Predictive value of FDG-PET imaging for relapse in metastatic melanoma patients treated with immunotherapy.

BACKGROUND

Anti-PD1 immunotherapy has shown a sustainable clinical activity in patients with metastatic melanoma. However, strong predictive factors of the long-term response or risk of relapse remain to be identified.

OBJECTIVES

To determine whether FDG-PET imaging could be superior to CT scan in distinguishing residual tumours versus the absence of tumour in patients with a partial response (PR) or stable disease (SD) and whether a complete metabolic response (CMR) was associated with better outcomes.

METHODS

Retrospective study conducted in all patients with metastatic melanoma treated with anti-PD1 immunotherapy between October 2014 and October 2017 considered to be in complete remission. The primary outcome was the occurrence of a relapse during the follow-up. CT scan and FDG-PET scan had to be performed within a maximum of 2 months of treatment discontinuation. For CT imaging, the Response Evaluation Criteria in Solid Tumours (RECIST) 1.1 were used and included progressive disease (PD), SD, PR and complete response (CR). For FDG-PET imaging, the metabolic responses were classified as progressive metabolic disease, stable metabolic disease, residual FDG avidity (RFA) and CMR.

RESULTS

Twenty-six patients were in complete remission after collegial decision. Two patients had a SD on CT scan and a CMR on FDG-PET scan, and none of them relapsed. Ten patients had a PR on CT scan and a CMR on FDG-PET scan, and none of them relapsed. The mean treatment duration to achieve a complete remission was 7 months (3-23). A univariate analysis showed that a RFA assessed on the FDG-PET scan was significantly associated with a relapse (P = 0.00231).

CONCLUSIONS

Most patients with a PR on the CT scan and a CMR on the FDG-PET scan should be considered with a CR. Our study showed that FDG-PET imaging could play a crucial role in predicting the long-term outcome and help to decide whether treatment should be discontinued.