Main Session
Sep 28
SS 22 - Some crHEME With Your Coffee? A Winning Blend of Radiation and Systemic Therapy in Hematologic Malignancies

207 - Patterns of Progression after BCMA CAR T-Cell Therapy in Relapsed/Refractory Multiple Myeloma: A Lesion-Level PET/CT Analysis

11:25am - 11:35am ET
Room 259

Presenter(s)

Alexandra Dreyfuss, MD, MS - University of Miami - Sylvester Cancer Center, Coral Gables, FL

A. Dreyfuss1, J. Yahalom2, D. Nemorovsky3, A. Derkach3, B. Fregonese3, G. Cederquist2, Z. R. Moore2, S. Mailankody3, S. Usmani4, B. S. Imber3, and H. Hashmi3; 1Department of Radiation Oncology, University of Miami Miller School of Medicine, Sylvester Comprehensive Cancer Center, Miami, FL, 2Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, 3Memorial Sloan Kettering Cancer Center, New York, NY, 4Hematology, Medical Oncology, Memorial Sloan Kettering Cancer Center, New York, NY

Purpose/Objective(s): BCMA CAR T-cell therapy can induce durable responses in relapsed refractory multiple myeloma (RRMM), yet most patients (pts) progress. Patterns of progression (POD) at the lesion and anatomic compartment level (extramedullary (EM), paramedullary (PM), and bone) remain poorly defined. We hypothesized that POD post-CAR T occurs in compartment-specific spatial patterns, and characterization may define high-risk sites for radiotherapy (RT) integration.

Materials/Methods: We analyzed 102 RRMM pts treated with ciltacabtagene autoleucel (2018-2024) who underwent serial FDG PET/CT. Four PET timepoints were analyzed: pre-apheresis (80 scans, 307 lesions), pre–lymphodepleting chemotherapy (LDC) (98, 336), day 30 post-CAR T (102, 134), and first POD (102, 178). Individual lesions were followed to assess metabolic response and POD using SUVs per Lugano criteria given no accepted standard for MM. Multivariable Cox regression (MVA) evaluated associations with overall (OS) and progression-free (PFS) survival.

Results: With a median follow-up of 19 months (95% CI 18-22), 24-month PFS and OS were 48% (37- 61) and 74% (64-85), respectively. Among pts with POD, relapse was most often both biochemical and radiologic (63%), followed by radiologic only (28%), and biochemical only (10%). Radiologic POD involved both pre-existing and new disease sites (42%), pre-existing sites only (36%), or new sites only (22%). Prior to CAR T, EM and PM disease (EMD, PMD) were present in 26% and 12% of pts, respectively. At POD, disease distribution shifted toward non-bone compartments, with EMD increasing to 47% and PMD to 25%, while bone-only disease declined from 62% to 28%. Pre–LDC PMD and EMD were independently associated with inferior PFS (PMD: HR 20.4, 95% CI 7.2-58.0) (EMD: 7.1, 2.9–17.1) and OS (PMD: 5.4, 1.4–21.0) (EMD: 3.9, 1.2–12.7). Lesion-level analysis demonstrated higher baseline SUV in EM/PM sites vs bone, with POD occurring in 41% of EM and 31% of PM lesions vs 11% of bone lesions. Early metabolic response predicted durability: only 2% of lesions achieving complete metabolic response at day +30 progressed, compared with 27% of partial responses and 40% of stable disease.

Conclusion: POD after CAR T preferentially occurs in EM/PM compartments and is associated with pre-infusion EM/PM burden. Early lesion-level metabolic response may identify sites at highest risk for POD. These findings suggest rational, lesion-level criteria to integrate bridging or consolidative RT to potentially improve CAR T outcomes in RRMM.

Abstract 207 - Table 1: MVA of PFS post-CAR T

Characteristic

HR

95% CI

p-value

Lines of Therapy

0.033

<5

—

—

5+

2.2

1.0, 4.6

Functionally High Risk Disease

(POD <24 mo post-first line)

2.7

1.4, 5.2

0.003

High Risk Cytogenetics

1.8

0.9, 3.6

0.120

High Marrow Burden (>50%)

1.5

0.6, 3.8

0.400

Pre-LDC PET

<0.001

No PET avid disease

—

—

Bone only

1.6

0.6, 4.4

PMD

20.4

7.2, 57.6

EMD

7.1

2.9, 17.1