Immunohistologic Study of Ulcerative Colitis With Monoclonal Antibodies Against Tumor-Associated and/or Differentiation Antigens

Transcription

1 GASTROENTEROLOGY 1988;95: Immunohistologic Study of Ulcerative Colitis With Monoclonal Antibodies Against Tumor-Associated and/or Differentiation Antigens HARRY S. COOPER and ZENON STEPLEWSKI Department of Pathology and Cell Biology, Jefferson Medical College and Thomas Jefferson University Hospital; The Wistar Institute; and Comprehensive Rectal Cancer Center, Thomas Jefferson University, Philadelphia, Pennsylvania We used monoclonal antibodies (MAbs) (19-9, 55-2, and 73-3) that detect tumor-associated or differentiation antigens, or both, to immunohistochemically study a well-defined group of patients with ulcerative colitis. Monoclonal antibody 19-9 detects the gastrointestinal cancer-associated antigen (sialylated Lewis A). Monoclonal antibody 55-2 detects the Lewis Y antigen and reacts with deep crypt cells in the normal colon. In the normal colon MAb 73-3 reacts with mature superficial columnar cells detecting the protein moiety of a 35,000-dalton glycoprotein. In cases of inactive or mildly active disease, MAbs 19-9, 55-2, and 73-3 had staining patterns similar to normal colon. In 72% and 44% of cases of severely active disease, MAb 19-9 and MAb 73-3, respectively, reacted with epithelial cells at all levels of the crypt, whereas MAb 55-2 reacted only with deep crypt cells. Monoclonal antibodies 19-9, 55-2, and 73-3 reacted with dysplastic epithelium in 70%, 10%, and 60% of cases, respectively. In severely active disease, proliferating epithelial cells "paradoxically" express markers of differentiated epithelium throughout the entire crypt. Similarly, colonic epithelial cells may have the ability to reversibly express tumor-associated antigens. Unfortunately, the MAbs used in this study cannot differentiate dysplasia from reactive epithelium. Ulcerative colitis (UC) is an inflammatory disorder of the large intestine in which there is known perturbation of epithelial cell kinetics and in which there is an increased risk for the development of carcinoma of the large intestine (1-5). A histologic spectrum is observed in UC ranging from inactive disease (quiescent) to mildly active disease, severely active disease with marked regenerative changes, and dysplasia. Dysplasia is a neoplastic lesion and is thought to be a marker for the development of cancer in the patient with UC (4). At times, it may be difficult to differentiate florid reactive epithelial changes from true dysplasia with the light microscope. Immunohistochemical studies with tumorassociated antigens and lectins have not been able to distinguish between dysplasia and reactive changes (6-9). In fact, these studies have noted that nonneoplastic epithelium from patients with UC will express both tumor-associated antigen and lectin binding patterns seen in dysplasia or colonic carcinoma. Similarly, studies using mucin histochemistry have been unrewarding and nonspecific (9-11). In an effort to learn more about the biology of normal and neoplastic colonic epithelium and to discern a possible marker for dysplasia, we chose to immunohistochemically study a well-defined group of patients with UC [i.e., those with (a) inactive or mildly active disease, (b) severely active disease, and (c) dysplasia] with three monoclonal antibodies (MAbs) against tumor-associated antigens or differentiation antigens, or both. The three MAbs we chose to use are (a) 19-9, which detects the gastrointestinal carcinoma-associated antigen (GICA), whose biochemical determinant is a monosialylated-lewis A structure (12,13); (b) 55-2, which detects a small population of epithelial cells limited to the deep crypts ("immature epithelium") and whose determinant is the Lewis Y antigen (a difucosylated type 2 blood group backbone) (i4-16); and (c) 73-3, which reacts with the surface or superficial mature colum- Abbreviations used in this paper: GICA, gastrointestinal carcinoma-associated antigen; MAb, monoclonal antibody; UC, ulcerative colitis by the American Gastroenterological Association /88/$3.50

