Tumor Markers CEA, CA19-9 and CA125 in Monitoring of Response to Systemic Chemotherapy in Patients with Advanced Gastric Cancer

Jpn J Clin OncoI1999;29(11)550-555 Tumor Markers CEA, CA19-9 and CA125 in Monitoring of Response to Systemic Chemotherapy in Patients with Advanced Gastric Cancer Takekazu Yamao, Shunkichi Kai, Akira Kazami, Koichi Koizumi, Takayoshi Handa, Norishige Takemoto and Masakazu Maruyama Department of Medicine, Cancer Institute Hospital, Tokyo, Japan Background: To evaluate whether tumor markers can be used to assess response to systemic chemotherapy, we analyzed preliminarily the relationship between the response to chemotherapy based on serial imaging and on change in serum tumor marker level of CEA, CA19-9 and CA125. Methods: We analyzed 26 patients with advanced gastric cancer in whom at least one of the tumor markers CEA, CA19-9 and CA125 was elevated before systemic chemotherapy with regard to the relationship between the change in serum tumor marker level and response assessment by imaging studies throughout the treatment course. A responder was defined as showing a ~50% drop in tumor marker level for more than 4 weeks. Results: The sensitivity and negative predictive value of falling tumor marker level after chemotherapy for a partial response in imaging was 100 % When patients were categorized as responders or non-responders, a significant correlation was observed between the assessment of response by tumor markers and by imaging studies. The survival time of responders assessed by tumor markers was significantly longer than that of non-responders. Conclusions: The measurement of tumor markers might be useful in monitoring response and in predicting the prognosis of patients with advanced gastric cancer treated with systemic chemotherapy. Tumor markers may be used as a means of monitoring treatment in patients when in an imaging study it is difficult to assess response to chemotherapy in clinical practice. Further studies are required to confirm these findings. Key words: tumor marker - carcinoembryonic antigen - carbohydrate antigen 19-9 - carbohydrate antigen 125 - gastric cancer - chemotherapy INTRODUCTION Chemotherapy for patients with unresectable or metastatic gastric cancer is still under investigation. However, some recent clinical trials which compared systemic chemotherapy with supportive care alone have revealed a significant benefit in survival and quality of life for patients treated with systemic chemotherapy (1-). At the present, the only drugs active against advanced gastric cancer are 5-fluorouracil, mitomycin C, doxorubicin, cis-platinum and newly developed irinotecan, Received May 12, 1999; accepted July 29,1999 For reprints and all correspondence: Takekazu Yamao, Department of Medicine, Cancer Institute Hospital, 1-7-1 Kami-ikebukuro, Toshima-ku, Tokyo 170-8455, Japan. E-mail: tyamao@jfcr.or.jp Abbreviations: CEA, carcinoembryonic antigen; CA19-9, carbohydrate antigen 19-9; CA125, carbohydrate antigen 125; CR, complete response; PR, partial response; NC, no change; PO, progressive disease although various combination chemotherapies have been investigated to increase the anti-tumor activity (4-6). In the clinical practice of chemotherapy against neoplasm, the assessment of clinical efficacy is very important, because the decision to continue or discontinue treatment is based on it. In general, the assessment of response to chemotherapy of various solid tumors including advanced gastric cancer is made by measurement of tumor visualized by serial imaging, commonly CT scanning (7). In addition, in Japan, X-ray series and endoscopy of the stomach are used for the assessment of the response to chemotherapy in the primary gastric lesion according to the criteria described in the general rules for the treatment of gastric cancer proposed by the Japan Society for Gastric Cancer (8). However, objective assessment of efficacy in chemotherapy based on imaging is not always easy because patients often have a disease which is not measurable by imaging study, such as diffuse peritoneal dissemination in advanced gastric cancer. Furthermore, response assessment by imaging study is expensive and time consuming for the patients. 1999 Foundation for Promotion of Cancer Research

