Dr Rotow on DESTINY-Lung04 Outcomes at WCLC 2026

In the end, the key message here is having access to multiple effective lines of treatment. Multiple effective classes of treatment, we all believe, improves survival of patients with lung cancer. The control arm just had access to a lot more different classes of effective therapy during their whole treatment course post-study.

Julia K. Rotow, MD, clinical director, Lowe Center for Thoracic Oncology, Dana-Farber Cancer Institute, and assistant professor of medicine, Harvard Medical School, discussed subgroup and post-progression findings from the phase 3 DESTINY-Lung04 trial in HER2-mutant non–small cell lung cancer (NSCLC), presented at the IASLC 2026 World Conference on Lung Cancer (WCLC) in Seoul.

Rotow presented primary results from DESTINY-Lung04, which randomized patients with previously untreated HER2-mutant metastatic NSCLC to trastuzumab deruxtecan (T-DXd; Enhertu) or platinum, pemetrexed, and pembrolizumab (Keytruda). Median progression-free survival was 14.3 months with T-DXd vs 8.3 months with chemoimmunotherapy (HR, 0.63), a statistically significant difference.

Median overall survival (OS) was 29 - 33 months across both arms, and the interim OS analysis did not reach significance. Approximately 93% of patients had HER2 exon 20 mutations, all within the tyrosine kinase domain.

Rotow said the control arm performed better than expected, with PFS similar to the overall population of KEYNOTE-189, which was a surprise in a disease not thought to be immunotherapy sensitive. Even so, prespecified subgroups consistently favored T-DXd, with most confidence intervals falling below 1. That included patients with a history of tobacco use, roughly 35 - 40% of the study population and a common proxy for immune sensitivity, where there was no clear chemoimmunotherapy signal.

PD-L1 subgroups were not shown in the presentation but were examined, she said. Results favored T-DXd in PD-L1–negative and –low patients, who made up the vast majority, and directionally in the small PD-L1–high group, where confidence intervals were wide. About one-third of patients had unknown PD-L1 status, and their outcomes matched the overall population. Given that nearly all patients had exon 20 mutations, she does not expect a detectable difference by mutation subtype.

Post-progression therapy, shown compressed across all subsequent lines, was substantially imbalanced. During the trial, T-DXd and HER2 TKIs became available in the previously treated setting, creating de facto crossover: control-arm patients were more likely to receive next-line HER2-directed therapy, next-generation HER2-selective TKIs, and multiple lines of HER2-directed treatment. Patients on the T-DXd arm mostly received chemotherapy next, and many did not receive immunotherapy in any subsequent line, likely reflecting a belief in the community that checkpoint inhibitors are inactive in HER2-mutant disease.

Whether that belief is correct remains open, Rotow said. Retrospective single-agent data show low response rates, more on par with EGFR-mutant disease, but real-world data on chemoimmunotherapy combinations show some activity, and DESTINY-Lung04 adds to that body of evidence. The median OS in both arms is the best yet seen in HER2-mutant lung cancer, she said, reflecting how far the field has come when patients can access every effective drug class.