Nasopharyngeal Carcinoma - American Journal of Clinical Pathology

Nasopharyngeal Carcinoma: Antikeratin Immunohistochemistry
and Electron Microscopy
JEROME B. TAXY, M.D., DENISE F. HIDVEGI, M.D., AND HECTOR BATTIFORA, M.D.
NASOPHARYNGEAL CARCINOMA (NPC) is a biologically distinct form of head and neck cancer occurring
frequently in Oriental people and for which there are
strong epidemiologic links to prior Epstein-Barr virus
(EBV) exposure and certain HLA antigens.1821'26 In the
United States, this tumor is far less common, representing
less than 7% of all head and neck cancers.5 Using the
WHO classification of NPC,25 the most common histologic variant both in Oriental and Western countries is
undifferentiated carcinoma (UC). 6710 ' 26 which, by description and illustration, corresponds to what many
surgical pathologists recognize as "lymphoepithelioma"26
(Table 1). Even in those studies not using the WHO
classification, the least well-differentiated tumors constitute the majority of the cases. I ' 91219
Perhaps related both to the relative rarity of NPC
Received May 9, 1984; accepted for publication May 31, 1984.
Presented at the International Academy of Pathology, U.S.-Canadian
Division, San Francisco, California, March, 1984.
Supported in part by funding from the Nathan J. Reese Foundation.
Address reprint requests to Dr. Taxy: Department of Pathology,
Lutheran General Hospital, 1775 Dempster Street, Park Ridge, Illinois
60068.
320
Division of Pathology, Lutheran General Hospital, Park
Ridge, Illinois, and the Department of Pathology,
Northwestern University Medical School, Chicago, Illinois,
and the Division of Anatomic Pathology, City of Hope
National Medical Center, Duarte, California
here and in other Western countries and the predominance of the UC type, the histopathologic diagnosis may
be problematic. In one report of 33 cases, an incorrect
initial diagnosis was made in 50% of the cases, lymphoma
being the most frequent primary choice.9 This specific
issue in differential diagnosis has been emphasized recently by a report of four cases in which a diagnosis of
Hodgkin's Disease originally was favored.4 Although
serologic titers of certain EBV antibodies are elevated
in NPC patients, the practical diagnostic utility of these
determinations is uncertain, and presently the resolution
of the diagnosis still depends on tissue examination.
Suggested special tissue studies have included the demonstration of EBV-DNA genomes,18 red blood cell adherence for A, B, and H antigens,10 and various histochemical stains,4 none of which has received much
attention. Electron microscopy has been diagnostically
useful in demonstrating desmosomes with or without
tonofilaments, thereby establishing the epithelial nature
of the tumor cells. 1315
The recent advent of immunohistochemistry and
antikeratin antibody has proven useful in the diagnosis
of a wide spectrum of epithelial tumors,8,16'20'22,23 including NPC. 1415 The staining technics as reported have
varied with respect to fixation versus frozen section,
employing enzymatic predigestion or not, as well as the
length of incubation time of the tissue with the primary
antikeratin antibody. To this list of variables should
now also be added the avidin-biotin complex (ABC)
technic of immunoperoxidase as opposed to immunofluorescence15 or the well-established peroxidase-antiperoxidase (PAP) method. Further, the antisera in the
published reports have been prepared by the respective
authors and therefore have been restricted in their use.