Jpn J Clin OncoI1999;29(11) 551 Table 1. Characteristics of the patients Total No. of patients Performance status oor 1 2 Gender Male Female Median age (range) (years) Macroscopic type Diffuse Non-diffuse Histological type Differentiated Undifferentiated Prior chemotherapy Yes No Prior gastrectomy Yes No Metastatic site Peritoneum Lymph nodes Liver Lung Chemotherapy regimen MTX+SFU CPT-ll + CDDP 5FU +CDDP CPT-II CPT-ll +MMC Elevated tumor marker before chemotherapy CEA alone CA19-9 alone CA125 alone CEA + CA19-9 CEA + CA12S CA19-9 + CAl25 CEA + CA19-9 + CA125 Recently, various tumors markers have been developed and it is known that the serum levels of markers such as carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CAI9-9) and carbohydrate antigen 125 (CA125) are elevated in patients with advanced gastric cancer (9-11). However, the clinical usefulness of tumor markers has not been well defined from a diagnostic and therapeutic point of view. Tumor markers are not useful in diagnosis or screening disease because of low sensitivity and specificity. Although reports have shown the 26 22 4 15 11 51 (5-76) 12 14 9 17 2 6 20 12 11 5 1 16 7 1 4 6 2 5 Table 2. Chemotherapy regimens used in this study Regimen Dosage Reference No. MTX + SFU Methotrexate (100 mg/m", day 1) 16 + 5FU (600 mg/m-, day 1) + leucovorin (10 mg/m-, days 2 and ) CPT-II + CDDP Irinotecan (70 mg/m-, days 1 and IS) S + cis-platinum (80 mg/m-, day 1) SFU +CDDP 5-Fluorouracil (800mg/m 2, days 1-5) 6 + cis-platinum (20 mg/m'', days 1-5) CPT-II Irinotecan (150 rng/m-, days I and 15) 17 CPT-II +MMC Irinotecan (80 mg/m", days 1 and 15) 18 + mitomycin C (5 mg/m-, days 1 and 15) value of tumor markers as a prognostic factor for patients with advanced gastric cancer, clinical studies evaluating the roles of tumor markers in the monitoring of chemotherapeutic efficacy are limited (12-15). In the present study, we analyzed the relationship between conventional assessment of response to chemotherapy by serial imaging and that by the change in serum level of CEA, CA19 9 and CA125, tumor markers commonly used in cases of advanced gastric cancer, in order to evaluate preliminarily whether tumor markers can be used to assess chemotherapeutic efficacy. PATIENTS AND METHODS PATIENTS Between January 1997 and June 1998, a total of 1 patients with unresectable or metastatic gastric cancer with measurable lesion were treated with systemic chemotherapy in the Cancer Institute Hospital, Tokyo. In 28 of these 1 patients (90%), the pretreatment serum level of at least one of CEA, CA19-9 and CA125 was elevated. In 26 patients, the serial serum tumor marker level was measured and chemotherapeutic response was assessed by serial imaging studies throughout the course of the chemotherapy. Two patients were excluded because their tumor marker measurements were insufficient for analysis. The characteristics of the remaining 26 patients are summarized in Table 1. Fifteen patients were male and 11 female and the median age was 51 years (range 5-76 years). The majority of patients had good performance status (PS 0 or 1 in 22 patients and PS 2 in four patients) and the functions of the main organs such as bone marrow, liver and kidney were well preserved prior to chemotherapy in all patients. Twentythree patients had no prior chemotherapy and three were previously treated with 5FU-based chemotherapy. The metastatic site was peritoneum in 12 subjects, lymph node in 11, liver in five, bone in three and lung in one. The chemotherapy regimens used in this period are listed in Table 2.

