Monoclonal Anti-keratin (AE1) Reactivity in Routinely Processed Tissue from 166 Human Neoplasms
DOMINIC V. SPAGNOLO, M.B., B.S., SARA A. MICHIE, M.D., GERTRUDE S. CRABTREE, M.D., ROGER A. WARNKE, M.D., AND ROBERT V. ROUSE, M.D.
Department of Pathology, Stanford University Medical Center, Stanford, California
A large number of human neoplasms were tested for their keratin expression in routinely processed tissues by a simple, three-stage immunoperoxidase method using a broadly reactive monoclonal anti-keratin antibody AE1, which recognizes a number of keratin polypeptides distributed in a wide variety of epithelia. All car- cinomas, with the exception of hepatocellular, adrenocortical, and basal cell carcinomas and occasional renal cell, pulmonary small-cell, and pulmonary large-cell anaplastic carcinomas, re- acted with this antibody irrespective of differentiation, in most instances displaying staining of strong or moderate intensity in the majority of tumor cells. Equivocal results were obtained in some seminomas and dysgerminomas. Malignant melanoma, large-cell lymphoma, Hodgkin’s disease, malignant histiocytosis, and stromal mesenchymal elements in all tumors did not show any reactivity with AEI. Even after routine processing, the de- terminant detected by AEl is conserved and restricted to epi- thelial neoplasms. This suggests that AE1 would be valuable in the diagnostic distinction of anaplastic carcinoma from lymphoma and melanoma in routinely processed tissues. (Key words: Monoclonal antibody; Immunohistochemistry; Keratin; Human neoplasms; Carcinoma; Lymphoma) Am J Clin Pathol 1985; 84: 697-704
KERATINS HAVE BEEN SHOWN to be useful markers for the identification of the epithelial nature of neo- plasms. 1-3,7-9,11,12,14.16,17,20,21.24 With few exceptions, 1-3,12,21 publications describing the characterization of anti-keratin antibodies on routinely processed tissues have employed polyclonal heteroanti-sera. Although such studies suggest great usefulness for anti-keratin antibodies as diagnostic tools, the variability in sensitivity and specificity of poly- clonal anti-sera as well as the limited supply of any one preparation restrict the usefulness of anti-keratin staining in diagnostic pathology. Monoclonal antibodies provide the advantage of reproducible sensitivity and specificity as well as unlimited supply. These features make the effort of extensive characterization possible and worthwhile.
Monoclonal antibody AEl has been well characterized by immunoblot analysis as reactive with the subfamily of acidic keratins, members of which are present in a wide
variety of epithelial tissues.6.22.23.26 The histopathologic characterization of AEI to date indicates its potentially useful reactivity with paraffin-embedded tissues. 1-3.21 The purpose of this article is to provide standardized, repro- ducible data on the reactivity of monoclonal antibody AEI relevant to the differential diagnosis of poorly dif- ferentiated carcinoma from other anaplastic neoplasms.
Materials and Methods
Case Selection
One hundred sixty-six cases of various human neo- plasms, nearly all malignant, were selected from the files of the Department of Surgical Pathology at the Stanford University Medical Center. All slides and, where appro- priate, clinical details were reviewed. Only cases in which an unequivocal diagnosis could be made using standard histologic criteria were considered for inclusion in the study.
A large number and wide spectrum of carcinomas were selected to define the range of reactivity of AEI in car- cinomas; these included predominantly poorly differen- tiated neoplasms that nonetheless were diagnosable as carcinomas on histologic grounds alone. Of the cases studied, 15 of 17 from the gastrointestinal tract, 22 of 28 from the respiratory tract, and 16 of 18 from the bladder. prostate, endometrium, cervix, and ovarian surface epi- thelium were poorly differentiated to undifferentiated carcinomas. All melanomas exhibited at least focal pig- mentation. The large-cell lymphomas had previously been studied with the use of panels of monoclonal antibodies in frozen sections; two were T-lineage and the rest were B-lineage tumors. Cases of Hodgkin’s disease showing clusters or sheet-like aggregates of Hodgkin’s cells (so- called “syncytial” pattern) most likely to be confused with carcinoma were selected and included examples of mixed cellularity, cellular phase of nodular sclerosis, and inter- follicular Hodgkin’s disease. As sarcomas do not fre- quently pose significant problems in the differential di- agnosis of large cell anaplastic tumors, we investigated
Received February 1, 1985; received revised manuscript and accepted for publication May 6, 1985.
