Introduction

Lung cancer remains the leading cause of cancer death worldwide, though the trajectory has shifted meaningfully: lung cancer mortality in the United States has fallen by roughly half over the past 25 years. Sustaining that progress in early-stage disease depends on getting two decisions right: which patients benefit from immune-based or targeted therapy after surgery, and when in the treatment sequence that therapy should be given. Four separate randomized trials of adjuvant immunotherapy, plus a fresh readout on adjuvant targeted therapy, offer a clearer picture of both questions than the field has had before.

Adjuvant Immunotherapy: Four Trials, One Disappointing Pattern

Checkpoint inhibition transformed metastatic non-small cell lung cancer, and expectations ran high that giving it after surgery would do the same for early-stage disease. Chemotherapy alone, the prior standard in the adjuvant setting, produces only a modest and fading survival advantage after three decades of use, so the bar for improvement was not high.

Four trials have now tested that hypothesis, and none have cleared it convincingly.

IMpower010, testing adjuvant atezolizumab, showed a disease-free survival (DFS) benefit concentrated in patients with high PD-L1 tumor proportion scores (TPS), but overall survival (OS) curves have not separated even after five years of follow-up. KEYNOTE-091 (also known as PEARLS), testing pembrolizumab, showed a DFS benefit with no OS benefit reported yet, and the biomarker signal ran in the opposite direction: patients with high PD-L1 expression did not fare better than those with none. AstraZeneca's BR.31, testing durvalumab, was negative for its primary endpoint in the PD-L1-high population entirely.

The fourth and most recent entrant, ANVIL (ECOG-ACRIN EA5142, part of the ALCHEMIST platform), randomized a higher-risk population, mostly stage IIA to IIIA, to a year of nivolumab or observation after standard adjuvant chemotherapy. Presented this year, the trial found essentially no DFS benefit in the intent-to-treat population and only a nonsignificant trend toward benefit in patients with TPS above 50 percent. Overall survival showed no meaningful separation, and there was a suggestion the control arm simply performed better than expected historically, which may partly explain the null result. Adverse events, notably, still occurred, meaning the safety cost was not offset by a clear benefit.

Four negative-to-marginal trials do not mean adjuvant immunotherapy has no future. They mean it may be working in the wrong sequence.

Rewriting the Sequence: Why Neoadjuvant Changes the Math

Giving immunotherapy before surgery, while the tumor and its antigens are still present, appears to generate a stronger and more durable immune response than giving the same drug afterward. CheckMate 816, three cycles of neoadjuvant chemotherapy plus nivolumab followed by surgery, produced a substantial overall survival separation. KEYNOTE-671, a similar perioperative design with pembrolizumab, showed comparable benefit. By contrast, a year of adjuvant immunotherapy after surgery alone has, across the four trials above, produced little of consequence.

Cross-trial comparisons carry real limitations, but the pattern is consistent enough to inform practice today: patients with T3 or larger tumors, nodal disease, and tumors without an EGFR or ALK driver mutation eligible for chemoimmunotherapy should be considered for neoadjuvant treatment as a first-line strategy, decided within a multidisciplinary team. For patients who are incidentally upstaged after surgery, adjuvant chemotherapy and immunotherapy remain reasonable, but expectations for benefit should be modest. A clinical trial in the adjuvant setting remains a more compelling option than routine administration of a year of immunotherapy to PD-L1-negative patients.

Targeted Therapy: A Different Story Entirely

Where adjuvant immunotherapy has struggled, adjuvant targeted therapy for oncogene-driven disease has delivered some of the largest effect sizes seen in early-stage lung cancer.

EGFR-Mutated Disease: Generation Matters

ADAURA, testing adjuvant osimertinib in resected stage II-IIIA EGFR-mutated NSCLC, showed a disease-free survival hazard ratio near 0.2 and a roughly 50 percent reduction in the risk of death by five years, translating to an estimated one in eight patients spared death from cancer in that window. Longer follow-up shows a modest decline in benefit once patients stop treatment after three years, raising an open question about whether three years remains the right duration.

By contrast, Alliance A081105, the ALCHEMIST platform's EGFR-mutant substudy testing adjuvant erlotinib, a first-generation EGFR inhibitor, against observation, reported final overall survival results this year: five-year OS rates of 78.6 percent with erlotinib versus 77.9 percent with observation, no meaningful difference despite an earlier disease-free survival trend. The contrast between erlotinib and osimertinib in this same adjuvant setting is a useful reminder that not all EGFR inhibitors are interchangeable, and that generation and CNS penetration likely matter as much as target selection.

RET Fusion-Positive Disease: LIBRETTO-432

For the smaller subset of patients with RET fusion-positive tumors, LIBRETTO-432 enrolled roughly 150 patients with stage IB-IIIA disease to three years of adjuvant selpercatinib versus placebo, though the primary efficacy analysis was restricted to the roughly 100 patients with stage II-III disease. The hazard ratio for disease-free survival in that primary population, around 0.17, closely mirrors what was seen with osimertinib in EGFR-mutant disease, and benefit appeared consistent across subgroups. Selpercatinib carries a distinct safety profile, including hypertension and liver enzyme elevations, that requires active management with dose adjustment, but it is manageable in most patients over time.

Together, these results argue for broadening molecular testing beyond EGFR and ALK. Biomarker-driven perioperative therapy is expanding, and NCCN- and payer-covered next-generation sequencing panels increasingly make it feasible to identify these smaller, high-benefit subgroups before deciding on adjuvant strategy.

