Cancer Letters. 60 (1991)
9-13
Elsevier Scientific Publishers Ireland Ltd.
Cancer associated antigen CA19-9 in colonic effluent of patients with neoplasia of the colon and inflammatory bowel disease M. Tobi”, W. Steinbergb, ‘Department of
J. Henryb and L. Nochomovitzc
Gastroenterology
Tel-Aoiu Medical Center, 6 Weizman St., Tel-Aviv 64239 (Israel), and ‘Surgical Pathology, George Washington Uniuersity Medical Center, 2300 Eye
Washington
20037
Gastroenterology,
D.C.
bDiuision Street
oj
NW,
(U.S.A.).
(Received 8 March 1991) (Revision received 4 July 1991) (Accepted 8 July 1991)
Keywords: CA19-9; monoclonal antibody
Summary We measured colonic effluent samples from 10 patients with colorectal cancer, 13 with adenomatous polyps, 14 with normal colons and compared them to 10 patients with inflammatory bowel disease by measuring this CA1 9-9 content. Results showed considerable overlap between the different pathologic categories, making differentiation impossible. A lower leuel of CA19-9 in the effluent samples from patients with adenomas was noted. These differences were reproducible for assays performed seueral months apart. CA19-9 may originate from the upper gastrointestinal tract since large amounts are present in pancreatico-biliary secretions. This antigen is therefore not useful in the diagnosis of neoplasia or inflammatory bowel disease using colonic effluent samples as the test material.
Correspondence
to:
Dr.
Gastroenterology,
Wayne
Harper
3990
Hospital,
M.
Department
Tobi.
of
State University School of Medicine, John
R Street,
Detroit,
MI
48201,
U.S.A. Abbr&iotions: substances; monoclonal
CRC, IBD,
colorectal inflammatory
antibody.
0304-3835/91/$03.50
0
1991
Published and Printed in Ireland
cancer;
BGS,
bowel
blood
disease;
group MAb,
colonic
effluent;
Introduction The advent of monoclonal antibody (MAb) technology has led to the discovery of antibody-defined tumor associated antigens [l]. One such antigen, CA19-9, is defined by MAb raised against a colon cancer cell line [2]. This antigen has been shown to have sensitivity in the diagnosis of pancreatic cancer [3], and the antigen has been characterized as a sialylated lacto-IV-fucopentaose II that shares structural features with Lewis blood group substances (BGS) [2]. Normally, BGS are only expressed in the proximal adult colon. However, distal left-sided colonic carcinomas (and occasionally the normal mucosa in their immediate vicinity) also express BGS [4]. CA19-9 has been shown to be expressed in colonic neoplastic tissue [5]. Because this antigen has also been demonstrated in the mucosa of patients with inflammatory bowel disease and in normal mucosa, its utility as a marker in the diagnosis of colonic neoplasia and inflammatory bowel disease has been questioned [6]. CA19-9 has not been evaluated in colonic washings or feces where positive correlations between the presence of
El sevier Scientific Publishers Ireland Ltd
10
neoplasia and carcinoembryonic antigen levels have been reported [7 - lo]. Precolonoscopic washings may contain shed antigens detectable by monoclonal antibodies, which may be more representative of the mucosal status overall and correlate with presence of CRC [9,10]. We undertook to investigate whether CA19-9 could be quantitatively detected in colonic effluent samples and whether the effluent levels of CA19-9 are useful in differentiating colonic neoplasia and inflammatory bowel disease from normals. Materials
and Methods
