RETaliation—Tackling Rare Resistance Alterations to Osimertinib

In this issue of Clinical Cancer Research, Rotow and colleagues report on a series of EGFR mutation (EGFRm)-positive lung cancer patients with RET fusion–mediated resistance to osimertinib and clinical activity of a selective RET inhibitor combined with osimertinib (1). While EGFR tyrosine kinase inhibitors (TKI) have revolutionized the treatment of EGFRm advanced non–small cell lung cancer (NSCLC), acquired resistance is inevitable, and a rational approach to understanding these biological mechanisms may result in additional treatment strategies, delaying time to chemotherapy. Such resistance mechanisms are mediated through either on-target or off-target effects. On-target resistance results from mutations mapping to the EGFR kinase domain, typically gatekeeper mutations (T790M in the case of first- and second-generation EGFR inhibitors), or solvent front mutations such as C797X (most common mutation with third-generation EGFR inhibitors) that cause steric hindrance to drug binding at the ATP-binding site. Off-target resistance is mediated either through genomic alterations in upstream or downstream effectors of the EGFR/MAPK signaling pathway; alterations within parallel signaling pathways such as MET, HER2, and RET; or through nongenomic mechanisms such as phenotypic transformation (Fig. 1; ref. 2). It is thus increasingly relevant to undertake broad panel genomic analyses at point of progression on EGFR TKIs to identify druggable genomic targets. While circulating tumor DNA (ctDNA) analysis can provide an easy method to screen for such somatic variants, it is not necessarily optimal for fusion detection (due to either coverage variation or sensitivity) nor copy-number detection and cannot inform on histologic changes. Thus, complimentary tissue testing should also be considered for pathology review and RNA/DNA analysis to cover the full spectrum of druggable alterations, especially in the case of noninformative ctDNA (3).

RET fusions are a primary genomic driver of cancer accounting for 1% to 2% of NSCLCs with the selective RET inhibitors selpercatinib and pralsetinib demonstrating marked clinical activity underpinning FDA and European Medicines Agency approvals (4, 5). RET fusions have also previously been identified as a rare off-target osimertinib acquired resistance mechanism, potentially leading to kinase inhibition vulnerability. Given its rarity, only limited cases of RET fusion–mediated resistance have been reported to date (6, 7) with RET inhibitor/osimertinib combination showing promising activity in case series of 2 patients or less (8). Progress in evaluating treatment combinations in these rare genomic settings is hindered by traditional drug development and allied regulatory processes, requiring safety evaluation, dose finding, cohort expansion, and ultimately randomized trials that are simply unfeasible and/or impractical in such settings.

It is against this backdrop that Rotow and colleagues report the clinical activity of selpercatinib/osimertinib combination in the largest retrospective international cohort to date (n = 14) of RET fusion–mediated resistance to osimertinib with an innovative approach to data collection (1). Patients were treated with selpercatinib via an expanded access or single-patient compassionate use program, combined with osimertinib. Dosing was initiated according to investigator discretion but the majority received osimertinib at the recommended phase II dose (RP2D) of 80 mg once daily and selpercatinib at a reduced dose of 80 mg twice daily (RP2D 160 mg twice daily) on the basis of potential overlapping toxicity and efficacy data being comparable with higher doses (4). While clearly not a formal dose-finding study, this is a pragmatic and acceptable approach to dose selection with two licensed therapies with established toxicity profiles.

Among 12 evaluable patients with measurable disease, an overall response rate of 50% was achieved, disease control rate of 83% and median duration of treatment 7.9 months. Over 70% patients exhibited disease control for over 6 months. The combination was well tolerated and while adverse events were not rigorously collected there were no treatment-related serious adverse events and toxicities were reassuringly in keeping with individual agent profiles. These data therefore provide the first meaningful cohort level data on the benefit and tolerability of this treatment combination in the RET resistance setting and warrant further evaluation.