2 September 1988 MONOCLONAL ANTIBODIES AND ULCERATIVE COLITIS 687 nar cells and whose determinant is the protein moiety of a 35,000-dalton glycoprotein (17). Materials and Methods Materials The study materials were obtained by a retrospective review of cases indexed as UC in the surgical pathology files of the Thomas Jefferson University Hospital. All cases (slides) were re-reviewed for correctness of diagnosis. Only those cases with congruent histology and clinical history (from the pathology requisition slip filled out by the submitting clinician) were considered useable for this study. The study group consisted of 50 tissue sections from 43 patients with a diagnosis of UC (37 resection specimens and 13 biopsy specimens). These UC cases were divided into the following three groups based on congruent histologic findings and clinical history as noted by the submitting clinician: (a) inactive or mildly active disease, or both; (b) severely active disease; and (c) dysplasia. Those cases with inactive or mildly active disease histologically had absent or mild acute inflammation, respectively. Severe disease histologically showed extensive acute and chronic inflammation. There were areas of ulceration, inflammatory pseudopolyps, and reactive or regenerative epithelium, or both. The criteria of Riddell et al. (4) were used to define those cases with dysplasia. Fifteen colon specimens resected for diverticular disease were used as normal controls. Sections from these controls were chosen from blocks (slides) that were devoid of inflammatory changes. Twenty cases of rectal adenocarcinoma were used as "neoplastic" controls. All sections examined in this study were from the descending colon, sigmoid colon, and rectum. All three MAbs are well characterized and are known to react extensively with colorectal carcinoma (12-17). In this study, we used MAbs from mouse ascites fluid. Figure 1. Control colon stained with MAb There is no staining of the epithelium with MAb Magnification, X100. Ulcerative colitis. There were 15 cases of DC with inactive or mildly active disease. Fourteen (93%) of these cases reacted with MAb 19-9 similar to controls (Figure 2), and 1 case (7%) expressed MAb 19-9 in both superficial and basal goblet cells (Table 1). Eighteen of 25 (72%) cases of severely active DC showed extensive staining with MAb 19-9 similar to carcinomas. In these cases, staining was noted throughout the entire length of the crypt (Figure 3). Seven of 25 cases (28%) of severely active DC reacted with MAb 19-9 similar to controls (1 case negative and 6 cases with focal staining of superficial epithelial cells) (Table 1). Dysplasia. There were 10 cases with dysplasia. Seven of these (70%) showed a staining with Methods The avidin-biotin complex was used to detect reactivity of all three MAbs. The methods were identical to those of our previous work (18), however, mouse ascites fluid (Cappel Laboratories, Cochranville, Pa.] was also used as a control. Results Monoclonal Antibody 19-9 [Gastrointestinal Carcinoma-Associated Antigen) Controls and cancers. Ten of 15 controls (67%) showed no reactivity with MAb The remaining controls (33%) expressed staining with MAb 19-9 limited to the surface crypt cells (Figure 1, Table 1). Eighteen of 20 (90%) adenocarcinomas stained with MAb The staining patterns of the cancers were glycocalyx, apical cytoplasm, and total cytoplasm (Table 1). Figure 2. Inactive UC stained with MAb There is marked crypt distortion. There is no staining with MAb Magnification, x 100.

