Prognostic significance of PET/CT for CAR T cell therapy in relapsed/refractory multiple myeloma

Abstract

PET/CT plays an important role in staging of multiple myeloma (MM) and detecting extramedullary disease (EMD); however, its role in patients treated with commercially available CAR T cell therapies is unclear. We evaluated 61 patients treated with CAR T cell products. In 53 patients, PET/CT was available before CAR T infusion, and 43 had follow-up PET/CT on day 30. Findings from PET/CT were correlated to (CAR) T single-cell dynamics, fitness and T cell receptor diversity after infusion, and serological markers of tumor burden and inflammation. Patients with bone-independent EMD had inferior median progression-free survival (PFS: 3 vs. 15 months, p = 0.01). Univariate and multivariate analysis showed that EMD but not the number of lesions or metabolic tumor volume (MTV) were associated with inferior PFS. High MTV was connected to higher baseline sBCMA and Interleukin-6 levels, but not associated with hampered CAR T cell expansion or decreased fitness of the bystander T cell compartment. Follow-up PET/CTs identified patients with metabolic complete remissions, which were associated with better PFS. PET/CT identifies patients with high risk of relapse after CAR T cell therapy.

Figure 1

PET/CT‐derived classification of extramedullary disease stratifies for outcomes after BCMA‐directed CAR T cell therapy in RRMM. (A) Study workflow. (B) The number of patients with FL detected at different anatomical localizations in the cohort. Examples for the three different categories of FL are presented in (C)–(E). (C) Two patients with intramedullary FDG‐avid focal lesion (FL) in the humerus (left) and acetabulum (right). (D) Examples for two patients with bone‐associated extramedullary disease (EMD) with cortical bone erosion in the iliac bone (left) and rib (right). (E) Examples of two patients with bone‐independent EMD of the pleural/thoracic soft tissue (left) and of the orbita (right). (F) Bar plot showing the numerical distribution of different types of FL in the entire cohort. Patients were hierarchically stratified by lesion type: (1) no lesions, (2) bone‐restricted lesions, (3) bone‐associated EMD with or without bone‐restricted lesions, and (4) bone‐independent EMD with or without additional lesions, reflecting worsening outcomes. (G) Violin plots showing correlation between types of FL (restricted to bone, bone‐associated EMD, and bone‐independent EMD) and FL counts (left), metabolic tumor volume (MTV, middle), and MTV of largest lesion (right). Multigroup comparison performed with Jonckheere‐Terpstra test. Progression‐free survival (PFS) according to the type of (H) FL and (I) MTV, with an optimal cuttoff of 54.3 mL. p values were obtained using the log‐rank test.

Figure 2

IMPeTUs criteria and impact on progression‐free survival. Progression‐free survival (PFS) analysis by (A) diffuse bone marrow uptake grouped by the metabolic activity level in comparison to the liver (Deauville 2 + 3 vs. 4 + 5), (B) number of PET‐positive focal lesions (FLs), (C) number of osteolytic lesions from CT, (D) presence of extramedullary disease (EMD), and (E) combination of metabolic risk factors diffuse BM uptake and EMD using the Kaplan–Meier estimator. p values were obtained using the log‐rank test. Longitudinal analysis of plasma level of (F) soluble BCMA and (G) IL‐6 compared in the previously defined risk groups. p values were calculated using the Wilcoxon‐rank sum test.

Figure 3

Impact of PET‐derived remission on outcome and impact on T cell compartment. (A) Bar plot showing distribution of metabolic CR (metCR in blue) and PET/CT‐positivity (in red) 30 days after CAR T cell infusion based on the presence of bone‐restricted lesions, bone‐associated extramedullary disease (EMD), bone‐independent EMD, or PET‐negativity. (B) Alluvial plot showing n = 43 patients with available PET/CT after CAR T cell re‐infusion and their response assessment according to IMWG criteria. (C) Progression‐free survival (PFS) based on remission according to PET and serological CR. Double‐negative (metabolic CR and serological CR, in blue) had superior outcomes compared to positive patients (red). p value was obtained using the log‐rank test. (D) Box plots showing peak CAR T cell expansion according to MTV (top) as well as type of lesions (bottom). (E) Box plots of CD3+CAR+/totalCD3+, CD4+CAR+/CD3+CAR+, and CD8+CAR+/CD3+CAR+ after CAR T cell infusion between EMD groups. (F) Box plots of longitudinal PD1 expression in CD4+ (top) and CD8+ (bottom) T cells in patients with high (red) and low (blue) metabolic tumor volume (MTV). (G) Longitudinal assessment of Shannon diversity indices in CD4+ (top) and CD8+ (bottom) T cells between high‐ and low‐MTV patients. Multigroup comparison was performed using the Kruskal–Wallis test. Pairwise comparisons were performed using the Wilcoxon‐rank sum test.