxiety, social
anxiety and depressive symptoms on self-selection
and reporting of adverse events. Eur J Clin Pharmacol. 64: 575-582.
Almeida L, Kashdan TB, Coelho R, Albino-Teixeira A,
Soares-da-Silva P (2008c). Healthy subjects volunteering for Phase I studies: influence of curiosity, exploratory tendencies and perceived self-efficacy. Int
J Clin Pharmacol Ther. 46(3): 109-18.
Bennet J, Wakefield J (1994). Covariate selection for population pharmacokinetic models. Comunicação
apresentada no "Second Meeting of the Population
Approach Group in Europe", Greenford, London,
UK, 13-14 June.
Bruno R, Iliadis MC, Lacarelle B, Cosson V, Mandema
JW, Le Roux Y, Montay G, Durand A, Ballereau M,
Alasia M, Albanese J, Francois G, Iliadis A, Frydman
A (1992). Evaluation of Bayesian estimation in
comparison to NONMEM for population pharmacokinetic data analysis: application to pefloxacin
in intensive care patients. J Pharmacokin Biopharm.
20(6): 653-669.
Driscoll MS, Ludden TM, Casto DT, Littlefield LC (1989).
Evaluation of theophylline pharmacokinetics in a
pediatric population using mixed effects models. J
Pharmacokin Biopharm. 17(2): 141-168.
Grasela TH, Sheiner LB. Pharmacostatistical modeling
for observational data (1991). J Pharmacokin Biopharm. 19(3; Suppl.): 25S-36S.
Maitre PO, Bührer M, Thomson D, Stanski DR (1991). A
three-step approach combining Bayesian regression
and NONMEM population analysis: application to
midazolam. J Pharmacokin Biopharm. 19(4): 377-384.
Metzler CM. Extended least squares (ELS) for pharmacokinetic models (1987). J Pharmaceut Sci. 76(7):
565-571.
Frequência (n)
10
8
6
4
2
0
0
20
40
60
80
100
Ka(h-1)
Figura 9 - Distribuição dos valores de Ka na população estudada
(IPRED).
As estimadas dos parâmetros do modelo fármacoestatístico final são apresentadas na Tabela 2. A elevada
variabilidade dos dados condiciona a modelização. Por
exemplo, o valor estimado para Ka não é fidedigno (precisão da variabilidade em Ka é de 183%) e a precisão no
cálculo dos parâmetros de efeitos aleatórios inter-individuais é baixa, com excepção da variabilidade residual que,
sendo relativamente bem estimada (20,2% de precisão),
apresenta, no entanto, um valor extremamente elevado
(129%). Embora os valores para CL e Vd possam ser fiáveis, é visível pelo restante modelo que existe uma grande
dificuldade em racionalizar o comportamento cinético.
A impossibilidade de construir um modelo cinético
de trans-resveratrol fiável constituiu uma limitação intransponível à análise, e condicionou a possibilidade de
sucesso da introdução das variáveis psicométricas no
modelo. Daí, o facto do modelo básico ser simultaneamente o modelo final.
Tabela 2 - Estimadas dos parâmetros do modelo fármacoestatístico final.
Parâmetros
Estimadas
Precisãoa
θ1
19,8
37,8%
3160
15,3%
θ2
12100
15,3%
θ3
25,2%
183,0%
ωKa
22,5%
157,0%
ωCL
77,3%
43,5%
ωVd
σ
129%
20,2%
a Magnitude do erro padrão da média expresso em percentagem.
102
Análise exploratória da influência das variáveis psicométricas relevantes no processo de auto-selecção de voluntários
para participação em ensaios de fase I sobre a farmacocinética dos fármacos estudados nos mesmos ensaios
Moore ES, Faix RG, Banagale RC, Grasela TH (1989).
The population pharmacokinetics of theophylline
in neonates and young infants. J Pharmacokin Biopharm. 17(1): 47-66.
Murphy EA, ed (1982). Biostatistics in Medicine. Baltimore; The Johns Hopkins University Press.
Wade JR, Beal SL, Sambol NC (1991). Interaction between structural, statistical, and covariate models in
population pharmacokinetic analysis. J Pharmacokin
Biopharm. 22(2): 165-177.
Wiest DB, Pinson JB, Gal PS, Brundage RC, Schall S,
Ransom JL, Weaver RL, Purihit D, Brown Y (1991).
Population pharmacokinetics of intravenous indomethacin in neonates with symptomatic patent ductus arteriosus. Clin Pharmacol Ther. 49: 550-557.
Sheiner LB, Rosenberg B, Marathe V (1977). Estimation
of population characteristics of pharmacokinetic parameters from routine clinical data. J Pharmacokin
Biopharm. 5(5): 445-479.
Sheiner LB, Beal SL (1980). Evaluation of methods for
estimating population pharmacokinetic parameters.
I. Michaelis-Menten model: routine clinical pharmacokinetic data. J Pharmacokin Biopharm. 8(6):
553-571.
Sheiner LB, Beal SL (1981). Evaluation of methods for estimating population pharmacokinetic parameters. II.
Biexponential model and experimental pharmacokinetic data. J Pharmacokin Biopharm. 9(5): 635-651.
Sheiner LB, Beal SL (1983). Evaluation of methods for
estimating population pharmacokinetic parameters.
III. Monoexponential model: routine clinical pharmacokinetic data. J Pharmacokin Biopharm. 11(3):
303-319.
Vaz-da-Silva M, Almeida L, Falcão A, Soares E, Maia J,
Soares-da-Silva P (2008). Effect of eslicarbazepine
acetate on the steady-state pharmacokinetics and
pharmacodynamics of warfarin in healthy subjects.
J Clin Pharmacol (submetido para publicação).
Yamaoka K, Tanaka H, Okumura K, Yasuhara M, Hori R
(1986). An analysis program MULTI(ELS) based on
extended nonlinear least squares method for microcomputers. J Pharmacobio-Dyn. 9: 161-173.
Yamaoka K, Tanaka H (1987). A new version of
MULTI(ELS) for extended nonlinear least squares. J
Pharmacobio-Dyn. 10: 26-34.
Yukawa E, Higuchi S, Aoyama T (1992). Phenobarbitone
population pharmacokinetics from routine clinical
data: role of patient characteristics for estimating
dosing regimens. J Pharm Pharmacol. 44: 755-760.
103
Discussão e Conclusões
Discussão e Conclusões
Apesar da importância fundamental dos ensaios clínicos de fase I no desenvolvimento
de medicamentos, conhece-se pouco sobre os indivíduos que neles participam e sobre a sua
validade externa, isto é, sobre o seu grau de representatividade em relação à população geral
[Tishler & Bartholomae, 2002]. Por outro lado, o conhecimento é também escasso no que
respeita ao possível impacto do eventual viés de selecção nos resultados dos ensaios de fase
I [Tishler et al., 2005].
A consulta da bibliografia publicada mostra que os trabalhos desta dissertação podem
considerar-se, tanto quanto conhecemos, a primeira tentativa sistematizada de caracterizar
o perfil demográfico, social e psicológico dos participantes em ensaios de fase I, de avaliar
o possível viés de selecção (pela comparação entre as características de participantes versus
os seus pares na população geral) e de estudar a relação entre as diferenças impostas pela
auto-selecção e os resultados dos ensaios de fase I, nomeadamente a nível da avaliação de
tolerabilidade e da farmacocinética.