Currently, antikeratin antibodies are commercially
available and potentially widely usable in diagnostic
settings. Given the paucity of published standardized
Downloaded from http://ajcp.oxfordjournals.org/ by guest on October 12, 2016
Eighteen examples of nasopharyngeal carcinoma (NPC), a
tumor with potential diagnostic difficulty, were studied retrospectively. Using the WHO classification, 16 cases were
undifferentiated carcinoma (UC). Immunohistochemistry for
each tumor was performed on paraffin sections using two
commercially available polyclonal antisera and a monoclonal
antibody, AE-I. Method 1 used trypsinization, overnight incubation with the primary antibody and the avidin-biotin
complex (ABC) technic. Method 2 used a 20-minute incubation
with the primary antibody without trypsinization and employed
the peroxidase-antiperoxidase (PAP) technic. Method 2 is the
one most frequently employed by pathologists who use immunohistochemistry as a diagnostic aid. Method 1 gave clear
positive results in each case with antibody AE-I and, in most
cases, with the polyclonal antisera. Electron microscopy in 10
cases demonstrated desmosomes in each case and easily demonstrable tonofilaments in five. The results of this study
indicate that in the diagnosis of UC, the most common variant
of NPC, squamous differentiation can be documented readily
by electron microscopy and immunohistochemistry for keratin
proteins. With the latter, optimization of technic is essential
for reliable results. (Key words: Nasopharyngeal carcinoma;
Lymphoepithelioma; Immunohistochemistry; Desmosomes;
Tonofilaments) Am J Clin Pathol 1985; 83: 320-325
NASOPHARYNGEAL CARCINOMA
Vol. 83 • No. 3
results and technical approaches using these preparations,
we report a retrospective study of 18 cases of NPC
comparing two standard immunoperoxidase methods
(PAP and ABC), using three antibody preparations and
correlating the results with ultrastructural examination.
The antibodies either are already or are soon to be
available to practicing pathologists. We conclude that
immunohistochemistry utilized as described here and
conventional transmission electron microscopy are definitive and sensitive modalities in confirming the diagnosis of NPC.
Materials and Methods
Table 1. WHO Classification of
Nasopharyngeal Carcinoma
1. Squamous cell carcinoma (SCC): Squamous differentiation in the
form of kcratinization and/or intracellular bridges.
2. Nonkcratinizing carcinoma (NKC): Pavement-like or plexiform
growth pattern in which cell borders are well defined; no
squamous differentiation by light microscopy.
3. Undifferentiated carcinoma (UC): Syncytial growth of polygonal,
occasionally spindled cells with vesicular nuclei, prominent
nucleoli, indistinct cell borders, and associated with a lymphoid
stroma.
Tissue for electron microscopy, available in 13 cases,
was fixed in glutaraldehyde and embedded in Epon®.
Screening the thick sections excluded three cases in
which tumor was not recovered. One representative
block was prepared for ultrastructural examination in
each of the remaining ten cases.
Results
Twelve males and six females, ranging from 21 to 65
years of age, constitute the patients in this study. There
were no black patients, but three were Oriental and the
remainder were white. Eight patients were younger than
35 years of age.
Light microscopically there was 1 squamous cell
carcinoma, 1 nonkeratinizing carcinoma, and 16 undifferentiated carcinomas. Immunohistochemically, clear
positive staining, generally uniform in distribution, was
exhibited by the tumor cells with monoclonal antibody
AE-1 in all 18 cases. Approximately three-fourths of the
polyclonal antibody treated sections gave negative reactions. In the one example of squamous cell carcinoma,
one preparation yielded a positive reaction, while the
second did not. Neither polyclonal antibody gave a
positive reaction in the one nonkeratinizing carcinoma.
Following restudy of the cases with the polyclonal
antisera using Method 1, the reactions in general were
much stronger, often approaching the intensity exhibited
by AE-1 (Fig. 1). In general, the background did not
consistently display the clarity observed with the monoclonal antibody. Using Method 1, the one case that was
Zenker fixed failed to react with one polyclonal antibody
but did react with the other.
Table 2. Antikeratin Staining by Immunoperoxidase
Method 1
Method 2
Monoclonal antibody
Trypsinization
Overnight incubation of primary
antibody
ABC technic
Polyclonal antibody
No enzymatic treatment
20-minute incubation of primary
antibody
PAP technic
Downloaded from http://ajcp.oxfordjournals.org/ by guest on October 12, 2016
Only NPC cases in which there was residual tumor
tissue in the paraffin blocks were included in this study.
Eighteen cases (7 nasopharyngeal biopsies and 11 metastatic lymph node excisions) were collected, of which
16 were fixed in 10% buffered neutral formalin and one
each in B5 and Zenker's. Conventional hematoxylin
and eosin stained sections were interpreted according to
the WHO classification (Table 1).