552 Use oftumor markers in monitoring chemotherapy MEASUREMENT OF TuMOR MARKERS Serum CEA, CA19-9 and CA125 levels were measured prior to initiation of chemotherapy (within 7 days) and then again every 2 weeks after initiation of chemotherapy. Serum CEA, CA19-9 and CA125 levels were measured by the enzyme immunoassay using an AxSYM CEA kit (Dainabot, Tokyo, Japan), Graozyme New CA19-9 kit (Wako Junyaku Kogyo, Osaka, Japan) and AxSYM CA125 kit (Dainabot, Tokyo), respectively. The cut-off values for CEA, CA19-9 and CA125 were 5.0 ng/ml 7 Uzrnl and 6 Urrnl, respectively. ASSESSMENT OF CHEMOTHERAPEUTIC RESPONSE BY CHANGE OF TuMOR MARKER LEVEL The assessment of response to chemotherapy was based on the change of individual markers and overall response on the combined assessment of multiple markers. The assessment of response to chemotherapy of each tumor marker was defined as follows. A complete response in tumor marker (CR-m) was defined as a drop to within the normal range of marker after chemotherapy for at least 4 weeks. A partial response in tumor marker (PR-m) was defined as a ~50% reduction in level prior to chemotherapy for at least 4 weeks. No change in tumor marker (NC-m) was defined as a <50% reduction or <50% increase in the tumor marker level for at least 4 weeks. Progressive disease in tumor marker (PD-m) was defined as a >50% increase in level. The assessment of overall response based on combined multiple tumor markers was defined as follows. Overall CR-m was when CR-m was achieved for all markers. Overall PR-m was defined as PR-m achieved for all markers. A combination of CR-m and PR-m was defined as overall PR-m. Overall PD-m was when at least one of three markers was assessed as PD-m. All other situations were defined as overall NC-m. ASSESSMENT OF CHEMOTHERAPEUTIC RESPONSE BY IMAGING Imaging studies were performed prior to and every 4 weeks after the initiation of chemotherapy by using the technique employed for baseline tumor measurement, which included computed tomography and chest X-ray. For those patients with measurable lesions, responses were evaluated according to the World Health Organization (WHO) criteria (15). A complete response (CR) was defined as the disappearance of all evidence of cancerfor at least 4 weeks. A partial response (PR) was defined as a ~50% reduction in the sum of the products of the perpendicular diameters of all lesions for at least 4 weeks, without any evidence of the development of new lesions or of the progression of any lesions. No change (NC) was defined as a <50% reduction or <25% increase in the sum of the products of the perpendicular diameters of all lesions, without any evidence of new lesions. Progressive disease (PD) was defined as a >25% increase in one or more lesions or the appearance of new lesions. Since there are no criteria for assessment of response in primary gastric tumor in the WHO criteria, the response of primary gastric lesion was evaluated according to Table. Correlation between assessment of response to chemotherapy by tumor marker and imaging Assessment by tumor marker Assessment by imaging CR PR NC PD CR-m 0 6 0 PR-m 0 2 0 NC-m 0 0 5 PD-m 0 0 4 4 the response assessment criteria of chemotherapy for gastric carcinoma outlined by the Japanese Research Society of Gastric Cancer, which are based on the macroscopic appearance on X-ray and/or endoscopy (7). Comparison of response assessment by tumor markers and by imaging was expressed in terms of sensitivity, specificity, positive and negative predictive value and diagnostic accuracy, which were defined as follows: sensitivity: true positive/(true positive + false negative); specificity: true negative/(true negative + false positive); positive predictive value: true positive/(true positive + false positive); negative predictive value: true negative/(true negative + false negative); diagnostic accuracy: (true positive + true negative)/(number of responders(cr-m +PR-m) and non-responders(nc-m + PD-m). STATISTICAL ANALYSIS The correlation of assessment