1. Os resultados relatados no Capítulo I mostram que as motivações dos indivíduos
saudáveis que se voluntariam para ensaios clínicos de fase I são influenciadas por algumas
características sóciodemográficas e diversos traços de personalidade. Em consonância com
o conhecimento empírico e o que está relatado na literatura [Anderson & Weijer, 2001; Ayd
& Calabresi, 1972; Bigorra & Baños, 1990; Dickert & Grady, 1999; Erb et al., 1994; Gribbin
& Schaie, 1976; Hyun, 2006; McGee, 1997; Russel et al., 2000; Tishler & Bartholomae, 2002;
Tishler & Bartholomae, 2003; Van Gelderen et al., 1993; Vere, 1991; Viens, 2001; Weijer,
1999; Zink, 2001], também na nossa população a motivação económica (isto é, a expectativa de receber uma compensação financeira pela participação) foi a principal razão para participação num ensaio clínico. Contudo, outras motivações foram também declaradas como
importantes para a decisão de voluntariar, como a curiosidade e a oportunidade de obter
uma avaliação clínica gratuita, bem assim como outras de natureza mais altruística (por
exemplo, a contribuição para o desenvolvimento de novos medicamentos).
As diferentes motivações para participação não variaram com o género, com a ocupação profissional ou com a idade; contudo, no que respeita à idade, deve ter-se em atenção o
facto da população-alvo ser relativamente homogénea, pelo que o nosso estudo não permite
conclusões definitivas a este respeito.
Tal como era esperado, encontrou-se uma correlação inversa entre a importância da
compensação financeira como motivo de participação e o rendimento mensal; ou seja, os
107
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
indivíduos com menores rendimentos tendem a valorizar mais a compensação financeira do
que os com maiores rendimentos. Este aspecto deve merecer especial atenção, dado que pode
encerrar questões éticas, já que as pessoas economicamente mais desfavorecidas podem ser
particularmente mais vulneráveis à indução monetária [Tishler & Bartholomae, 2002; Viens,
2001]. Por outro lado, encontrámos uma correlação inversa entre a importância da compensação financeira como motivação e o grau de escolaridade, mostrando que as pessoas com
maior instrução académica são menos susceptíveis à indução monetária.
De modo interessante, embora não inesperado, mostrou-se que as motivações para a
participação são influenciadas pela personalidade do indivíduo (Figura 4.1).
Auto-selecção
Potencial voluntário
Modéstia
Compensação financeira
Autodisciplina
Progresso da medicina
Competência
Valores
Motivações
Curiosidade
Rectidão
Sensibilidade
Confiança
Check-up gratuito
Voluntário
Assertividade
Figura 4.1. Influência das características de personalidade nas motivações para participação em ensaios
de fase I. [ + = correlação positiva; – = correlação negativa).
A compensação financeira pela participação foi mais valorizada por indivíduos com
pontuação elevada em “modéstia” (ou seja, humildes e pouco preocupados consigo próprios, o que não significa necessariamente que tenham baixa autoestima ou baixa autoconfiança). O progresso da medicina ou ciência foi especialmente valorizado como motivação
por indivíduos com pontuação elevada em “autodisciplina” (capacidade de se motivar para
a prossecução de um objectivo ou iniciar uma tarefa e levá-la a cabo, apesar do enfado e
distracções) e “competência” (sentir-se capaz e bem preparado para lidar com a vida; sentimento de que se é capaz, sensível, prudente e eficaz) e com pontuação baixa em “valores”
(disposição para reexaminar os valores sociais, políticos e religiosos; não conformista, aberto
de espírito, antidogmático).
Os indivíduos predispostos a mostrarem-se motivados pela curiosidade apresentaram
pontuações mais elevadas em “rectidão” (franqueza, frontalidade, sinceridade e naturalidade
108
Discussão e Conclusões
ao lidar com os outros), “sensibilidade” (tendência a ser guiado pelos sentimentos) e “confiança” (tendência a atribuir intenções benévolas aos outros e a considerá-los honestos e
bem intencionados), e pontuações mais baixas em “valores”.
Indivíduos com elevada “assertividade” (dominantes, com ascendente social e força de
vontade, confiantes e decididos) declararam-se menos impressionados pela possibilidade de
realizar uma avaliação clínica gratuita do que os indivíduos pouco assertivos.
2. De um modo algo surpreendente, cerca de 9% dos voluntários afirmaram que aceitariam participar, e outros 53% afirmaram que talvez pudessem considerar participar, mesmo
que não lhes fosse concedida qualquer compensação financeira. Contudo, apesar da resposta ter sido dada com carácter anónimo e confidencial e, portanto, não haver razões objectiváveis condicionantes de uma resposta genuína, não é certo que os indivíduos efectivamente
concretizassem a intenção de participação no caso de não haver compensação financeira.
Aliás, é de admitir que a natureza da resposta possa, ela própria, ter sido condicionada pelas
características individuais ao nível de valores e personalidade. Com esta ressalva, as facetas
da personalidade que apresentaram uma correlação significativa com a declarada disponibilidade para se voluntariarem sem compensação financeira são apresentadas na Figura 4.2.
Neuroticismo
Ansiedade
Obediência ao dever
Depressão
Autodisciplina
Autoconsciência
Deliberação
Impulsividade
Conscienciosidade
Participação sem
compensação financeira
Vulnerabilidade
Confiança
Valores
Sim/Talvez
Figura 4.2. Influência das características de personalidade na disponibilidade declarada para participar num
ensaio clínico de fase I sem compensação económica. [ + = correlação positiva; – = correlação negativa).
O domínio “neuroticismo” avalia a maior ou menor instabilidade emocional, nomeadamente “ansiedade” (tendência para a apreensão, tensão, medo e preocupação), “depressão”
(tendência para experienciar excessivamente afecto depressivo), “autoconsciência” (tendência para sentimentos de inferioridade, vergonha, timidez e ansiedade social), “impulsividade”
(incapacidade de resistir aos impulsos, às tentações) e “vulnerabilidade” (incapacidade de
109
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
lidar com a tensão, facilmente enerváveis). Apesar de, conforme se documenta no Capítulo
III.a, os indivíduos com pontuações elevadas em “neuroticismo” terem tendido a auto-excluírem-se de participação em ensaios clínicos de fase I, o estudo apresentado no Capítulo I
mostra que, entre os participantes, os voluntários com pontuações mais elevadas em “neuroticismo” declararam-se menos motivados pela compensação financeira do que os com
pontuações mais baixas.
No domínio da “conscienciosidade” (que avalia o grau de organização, persistência e
motivação pelo comportamento orientado para um objectivo), verificou-se uma tendência
oposta: a disponibilidade declarada para participar sem compensação financeira aconteceu
sobretudo nos indivíduos com pontuações mais baixas nas facetas “obediência ao dever”
(adesão estrita aos seus padrões de conduta), “autodisciplina” e “deliberação” (tendência a
pensar com cautela, planificar e ponderar antes de agir).
3. Conhece-se muito pouco sobre o processo de decisão de indivíduos saudáveis se
voluntariarem para um ensaio clínico, nomeadamente sobre se consultam terceiros (por
exemplo, a família, amigos ou mesmo um profissional de saúde). Na nossa população, a
percentagem de indivíduos que afirmaram ter pedido opinião a terceiros antes de se voluntariarem para ensaios clínicos de fase I foi elevada (88%) e, apesar da opinião das pessoas
consultadas ser geralmente desfavorável à participação (normalmente, invocando o receio de
complicações graves), mesmo assim os indivíduos decidiram participar. Ou seja, não tomaram em consideração a opinião dos terceiros consultados que recomendaram a não participação. Contudo, resta aqui a dúvida sobre qual a proporção de indivíduos que perguntaram
a opinião a terceiros e, perante uma opinião desfavorável à participação, decidiram segui-la.
É de admitir, aliás, que um “auto-selector” muito importante poderá situar-se a este nível:
pode sugerir-se que, perante uma opinião desfavorável do terceiro a quem se pediu opinião,
a decisão de aceitar ou não a recomendação de não participação no ensaio clínico possa ser
influenciada pelas características do próprio indivíduo.
