Using 18F-FDG PET/CT to predict programmed cell death ligand expression in non-small cell lung cancer via metabolic...

The purpose of this study is to evaluate the effectiveness of using 18F-FDG PET/CT metabolic heterogeneity to assess the programmed cell death ligand (PD-L1) expression in primary tumours.

Data from 103 non-small cell lung cancer (NSCLC) patients undergoing 18F-FDG PET/CT were collected. PD-L1 expression was verified via biopsy or surgical specimens. The coefficient of variation (COV) assessed metabolic heterogeneity of the primary tumour. ROC curves evaluated the predictive potential of metabolic metrics and defined thresholds. Logistic regression examined predictors of PD-L1 expression.

The study included 103 patients (mean age: 63.65 ± 9.28 years), of whom 60 were male. Sixty-four patients had positive PD-L1 expression, while 39 had negative PD-L1 expression. COV was significantly higher in the PD-L1-positive group (Z = -2.529, P = 0.011), while no significant differences were noted in other parameters between the groups (P > 0.05 for all). The optimal cut-off value was proposed as 28.9, with sensitivity and specificity of 46.9% (34.3%-59.8%) and 82.1% (66.5%-92.5%), respectively (AUC: 0.649 (0.549, 0.741)) which can more effectively identify PD-L1-negative patients. Other metabolic parameters are less effective than COV (AUC< 0.6). In addition, COV-defined metabolic heterogeneity outperformed other metabolic parameters in predicting PD-L1 expression (P = 0.049) and emerged as an independent predictor.

Metabolic heterogeneity, described by the COV of the primary lesion, is a marker for predicting PD-L1 expression in NSCLC patients. Therefore, the COV of the primary tumour may complement conventional imaging in providing immunohistochemical information before biopsy.

COV of the primary tumour can predict PD-L1 expression, potentially complementing conventional imaging for immunohistochemical information prior to biopsy.