Dynamics of peripheral blood inflammatory index predict tumor pathological response and survival among patients with locally advanced non-small cell lung cancer who underwent neoadjuvant immunochemotherapy: a multi-cohort retrospective study.

BACKGROUND

Static tumor features before initiating anti-tumor treatment were insufficient to distinguish responding from non-responding tumors under the selective pressure of immuno-therapy. Herein we investigated the longitudinal dynamics of peripheral blood inflammatory indexes (dPBI) and its value in predicting major pathological response (MPR) in non-small cell lung cancer (NSCLC).

METHODS

A total of 147 patients with NSCLC who underwent neoadjuvant immunochemotherapy were retrospectively reviewed as training cohort, and 26 NSCLC patients from a phase II trial were included as validation cohort. Peripheral blood inflammatory indexes were collected at baseline and as posttreatment status; their dynamics were calculated as their posttreatment values minus their baseline level. Least absolute shrinkage and selection operator algorithm was utilized to screen out predictors for MPR, and a MPR score was integrated. We constructed a model incorporating this MPR score and clinical predictors for predicting MPR and evaluated its predictive capacity via the area under the curve (AUC) of the receiver operating characteristic and calibration curves. Furthermore, we sought to interpret this MPR score in the context of micro-RNA transcriptomic analysis in plasma exosomes for 12 paired samples (baseline and posttreatment) obtained from the training cohort.

RESULTS

Longitudinal dynamics of monocyte-lymphocyte ratio, platelet-to-lymphocyte ratio, platelet-to-albumin ratio, and prognostic nutritional index were screened out as significant indicators for MPR and a MPR score was integrated, which was further identified as an independent predictor of MPR. Then, we constructed a predictive model incorporating MPR score, histology, and differentiated degree, which discriminated MPR and non-MPR patients well in both the training and validation cohorts with an AUC value of 0.803 and 0.817, respectively. Furthermore, micro-RNA transcriptomic analysis revealed the association between our MPR score and immune regulation pathways. A significantly better event-free survival was seen in subpopulations with a high MPR score.

CONCLUSION

Our findings suggested that dPBI reflected responses to neoadjuvant immuno-chemotherapy for NSCLC. The MPR score, a non-invasive biomarker integrating their dynamics, captured the miRNA transcriptomic pattern in the tumor microenvironment and distinguished MPR from non-MPR for neoadjuvant immunochemotherapy, which could support the clinical decisions on the utilization of immune checkpoint inhibitor-based treatments in NSCLC patients.