Significance of the tumour markers CA 125 II, CA 72-4, CASA and CYFRA 21-1 in ovarian carcinoma.

We compared the tumour marker CA 72-4 and the new markers CASA and CYFRA 21-1 with the established marker CA 125II in follow-up care and control of efficacy of treatment in ovarian cancer in order to determine whether there are differences in recognizing ovarian carcinomas of different histological type. The investigation was done retrospectively on serum samples frozen at -80 degrees C obtained from 262 subjects, among them 50 healthy women, 53 sera with benign gynecological diseases and 159 sera with ovarian cancer, 72 of them at the time of primary diagnosis. We used commercially available kits: CA 125 II: Centocor RIA, CASA: Medac EIA, CA 72-4: Centocor RIA and CYFRA 21-1: Boehringer EnzymunR ELISA. Fixing specificity at 95% versus benign gynecological diseases as a clinically relevant reference group, we found cut-off values of 160 U/mL for CA 125II, 6.5 U/mL for CASA, 6.8 U/mL for CA 72-4 and 2.4 ng/ml for CYFRA 21-1. Based on this specificity we can compare the corresponding sensitivities at the time of primary diagnosis (n = 72): CA 125 II 47%, CA 72-4 47%; CASA 31% and CYFRA 21-1 44%. Regarding histological types of ovarian carcinomas, we found a sensitivity of 50% for CA 125 II and CASA, 36% for CA 72-4 and 33% for CYFRA 21-1 in serous ovarian cancer (n = 53), 21% for CA 125 II and CASA, 36% for CYFRA 21-1 and 43% for CA 72-4 for mucinous ovarian cancer (n = 27). Serous ovarian carcinomas were classified by higher FIGO-stages than mucinous ovarian carcinomas. Additional sensitivities were found at the time of primary diagnosis for the combination of CA 125 II and CA 72-4 and in serous ovarian cancer for CA 125 II and CASA. According to our results, at the time of primary diagnosis the combined determination of CA 125 II and CA 72-4 is useful. If both are negative, determination of CASA can be helpful. For follow-up care and control of efficacy of treatment the preoperative positive or leading marker is sufficient.