For the hundreds of thousands of patients worldwide living with EGFR-mutant non-small cell lung cancer, osimertinib has become a standard first-line therapy — but nearly all patients eventually develop resistance. Understanding precisely how that resistance arises, and whether it varies from patient to patient, is now one of the central challenges in thoracic oncology.
The ELIOS trial (NCT03239340), a prospective multicenter study published in Cancer Discovery, characterized acquired resistance mechanisms by comparing pre-treatment and post-progression tumor biopsies. Among 52 patients with evaluable paired tissue samples from an enrolled cohort of 154, MET amplification emerged as the most frequent acquired alteration at progression (17%), followed by deletions of the tumor suppressors CDKN2A/CDKN2B (15%) and MTAP (13%), and the gatekeeper mutation EGFR C797S (13%). Critically, a proteogenomic analysis — the first of its kind in the osimertinib setting — found that TROP2 protein was robustly expressed both before and after treatment, independent of the underlying genetic resistance mechanism. In a parallel matched tissue-plasma analysis of 51 patients, liquid biopsy identified potential resistance alterations in 82% of cases, underscoring the complementary diagnostic value of combining tissue and circulating tumor DNA.
This study carries several layers of significance worth parsing carefully. The co-occurrence of MET amplification and CDKN2A/CDKN2B loss as leading resistance drivers aligns with prior retrospective and real-world data, lending prospective confirmatory weight to findings that have shaped combination therapy trials already underway. However, the paired biopsy cohort of only 52 patients reflects a recurring limitation in resistance biology research: the difficulty of obtaining repeat tissue at progression. The heterogeneity of resistance mechanisms observed — with no single alteration dominating — reinforces why single-target salvage strategies are unlikely to broadly succeed. The TROP2 proteomic finding is arguably the most novel element; it suggests that antibody-drug conjugates targeting TROP2, such as sacituzumab govitecan, may have mechanistic rationale across diverse resistance contexts, regardless of genotype. This is an incremental but well-designed confirmatory study with one genuinely novel proteomic thread worth watching.