Forty-seven patients undergoing routine diagnostic and surveillance colonoscopic exat The George Washington aminations Medical Center were evaluated. They were divided as follows: (1) endoscopically normal colon - 14 patients (average age 50.9 years); (2) colon cancer - 10 patients (average age 64.9 years); (3) adenomatous polyps - 13 patients (average age 58 years) ; (4) IBD - 10 patients (average age 42.7 years). The latter group consisted of 3 patients with Crohn’s disease and 7 with ulcerative colitis. The subjects underwent conventional bowel preparation, namely a 1.5-day liquid diet, magnesium citrate purge (300 ml) and 5 bisacodyl tablets the night before the procedure. On the day of the examination a tap water enema was administered; the first return was discarded and 100 ml of the second return was collected and centrifuged at 1500 rev./min in a clinical centrifuge. The supernatant was then filtered through a 0.8 pm filter (Nalgene). The filtrate was assayed for protein content by the method of Lowry et al. [ll], divided into aliquots and stored at - 2OOC. The visual interpretation during colonoscopy and review of all biopsy material constituted the evaluation of the colon. The colonoscopy procedure was regarded as complete if the cecum was reached; the study required that an unequivocal diagnosis of cancer, adenomatous polyps or inflammatory bowel disease be made. Most of the cancer patients had left-sided de
novo lesions, 6 had malignant degeneration in a pre-existing polyp and 1 had widespread carcinomatosis with a primary lesion in the ascending colon. There was no evidence of dysplasia in any of the multiple biopsy samples in the IBD patients. All effluent samples were analyzed in duplicate for CA19-9 in a solid phase radioimmunoassay according to the distributor’s recommendation (Centocor CA19-9, Malvern, PA). We used 200 ~1 of sample per test. Quantities of CA19-9 were expressed in units per ml, which is an arbitrary activity level corresponding to approximately 0.8 ng of purified antigen material [ 121. The results were also expressed in units per mg protein to allow for standardization of different samples [9]. Assays were repeated at least twice at an interval of several months on all patients to assess reproducibility after the stored frozen samples had been thawed. Statistical analysis was performed by Student’s two-tailed t-test. Results The mean levels of CA19-9 revealed no clinically significant differences between the diagnostic groups; the overlap is demonstrated in Fig. 1. The resultant means ( * standard deviations) were: cancer - 381 f 389 units per ml; polyp - 62 l 134; normal - 205 * 252; IBD groups - 198 f 278. There is a bimodal expression of CA19-9 in the effluent of cancer cases and generally lower levels in the samples of patients with polyps as compared to those with cancer. When the results are standardized in units/mg protein (Fig. 2), the scatter in the cancer group is unimodal and the differences between groups follow the same trend: mean l standard deviation in cancer - 402 * 457; polyp - 125 f 221; normal - 247 l 327; IBD groups - 207 f 312. In this instance the difference between the cancer and polyp groups is not statistically significant (P = 0.06). With regard to reproducibility, similar trends were evident in an assay performed at an interval of several months from the original: cancer (5 patients) 132 * 183 (mean f standard de-
11
”
.
Fig. 1. Scattergram of ent samples. The short means for each group. units/ml. Patient disease horizontal axis.
...
previously, and 3 were found to have levels of greater than 37 units/ml. In these 4 patients, the outcome of the earlier assay was confirmatory of that performed later.
:
CA19-9 levels in colonic effluhorizontal lines designate the The results are expressed in categories are indicated on the
viation); polyps (8 patients) 88 f 100; normal (14: patients) 120 f 170 units/ml. Again, the overlap between groups does not allow for clinical differentiation. In the IBD group, only 4 of the 10 IBD cases had been assayed
Fig. 2. Scattergram of CA19-9 levels in colonic effluent samples expressed in units/mg of protein. Mean binding levels are indicated in each group by short horizontal lines. There is a trend to lower levels when comparing results in the cancer and polyp groups (P = 0.06).
