Anti-BCMA bispecific antibodies (BiAbs) improved outcomes of patients (pts) with relapsed/refractory multiple myeloma (RRMM). In this setting, predicting response and toxicities such as cytokine release syndrome (CRS) or immune effector cell-associated neurotoxicity syndrome (ICANS) is still a challenge. Extracellular vesicles (EVs) are membrane-bound particles with emerging role in intercellular communication and modulation of biological functions. EVs have also been explored as markers of minimal residual disease and predictor of ICANS following CAR T cell therapy.

We prospectively assessed BCMA+ EVs released from plasma cell and CD3+, CD4+, and CD8+ EVs released from lymphocytes in peripheral blood (PB) of RRMM treated with anti-BCMA BiAbs. Pts received either teclistamab (tec) or elranatamab (elra) as per clinical practice. PB samples were collected at different timepoints: baseline (T0), +24h after 1st step up dose (T1), +24h after 2nd step up dose (T2), +24h after 1stfull-dose (T3), then at D1C4 (T4), D1C7 (T5) and D1C13 (T6).

We report preliminary results of 13 RRMM pts treated at our Institution between October 1st, 2024, and June 30th, 2025. At baseline median age was 65 years (range 50 – 77), 8 pts (61%) were male, 1 pts (8%) had ISS 3, 4 pts (31%) had R-ISS 3, 9 pts (69%) had high-risk FISH abnormalities, and 4 pts (31%) had extramedullary disease. Median number of prior lines of therapy was 3 (range 3 – 5), 13 pts (100%) and 11 pts (85%) were triple-class exposed and refractory, respectively. 7 pts (54%) received tec and 6 pts (46%) received elra. Median follow up was 6 months (range 1 – 9). Median number of BiAbs cycles was 4 (range 0 – 10). At 1-month, overall response rates (ORR) and ≥very good partial response (VGPR) were 77% and 38%, respectively. After a median of 2.4 months (range 0.5 – 7.5), best ORR and ≥VGPR were 77% and 54%, respectively. Median PFS and OS were both not reached. CRS occurred in 7 pts (54%), all grade 1, tocilizumab was given in 3/7 pts. ICANS occurred in 2 pts (15%), 1 grade 2 and 1 grade 3, both resolved with dexamethasone. Infections occurred in 8 pts (61%), 6 events (46%) were grade 3/4, 12 pts (92%) regularly supplemented IVIG. At data cut-off, 4 pts discontinued BiAbs (3 disease progression, 1 toxicity).

At T0, median number (N) of total EVs/μL was 10857 (range 1820 – 30816.8), median N of BCMA+ EVs/μL was 473.2 (range 2.8 – 2088.8), median N of CD4+ EVs/μL was 383.6 (range 0 – 10123.4), median N of CD8+ EVs/μL was 217 (range 0 – 1064.2). At T1, median N of total EVs/μL was 13119.4 (range 153.72 – 49400.4), median N of BCMA+ EVs/μL was 115.2 (range 25.2 – 1495.2), median N of CD4+ EVs/μL was 333.2 (range 0 – 21585.2), median N of CD8+ EVs/μL was 67.2 (range 2.8 – 5870.2). At T2, median N of total EVs/μL was 16144.8 (range 2067.8 – 57673), median N of BCMA+ EVs/μL was 210.7 (range 26.6 – 1145.2), median N of CD4+ EVs/μL was 727.3 (range 37.8 – 22786.4), median N of CD8+ EVs/μL was 200.2 (range 0 – 1586.2). At T3, median N of total EVs/μL was 5644.8 (range 1915.2 – 44252.6), median N of BCMA+ EVs/μL was 93.8 (range 0 – 3694.6), median N of CD4+ EVs/μL was 576.8 (range 0 – 47069.4), median N of CD8+ EVs/μL was 148.4 (range 0 – 2125.2).

Pts experiencing any grade CRS had significantly lower BCMA+ EVs/μL at T2 compared to those without CRS (72.8 vs 429.8, respectively; Mann-Whitney test p=0.048). Lower BCMA+ EVs/μL at T2 also predicted ORR at C1 and best ORR compared to those pts not responding to BiAbs (189 vs 817.4, respectively; point biserial correlation p=0.02). Similarly, lower CD4+ EVs/μL at T2 and T3 predicted achievement of ≥VGPR at C1 compared to pts not obtaining at least VGPR (T2: 144.9 vs 1978.2, respectively; Mann-Whitney test p=0.01) (T3: 147 vs 2248.4, respectively; Mann-Whitney test p=0.02). A trend for higher risk of ICANS was observed among pts with lower %CD4+ EVs at T1 compared to pts without ICANS (0.04 vs 14.25, respectively; Mann-Whitney test p=0.07).

Preliminary results from our prospective cohort showed that EVs appear as novel promising markers of response and toxicity in RRMM treated with anti-BCMA BiAbs. BMCA+ EVs assessed at T2 predicted CRS and disease response. Those pts obtaining rapid and deep response had lower CD4+ EVs both at T2 and T3. A trend for higher risk of ICANS was observed among pts with lower %CD4+ EVs at T1. Longer follow up and greater population might further confirm these initial findings.

This content is only available as a PDF.
Sign in via your Institution