Autoconsciência
Ordem
Impulsividade
Esforço de realização
Vulnerabilidade
Autodisciplina
Decisão de participar
Figura 4.3. Influência das características de personalidade na decisão de pedir opinião a terceiros antes
de decidir participar num ensaio clínico de fase I. [ + = correlação positiva; – = correlação negativa).
110
Conscienciosidade
Neuroticismo
Consulta da opinião de terceiros
antes de decidir participar
Discussão e Conclusões
Dentro da população de participantes, verificaram-se diferenças nas características de
personalidade entre os que pediram opinião a terceiros e os que não a pediram (Figura 4.3).
Indivíduos com pontuações elevadas em várias facetas do domínio “neuroticismo”, nomeadamente “autoconsciência”, “impulsividade” e “vulnerabilidade”, tenderam a pedir mais vezes
a opinião a terceiros antes de decidir a participação em ensaios clínicos. No domínio da “conscienciosidade”, verificou-se uma tendência oposta: indivíduos com pontuações elevadas nas
facetas “ordem”, “autodisciplina” e “esforço de realização” (níveis de realização elevados e forte
motivação para os atingir; pessoas diligentes, com objectivos e sentido de vida; relacionado com
alguns dos traços do padrão comportamental tipo A) tenderam a não pedir opinião a terceiros.
No que respeita à opinião dos terceiros consultados, ela foi predominantemente desfavorável à participação, sendo o risco de complicações e “fazer o papel de cobaia” as objecções mais frequentes. Isto traduz uma percepção social predominantemente negativa em
relação à participação em ensaios clínicos de fase I, constituindo um selector situacional ou
de contexto relevante e justificando medidas de informação pública sobre o papel fundamental dos ensaios clínicos no desenvolvimento de novos medicamentos e sobre o risco real
(mínimo) que a participação nos mesmos encerra.
Em resumo, o estudo apresentado no Capítulo I, em participantes em ensaios clínicos
de fase I na Unidade de Farmacologia Humana de BIAL, permitiu concluir que a compensação financeira é um factor de motivação muito importante mas não é o único motivador.
Permitiu também perceber que as motivações e percepções em relação à participação são
influenciadas pelas características individuais, nomeadamente os traços de personalidade.
4. O acto de decidir participar num ensaio clínico é uma decisão que deve ser individual,
livre e informada. Como se apresenta no Capítulo II, os valores interpessoais são importantes
na auto-selecção de voluntários para ensaio clínico de fase I. Verificámos que os indivíduos participantes valorizam particularmente o “estímulo” (ser tratado com compreensão e
amabilidade, receber o estímulo das outras pessoas), a “independência” (poder fazer tudo
aquilo que têm vontade de fazer; gosto pela liberdade de acção e por novas experiências) e
a “benevolência” (ser compreensivo com os mais desfavorecidos, partilhar, ajudar os mais
necessitados, ser generoso), e valorizam menos a “conformidade” (fazer o que é aceitável e
correcto, agir sempre em conformidade com as regras, fazer aquilo que é considerado socialmente correcto, agir de forma convencional, ser conformista; valorização da organização) e
a “liderança” (coordenar o trabalho dos outros, estar em posição de dizer aos outros o que
devem fazer, ser seleccionado para uma posição de liderança; valorização do desafio) do que
os seus pares do grupo normativo (Figura 4.4).
Não se detectaram diferenças significativas entre os voluntários e a população normativa no que respeita à valorização do “reconhecimento” (ser benquisto e admirado, ser considerado importante, ser apreciado e reconhecido pelos outros, receber comentários favoráveis a seu respeito).
5. No Capítulo I, verificámos que as motivações e percepções se correlacionam de modo
diferente com as motivações e percepções relativas à participação em ensaios de fase I e os
resultados descritos no Capítulo III.a confirmaram a hipótese de que a população de participantes é auto-seleccionada com base nos traços de personalidade.
111
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Potencial voluntário
Conformidade
Estímulo
Independência
Liderança
Reconhecimento
Benevolência
Voluntário
Figura 4.4. Valorização dos valores interpessoais como factores de auto-selecção para participação num
ensaio clínico de fase I. [ + : selector positivo; – : selector negativo; = : sem diferença entre voluntários
e população normativa].
A aplicação do NEO-PI-R a uma população de participantes em ensaios de fase I revelou
que os mesmos diferiam da população normativa em quase todos os traços de personalidade
avaliados (Figura 4.5).
Os participantes em ensaios de fase I mostraram pontuações mais baixas de “neuroticismo” e pontuações mais elevadas de “extroversão” e “abertura à experiência” do que a
norma. No domínio “neuroticismo”, os participantes apresentaram pontuações mais baixas de “ansiedade”, “hostilidade” (tendência a vivenciar raiva ou estados afins), “depressão”, “autoconsciência” e “vulnerabilidade”, e pontuações mais elevadas de “impulsividade”.
Os participantes apresentaram pontuações mais elevadas em todas as facetas do domínio
“extroversão”: “acolhimento caloroso” (uma pessoa calorosa é amigável, conversadora e
afectuosa; gosta verdadeiramente dos outros), “gregariedade” (gosta do convívio, procura o contacto social, tem muitos amigos), “assertividade”, “actividade” (pessoa enérgica,
necessita de estar ocupada), “procura de excitação” (procura estímulos fortes, aceita riscos) e “emoções positivas” (pessoa alegre, espirituosa, divertida, exuberante). No domínio
“abertura à experiência”, os participantes mostraram pontuações mais elevadas em abertura à “fantasia” (imaginação viva e fantasia activa, sonhar acordado), “sentimentos” (pessoa que responde emocionalmente às situações, é sensível, empática e valoriza os próprios
sentimentos), “acções” (procura a variedade e novidade, novas experiências e novas actividades), “ideias” (pessoa com curiosidade intelectual, orientada teoricamente, aprecia a
argumentação filosófica) e “abertura à experiência estética” (apreço pela arte e beleza).
6. Os resultados da aplicação das escalas SES, CEI-T, STAI, SAD, FNE e BDI-II descritos
nos trabalhos incluídos no Capítulo III, secções b), c) e d) são consistentes com os resultados do NEO-PI-R relatados no Capítulo III.a.
A aplicação do SES e CEI-T ao grupo de participantes em ensaios de fase I e ao grupo
de controlo mostrou pontuações de auto-eficácia percepcionada e de curiosidade mais ele-
112
Discussão e Conclusões
Potencial voluntário
Ansiedade
Acolhimento caloroso
Impulsividade
Vulnerabilidade
Gregariedade
Assertividade
Actividade
Fantasia
Estética
Procura de excitação
Sentimentos
Emoções positivas
Acções
Confiança
Ideias
Rectidão
Valores
Altruísmo
Competência
Complacência
Ordem
Obediência ao dever
Esforço de realização
Modéstia
Extroversão
Hostilidade
Autoconsciência
Amabilidade
Conscienciosidade
Abertura à
Experiência
Neuroticismo
Depressão
Sensibilidade
Autodisciplina
Deliberação
Voluntário
Figura 4.5. Valorização dos traços de personalidade avaliados pelo NEO-PI-R como factores de autoselecção de participação num ensaio clínico de fase I. [ + : selector positivo; – : selector negativo; = : sem
diferença entre voluntários e população normativa].
vadas nos participantes do que na população de comparação (Capítulo III.b). Resultados
coerentes com esses obtiveram-se quando o subgrupo dos sujeitos que declararam potencial
disponibilidade para se voluntariarem para participação num ensaio clínico foi comparado
com o subgrupo dos que recusaram (Capítulo III.d). A aplicação de modelos de regressão
logística mostrou que o componente “iniciativa e persistência” da escala SES e o componente
“tendências exploratórias” da escala CEI-T são selectores positivos de participação em ensaio
clínico (Figura 4.6).
