REFINE-Lung: A multicentre phase III study to determine the optimal frequency of pembrolizumab in non-small cell lung cancer utilising a novel multi-arm design.

TPS9145 Background: Pharmacological and clinical data suggest standard regimens for anti-PD1 agents nivo and pembro result in overtreatment. These agents have high target affinity, saturate PD1 at very low concentrations and maintain receptor occupancy for up to 30 wks following discontinuation. Clinical studies suggest no dose-response relationship in the wide range tested (0.1 – 10 mg/kg) and retrospective data suggest the efficacy of lower doses or longer administration frequency. Given their high costs and important implications for pt quality of life and toxicity, new approaches to optimise treatment regimens are required. However, conventional 2-arm non-inferiority designs require large numbers of patients, explore a single alternative to the standard of care and potentially require multiple trials to establish optimal parameters such as dose or frequency. Here we present a new multi-arm trial design to tackle this problem, implemented in the UK multicentre phase III REFINE-Lung study that aims to determine the optimal treatment frequency of pembro in NSCLC. Pts who are progression free after 6 months of standard therapy are randomised to 5 pembro frequency reduced arms (q6w, q9w, q12w, q15w and q18w). By evaluation of the frequency-response relationship, we will determine the longest frequency that is equivalent to standard of care within a pre-defined margin. Methods: The REFINE-Lung trial (NCT05085028) is a multicentre, randomised open-label, phase III trial in advanced NSCLC. REFINE-Lung utilises a novel Multi-Arm Multi-Stage Response Over Continuous Interventions (MAMS-ROCI) design to determine the optimal dose frequency of pembro. 1750 pts aged ≥18 and progression free at six months into 1st line pembro and planning to continue, will be enrolled from up to UK 35 hospital groups. Pts will be randomised across 5 arms to determine the optimal dose frequency. To assess safety, pts will initially be randomised 1:1 into q6w (control) and q12w arms. Interim analysis will be conducted once 37 progression free events are observed in the control arm. If PFS in the q12w arm is not significantly reduced, the remaining q9w, q15w and q18w arms will be opened. Pts who progress on a frequency reduced arm may re-escalate to q6w therapy. The primary objective is to determine the optimal continuing dose frequency of pembro, defined as the longest dose interval non-inferior to standard therapy using 2-year survival as the primary outcome. Secondary endpoints include OS, PFS, ORR, duration of response, adverse events, quality of life and cost effectiveness. An exploratory sub-study will explore fundamental aspects of cancer immunotherapy and develop novel biomarkers of response, resistance, toxicity and patient suitability for dose frequency de-escalation. Recruitment is ongoing and the trial is open at 16 centres at time of submission. Clinical trial information: NCT05085028 .