Regorafenib combined with irinotecan as second-line treatment in metastatic gastro-oesophageal adenocarcinomas: results of PRODIGE 58-UCGI35-REGIRI Unicancer randomised phase II study.

BACKGROUND

Several options have been evaluated in metastatic gastro-oesophageal adenocarcinomas (mGA) after failure of first-line fluoropyrimidine and platinum-based chemotherapy. Regorafenib (REGO), a receptor tyrosine kinase inhibitor, has shown promising activity as second- and third-line treatment of mGA.

PATIENTS AND METHODS

PRODIGE58-UCGI35-REGIRI was a comparative, prospective, phase II, open-label study evaluating the safety and efficacy of REGO [160 mg/day on day 2 (D2)-D8/D16-D22] plus irinotecan (IRI: 180 mg/m intravenously on D1/D15 every 28 days) versus IRI alone in patients with mGA (gastric or gastro-oesophageal junction/tumour Siewert II and III) after failure of first-line fluoropyrimidine and platinum-based chemotherapy. Primary endpoint was overall survival (OS).

RESULTS

Forty-four patients were included in the REGIRI arm and 45 in the IRI arm, primary tumours (67.4%) were mainly localised in the gastro-oesophageal junction, and 60.7% patients had synchronous metastases. With a median follow-up of 19.4 months [95% confidence interval (CI) 16.8-29.9 months], median OS was 6.3 months (95% CI 5.2-7.1 months) versus 8.2 months (95% CI 5.2-9.7 months) in the REGIRI versus IRI arms (hazard ratio 1.11, 95% CI 0.70-1.74, P = 0.66). Median progression-free survival was 2.2 months versus 1.9 months, objective response rate 15.9% versus 13.3%, and disease control rate 45.5% versus 33.3%. Grade 3 treatment-related adverse events (AEs) were reported for 52.3% of patients in the REGIRI arm versus 23.3% in the IRI arm with four toxic deaths (two homozygous UGT1A1∗28 patients died from sepsis and thrombotic microangiopathy, and two heterozygous UGT1A1∗1/∗28 patients from diarrhoea and pulmonary embolism), versus one (UGT1A1∗1 wild-type patient died from primary tumour perforation). Main grade ≥3 AEs were diarrhoea (18.2% versus 7.0%), hypertension (9.1% versus 0.0%), asthenia (6.8% versus 0.0%), febrile neutropenia (6.8% versus 0.0%), neutropenia (6.8% versus 11.6%), and weight decrease (6.8% versus 0.0%).

CONCLUSIONS

The study was stopped early because of limited efficacy and increased toxicities in the REGIRI arm, possibly due to drug interactions. No optimal sub-population that could benefit from a REGIRI regimen exposure was identified.