Background: Mobocertinib (mobo) and amivantamab (ami) are FDA-approved treatments for patients (pts) with locally advanced or metastatic NSCLC with EGFR ex20ins whose disease progressed on or after platinum-based chemotherapy. An unanchored MAIC was used to compare confirmed overall response rate (cORR), duration of response (DoR), progression-free survival (PFS) and overall survival (OS) between mobo and ami. Methods: Clinical outcomes were compared in platinum-pretreated pts with EGFR ex20ins+ NSCLC treated with mobo 160 mg QD in a phase I/II single-arm study (NCT02716116, cut-off 1 Nov 2020, n=114) or with ami 1,050 mg (1,400 mg, ≥80 kg) in a phase I single-arm study (NCT02609776, cut-off 8 June 2020, n=81). Differences in baseline characteristics reported in both studies, including age, race, sex, smoking status, Eastern Cooperative Oncology Group, histology, sites of metastasis (brain, bone and liver), time from advanced diagnosis, number of prior lines of therapy, prior immuno-oncology therapy, prior EGFR tyrosine kinase inhibitor treatment and prior EGFR ex20ins targeted therapy, were adjusted with MAIC. Results: After MAIC weighting all reported baseline characteristics were balanced between mobo and ami. OS and cORR per investigator assessment (INV) were similar between mobo and ami (Table). cORR per independent review committee (IRC) was numerically higher for ami (odds ratio [OR]=0.64, p value=0.230). For PFS per IRC, the adjusted hazard ratio (HR) was numerically favorable for mobo (HR=0.82, p value=0.417). Among the responders, DoR was longer for mobo (DoR per INV: HR=0.44, p value=0.049; DoR per IRC: HR=0.56, p value=0.149). Conclusions: Mobo and ami appear to have overall similar efficacy. As each has a different mechanism of action and route of administration, they provide multiple options in the treatment of EGFR ex20ins+ NSCLC. Clinical trial information: NCT02716116.

Abbreviations: ESS=effective sample size; mos=months; NE: not estimable. [1] Park K, et al. JCO 2021;39(30):3391-3402. ^a estimated using reconstructed patient-level data from swimmer and survival plots for ami.