3 688 COOPER AND STEPLEWSKI GASTROENTEROLOGY Vol. 95, No.3 Table 1. Expression of Monoclonal Antibody 19-9 and Its Distribution Within the Crypt Expression of MAb and staining loc ation Inactive DC and/or within crypt Control mildly acti ve DC Severely active DC Dysplasia Canc ers " Negative 10 (67%) 11 (73%) 1 (4%) 3 (30%) 2 (10%) Base of cryp t 0(0%) 0 (0%) 0(0%) 0(0%) NA Surface of crypt 5 (33%) 3 (20%) 6 (24%) 0 (0%) NA Surface and bas e 0 (0%) 1 (7%) 18 (72%) 7 (70%) NA Positi ve NA NA NA NA 18 (90%) Total No. 15 (100%) 15 (100%) 25 (100%) 10 (100%) 20 (100%) MAb, monoclonal antibody; NA, not applica ble; DC, ulcerati ve colitis. a Can cers are evaluated only as positive or negati ve for monoclona l antibody MAb 19-9 similar to cancers (Figure 4), and 3 cases (30 %) failed to stain with MAb One of our negative dysplasia cases also had a carcinoma that stained with MAb 19-9 (Table 1). Monoclonal Antibody 55-2 (Lewis Y Blood Group Antigen) Controls and cancers. Six of 15 (40%) controls showed no reactivity with MAb Eight of 15 controls (53%) expressed the Lewis Y antigen focally among the apical cytoplasm or glycocalyx of basal crypt cells (Figure 5), and 1 control (7%) stained w ith MAb 55-2 in both basal and surface columnar cells (Table 2). Nineteen of 20 (95 %) cancers expressed the Lewis Y antigen with staining observed in the glycoc alyx, apical cytoplasm, total cytoplasm, and cell membrane (Table 2). Ulcerative colitis. In 14 of 15 (93 %) cases of inactive or mildly active DC the staining reaction with MAb 55-2 was similar to controls (Figure 6), One case (7%) showed focal staining of superficial columnar cells. Twenty-one of 25 (84 %) cas es of severely active DC showed staining similar to controls (Figure 7), and 4 cases (16 % ) stained focally with MAb 55-2 in the supranuclear region of superficial columnar cells and glycocalyx of basal crypt cells (Table 2). Dysplasia. Onl y 1 of 10 (10%) cases of dysplasia showed a staining pattern with MAb 55-2 similar to cancers. In this case, there was staining of the total cytoplasm throughout most of the crypt. Nine of 10 (90%) cases of dysplasia stained with MAb 55-2 similar to controls. Two cases of dysplasia with a " control" staining pattern for MAb 55-2 had associated cancers that both extensively stained for MAb 55-2 (Table 2). Monoclonal Antibody 73-3 Controls and can cers. All 15 control colons (100%) reacted extensively with MAb 73-3, with staininglocalizedto the surface or superficial colum- Fi gure 3. Severel y active DC stained with Mab The epithe lium is of a regenerative nature an d extensively stains with MAb 19-9, similar to cancers and dysplasia. Magnification, x 100. Figure 4. Dysplastic epithelium stained w ith MAb There is int ensive staining of epithelium at all levels of the crypt. Magn ification, x 100.

4 September 1988 MONOCLONAL ANTIBODIES AND ULCERATIVE COLITIS 689 Figure 5. Control colon stained with MAb The staining is limited only to the glycocalyx or apical cytoplasm of the deep crypt epithelial cells (arrows). The rest of the epithelium fails to stain with MAb Magnification, Xl00. Figure 6. Mildly active UC stained with MAb The staining with MAb 55-2 is limited to the base of crypts (arrows), similar to control colon (Figure 5). Magnification, Xl00. nar cells, or both. The staining pattern was mainly along the lateral cell membrane; however, occasional staining of the apical cytoplasm and supranuclear and subnuclear region of columnar cells was also noted (Figure 8). Five cases (33%), besides showing extensive lateral membrane staining of mature surface columnar cells, also showed focal staining of the subnuclear portion of basal crypt goblet cells. In all instances, the membrane staining pattern was limited to only the mature surface columnar cells (Table 3). Thirteen of 20 cancers (65%) stained with MAb The staining pattern was cell membrane, apical cytoplasmic, supranuclear, and subnuclear. In the vast majority of cancers that reacted with MAb 73-3, the staining was usually only focal and much less intense than seen in controls (Table 3). Ulcerative colitis. All 15 cases (100%) of inactive or mildly active DC (Figure 9) and 14 of 25 (56%) severely active DC cases reacted with MAb 73-3 with a membranous staining pattern similar to controls. Eleven of 25 cases (44%) of severely active DC showed staining with MAb 73-3 in a membrane pattern throughout the entire length of the crypt. The membrane staining pattern correlated with columnar cell differentiation and was distinctly different from the subnuclear pattern occasionally seen in the basal crypt goblet cells of controls (Figure 10) (Table 3). Dysplasia. Six of 10 cases (60%) of dysplasia stained with MAb 73-3 at all levels of the crypt. The staining patterns were membrane, apical cytoplasm, subnuclear, and total cytoplasm. The membrane patterns could be correlated with columnar cell differentiation. Four (40%) of the dysplasia cases reacted with MAb 73-3 similar to control colons (Table 3). Discussion In this study we have examined a welldefined group of DC patients with three MAbs against tumor-associated antigens or differentiation antigens, or both, and compared their expression with cancers and controls. Our results show that Table 2. Expression of Monoclonal Antibody 55-2 and Its Distribution Within the Crypt Expression of MAb and staining location Inactive UC and/or within crypt Control mildly active UC Severely active UC Dysplasia Cancers" Negative 6 (40%) 6 (40%) 12 (48%) 9 (90%) 1 (5%) Base of crypt 8 (53%) 8 (53%) 9 (36%) 0(0%) NA Surface of crypt 0(0%) 0(0%) 0(0%) 0(0%) NA Surface and base 1 (7%) 1 (7%) 4 (16%) 1 (10%) NA Positive NA NA NA NA 19 (95%) Total No. 15 (100%) 15 (100%) 25 (100%) 10 (100%) 20 (100%) MAb, monoclonal antibody; NA, not applicable; UC, ulcerative colitis. Cancers are evaluated only as positive or negative for monoclonal antibody 55-2.