For immunohistochemistry, each case was studied
with two polyclonal antisera prepared in rabbits from
human stratum corneum (Dako Corporation, Santa
Barbara, CA, and Ortho Diagnostics, Raritan, NJ) and
a monoclonal mouse antibody, AE-1 (Hybritech, La
Jolla, CA), also prepared from human calluses by hybridoma technics. AE-1 (AE being an abbreviation for
antiepithelium) recognizes a spectrum of keratin classes
including a 40,000-dalton class that is expressed by
many types of epithelia and epithelial neoplasms, a
50,000-dalton keratin class that is common to all stratified squamous epithelia, and a 56,500-dalton class
highly characteristic of keratinizing epithelium.17 The
specifics of the two methods employed are listed in the
Appendix, but the major differences are summarized in
Table 2. Briefly, Method 1 initially was employed only
with the monoclonal antibody and is an adaptation of
the avidin-biotin-complex (ABC) method of Hsu and
associates." There are two modifications. First, after
quenching endogenous peroxidase activity but prior to
applying nonimmune normal goat serum to remove
nonspecific background staining, the tissue sections were
treated with 0.1% trypsin. Second, the primary antibody
was incubated with the tissue overnight. Method 2, used
initially for both polyclonal antisera, is a standard PAP
method, is suggested by both manufacturers in the
package inserts, and is the method most likely to be
adopted by the users of these products. Enzymatic
treatment is not advised routinely and therefore was not
employed; primary antibody incubations were 20 minutes. Subsequently, the polyclonal antisera were studied
using Method 1.
321
322
TAXY, HIDVEGI, AND BATTIFORA
A.J.C.P. • March 1985
Downloaded from http://ajcp.oxfordjournals.org/ by guest on October 12, 2016
FIG. 1. A (upper, left). Nasopharyngeal biopsy from a 34-year-old man, demonstrating typical island of undifferentiated carcinoma with heavy
lymphoid infiltrate. Hematoxylin and eosin (X400). B (upper, center). Section stained by Method 1 using monoclonal antibody AE-1. Dark linear
bands of positive material within the cytoplasm. Immunoperoxidase (avidin-biotin complex) (X400). C and D (upper and center, right). Sections
stained by Method 2 using each of the polyclonal antisera. No staining observed. Immunoperoxidase (peroxidase-antiperoxidase) (X400). E
(lower, right). Section stained by Method 1 with one of the polyclonal antibodies demonstrating strong positivity. Similar result obtained with
other polyclonal antiserum. Immunoperoxidase (avidin-biotin complex) (X400).
FIG. 2 (lower, left). Undifferentiated carcinoma with tonofilaments and desmosomes (X4,000). Inset. Immunoperoxidase stain using AE-1
antibody, Method 1. Note linear distribution of staining perhaps reflecting distribution of desmosomes. Immunoperoxidase (avidin-biotin
complex) (X400).
NASOPHARYNGEAL CARCINOMA
Vol. 83 • No. 3
The ten cases examined ultrastructurally were all UC.
Each case had easily discernible intercellular junctions;
however, tonofilaments were not identified in two cases
and were found with difficulty in three other cases (Fig.
2). In five cases, tonofilaments were readily apparent.
Discussion
Being a retrospective study, 16 of the 18 cases were
fixed in formalin and embedded in paraffin for periods
ranging from one to eight years. It is an unfortunate
aspect of immunoperoxidase regarding keratin antigens
that formalin is probably the least desirable fixative,
due, at least in part, to the masking effect of aldehyde
linkages inherent in formalin fixation. Paraffin embedding also may contribute to this masking effect.21'22
However, the practice of surgical pathology undoubtedly
will continue to rely on traditional methods of fixation
and embedding, especially when small tissue fragments
are submitted, as is to be expected from nasopharyngeal
biopsies. Technical adjustments can be effective under
these circumstances. Alternatively, the potential deleterious effects may be avoided by using frozen sections
and/or ethanol fixation, which have given technically
superior and more sensitive results. 215-2 '
The key elements in the success of Method 1 were,
we believe, trypsinization and the ABC technic. Enzymatic digestion has been used to reduce aldehyde linkages
and has been suggested by several authors as well as the
manufacturers of the polyclonal antibodies in instances
in which the routine procedure fails to demonstrate
positive keratin staining.3-81516-20'22 We suggest that this
step be made part of routine procedure. Trypsinization
was used here in Method 1, initially with the monoclonal
antibody and later with both polyclonal antisera. The
resulting demonstration of positive staining with the
latter indicates the value of this step.