of efficacy by tumor markers and by imaging studies was analyzed using the chi-squared test. Survival curves were drawn using the Kaplan-Meier method and analyzed by the log rank test. Comparisons between unpaired groups were made using the Mann-Whitney U-test and correlations assessed using the Spearman rank order correlation. Results were considered significant if the p value was <0.05. RESULTS In 26 patients, the pretreatment serum CEA, CA19-9 and CAl25 values were elevated beyond the normal cut-off value in 14 (54%), 11 (42%) and 17 (65%) patients, respectively. The number of patients in which levels were elevated for a single marker of CEA, CA19-9 and CA125 was four, three and six, respectively. Half of the patients (1 patients) had more than two elevated markers. The combination patterns of elevated markers are summarized in Table 1. The median value and the range of pretreatment CEA, CA19-9 and CA125 were 4.0 (5.5-2028) ng/ml, 16 (5.8-87.5) Vim I and 140.7 (42.5-577) Uzml, respectively. There was no correlation between an elevated marker and site of metastatic disease. In the subpopulation of patients who had more than two elevated markers prior to chemotherapy, there was a moderate correlation between the level of CEA and CA19-9 or CA125

Jpn J Clin OncoI1999;29(11) 55 Table 4. Sensitivity, specificity and positive and negative predictive value of serial tumor marker measurement in evaluating a partial response (PR) as demonstrated by imaging studies CEA CA19-9 CA125 Combined markers Sensitivity (%) 4/4 (l00) / (l00) 2/2 (100) 8/8 (l00) Specificity (%) 6/10 (60) 6/8 (75) 8/15 (5) 11/18 (72) Positive predictive value (%) 4/8 (50) /5 (60) 2/9 (22) 8/12 (67) Negative predictive value (%) 6/6 (l00) 6/6 (l00) 8/8 (100) 14/14 (100) Diagnostic accuracy (%) 10/14 (71) 9/11 (82) 10/17 (59) 19/26 (7) Table 5. Sensitivity, specificity and positive and negative predictive value of serial tumor marker measurement in evaluating a progressive disease (PD) as demonstrated by imaging studies CEA CA19-9 CA125 Combined markers Sensitivity (%) 1/2 (50) 2/ (66) 2/5 (40) 4/5 (80) Specificity (%) 12/12 (100) 7/8 (88) 9/12 (75) 17/21 (81) Positive predictive value (%) 1/1 (100) 2/ (67) 2/5 (40) 4/8 (50) Negative predictive value (%) 12/1 (92) 7/8 (88) 9/12 (75) 17/18 (94) Diagnostic accuracy (%) 1/14 (9) 9/11 (82) 11/17 (65) 21/26 (81) throughout the chemotherapy (Spearman p = 0.46, p < 0.001 and p =0.42, p < 0.001, respectively). However, there was no significant correlation between level of CAl25 and CA19-9 (Spearman p =0.20,p =0.14). The correlation between overall assessment of response by tumor markers and imaging studies is summarized in Table. Overall CR-m and PR-m were achieved in seven (27%) and five patients (19%), respectively, and the response rate by tumor marker assessment was 46.1%. The median period from initiation of chemotherapy to drop in tumor marker level by <50% was 4 days (range 9-77 days). Eight PRs (1%) were achieved by imaging studies but there were no CRs by imaging. Agreement in assessment by tumor markers and imaging study was observed in 11 patients (42%). However, when patients were categorized as responders (CR and PR or CR-m and PR-m) or non-responders (NC and PD or NC-m and PD-m), a significant correlation was observed between the assessment of response by tumor markers and by imaging (p = 0.0012; chi-squared test). There was no significant difference in clinicopathological features and baseline value of each tumor marker between the responders and non-responders (data not shown). Comparisons of response assessment by tumor markers with that by imaging are shown in Tables 4 and 5. A fall in any of the markers (CR-m and PR-m) was highly sensitive in the prediction of a partial response (PR) in the imaging study (100% sensitivity for all three markers). However, specificity was lower with a positive predictive value of 50% for fall in CEA, 60% for fall in CA19-9 and 22% for fall in CA125. The negative predictive value of drop in marker level for partial