2488 Serum CA 125 Level as a Predictor of Peritoneal Dissemination in Patients with Gastric Carcinoma Bunzo Nakata, M.D. Kosei Hirakawa-YS Chung, Yasuyuki Kato, M.D. Yoshito Yamashita, M.D. Kiyoshi Maeda, M.D. Naoyoshi Onoda, M.D. Tetsuji Sawada, M.D. Michio Sowa, M.D. M.D. First Department of Surgery, Osaka City University Medical School, Osaka, Japan. BACKGROUND. Prediction of peritoneal dissemination is very difficult using current diagnostic tools such as computed tomography, ultrasonography, or various tumor markers. The predictive value of serum CA 125 levels for peritoneal metastasis from gastric carcinoma was studied. METHODS. The sera from 384 patients with gastric carcinoma were measured for CA 125 titer using an immunoradiometric assay. Carcinoembryonic antigen, carbohydrate antigen 19-9, and sialyl-Tn antigen were measured in the same samples. RESULTS. The serum CA 125 level was elevated according to the degree of peritoneal dissemination. The reference value for peritoneal dissemination was determined to be 35 U/mL, resulting in a sensitivity of 39.4%, specificity of 95.7%, and diagnostic accuracy of 90.8%. The diagnostic ability was more reliable than the other imaging modalities including computed tomography and ultrasonography and the other useful tumor markers for gastric carcinoma. The serum CA 125 level was elevated after gastrectomy for approximately 2 months, most likely due to the continuous inflammation of the peritoneum and lost predictive significance for peritoneal dissemination during this period. CONCLUSIONS. Measurement of the serum CA 125 titer may be a powerful predictor of peritoneal metastases in patients with gastric carcinoma. Cancer 1998;83: 2488 –92. © 1998 American Cancer Society. KEYWORDS: carbohydrate antigen 125 (CA 125), gastric carcinoma, peritoneal dissemination, gastrectomy. P Address for reprints: Bunzo Nakata, M.D., First Department of Surgery, Osaka City University Medical School, 1-5-7 Asahimachi, Abeno-ku, Osaka 545-8585, Japan. Received January 29, 1998; revision received April 29, 1998; accepted April 29, 1998. © 1998 American Cancer Society eritoneal dissemination is one of the most lethal events in patients with gastric carcinoma. Gastrectomy generally is not indicated for cases of severe peritoneal dissemination because it does not improve prognosis and can impair the patient’s quality of life. Accordingly, the accurate prediction of peritoneal dissemination is essential to ensure appropriate surgical referral or to plan appropriate treatments including intraperitoneal chemotherapy and hyperthermia. Diagnosis for the disease currently involves recognition of ascites on physical examination, recognition of induration of the pouch of Douglas by digital examination, or the use of abdominal computed tomography (CT) or ultrasonography (US). However, none of these methods has demonstrated a high predictive value. Recently laparoscopy, a minimal invasive procedure, has been utilized to detect peritoneal dissemination. When a patient with gastric carcinoma is believed to have peritoneal dissemination, laparoscopy can be a better method to confirm it than laparotomy. The diagnostic method presented in this article is based on the known distribution of carbohydrate antigen 125 (CA 125) in mesothelial cells of the peritoneum, pleura, and pericardium, as well as in the epithelium of the fallopian tubes, endometrium, and endocervix,1 which suggests that dissemination of gastric Serum CA 125 and Peritoneal Metastasis/Nakata et al. 2489 carcinoma to the peritoneum may affect the serum levels of CA 125. Therefore we investigated whether elevation of the serum CA 125 titer has predictive value for peritoneal metastasis from gastric carcinoma. MATERIALS AND METHODS Patients Three hundred and eighty-four patients with histologically proven gastric carcinoma who underwent surgery between January 1993 and June 1997 at the First Department of Surgery of the Osaka City University Hospital were studied. The patients were comprised of 271 males and 113 females (mean age ⫾ standard deviation [SD], 60.2 ⫾ 10.9 years [range, 30 –93 years]). According