34P - Heterogeneous response to Immune Checkpoint Inhibitors in metastatic melanoma patients - assessment of lesion-level response with 18F-FDG PET/CT

Background

Existing therapy response assessments are based largely on summing changes in target lesion diameters assessed on CT or MRI (e.g., RECIST) and do not consider lesion response heterogeneity. Evaluating lesion-level response of all lesions enables identification of novel patterns, such as heterogeneous response (HeR). In our study, we use a lesion-level response model, based on ¹⁸F-FDG PET/CT analysis to correlate lesion response patterns and treatment outcomes in metastatic melanoma (MM) patients treated with immune checkpoint inhibitors (ICI).

Methods

We retrospectively assessed 27 MM patients treated with anti-PD-1 therapy in two academic centers. Lesions were manually segmented on baseline (PET1) and first follow-up (PET2, month 3-4) ¹⁸F-FDG PET/CT. Lesions were categorized based on relative change of SUV_total from PET1 to PET2: complete response (iCR, disappeared on PET2), partial response (iPR, ΔSUV_total<-30%), stable disease (iSD, |ΔSUV_total | ≤ 30%), progressive disease (iPD, ΔSUV_total>+30%), new disease (iND, new on PET2). HeR was defined as patients having at least one iCR or iPR and iPD or iND lesion. Survival analysis was done using Cox regression analysis. Patients were dichotomized into two groups based on their clinical status at the end of treatment (EOT): progressive disease (PD) and non-progressive disease patients (non-PD). Two-proportion z-test was used to compare proportion of patients with HeR in both groups.

Results

426 lesions were segmented and categorized as iCR (56%), iPR (10%), iSD (3%), iPD (8%), iND (23%). Patients with one or more iPD lesion (HR = 1.8, p= 0.003), two or more iND or iPD lesions (HR = 1.9, p=0.003), or identified with HeR (HR = 1.42, p=0.04) on PET2 had significantly shorter PFS. 37% (10/27) patients were identified to have HeR: 2/13 non-PD and 8/14 PD patients. Patients with HeR at PET2 have a significantly higher probability of progression at EOT (p=0.02).

Conclusions

Lesion-level analysis of ¹⁸F-FDG PET/CT response can recognize non-responding lesions. We found that patients with HeR had significantly shorter PFS. We anticipate a future interventional study where early identification of non-responding lesions may be amendable to local therapies.

Clinical trial identification

At University of Wisconsin Carbone Cancer Center, the study was approved by Institutional Review Board (UW14084). At Oncology Institute Ljubljana, the study was approved by the Ethics comitee of OIL and Comitee for assessment of Clinical protocols (ERDIKE-005/2020, ERID-KSOPKR-002/2020).

Legal entity responsible for the study

The authors.

Funding

University of Wisconsin Carbone Cancer Center (UWCCC), National Cancer Institute of the National Institutes of Health, The Slovenian Research Agency (ARRS).

Disclosure

R. Jeraj: Financial Interests, Personal and Institutional, Funding: AIQ Solutions. All other authors have declared no conflicts of interest.