5 690 COOPER AND STEPLEWSKI GASTROENTEROLOGY Vol. 95, No.3 Figure 8. Control colons stained with MAb The staining is limited to the superficial (surface) columnar cell s. Magnification, x 100. Figure 7. Severely active UC stained with MAb The staining is localized only to basal crypt cells. Note similarity of this pattern to controls (Figure 5) and mildly active UC (Figure 6). Magnification, x 100. epithelial cells from patients with severely active UC express certain tumor-associated antigens (GICA) similar to cancers. Interestingly, when the disease is inactive or mildly active, epithelial cells express the GICA similar to controls, suggesting that the expression of tumor-associated antigens may be a reversible phenomena. Our data also show that proliferating epithelia of patients with severely active UC express features of mature epithelial cells (MAb 73-3 and MAb 19-9) while lacking certain features of immature basal crypt cells (MAb 55-2). Monoclonal antibody 19-9 detects the GICA (a sialylated Lewis A antigen) (12,13). Investigators have reported that this antigen is absent from normal colorectal epithelium (19,20). However, recently Bara et al, (21) and Cooper et al. (18) have published data indicating that the GICA is expressed in the superficial epithelial cells of the normal colon, indicating that the GICA is also a marker for mature colonic epithelial cells. Investigators have noted that the GICA is present in colorectal cancers, adenomas, and fetal colon (19,20,22-24). Recently, Olding et al, (25) noted that the GICA was present in dysplastic UC epithelium but was absent from inflamed or hyperplastic UC epithelium. Our data show that the pattern and degree of expression of GICA in UC epithelium is directly related to the degree of inflammation present and not limited to dysplastic epithelium. In UC there is a disturbance of normal colonic epithelial cell kinetics (1-3,5). In UC colonic epithelium there is " lack of maturation," with deoxyribonucleic acid synthesis and mitoses occurring at all levels of the crypt, findings similar to those noted in colonic adenomas and carcinomas (26). Also noted Table 3. Expression of Monoclonal Antibody 73-3 and Its Distribution Within the Crypt Expression of MAb with membranous staining pattern within Inactive UC and/or the crypt" Control mildly active UC Severely active UC Dysplasia Cancers" Negative 0(0%) 0(0%) 0(0%) 0(0%) 7 (35%) Base of crypt 0(0%) 0(0%) 0(0%) 0(0%) NA Surface of crypt 15 (100%)C 15 (100%)C 14 (56%)C 4 (40%) NA Surface and base 0(0%) 0(0%) 11 (44%) 6 (60%) NA Positive NA NA NA NA 13 (65%) Total No. 15 (100%) 15 (100%) 25 (100%) 10 (100%) 20 (100%) MAb, monoclonal antibody; NA, not applicable; UC, ulcerative colitis. a Results as they related only to membranous staining pattern of epithelial cells. See text for description. b Cancers are evaluated only as positive or negative for monoclonal antibody C In controls, 5 cases also had subnuclear staining of basal crypt goblet cells. Three cases of inactive/mildly active UC also had subnuclear staining of basal crypt goblet cells. Seven cases of severely active UC also had subnuclear staining of basal crypt goblet cells.

