Significance of clinical parameters and biomarkers to predict nintedanib-induced diarrhea: an interview-based retrospective study.

Background: Idiopathic pulmonary fibrosis (IPF) is a poor prognostic fibrotic interstitial lung disease (ILD) of unknown etiology. Similarly, the prognosis of patients with ILD and progressive pulmonary fibrosis (PPF) is poor. Nintedanib reduces the decline in forced vital capacity (FVC) and improves the survival of patients with IPF and non-IPF ILDs meeting PPF criteria (hereafter "PPF"). Diarrhea is a significant adverse event associated with nintedanib, and it is sometimes the reason for the discontinuation of the drug. In this study, we aimed to identify clinical predictors of nintedanib-induced diarrhea and to clarify significance of ILD biomarkers to predict their occurrence after the adjustment using the significant clinical predictors. Methods: Seventy-nine patients with ILDs treated with nintedanib were included in this study. Diarrhea was retrospectively evaluated based on interviews, and medical records were reviewed for other clinical findings. Furthermore, biomarkers including surfactant protein-D (SP-D) serum levels and peripheral blood monocyte counts were examined. Parameters' predictive abilities were examined using univariate and multivariate logistic regression analyses. Results: Participants comprised 57 males and 22 non-smokers. The underlying ILDs included IPF (n=39) and PPF (n=40). PPF included idiopathic interstitial pneumonia (IIP) other than IPF (n=19), fibrotic hypersensitivity pneumonitis (FHP) (n=8), connective tissue disease-related ILDs (CTD-ILDs) (n=8), and other ILDs (n=5). Fourteen patients underwent corticosteroid therapy at the initiation of nintedanib. Nintedanib-induced diarrhea occurred within 3 months in 47 patients (IPF, n=30; PPF, n=17). IPF, no corticosteroid therapy, nintedanib per body surface area (BSA), and %FVC ≤80% were associated with the occurrence of diarrhea within 3 months after commencing nintedanib treatment by multivariate logistic regression analysis. Additionally, monocyte counts ≤650/µL and serum SP-D >157.5 ng/mL were associated with occurrence of diarrhea after the adjustment of other factors. Conclusions: Nintedanib-induced diarrhea is significantly associated with various complex factors. IPF, no corticosteroid therapy, a higher nintedanib dose per BSA, and a lower %FVC were associated with the occurrence of diarrhea within 3 months of initiating nintedanib therapy. Lower monocyte counts and higher levels of serum SP-D at the initiation of nintedanib might suggest occurrence of diarrhea. Although large-scale studies are needed to draw definite conclusions regarding our hypothesis, the results of our study and hypothesis of nintedanib-induced diarrhea might suggest future research direction and lead to new management of nintedanib-induced diarrhea.