Impact of primary (1º) tumor location on overall survival (OS) and progression-free survival (PFS) in patients (pts) with metastatic colorectal cancer (mCRC): Analysis of CALGB/SWOG 80405 (Alliance).

3504 Background: 80405 found no OS or PFS difference when Bevacizumab (BV) or Cetuximab (Cet) was added to 1st-line FOLFOX or FOLFIRI in mCRC pts. As location of the 1° may affect mCRC outcome, we assessed the impact of 1° side (R v L) on OS and PFS in 80405 pts. Methods: 1° location was determined by chart review: 1137 pts w/KRAS wt (codons 12 and 13) in main cohort; 252 pts w/ KRAS mut tumors treated w/ BV or Cet pre-amendment. R-sided = cecum to hepatic flexure; L-sided = splenic flexure to rectum. Transverse (T) = hepatic to splenic flexure. PFS per investigator. Kaplan Meier and Cox regression methods used. Results: KRAS wt pts: Median age = 59; synchronous = 78%. 1° site: R – 280 (25%); L – 689 (61%); T- 62 (5%); unsure – 106 (9%). OS & PFS (Table) difference by side statistically significant if adjusted for age, gender, BV / Cet, chemotherapy, prior therapy. There was a significant 1° side by biologic interaction (P int = 0.003, PFS and OS) but not by chemo, gender or RAS. OS, L-sided: Cet v BV, superiority (Log rank p = 0.04); R-sided: BV v Cet, superiority (p = 0.03). Results similar for PFS and if T colon allocated to R side. KRAS mut pts: 1°s: R - 35%; L- 50%. No statistically significant difference in any subset although OS favors L > R (only OS data shown). Conclusions: mCRC arising in the R v L colon are clinically different. In KRAS wt mCRC, pts w/ L-sided 1° tumor have superior OS and PFS v pts w/ R-sided 1°. Though not pre-planned analyses, OS and PFS were prolonged w/ Cet in L and w/BV in R but were poorer w/ Cet in R. Forthcoming molecular analysis of 1°s - e.g. BRAF, MSI, methylation - may provide a biological explanation. For now, stratification in mCRC studies by R v L 1° sidedness is indicated. These data support BV in 1st line treatment for mCRC pts w/R-sided 1° tumor regardless of KRAS status. Support: U10CA180821, U10CA180882. Clinical trial information: NCT00265850. Right 1° Left 1° Hazard Ratio, 95% Confidence Interval (R v L, adjusted) Log Rank p-value (adjusted) OS (mos) All KRAS wt 19.4 34.2 1.56 (1.32, 1.84) < 0.0001 Cet 16.4 37.5 1.97 (1.56, 2.48) BV 24.5 32.1 1.26 (1.00, 1.58) All KRAS mut 23.1 30.3 1.28 (0.95, 1.73) PFS (mos) All KRAS wt 8.9 11.5 1.25 (1.08, 1.46) .002 Cet 7.7 12.0 1.54 (1.25, 1.91) BV 9.5 11.1 1.03 (0.83, 1.28)