Summary Plasma prolactin (PRL) and carcinoembryonic antigen (CEA) were measured by radioimmunoassay in 74 patients with adenocarcinoma of colon and rectum. The markers were correlated with disease stage, histological grade and progression/remission of disease. The circulating preoperative median PRL and CEA levels were significantly higher in colorectal cancer patients than in their respective controls. PRL was elevated in all Dukes stages and in all histological grades of the tumour whereas the rise in CEA was more pronounced in Dukes D. Out of 74 patients, 29% (21/74) developed recurrent disease and 31% (23/74) responded to the treatment. With regard to monitoring recurrence(s), the predictive value of PRL was 94% which was significantly greater than that of CEA which was only 62%. In patients who developed liver metastases PRL remained elevated whereas CEA showed more than 100-fold increase. Therefore, we feel that CEA is a better marker for monitoring patients who developed liver metastases. From our results, we suggest that PRL can be used as a better overall marker for detecting recurrence(s) in patients with colorectal adenocarcinoma.Recently, we have published data on circulating prolactin levels in patients with breast cancer (80% of these had advanced disease i.e. with stage III and IV). The data mainly concern relationship between circulating prolactin and histologic grade, estrogen-and progesterone-receptor (ER, PR) and 2 years postoperative survival (Bhatavdekar et al., 1990a). We have also found plasma prolactin useful both as an indicator of disease progression and as short-term prognosticator in patients with advanced breast cancer (Bhatavdekar et al., 1990b;1992). In light of the interesting and convincing results obtained by us in breast cancer patients, we have now tested the significance of prolactin in colorectal cancer, another common cancer in this region, by comparing simultaneously prolactin results with those of CEA.In this study therefore, we have compared the sensitivity and specificity of prolactin and CEA and thus the relative usefulness of these markers in monitoring recurrences in patients with colorectal adenocarcinomas. In addition, plasma prolactin and CEA levels were also correlated with disease stage and histologic grade. Plasma PRL and CEA were assayed using double antibody RIA kits (Diagnostic Products Co., USA). The assays were performed in duplicate with an intra-and inter-assay coefficient of variation (CV) of 3-5% and 5-8% respectively. PRL values > 15.0 ng ml-' for males, > 20.0 ng ml-' for premenopausal and > 10.0 ng ml-' for postmenopausal females were considered for % elevation. CEA levels above 5.0 ng ml-' was regarded as % elevated.Criteria for positive tests were: continual rise in the marker level after an initial fall or persistent high level of the marker as an indicator of relapse and/or no response to treatment.
Statistical analysisThe statistical significance of differences between various groups was calculated by Mann-Whitney U-test. a-value