ALKTERNATE: a pilot study alternating lorlatinib with crizotinib in ALK-positive NSCLC with prior ALK inhibitor resistance

Introduction: ALK-positive lung cancers represent a molecularly diverse disease. With drug exposure, driving selection pressure, and resistance pathways, disease relapse will emerge. There is compelling rationale to investigate novel treatment strategies, informed by dynamic circulating tumor DNA (ctDNA) monitoring. Methods: The single-arm, pilot study ALKTERNATE investigated fixed alternating cycles of lorlatinib intercalated with crizotinib in individuals resistant to second-generation ALK inhibitors. Dynamic ctDNA explored the correlation with disease response and disease recurrence and defined disease resistance. The primary outcome was time-to-treatment failure, a composite of tolerability, feasibility, and efficacy. Secondary outcomes included standard survival measures, toxicity, pharmacokinetic analysis, and patient-reported outcomes. Tertiary outcomes were proteogenomic analyses of tissue and plasma. Results: A total of 15 individuals were enrolled; three encountered primary resistance to lorlatinib induction. There were 12 participants who received alternating therapy, and this approach revealed safety, feasibility, and effectiveness. Patient-reported outcomes were maintained or improved on therapy, and toxicity was consistent with previous reports. The pharmacokinetic measures were similar to the single-arm drug experience. Median time-to-treatment failure was 10 months; overall survival was 23 months. ctDNA profiles indicated inferior survival in those with preexistent TP53 mutations and those without clear or cleared ctDNA at trial induction. The study defined a vastly heterogeneous population with an abundance of ALK coexisting with non-ALK resistance variants. Conclusions: ALKTERNATE revealed feasibility with a novel alternating ALK inhibitor strategy in ALK-positive NSCLC. Results support progressing inquiry into this approach and propose a flexible design with drug(s) selected and alternating time frames, informed by real-time plasma profiling. Moving this concept to treatment naive may also optimize impact.