P-105: PET/CT has prognostic significance for CAR T cell therapy in relapsed/refractory multiple myeloma

Introduction: Positron emission tomography–computed tomography (PET/CT) is pivotal in staging of multiple myeloma (MM), monitoring treatment response and detecting extramedullary disease (EMD). However, its prognostic role for heavily pretreated patients receiving CAR-T cell therapies is unclear. We retrospectively assessed PET/CT images of 53 patients who received commercially available CAR-T cell therapies at our center.

Methods: Idecabtagene vicleucel (Ide-cel) was administered in 33 patients and Ciltacabtagene autoleucel (Cilta-cel) in 20 patients. PET/CTs were performed prior to lymphodepletion and retrospectively analyzed in joint sessions by 2 experts applying the four-eyes principle. Total metabolic tumor volume (MTV) was determined using a standardized uptake value (SUV) threshold adjusted to the degree of metabolic activity in order to precisely encompass the lesions. Response to CAR-T was assessed monthly according to IMWG guidelines. In 43 cases, follow-up (FU) PET/CT was available on day 30 (D30) after CAR-T re-infusion.

Results: Fifty-nine percent of the patients prior to CAR-infusion showed focal FDG-avid lesions (FL). These patients were classified according to EMD status, with 23% of cases showing bone restricted osteolytic lesions, 11% had bone associated EMD defined as FL contiguous to bone but infiltrating stromal or soft tissue and 25% of cases with bone independent EMD defined as isolated extra-osseous FL.

With a median FU of 11 [95%CI, 8-12] months, univariate COX-PH regression revealed that progression-free survival (PFS) was significantly affected by bone independent EMD (PFS: HR 3.8 [1.4-10.6], p=0.007), but not by bone restricted lesions (PFS: HR 1.3 [0.4-3.9], p=0.6), or bone associated EMD (PFS: HR 0.7 [0.1-2.8], p=0.7). Overall MTV ranged from 0 ml in PET-negative patients up to a value of 5313 ml. Optimal cut-point for was defined as 54,33 ml by ROC analysis (AUC=0.71). Patients with high MTV above this threshold had significantly worse PFS (log-rank test, p< 0.001). In multivariate analysis including the type of CAR-T as covariate, the negative prognostic value of bone independent EMD remained significant for PFS (p=0.01, OS: p=0.08).

The proportion of initial responders defined as PR or better on D30 was equal amongst EMD groups (p=0.9). However, cases with bone independent EMD had a higher risk of relapse (OR=6, p=0.01).

PET/CT on D30 revealed that 37.2% of patients were in complete metabolic remission (metCR), of which 69% were classified as CR, 25% as VGPR/PR and 6% as SD/PD by IMWG criteria. MetCR was associated with a trend towards improved PFS (log-rank test, p=0.06) and OS (p=0.12).

Conclusions: PET/CT is of prognostic significance for risk stratification and response evaluation in MM patients treated with CAR-T cell therapy. PET/CT derived classification of EMD status shows that patients with bone independent EMD have inferior outcomes to CAR-T cell therapy due to early relapses, which has important implications for future trial strategies.