Evaluation of 18F-FDG-PET/CT for Response Assessment in Patients with Advanced Melanoma Treated with Pembrolizumab Checkpoint Inhibitor Monotherapy

133 Objectives: Pembrolizumab is an immune-checkpoint inhibitor that blocks the programmed cell death receptor 1 (PD-1) and improves treatment outcomes for patients with advanced melanoma. The aim of the study was to evaluate the changes seen in quantitative 18F-FDG PET/CT in assessing response of melanoma patients receiving pembrolizumab anti-PD-1 monotherapy. Methods: Under an IRB-approved protocol, we identified 30 patients diagnosed with advanced melanoma and treated with single-agent pembrolizumab between January 2008 and December 2017 with baseline (PET0) and interim response (PET1) 18F-FDG-PET/CT scans. Three patients were excluded as their PET0 scan showed no visible disease. Of the remaining 27 patients, PET0 scans were performed at a median 18 days (range 0-58) before treatment initiation and PET1 scans at a median of 108 days (range 81-229) after PET0 scans. These scans were analyzed for SUVmax and SULpeak of the most FDG-avid lesion. An additional six patients were excluded because they had non-measurable disease on the non-contrast CT (bone or liver lesions or resected disease), leaving 21 patients with scans analyzed for short-axis and perpendicular long-axis measurements of the most FDG-avid lesion. Patients were categorized as responders (stable or no evidence of disease) or nonresponders (progression) based on the oncologist’s clinical assessment using all available clinical data at the last documented follow-up office visit, which occurred at a median 5.4 months (range 0-19) after PET1. Repeated analysis of variance (SigmaPlot 12.5, Systat Software, Inc.) determined differences in the measurements in each response category. Data are represented as mean±standard deviation. RESULTS: We identified 16 responders (age 66±12 years, 6F/10M) and 11 nonresponders (age 68±20 years, 3F/8M) with evaluable PET0 and PET1 scans. Only five patients in the nonresponder group had evaluable lesions on CT; all 16 responders had evaluable lesions on CT. The SUVmax (PET0 9.0±7.8, PET1 15.3±13.4, p=0.01) and SULpeak (PET0 5.8±5.0, PET1 9.5±7.3, p=0.003) increased significantly in the nonresponders (6.3±11 absolute/96±176% increase in SUVmax, 3.7±5.3 absolute/79±140% increase in SULpeak). Neither SUVmax (PET0 6.4±3.7, PET1 4.7±4.4, p=0.36) nor SULpeak (PET0 4.6±3.2, PET1 3.0±3.1, p=0.097) changed significantly in responders, though there was a trend toward a significant decrease in SULpeak (1.8±3.6 absolute/22±61% decrease in SUVmax; 1.6±1.9 absolute/33±43% decrease in SULpeak). The PET1 SUVmax and SULpeak were significantly lower in the responders compared to the nonresponders (p=0.001 for both). Though the axis diameters decreased more noticeably in the responder group, these changes were not statistically significant (Table 1). Conclusion: Markedly increased 18F-FDG uptake on interim response PET/CT scans in patients treated with pembrolizumab appears to identify nonresponding patients. Figure 1. 18F-FDG positron emission tomography/computed tomography (PET/CT) showing progression and response in patients with advanced melanoma treated with the immune checkpoint inhibitor pembrolizumab. Maximum intensity projection (MIP) images of an 88 year-old patient with metastatic melanoma to the liver before initiation of therapy (1a) and after 3 cycles of treatment (1b) show progressive disease on the scan. This patient ultimately progressed on therapy (nonresponder). MIP images of a 58 year-old patient with metastatic disease to an axillary lymph node before initiation of therapy (2a) and after 3 cycles (2b) showed complete response on the images. This patient was ultimately classified as having no evidence of disease (responder).