Abstract:

Objective To investigate the risk factors of pulmonary invasive fungal disease （IFD） in patients with primary non-small cell lung cancer （NSCLC） after radiotherapy and to construct predictive model. Methods A total of 298 patients with primary NSCLC who received radiotherapy in the Second People's Hospital of Yibin of Sichuan from January 2020 to January 2024 were retrospectively included as the study objects. The incidence of pulmonary IFD after radiotherapy was analyzed. Univariate and multivariate analyses were performed on the risk factors of pulmonary IFD in patients with primary NSCLC after radiotherapy. A logistic prediction model was constructed according to the results of multivariate analysis，and the predictive efficacy of each index was evaluated by receiver operator characteristic （ROC） curve. Results There were 61 cases with pulmonary IFD after radiotherapy in all 298 patients，with the incidence of 20.47%. And 73 strains fungi were detected，including 57 strains for Candida and 16 strains for Aspergillus. There were statistically significant differences in age （χ²=23.13，P<0.001），whether they had type 2 diabetes （χ²=19.28，P<0.001），whether they underwent invasive procedures （χ²=17.49，P<0.001），and concurrent chemoradiotherapy （χ²=18.48，P<0.001） between IFD patients and non-IFD patients. Multivariate analysis showed that age≥65 years （OR=4.64，95%CI：2.12-10.13，P<0.001），combined type 2 diabetes （OR=5.63，95%CI：2.19-14.48，P<0.001），concurrent chemoradiotherapy （OR=3.73，95%CI：1.74-8.02，P=0.001） and invasive procedures （OR=5.11，95%CI：2.33-11.19，P<0.001） were independent risk factors for pulmonary IFD in patients with primary NSCLC after radiotherapy. Based on the above indexes，the logistic prediction model was constructed as follows：logit（P）=-4.59+1.53×age+1.73×combined type 2 diabetes+1.32×concurrent chemoradiotherapy+ 1.63×acceptance of invasive procedures （R²=0.852）. ROC curve analysis showed that the area under the curve of pulmonary IFD in patients with primary NSCLC who were≥65 years old，combined with type 2 diabetes，receiving invasive procedures，concurrent chemoradiotherapy，and regression model P value were 0.68，0.63，0.68，0.68，0.82，respectively. Conclusion The incidence of pulmonary IFD in patients with primary NSCLC after radiotherapy is independently related to age，type 2 diabetes，invasive procedures and concurrent chemoradiotherapy. The prediction model constructed by using the above four factors has good efficacy in predicting IFD in patients' lungs.

Key words: Carcinoma, non-small-cell lung, Radiotherapy, Invasive fungal infections, Risk factors, Prediction model