142P Exploring 18F-FDG PET as a predictior of immune-related adverse events (irAEs) in early-stage triple-negative breast cancer (eTNBC)

Immune checkpoint inhibitors (ICI) have improved outcomes in patients with eTNBC. However, irAEs are a major concern in clinical practice. 18F-FDG PET/CT scan is routinely performed for initial staging before neoadjuvant treatment with immunotherapy and chemotherapy (NACT-IO). This study aims to evaluate whether FDG-PET features could predict the occurrence of irAE from NACT-IO in eTNBC. We conducted a single-center retrospective study on eTNBC patients treated with NACT and pembrolizumab, between April 2022 and April 2023. Adverse events were recorded and graded according to NCI-CTC AE version 5. Extracted FDG-PET parameters were primary tumor and metastatic loco-regional lymph node-related features (SUVpeak, SUVmax, MVT, TLG, necrosis [only primary tumor]), and healthy organs features (SUVmean of liver, spleen, bone marrow, blood pool, digestive organs, endocrine glands and breasts). Data cut-off date was November 2023. Primary endpoint was the incidence of grade 3-4 irAEs. Logistic regression analysis was performed using R software version 4.1.1. Of 86 identified patients, 50 had baseline FDG-PET raw data and toxicity data available at our institution. 64% (32/50) of patients experienced any-grade irAEs, and 26% (13/50) severe irAEs. Primary lesion SUVpeak ranged between 1.9 and 27.8 (median: 9.7). 44% (22/50) of primary lesions had central necrosis. 52% (26/50) patients had FDG-avid locoregional lymph nodes. Among the FDG-PET parameters explored, the univariable analysis found a unique significant association between the primary lesion SUVpeak/SUVmax and an increased risk of severe irAE (OR 1.13 IC95 [1.02-1.27]; per unit of SUV). Neither lymph node related features nor healthy organ basal uptake value were associated with severe toxicity. Our findings suggest that baseline FDG-PET SUVpeak/SUVmax of the primary lesion are significantly associated with an increased risk of severe ICIs irAEs. If these findings are independently validated, baseline FDG-PETmay contribute to personalized risk assessment of irAE in eTNBC, ultimately improving clinical decision-making.