PET May Predict Long-Term Benefit With Anti-PD-1 Therapy in Melanoma

PD-1 targeting immune checkpoint inhibitors have revolutionized treatment of metastatic melanoma, resulting in long-term survival in many patients. However, there are few reliable predictors of long-term response to anti-PD-1 therapy in these patients. Currently, complete response (CR) by computed tomography (CT) imaging is considered the best indicator for long-term outcomes, and it appears that these patients can safely discontinue therapy with low risk of recurrence. However, many patients with partial response (PR) or even stable disease (SD) by CT may also have durable response to anti-PD-1 based therapy, but identifying which of these patients are at risk for progression and what is optimal duration of therapy remains a challenge. Molecular imaging with 18F fludeoxyglucose positron emission tomography (18F-FDG-PET) may be a more sensitive and accurate method for monitoring response to immunotherapy and predicting long-term outcomes, particularly in patients with residual disease on CT.

This was investigated in a retrospective analysis (N = 104) that compared the 18F-FDG-PET to CT imaging in measuring response to treatment at 1 year in patients with metastatic melanoma who received anti-PD-1 based therapy . Patients were included in the study if they received PD-1 inhibitor monotherapy (67%) or a PD-1 inhibitor in combination with ipilimumab (31%). Response to treatment was assessed by CT using response evaluation criteria in solid tumors (RECIST) criteria, coded as CR, PR, SD, or progressive disease (PD), and by PET using modified European Organization for Research and Treatment of Cancer (EORTC) criteria, coded as complete metabolic response (CMR), partial metabolic response (PMR), stable metabolic disease (SMD), or progressive metabolic disease (PMD).

At one year of treatment, 28% had CR, 66% had PR, and 6% had SD on CT, while 75% had CMR, 16% PMR, and 9% SMD or PMD. Importantly, 68% of patients who achieved PR by CT were identified as having a CMR on PET scan. Lung lesions (59%) and lymph nodes (33%) were the most common residual disease sites seen on CT that were not avid by PET. Of the 27 patients with CMR who had follow-up >24 months, all remained in CMR on PET at 2 years. On CT correlation, 13 patients were in CR, 13 in PR, and 1 had SD. Nine patients have longer follow-up, and at the 3-year PET scan, all remained in CMR while CT responses were either the same or better than at the 1-year CT.

At a median follow-up of 30.1 months, 98% of patients remained alive, and 87% were progression-free. There was no significant difference in progression-free survival (PFS) in patients who were in CR versus PR/SD at the 1-year analysis (not reached for both; HR 0.18, P = .06), although outcomes favored patients who achieved a CR. One year following imaging, rates of PFS were 100% in patients achieving a CR and 86% in patients achieving PR/SD. In contrast, patients who achieved a CMR by PET had a significantly improved PFS compared to those who did not achieve CMR (not reached vs 12.8 months; HR 0.06; P<.01). One-year post-imaging, the PFS rates were 100% in those with a CMR and 57% in those without. All patients with a CMR by PET had similar PFS, regardless of CT scan results (CR vs PR/SD). Of the 78 patients with CMR at 1-year, 78% had discontinued treatment with median follow-up post discontinuation of 14.5 months, and 96% of patients remained progression free.

The authors concluded that 18F-FDG-PET may have an important role in response evaluation to immune checkpoint inhibitors with the ability to predict long-term benefit and help guide discontinuation of therapy.

Ann Oncol. 2018 Aug 21 [Epub ahead of print].
PET-predicts-melanoma-PD1-response