Discussion This study shows that the monoclonal antibody-defined tumor associated antigen CA19-9 is detectable in pre-colonoscopic effluent samples. The levels of CA19-9, whether expressed as units/ml or standardized as units/mg protein, are similar and do not differentiate between the different pathologic categories. This test is not diagnostic for CRC using effluent samples, despite the trend toward higher levels of CA19-9 in cancer patients. Fifty percent of cancer patients had levels greater than 500 units/ml as opposed to 28% of normal subjects. It is unclear why CA19-9 levels are lower in effluent samples of patients with polyps since CA19-9 can be demonstrated in polyps [5]. It may be that the form of CA19-9 in these polyps is more tightly bound and less able to be shed. Another possibility is that higher CA19-9 levels may recurrent predict a tendency towards neoplasia. Follow-up in these patients is therefore necessary to determine the possible prognostic significance of this finding. It has been shown that CA19-9 is structurally identical to a sialylated Lewis A blood group substance (BGS) [3]. Patients who do not express this BGS would therefore not be expected to show CA19-9 binding activity in their effluent sample, and this may theoretically account for the bimodal distribution of CA19-9 levels seen in patient with cancer but is unlikely in that only 5% of the population lacks the Legene [13]. We did not evaluate Lewis A status in these subjects in this study. We have shown that the overall differences between the pathologic groups are consistently preserved when the assay is repeated after a substantial time period. This is inclusive of some of the patients with IBD, despite the limited number of patients involved. The CEA levels extracted from feces have
12
generally correlated well with the presence of cancer as summarized [lo]. In previous work performed with colonic lavage, CEA levels correlated with a presence of polyps greater than 1 cm in size, but not cancers [7]. Similarly, where polyclonal antibodies were used to evaluate CEA levels, rectal cancers showed no correlation, but levels in IBD patients were high [8]. We have shown that CEA in effluent samples may be detected by MAbs and that a trend exists between the different high risk groups [lo]. Recently we have shown a correlation between the presence of CRC and an early tumor antigen in colonic effluent material
PI. In IBD patients, CA19-9 staining is demonstrable in 50 -90% of cells in the involved mucosa, except in the cecum [6]. The patients in this study were extensively sampled. No clear difference was seen in the CA19-9 staining between patients with ulcerative colitis and Crohn’s disease. The staining was distributed heterogeneously at intracellular, membranous and luminal locations and hence, could be expected to be shed into the effluent. Our results show lower levels of CA19-9 in the effluent of patients with IBD, which appears to be inconsistent with what one would expect. One explanation may be that the antigens are washed out by the pre-colonoscopic bowel preparation. Alternatively, since inflammation induces goblet cell mucin depletion in IBD [14], the lower levels of CA19-9, a constituent of gastrointestinal mucin, [2] may result from this depletion. The results of a recent study on CA19-9 staining in ulcerative colitis also failed normal from dysplastic to distinguish epithelium [15] suggesting that CA19-9 has limited diagnostic use in this setting. The similarity of CA19-9 levels in colonic effluent samples from patients with CRC and subjects with a macroscopically normal colon may be explained by the profusion of CA19-9 which probably emanates from normal pancreaticobiliary secretions [16] and percolates down to the colon. Our attempt to standardize for protein content, to enhance diagnostic yield as has been
shown for CA19-9 in pancreatic secretions from patients with pancreatic cancer [16], was unsuccessful. Thus it may be more appropriate to evaluate antigens that originate directly from the colon such as CEA in the diagnosis of colorectal disease [lo]. Acknowledgments The authors wish to thank Dr. Terence M. Phillips for his invaluable assistance with the manuscript, Drs. Trujillo and O’Kieffe for the collection of clinical material, and A.P. Spence for technical assistance. This paper is dedicated to the memory of Albert Levy M.D. References 1
Arends J.W.,
Bosman F.T.
antigens in large-bowel Magnani J.L., V.
Tissue
Biochim. Biophys Ac-
1 - 19.
ta., 780, 2
and Hilgers J. (1985)
carcinoma.
(1983)
Steplewski i!., Koprowski H. and Ginsburg
Identification
of the gastrointestinal
creatic cancer-associated antibody
and pan-
antigen detected by monoclonal
19-9 in the sera of patients as a mucin. Cancer
Res., 43, 5489-92. 3
Steinberg tumor
W.
(1990)
associated
The clinical utility of the CA19-9
antigen.
Am.
J.
Gastroenterol.,
85,
350 - 355. 4
Yuan
M.,.ltzkowitz
S.H.
and Palekar
A.,
et al. (1985)
Distribution of blood group antigens A, B, H, human,
normal,
and
malignant
colonic
Lewis in
tissue.
Cancer
Res., 45, 4499-4511. 5
Afdal
N.H.,
Long
O’Donoghue
D.P.
A.,
Tobbia
(1987)
I.,
Cullen
A.
lmmunohistochemical
and
CA19-9
in primary colonic polyps and polyps synchronous with colorectal cancer. 6
Gut, 28, 594 - 600.
Frykholm G., Enblad P., Pahlman L. and Busch C. (1987) Expression of the carcinoma-associated and CA50
in inflammatory
antigens CAlV-9
bowel disease.
Dis. Colon.
Rect., 30. 545-548. 7
Winawer
S.J.,
Fleischer M. and Green
Carcinoembryonic Gastroenterology, 8
Vellacott
K.D.,
in
antigen 92, 719-
Groom
and Hardcastle J.D.