A aplicação das escalas STAI, SAD, FNE e BDI-II (Capítulo III.c) mostrou que as pontuações da ansiedade-traço (STAI-T) e o componente “evitamento e mal-estar psicológico e
social” (SAD) da ansiedade social correlacionaram-se negativamente com a disponibilidade
em participar num ensaio clínico, sugerindo que são auto-selectores negativos (Figura 4.6).
Não se observou correlação entre as pontuações do BDI-II e a disponibilidade para participar, mas verificou-se que os indivíduos com pontuações elevadas no BDI-II tenderam a autoexcluir-se da participação. O componente “medo de avaliação negativa” (FNE) da ansiedade
social não mostrou ser um auto-selector relevante. Os resultados permitiram concluir que
indivíduos tímidos e receosos das interacções sociais bem como os com ansiedade-traço
elevado tendem a auto-excluir-se da participação em ensaios de fase I.
113
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Potencial voluntário
Ansiedade-traço
Iniciativa e persistência
Evitamento e mal-estar
psicológico e social
Tendências exploratórias
Voluntário
Figura 4.6. Traços de personalidade com capacidade preditiva de participação em
ensaios de fase I, entre os avaliados pelas escalas SES, CEI-T, STAI, SAD, FNE e BDIII. [ + : selector positivo; – : selector negativo ].
Numa análise multivariada no grupo de controlo (Capítulo III.d), na qual se incluíram
os resultados de todas as escalas aplicadas, as tendências exploratórias (CEI-T), a iniciativa
e persistência (SES) e a ansiedade-traço (STAI-T) foram os únicos factores determinantes na
resposta ao convite para participar num ensaio clínico de fase I. Os indivíduos que responderam afirmativamente ao convite apresentaram pontuações mais elevadas em tendências
exploratórias e em iniciativa e persistência e pontuações mais baixas em ansiedade-traço do
que os que recusaram.
7. A aplicação de uma escala (Capítulo IV) a um grupo de participantes em ensaios de
fase I e a um grupo de controlo apresentou uma incidência relativamente elevada de queixas
de origem não medicamentosa, muitas das quais constam entre as que são frequentemente
relatadas como efeitos adversos durante ensaios de fase I. As consequências deste facto para
os ensaios clínicos de fase I são óbvias, dado que complicam e complexificam a avaliação de
tolerabilidade durante os mesmos.
O número de queixas foi significativamente mais baixo nos participantes em ensaios de
fase I do que na população de controlo. Quando se testou a relação entre as pontuações das
escalas SES, STAI-T (ansiedade-traço) e BDI-II e a incidência das queixas apresentadas, verificaram-se correlações positivas com o STAI-T e o BDI-II e correlações negativas com o SES. Os
resultados sugerem que os participantes em ensaios de fase I apresentam menor tendência
para relatar efeitos adversos de origem não medicamentosa do que a população geral com
as mesmas características sóciodemográficas, e que tal deve-se, com elevada probabilidade, à auto-selecção de indivíduos com características de personalidade que predispõem ao
baixo relato de queixas, nomeadamente ansiedade-traço baixa e auto-eficácia geral elevada.
114
Discussão e Conclusões
Ou seja, os participantes em ensaios de fase I são uma amostra selectiva de indivíduos com
menor tendência para relatar queixas de origem não medicamentosa do que a população de
controlo e a causa provável é a auto-selecção de indivíduos com nível baixo de ansiedadetraço e nível elevado de auto-eficácia percepcionada.
8. Como se apresenta nas secções a) e c) do Capítulo III, a influência potencial das características de personalidade dos participantes em ensaios de fase I na avaliação de tolerabilidade durante ensaios de fase I foi explorada através do teste da relação entre as pontuações
das avaliações psicométricas utilizadas e a incidência de acontecimentos adversos relatados
durante os ensaios de fase I. Apesar da auto-selecção conduzir a amostras relativamente homogéneas em termos de uma determinada característica de personalidade e esse facto dificultar as análises de correlação relativas a essa característica, foi possível mostrar correlações
significativas entre vários traços de personalidade mensurados pelo NEO-PI-R e a incidência
dos acontecimentos adversos. Indivíduos extrovertidos mostraram relatar menos efeitos adversos do que indivíduos introvertidos. Embora se tenha encontrado uma tendência para
indivíduos ansiosos relatarem mais efeitos adversos, essa tendência não atingiu significância
estatística. Contudo, a falta de significado estatístico pode dever-se ao facto da população
de participantes avaliados pelo NEO-PI-R ser uma amostra relativamente pequena e, particularmente, ao facto de ser uma amostra auto-seleccionada de indivíduos com níveis baixos em
neuroticismo. Ou seja, a sub-representação de indivíduos ansiosos diminui a probabilidade
de se tornarem aparentes potenciais correlações entre o neuroticismo e o relato de efeitos
adversos.
Não se encontraram correlações significativas entre as pontuações das escalas BDI-II,
SAD e FNE e a ocorrência de efeitos adversos. Contudo, apesar da auto-selecção conduzir a
uma população com níveis baixos em ansiedade-traço, verificaram-se correlações positivas,
embora fracas, entre as pontuações do STAI-T e os acontecimentos adversos relatados durante a participação em ensaios clínicos de fase I.
Em termos globais, os resultados são, pois, sugestivos de que a auto-selecção conduz
a uma população com tendência mais baixa para relatar acontecimentos adversos durante
a participação em ensaios clínicos de fase I do que a população normal. Devido ao facto
da auto-selecção se constituir como um factor de confusão na análise, não se podem tirar
conclusões definitivas sobre quais são os traços de personalidade que realmente têm influência relevante no relato de efeitos adversos. Conclusões definitivas exigiriam o desenho de
estudos prospectivos nos quais os indivíduos fossem estratificados de acordo com as suas
características de personalidade e com amostras com poder estatístico adequado. Contudo,
tais estudos parecem difíceis de implementar porque a participação exige o consentimento
informado e os voluntários com certos traços de personalidade (por exemplo, neuroticismo
elevado) tendem a recusar participação. De qualquer modo, o elevado grau de consistência
entre a informação obtida pela correlação entre as características de personalidade e os efeitos adversos durante a participação em ensaio clínico e a informação colhida no estudo das
queixas de origem não medicamentosa relatadas no grupo de controlo e no grupo de parti-
115
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
cipantes são altamente sugestivos de que a auto-selecção de indivíduos com nível baixo de
neuroticismo e, particularmente, reduzida ansiedade-traço, pode ter uma influência marcada
no relato de efeitos adversos em ensaios de fase I, ao conduzir a uma população com menor
tendência para relato de acontecimentos adversos do que a população normal.
9. Entre todas as variáveis sócio-demográficas estudadas, apenas o género demonstrou
ter uma influência na probabilidade de relatar efeitos adversos durante a participação em
ensaios de fase I. As mulheres tenderam a relatar mais efeitos adversos do que os homens.
10. A exploração da possível influência dos factores psicológicos sobre a cinética dos
fármacos estudados nos ensaios de fase I objecto desta dissertação encontra-se relatada no
Capítulo V. Devido ao facto do tamanho das amostras ser relativamente pequeno, a caracterização do comportamento cinético dos fármacos foi efectuada através do desenvolvimento
de um modelo de efeitos mistos baseada numa aproximação paramétrica. Recorreu-se, para
o efeito, ao programa estatístico padrão para este tipo de análise (NONMEM - Nonlinear
Mixed Effects Model).
