LEAP-008: pembrolizumab plus lenvatinib after immunotherapy in non-small cell lung cancer Authors: Koji Fukuda Published: 2025 The advent of immune checkpoint inhibitors (ICIs) has transformed first-line therapy for advanced non-small cell lung cancer (NSCLC), yet most patients eventually experience disease progression on or after programmed cell death protein 1 (PD-1)/programmed death ligand 1 (PD-L1) inhibitor therapy. Optimal management of these patients remains an urgent clinical challenge. Leighl et al. address this issue in the LEAP-008 trial, a phase III study evaluating pembrolizumab plus lenvatinib versus standard docetaxel chemotherapy in metastatic NSCLC patients who progressed after platinum-based chemotherapy and prior anti-PD-1/PD-L1 therapy (1). In LEAP-008, docetaxel was chosen as the control arm given that second-line docetaxel (with or without an antiangiogenic agent) is an established standard in this setting—prior trials combining docetaxel with ramucirumab (REVEL) or nintedanib (LUME-Lung1) have demonstrated modest improvements in overall survival (OS) compared to docetaxel alone (2,3). The rationale for the experimental arm was to continue PD-1 blockade beyond progression while adding an antiangiogenic agent (lenvatinib) to potentially re-sensitize tumors to immunotherapy by modulating the tumor microenvironment. This strategy has proven effective in other malignancies (e.g. renal cell and endometrial cancers), prompting investigation in NSCLC. What LEAP-008 showed Unfortunately, LEAP-008 did not show any significant benefit from the pembrolizumab-lenvatinib combination. The trial’s co-primary endpoints of progression-free and OS were not met (1). Median progression-free survival (PFS) was similar between the lenvatinib-pembrolizumab arm and the docetaxel arm [hazard ratio (HR) =0.89, P not significant], and median OS was virtually identical (approximately 11.3 versus 12.0 months; HR =0.98) (1). The objective response rate (ORR) with pembrolizumab-lenvatinib was modest (23%) and not significantly higher than with docetaxel (1). In addition, toxicity was greater with the combination therapy, as expected: rates of grade ≥3 adverse events (e.g., hypertension, fatigue, hepatic and gastrointestinal toxicities) were higher in the lenvatinib-containing arm, leading to more frequent treatment discontinuations (1). Thus, continuing immunotherapy beyond initial progression, even with the addition of a multitargeted VEGF inhibitor, failed to improve outcomes in an unselected NSCLC population. These results underscore the difficulty of surpassing the efficacy of cytotoxic chemotherapy in the post-immunotherapy setting. Why might LEAP‑008 have been negative? From a mechanistic and practical standpoint, three non-exclusive factors likely contributed. First, the biology of post-ICI resistance—including impaired antigen presentation, entrenched T-cell exhaustion, and myeloid-driven exclusion—may not be sufficiently reversed by broad VEGF/FGFR blockade alone, explaining the ORR-OS disconnect despite modest shrinkage signals. Second, the therapeutic index matters: the higher rates of grade ≥3 toxicities and the frequent lenvatinib interruptions (59%) and dose reductions (46%) in LEAP-008 plausibly attenuated the sustained exposure needed to realize immunomodulatory synergy. Third, selectivity versus polypharmacology may be pivotal; emerging patterns suggest that antibody-based VEGF/VEGFR strategies might condition the tumor microenvironment more cleanly—and with better tolerability—than multi-target tyrosine kinase inhibitors (TKIs) when combined with PD-1/PD-L1 blockade. Alignment with other phase III evidence The negative outcome of LEAP-008 is consistent with several other recent phase III trials that attempted to improve post-immunotherapy outcomes by adding novel agents to standard therapy. For example, the CONTACT-01 trial evaluated atezolizumab (a PD-L1 inhibitor) plus the multi-kinase inhibitor cabozantinib versus docetaxel in metastatic NSCLC patients previously treated with chemoimmunotherapy—and found no improvement in survival with the combination (4). Similarly, the SAPPHIRE trial compared nivolumab (PD-1 inhibitor) plus the TKI sitravatinib against docetaxel in advanced non-squamous NSCLC after immunotherapy and reported no significant OS benefit (median OS =12.2 months in both arms) (5). A different approach was taken in TROPION-Lung01, which tested an antibody-drug conjugate (ADC): datopotamab deruxtecan (targeting TROP2) versus docetaxel in previously treated advanced NSCLC. TROPION-Lung01 did meet its primary endpoint of improved PFS (median PFS 4.4 versus 3.7 months, HR =0.75), but so far has shown only a trend toward better OS with the ADC and no statistically significant survival gain (6). Taken together, these trials—along with LEAP-008—suggest that simply adding another agent [whether a VEGF pathway inhibitor, a Tyro3, Axl, and Mer (TAM) receptor inhibitor, or even a novel cytotoxic conjugate] to second-line therapy has not yielded major advances in an unselected patient