Plasma prolactin in patients with colorectal cancer. Value in follow-up and as a prognosticator.

Background: Preoperative plasma prolactin and carcinoembryonic antigen (CEA) levels were assessed to monitor disease recurrence and to identify low-risk and high-risk patients with Dukes B or C colorectal cancer.
Methods: Prolactin and CEA were estimated by radioimmunoassay method. Blood samples were collected preoperatively and sequentially thereafter from patients with colorectal cancer (N = 114); the samples were compared with samples from age-matched healthy control subjects (smokers and nonsmokers, N = 45). For rest of the analysis, patients with Dukes A disease (N = 7) were not included because of the small number. In monitoring recurrences, the criteria for positive test for the two markers was a continual increase in the marker level after an initial decrease or persistent high level of the marker. These were the indicators of relapse or no response to treatment. To determine the efficacy of the preoperative markers, the patients were grouped according to disease status at the end of 3 years, i.e., patients who had response to the treatment modalities (N = 52) and patients who later had progressive disease (N = 55). To determine the prognostic significance of preoperative marker levels, the patients were divided according to the cutoff levels (upper normal limits); for prolactin the cutoff level was 20.0 ng/ml plasma, and for CEA it was 5.0 ng/ml plasma.
Results: Both of the markers were significantly high in patients with colorectal cancer compared with the markers of their respective control subjects (P < 0.0001). In monitoring disease course, the predictive power of prolactin was 100%, whereas that of CEA was 66%. Prolactin showed a lead time of 2-3 months. Preoperative prolactin levels were significantly higher in patients who later had progressive disease (P < 0.001) than in patients who had response to the treatments. However, such an intergroup variation was not observed for CEA. Patients with preoperative levels of prolactin greater than 20.0 ng/ml had shorter overall survival times than did those with prolactin levels less than 20.0 ng/ml plasma; such a trend was not observed for patients with CEA levels less than 5.0 ng/ml and those with CEA levels greater than 5.0 ng/ml plasma.
Conclusion: Prolactin is a better overall marker than is CEA in patients with Dukes B or C colorectal cancer. The authors recommend the use of plasma prolactin levels to help identify low-risk and high-risk patient subgroups so that high-risk patients may be followed up more intensely and treated accordingly. Hyperprolactinemic patients with Dukes B or C disease have shortened survival time.