Pelas razões justificadas no Capítulo V, o nebicapone não foi incluído na análise, dado
que a sua cinética é marcadamente influenciada pelo seu efeito farmacodinâmico. Assim, a
análise centrou-se na cinética dos dois outros fármacos estudados nos ensaios de fase I com
os indivíduos que constituíram a população de participantes desta dissertação: BIA 2-005
(metabolito activo do acetato de eslicarbazepina cuja formação ocorre por simples hidrólise)
e trans-resveratrol.
Devido a marcada sulfatação e glucuronoconjugação durante a primeira passagem, a
biodisponibilidade do trans-resveratrol é muito baixa e a sua cinética é muito variável [Almeida et al., 2008]. Assim, não foi surpreendente a impossibilidade de criar um modelo farmacocinético populacional satisfatório para o trans-resveratrol e, consequentemente, não
se mostrou viável o estudo do efeito da possível influência dos factores psicológicos sobre a
cinética do mesmo.
Todavia, revelou-se viável proceder à análise da influência das variáveis psicométricas
na farmacocinética do BIA 2-005. Como se apresenta no Capítulo V, testou-se a influência
das variáveis psicométricas que revelaram poder discriminante auto-selectivo de participação
em ensaios de fase I nas análises multivariadas apresentadas no Capítulo IV: SES, CEI-T e
STAI. No que respeita ao STAI, para além da ansiedade-traço (avaliada pela escala STAI-T),
procedeu-se também à análise dos resultados da aplicação da ansiedade-estado (avaliada
pela escala STAI-S). Na análise realizada, não se detectou influência das variáveis psicométricas avaliadas pelo SES e CEI-T nos parâmetros farmacocinéticos populacionais estudados (constante de absorção, depuração e volume de distribuição). No que respeita ao STAI,
apesar do facto da auto-selecção ter conduzido a uma população pouco heterogénea em
relação à ansiedade (o que, como já foi referido anteriormente, é uma limitação à análise de
possíveis relações entre essa variável e outras variáveis), foi aparente a influência dos níveis
de ansiedade-estado na constante de absorção (Ka) do acetato de eslicarbazepina (mediante
quantificação do BIA 2-005 em circulação), sem influência dos mesmos na depuração e no
volume de distribuição do BIA 2-005. Ou seja, a ansiedade-estado mostrou ser um factor
preditivo discriminante relativamente à velocidade a que ocorre a absorção.
116
Discussão e Conclusões
A dicotomização em função da mediana dos valores observados de ansiedade-estado (“estar ansioso”) e ansiedade-traço (“ser ansioso”, isto é, predisposto a experienciar excessivamente estados de ansiedade) permitiu verificar que os indivíduos com valores de ansiedade-estado
abaixo da mediana apresentavam um factor multiplicativo de 0,60 relativamente à Ka (o que
corresponde a uma diminuição de 40% na Ka) em relação aos valores totais observados e que,
no caso da variável ansiedade-traço, o factor multiplicativo era de 0,75 (diminuição de 25% na
Ka). Para além desta observação relativamente à associação entre estas variáveis psicométricas
e a velocidade de absorção, merece igualmente destaque o facto da introdução da ansiedadeestado no modelo final apresentar estimadas extremamente precisas e robustas para a generalidade dos parâmetros de efeitos fixos (Ka, CL e Vd) e de efeitos aleatórios (ωKa, ωCL, ωVd e
σ). Este facto revela que o modelo determinístico se encontra estabilizado e, por esse motivo,
o efeito mostrado pela ansiedade-estado na Ka não resulta de um artefacto matemáticoestatístico, mas sim de uma evolução consistente do modelo que explica os dados observacionais.
Há, aparentemente, uma subalternização da influência da ansiedade-traço em relação
à ansiedade-estado; contudo, a ansiedade-traço predispõe a eventos de ansiedade-estado, e
é esta que, através dos fenómenos neuroendócrinos de que se acompanha, potencialmente
interfere na farmacocinética. Por isso, o facto da ansiedade-traço ter revelado menor influência na constante de absorção do que a ansiedade-estado não surpreende, já que se pode
presumir que a ansiedade-traço se manifesta sobretudo através da ansiedade-estado.
Conhece-se que a idade, género, variações circadianas e sazonais, factores dietéticos,
função imunológica, ingestão de álcool, tabagismo, entre outros factores, podem afectar a
avaliação dos resultados farmacocinéticos e farmacodinâmicos [Vessel, 1982]. Existe também alguma evidência de que os traços de personalidade podem afectar a resposta farmacológica, aspecto que Meyer designa de “psicofarmacologia diferencial” e à qual se tem dado
atenção insuficiente [Meyer, 2001]. Indivíduos com traços de padrão comportamental tipo
A eliminam o captopril e absorvem e eliminam o sildenafil de uma forma mais rápida do que
indivíduos de traços de personalidade tipo B [Xu, 2007]. A absorção e a biotransformação
podem também ser afectadas pelos níveis de neuroticismo ou ansiedade, como demonstrado
para o diazepam [Nakano et al., 1980], cafeína [Meyer, 1996], cisapride [Xu, 2007] e teofilina [Hartazema et al., 1982, Jackson et al., 1985].
Indivíduos emocionalmente instáveis absorvem os fármacos mais rápida e completamente do que indivíduos emocionalmente estáveis [Meyer, 2001]. O esvaziamento gástrico
ocorre, geralmente, de um modo mais rápido em indivíduos com elevados níveis de neuroticismo em relação a indivíduos com níveis baixos de neuroticismo [Tishler et al., 2005] e
verificou-se que a ansiedade se acompanha de aceleração do trânsito gastrointestinal [Gorard et al., 1996]. A menor rapidez de absorção do acetato de eslicarbazepina verificada nos
nossos participantes em ensaios de fase I com níveis de ansiedade mais baixos, em relação
aos com níveis de ansiedade mais elevados, pode dever-se a um trânsito gastrointestinal
mais lento e, eventualmente, outras alterações regionais, por exemplo menor fluxo sanguíneo
gastrointestinal. Dado que não se verificou diferença entre a depuração do BIA 2-005 nos
indivíduos com níveis de ansiedade baixos em relação aos indivíduos com níveis de ansiedade
mais elevados, pode concluir-se que a ansiedade tem um efeito na absorção do acetato de
eslicarbazepina mas não afecta a eliminação do seu metabolito activo BIA 2-005.
117
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Em conclusão, indivíduos mais ansiosos apresentaram maior rapidez de absorção do
acetato de eslicarbazepina, apesar do efeito de auto-selecção ter conduzido a uma população de indivíduos pouco ansiosos. Dado que o efeito de auto-selecção nos ensaios de fase I
não é possível de ultrapassar porque a participação depende sempre do consentimento informado e livre de coacção, a avaliação da real dimensão do efeito dos factores psicológicos nas
variáveis farmacocinéticas exigirá o estudo de doentes em rotina terapêutica, porque nesses
doentes provavelmente o viés de auto-selecção não será tão marcado como em voluntários
de fase I.
De qualquer modo, os resultados do nosso estudo parecem encerrar implicações práticas importantes para os ensaios de fase I. Para além do potencial efeito da ansiedade na variabilidade interindividual em estudos de farmacocinética, é de admitir que a ansiedade possa contribuir para a variabilidade intra-individual, isto é, a variação da cinética verificada em
diferentes alturas no mesmo indivíduo, como consequência de variação na ansiedade-estado.
Assim sendo, a caracterização dos níveis de ansiedade dos participantes pode ser importante
para uma adequada interpretação dos resultados de estudos de farmacocinética, incluindo
os estudos de bioequivalência e outros ensaios que adoptem um desenho cruzado e em que
a variabilidade intra-individual pode ser um factor de confusão relevante na análise.
