A Risk Model to Predict the Delivery of Adjuvant Chemotherapy Following Lung Resection in Patients With Pathologically Positive Lymph Nodes

To investigate factors associated with the ability to receive adjuvant chemotherapy in patients with pathological N1 and N2 stage after anatomic lung resections for non-small cell lung cancer (NSCLC). Multicenter retrospective analysis on 707 consecutive patients found pathologic N1 (pN1) or N2 (pN2) disease following anatomic lung resections for NSCLC (2014-2019). Multiple imputation logistic regression was used to identify factors associated with adjuvant chemotherapy and to develop a model to predict the probability of starting this treatment. The model was externally validated in a population of 253 patients. In the derivation set, 442 patients were pN1 and 265 pN2. 58% received at least 1 cycle of adjuvant chemotherapy. The variables significantly associated with the probability of starting chemotherapy after multivariable regression analysis were: younger age (p < 0.0001), Body Mass Index (BMI) (p = 0.031), Forced Expiratory Volume in 1 second (FEV1) (p = 0.037), better performance status (PS) (p < 0.0001), absence of chronic kidney disease (CKD) (p = 0.016), resection lesser than pneumonectomy (p = 0.010). The logit of the prediction model was: 6.58 -0.112 x age +0.039 x BMI +0.009 x FEV1 -0.650 x PS -1.388 x CKD -0.550 x pneumonectomy. The predicted rate of adjuvant chemotherapy in the validation set was 59.2 and similar to the observed 1 (59%, p = 0.87) confirming the model performance in external setting. This study identified several factors associated with the probability of initiating adjuvant chemotherapy after lung resection in node-positive patients. This information can be used during preoperative multidisciplinary meetings and patients counseling to support decision-making process regarding the timing of systemic treatment. [Display omitted]