The two methods employed in this study, i.e., PAP
and ABC, were chosen because they are the two most
widely used by practicing pathologists. In this study,
however, neither AE-1 nor trypsinization was used with
Method 2. The reasons for this decision were based first,
on the empiric observation that ABC is a method in
which the sensitivity of the antibody is greatly enhanced,''
second, this method is also best suited for use with
monoclonal antibodies. Since monoclonal antibodies
are usually of murine origin, their use with the PAP
technic would require mouse PAP, an expensive and
not readily obtainable reagent. When Method 1 employing ABC was used with the polyclonal antisera, the
quality and sensitivity of the reactions was enhanced
maximally. This was especially apparent in the majority
of cases, which were initially negative with Method 2.
The consistent positive reactions with striking background clarity with the monoclonal antibody resulted
in slides that were interpreted more easily. This was the
chief advantage of this preparation over the polyclonal
antibodies. Nevertheless, the general enhancement of
the reactions of the polyclonal antibodies by Method 1
indicates that they may be applied effectively in the
diagnosis of UC.
Incubation times with the primary antibody have
varied, with 20 minutes, recommended by the manufacturer of the polyclonal antibodies, and 30 minutes, 2
hours, or overnight times being reported.816'20-23-24 Testing
for the monoclonal antibody used in this study has
shown that overnight incubation will ensure optimal
sensitivity. Shorter incubation times, e.g., 60 minutes,
can of course be tried if duplicate slides are run. At the
end of a 60-minute incubation, one slide can be stained
and viewed. If positive, the overnight incubation can be
discontinued.
The present report demonstrates that, despite the
undifferentiated histologic features of most cases of
NPC, the early demonstrable expression of the ultrastructural and immunohistochemical development of
squamous features is quite advanced. Electron microscopy of this tumor is an established modality in demonstrating desmosomes with or without intracytoplasmic
tonofilaments.13 The ten cases here, each with desmosomes and five with easily recognized tonofilaments,
support this observation. Since only one representative
block of tumor was cut for examination in the electron
microscope, the results suggest that sampling for ultrastructural study is not an issue in NPC provided that
neoplastic cells are present on the thick sections. Furthermore, UC is one of the few instances in which the
electron microscopic examination is quick and definitive.
If obvious desmosomes are not readily apparent, the
Downloaded from http://ajcp.oxfordjournals.org/ by guest on October 12, 2016
Lymphoepithelioma is a term well recognized by
surgeons and pathologists in the diagnosis of NPC. The
histologic picture it connotes, e.g., & poorly differentiated
or undifferentiated carcinoma with a prominent lymphoid infiltrate, is sufficiently distinctive to have been
adapted for poorly differentiated squamous cell carcinomas primary in the thymus, the most common variant
of thymic carcinoma, which is histologically identical
with its NPC counterpart.27 In the nasopharynx, UC is
basically a squamous carcinoma and must be distinguished from varieties of lymphoma, although occasionally melanoma or rhabdomyosarcoma may be considered. The results of this study emphasize the important
and complementary ancillary roles of immunohistochemistry and electron microscopy in resolving the
differential diagnosis.
