Abstract
Preoperative CEA and CA 19–9 levels have been used in the past as prognostic indicators in colorectal cancer, but Dukes’ stage is still considered to be the most important prognostic factor. Recent survival estimates may have been influenced by the fact that in the last decade adjuvant chemotherapy and postoperative irradiation have been included in the routine management of advanced-stage disease. In a heterogeneous Kuwaiti population higher reference levels (95th percentile) of CEA and CA 19–9 have been found than those usually employed. In the present study 62 patients with Dukes’ stage B + C could be analyzed for two-year disease-free survival (DFS). Relapse was observed in 19 patients, 28 patients were disease free and 15 patients with censored observations were included. No significant difference in DFS was observed in Dukes’ B (69%) versus Dukes’ C (48%) patients (p=0.09). On the other hand, Dukes’ stage B+C patients with elevated preoperative levels of CEA or CA 19–9 had a significantly poorer DFS than patients with normal levels. For CEA levels below or above the cutoff the DFS was 74% versus 23% (p=0.003); for CA 19–9 levels below or above the cutoff the DFS was 71% versus 33% (p=0.004). In 54 patients with Dukes’ stage B+C for whom preoperative levels of both CEA and CA 19–9 were available multivariate analysis revealed a decreasing risk of relapse in the following order: CEA and/or CA 19–9 elevated (chi-square 7.09; p=0.008), CA 19–9 elevated (chi-square 6.27; p=0.01), CEA elevated (chi-square 5.47; p=0.02), and Dukes’ C (chi-square 2.08; p=0.15 n.s.). Hence, novel treatment protocols may have improved the disease-free survival, but the use of adjuvant chemotherapy and/or radiotherapy is of questionable benefit in patients who have elevated levels of CEA and/or CA 19–9 prior to treatment.