to the Japanese classification of gastric carcinoma,2 205 patients had Stage I disease, 48 had Stage II disease, 73 had Stage III disease, and 58 had Stage IV disease. Peripheral blood samples were obtained from each patient within 1 week before surgery and again at the serial times after surgery. Assay The sera were assayed for CA 125 with an immunoradiometric assay using a Centrocor CA 125 II IRMA kit (Centocor Diagnostics Division, Malvern, PA). Serum carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA 19-9) were measured by a counting immunoassay using a Ranream CEA kit (TOA Medical Electronics Co., Kobe, Japan) and a Ranream CA 19-9 kit (Toray-Fuji Bionics, Tokyo). Serum sialyl-Tn (STN) levels were measured by an immunoradiometric competitive inhibition assay using an S-Tn Otsuka kit (Otsuka Assay Laboratories, Tokushima, Japan). The cutoff values of CA 125, CEA, CA 19-9, and STN were set as recommended by the respective manufacturers: 35 U/mL, 6.5 ng/mL, 37 U/mL, and 45 U/mL, respectively. Statistical Analysis The nonparametric Mann–Whitney U test was used for comparison of the two independent groups and Kruskal–Wallis one-way analysis was performed for the multiple comparison tests. The significance of association was determined by the chi-square test. Survivor analysis was estimated by the Kaplan–Meier method and examined by the log rank test. P values ⬍ 0.05 were considered statistically significant. RESULTS Correlation between Serum CA 125 Values and Peritoneal Dissemination In the patients with gastric carcinoma, serum CA 125 values ranged from 2– 620 U/mL with a median value FIGURE 1. Serum CA 125 distribution according to peritoneal dissemination. There was a significant difference in the serum levels between the patients with and without peritoneal disseminations by the Mann–Whitney U test (P ⬍ 0.0001). The nonparametric multiple comparison tests (Kruskal–Wallis oneway analysis) indicated serum CA 125 levels were increased significantly according to the presence or absence of peritoneal dissemination (P ⬍ 0.0001). Horizontal line: median value; column: interquatile range; top bar: the 90th percentile value; bottom bar: the 10th percentile value; P0: no peritoneal dissemination; P1: metastases to the adjacent peritoneal but not the distant peritoneum; P2: a few metastases to the distant peritoneum; P3: numerous metastases to the distant peritoneum. of 9 U/mL. Twenty-eight patients (7.3%) had values above the cutoff value of 35 U/mL. Thirty-three patients were diagnosed with peritoneal disseminations by laparotomy. Figure 1 shows serum CA 125 values according to peritoneal dissemination. These values were distributed nonparametrically; therefore, Kruskal– Wallis one-way analysis was performed for the multiple comparison of CA 125 values among these various levels of peritoneal dissemination. Serum CA 125 levels were elevated significantly between P0 and P3 (P ⬍ 0.0001). The median value of serum CA 125 for patients with peritoneal dissemination was significantly higher than that of patients without peritoneal metastases using the Mann–Whitney U test (23 U/mL vs. 9 U/mL; P ⬍ 0.0001). The mean ⫾ SD values of serum CA 125 for patients without and those with peritoneal metastases were 12.9 ⫾ 14.5 U/mL and 67.5 ⫾ 129.5 U/mL (P1, 25.0 ⫾ 27.2 U/mL; P2, 28.8 ⫾ 35.9 U/mL; and P3, 105.4 ⫾ 171.6 U/mL), respectively. Comparison of Diagnostic Abilities of Various Examinations for Peritoneal Dissemination CT and US revealed peritoneal dissemination in 7 and 6 respectively, of 33 patients. By physical examination, including digital examination, only two patients had peritoneal dissemination detected. When the patients were divided by degree of peritoneal involvement, those with metastases to the adjacent peritoneum but 2490 CANCER December 15, 1998 / Volume 83 / Number 12 TABLE 1 Comparison of Various Examinations for Patients with Peritoneal Dissemination from Gastric Carcinoma Degree of dissemination CA 125 CT US Digital examination Palpation of abdomen P1 (7) P2 (9) P3 (17) Total (33) 28.6 22.2 53.0 39.4 0 22.2 29.4 21.2 0 11.1 29.4 18.2 0 11.1 5.9 6.1 0 0 11.8 6.1 