323
A.J.C.P. • March 1985
TAXY, HIDVEGI, AND BATTIFORA
324
Acknowledgments. The authors acknowledge the technical assistance
of Abbyann Sisk and Mary Kopinski and the clerical assistance of
Susan Diederich.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
References
1. Baker SR, Wolfe RA: Prognostic factors of nasopharyngeal malignancy. Cancer 1982; 49:163-169
2. Battifora H, Sheibani K, Tubbs RR, Kopinski MI, Sun TT:
Antikeratin antibodies in tumor diagnosis: Distribution between
seminoma and embryonal carcinoma. Cancer 1984; 54:843848
3. Bourne JA: Handbook of immunoperoxidase staining methods.
Santa Barbara, Dako Corporation, 1983
4. Carbone A, Micheau C: Pitfalls in microscopic diagnosis of
undifferentiated carcinoma of nasopharyngeal type (lymphoepithelioma). Cancer 1982; 50:1344-1351
5. Decker J, Goldstein JC: Current concepts in otolaryngology: Risk
factors in head and neck cancer. N Engl J Med 1982; 306:11511155
6. Easton JM, Levine PH, Connelly RR, Day NE: Studies on
nasopharyngeal carcinoma in the United States: A model for
international comparison. Comp Immun Microbiol Infect Dis,
vol 2. Great Britain, Pergamon Press, Ltd., 1979, pp 221-228
7. Easton JM, Levine PH, Hyams VJ: Nasopharyngeal carcinoma in
the United States: A pathologic study of 177 U.S. and 30
foreign cases. Arch Otolaryngol 1980; 106:88-91
8. Espinoza CG, Azar HA: Immunohistochemical localization of
keratin-type proteins in epithelial neoplasms: Correlation with
23.
24.
25.
26.
27.
electron microscopic findings. Am J Clin Pathol 1982; 78:500507 _
Giffle'r RF, Gillespie JJ, Ayala AG, Newland JR: Lymphepithelioma
in cervical lymph nodes of children and young adults. Am J
Surg Pathol 1977; 1:293-302
Gupta RK: Isoantigens, A, B, H(O) in nasopharyngeal carcinoma.
Cancer 1983;51:256-259
Hsu SM, Rami L, Fanger H: Use of avidin-biotin peroxidase
complex (ABC) in immunoperoxidase techniques: A comparison
between ABC and unlabeled antibody (PAP) procedures. J
Histochem Cytochem 1981;29:577-580
Jenkin RDT, Anderson JR, Jereb B, et al.: Nasopharyngeal
carcinoma: A retrospective review of patients less than thirty
years of age: A report from children's cancer study group.
Cancer 1981;47:360-366
Lin HS, Liu CS, Yeh S, Tu SM: Fine structure of nasopharyngeal
carcinoma with special reference to the anaplastic type. Cancer
1969; 23:390-405
Madri J A, Barwick KW: An immunohistochemical study of nasopharyngeal neoplasms using keratin antibodies: Epithelial vs.
nonepithelial neoplasms. Am J Surg Pathol 1982; 6:143-149
Miettinen M, Lehto VP, Virtanen I: Nasopharyngeal lymphoepithelioma: Histological diagnosis as aided by immunohistochemical demonstration of keratin. Virchows Arch Cell Pathol 1982;
40:163-169
Nagle RB, McDaniel KM, Clark VA, Payne CM: The use of
antikeratin antibodies in the diagnosis of human neoplasm.
Am J Clin Pathol 1983; 79:458-466
Nelson WG, Sun TT: The 50 and 58-kdalton keratin classes as
molecular markers for stratified squamous epithelia: Cell culture
studies. J Cell Biol 1983; 97:244-251
Pearson GR, Weiland LH, Neel HB III, et al: Application of
Epstein-Barr virus (EBV) serology to the diagnosis of North
American nasopharyngeal carcinoma. Cancer 1983; 51:260268
Pick T, Mauer HM, McWilliams NB: Lymphoepithelioma in
childhood. J Pediatr 1974; 84:96-100
Ramaekers F, Puts J, Moesker O, Kant A, Jap P, Vooijs P:
Demonstration of keratin in human adenocarcinomas. Am J
Pathol 1983; 111:213-223
Ringborg V, Henle W, Henle G, et al: Epstein-Barr virus specific
serodiagnostic tests in carcinomas of the head and neck. Cancer
1983;52:1237-1243
Said JW: Immunohistochemical localization of keratin proteins in
tumor diagnosis. Hum Pathol 1983; 14:1017-1019
Schlegel R, Banks-Schlegel S, McLeod JA, Pinkus GS: Immunoperoxidase localization of keratin in human neoplasm. Am J
Pathol 1980; 101:41-50
Schlegel R, Banks-Schlegel S, Pinkus G: Immunohistochemical
localization of keratin in normal human tissues. Lab Invest
1980; 42:91-96
Shanmugaratnam K, Sobin LH: Histological typing of upper
respiratory tract tumors, International histological classification
of tumors, No 19, WHO, Geneva, 1978, p 32-33
Shanmugaratnam K, Chan SH, de-The G, et al: Histopathology
of nasopharyngeal carcinoma: Correlations with epidemiology,
survival rates and other biological characteristics. Cancer 1979;
44:1029-1044
Wick MR, Scheithauer BW, Weiland LH, Bernatz PE: Primary
thymic carcinomas. Am J Surg Pathol 1982; 6:613-630
APPENDIX
Method
1
1. Cut paraffin section 4 - 6 /urn thick and apply to glue
covered slides (5% Elmer's® glue). Dry at 60 °C overnight.
