A simple risk scoring system for predicting acute exacerbation of interstitial pneumonia after pulmonary resection in lung cancer patients.

Abstract

Lung cancer patients with interstitial lung diseases (ILDs) who have undergone pulmonary resection often develop acute exacerbation of interstitial pneumonia (AE) in the post-operative period. To predict who is at high risk of AE, we propose a scoring system that evaluates the risk of AE in lung cancer patients with ILDs. We derived a score for 30-day risk of AE onset after pulmonary resection in lung cancer patients with ILDs (n = 1,022; outcome: risk of AE) based on seven risk factors for AE that were identified in a previous retrospective multi-institutional cohort study. A logistic regression model was employed to develop a risk prediction model for AE. A risk score (RS) was derived: 5 × (history of AE) + 4 × (surgical procedures) + 4 × (UIP appearance in CT scan) + 3 × (male sex) + 3 × (preoperative steroid use) + 2 × (elevated serum sialylated carbohydrate antigen, KL-6 level) + 1 × (low vital capacity). The RS was shown to be moderately discriminatory with a c-index of 0.709 and accurate with the Hosmer-Lemeshow goodness-of-fit test (p = 0.907). The patients were classified into three groups: low risk (RS: 0-10; predicted probability <0.1; n = 439), intermediate risk (RS: 11-14; predicted probability 0.1-0.25; n = 559), and high risk (RS: 15-22; predicted probability >0.25; n = 24). Although further validation and refinement are needed, the risk score can be used in routine clinical practice to identify high risk individuals and to select proper treatment strategies.