Supported by PHS Grants Numbers CA-34233 and CA-37083 awarded by the National Cancer Institute, DHHS. Dr. Spagnolo was supported by a W. A. and M. G. Saw Medical Research Fellowship from the Uni- versity of Western Australia.
| Tumor Type | Number Examined | Number Negative with AEI | Total Number Positive with AEI | Extent of staining* in Positive Cases | ||
|---|---|---|---|---|---|---|
| 3+ | 2+ | 1+ | ||||
| Esophagus | ||||||
| Squamous carcinoma | 2 | 0 | 2 | 2 | - | |
| Stomach | ||||||
| Adenocarcinoma | 5 | 0 | 5 | 5 | ||
| Colon | ||||||
| Adenocarcinoma | 5 | 0 | 5 | 5 | ||
| Pancreas | ||||||
| Adenocarcinoma (ductal) | 5 | 0 | 5 | 5 | ||
| Breast | ||||||
| Infiltrating ductal carcinoma | 5 | 0 | 5 | 4 | 1 | |
| Infiltrating lobular carcinoma | 2 | 0 | 2 | 2 | ||
| Medullary carcinoma | 2 | 0 | 2 | 2 | ||
| Paget's disease of nipple | 2 | 0 | 2 | 2 | ||
| Lung | ||||||
| Adenocarcinoma | 5 | 0 | 5 | 3 | 2 | |
| Squamous carcinoma | 5 | 0 | 5 | 3 | 1 | 1 |
| Small-cell carcinoma | 7 | 3 | 4 | 2 | 1 | 1 |
| Large-cell anaplastic | ||||||
| carcinoma | 6 | 1 | 5 | 3 | 1 | 1 |
| Nasopharynx | ||||||
| Nasopharyngeal carcinoma | St | 0 | 5 | 5 | - | |
| Pleura and omentum | ||||||
| Malignant mesothelioma | 3 | 0 | 3 | 3 | ||
| Thymus | ||||||
| Thymoma | 13 | 0 | 13 | 12 | 1 | |
| Skin | ||||||
| Squamous carcinoma | 3 | 0 | 3 | 2 | 1 | |
| Basal cell carcinoma | 2 | 2 | 0 | |||
| Oral mucosa | ||||||
| Squamous carcinoma | İ | 0 | 1 | 1 | ||
| Kidney | ||||||
| Renal cell adenocarcinoma | 5 | 1 | 4 | 3 | 1 | |
| Bladder | ||||||
| Transitional cell carcinoma | 3 | 0 | 3 | 3 | ||
| Prostate | ||||||
| Adenocarcinoma | 3¢ | 0 | 3 | 2 | 1 | |
| Testis | ||||||
| Seminoma | 7 | 6 | 1 | 1§ | ||
| Embryonal carcinoma | 3' | 0 | 3 | 2 | 1 | |
| Endodermal sinus tumor | 1 | 0 | 1 | 1 | ||
| Choriocarcinoma | 1 | 0 | 1 | 1 | ||
| Leydig cell tumor | 1 | 1 | 0 | |||
| Cervix | ||||||
| Squamous carcinoma | 2 | 0 | 2 | 1 | 1 | |
| Endometrium | ||||||
| Endometrial adenocarcinoma | 3 | 0 | 3 | 1 | 1 | 1 |
| Uterine papillary serous (CA) | 1 | 0 | 1 | 1 | ||
| Ovary | ||||||
| Serous carcinoma | 2 | 0 | 2 | 2 | ||
| Endodermal sinus tumor | 1 | 0 | 1 | 1 | ||
| Dysgerminoma | 3 | 1 | 2 ** | 2 | ||
| Thyroid | ||||||
| Papillary carcinoma | 2 | 0 | 2 | 2 | ||
| Follicular carcinoma | 1 | 0 | 1 | 1 | ||
| Anaplastic/follicular | ||||||
| carcinoma | 1 | 0 | 1 | 1 | ||
| Neuroendocrine/neuroepithelial | ||||||
| Medullary carcinoma thyroid | 3 | 0 | 3 | 1 | 2 | |
| Islet cell tumor (insulinoma) | 1 | 0 | 1 | 1 | ||
| Islet cell tumor (malignant) | 1 | 0 | 1 | 1 | ||
| Carcinoid tumor | 2++ | 0 | 2 | 2 | ||
| Merkel cell tumor | 2 | 0 | 2 | 1 | 1 | |
| Pheochromocytoma | 3 | 3 | 0 | |||
| Liver | ||||||
| Hepatocellular carcinoma | 2 | 2 | 0 | |||
| Tumor Type | Number Examined | Number Negative with AEI | Total Number Positive with AE1 | Extent of staining* in Positive Cases | ||
|---|---|---|---|---|---|---|
| 3+ | 2+ | 1+ | ||||
| Adrenal | ||||||
| Adrenocortical carcinoma | 3 | 3 | 0 | - | - | - |
| Malignant melanoma | 7 | 7 | 0 | - | - | - |
| Hodgkin's disease | 5 | 5 | 0 | - | - | - |
| Large-cell lymphoma (diffuse) | 15 | 15 | 0 | - | - | - |
| Malignant histiocytosis | 2 | 2 | 0 | - | - | - |
| Soft tissues | ||||||