D., Balfour T. W.,
(1982)
Tobi M.,
Darmon
Nochomovitz expression
L.E.
10
Darmon
8, 61-67. Heller
and Steinberg W.M. adenoma
T.,
Rosen P.,
(1990)
Increased
associated antigen in
effluent of patients with colorectal cancer.
Lett., 50, 21-
25.
E., Tobi M. and Rozen P. (1990)
indentifying
Baldwin R.W.
Tumor associated products in
Phillips T.,
of a putative
precolonoscopic Cancer
E.,
lavage.
722.
colonic lavage fluid. Clin. Oncol., 9
S., et al. (1977) colonic
persons
with colorectal
Newer tests for
neoplasia.
In: Ad-
13
vances in large bowel cancer: Policy, prevention, research and treatment. Frontiers of Gastrointestinal Research, 18.
11
12
13
pp. 236- 247. Editors: P. Rozen, C.B. Reich and S.J. Winawer. Karger, Basel. Lowry O.H., Rosenbrough N.J., Farr A.L. and Randall R.J. (1951) Protein measurements with folin phenol reagent. J. Biol. Chem., 193, 265-275. Del Villano B.C., Brennan S. and Brock P., et al. (1983) Radioimmunometric assay for a monoclonal antibodydefined tumour marker, CA19-9. Clin. Chem., 29, 549 - 52. ltzkowito S.H. and Kim Y.S. (1986) New carbohydrate tumor markers. Gastroenterology, 90, 491-494.
14
15
16
Alpert M.B. and Nochomovitz
L.E. (1989) Dysplasia
and
cancer surveillance in inflammatory bowel disease. Gastroenterol. Clin. North Am., 18 (l), 83-97. Cooper H.S. and Steplewski 2. (1988) lmmunohistologic study of ulcerative colitis with monoclonal antibodies against tumor associated and/or differentiation antigens. Gastroenterology, 95, 686 - 93. Schmiegel W. (1989) Tumor markers in pancreatic cancer current concepts. Hepato-gastroenterology. 36, 446-49.
9-13
Elsevier Scientific Publishers Ireland Ltd.
Cancer associated antigen CA19-9 in colonic effluent of patients with neoplasia of the colon and inflammatory bowel disease M. Tobi”, W. Steinbergb, ‘Department of
J. Henryb and L. Nochomovitzc
Gastroenterology
Tel-Aoiu Medical Center, 6 Weizman St., Tel-Aviv 64239 (Israel), and ‘Surgical Pathology, George Washington Uniuersity Medical Center, 2300 Eye
Washington
20037
Gastroenterology,
D.C.
bDiuision Street
oj
NW,
(U.S.A.).
(Received 8 March 1991) (Revision received 4 July 1991) (Accepted 8 July 1991)
Keywords: CA19-9; monoclonal antibody
Summary We measured colonic effluent samples from 10 patients with colorectal cancer, 13 with adenomatous polyps, 14 with normal colons and compared them to 10 patients with inflammatory bowel disease by measuring this CA1 9-9 content. Results showed considerable overlap between the different pathologic categories, making differentiation impossible. A lower leuel of CA19-9 in the effluent samples from patients with adenomas was noted. These differences were reproducible for assays performed seueral months apart. CA19-9 may originate from the upper gastrointestinal tract since large amounts are present in pancreatico-biliary secretions. This antigen is therefore not useful in the diagnosis of neoplasia or inflammatory bowel disease using colonic effluent samples as the test material.
Correspondence
to:
Dr.
Gastroenterology,
Wayne
Harper
3990
Hospital,
M.
Department
Tobi.
of
State University School of Medicine, John
R Street,
Detroit,
MI
48201,
U.S.A. Abbr&iotions: substances; monoclonal
CRC, IBD,
colorectal inflammatory
antibody.
