The value of F-FDG PET/CT for predicting or monitoring immunotherapy response in patients with metastatic melanoma: a systematic review and meta-analysis.

PURPOSE

To investigate the ability of F-FDG PET/CT to assess the response of patients with metastatic melanoma to immunotherapy.

METHODS

A comprehensive search of the literature for studies examining the prognostic value of F-FDG PET/CT in monitoring the response of patients with metastatic melanoma to immunotherapy was performed. We also screened the references of the selected articles to identify any other relevant studies. Detailed data were extracted and categorized. Comprehensive meta-analysis software was used for analysis.

RESULTS

Twenty four eligible articles were included in the systematic review. Based on the baseline F-FDG PET/CT imaging, the pooled hazard ratios of MTV, SLR, SUV/SULmax, SUV/SULpeak, and TLG for overall survival (OS) were 1.777 (95%CI: 1.389-2.275, p < 0.001), 3.425 (95%CI: 1.707-6.869, p = 0.001), 0.941 (95%CI: 0.599-1.477, p = 0.791), 1.704 (95%CI: 1.253-2.316, p = 0.016), and 1.755 (95%CI: 1.315-2.342, p < 0.001), respectively. The conventional and modified response assessment criteria exhibited a pooled sensitivity of 64% (95%CI: 46-79%) and 94% (95%CI: 81-99%) and a pooled specificity of 80% (95%CI: 59-93%) and 84% (95%CI: 64-95%), respectively, for the early F-FDG PET/CT scan. On the other hand, based on the late F-FDG PET/CT scan, the pooled sensitivity of 67% (95%CI: 35-90%) and 92% (95%CI: 73-99%) and pooled specificity of 77% (95%CI: 56-91%) and 76% (95%CI: 50-93%) were observed for the conventional and modified criteria, respectively. PET-detectable immune-related adverse events (irAEs) were associated with the response to therapy.

CONCLUSIONS

The baseline SUVpeak, MTV, and TLG parameters represent promising predictors of the final response of metastatic melanoma patients to immunotherapy. Modified response assessment criteria are potentially an appropriate method for monitoring immunotherapy. irAEs are also valuable for predicting eventual clinical benefit of treatment.