A phase II prospective multi-institutional trial of adjuvant active specific immunotherapy following curative resection of colorectal cancer hepatic metastases: cancer and leukemia group B study 89903.

BACKGROUND: Patients with curatively resected colorectal cancer hepatic metastases often harbor occult metastatic disease and are at high risk of experiencing recurrence. This patient cohort is ideally suited to test novel therapies such as immunotherapy. We treated patients-post-hepatic resection-with anti-idiotype monoclonal antibody vaccines to the tumor-associated antigens carcinoembryonic antigen (CeaVac) and human milk fat globule (TriAb), both of which are co-expressed in more than 90% of colorectal cancer patients. METHODS: Vaccinations commenced 6-12 weeks post-hepatic resection and consisted of four biweekly treatments of 2 mg CeaVac and TriAb, then monthly treatments for 2 years, then on every other month for 3 years. The primary endpoint was to investigate the proportion of patients recurrence-free at 2 years, and the objective of the study was to demonstrate that at least 58% would be recurrence-free at this time to consider the regimen worthy of further study. RESULTS: Between July 2001 and October 2004, 56 patients were accrued; 52 patients with margin-negative resection were eligible for analysis. Hepatic lobectomy was performed in 56% of patients with a median of one metastasis (range 1-3). Of the 52 eligible patients, 49 were evaluable for the primary end point. Median follow-up was 3.1 years. The proportion of patients recurrence-free at 2 years was 39%, with a lower confidence bound (LCB) of 0.29. Median recurrence-free survival was 16 months. The 2-year overall survival was 94% (95% CI, 0.81, 0.98). Only 10% of patients had documented grade-3 adverse events. CONCLUSIONS: Anti-idiotype monoclonal antibody vaccine therapy with CeaVac and TriAb as an adjuvant to curative resection of colorectal cancer hepatic metastases is well tolerated but did not improve 2-year recurrence-free survival when compared with the expected value of 40% reported for hepatic resection alone.

Duke Scholars

Author Donna Niedzwiecki Biostatistics & Bioinformatics, Division of Biostatistics