Background: We aimed to investigate the predictive value of a systematic serum inflammation index, pan-immune-inflammatory value (PIV), in pathological complete response (pCR) of patients treated with neoadjuvant immunotherapy to further promote ideal patients’ selection.
Methods: The clinicopathological and baseline laboratory information of 128 NSCLC patients receiving neoadjuvant immunochemotherapy between October 2019 and April 2022 were retrospectively reviewed. We performed least absolute shrinkage and selection operator (LASSO) algorithm to screen candidate serum biomarkers for predicting pCR, which further entered the multivariate logistic regression model to determine final biomarkers. Accordingly, a diagnostic model for predicting individual pCR was established. Kaplan–Meier method was utilized to estimate curves of disease-free survival (DFS), and the Log rank test was analyzed to compare DFS differences between patients with and without pCR.
Results: Patients with NSCLC heterogeneously responded to neoadjuvant immunotherapy, and those with pCR had a significant longer DFS than patients without pCR. Through LASSO and the multivariate logistic regression model, PIV was identified as a predictor for predicting pCR of patients. Subsequently, a diagnostic model integrating with PIV, differentiated degree and histological type was constructed to predict pCR, which presented a satisfactory predictive power (AUC, 0.736), significant agreement between actual and our nomogram-predicted pathological response.
Conclusion: Baseline PIV was an independent predictor of pCR for NSCLC patients receiving neoadjuvant immunochemotherapy. A significantly longer DFS was achieved in patients with pCR rather than those without pCR; thus, the PIV-based diagnostic model might serve as a practical tool to identify ideal patients for neoadjuvant immunotherapeutic guidance.
Keywords: non-small-cell lung cancer, pan-immune-inflammatory value, neoadjuvant immunochemotherapy, pathological complete response, survival benefits