A fascinating insight into emergent resistance patterns to the selpercatinib/osimertinib combination was also presented. As predicted, both on-target (EGFR and RET gatekeeper or solvent front mutations) and off-target resistance patterns were identified (Fig. 1). Loss of RET fusion and mutations in parallel MAPK signaling pathway accounted for the majority of off-target events. Interestingly, both on-target and off-target resistance coexisted in several patients. Perhaps, most intriguing was the emergence in 1 patient of two ALK fusions with maintained RET fusion/EGFR mutation and emergent RET (V804E) gatekeeper mutation. The evolutional biology underpinning development of multiple coexisting pathogenic fusions is bewildering and serves to illustrate the challenges faced in tackling resistance in real time in the clinical setting. Whether resistance reflects selective pressure of treatment in preexisting subclones or de novo genomic changes in existing cells or a combination of both, developing strategies for preventing and ultimately bypassing resistance will become increasingly complex (9, 10).

However, several study limitations should be borne in mind including the overall low patient numbers, absence of formal prospective data collection and standardized follow-up, differences in dosing, upward biases in unblinded investigator efficacy reporting and indeed how additional data can be obtained to support potential regulatory approvals and develop a roadmap for other similar rare genomic resistance settings. Authors flag the ongoing phase II nonrandomized ORCHARD study (NCT03944772); another approach may be for Industry partners/academic collaboratives/health care systems to formalize processes and data collection within the context of compassionate use drug access programs for formal outcomes and toxicity evaluation. In Europe, relevant combination treatment arms could be included in DRUP-like studies aimed at evaluating licensed treatments in unlicensed indications, where single-arm data are planned to be pooled across Europe and with each country working with regulators to inform a route to reimbursement in rare settings (11). While differences exist between countries in requirements and approach to reimbursement decisions in the rare cancer setting (12), and greater clarity is needed regarding use of real-world data for comparator arms (13), there are several examples of agents receiving regulatory approval in the absence of traditional randomized data and with single-arm phase II data (14, 15).

The clinical efficacy of selpercatinib/osimertinib should also be considered in context of chemotherapy as a comparator in those whom are chemo-naïve. The CheckMate722 trial has benchmarked platinum-pemetrexed in the post-osimertinib setting reporting a median progression-free survival of 5.4 months, a target that all rationally designed drug combinations should aim to beat (16). Given the complexities of emerging resistance mechanisms, generic approaches to treatment such as antibody–drug conjugates (ADC) should also be considered and evaluated. Indeed, promising activity has been demonstrated with patritumab deruxtecan (HER3-directed ADC) and datopotamab deruxtecan (TROP2-directed ADC) in EGFRm patients (17, 18). Phase III clinical trials are ongoing to further evaluate these agents and help determine their optimal position in the treatment pathway (NCT05338970; NCT04656652).

Rotow and colleagues (1) present a valuable dataset and highlight the challenges faced in drug development for rare druggable emergent resistance profiles to TKIs. While further data are collated, next-generation sequencing testing at the point of resistance and consideration to clinical trials or formal data collection in the context of compassionate use requests should be encouraged as we are now firmly entrenched in an era of genomic medicine, exploring additional combinations in similar resistance settings. Academia and industry should work together with regulatory and health technology agencies and identify a unified framework for assessing such rare patient groups and identify acceptable suitable comparator arms when traditional drug development processes are unfeasible.

M.G. Krebs reports personal fees from Bayer, Janssen, Guardant Health, Roche, Seattle Genetics, AstraZeneca, BerGenBio, and Immutep and grants from Novartis and Roche outside the submitted work. S. Popat reports personal fees from Amgen, AstraZeneca, Bayer, BeiGene, Blueprint, BMS, Boehringer Ingelheim, Daiichi Sankyo, EQRx, GSK, Guardant Health, Incyte, Lilly, Merck Serono, MSD, Novartis, Pfizer, Seattle Genetics, Takeda, and Turning Point Therapeutics and personal fees and non-financial support from Janssen and Roche outside the submitted work. No other disclosures were reported.