Primary tumor resection in metastatic colorectal cancer (mCRC): A prospective cohort study.

Abstract

3584 Background: The role of primary tumor resection in patients presenting with mCRC remains controversial. Previously reported survival benefits associated with primary tumor resection may not translate in the modern era of systemic therapies. We examined the impact of primary tumor resection on survival in a modern cohort of mCRC patients. Methods: Patients were identified using a clinician-designed mCRC registry involving 15 participating Australian sites from mid 2009. Patients were excluded if planned for curative metastasectomy or had incomplete data. Univariate logistic regression and multivariate cox regression was utilized to identify significant associations between resection, clinical variables and survival outcomes. Results: We identified 682 mCRC patients with median follow up 20 months. 40% (n = 275) had their primary in-situ. Rates of primary resection were higher for age > 70 years (OR 1.66 95% CI [1.22 – 2.26], p = 0.001) and Charlson score ≥3 (OR 1.50 [1.10 – 2.06], p = 0.011). Lower resection rates were observed for rectal v colon primary (OR 0.39 [0.28 – 0.55], p < 0.001), liver metastases (OR 0.59 [0.42 – 0.82], p = 0.002) and ECOG 2 - 4 (OR 0.64 [0.45 – 0.92], p = 0.015). There was a significant survival advantage for pts with primary tumor resection (median OS 21.3 vs 16.8 months; HR 0.63, p < 0.001), even when adjusting for known prognostic factors in a multivariate analysis (HR 0.56 [0.44 – 0.72], p < 0.001). Multivariate analyses also demonstrated that age > 70 years (HR 1.32 [1.03 – 1.71], p = 0.031) and ECOG ≥ 2 (HR 3.17 [2.43 – 4.15], p < 0.001) were significantly associated with poorer outcomes; whereas chemotherapy use (HR 0.61 [0.45 – 0.84], p = 0.002), bevacizumab use (HR 0.68 [0.52 – 0.89], p = 0.005) and rectal primary (HR 0.69 [0.53 – 0.91], p = 0.009) predicted improved survival. Conclusions: Our study suggests that primary tumor resection is associated with significant survival advantages for mCRC patients in the modern era of systemic therapies. The 40% of primary cancers in-situ is higher than previous mCRC studies and suggests a tendency for non-operative intervention in Australia. Further analysis aimed at examining the impact of other confounding variables such as tumor burden is ongoing and will be presented.