FDG-PET associations with pathological response and survival with neoadjuvant immunotherapy for melanoma

Abstract

Background: Neoadjuvant immunotherapy has become the new standard of care for stage III melanoma. This study sought to describe the metabolic changes seen with fludeoxyglucose-18-positron emission tomography (FDG-PET) following neoadjuvant immunotherapy in patients with melanoma and explore associations with pathological response and recurrence-free survival (RFS).

Methods: Data from patients with macroscopic stage III nodal melanoma treated with neoadjuvant checkpoint inhibitor therapy were pooled from five melanoma centers. All patients underwent baseline and preoperative FDG-PET and CT assessments, and all had surgery. Pathological response was determined using the International Neoadjuvant Melanoma Consortium criteria, radiological response using Response Evaluation Criteria in Solid Tumors (RECIST) criteria, and FDG-PET response using European Organization for Research and Treatment of Cancer (EORTC) criteria. The primary endpoint was to explore the associations of metabolic and radiological responses with pathological response; secondary endpoints were RFS outcomes stratified by each response category.

Results: 115 patients were included, 69% male, median age 59 years (27-92), 43% BRAF mutant, and median follow-up was 22.2 months (95% CI 13.7 to 26.4). 40 patients received anti-PD-1 monotherapy, 20 patients received pembrolizumab combined with lenvatinib, and 55 patients received ipilimumab and nivolumab. The major pathological response (MPR) rate was 62%, and the pathological complete response rate was 51%. RECIST response underestimated pathological response; patients achieving RECIST stable disease (38%) had a 50% chance of achieving MPR. The FDG-PET metabolic response rate was 73%, with most achieving an MPR (80%), especially in patients with a complete metabolic response (CMR, 96% MPR). A small proportion of patients (10%) had stable metabolic disease on FDG-PET, and all these patients were non-MPR. Patients with progressive metabolic disease were also in the majority non-MPR (79%). Patients with MPR, complete response/partial response on CT, and CMR/partial metabolic response on FDG-PET had a favorable 24-month RFS (95.6%, 97.3%, and 93.7%, respectively), with FDG-PET able to identify a greater proportion of patients with favorable progression-free survival (PFS) than pathology or CT (73%, 62%, and 43%, respectively).

Conclusion: Neoadjuvant immunotherapy has high FDG-PET response rates in melanoma. FDG-PET response associates with pathological response and confers impressive RFS, suggesting this could be an important clinical tool.

Keywords: Immunotherapy; Melanoma; Neoadjuvant.

Figure 1. Response rates and recurrence-free survival (RFS) by (A) pathology, (B) CT, and (C) fludeoxyglucose-18-positron emission tomography (FDG-PET). CMR, complete metabolic response; CR, complete response; npCR, near-complete pathological response; pCR, pathological complete response; PD, progressive disease; PMD, progressive metabolic disease; PMR, partial metabolic response; pNR, pathological non-response; pPR, pathological partial response; PR, partial response; SD, stable disease; SMD, stable metabolic disease.

Figure 2. Categorical response rates and recurrence-free survival (RFS) by (A) pathology, (B) CT, and (C) fludeoxyglucose-18-positron emission tomography (FDG-PET). MPR, major pathological response (≤10% tumor cells); MR, metabolic response (complete metabolic response (CMR)+partial metabolic response (PMR)); RR, radiological response (complete response (CR)+partial response (PR)).

Figure 3. Waterfall plot by SUVchange associated with pathological response and recurrence. Cases A–D details are shown in figure 4. npCR, near-complete pathological response; pCR, pathological complete response; pNR, pathological non-response; pPR, pathological partial response; SUV, standardized uptake value.

Figure 4. Representative discordant cases with images of fludeoxyglucose-18-positron emission tomography (FDG-PET) and pathology. (A) Patient (Pt) 37 treated with two cycles of pembrolizumab with progressive metabolic disease (PMD) (a new avid node while the initial node was also more avid) developed redness, warmth, and tenderness over the inguinal fossa, suggestive of an inflammatory process. This patient achieved a pathological complete response (pCR) (necrosis 40%, melanosis 10%, and 50% fibrosis) and did not have any adjuvant therapy. (B) Pt 6 treated with pembrolizumab plus lenvatinib had PMD with pCR (20% necrosis and 80% fibrosis/melanosis) and no adjuvant therapy, recurring with brain metastasis 15.4 months later. (C) Pt 64 treated with nivolumab and ipilimumab had a complete metabolic response (CMR) with pathological non-response (pNR) (subcapsular intact neoplastic cells) and received adjuvant immunotherapy afterwards. Immunohistological chemistry (IHC) images: HMB45 Ventana antibody. (D) Pt 78 treated with nivolumab and ipilimumab had both a CMR and pCR without adjuvant therapy and recurred locally 5 months later. SUVmax, maximum voxel value of standardized uptake value.