Moving Targeted Therapy Earlier, Too

The same logic driving immunotherapy toward the neoadjuvant setting is now being tested with targeted agents. The phase 3 NeoADAURA trial randomized patients with resectable EGFR-mutated NSCLC to neoadjuvant osimertinib alone, osimertinib plus chemotherapy, or chemotherapy alone. Major pathological response rates were 25 to 26 percent with the osimertinib-containing arms versus 2 percent with chemotherapy alone, a striking difference given how limited chemotherapy's neoadjuvant activity has historically been in this population. Whether this pathological response advantage will translate into an event-free or overall survival benefit remains to be confirmed with longer follow-up.

A separate, smaller single-arm study in a Chinese cohort with borderline resectable, node-positive ALK positive NSCLC- reportedly achieved pathological complete response in close to half of patients who went on to surgery after a short course of the third-generation ALK inhibitor, lorlatinib with substantial nodal downstaging. These truly unprecedented findings are provocative but need replication in larger, more diverse populations before they inform practice; the specific regimen and results should be independently verified against the primary presentation before publication.

For Patients

Two clear messages have emerged from this year's data for anyone facing a diagnosis of early-stage lung cancer. First, molecular testing (checking the tumor for EGFR, ALK, RET, and other targetable changes) should happen for every patient with lung cancer, regardless of stage, before starting any treatment, including immunotherapy. In several states, including Nebraska and Iowa, this kind of testing is now required by law for exactly this reason. Second, timing matters: for many patients, receiving treatment before surgery, rather than only afterward, is proving to offer a meaningfully better chance at long-term survival. Patients should feel comfortable asking their care team whether a multidisciplinary tumor board has reviewed their case and whether neoadjuvant treatment was considered before their surgery date was set.

Key Takeaways

  • Across four randomized trials (IMpower010, KEYNOTE-091/PEARLS, BR.31, and ANVIL/EA5142), adjuvant immunotherapy after surgery alone has produced, at best, modest disease-free survival benefit and no consistent overall survival advantage in early-stage NSCLC.

  • Giving chemotherapy plus immunotherapy before surgery (CheckMate 816, KEYNOTE-671) has produced substantially larger overall survival benefit than the same drugs given after surgery, supporting a shift toward neoadjuvant treatment for eligible patients.

  • Adjuvant targeted therapy for oncogene-driven disease remains the clearest adjuvant success story: osimertinib (ADAURA) and selpercatinib (LIBRETTO-432) each reduced the risk of disease-free survival events by 80 percent or more (hazard ratios of roughly 0.17-0.20) in their respective biomarker-defined populations; the roughly 50 percent reduction in the risk of death applies specifically to ADAURA's overall survival data, which LIBRETTO-432 has not yet reported.

  • Not all EGFR inhibitors perform equally in the adjuvant setting; a first-generation agent (erlotinib, Alliance A081105) failed to improve overall survival, unlike third-generation osimertinib.

  • Broad molecular testing at diagnosis, regardless of stage, is now essential to identify the patients most likely to benefit from these targeted approaches, and several states have passed laws mandating access to this testing.

  • Early NeoADAURA data show a strong pathological response signal with neoadjuvant osimertinib, suggesting it may be an appropriate option to consider in select patients with resectable EGFR-mutated disease; this does not yet establish neoadjuvant as superior to adjuvant targeted therapy, and survival data remain immature.

References

  1. Felip E, et al. Adjuvant atezolizumab after chemotherapy in resected stage II-IIIA non-small-cell lung cancer (IMpower010): a randomised, multicentre, open-label, phase 3 trial. Lancet. 2021.

  2. O'Brien M, et al. Pembrolizumab versus placebo as adjuvant therapy for completely resected stage IB-IIIA non-small-cell lung cancer (PEARLS/KEYNOTE-091). Lancet Oncol. 2022.

  3. Spigel DR, et al. Adjuvant durvalumab in resected stage II-III non-small-cell lung cancer (BR.31). Presented ASCO 2023.

  4. Chaft JE, et al. EA5142 adjuvant nivolumab in resected lung cancers (ANVIL): results of a randomized phase III ALCHEMIST trial. Presented ASCO 2026.

  5. Forde PM, et al. Neoadjuvant nivolumab plus chemotherapy in resectable lung cancer (CheckMate 816). N Engl J Med. 2022.

  6. Wakelee H, et al. Perioperative pembrolizumab for early-stage non-small-cell lung cancer (KEYNOTE-671). N Engl J Med. 2023.

  7. Herbst RS, et al. Adjuvant osimertinib for resected EGFR-mutated non-small-cell lung cancer (ADAURA). N Engl J Med. 2020; updated survival analyses through 2023.

  8. Govindan R, et al. Adjuvant erlotinib versus observation after complete resection of EGFR-mutant NSCLC: final overall survival results of Alliance A081105. J Clin Oncol. 2026.

  9. Zhou C, et al. LIBRETTO-432: adjuvant selpercatinib in resected stage IB-IIIA RET fusion-positive non-small-cell lung cancer. Presented ASCO Plenary 2026.

  10. Tsuboi M, et al. Neoadjuvant osimertinib with or without chemotherapy for resectable EGFR-mutated non-small-cell lung cancer (NeoADAURA). Presented ASCO 2025-2026.