population. Docetaxel-based chemotherapy, possibly combined with an antiangiogenic, remains a tough comparator that has yet to be clearly surpassed in the immunotherapy-refractory setting. It is worth noting that disappointing combination results have not been limited to the salvage setting. Parallel efforts to intensify first-line treatment with immunotherapy plus antiangiogenics in NSCLC have also failed to improve patient outcomes. In the phase III LEAP-006 trial, adding lenvatinib to first-line pembrolizumab + chemotherapy (pemetrexed–platinum) did not significantly prolong PFS or OS compared to pembrolizumab + chemotherapy alone (7). Likewise, the LEAP-007 trial tested pembrolizumab combined with lenvatinib versus pembrolizumab monotherapy in treatment-naïve metastatic NSCLC with PD-L1 tumor proportion score ≥1%; this study again showed no benefit from the addition of lenvatinib, as neither median survival nor response rates improved with the combination (8). These concordant results in the frontline setting indicated early on that lenvatinib was not contributing meaningful additive efficacy to PD-1 blockade in NSCLC—a finding now reinforced by the negative second-line LEAP-008 outcome. The lack of synergy may stem from underlying tumor biology; it is possible that once NSCLC has progressed on immunotherapy, simply continuing PD-1 inhibition (even with enhanced anti-angiogenic effects) is insufficient to overcome resistance mechanisms. Furthermore, overlapping toxicities can limit the ability to deliver combination therapy at full intensity, potentially eroding any incremental efficacy. Positioning LEAP‑008 within the VEGF-immuno-oncology landscape Across settings, outcomes diverge by how the VEGF axis is targeted. In the first-line metastatic non-squamous NSCLC population, IMpower150 showed that adding bevacizumab to atezolizumab + chemotherapy (ABCP) improved OS versus bevacizumab + chemotherapy (BCP), supporting the immunologic complementarity of antibodybased VEGF blockade with PD-L1 inhibition (9). Conversely, three lenvatinib-containing phase III trials (LEAP-006/-007/-008) and other multitarget TKI + PD-1/PD-L1 combinations (CONTACT-01, SAPPHIRE) were uniformly negative for OS benefit in modern, ICI-exposed populations (1,2,5,7,8). In contrast, the Lung-MAP S1800A phase II trial reported significantly improved OS with pembrolizumab + ramucirumab versus standard options postICI, generating the hypothesis that selective, antibody-based VEGFR2 blockade may partner better with PD-1 inhibition than polypharmacologic TKIs (10). Emerging PD-1/VEGF bispecific antibodies A next step toward “on-target” synergy is bispecific antibodies that co-engage PD-1 and VEGF-A within a single molecule. Ivonescimab has recently demonstrated PFS superiority against pembrolizumab in previously untreated PD-L1-positive NSCLC (HARMONi-2), with OS still maturing and global validation ongoing (11). In the EGFRmutated, post-TKI setting, ivonescimab + chemotherapy (HARMONi-A) significantly prolonged PFS versus chemotherapy alone, with press-release OS data now suggesting an OS benefit in that China-only study; peer-reviewed OS details are pending (12). Although early, these results—together with the S1800A signal—support the notion that antibody-centric VEGF modulation may achieve a more favorable therapeutic index and immune conditioning than multitarget TKIs when paired with PD-1/PD-L1 blockade. Implications and next steps To make progress after anti-PD-1/PD-L1 plus platinum, we likely need to move beyond one-size-fits-all add-ons. A biomarker‑directed approach (e.g., T-cell-inflamed signatures as explored in KEYNOTE-495) should guide mechanism-focused regimens—PD-1 + selective anti-VEGFR antibodies or PD-1 + ADCs—with rigorous patientreported outcomes to capture clinical value (1,10,11,13). Conclusions LEAP-008 trial provides a sobering reminder that more therapy is not always better, especially without patient selection. Pembrolizumab plus lenvatinib failed to improve survival over standard chemotherapy in PD-1/PD-L1-pretreated NSCLC (1), aligning with other phase 3 studies that have tested add-on agents in this setting (4-6). While these results are disappointing, they offer valuable insights to inform future research. The focus now shifts toward selective, antibody‑based VEGF strategies and biomarker‑guided combinations as potentially more fruitful avenues to tackle immunotherapyresistant lung cancer (10-12). Supplementary The article’s supplementary files as Acknowledgments None. Ethical Statement: The author is accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Footnotes Provenance and Peer Review: This article was commissioned by the editorial office, Journal of Thoracic Disease. The article has undergone external peer review. Funding: None. Conflicts of Interest: The author has completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1815/coif). The author has no conflicts of interest to declare. References Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1815/prf Supplementary Materials The article’s supplementary files as