CT: computed tomography; US: ultrasonography; P1: metastases to the adjacent peritoneal but not distant peritoneum; P2: a few metastases to the distant peritoneum; P3: numerous metastses to the distant peritoneum. Numbers are percentages of correct diagnosis by each examination. Numbers in parentheses are the numbers of patients in each peritoneal dissemination status. Peritoneal dissemination was diagnosed for patients whose CA 125 value was the cutoff value of 35 U/mL; whose computed tomography or ultrasonography showed peritoneal tumor or ascites; whose digital examination showed induration of the pouch of Douglas; or whose abdominal palpation showed ascites or induration other than from the main tumor. not the distant peritoneum (P1) could not be diagnosed by any of these methods. Even for patients with numerous metastases to the distant peritoneum (P3), the diagnostic ability of these examinations was very poor. However, using serum CA 125 with a cutoff value of 35 U/mL, peritoneal metastases were predicted in 2 patients with P1, 2 patients with P2 (a few metastases to the distant peritoneum), and 9 patients with P3 (Table 1). FIGURE 2. Comparison of diagnostic ability of peritoneal dissemination by each tumor marker. Solid bar: sensitivity; Open bar: specificity; Shaded bar: diagnostic accuracy. Sensitivity ⫽ (patients with peritoneal dissemination with positive tests/all patients with peritoneal dissemination tested) ⫻ 100; Specificity ⫽ (patients without peritoneal dissemination with negative tests/all patients without peritoneal dissemination tested) ⫻ 100; Diagnostic accuracy ⫽ ((patients with peritoneal dissemination with positive tests ⫹ patients without peritoneal dissemination with negative tests)/all patients tested) ⫻ 100. CEA: carcinoembryonic antigen; STN: sialyl-Tn antigen. compared with patients with a low CA 125 serum level (Fig. 3). Serum CA 125 Value after Gastrectomy Comparison of Tumor Markers for the Diagnosis of Peritoneal Dissemination Figure 2 shows a comparison of various tumor markers for peritoneal dissemination. For CA 125, CEA, CA 19-9, and STN, each cutoff value was used as a reference value for peritoneal dissemination. The sensitivity of CA 125 was 39.4% (13 of 33 patients) and the sera from 20 patients with peritoneal dissemination were found to be false-negative. The rate of elevated CA 125 levels in the patients without peritoneal dissemination was 4.3% (15 of 351 patients). CA 125 had the best specificity (95.7%) and accuracy (90.9%), and the second best sensitivity next to STN among these markers. STN had the worst specificity and accuracy and CA 19-9 had the worst sensitivity among the markers tested. CEA had the same sensitivity as CA 125; however, the specificity and accuracy were not as high. Moreover, CA 125 had the highest odds ratio for predicting peritoneal dissemination among the markers tested (chi-square test) (Table 2). Prognostic Value of Serum CA 125 Kaplan–Meier curves indicated a poorer prognosis for patients with a high CA 125 serum level (ⱖ 35 U/mL) The sera of 20 patients without peritoneal metastases who underwent curative surgery were measured for CA 125 levels. All patients showed low levels of serum CA 125 (⬍ 20 U/mL) prior to the surgery; however, in 18 patients serum CA 125 was highly elevated postoperatively. The CA 125 level was shown to peak at 2–3 weeks after gastrectomy, and this elevation continued for approximately 2 months (Fig. 4). DISCUSSION The CA 125 antigen is detectable by a monoclonal antibody OC125, which is produced by the immunization of mice with ovarian carcinoma cells.3 This antigen has been detected in the serum of 80% of patients with epithelial ovarian carcinoma.4,5 It also has been established that serum CA 125 is useful for monitoring disease recurrence, determining the chemotherapeutic effects and predicting the prognosis of epithelial ovarian carcinoma.6,7 Patients with advanced stage lung carcinoma,8,9 pancreatic carcinoma,10,11 or endometrial carcinoma12 also have shown elevated levels of serum CA 125. In the case of gastric carcinoma, some investigators have demonstrated that