0304-3835/91/$03.50
0
1991
Published and Printed in Ireland
cancer;
BGS,
bowel
blood
disease;
group MAb,
colonic
effluent;
Introduction The advent of monoclonal antibody (MAb) technology has led to the discovery of antibody-defined tumor associated antigens [l]. One such antigen, CA19-9, is defined by MAb raised against a colon cancer cell line [2]. This antigen has been shown to have sensitivity in the diagnosis of pancreatic cancer [3], and the antigen has been characterized as a sialylated lacto-IV-fucopentaose II that shares structural features with Lewis blood group substances (BGS) [2]. Normally, BGS are only expressed in the proximal adult colon. However, distal left-sided colonic carcinomas (and occasionally the normal mucosa in their immediate vicinity) also express BGS [4]. CA19-9 has been shown to be expressed in colonic neoplastic tissue [5]. Because this antigen has also been demonstrated in the mucosa of patients with inflammatory bowel disease and in normal mucosa, its utility as a marker in the diagnosis of colonic neoplasia and inflammatory bowel disease has been questioned [6]. CA19-9 has not been evaluated in colonic washings or feces where positive correlations between the presence of
El sevier Scientific Publishers Ireland Ltd
10
neoplasia and carcinoembryonic antigen levels have been reported [7 - lo]. Precolonoscopic washings may contain shed antigens detectable by monoclonal antibodies, which may be more representative of the mucosal status overall and correlate with presence of CRC [9,10]. We undertook to investigate whether CA19-9 could be quantitatively detected in colonic effluent samples and whether the effluent levels of CA19-9 are useful in differentiating colonic neoplasia and inflammatory bowel disease from normals. Materials
and Methods
Forty-seven patients undergoing routine diagnostic and surveillance colonoscopic exat The George Washington aminations Medical Center were evaluated. They were divided as follows: (1) endoscopically normal colon - 14 patients (average age 50.9 years); (2) colon cancer - 10 patients (average age 64.9 years); (3) adenomatous polyps - 13 patients (average age 58 years) ; (4) IBD - 10 patients (average age 42.7 years). The latter group consisted of 3 patients with Crohn’s disease and 7 with ulcerative colitis. The subjects underwent conventional bowel preparation, namely a 1.5-day liquid diet, magnesium citrate purge (300 ml) and 5 bisacodyl tablets the night before the procedure. On the day of the examination a tap water enema was administered; the first return was discarded and 100 ml of the second return was collected and centrifuged at 1500 rev./min in a clinical centrifuge. The supernatant was then filtered through a 0.8 pm filter (Nalgene). The filtrate was assayed for protein content by the method of Lowry et al. [ll], divided into aliquots and stored at - 2OOC. The visual interpretation during colonoscopy and review of all biopsy material constituted the evaluation of the colon. The colonoscopy procedure was regarded as complete if the cecum was reached; the study required that an unequivocal diagnosis of cancer, adenomatous polyps or inflammatory bowel disease be made. Most of the cancer patients had left-sided de
novo lesions, 6 had malignant degeneration in a pre-existing polyp and 1 had widespread carcinomatosis with a primary lesion in the ascending colon. There was no evidence of dysplasia in any of the multiple biopsy samples in the IBD patients. All effluent samples were analyzed in duplicate for CA19-9 in a solid phase radioimmunoassay according to the distributor’s recommendation (Centocor CA19-9, Malvern, PA). We used 200 ~1 of sample per test. Quantities of CA19-9 were expressed in units per ml, which is an arbitrary activity level corresponding to approximately 0.8 ng of purified antigen material [ 121. The results were also expressed in units per mg protein to allow for standardization of different samples [9]. Assays were repeated at least twice at an interval of several months on all patients to assess reproducibility after the stored frozen samples had been thawed. Statistical analysis was performed by Student’s two-tailed t-test. Results The mean levels of CA19-9 revealed no clinically significant differences between the diagnostic groups; the overlap is demonstrated in Fig. 1. The resultant means ( * standard deviations) were: cancer - 381 f 389 units per ml; polyp - 62 l 134; normal - 205 * 252; IBD groups - 198 f 278. There is a bimodal expression of CA19-9 in the effluent of cancer cases and generally lower levels in the samples of patients with polyps as compared to those with cancer. When the results are standardized in units/mg protein (Fig. 2), the scatter in the cancer group is unimodal and the differences between groups follow the same trend: mean l standard deviation in cancer - 402 * 457; polyp - 125 f 221; normal - 247 l 327; IBD groups - 207 f 312. In this instance the difference between the cancer and polyp groups is not statistically significant (P = 0.06). With regard to reproducibility, similar trends were evident in an assay performed at an interval of several months from the original: cancer (5 patients) 132 * 183 (mean f standard de-
11
”
.