response was 100% for all three markers. Assessment by multiple markers improved the specificity and positive predictive value with sensitivity and negative predictive value remaining the same. The diagnostic accuracy of markers in predicting PR in imaging study was 7% (19/26) using three markers in combination, while it was as high as 82% (9/11) using CA19-9. (Table 4) Considering the prediction of progressive disease (PD) in the imaging study, the sensitivity of rising marker levels was 50% for CEA, 66% for CA19-9 and 40% for CA125. The specificity was 100% for CEA, 88% for CA19-9 and 75% for CA125, higher than the sensitivity in each case, resulting in a high negative predictive value, 92% for CEA, 88% for CA19-9 and 75% for CA125. Assessment by multiple markers improved the sensitivity (80%) in the prediction of progressive disease. The diagnostic accuracy of markers in predicting PD in the imaging study was 81% (21/ 26) using three markers in combination, while it reached up to 9% (1/14) in CEA (Table 5). The median survival time of all 26 patients was 225 days and the 1 year survival rate was 20%. Fig. 1 compares the survival times for responders (CR-m and PR-m) and non-responders (NC-m and PD-m) defined by efficacy assessment by tumor marker. A significant difference in survival was observed between responders and non-responders (p = 0.0056; log rank test) with a median survival time of 01 versus 127 days. DISCUSSION The WHO criteria are commonly used to assess therapeutic efficacy in the treatment of solid tumors. Criteria for assessing the response to chemotherapy based on the objective tumor shrinkage are essential for clinical trials to select and develop

554 Use oftumor markers in monitoring chemotherapy... itj..~ S0.8 I,... IzI ~0.6 ~ 0.4 f'j.0 ~ 0.2 o NC-m + PD-m P=0.0056 o 100 200 00 400 500 600 700 Days after initiation of chemotherapy Figure 1. Survival curves using the Kaplan-Meiermethod for responders (CRm and PR-m; open triangle) and non-responders (NC-m and PD-m; open circle) to systemic chemotherapy assessed by change in tumor markers, which show a significant difference between responders and non-responders. (p = 0.0056; log rank test). new anti-tumor agents. The RECIST criteria, which are a newly revised version of the WHO criteria, is also based on the objective tumor shrinkage. A 'measurable' lesion is required to assess the objective response and tumor markers alone cannot be used to assess response in such situations. However, in daily clinical practice of oncology, patients do not always have a 'measurable' lesion. Clinicians sometimes have to evaluate the tumor response based on a subjective medical judgement from clinical symptoms and laboratory data. It is known that the serum levels of various tumor markers such as CEA, CAI9-9, CA12S and others are elevated in patients with gastrointestinal cancers. However, the clinical use of such markers in the management of these patients still remains controversial (1,14). Elevated CEA, CA19-9 and CA 125 levels have been reported in around 0-60% of patients with advanced gastric cancer and our results were also within this range (8-10). Although the positive rate of each tumor marker was around 50%, in 90% (28 out of 1) of the patients who received systemic chemotherapy, the level of at least one of the tumor markers CEA, CA19-9 and CA12S was elevated. It is well known that elevated marker levels decrease after curative resection of tumor only to elevate again on recurrence (19,20). We therefore postulated that the change in these tumor markers would reflect the relative tumor burden in individual patients after chemotherapy when the marker levels had been elevated prior to chemotherapy. We analyzed here the serial change in tumor marker values from the viewpoint of chemotherapeutic response. Although the number of patients evaluated here was too small to confirm the reliability of the response assessment to chemotherapy based on serial change of tumor marker level, the results suggest the clinical