| Epithelioid sarcoma | 1 | 0 | 1 | - | 1 | - |
| Epithelioid leiomyosarcoma | 1 | 1 | 0 | - | - | - |
* 3+ indicates >80% tumor cells show positive reaction. irrespective of staining intensity: 2+ indicates 30-80% tumor cells positive: 1+ indicates <30% tumor cells positive. t One case was metastatic to lung and histologically identical to the primary.
# Biopsy of prostate primary and subsequent bone metastasis (formic acid/formalin decal) 3 years later from same patient counted as I specimen. All cells in both stained, although intensity was significantly less in marrow than in primary tumor.
§ In one seminoma, about 20% of the cells demonstrated diffuse cytoplasmic staining: rare cells
only two sarcomas, one an epithelioid sarcoma the other an epithelioid leiomyosarcoma.
The AEI Antibody
The monoclonal antibody AEI (lg G1) was kindly sup- plied by Hybritech (San Diego, CA). It is one of three monoclonal antibodies (AE1, AE2, AE3) prepared against human epidermal keratins, the production and charac- terization of which are fully documented elsewhere.6.22.23.26 The AEI antibody has been shown to react with acidic keratins of four different molecular weights (40, 48, 50/ 50’, and 56.5 Kd) present in a wide variety of human and other mammalian epithelia including simple, stratified, nonkeratinizing, and keratinizing epithelia.
Immunologic Staining
Staining was carried out on 5-um sections of formalin- fixed and paraffin-embedded tissue using a three-stage immunoperoxidase method. One, or occasionally two, representative blocks were cut from each case. Slides were placed in a 60° C oven for two hours before use.
After deparaffinizing and rehydrating the sections, en- dogenous peroxidase activity was blocked using 3% H2O2 in distilled water for 5 minutes. After washing in pH 7.4 phosphate-buffered saline (PBS), sections were placed in a solution of 0.005% protease (Type VII Sigma, St. Louis, MO) for 7 minutes at 37° C (0.05 M TRIS buffer, pH 7.6) and then in 4º C PBS for several minutes. The concen- tration, time, and temperature of protease solution were chosen because they allowed good tissue staining without detaching the tissue section from the slide (unpublished data).
Slides were then stained as described.18 Briefly, slides were sequentially incubated with primary antibody AE1
showed peripherally accentuated staining.
’ One embryonal carcinoma contained a microscopic focus of seminoma in which rare cells showed definite positive staining with peripheral accentuation.
** One dysgerminoma contained rare cells showing definite peripherally accentuated staining. The second case contained scattered cells (<5%) showing strong diffuse cytoplasmic staining. tt One case was metastatic to liver and histologically identical to the primary.
(1:10 dilution of culture supernate), goat anti-mouse Ig (1:15, Tago, Inc., Burlingame, CA), normal swine serum (1:10), and peroxidase conjugated swine anti-goat Ig (1:15, Tago, Inc.). All incubations were for 30 minutes at room temperature, followed by 5-minute washes in PBS. The sections were then sequentially incubated with di- aminobenzidine and copper sulfate, fixed in absolute methanol, and lightly counterstained. As a negative con- trol, AEI was replaced by an unrelated IgG, mouse monoclonal antibody.
