Pleural metastases carried worse two-year survival than contralateral lung metastases in stage M1a NSCLC

On chemotherapy, two-year survival was 18.5% with pleural metastases against 26.9% with contralateral lung disease

Among patients treated with chemotherapy, two-year overall survival was 26.9% (95% CI 24.9 to 28.9) for those with contralateral pulmonary metastases and 18.5% (95% CI 16.6 to 20.4) for those with pleural metastases. The confidence intervals do not overlap, and the cohort is large enough that both estimates are tight: 2701 patients in the contralateral group and 2590 in the pleural group.

The absolute difference of 8.4 percentage points is the figure to hold on to, because the same two patients carry the same stage. Under the 9th edition, pleural metastases and a metastasis in the opposite lung are both M1a, and a staging category exists to group patients whose outlook is comparable. These data report that the outlook of the two groups was not comparable.

The difference held across chemotherapy, immunotherapy and chemo-immunotherapy

With immune checkpoint inhibitor monotherapy, two-year survival was 49.5% (95% CI 43.4 to 55.2) with contralateral pulmonary metastases and 41.6% (95% CI 36.6 to 46.5) with pleural metastases. With chemo-immunotherapy the values were 40.4% (95% CI 36.5 to 44.2) and 28.6% (95% CI 24.9 to 32.4). The authors report that the difference in overall survival was present irrespective of the type of systemic treatment.

The consistency across three treatment strategies is what gives the finding its weight. A difference appearing under one regimen alone could be a feature of that regimen. A difference of 7.9 to 11.8 percentage points under all three points instead to the disease site itself, and the widest of the three gaps appeared with chemo-immunotherapy.

Patients with metastases at both sites had two-year survival of 12.1% on chemotherapy

The 341 patients who had both pleural and contralateral pulmonary metastases did worse than either single-site group: 12.1% (95% CI 8.2 to 16.7) with chemotherapy, 31.1% (95% CI 18.4 to 44.7) with immunotherapy alone and 25.0% (95% CI 15.7 to 35.4) with chemo-immunotherapy.

These estimates rest on a much smaller group, and the confidence intervals show it. The immunotherapy estimate of 31.1% spans 18.4 to 44.7, a range of 26.3 percentage points, so the ordering between the two-site group and the pleural-only group under immunotherapy is not established by these numbers. The chemotherapy estimate has the narrowest interval of the three, at 8.5 percentage points wide, and is the firmest of them.

Worse performance score in the pleural group, with no adjusted analysis in the abstract

The two groups differed before treatment started. Pleural metastases were more common in men, in patients with a worse performance score and in patients who received immune checkpoint inhibitor monotherapy, each at p<0.001. Performance score is a recognised prognostic factor in advanced NSCLC, so part of the survival gap may reflect who had pleural disease rather than the pleural disease itself.

The abstract reports these imbalances and reports the survival figures, and it does not state whether the comparison was adjusted for them. No hazard ratios and no median overall survival values appear in it. Registry data of this kind also carries the era effect of a study period running from 2010 to 2022, across which immune checkpoint inhibitors entered first-line treatment, as the study’s own three treatment groups reflect. Whether the full paper reports an analysis adjusted for any of this cannot be told from the abstract.

“Patients with pleural metastases have inferior OS compared to those with contralateral lung metastases. These findings indicate prognostic heterogeneity between M1a descriptors.”Noordhof et al., European Respiratory Journal (2026)

What follows from that conclusion is a staging question rather than a treatment question. The study compared treatments a patient had already been given and was not designed to show that a patient with pleural metastases should receive something different. Its finding concerns the descriptor: two conditions that the 9th edition treats as one category behaved differently over two years, in more than five thousand patients, under every systemic treatment examined.