Tumor burden (TB) as a prognostic indicator in patients with metastatic colorectal cancer (mCRC).

Abstract

572 Background: At initial diagnosis of mCRC, accurate assessment of patient prognosis is critical to defining the optimal initial treatment strategy. While performance status (PS) and number of metastatic sites are presumed surrogates of TB, to date TB assessment has not been formally examined as a potential prognostic marker in mCRC, despite supportive evidence in other malignancies. Methods: We developed a CT-based scoring system for TB on mCRC patients from the TRACC (Treatment of Recurrent and Advanced Colon Cancer) database. Objective (size and number of metastases) and subjective (clinical judgment) scoring systems were used to separate patients into high, moderate, or low TB subsets. Subjective scoring was repeated by a blinded second assessor on a subset of cases. Clinicopathologic correlates with TB were then assessed, along with survival outcomes for each subset. Results: 151 patients were assessed, median age was 68 years and 51% were male. Using subjective scoring, 55 (36%), 33 (22%), and 63 (42%) were categorized to high, moderate, and low TB subsets. Concordance between assessors using subjective scoring for 30 selected cases was 80% (high), 60% (moderate), and 90% (low). Factors correlating with increasing TB (high vs. low) on objective scoring were PS (31 vs 5% ≥ 2, p = 0.0002), number of disease sites (57 vs 9% ≥2, p < 0.0001) and for patients with mCRC at presentation, reduced likelihood of primary CRC resection. Median overall survival was superior in low compared to high TB subsets for subjective (not reached vs. 15.9mths, p < 0.0001) and objective (not reached vs. 14.7mths, p < 0.0001) scores. Survival difference for low vs high TB were maintained on objective scoring even when patients with resectable metastases were excluded (22.7 vs. 11.9mths, p = 0.0017). Conclusions: TB measured subjectively or objectively is associated with overall survival in mCRC, and is a potential novel prognostic marker. Multi-variate analyses will be performed, and an independent validation cohort is being examined. As TB inversely correlated with primary resection for patients with mCRC at presentation, including TB assessment in further analyses of the survival impact of primary CRC resection is one potential application of this measure.