The treatment paradigm for advanced non-small cell lung cancer (NSCLC) harboring EGFR mutations is undergoing a significant transition. While third-generation tyrosine kinase inhibitors (TKIs) like osimertinib have long served as the frontline standard, the emergence of heterogeneous resistance mechanisms requires more robust therapeutic strategies. This review evaluates the clinical impact of the MARIPOSA trial, which demonstrated the superior efficacy of combining the bispecific antibody amivantamab with lazertinib. Beyond improving progression-free and overall survival, this dual-inhibition approach fundamentally alters the clonal evolution of the disease by suppressing common escape routes, such as MET amplifications and secondary EGFR mutations. Furthermore, we explore the diversifying landscape of second-line interventions, including the rise of antibody-drug conjugates (ADCs) like Sac-TMT and patritumab-deruxtecan, dual PD-1/VEGF inhibitors, and novel fourth-generation TKIs. By integrating preclinical insights on drug-tolerant persister cells with late-phase clinical data, this article outlines a future for EGFR-mutant NSCLC management defined by precision sequencing and the proactive mitigation of molecular resistance.

This commentary discusses the FDA’s drug approvals in 2025, with a particular focus on cancer therapies and the role of companion diagnostics (CDx). Cancer has emerged as the leading therapeutic area, accounting for 35% of all new drug approvals, largely driven by targeted therapies, with kinase inhibitors representing nearly half of these drugs. Many of the drugs have received orphan drug designations and/or have utilized the Accelerated Approval Program. A key finding was the widespread adoption of the drug-diagnostic co-development model, in which a CDx assay is developed along with the drug and used for patient selection in clinical trials. However, a significant challenge is the frequent lack of concurrent drug and CDx assay approvals. The absence of an analytically and clinically validated CDx assay may pose a challenge for healthcare providers in accurately identifying eligible patients, potentially delaying access to appropriate therapy. The FDA’s cancer drug approvals for 2025 highlight an ongoing commitment to precision medicine, with several new targeted treatments, such as antibody-drug conjugates and kinase inhibitors, where CDx assays play an important role in identifying the appropriate patient population.

Machine learning (ML) and deep learning (DL) models applied to electronic health records (EHRs) have substantial potential to improve oncology care across diagnosis, prognosis, treatment selection, and trial recruitment. However, opacity of many high-performing models limits clinician trust, regulatory acceptance, and safe deployment. Explainable artificial intelligence (XAI) methods aim to make model behavior understandable and actionable in clinical contexts. The present perspective summarizes current XAI approaches applied to EHR-based oncology tasks, identifies key challenges in evaluation, reproducibility, clinical utility, and equity, and proposes pragmatic recommendations and research directions to accelerate safe adoption in oncology. Common XAI categories used with EHR data include feature importance/interaction methods, intrinsically interpretable models, attention mechanisms, dimensionality reduction, and knowledge distillation or rule extraction. Tree-based models with SHapley Additive exPlanations (SHAP) explanations dominate recent EHR studies. Other interpretable strategies, such as generalized additive models and rule sets, appear in settings where transparency is prioritized. Gaps include inconsistent reporting, scarce formal evaluation of explanations for clinical utility, limited reproducibility for data and code availability, inadequate external validation, and insufficient consideration of fairness and equity that these issues are particularly important in oncology, where heterogeneity and stakes are high. Overall, integrating XAI with EHR-driven oncology models is promising but underdeveloped, which requires further progress by multi-stakeholder evaluation frameworks, reproducible pipelines, prospective and multicenter validations, and equity-aware design. The field should prioritize clinically meaningful explanations beyond ranking features and study how explanations affect clinician decision-making and patient outcomes.