usefulness of tumor markers in the monitoring of chemotherapeutic efficacy which enables us to distinguish clearly between responders and non-responders. In the prediction of partial response, the sensitivity and negative predictive values of falling tumor marker levels were 100% for all three markers and tumor shrinkage always accompanied a drop in marker levels in the present study. In contrast to the high sensitivity of the three markers for prediction of tumor shrinkage, the specificity and positive predictive value of single markers was low. Additionally, the sensitivity of each marker for the prediction of progressive disease was also low. These observations seem to be due to the heterogeneity of tumor cells in individual patients and the assessment of migration between PR and NC in patients with falling marker levels and between NC and PD in patients with rising levels. Hence it should be kept in mind that both overestimation of tumor response and underestimation of progressive disease can occur when the chemotherapeutic efficacy is assessed by change in tumor marker level. As regards the overestimation of responses, this is perhaps explicable on the basis that the treatment had a significant inhibitory effect on the tumor which was insufficient to result in a response on imaging. The underestimation of disease progression, i.e. disease progression without rise in marker level, is a serious problem in clinical practice and is probably due to the heterogeneity of tumor cells. CEA, CA19-9 and CA12S combined were elevated in a variety of patterns in the patients studied. Interestingly, when more than two markers were elevated, changes did not always occur in synchrony. This observation may reflect in part the biological heterogeneity of tumor cells and the heterogeneity of responsiveness to chemotherapy in the same patient. Concerning the diagnostic accuracy, CA19-9 was the most reliable marker for the prediction of PR (82%) and CEA for the prediction for PD (9%). However, CAl2S was less reliable in the prediction of clinical response in imaging than CEA and CAI9-9. In clinical practice, one should select the most useful marker from the viewpoint of cost effectiveness when the patient is positive for multiple markers. We recommend that one of CEA and CA 19-9 should be measured when both markers are increased before treatment. The responders to chemotherapy showed falling tumor marker levels in the early period from initiation of treatment. The median interval between initiation of treatment and 50% decrease in tumor marker was 4 days (range 9-77 days). In addition, a fall in tumor marker level was significantly associated with survival of the patients. Falling tumor marker levels may be an early predictor of long-term survival in patients treated with systemic chemotherapy. The nunber of patients analyzed here is very small and it is clear that tumor markers cannot replace the use of imaging in the management and assessment of patients with advanced gastric cancer treated with systemic chemotherapy, especially in clinical trials to evaluate anti-tumor activity. However, the present study suggests that tumor markers may be sensitive and useful for monitoring the response to chemotherapy and can be used to assess chemotherapeutic efficacy in daily clinical practice. In those cases in which disease is difficult to evaluate by imaging study, such as diffuse peritoneal dissemination, tumor markers may be used as a helpful means