Fig. 1. Scattergram of ent samples. The short means for each group. units/ml. Patient disease horizontal axis.
...
previously, and 3 were found to have levels of greater than 37 units/ml. In these 4 patients, the outcome of the earlier assay was confirmatory of that performed later.
:
CA19-9 levels in colonic effluhorizontal lines designate the The results are expressed in categories are indicated on the
viation); polyps (8 patients) 88 f 100; normal (14: patients) 120 f 170 units/ml. Again, the overlap between groups does not allow for clinical differentiation. In the IBD group, only 4 of the 10 IBD cases had been assayed
Fig. 2. Scattergram of CA19-9 levels in colonic effluent samples expressed in units/mg of protein. Mean binding levels are indicated in each group by short horizontal lines. There is a trend to lower levels when comparing results in the cancer and polyp groups (P = 0.06).
Discussion This study shows that the monoclonal antibody-defined tumor associated antigen CA19-9 is detectable in pre-colonoscopic effluent samples. The levels of CA19-9, whether expressed as units/ml or standardized as units/mg protein, are similar and do not differentiate between the different pathologic categories. This test is not diagnostic for CRC using effluent samples, despite the trend toward higher levels of CA19-9 in cancer patients. Fifty percent of cancer patients had levels greater than 500 units/ml as opposed to 28% of normal subjects. It is unclear why CA19-9 levels are lower in effluent samples of patients with polyps since CA19-9 can be demonstrated in polyps [5]. It may be that the form of CA19-9 in these polyps is more tightly bound and less able to be shed. Another possibility is that higher CA19-9 levels may recurrent predict a tendency towards neoplasia. Follow-up in these patients is therefore necessary to determine the possible prognostic significance of this finding. It has been shown that CA19-9 is structurally identical to a sialylated Lewis A blood group substance (BGS) [3]. Patients who do not express this BGS would therefore not be expected to show CA19-9 binding activity in their effluent sample, and this may theoretically account for the bimodal distribution of CA19-9 levels seen in patient with cancer but is unlikely in that only 5% of the population lacks the Legene [13]. We did not evaluate Lewis A status in these subjects in this study. We have shown that the overall differences between the pathologic groups are consistently preserved when the assay is repeated after a substantial time period. This is inclusive of some of the patients with IBD, despite the limited number of patients involved. The CEA levels extracted from feces have
12
generally correlated well with the presence of cancer as summarized [lo]. In previous work performed with colonic lavage, CEA levels correlated with a presence of polyps greater than 1 cm in size, but not cancers [7]. Similarly, where polyclonal antibodies were used to evaluate CEA levels, rectal cancers showed no correlation, but levels in IBD patients were high [8]. We have shown that CEA in effluent samples may be detected by MAbs and that a trend exists between the different high risk groups [lo]. Recently we have shown a correlation between the presence of CRC and an early tumor antigen in colonic effluent material
PI. In IBD patients, CA19-9 staining is demonstrable in 50 -90% of cells in the involved mucosa, except in the cecum [6]. The patients in this study were extensively sampled. No clear difference was seen in the CA19-9 staining between patients with ulcerative colitis and Crohn’s disease. The staining was distributed heterogeneously at intracellular, membranous and luminal locations and hence, could be expected to be shed into the effluent. Our results show lower levels of CA19-9 in the effluent of patients with IBD, which appears to be inconsistent with what one would expect. One explanation may be that the antigens are washed out by the pre-colonoscopic bowel preparation. Alternatively, since inflammation induces goblet cell mucin depletion in IBD [14], the lower levels of CA19-9, a constituent of gastrointestinal mucin, [2] may result from this depletion. The results of a recent study on CA19-9 staining in ulcerative colitis also failed normal from dysplastic to distinguish epithelium [15] suggesting that CA19-9 has limited diagnostic use in this setting. The similarity of CA19-9 levels in colonic effluent samples from patients with CRC and subjects with a macroscopically normal colon may be explained by the profusion of CA19-9 which probably emanates from normal pancreaticobiliary secretions [16] and percolates down to the colon. Our attempt to standardize for protein content, to enhance diagnostic yield as has been
shown for CA19-9 in pancreatic secretions from patients with pancreatic cancer [16], was unsuccessful. Thus it may be more appropriate to evaluate antigens that originate directly from the colon such as CEA in the diagnosis of colorectal disease [lo]. Acknowledgments The authors wish to thank Dr. Terence M. Phillips for his invaluable assistance with the manuscript, Drs. Trujillo and O’Kieffe for the collection of clinical material, and A.P. Spence for technical assistance. This paper is dedicated to the memory of Albert Levy M.D. References 1
Arends J.W.,
Bosman F.T.
antigens in large-bowel Magnani J.L., V.
