Single-arm multicentre phase II study of bevacizumab (B) combined with 9-weekly alternating CAPOX and CAPIRI as first-line treatment of patients with metastatic colorectal cancer (mCRC): Main results of the AXOAXI-trial

Bevacizumab (B) combined with capecitabine and oxaliplatin (B-CAPOX) or irinotecan (B-CAPIRI) are the cornerstones of mCRC treatment but linked to cumulative neuropathy and limited treatment duration and efficacy, especially in the conversion setting. The aim was to improve resectability, efficacy, and tolerability by a 9-weekly alternating four-drug schedule of bevacizumab with CAPOX or CAPIRI, instead of the most efficacious but more toxic FOLFOXIRI. Finnish mCRC patients at 3 centres were enrolled from 12/2012 to 10/2018. The objective was to investigate 1-year progression-free survival (PFS) rate, which was defined as clinically meaningful if over 44%, in patients having received at least 3 cycles of therapy (n=75) in the AXOAXI-trial (NCT01531595, EudraCT2011-003137-33). In the intention to treat (ITT, n=77) population, resectability by repeated central assessment, PFS 1st from initiation of therapy to 1st progression (including progression on treatment pause), PFS 2nd progression on doublet, PFS 3rd to progression on all study drugs, overall survival (OS), response rates (RR) and disease control rate (DCR), safety graded by NCI-CTCAEv4.0 (patient diary used) and biomarkers, were analysed as secondary endpoints. Median age was 64 years and 71% were male. ECOG status was 0/1/2 in 34%/55%/12%, respectively. Metastases were synchronous in 83%; with liver (72%), lymph nodes (36%), lungs (32%) and peritoneum (23%) being most frequent and 1/2/3-6 sites present in 32%/39%/30%, respectively. Median duration of 1st/2nd/3rd/4th-line therapy was 11.3/6.1/2.1/1.8 months, respectively. Median number of AXOAXI cycles was 18 (range 1-68) and post-AXOAXI 4 (range 1-20). Post-AXOAXI therapy in 3rd and 4th line were given in 36% and 8%, respectively. At 12 months, 28 out of 75 treated had progressed and PFS-rate was 62% (>44%). Median PFS 1st in ITT (n=77) was 16.6 (CI95% 13-20) months, PFS 2nd 9.7 (CI95% 15-24) months and PFS 3rd 22.4 (CI95% 16-29) months. OS was 28.2 (CI95% 20-37) months. RR-rate for 1st-line was 59% (CR 3%, PR 38%, PR→NED 18%) and DCR 99%, and in 2nd-line RR was 21% and DCR 74%. In central assessment, 10 (13%) were resectable upfront and 10 (13%) were converted, of which 19 (95%) were R0/1/2-resected. A 25% resection rate, with estimated OS from 1st resection of 48.1 (CI95% 19-77) months was noted, and 1/3-year OS-rates were 100%/80%, respectively (median follow-up 34 months). One died of intra-abdominal bleed during lymphadenectomy. RAS/RAFwt were seen in 36% with mPFS/mOS of 15.2/41.6 months, RASmt in 51% with 17.7/23.6 months and BRAFmt in 13% with 11.5/19.6 months, respectively. grade 3-4 adverse events were seen in 54%, with infection 22%, diarrhoea 17%, neutropenia 12% and nausea/vomiting 7% being most common. Neuropathy grade 1-2 was seen in 87% and 3-4 in 5%. Bevacizumab-related toxicities were any grade proteinuria in 25%, hypertension in 18% and perforation in 3%. One death due to sepsis was noted. Alternating B-CAPOX and B-CAPIRI clearly met its primary endpoint with a 62% 12-month PFS-rate, and high conversion and resection rates giving promising curation. PFS was 16.6 months and OS >40 months in RAS/BRAFwt, 24 months in RASmt and 20 months in BRAFmt. Toxicity was acceptable with grade 3-4 overall toxicity in 54% and neuropathy in only 5%.