Results
A summary of the results appears in Table 1, from which several important observations can be made.
Nearly all carcinomas, irrespective of histologic type or differentiation, stained strongly with AEI. The only carcinomas that did not react with the antibody were all hepatocellular, basal cell, and adrenocortical carcinomas as well as one of five renal cell, three of seven pulmonary small-cell, and one of six pulmonary large-cell anaplastic carcinomas. All other carcinomas originating in many different sites stained with AE1. In most instances, the majority of tumor cells were stained, usually intensely and sharply contrasted to the clean stromal background, a constant feature in all cases. The pattern was quite char- acteristic and appeared as cytoplasmic staining usually accentuated at the cell periphery and also apically where glands were formed. In many cases, staining was heavy and virtually uniformly distributed throughout the entire cytoplasm (Fig. 1A). Moderate or weak staining was ob- served in some carcinoma cells and, clearly, unstained cells could also be observed in varying numbers in some cases (Fig. 1B). Necrosis and poor fixation (as evidenced by poor morphology) seemed to be important factors that
could account for some variability of staining in a few cases.
All pulmonary adenocarcinomas and squamous car- cinomas were reactive with AE1. In general, the majority of cells within each tumor showed strong staining, the lone exception being a moderately differentiated squa- mous carcinoma without any unusual histologic features in which only 10% of the cells stained. Six large cell an- aplastic carcinomas of lung showed a spectrum of staining ranging from completely unreactive (one case) to scat- tered, strongly stained cells (two cases, Fig. 1 B) to all cells strongly stained (three cases). Of seven pulmonary small- cell carcinomas tested, one of one lymphocyte-like type and two of five intermediate-type tumors (one predomi- nantly spindled, Fig. 1C) stained. In one small-cell car- cinoma showing both lymphocyte-like and intermediate- type areas, the majority of cells in both areas stained.
The Merkel cell tumors showed distinctive staining patterns, either diffuse, coarsely granular cytoplasmic staining, or crescentic perinuclear staining. Some of the positive cells in the insulinoma showed similar patterns as well as accentuation of staining at intercellular inter- faces in tumor trabeculae or gland-like formations. In the latter situation, apical accentuation of the staining was also prominent.
The thymomas and nasopharyngeal carcinomas (most were of the “lymphoepithelial” type) also displayed similar intense staining. Staining with the AEI antibody revealed many more tumor cells than could be appreciated in he- matoxylin and eosin (H and E) stained sections alone where single or small clusters of tumor cells were obscured by the dense lymphoid stroma (Fig. ID). A similar situ- ation was noted in diffusely infiltrating (linitis plastica) gastric carcinomas, where single tumor cells insidiously infiltrated the gastric musculature (Fig. 1E).
Two basal cell carcinomas were tested. These were both unreactive, although portions of the overlying epithelium and adjacent adnexa were strongly stained with AE1. All squamous cell carcinomas of the skin and oral mucosa stained.
The renal cell carcinomas showed various patterns of staining, both in tumors predominantly composed of clear cells as well as in those with granular cells. In addition to the peripherally accentuated pattern seen in carcinomas at other sites, there was also variable, coarse granular cy- toplasmic staining. The former pattern seemed to occur mainly in the clear cells and the latter in granular cells, but there was considerable overlap. In one case containing a minor spindle cell component, there was diffuse cyto- plasmic staining of some of the spindled tumor cells (Fig. 1F). Another case was an exclusively oncocytic neoplasm in which there were scattered, strongly stained cells oc- curring singly or in small clusters, cords, or tubules. The one tumor that did not show any reactivity with the an-
tibody was a histologically classic renal cell carcinoma containing both granular and clear cell areas.
All three malignant mesotheliomas, one pleural and biphasic, one pleural and epithelial, the third omental and purely epithelial, displayed intense staining of the epithe- lial areas in a pattern indistinguishable from that already described. In the biphasic tumor, definite staining of many of the spindle cells was present albeit less intense than in the epithelial areas (Fig. 2A).