6 September 1988 MONOCLONAL ANTIBODIES AND ULCERATIVE COLITIS 691 Figure 9. Inactive area of UC from patient with dysplasia and cancer stained with MAb The staining is limited to surface epithelium. Compare with controls (Figure 8) and severely active colitis (Figure 10). Magnification, Xl00. is a more rapid migration of the epithelial cell to the crypt surface (1,2). The similarity of cell kinetics between active DC and colonic neoplasms may explain why DC epithelium extensively expresses GICA. The "fetallike nature" of DC epithelium has been noted with the similar expression of carcinoembryonic antigen (8,9), peanut lectin (27), and second-trimester fetal antigen by DC epithelium and neoplasms (28). This may be another reason why DC epithelium may extensively express GICA, which is also expressed by fetal colonic epithelium (24). Monoclonal antibody 55-2 detects the Lewis Y blood group antigen (a difucosylated type 2 blood group backbone) (14-16). This antigen is expressed focally in a small population of basal crypt (immature) epithelial cells in the normal distal colon and is expressed by colonic cancers and adenomas (29,30). In our study, we noted that the Lewis Y antigen was expressed in cancers and controls similar to that previously reported. However, unlike our findings with MAbs 19-9 and 73-3, the expression of Lewis Y antigen in severely active and dysplastic DC epithelium was similar to that of controls. Similar results have been noted by Shi et al. (31), who studied DC patients for expression of stage-specific embryonic antigen 1 (a monofucosylated type 2 blood group backbone). The fact that dysplastic DC epithelium expresses MAb 55-2 similar to controls correlates with our findings of MAb 55-2 (Lewis Y) expression in colonic adenomas (32). Monoclonal antibody 73-3 detects the protein moiety of a 35,000-dalton glycoprotein (17). We have noted that in the normal colon there is extensive membranous staining of superficial and surface epithelial cells with MAb 73-3, and that this staining pattern correlated with mature columnar cell differentiation. In our study, we found that 65% of adeno- carcinomas stained with MAb 73-3, data that are similar to those of Herlyn et al. (17). We also found that MAb 73-3 stained inactive and mildly active DC epithelium similar to controls; however, 44% of severely active DC and 60% of dysplastic DC epithelium stained with MAb 73-3 throughout the entire length of the crypt. In cases of severely active DC, this staining pattern was membranous in nature and could be correlated with columnar cell differentiation. In cases of severely active DC, there is marked depletion of mucin-producing cells and a replacement by nonmucin columnar-type cells. The extensive membrane staining by MAb 73-3 seen throughout the entire length of the crypt may be a biochemical expression of this loss of mucinproducing cells and replacement by columnar cells. Similarly, the abnormal cell kinetics seen in DC may explain why there is membrane staining with MAb 73-3 throughout the entire length of the crypt. It is known that in DC there is synthesis of incomplete glycoproteins, possibly due to an increased rate of migration of crypt epithelium to the surface, hence leaving less time for complete glycosylation (1,2,33). These incomplete glycoproteins lack certain carbohydrate moieties that are normally synthesized in the Golgi apparatus after synthesis of the protein backbone in the endoplasmic reticulum (34,35). As MAb 73-3 detects the protein moiety of a 35,000 dalton glycoprotein, the finding of membrane staining by MAb 73-3 throughout the entire crypt length in active DC may indicate the more rapid transport of an underglycosylated glycoprotein through the cell organelles to the cell membrane. The finding that actively proliferating DC epithe- Figure 10. Severely active UC stained with MAb The epithelium shows "regenerative-type changes," with intense membrane staining throughout the entire length of the crypt. Magnification, Xl00.

