Abstract Rationale As a novel molecular imaging technique, 68Ga-grazytracer PET/CT specifically targets granzyme B, a key enzyme released by cytotoxic T lymphocytes and natural killer cells to initiate tumor cell apoptosis. Bronchial arterial infusion chemotherapy (BAIC) is a treatment option for advanced lung cancer. This study aimed to evaluate the predictive value of 68Ga-grazytracer PET/CT for treatment response in advanced lung cancer patients receiving BAIC. Methods A total of 28 patients (mean age 64±7 years, 92.9% male) with advanced non-small cell lung cancer were enrolled. Most had squamous cell carcinoma (82.1%) and stage III disease (67.9%). Dynamic PET/CT scans were performed after the 1st and 2nd treatment cycles in 7 patients to assess parameter stability. Patients were followed for 6 months and categorized into a benefit group (complete response/partial response, n = 17) and a non-benefit group (progressive disease/stable disease, n = 11). Metabolic parameters (SUVmax) and CT-based morphological changes were compared. Predictive performance was evaluated using ROC analysis, Kaplan-Meier survival curves, and Cox regression. Results The 68Ga-grazytracer PET/CT metabolic parameters demonstrated excellent stability, with no significant differences in primary lesion SUVmax (2.98±0.62 vs. 3.05±0.51, P = 0.81) or metastatic lymph node SUVmax (2.55±0.44 vs. 2.60±0.47, P = 0.86) between the two cycles. At the 6-month follow-up, the overall progression-free survival (PFS) rate was 71.4%. The benefit group exhibited a significantly higher primary lesion SUVmax (3.27±0.61 vs. 2.23±0.19, P = 0.001) and a greater reduction in primary lesion size (ΔCT: -33.82±15.56% vs. -14.78±21.19%, P = 0.01). In ROC analysis, the primary lesion SUVmax showed outstanding predictive efficacy (AUC=0.989, P 0.001), with an optimal cutoff of 2.47 (100% sensitivity, 94.1% specificity). Kaplan-Meier analysis, using the median primary SUVmax value of 2.59 as the cutoff, revealed a significantly superior 6-month PFS for the high-uptake group (100.0% vs. 42.9%, Log-rank P = 0.001). Univariate COX analysis further confirmed primary SUVmax as a significant predictor of PFS (HR = 0.02, P = 0.01). Conclusion 68Ga-grazytracer PET/CT metabolic parameters demonstrate high stability and strong predictive value for BAIC efficacy in advanced non-small cell lung cancer. Primary lesion SUVmax is a robust early predictor of treatment benefit. This abstract is funded by: None
Zhang et al. (Fri,) studied this question.