In the germ-cell tumor group, AEI reactivity was pres- ent in the embryonal carcinomas (including any intra- tubular component). The number of positive cells varied greatly, and staining was focal and generally of moderate or weak intensity. The cytotrophoblastic element of the choriocarcinoma displayed intense staining. Syncytiotro- phoblasts were generally unstained, but weak staining was noted in a few cells. Two endodermal sinus tumors showed strong staining of all cells. In addition, microscopic foci of endodermal sinus tumor in an embryonal carcinoma and in a dysgerminoma (not appreciated on H and E sec- tions initially) showed strong staining (Fig. 2B).
The seminomas and dysgerminomas were more diffi- cult to interpret (see Table 1). Of the ten cases, seven did not show any reactivity with AEI. The other three cases showed diffuse cytoplasmic staining in 5-20% of cells and/ or rarely single cells showing distinct peripherally accen- tuated cellular staining. This latter feature was also noted in an incidental microscopic seminomatous focus in an embryonal carcinoma.
AEI did not react with any nonepithelial neoplasms (Fig. 2C) with the exception of epithelioid sarcoma, a tumor of uncertain histogenesis or differentiation (see Table 1).
Discussion
The AEl antibody is one of a series of three mouse monoclonal anti-keratin antibodies (AE1, AE2, AE3) produced by using human callus keratin as the immu- nogen.6.22.23.26 The antibodies have been extensively char- acterized with the use of the immunoblot technic and shown to be highly specific for keratin filaments with no reactivity with nonepithelial cells. AE1 reacts with the subfamily of relatively acidic (pH = 5.5) keratins of 40, 48, 50/50’, and 56.5 Kd. Although no single one is present in all epithelia, virtually all epithelia tested contain at least one member of this acidic subfamily.22,23 Hence, because of its broad reactivity with a wide variety of epithelia, we felt AEI could be a useful antibody to use in the study of anaplastic tumors.
Our results suggest that AEI would be a useful antibody for this purpose. Virtually all types of carcinomas, apart from hepatocellular, basal cell, and adrenocortical carci- nomas, showed reactivity with this antibody. In nearly all
other carcinomas regardless of site of origin or differen- tiation, in thymomas, and in mesotheliomas, a charac- teristic pattern of strong cytoplasmic staining with accen- tuation around the cell periphery was noted. As most of
the tumors studied were poorly differentiated, this suggests that AEI would be useful in detecting epithelial differ- entiation in anaplastic tumors. Furthermore, consistent results were obtained in these neoplasms with the use of
formalin-fixed and paraffin-embedded material with a sensitive, yet simple, immunoperoxidase detection system coupled with enzyme predigestion.
In some previous studies using formalin-fixed paraffin sections, a narrower spectrum of reactivity with anti-ker- atin antibodies has been noted.5,7,9,11,19,20 Such differences may be due to lack of proteolytic digestion or to the use of polyclonal anti-sera or monoclonal antibodies of dif- ferent specificity.
In the endocrine and neuroepithelial group of neo- plasms, in general, the staining obtained was more focal (i.e., fewer positive tumor cells) and of more variable in- tensity in contrast to the group of squamous carcinomas and adenocarcinomas (Fig. 2D). Therefore, AEI may not be as reliable in cases in which neuroendocrine neoplasms enter into the differential diagnosis. Similarly, the reac- tivity with epithelioid sarcoma, also reported by others,4 indicates that AEI may not be useful in the distinction of this lesion from carcinomas.
One of five renal carcinomas did not display reactivity with AEI. The positivity in this case of adjacent normal tubules and collecting ducts suggests that this lack of reac- tivity is not secondary to processing but instead may be due to variable differentiation of renal cell carcinoma. Variable results have been reported for the reactivity of renal cell carcinomas with polyclonal and monoclonal anti-keratin antibodies.1,7,9,12,14,20,24 Similarly, there are variable observations of keratin staining of basal cell car- cinomas, hepatocellular carcinomas, and normal liver with polyclonal and monoclonal antibodies. 1.8.11,12,23,24
This study on a wide variety of routinely processed tis- sues, using a single, easily reproducible procedure, con- firms the diagnostic utility of the AE1 antibody suggested by the report of its reactivity under a variety of fixation and staining conditions.’ Specific reported reactivities of AEI with routinely processed mesotheliomas1,2 (in con- trast to some results using polyclonal anti-sera5,19) and carcinomas of pulmonary, nasopharyngeal, and breast origin are confirmed1.2,21 (Garcia C: personal communi- cation).
