Comparison of the predictive value among inflammation-based scoring systems for bleomycin pulmonary toxicity in patients with germ cell tumors.

Objective: To compare the predictive value of pretreatment inflammation-based scoring systems in patients with germ cell tumors receiving first-line bleomycin-based chemotherapy.

Methods: Retrospectively, we evaluated 57 patients with germ cell tumors. Bleomycin pulmonary toxicity was defined as the presence of asymptomatic decline in pulmonary function tests, pulmonary symptoms or interstitial pneumonia on computed tomography in the absence of infection. The neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, systemic immune-inflammation index, albumin-to-globulin ratio, Prognostic Nutritional Index, Glasgow Prognostic Score and C-reactive protein were measured in all patients. To assess the predictive ability of each scoring system, the area under the receiver operating characteristic curve was calculated, and multivariate analysis was carried out to identify the predictive scores associated with bleomycin pulmonary toxicity.

Results: Of the 57 patients, 15 patients developed bleomycin pulmonary toxicity. The neutrophil-to-lymphocyte ratio had the highest area under the curve value (0.763) of all inflammation-based scoring systems, followed by the Prognostic Nutritional Index (0.749). In multivariate analysis, the neutrophil-to-lymphocyte ratio (odds ratio 11.5; P = 0.009) and Prognostic Nutritional Index (odds ratio 9.07; P = 0.013) were independently associated with development of bleomycin pulmonary toxicity. As these two independent markers were combined, the area under the curve achieved the highest value (0.822).

Conclusions: The present study shows that the neutrophil-to-lymphocyte ratio and Prognostic Nutritional Index are independent risk factors for development of bleomycin pulmonary toxicity. The combination of the neutrophil-to-lymphocyte ratio and Prognostic Nutritional Index seems to have superior predictive value compared with other inflammation-based scoring systems.