Utility of F-FDG PET/CT uptake values in predicting response to neoadjuvant chemoimmunotherapy in resectable non-small cell lung cancer.

INTRODUCTION

Reliable predictive markers are lacking for resectable non-small cell lung cancer (NSCLC) patients treated with neoadjuvant chemoimmunotherapy. The present study investigated the utility of SUV values acquired from PET/CT to predict the response to neoadjuvant chemoimmunotherapy for resectable NSCLC.

MATERAL AND METHODS

SUV, clinical and pathological outcomes, were collected from patients in 5 hospitals. Patients who received dynamic PET/CT surveillance were divided into cohorts A (chemoimmunotherapy) and B (chemotherapy), respectively, while cohort C (chemoimmunotherapy) comprised patients undergoing post-therapy PET/CT. Associations between SUV and major pathologic response (MPR) were evaluated through receiver operating characteristic (ROC) curves.

RESULTS

A total of 129 cases with an MPR rate of 46.5 % was identified. In neoadjuvant chemoimmunotherapy, ΔSUV% (AUC: 0.890, 95 % CI: 0.761-0.949) and post-therapy SUV (AUC: 0.933, 95 % CI: 0.802-0.959) could accurately predict MPR. On the contrary, the baseline SUV was not associated with MPR (p = 0.184). Furthermore, an independent cohort C proved that post-therapy SUV could serve as an independent predictor (AUC: 0.928, 95 % CI: 0.823-0.958). In addition, robust predictive performance could be observed when we use the optimal cut-off point of both ΔSUV% (54.4 %, AUC: 0.912, 95 % CI: 0.824-0.994) and post-therapy SUV (3.565, AUC: 0.912, 95 % CI: 0.824-0.994) in neoadjuvant chemoimmunotherapy. The RNA data revealed that the expression of PFKFB4, a key enzyme in glycolysis, was positively correlated with SUV value and tumor cell proliferation after neoadjuvant chemoimmunotherapy.

CONCLUSION

These findings highlighted that the ΔSUV% and remained SUV were accurate and non-invasive tests for the prediction of MPR after neoadjuvant chemoimmunotherapy.