Similar to the recent results of Battifora and associates,3 in a study of germ cell tumors using the AE1 antibody, our three embryonal carcinomas contained detectable keratins, although in a patchy distribution. Comparable results were also obtained in the choriocarcinoma and in the microscopic foci of endodermal sinus tumor noted in two of our cases. In contrast to the apparent unequivocal lack of detectable keratins in the seminomas reported by Battifora and associates,3 we found our results more dif- ficult to interpret. In three of ten seminomas and dysger- minomas, a variable number of cells showed diffuse cy- toplasmic staining, and, rarely, single cells displayed the type of peripherally accentuated staining that typified the
carcinomas. We are unsure how to reconcile these results with the previous study. We repeated the procedure using two other monoclonal anti-keratin antibodies (unpub- lished observations) and obtained the same results with one of these (PPK1), while no reactivity was noted with the other antibody (BD-M324). Study of a larger group of germ cell tumors may be necessary to clarify these re- sults. Based on our results at the present time, we are not confident that AEI reactivity can reliably distinguish beween seminoma and carcinoma with the use of the de- tection system described.
We have found panleukocyte-reactive monoclonal an- tibodies25 and anti-S10010,15 to be useful in the differential diagnosis of lymphomas and melanomas from anaplastic carcinomas on routinely processed tissues.13 A need exists for a similar antibody that reliably reacts only with a broad range of carcinomas in a complementary fashion. The general approach to the testing of antibodies for such use should involve at least two steps of characterization. First, is characterization on large numbers of neoplasms of known diagnosis that fall into the differential diagnosis to be addressed. It is to this first step that the present article is addressed. The second step is characterization on diagnostically difficult cases by correlation with clinical behavior, electron microscopy, or other immunologic studies, all of which are beyond the scope of this study.
In summary, we suggest that the broad range of reac- tivity of AE1 seen in the wide variety of poorly differen- tiated carcinomas studied and lack of reactivity with non- epithelial tumors make it a useful marker of epithelial differentiation in the evaluation of anaplastic tumors. The method we have described works well in formalin-fixed paraffin sections and uses a simple immunoperoxidase system. As AEl is a monoclonal antibody, the results of this characterization should be widely reproducible.
Acknowledgments. The contributions of Linda Ivor of Hybritech for supplying the AEI monoclonal antibody and Dr. T. T. Sun for helpful information are acknowledged.
References
1. Battifora H: Recent progress in the immunohistochemistry of solid tumors. Seminars in Diagnostic Pathology 1984; 1:251-271
2. Battifora H, Kopinski MI: Distinction of mesothelioma and reactive mesothelial cells from adenocarcinoma: an immunohistochemical study. Lab Invest 1984; 50:4A-5A
3. Battifora H, Sheibani K, Tubbs RR, Kopinski MI, Sun T-T: Anti- keratin antibodies in tumor diagnosis. Distinction between sem- inoma and embryonal carcinoma. Cancer 1984; 54:843-848
4. Chase DR, Enzinger FM, Weiss SW, Langloss JM: Keratin in epi- thelioid sarcoma. An immunohistochemical study. Am J Surg Pathol 1984; 8:435-441
5. Corson JM, Pinkus GS: Mesothelioma: profile of keratin proteins and carcinoembryonic antigen. An immunoperoxidase study of 20 cases and comparison with pulmonary adenocarcinomas. Am J Pathol 1982; 108:80-87
6. Eichner R, Bonitz P, Sun T-T: Classification of epidermal keratins
according to their immunoreactivity, isoelectric point, and mode of expression. J Cell Biol 1984; 98:1388-1396