2. Deparaffinize in two changes of xylene (5 minutes each)
and two changes of 100% ethanol (ten dips each).
Downloaded from http://ajcp.oxfordjournals.org/ by guest on October 12, 2016
tumor is most likely something other than UC. Lastly,
and perhaps most important for pathologists attempting
to correlate the ultrastructure with the immunohistochemistry, the absence of tonofilaments does not preclude
positive immunologic staining for keratin.
The success of immunohistochemistry in this or any
tumor may be related to the nature of the keratin
antigen(s) being expressed as well as the technics employed to stain it (them).22 Although the exact identity
of the antigen(s) being stained in UC is not known, it
has been established that AE-1 antibody recognizes three
keratin classes: a 40,000-dalton class present in all
epithelia except epidermis, 50,000-dalton class common
to all stratified epithelia, and a 56,000-dalton class
associated with keratinizing epithelium.17 The polyclonal
preparations are, as this study shows, also capable of
recognizing one or more of these antigens. However,
both polyclonal and monoclonal antibodies require the
technical modifications as outlined in Method 1. In
most examples of this tumor, it is possible that one or
more of these keratin classes are more readily polymerized as aggregated so that they are ultrastructurally
recognizable as tonofilaments. From the practical viewpoint, antikeratin staining and electron microscopy can
provide definitive information in the diagnosis of NPC.
Immunoperoxidase is cheaper and requires no special
equipment. Both modalities are widely available, are
technically easy, are relatively rapid, and provide permanent records of examination.
Vol. 83 • No. 3
NASOPHARYNGEAL CARCINOMA
325
Method 2
1. Cut sections, as in Method 1.
2. Deparaffinize and hydrate to distilled H 2 0.
3. Incubate slides with 3% H 2 0 2 in methanol, 5 minutes.
4. PBS, 5 minutes.
5. Normal sheep serum (1:5), 20 minutes.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Antikeratin (1:150), 20 minutes.
PBS, 5 minutes.
Swine anti-rabbit globulin (1:100), 20 minutes.
PBS, 5 minutes.
PAP (1:100), 20 minutes.
PBS, 5 minutes.
DAB chromogen, 10 minutes.
H2Q rinse.
Counterstain, Mayer's Hematoxylin 10 minutes.
Wash in tap H 2 0, 10 minutes.
16. Coverslip.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on October 12, 2016
3. Incubate slides with 6% H 2 0 2 in methanol, room temperature, 30 minutes.
4. Rinse in distilled water and follow by wash in PBS
(phosphate-buffered saline, pH 7.3), 5 minutes.
5. Trypsinize (0.1% trypsin in PBS), 37 °C for 60 minutes.
6. Wash with distilled H 2 0-PBS. ,
7. Normal goat serum (1:20), 20 minutes.
8. Antikeratin (1:300), room temperature, overnight.
9. PBS, 5 minutes.
10. Biotinylated horse anti-mouse (1:1000), room temperature,
60 minutes.
11. PBS, 5 minutes.
12. ABC, room temperature, 60 minutes (Vectastain® ABC
kit, Vector Laboratories, Burlingame, CA).
13. PBS, 5 minutes.
14. DAB (diaminobenzidine) chromogen, 10 minutes.
15. PBS, 5 minutes.
16. Counterstain with Mayer's Hematoxylin, 10 minutes.
17. Wash in tap water, coverslip.