Concluímos assim que:
1. Numa população predominantemente composta por estudantes e jovens licenciados,
não se verificaram diferenças relevantes entre indivíduos que se voluntariaram para ensaio clínico de fase I e outros que recusaram a participação, em qualquer das variáveis
estudadas, incluindo o rendimento mensal.
2. A motivação para participar num ensaio clínico de fase I deriva sobretudo da compensação financeira esperada, a qual se correlaciona inversamente com o rendimento mensal,
mas é também afectada pelo grau académico (sujeitos com maior escolaridade são menos sensíveis à indução financeira). Outras motivações são também valorizadas, nomeadamente a curiosidade e o desejo de contribuição para o progresso da Medicina, sendo
estas mais susceptíveis à influência dos traços de personalidade do que a motivação pela
compensação financeira.
3. A maior parte dos voluntários pede a opinião a terceiras pessoas antes de decidir voluntariar-se para participação num ensaio clínico de fase I. Os traços de personalidade (particularmente, o nível de neuroticismo) influenciam a decisão de consultar outrem, sendo
de admitir que influenciem também a decisão em se voluntariar, perante uma opinião
desfavorável à participação por parte da pessoa consultada.
4. A percepção social sobre o risco relativo à participação num ensaio clínico de fase I é
negativa e desfavorável à participação pelo receio de complicações. Um modo de reduzir
a importância deste selector negativo de participação é promover estratégias de informação pública sobre ensaios clínicos de fase I que minorem a percepção social negativa
sobre os mesmos.
118
Discussão e Conclusões
5. Os valores interpessoais constituem-se como factores de auto-selecção de participação
em ensaios de fase I. Os participantes em ensaios de fase I valorizam a benevolência, generosidade e compreensão, serem tratados com amabilidade e receberem o estímulo das
outras pessoas, poderem proceder conforme sua vontade e terem liberdade de acção e
acesso a novas experiências; tendem a valorizar pouco a conformidade com as normas ou
convenções e, especialmente, a ascensão a posições de liderança.
6. Os traços de personalidade constituem-se como factores de auto-selecção dos participantes. Os voluntários para ensaios de fase I são indivíduos com menor grau de neuroticismo
e mais extrovertidos e abertos a novas experiências do que a população normativa.
7. Queixas que são frequentemente relatadas como efeitos adversos durante ensaios clínicos de fase I são relativamente frequentes em indivíduos da população “saudável” não
participante. A incidência dessas queixas não relacionadas com medicamentos é significativamente mais comum em indivíduos ansiosos e menos frequente em indivíduos com
percepção elevada de auto-eficácia.
8. Os participantes em ensaios clínicos de fase I apresentam uma incidência de queixas não
relacionadas com medicamentos significativamente menor do que aquela verificada nos
indivíduos da população geral com as mesmas características sóciodemográficas. A causa provável dessa diferença é a auto-selecção de indivíduos pouco ansiosos e com elevada
percepção de auto-eficácia para participação nos ensaios de fase I.
9. A incidência de efeitos adversos durante a participação em ensaios de fase I é influenciada
pelos traços de personalidade dos participantes. Participantes mais ansiosos e menos extrovertidos tendem a relatar mais efeitos adversos. Contudo, o processo de auto-selecção
dos participantes conduz a uma população relativamente homogénea em relação às características de personalidade, limitando o poder da análise. Conclusões mais definitivas
podem, eventualmente, ser obtidas em ensaios em voluntários saudáveis estratificados
em função dos traços de personalidade, ou em estudos em doentes em rotina terapêutica, pois é de presumir que nestes o fenómeno de auto-selecção seja menos intenso.
10.Nenhuma das variáveis sóciodemográficas estudadas mostrou um efeito relevante na incidência de efeitos adversos durante a participação em ensaios clínicos de fase I, excepto
o género. Indivíduos do sexo feminino tenderam a relatar mais efeitos adversos do que os
indivíduos do sexo masculino participantes nos mesmos ensaios.
11.As análises exploratórias realizadas entre os factores psicológicos e a cinética dos fármacos estudados foi inconclusiva para o trans-resveratrol. Contudo, no caso do BIA 2-005
(metabolito activo do acetato de eslicarbazepina), verificou-se uma correlação entre a
rapidez de absorção e os níveis de ansiedade, particularmente da ansiedade-estado. Para
além do potencial efeito da ansiedade na variabilidade interindividual, é de admitir que a
ansiedade-estado possa contribuir para a variabilidade intra-individual (isto é, variação
da cinética em diferentes alturas, no mesmo indivíduo), o que pode ter implicações práticas importantes na metodologia de ensaios clínicos de fase I. Conclusões mais definitivas
sobre a dimensão do impacto da ansiedade na farmacocinética podem ser obtidas em
ensaios em voluntários saudáveis estratificados em função dos traços de personalidade,
ou em estudos em doentes em rotina terapêutica, pois é de presumir que em doentes o
fenómeno de auto-selecção seja menos intenso.
119
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
12.Os vários factores que actuam sobre a selecção de participantes para ensaios de fase I
fazem com que as amostras de participantes não se possam considerar representativas
da população geral. Parece, contudo, que uma fonte importante de não representatividade, e consequentemente uma potencial ameaça à validade externa (isto é, do potencial
de generalização dos resultados), reside na auto-selecção dependente do consentimento
informado.
13.O modo de controlar, embora não eliminando, o factor de confusão representado pela
auto-selecção é a estratificação dos auto-selectores mais relevantes e aleatorização dos
voluntários em função da mesma, ou o controlo estatístico dos factores de confusão ao
tempo da análise dos resultados. Em qualquer dos casos, implica que as características
relevantes sejam mensuradas sistematicamente em cada participante em ensaios de fase I.
14.De modo a aumentar o papel dos ensaios de fase I no apoio a decisões de descontinuação de desenvolvimento precoces e no fornecimento de informação que possa gerar
marcadores de resposta ou de tolerabilidade/segurança (no imediato ou retrospectivamente, a prazo) que permitam a estratificação dos ensaios clínicos terapêuticos e o desenvolvimento da medicina personalizada, parece recomendável a recolha do máximo de
informação individual possível durante os ensaios de fase I, nomeadamente informação
psicométrica.
15.Os resultados mostram que a ansiedade é um auto-selector importante e tem potencial
impacto nos resultados dos ensaios, tanto a nível das avaliações de tolerabilidade como
das avaliações farmacocinéticas. Dado que é de fácil implementação, parece-nos defensável a inclusão da avaliação da ansiedade na bateria de avaliações dos ensaios clínicos
de fase I.
120
Resumo, Summary
Resumo, Summary
RESUMO:
No presente trabalho são descritas as características demográficas, de estilo de vida
e socio-económicas dos participantes em ensaios de fase I, em voluntários saudáveis, na
Unidade de Farmacologia Humana de BIAL. Estudou-se também quais são as motivações e
as percepções dos voluntários relativamente à sua participação. Fez-se a caracterização dos
participantes em termos de valores interpessoais e de traços de personalidade, em comparação com a população normativa, através de escalas adaptadas à população portuguesa, de
forma a identificar quais os valores interpessoais e os traços de personalidade que têm potencial influência na auto-selecção dos participantes, como consequência do consentimento
informado. Estudou-se a prevalência de queixas não relacionadas com medicamentos, estudaram-se várias características de personalidade através de escalas psicométricas específicas,
e correlacionaram-se os respectivos resultados com a incidência das queixas. Investigou-se,
ainda, a influência dos traços de personalidade na probabilidade de relato de acontecimentos adversos durante a participação em ensaios de fase I e a influência de factores psicológicos na cinética dos fármacos estudados.