7. Espinoza CG, Azar HA: Immunohistochemical localization of ker- atin-type proteins in epithelial neoplasms. Correlation with elec- tron microscopic findings. Am J Clin Pathol 1982; 78:500-507
8. Gown AM, Vogel AM: Monoclonal antibodies to human inter- mediate filament proteins. II. Distribution of filament proteins in normal human tissues. Am J Pathol 1984; 114:309-321
9. Kahn HJ, Huang S-N. Hanna WM, Baumal R. Phillips MJ: Im- munohistochemical localization of epidermal and Mallory body cytokeratin in undifferentiated epithelial tumors. Comparison with ultrastructural features. Am J Clin Pathol 1984; 81:184- 191
10. Kahn HJ. Marks A, Thom H, Baumal R: Role of antibody to S- 100 protein in diagnostic pathology. Am J Clin Pathol 1983; 79: 341-347
11. Krepler R, Denk H, Artlieb U, Fichtinger E, Davidovits A: Antibodies to intermediate filament proteins as molecular markers in clinical tumor pathology. Differentiation of carcinomas by their reaction with different cytokeratin antibodies. Pathol Res Pract 1982; 175: 212-226
12. Makin CA, Bobrow LG, Bodmer WF: Monoclonal antibody to cy- tokeratin for use in routine histopathology. J Clin Pathol 1984; 37:975-983
13. Michie SA, Spagnolo DV, Dunn KA, Warnke RA, Rouse RV: Sen- sitivity and specificity of antibody panels applied to 120 routinely processed poorly differentiated human neoplasms. (Submitted for publication)
14. Nagle RB, McDaniel KM, Clark VA, Payne CM: The use of antiker- atin antibodies in the diagnosis of human neoplasms. Am J Clin Pathol 1983; 79:458-466
15. Nakajima T, Watanabe S, Sato Y, Kameya T, Hirota T, Shimosato Y: An immunperoxidase study of S-100 protein distribution in normal and neoplastic tissues. Am J Surg Pathol 1982; 6:715- 727
16. Osborn M. Weber K: Tumor diagnosis by intermediate filament typing: a novel tool for surgical pathology. Lab Invest 1983; 48: 372-394
17. Ramaekers F, Huysmans A, Moesker O, et al: Monoclonal antibody to keratin filaments, specific for glandular epithelia and their tu- mors. Use in surgical pathology. Lab Invest 1983; 49:353-361
18. Rouse RV, Warnke RA: Special applications of tissue section im- munologic staining in the characterization of monoclonal anti- bodies and in the study of normal and neoplastic tissues. Hand- book of Experimental Immunology, fourth edition. Edited by D Weir, L Herzenberg, C Blackwell, L Herzenberg. Edinburgh, Blackwell Scientific Publications Ltd (In press)
19. Said JW, Nash G, Tepper G, Banks-Schlegel S: Keratin proteins and carcinoembryonic antigen in lung carcinoma: An immu- noperoxidase study of fifty-four cases, with ultrastructural cor- relations. Hum Pathol 1983: 14:70-76
20. Schlegel R, Banks-Schlegel S, McLeod JA, Pinkus GS: Immuno- peroxidase localization of keratin in human neoplasms. Am J Pathol 1980; 101:41-50
21. Shi S-R, Goodman ML, Bhan AK, Pilch BZ, Chen LB, Sun T-T: Immunohistochemical study of nasopharyngeal carcinoma using monoclonal keratin antibodies. Am J Pathol 1984; 117:53-63
22. Sun T-T, Eichner R, Nelson WG, et al: Keratin classes: molecular markers for different types of epithelial differentiation. J Invest Dermatol 1983; 81:109s-115s
23. Tseng SCG, Jarvinen MJ, Nelson WG, Huang J-W, Woodcock- Mitchell J, Sun T-T: Correlation of specific keratins with different types of epithelial differentiation: monoclonal antibody studies. Cell 1982; 30:361-372
24. Van Muijen GNP, Ruiter DJ, Ponec M, Huiskens-van der Mey C, Warnaar SO: Monoclonal antibodies with different specificities against cytokeratins. An immunohistological study of normal tissues and tumors. Am J Pathol 1984; 114:9-17
25. Warnke RA. Gatter KC, Falini B. et al: Diagnosis of human lym- phoma with monoclonal anti-leukocyte antibodies. N Engl J Med 1983; 309:1275-1281
26. Woodcock-Mitchell J. Eichner R, Nelson WG. Sun T-T: Immu- nolocalization of keratin polypeptides in human epidermis using monoclonal antibodies. J Cell Biol 1982; 95:580-588