Tissue
Biochim. Biophys Ac-
1 - 19.
ta., 780, 2
and Hilgers J. (1985)
carcinoma.
(1983)
Steplewski i!., Koprowski H. and Ginsburg
Identification
of the gastrointestinal
creatic cancer-associated antibody
and pan-
antigen detected by monoclonal
19-9 in the sera of patients as a mucin. Cancer
Res., 43, 5489-92. 3
Steinberg tumor
W.
(1990)
associated
The clinical utility of the CA19-9
antigen.
Am.
J.
Gastroenterol.,
85,
350 - 355. 4
Yuan
M.,.ltzkowitz
S.H.
and Palekar
A.,
et al. (1985)
Distribution of blood group antigens A, B, H, human,
normal,
and
malignant
colonic
Lewis in
tissue.
Cancer
Res., 45, 4499-4511. 5
Afdal
N.H.,
Long
O’Donoghue
D.P.
A.,
Tobbia
(1987)
I.,
Cullen
A.
lmmunohistochemical
and
CA19-9
in primary colonic polyps and polyps synchronous with colorectal cancer. 6
Gut, 28, 594 - 600.
Frykholm G., Enblad P., Pahlman L. and Busch C. (1987) Expression of the carcinoma-associated and CA50
in inflammatory
antigens CAlV-9
bowel disease.
Dis. Colon.
Rect., 30. 545-548. 7
Winawer
S.J.,
Fleischer M. and Green
Carcinoembryonic Gastroenterology, 8
Vellacott
K.D.,
in
antigen 92, 719-
Groom
and Hardcastle J.D.
D., Balfour T. W.,
(1982)
Tobi M.,
Darmon
Nochomovitz expression
L.E.
10
Darmon
8, 61-67. Heller
and Steinberg W.M. adenoma
T.,
Rosen P.,
(1990)
Increased
associated antigen in
effluent of patients with colorectal cancer.
Lett., 50, 21-
25.
E., Tobi M. and Rozen P. (1990)
indentifying
Baldwin R.W.
Tumor associated products in
Phillips T.,
of a putative
precolonoscopic Cancer
E.,
lavage.
722.
colonic lavage fluid. Clin. Oncol., 9
S., et al. (1977) colonic
persons
with colorectal
Newer tests for
neoplasia.
In: Ad-
13
vances in large bowel cancer: Policy, prevention, research and treatment. Frontiers of Gastrointestinal Research, 18.
11
12
13
pp. 236- 247. Editors: P. Rozen, C.B. Reich and S.J. Winawer. Karger, Basel. Lowry O.H., Rosenbrough N.J., Farr A.L. and Randall R.J. (1951) Protein measurements with folin phenol reagent. J. Biol. Chem., 193, 265-275. Del Villano B.C., Brennan S. and Brock P., et al. (1983) Radioimmunometric assay for a monoclonal antibodydefined tumour marker, CA19-9. Clin. Chem., 29, 549 - 52. ltzkowito S.H. and Kim Y.S. (1986) New carbohydrate tumor markers. Gastroenterology, 90, 491-494.
14
15
16
Alpert M.B. and Nochomovitz
L.E. (1989) Dysplasia
and
cancer surveillance in inflammatory bowel disease. Gastroenterol. Clin. North Am., 18 (l), 83-97. Cooper H.S. and Steplewski 2. (1988) lmmunohistologic study of ulcerative colitis with monoclonal antibodies against tumor associated and/or differentiation antigens. Gastroenterology, 95, 686 - 93. Schmiegel W. (1989) Tumor markers in pancreatic cancer current concepts. Hepato-gastroenterology. 36, 446-49.