Os resultados mostram que a motivação económica é a principal razão para a participação em ensaios clínicos, mas outras motivações (como a curiosidade e o desejo de contribuir
para o progresso da Medicina) são também valoradas. A motivação económica correlacionase inversamente com o rendimento económico e com o grau de escolaridade. Mostrou-se,
ainda, que as características de personalidade e os valores interpessoais são auto-selectores
fortes na decisão de participação num ensaio clínico. Num grupo de controlo convidado a
participar num ensaio clínico, a intenção declarada de participar não se correlacionou de
modo relevante com qualquer das variáveis sóciodemográficas avaliadas, aparentando depender sobretudo dos traços de personalidade.
Os participantes em ensaios clínicos são uma população de indivíduos mais benevolentes, mais extrovertidos e com menor neuroticismo do que a norma. Como se verificou uma
correlação directa entre a ansiedade e a incidência de queixas não relacionadas com medicamentos, e entre a ansiedade e a incidência de efeitos adversos durante a participação em
ensaios clínicos de fase I, pode concluir-se que os participantes em ensaios clínicos de fase
I são uma população auto-seleccionada de indivíduos com menor predisposição a relatar
efeitos adversos do que a população normativa.
123
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
No caso do fármaco acetato de eslicarbazepina, verificou-se uma correlação directa
entre a rapidez de absorção e os níveis de ansiedade.
De modo a aumentar o papel dos ensaios de fase I no apoio a decisões de descontinuação de desenvolvimento precoces e no fornecimento de informação que possa gerar marcadores de resposta ou de tolerabilidade/segurança (no imediato ou retrospectivamente, a prazo) que permitam a estratificação dos ensaios clínicos terapêuticos e o desenvolvimento da
medicina personalizada, parece recomendável a recolha do máximo de informação individual
possível durante os ensaios de fase I, nomeadamente informação psicométrica. A avaliação
da ansiedade dos participantes em ensaios de fase I é relevante e é de fácil implementação.
124
Resumo, Summary
SUMMARY:
In the present work the demographic, life-style and socioeconomic characteristics of participants in phase I studies at the Human Pharmacology Unit of BIAL are described, together
with their motivations and perceptions regarding participation. Also, phase I participants
were characterised in terms of their interpersonal values and personality traits, in comparison
with the normative population, through validated scales for the Portuguese population, in
order to identify those interpersonal values and personality traits that actually have an influence in the participant’s self-selection mediated by the informed consent. The incidence of
non-drug complaints was studied, several personality and psychological characteristics were
assessed through specific scales, and correlations between personality scores and the incidence of complaints were explored. The influence of personality traits on the probability of
reporting adverse events during participation in phase I studies was also investigated.
The results show that economic rewarding is the main motivation for participation in
phase I studies in most participants, but other motivations (such as curiosity and the desire to
contribute the progress of medicine) are also valued. Economic rewarding motivation showed
to be inversely correlated with the subject’s monthly income and educational degree. Personality traits and interpersonal values showed to be strong self-selectors in the decision to
volunteer to phase I studies participation. In a control group invited to participate in a phase
I study, no influence of any of the socio-economic variables on the willingness to participate
was found, but some personality traits showed to be significant predictors of participation.
Participants in phase I studies are a population of more benevolent, more extraverted
and less anxious subjects than the normative population. Since a direct correlation was demonstrated between anxiety and the incidence of non-drug complaints, and between anxiety
and the incidence of adverse events during participation in phase I studies, it can be concluded that phase I participants are a self-selected population of subjects less prone to report
complaints than the normative population.
For the drug eslicarbazepine acetate, a correlation was found between the absorption
rate and the anxiety levels.
In order to increase the contribution of phase I studies to informed early “go/no go”
decisions during clinical development of new medicines and to the generation of information
that can generate response or tolerability/safety markers that will allow stratification during
125
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
the therapeutic clinical trials and the development of personalised medicine, it appears to be
highly recommendable to collect as much as possible individual information during phase I
studies, including psychometric data. The assessment of the participant’s anxiety appears to
be relevant and easy to implement.
126
Bibliografia
Discussão e Conclusões
Allison M (2008). Is personalized medicine finally arriving? Nat Biotechnol 26:509-517.
Almeida L, Falcão A, Vaz-da-Silva M, Soares E, Costa
R, Loureiro AI, Fernandes-Lopes C, Rocha JF, Nun
es T, Wright L, Soares-da-Silva P (2008). Pharmacokinetic and safety profile of trans-resveratrol in
a rising multiple-dose study in healthy volunteers.
Mol Nutr Food Res (in press).
Anderson JA, Weijer C (2001). The research subject as entrepeneur. Am J Bioethics 1:67-69.
Ayd FJ, Calabresi P (1972). Motivations and rewards
for volunteering to be an experimental subject. Clin
Pharmacol Ther 13:771-781.
Ball CJ, McLaren PM, Morrison PJ (1993). The personality structure of 'normal' volunteers. Br J Clin Pharmacol 36:369-371.
Bandura A (1977). Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev 84:191-215.
Bandura A (1994). Self-efficacy. In Ramachaudran VS
(Ed.) Encyclopedia of Human Behavior. New York: Academic Press, pp. 71-81.
Beck AT, Steer RA, Brown GK (1996). BDI-II: Beck Depression Inventory - second edition manual. San Antonio: The Psychological Corporation.
Beukman TL (2005). Culture values and work-related
values - A theoretical overview. In The effect of selected variables on leadership behaviour within the framework of a transformational organisation paradigm, PhD
Thesis, University of Pretoria.
Bigorra J, Baños JE (1990). Weight of financial reward
in the decision by medical students and experienced healthy volunteers to participate in clinical
trials. Eur J Clin Pharmacol 38:443-446.
Buckman S, Huang SM, Murphy S (2007). Medical
product development and regulatory science for
the 21st century: the critical path vision and its
impact on health care. Clin Pharmacol Ther 81:141144.
Cami J, Llorente M, Farre M, Badenas J-M, Ugena B
(1989). Personality of healthy volunteers participating in phase I clinical trials. Person Individ Diff
10:1199-1200.
Costa PT, McCrae RR (1995). NEO Personality Inventory - Revised. Odessa: Research Psychologists
Press, Inc.
Costa PT, McCrae RR (2000). Inventário de Personalidade NEO Revisto (NEO-PI-R): Manual Profissional. Lisbon: CEGOC-TEA Ltd.
Dickert N, Grady C (1999). What's the price of a research subject? Approaches to payment for research participation. N Engl J Med 341:198-203.
Drici MD, Raybaud F, De Lunardo C, Iacono P, Gustovic P (1995). Influence of the behaviour pattern
on the nocebo response of healthy volunteers. Br J
Clin Pharmacol 39:204-206.
EFPIA (2008). The Pharmaceutical Industry in Figures. European Federation of Pharmaceutical Industries
and Associations (EFPIA), Brussels.
EFPIA/EU (2007). Innovative Medicines Initiative. European Federation of Pharmaceutical Industries and
Associations (EFPIA) and the European Communities represented by the European Commission.
http://imi.europa.eu, acedido em 12 de Agosto de
2008.
Erb K, Staks T, Su C-A (1994). Trends in the remuneration of volunteers. Appl Clin Trials 3:54-58.
Esecover H, Malitz S, Wilkens B (1961). Clinical profiles
of paid normal subjects volunteering for hallucinogen drug studies. Am J Psychiatry 117:910-915.
Farre M, Lamas X, Cami J (1995). Sensation seeking
amongst healthy volunteers participating in phase
I clinical trials. Br J Clin Pharmacol 39:405-409.
FDA (2004). Food and Drug Administration. FDA
Critical Path Initiatives white paper: innovation or stagnation? Challenge and opportunity on the critical path to
new medical products. http://www.fda.gov/oc/initiatives/criticalpath/whitepaper.html, acedido em 12
de Agosto de 2008.