elevated serum CA 125 implied poor prognosis Serum CA 125 and Peritoneal Metastasis/Nakata et al. 2491 TABLE 2 Comparison of the Diagnostic Ability of Serum Tumor Marker Levels for Peritoneal Dissemination Peritoneal dissemination Tumor marker CA 125 CEA CA 19-9 STN Positivea Negativeb Positive Negative Positive Negative Positive Negative Positive (no. of patients) Negative (no. of patients) 13 20 13 20 9 24 17 16 15 336 81 270 37 314 86 265 Odds ratio 95% CI Chi-square test P value 14.56 3.23–65.5 12.17 ⬍ 0.0001 2.17 1.05–4.50 4.34 0.0371 3.18 1.43–7.08 8.01 0.0047 3.27 1.68–6.37 12.17 0.0005 95% CI: 95% confidence interval; CEA: carcinoembryonic antigen; STN: sialyl-Tn antigen. a Positive indicates the serum levels of CA 125, carcinoembryonic antigen, CA 19-9, and sialyl-Tn antigen were ⱖ the cutoff value of 35 U/mL, 6.5 ng/mL, 37 U/mL, and 45 U/mL, respectively. b Negative indicates the serum levels of CA 125, carcinoembryonic antigen, CA 19-9, and sialyl-Tn antigen were ⬍35 U/mL, 6.5 ng/mL, 37 U/mL, and 45 U/mL, respectively. FIGURE 3. Probability of survival in patients with gastric carcinoma in relation to their serum CA 125 levels. and aggressive biology.13,14 However, the association between the serum CA 125 level and peritoneal dissemination from gastric carcinoma has not been elucidated fully. The CA 125 antigen has been observed in the peritoneum, particularly in areas of inflammation and adhesion.1 Peritoneal dissemination may cause inflammation of the peritoneum; therefore, one would expect that an elevation of serum CA 125 would be observed in patients with peritoneal dissemination. Our data indicated a strong association between the serum CA 125 level and peritoneal dissemination. We also demonstrated that the serum CA 125 level was a more reliable predictor of peritoneal dissemination than the other imaging studies and tumor markers currently in use. Moreover, a cutoff value of 35 U/mL was found to be useful for predicting poor prognosis in patients with this disease. FIGURE 4. The variance in the serum CA 125 titer after gastrectomy in the patients with gastric carcinoma who had no peritoneal dissemination. Pre: preoperative period; 1M: 1 month postoperatively; 2M: 2 months postoperatively; 3M: 3 months postoperatively; 4M: 4 months postoperatively. In our 20 patients without peritoneal dissemination, serum levels of CA 125 were elevated for approximately 2 months postoperatively, after which time they returned to within normal range. The literature shows that the natural half-life of CA 125 is approximately 5 days.8,15 The long postoperative duration of the high serum CA 125 titer may be a result of continuous inflammation in the peritoneum and serosa after gastrectomy. These results suggest that serum CA 125 levels are not a specific indicator of peritoneal dissemination during the first 2 months after gastrectomy, and therefore this marker should be measured ⱖ2 months postoperatively. It recently has been reported that molecular biologic substances such as CD44H and integrin may 2492 CANCER December 15, 1998 / Volume 83 / Number 12 mediate the attachment of gastric carcinoma cells to mesothelial cells.16,17 It also has been reported that transforming growth factor-␤18 and hepatocyte growth factor19 may play a role in promoting peritoneal dissemination. The decrement of E-cadherin, which plays a major role in the maintenance of intercellular adhesion in epithelial cells, has been reported to be associated with peritoneal dissemination.20 In the future, these substances also may prove useful as markers for peritoneal dissemination, although currently these findings have been observed only at an experimental level. The serum CA 125 level may be useful in predicting peritoneal dissemination, a negative value predicts with 95.7% confidence that there will not be carcinomatosis at exploration, and an elevated CA 125 level will predict peritoneal dissemination in 39.4% of those patients. Therefore, surgeons can be better prepared for alternative therapy at the time of exploration, such as intraperitoneal chemotherapy. Laparoscopic examination may be another setting prior to chemotherapy in which those patients with highly suspicious peritoneal dissemination may be detected by elevated serum CA 125 levels. However, because the serum CA 125 level increases for approximately 2 months after gastrectomy, it cannot be used for the prediction of peritoneal dissemination during this period. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES 1. 2. 3. 4. 5. 6. Kabawat SE, Bast RC, Bhan AK, Welch WR, Knapp RC, Colvin RB. Tissue distribution of a coelomic-epitheliumrelated antigen recognized by the monoclonal antibody OC125. Int J Gynecol Pathol 1983;2:273– 85. Japanese Research Society for Gastric Cancer. Japanese classification of gastric carcinoma. 1st English edition. Tokyo: Kanehara, 1995. Bast RC, Feeney M, Lazarus H, Nadler LM, Colvin RB, Knapp RC. Reactivity of a monoclonal antibody with human ovarian carcinoma. J Clin Invest 1981;68:1331–7. Canney PA, Moore M, Wilkinson PM, James RD. Ovarian cancer antigen CA125: a prospective clinical assessment of its role as a tumour marker. Br J Cancer 1984;50:765–9. Klung TL, Bast RC, Niloff JM, Knapp RC, Zurawski VR Jr. Monoclonal antibody immunoradiometric assay for an antigenic determinant (CA125) associated with human epithelial ovarian carcinomas. Cancer Res 1984;44:1048 –53. Fisken J, Leonard RC, Stewart M, Beattie GJ, Sturgeon C, Aspinall L, et al. The prognostic value of early CA125 serum assay in epithelial ovarian carcinoma. Br J Cancer 1993;68: 140 –5. 16. 17. 18. 19. 20. Parker D, Bradley C, Bogle SM, Lay J, Masood M, Hancock AK, et al. Serum albumin and CA125 are powerful predictors of survival in epithelial ovarian cancer. Br J Obstet Gynaecol 1994;101:883–93. Kimura Y, Fujii T, Hamamoto K, Miyagawa N, Kataoka M, Iio A. Serum CA125 level is a good prognostic indicator in lung cancer. Br J Cancer 1990;62:676 – 8. Diez M, Gomez A, Hernando F, Ortega MD, Maestro ML, Torres A, et al. Serum CEA, CA125, and SCC antigens and tumor recurrence in resectable non-small cell lung cancer. Int J Biol Markers 1995;10:5–10. Haglund C. Tumour marker antigen CA125 in pancreatic cancer: a comparison with CA19-9 and CEA. Br J Cancer 1986;54:897–901. Sakamato K, Haga Y, Yoshimura R, Egami H, Yokoyama Y, Akagi M. Combination effectiveness of the tumour diagnostics, CA19-9, CA125 and carcinoembryonic antigen in patients with diseases of the digestive system. Gut 1987;28: 323–9. Takeshima N, Shimizu Y, Umezawa S, Hirai Y, Chen JT, Fujimoto I, et al. Combined assay of serum levels of CA125 and CA19-9 in endometrial carcinoma. Gynecol Oncol 1994; 54:321– 6. Haga Y, Sakamoto K, Egami H, Yoshimura R, Mori K, Akagi M. Clinical significance of serum CA125 values in patients with cancers of the digestive system. Am J Med Sci 1986;292: 30 – 4. Webb A, Scott-Mackie P, Cunningham D, Norman A, Andreyev J, O’Brien M, et al. The prognostic value of serum and immunohistochemical tumor markers in advanced gastric cancer. Eur J Cancer 1996;32:63– 8. Cruickshank DJ, Terry PB, Fullerton WT. The potential value of CA125 as a tumor marker in small volume, non-evaluable epithelial ovarian cancer. Int J Biol Markers 1991;6:247–52. Nishimura S, Chung YS, Yashiro M, Inoue T, Sowa M. CD44H plays an important role in peritoneal dissemination of scirrhous gastric cancer cells. Jpn J Cancer Res 1996;87: 1235– 44. Nishimura S, Chung YS, Yashiro M, Inoue T, Sowa M. Role of ␣2␤1- and ␣3␤1-integrin in the peritoneal implantation of scirrhous gastric carcinoma. Br J Cancer 1996;74:1406 –12. Nakashio T, Narita T, Akiyama S, Kasai Y, Kondo K, Ito K, et al. Adhesion molecules and TGF-␤ are involved in the peritoneal dissemination of NUGC-4 human gastric cancer cells. Int J Cancer 1997;70:612– 8. Yashiro M, Chung YS, Inoue T, Nishimura S, Matsuoka T, Fujihara T, et al. Hepatocyte growth factor (HGF) produced by peritoneal fibroblasts may affect mesothelial cell morphology and promote peritoneal dissemination. Int J Cancer 1996;67:289 –93. Yonemura Y, Ninomiya I, Kaji M, Sugiyama K, Fujimura T, Tsuchihara K, et al. Decreased E-cadherin expression correlates with poor survival in patients with gastric cancer. Anal Cell Pathol 1995;8:177–90.