Eligibility of real-world patients with stage II/III colorectal cancer (CRC) in adjuvant chemotherapy (AC) trials.

Abstract

50 Background: The results of AC trials in stage II/III CRC are often generalized to real-world patients. However, clinical trials have stringent inclusion and exclusion criteria, which can potentially lead to poor generalizability of results and slow accrual. This study was conducted to determine the proportion of real-world patients with stage II/ III CRC who would be eligible for AC trials based on common eligibility criteria and to compare the outcomes in eligible and ineligible patients. Methods: We identified all patients diagnosed with stage II/III CRC in 2004-2015 from the Alberta Cancer Registry. Patients meeting any one of the following criteria were considered ineligible: age >75 years, anemia, comorbid conditions (heart disease, uncontrolled diabetes, kidney disease, liver disease) and history of a prior malignancy or immunosuppression. Logistic regression was used to describe the likelihood of receiving AC and Cox regression models were constructed to determine overall survival (OS). Results: A total of 7841 patients with stage II/III CRC were identified, of whom 52% were men and median age at diagnosis was 71 years (IQR: 61-79 years). Approximately 59% patients were deemed trial-ineligible and the most common reasons for ineligibility were advanced age (36%), renal dysfunction (27%), and cardiac disease (17%), respectively. In the real-world, 54% of eligible patients received AC as compared to 23% of ineligible patients [odds ratio 3.89, 95% confidence interval (CI) 3.53-4.28, P< 0.0001]. The 5-year OS of trial-ineligible patients who received AC was significantly better than those treated with surgery alone (table). Conclusions: Majority of real-world patients with stage II/ III CRC are unable to participate in AC trials due to strict exclusion criteria, but a fair proportion of these patients still derive some benefit from AC. The eligibility criteria of AC trials in CRC should be broadened to be more representative of real-world patients. [Table: see text]