Baseline plasma KL-6 level predicts adverse outcomes in patients with idiopathic pulmonary fibrosis receiving nintedanib: a retrospective real-world cohort study.

Background: Nintedanib is effective for treating idiopathic pulmonary fibrosis (IPF), but some patients may exhibit a suboptimal response and develop on-treatment acute exacerbation (AE-IPF), hepatic injury, or mortality. It remains unclear which patients are at risk for these adverse outcomes.

Methods: We analysed the demographic and clinical data, baseline plasma levels of Krebs von den Lungen-6 (KL-6) and surfactant protein A (SPA), and longitudinal clinical courses of a real-world cohort of IPF patients who received nintedanib ≥ 14 days between March 2017 and December 2020. Cox proportional-hazards regression, subdistribution hazards regression, and sensitivity analyses were performed to investigate the association between baseline predictors and AE-IPF, mortality, and nintedanib-related hepatic injury. The relationship between baseline predictors and pulmonary function decline was determined.

Results: Fifty-seven patients were included, of whom 24 (42%) developed hepatic injury, 20 (35%) had AE-IPF, and 16 (28%) died on-treatment. A baseline plasma KL-6 level ≥ 2.5 ng/mL, and diffusion capacity for carbon monoxide (D) < 55% predicted, were associated with increased risk of hepatic injury (adjusted hazard ratio [aHR] was 3.46; 95% CI 1.13-10.60; p = 0.029 for KL-6, and 6.05; 95% CI 1.89-19.32; p = 0.002 for D). Both factors also predicted severe and recurrent hepatic injury. Patients with baseline KL-6 ≥ 2.5 ng/mL also had a higher risk of AE-IPF (aHR 4.52; 95% CI 1.63-12.55; p = 0.004). For on-treatment mortality, baseline KL-6 ≥ 3.5 ng/mL and SPA ≥ 600 pg/mL were significant predictors (aHR 5.39; 95% CI 1.16-24.97; p = 0.031 for KL-6, and aHR 12.28; 95% CI 2.06-73.05; p = 0.006 for SPA). Results from subdistribution hazard regression and sensitivity analyses supported these findings. Patients with elevated baseline plasma KL-6 levels also exhibited a trend towards faster pulmonary function decline.

Conclusions: For patients with IPF who are receiving nintedanib, we have identified baseline predictors, in particular plasma KL-6 levels, for the risk of adverse outcomes. Patients with these predictors may require close monitoring for unfavourable responses during treatment. Our findings also support the prognostic role of molecular markers like KL-6 and may contribute to future formulation of more individualized therapeutic strategies for IPF.