7 692 COOPER AND STEPLEWSKI GASTROENTEROLOGY Vol. 95, No.3 lium expresses markers of mature columnar epithelium throughout the entire length of the crypt, as evidenced by staining with MAbs 19-9, 55-2, and 73-3, was surprising. We have recently shown that hyperplastic colonic polyps, which consist of mature colonic epithelial cells at a lower point of the crypt than controls, extensively express GICA throughout the entire crypt (18). Recently, Chesa et al. (36) have shown similar paradoxical expression of mature colonic epithelium cytokeratin by villous adenomas. Unfortunately, the three MAbs used in this study have not provided us with a marker to differentiate dysplasia from reactive epithelium. However, they have provided us with new data about the biochemical changes that perturbed colonic epithelium may undergo. The staining patterns with MAbs 19-9, 55-2, and 73-3 in cases of severely active UC indicate that these proliferating cells have certain features of mature colonic epithelium. Our data also show that extensive expression of the GICA is not limited to neoplastic states. In UC its expression is directly related to the degree of disease activity, and this expression may be reversible when the inflammatory process is more quiescent. References 1. Bleiberg H, Mainguet P, Galand P, Chretien J, Dupont Mairesse N. Cell renewal in the human rectum. In vitro autoradiographic study in active ulcerative colitis. Gastroenterology 1970;58: Eastwood GL, Trier IS. Epithelial cell renewal in cultured rectal biopsies in ulcerative colitis. Gastroenterology 1973;64: Franklin WA, McDonald GB, Stein HO, et al. Immunohistological demonstration of abnormal colonic crypt cell kinetics in ulcerative colitis. Hum PathoI1985;16: Riddel RH, Goldman H, Rausohoff DF, et al. Dysplasia in inflammatory bowel disease. Standardized classification with provisional clinical application. Hum PathoI1985;14: Serafini EP, Kirk AP, Chambers TJ. Rate and pattern of epithelial cell proliferation in ulcerative colitis. Gut 1981;22: Cooper HS, Farano P, Coapman RA. Peanut lectin binding sites of colons of patients with ulcerative colitis. Arch Pathol Lab Med 1987;111: Pihl E, Peura A, Johnson WR, McDermott FT, Hughes ESR. T-antigen expression by peanut agglutinin staining relates to mucosal dysplasia in ulcerative colitis. Dis Colon Rectum 1985;28: Roagnum T, Elgjo K, Fausa 0, Brandtzaeg P. Immunohistochemical evaluation of carcinoembryonic antigen, secretory component, and epithelial IgA in ulcerative colitis with dysplasia. Gut 1982;23: Ahren DJ, Warren GH, Greene LJ, Singleton JW, Brown WR. Search for a specific marker of mucosal dysplasia in chronic ulcerative colitis. Gastroenterology 1987;93: Ehsanullah M, Filipe MI, Gazzard B. Mucin secretion in inflammatory bowel disease. Correlation with disease activity and dysplasia. Gut 1982;23: Ehsanullah M, Filipe MI, Gazzard B. Morphological and mucus secretion criteria for the differential diagnosis of solitary ulcer syndrome and non-specific proctitis. J Clin PathoI1982;35: Koprowski H, Steplewski Z, Mitchell K, Herlyn M, Herlyn D, Fuhrer P. Colorectal carcinoma antigens detected by hybridoma antibodies. Somat Cell Genet 1979;5: Magnani JL, Steplewski Z, Koprowski H, Ginsburg V. Identification of the gastrointestinal and pancreatic cancerassociated antigen detected by monoclonal antibody 19-9 in the sera of patients as a mucin. Cancer Res 1983;43: Hakomori S, Kabata A. Blood group antigens. In: Sela M, ed. The antigens. Volume 279. New York, Academic, Steplewski Z, Blaszezyle M, Herlyn D, Herlyn M, Koprowski H. Effector cells in ADCC with antibreast cancer monoclonal antibodies. In: Cernani M, ed. Monoclonal antibodies and breast cancer. Boston: Nijhoff 1985: Steplewski Z, Koprowski H. Glycolipid and glycoprotein markers of gastrointestinal cancers. In: Koprowski H, Ferrone S, Albertini A, eds. Biotechnology in diagnostics. New York: Elsevier Science, 1985: Herlyn D, Herlyn R, Ross AH, Ernst C, Atkinson B, Koprowski H. Efficient selection of human tumor growth inhibiting monoclonal