For years after the FLAURA trial established osimertinib, a selective third-generation EGFR tyrosine kinase inhibitor with brain penetrance, as the preferred first-line treatment, the clinical calculus was straightforward. Osimertinib monotherapy delivered a median progression-free survival of 18.9 months and a final overall survival of 38.6 months, far exceeding first-generation comparators, with a manageable toxicity profile. The main question was access, not selection. That simplicity ended with the MARIPOSA and FLAURA2 phase 3 trials, both of which demonstrated that adding a second therapeutic modality to osimertinib extends both progression-free and overall survival.

The numbers are striking. In MARIPOSA, the chemotherapy-free pairing of amivantamab, a bispecific antibody targeting both EGFR and MET, with the brain-penetrant TKI lazertinib produced a median progression-free survival of 23.7 months and, at final analysis, a statistically significant overall survival advantage over osimertinib alone, with three-year survival of 60 percent versus 51 percent. FLAURA2, which added platinum-pemetrexed chemotherapy to osimertinib, achieved a median progression-free survival of 25.5 months and final overall survival of 47.5 versus 37.6 months. Both regimens now carry regulatory approval and the highest guideline recommendations. Yet the price is steep: grade 3 or worse adverse events occurred in 70 percent of FLAURA2 combination patients and 80 percent of MARIPOSA patients, compared with far lower rates on monotherapy.

Neither pivotal trial, the review notes, offers clinicians a validated framework for choosing between these strategies. Both enrolled only patients with good performance status and median ages in the early-to-mid sixties, while real-world patients are older, frailer, and more comorbid. Decision-making currently rests on trial eligibility criteria and institutional habit rather than individualized risk stratification. The authors argue that the solution is a structured, four-step algorithm integrating tumor biology, molecular co-alterations, patient fitness, and logistical reality.

The first pillar of the framework is tumor biology. Certain anatomic features consistently predict inferior outcomes on monotherapy and may identify patients who benefit disproportionately from intensification. In a large United States real-world cohort of 1,323 patients starting first-line osimertinib, median overall survival was markedly shorter in those with liver metastases, at 19.3 months, or poor performance status, at 18.1 months, compared with 28.6 months overall. Subgroup analyses from both MARIPOSA and FLAURA2 suggest that patients with liver or bone metastases derive greater relative benefit from combination therapy. Baseline brain metastases tell a similar story: FLAURA2 showed median intracranial progression-free survival of 24.9 months with the combination versus 13.8 months with monotherapy, while MARIPOSA reported three-year intracranial progression-free survival of 38 percent versus 18 percent, though the review cautions these endpoints are not directly comparable across trials.

Molecular co-alterations add a second, more contested layer. TP53, the most frequently co-mutated tumor suppressor in EGFR-mutant disease, appears in roughly half to two-thirds of tumors and is consistently prognostic, shortening TKI benefit duration and overall survival on monotherapy. Whether it predicts differential benefit from combination therapy is murkier: MARIPOSA biomarker analyses hinted at a stronger progression-free survival benefit from amivantamab-lazertinib in TP53 co-mutated patients, while FLAURA2 found no differential effect. The most compelling prospective evidence comes from the phase 3 TOP trial, which randomized 294 patients with concurrent EGFR and TP53 mutations to osimertinib plus carboplatin-pemetrexed or monotherapy, and found median progression-free survival of 34.0 versus 15.8 months, effectively doubling disease control in this molecularly defined subgroup, albeit with substantially higher hematologic toxicity and immature overall survival data. The review concludes that TP53 status should be treated as a prognostic marker supporting heightened vigilance, not a definitive selection criterion.

Circulating tumor DNA offers a third emerging risk axis. Detectable baseline ctDNA correlates with greater systemic dissemination and shorter osimertinib benefit, and MARIPOSA analyses suggested a trend toward greater benefit from the combination in ctDNA-positive patients. But the authors are candid about the practical barriers: quantitative plasma profiling is not universally available, assay thresholds are not standardized across trials, and low or undetectable ctDNA does not reliably indicate low biological risk, particularly in CNS-only disease where tumor shedding is limited. Until a validated threshold exists, ctDNA should inform, never dictate, treatment selection.

The algorithm’s remaining steps address the patient rather than the tumor. Step three assesses fitness: patients with poor performance status, significant organ dysfunction, frailty by validated tools such as the G8 questionnaire, or age over 75 with meaningful comorbidities should default to monotherapy regardless of tumor biology. Supporting data are reassuring here, with studies in patients 75 and older showing osimertinib monotherapy outcomes closely approximating those of younger trial populations. Step four weighs preference and logistics. The subcutaneous formulation of amivantamab, validated in the PALOMA-3 trial, cuts infusion-related reactions from 66 percent to 13 percent and administration time from roughly five hours to under five minutes, potentially transforming the feasibility of MARIPOSA-based therapy, though annual costs exceeding 200,000 dollars remain a real barrier requiring proactive financial navigation.

Prophylactic management has become integral to the combination strategies themselves. The COCOON trial showed that structured dermatologic prophylaxis, including doxycycline, clindamycin lotion, chlorhexidine, and ceramide moisturizers, cut grade 2 or worse skin toxicity from 75 percent to 42 percent. Venous thromboembolism, occurring in roughly 40 percent of MARIPOSA patients versus 11 percent on osimertinib, can be approximately halved with four months of prophylactic anticoagulation, as demonstrated across the PALOMA program. The ongoing COPERNICUS study is now testing whether these toxicity-mitigated subcutaneous regimens hold up in a more representative real-world population.

The review is careful to acknowledge its own limits: the algorithm is a narrative synthesis, not prospectively validated evidence, and its biomarker arguments rest partly on unplanned subgroup analyses with conflicting results. But its central message is clear and timely. Combination therapy delivers genuine, statistically significant survival gains that should not be withheld from fit, high-risk patients, yet universal intensification would expose many patients to unnecessary toxicity, cost, and burden. Resistance biology reinforces the stakes: dual EGFR-MET blockade in MARIPOSA suppressed the dominant osimertinib resistance mechanisms, with MET amplification emerging in 13 percent of monotherapy patients versus 3 percent on the combination. Until prospective biomarker-stratified trials mature, the authors contend, structured shared decision-making anchored to tumor biology, patient fitness, and honest conversations about benefit and burden is the most defensible path through lung cancer’s new, more complicated first-line landscape.

Subject of Research: Risk-adapted first-line treatment selection for advanced EGFR-mutant non-small cell lung cancer

Article Title: Risk-adapted first-line therapy selection in advanced EGFR-mutant NSCLC: a practical clinical algorithm integrating tumor biology and patient factors

Article References: Sharma, A., Tan, J. K., Kumar, H., & Bharadwaj, H. R. (2026). Risk-adapted first-line therapy selection in advanced EGFR-mutant NSCLC: a practical clinical algorithm integrating tumor biology and patient factors. Medical Oncology, 43(10), Article 258. https://doi.org/10.1007/s12032-026-03338-y

Image Credits: AI Generated

DOI: 10.1007/s12032-026-03338-y

Keywords: EGFR-mutant NSCLC, osimertinib, amivantamab-lazertinib, MARIPOSA trial, FLAURA2 trial, risk-adapted therapy, TP53 co-mutation, circulating tumor DNA, subcutaneous amivantamab, venous thromboembolism, shared decision-making, targeted therapy