Intermittent or continuous panitumumab plus FOLFIRI (FOLFIRI/PANI) for first-line treatment of patients (pts) with RAS/BRAF wild-type (wt) metastatic colorectal cancer (mCRC): An update of survival/toxicity and preliminary results of genomic alterations from IMPROVE study.

Abstract

3571 Background: IMPROVE is a randomized, non-comparative multicenter, phase 2 study which evaluated continuous, until progressive disease (PD), or intermittent FOLFIRI/PANI, in pts with unresectable and previously untreated RAS/BRAF wt mCRC. In the intermittent schedule 8 cycles were followed by a treatment free interval lasting until PD when another 8 cycles were restarted, continuing this strategy until PD occurred on treatment. Primary endpoints was progression-free survival on treatment (PFSOT) at 1 year. In a previous analysis the primary endpoint of the study was met and a reduced skin toxicity was observed in the intermittent arm. Here we report an updated analysis. Methods: Updated analysis includes PFSOT and toxicity with 9 months (mo) of additional follow-up. Moreover we report overall survival (OS), skin toxicity burden (STB) score and preliminary data on genomic alterations (GAs) in 46 pts with PD. STB score, incorporating both the frequency and the severity of skin toxicity, was obtained by summing all grades that the pts experienced across all treatment cycles. GAs were performed by NGS on circulating tumor DNA (ctDNA) from paired plasma samples (baseline and progression), focusing specifically on RAS, BRAF and PI3K mutations. Results: Pts received FOLFIRI/PANI continuously (arm A, 69 pts) or intermittently (arm B, 67 pts). Median number of FOLFIRI/PANI cycles administered per pt were (arm A/B) 13/16. In arm B, after induction treatment, 24/53 pts (45%) without PD had ≥ 2 rounds of 8 FOLFIRI/PANI cycles. ORR (arm A/B) was 63/57%. At a median follow-up of 28 mo (IQR: 21-37), median PFSOT (77% of events) was 11.4 mo (95% CI: 9.1-13.7) in arm A, and 18.1 mo (95% CI: 6.8-29.3) in arm B. Median PFSOT in left sided tumors was 11.7 mo (95% CI: 9.1-14.3) in arm A and 23.9 mo (95% CI: 15.0-32.9) in arm B, compared with 10.7 mo (95% CI: 7.3-14.1) and 7.9 mo (95% CI: 5.7-10.1) in right sided pts, respectively. OS (46% of events) was 31.0 mo (95% CI: 24.7-37.2) in arm A and 32.2 mo (95% CI: 23.6-40.8) in arm B. Main grade ≥ 3 toxicities were (arm A/B): skin 30/18%, neutropenia 25/24%; diarrhea 13/15%. Median STB score was 0.77/cycle (IQR: 0.20-1.06) in arm A and 0.36/cycle (IQR: 0-0.77) in arm B. GAs in baseline ctDNA were evidenced in 12/46 (26%) pts, persisting to PD in all but one pt. Among the 34 pts without baseline GAs, only 8 (23%) developed ≥ 1 acquired GAs (Acq-GAs) to PD. Conclusions: Updated analyses confirmed that intermittent FOLFIRI/PANI strategy produces a long PFSOT and a reduced skin severe and skin burden toxicity without any detrimental effect on OS. Preliminary data on Acq-GAs suggest that classical mutations associated with anti-EGFR resistance are infrequent with up-front use of anti-EGFR/chemotherapy. Clinical trial information: NCT04425239 .