antibodies. J Immunol Methods 1984;73: Cooper HS, Marshall C, Ruggerio F, Steplewski Z. Hyperplastic polyps of the colon and rectum. An immunohistochemical study with monoclonal antibodies against blood group antigens [Sialosyl-Le", Lea, Le b, Lex, LeY, A, B, H). Lab Invest 1987;57: Arends JW, Verstynen C, Bosman FT, Hilgres J, Steplewski Z. Distribution of monoclonal antibody-defined monosialoganglioside in normal and cancerous human tissues. An immunoperoxidase study. Hybridoma 1983;2: Atkinson BF, Ernst CS, Herlyn M, Steplewski Z, Sears HF, Koprowski H. Gastrointestinal cancer-associated antigen in immunoperoxidase assay. Cancer Res 1982;42: Bara J, Zabaleta EH, Mollicone R, Nap M, Burtin P. Distribution of GICA in normal gastrointestinal and endocervical mucosa and in mucinous ovarian cysts using antibody Am J Clin Pathol 1986;85: Gong E, Hirohashi S, Shimosato Y, et al. Expression of carbohydrate antigen 19-9 and stage specific embryonic antigen 1 in non-tumorous epithelium of the human colon and rectum. JNCI 1985;75: Olding LB, Thurin J, Svalander C, Koprowski H. Expression of gastrointestinal carcinoma-associated antigen (GICA) detected in human fetal tissues by monoclonal antibody NS Int J Cancer 1984;34: Rauz H, Labbe F, Fondaneche MC, Koprowski H, Burtin P. A study of gastrointestinal cancer-associated antigen (GICA) in human fetal organs. Int J Cancer 1983;32: Olding LB, Ahren C, Thurin J, Karlsson KA, Svalander C, Koprowski H. Gastrointestinal carcinoma-associated antigen detected by a monoclonal antibody in dysplasia and adenocarcinoma associated with chronic ulcerative colitis. Int J Cancer 1985;36: Deschner EE. Cell proliferation as a biological marker in human colorectal neoplasia. In: Winawer S, Schottenfeld D, Sherlock P. Colorectal cancer, prevention, epidemiology, and screening. New York: Raven, 1980: Coapman RA, Cooper HS. Peanut lectin binding sites in human fetal colon. Arch Pathol Lab Med 1986;110: Biasco G, Lipkin M, Minarini A, Higgins P, Miglioli M, Barbara L. Proliferative and antigenic properties of rectal cells in patients with chronic ulcerative colitis. Cancer Res 1984;44: Brown A, Ellis 10, Embleton MJ, Baldwin RW, Turner DR,

8 September 1988 MONOCLONAL ANTIBODIES AND ULCERATIVE COLITIS 693 Hardcastle JD. Immunohistochemical localization of Y hapten and the structurally related H type-2 blood group antigen on large bowel tumors and normal adult tissues. Int J Cancer 1984;33: Abe K, Hakomon SI, Ohshiba S. Differential expression of difucosyl type 2 chain (LeY) defined by monoclonal antibody AH6 in different locations of colonic epithelium in various types of colonic polyps and adenocarcinoma. Cancer Res 1986;46: Shi ZR, McIntyre LJ, Knowles BB, SoIter D, Kim YS. Expression of a carbohydrate differentiation antigen, stage-specific embryonic antigen 1, in human colonic adenocarcinoma. Cancer Res 1984;44: Ruggiero F, Cooper HS, Steplewski Z. Immunohistochemical study of colorectal adenomas with monoclonal antibodies against blood group antigens [Sialosyl-Le", Lea, Le b, LeX, LeY, A, B, H). Lab Invest (in press). 33. Clamp JR, Fraser G, Read AE. Study of carbohydrate content of mucus glycoproteins from normal and diseased colons. Clin Sci 1981;61: Neurtra MR, Grand RIo Trier JS. Glycoprotein synthesis, transport, and secretion by epithelial cells in human rectal mucosa. Normal and cystic fibrosis. Lab Invest 1977;36: Spiro RG. Glycoproteins, their biochemistry, biology, and role in human disease. N Engl J Med 1969;281: Chesa PG, Rettig WI. Melamed MR. Expression of cytokeratins in normal and neoplastic colonic epithelial cells. Implications for cellular differentiation and carcinogenesis. Am J Surg Pathol 1986;10: Received November 13, Accepted April 12, Address requests for reprints to: Harry S. Cooper, M.D., Department of Pathology, Thomas Jefferson University Hospital, 111 South 11th Street, Philadelphia, Pennsylvania The authors thank Carol Bass, [o Ann Feola, and Debbie Sedergren for technical assistance and Theresa Calabro and Debbie Paul for secretarial assistance.