Jpn J Clin OncoI1999;29(l1) 555 of monitoring chemotherapeutic efficacy. When these findings are confirmed by the analysis with a large sample size, the assessment of chemotherapeutic efficacy by tumor markers will provide benefits in cost and convenience. In conclusion, the measurement of tumor markers might be useful in the monitoring of response and in the prediction of prognosis in patients with advanced gastric cancer treated with systemic chemotherapy. Further studies are required to confirm these findings. References 1. Murad AM, Santiago FF, Petroianu A, Rocha PR, Rodrigues MA, Rausch M. Modified therapy with 5-fluorouracil, doxorubicin and methotrexate in advanced gastric cancer. Cancer 199;72:7-41. 2. Glimelius B, Ekstrom K, Hoffman K, Graf W, Sjoden PO, Haglund U, et al. Randomized comparison between chemotherapy plus best supportive care with best supportive care in advanced gastric cancer. Ann Oncol 1997;8:16-8. 2. Pyrhonen S, Kuitunen T, Nyandoto P, Kouri M. Randomised comparison of fluorouracil, epidoxorubicin and methotrexate (FEMTX) plus support ive care with supportive care alone in patients with non-resectable gastric cancer. Br J Cancer 1995;71 :587-91. 4. Wils J. The treatment of advanced gastric cancer. Semin Oncol 1996;2:97-406. 5. Shirao K, Shimada Y, Kondo H, Saito D, Yamao T, Ono H, et al. Phase I n study of irinotecan hydrochloride combined with cisplatin in patients with advanced gastric cancer. J Clin Onco11997; 15:921-7. 6. Ohtsu A, Shimada Y, Yoshida S, Saito H, Seki S, Morise K, et al. Phase II study of protracted infusional 5-fluorouracil combined with cisplatinum for advanced gastric cancer: report from the Japan Clinical Oncology Group (JCOG). Eur J Cancer 1994;0A:2091-. 7. World Health Organization: WHO Handbook for Reporting Results of Cancer Treatment. WHO Offset Publication No. 48. Geneva: World Health Organization 1979. 8. Japanese Research Society for Gastric Cancer. Japanese Classification of Gastric Cancer. Tokyo: Kanehara 1995. 9. Sakamoto K, Haga Y, Yoshimura R, Egami H, Yokoyama Y, Akagi M. Comparative effectiveness of the tumour diagnostics, CA 19-9, CA125 and carcinoembryonic antigen in patients with diseases of the digestive system. Gut 1987;28:2-9. ro. Dittrich C, Jakesz R, Havelec L, Lenzhofer R, Breyer S, Moser K. Carcinoembryonic antigen (CEA) plasma level determination in the management of gastric cancer patients. Cancer Detect Prev 1985;8:181-7. 11. Wobbes T, Thomas CM, Segers MF, Nagengast FM. Evaluation of seven tumor markers (CA 50, CA 19-9, CA 19-9TruQuant, CA 72-4, CA 195, carcinoembryonic antigen and tissue polypeptide antigen) in the pretreatment sera of patients with gastric carcinoma. Cancer 1992;69:206-41. 12. Ward U, Primrose IN, Finan PJ, Perren TJ, Selby P, Purves DA, et al. The use of tumour markers CEA, CA-195 and CA-242 in evaluating the response to chemotherapy in patients with advanced colorectal cancer. Br J Cancer 199;67: 112-5. 1. Komek G, Schenk T, Raderer M, Djavammad M, Scheithauer W. Tissue polypeptide-specific antigen (TPS) in monitoring palliative treatment response of patients with gastrointestinal tumours. Br J Cancer 1995;71:182-5. 14. Moertel CG, Fleming TR, Macdonald JS, Haller DG, Laurie JA, Tangen C. An evaluation of the carcinoembryonic antigen (CEA) test for monitoring patients with resected colon cancer. JAm Med Assoc 199;270:94-7. 15. Grem J. The prognostic importance of tumor markers in adenocarcinomas of the gastrointestinal tract. Curl' Opin OncoI1997;9:80-7. 16. Murakami M, Ota K, Miyazaki T, Niitsu Y, Wakui A, Yokoyama M, et al. Sequential methotrexate-5-fluorouracil (MTX-5-FU) treatment of patients with advanced gastric and colorectal cancer. Sequential Methotrexate-5-FU Study Group. Gan To Kagaku Ryoho 1987;14:2482-90 (in Japanese). 17. Sakata Y, Shimada Y, Yoshino M, Kambe M, Futatsuki K, Nakao I, et al. A late phase II study of CPT-II, irinotecan hydrochloride, in patients with advanced pancreatic cancer. CPT-II Study Group on Gastrointestinal Cancer. Gan To Kagaku Ryoho 1994;21:109-46 (in Japanese). 18. Shimizu Y, Umezawa S, Hasumi K. Successful treatment of clear cell adenocarcinoma of the ovary (OCCA) with a combination of CPT-II and mitomycin C. Gan To Kagaku Ryoho 1996;2:587-9 (in Japanese). 19. Wanebo JH, Steams M, Schwartz MK. Use of CEA as an indicator of early recurrence and as a guide to a selected second-look procedure in patients with colorectal cancer. Ann Surg 1978;188:481-9. 20. Attiyeh FF, Steams MW Jr. Second-look laparotomy based on CEA elevations in colorectal cancer. Cancer 1981;47:2119-25.