Which side is better? The impact of primary tumor side in early stage colorectal cancer (CRC)

e15106 Background: Multiple studies have indicated the prognostic and potential predictive significance of primary tumor side in metastatic CRC. To date, the few studies examining its impact in early stage disease have either combined data across multiple stages or restricted analysis to overall survival (OS) data. A by stage analysis of the impact of tumor side on recurrence risk is critical if it is to impact adjuvant therapy decisions. Methods: We examined data from a multi-site Australian registry of consecutive patients diagnosed from 2003-2016. Tumors at and distal to the splenic flexure, including the rectum, were considered a left primary (LP). Rectal patients treated with initial chemoradiation were excluded. Clinico-pathologic and outcome data were examined. Data analysis was provided by the healthcare group at IBM Research Australia. Results: A total of 6123 patients were identified, of which 1046 (17.1%) had initial stage I, 1892 (30.9%) had stage II, 1708 (27.9%) had stage III, and 1477 (24.1%) had stage IV disease. Most patients were male (55.2%), and had a LP (n = 3818, 62.4%). Median age at diagnosis was 68.8 years, was higher in patients with a right primary (RP) (71.6 versus 67.0 years for LP, p < 0.001), with more females in the RP group (51.1% vs 41.0% for LP, p < 0.001). The proportion of RP varied by stage, highest in stage II (44.9%), lowest in stage IV (31.5%). For all stage IV disease, including metachronous cases, OS was worse with a RP (HR 1.32, 95% CI 1.14-1.53). For early stage cases, distant recurrence free survival (DRFS) was similar for RP vs LP for stage I (HR 0.63, 95% CI 0.32-1.23), better for stage II RP (HR 0.72, 95% CI 0.55-0.95) and worse for stage III RP disease (HR 1.22, 1.01-1.48). OS did not differ for RP vs LP for stage I or II disease, but was worse for stage III disease with a RP (HR 1.39, 95% CI 1.13-1.70). Furthermore, post recurrence survival was poorer in stage III RP disease (HR 1.61, 95% CI 1.33-1.96). Conclusions: Primary tumor side has potential as an important prognostic marker in early stage CRC. Our novel finding of a variable impact by stage indicate that an assessment of cohorts where recurrence data is available is critical to fully understanding the implications of tumor side for adjuvant therapy decision making.