Foernzler D (2008). Effective use of biomarkers in clinical decision making. Biomarkers Congress 2008, 14th
& 15th May, Manchester.
Gaw A (2008). Selecting for success early in drug development. Biomarkers Congress 2008, 14th & 15th
May, Manchester.
129
Caracterização Psico-social de Voluntários Saudáveis em Ensaios Clínicos de Farmacologia
Humana e sua Correlação com os Resultados de Tolerabilidade e Farmacocinética
Gorard DA, Gomborone JE, Libby GW, Farthing MJ
(1996). Intestinal transit in anxiety and depression. Gut 39:551-555.
Gordon LV (1993). Survey of Interpersonal Values, Technical
Manual, third edition. Chicago: Science Research
Associates.
Gordon LV (2001). Inventário de Valores Interpessoais (SIV):
Manual Técnico. Lisbon: CEGOC-TEA Ltd.
Gribbin K, Schaie KW (1976). Monetary incentive, age,
and cognition. Exp Aging Res 2:461-468.
Hartazema AG, Pancorbo S, Davis S (1982). Personality traits and theophylline metabolism. Biopharm
Drug Dispos 3:311-316.
Hassar M, Pocelinko R, Weintraub M, Nelson D, Thomas G, Lasagna L (1977). Free-living volunteer's
motivations and attitudes toward pharmacologic
studies in man. Clin Pharmacol Ther 21:515-519.
Hyun I (2006). Fair payment or undue inducement? Nature 442:629-630.
Jackson SHD, Cooper CAP, Turner TH (1985). Personality traits and theophylline pharmacokinetics. Eur
J Clin Pharmacol 28:429-431.
Kashdan TB (2004). Curiosity: Interest, novelty-seeking, openess to experience. In Peterson C, Seligman MEP (Eds.) Character Strengths and Virtues: A
Handbook and Classification. Oxford: Oxford University Press, pp.125-141.
Kashdan TB, Roberts JE (2004). Social anxiety's impact
on affect, curiosity, and social self-efficacy during
a high self-focus social threat situation. Cogn Ther
Res 28:119-141.
Kola I, Landis J (2004). Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov
3:711-715.
Lazarou J, Pomeranz BH, Corey PN (1998). Incidence
of adverse drug reactions in hospitalized patients:
a meta-analysis of prospective studies. JAMA
279:1200-1205.
Lesko LJ (2007). Personalized medicine: elusive dream
or imminent reality? Clin Pharmacol Ther 81:807816.
Marcus B, Schutz A (2005). Who are the people reluctant to participate in research? Personality correlates of four different types of nonresponse as
inferred from self- and observer ratings. J Personality
73:959-984.
Martin DC, Arnold JD, Zimmerman TF, Richart RH
(1968). Human subjects in clinical research: a report of three studies. N Engl J Med 279:1426-1431.
McGee C (1997). A piece of my mind: subject to payment? JAMA 278:199-200.
Meyer FP (1996). Pharmacokinetic responses to caffeine in volunteers with higher scores for neuroticism. Psychopharmacol 126:275-276.
Meyer FP (2001). Psycho- and immunopharmacological factors relevant to selection of volunteers in
clinical studies. Int J Clin Pharmacol Ther 39:300310.
Meyer FP, Haschke R, Rohl FW (1995). The healthy volunteer in clinical pharmacology: personality and
motivation. Eur J Clin Pharmacol 48:91-96.
Meyer FP, Troger U, Rohl FW (1996). Adverse nondrug
reactions: an update. Clin Pharmacol Ther 60:347352.
Nada A, Somberg J (2007). First-in-man (FIM) clinical
trials post-TeGenero: A review of the impact of the
TeGenero trial on the design, conduct, and ethics
of FIM trials. Am J Ther 14:594–604.
Nakano S, Ogawa N, Kawazu Y (1980). Influence of
neuroticism on oral absorption of diazepam. Clin
Pharmacol Ther 27:370-374.
Novak E, Seckman CE, Stewart RD (1977). Motivations for volunteering as research subjects. J Clin
Pharmacol 17: 365-371.
Ong D, Popat A, Knowles SR, Arrowood JS, Shear NH,
Binkley KE (2004). Objective psychological measurement and clinical assessment of anxiety in
adverse drug reactions. Can J Clin Pharmacol 11:e8e16.
Pieters MS, Jennekens-Schinkel A, Schoemaker HC, Cohen AF (1992). Self-selection for personality variables among healthy volunteers. Br J Clin Pharmacol
33:101-106.
Ribeiro JLP (1995). Adaptação de uma escala de avaliação da auto-eficácia geral. In Almeida L, Ribeiro
I (Eds.) Avaliação Psicológica: formas e contextos.Braga:
APPORT, pp. 163-176.
Rolan P, Atkinson AJ, Lesko LJ (2003). Use of biomarkers from drug discovery through clinical practice:
report of the Ninth European Federation of Pharmaceutical Sciences Conference on Optimizing
Drug Development. Clin Pharmacol Ther 73:284291.
Russel ML, Moralejo DG, Burgess ED (2000). Paying
research subjects: participant's perspectives. J Med
Ethics 26:126-130.
Sherer M, Maddux JE, Mercandante B, Prentice-Dunn
S, Jacobs B, Rogers RW (1982). The Self-Efficacy
Scale: Construction and validation. Psychol Rep
51:663-671.
Spielberger CD, Gorsuch RL, Lushene RE (1970). STAI
manual for the state-trait anxiety inventory. Palo Alto:
Consulting Psychologists Press, Inc.
Tishler CL, Bartholomae S (2002). The recruitment of
normal healthy volunteers: A review of the literature on the use of financial incentives. J Clin Pharmacol 42:365-375.
130
Discussão e Conclusões
Tishler CL, Bartholomae S (2003). Repeat participation among normal healthy research volunteers:
professional guinea pigs in clinical trials? Perspect
Biol Med 46:508-520.
Tishler CL, Bartholomae S, Rhodes AR (2005). Personality profiles of normal healthy research volunteers:
a potential concern for clinical drug trial investigators? Med Hypotheses 65:1-7.
Van Gelderen CEM, Savelkoul TJF, Van Dokkum W,
Meulenbelt J (1993). Motives and perception of
healthy volunteers who participate in experiments.
Eur J Clin Pharmacol 45:15-21.
Vere DW (1991). Payments to healthy volunteers: ethical problems. Br J Clin Pharmacol 32:141-142.
Vessel ES (1982). On the significance of host factors
that affect drug disposition. Clin Pharmacol Ther
31:1-7.
Viens AM (2001). Socio-economic status and inducement to participate. Am J Bioethics 1:69-69.
Vrlovac R, Francetic I, Rotim K (1990). Drug trials on
healthy volunteers in Yugoslavia. Int J Clin Pharmacol
Ther Toxicol 28:375-379.
Watson D, Friend R (1969). Measurement of socialevaluative anxiety. J Consult Clin Psychol 33:448457.
Weijer C (1999). What's the price of a research subject? N Engl J Med 341:206.
Woodcock J (2007). The prospects for "personalized
medicine" in drug development and drug therapy.
Clin Pharmacol Ther 81:164-169.
Woosley RL (2004). Discovering adverse reactions: why
does it take so long? Clin Pharmacol Ther 76:287289.
Woosley RL, Cossman J (2007). Drug development and
the FDA's Critical Path Initiative. Clin Pharmacol
Ther 81:129-133.
Xu F (2007). Effect of personality type on pharmacodynamics through changing pharmacokinetics. Med
Hypotheses 69:1131-1134.
Zink S (2001). Maybe we should pay them more. Am J
Bioethics 1:88-89.
Zuckerman M (1991). Psychobiology of personality. New
York: Cambridge University Press.
131