Prospective randomised trial of early cytotoxic therapy for recurrent colorectal carcinoma detected by serum CEA.

Of 663 patients treated with radical surgery for colorectal cancer, 52 showed a progressive rise in serum carcinoembryonic antigen (CEA) with no other evidence of recurrent disease and were randomised in a prospective study of chemotherapy. Twenty six patients in the treatment group received 5FU and methyl CCNU from the time of randomisation and the remaining 26 controls were given further therapy only if there were clinical indications. All patients were followed for five years or until their death and all but one (control) developed clinical evidence of recurrence. Overall there was no significant difference between the two groups with respect to disease free interval and survival. Whereas the rise in CEA in controls was generally progressive, marked inflections on the CEA curves were seen in the majority of patients receiving early treatment. Eight of 26 treated patients showed a fall in CEA of greater than 20% two months after starting therapy. These patients had a median disease free interval of 90 weeks and a median survival of 107 weeks, these figures being longer than those of treated patients who did not show a fall in CEA and control patients. The serum CEA therefore appeared to give important prognostic information in patients receiving cytotoxic treatment. Early therapy was generally well tolerated.