Absolute number of new lesions on 18F-FDG PET/CT is more predictive of clinical response than SUV changes in metastatic melanoma patients receiving ipilimumab.

Abstract

Evaluation of response to immunotherapy is a matter of debate. The aim of the present study was to evaluate the response of metastatic melanoma to treatment with ipilimumab by means of 18F-FDG PET/CT, using the patients' clinical response as reference. The final cohort included in the analyses consisted of 41 patients with metastatic melanoma who underwent 18F-FDG PET/CT before and after administration of ipilimumab. After determination of the best clinical response, the PET/CT scans were reviewed and a separate independent analysis was performed, based on the number and functional size of newly emerged 18F-FDG-avid lesions, as well as on the SUV changes after therapy. The median observation time of the patients after therapy was 21.4months (range 6.3-41.9months). Based on their clinical response, patients were dichotomized into those with clinical benefit (CB) and those without CB (No-CB). The CB group (31 patients) included those with stable disease, partial remission and complete remission, and the No-CB group (10 patients) included those with progressive disease. The application of a threshold of four newly emerged 18F-FDG-avid lesions on the posttherapy PET/CT scan led to a sensitivity (correctly predicting CB) of 84% and a specificity (correctly predicting No-CB) of 100%. This cut-off was lower for lesions with larger functional diameters (three new lesions larger than 1.0cm and two new lesions larger than 1.5cm). SUV changes after therapy did not correlate with clinical response. Based on these findings, we developed criteria for predicting clinical response to immunotherapy by means of 18F-FDG PET/CT (PET Response Evaluation Criteria for Immunotherapy, PERCIMT). Our results show that a cut-off of four newly emerged 18F-FDG-avid lesions on posttherapy PET/CT gives a reliable indication of treatment failure in patients under ipilimumab treatment. Moreover, the functional size of the new lesions plays an important role in predicting the clinical response. Validation of these results in larger cohorts of patients is warranted.