To the Editor:

Chimeric antigen receptor T-cell (CAR-T) therapy targeting B-cell maturation antigen (BCMA) is now a standard of care for relapsed and refractory multiple myeloma (MM) with two approved drug products (i.e., idecabtagene vicleucel [ide-cel], ciltacabtagene autoleucel [cilta-cel]) [1,2,3,4,5]. Treatment with BCMA CAR-T in MM patients is characterized by rapid response kinetics, with a median time to first response and complete response (CR) of 1 and 3 months, respectively [1,2,3,4,5]. Patients who achieve a CR as the best overall response to BCMA CAR-T have significantly longer PFS [1,2,3,4,5], but these analyses are limited by immortal time bias and the prognostic significance of early categorical response attainment at specific landmark time points is unclear.

18F-fluorodeoxyglucose positron emission tomography-computer tomography (PET/CT) is increasingly being used in MM patients treated with BCMA CAR-T. The International Myeloma Working Group (IMWG) also recently incorporated PET/CT into the revised response criteria for minimal residual disease (MRD) negativity [6]. Several prospective studies have demonstrated that PET/CT negativity is associated with prolonged survival in newly diagnosed, transplant-eligible MM patients treated with triplet therapy +/− autologous stem cell transplantation [7,8,9]. However, prospective data validating PET/CT in relapsed and refractory MM patients treated with BCMA CAR-T is limited to a phase 2 clinical trial (n = 63) [10]. Zugasti and colleagues showed that PET/CT negativity at day 100 was associated with longer overall survival (OS) after treatment with ARI0002h, an academic BCMA CAR-T [10]. We sought to investigate further the prognostic impact of early serologic and PET/CT response attainment in MM patients treated with BCMA CAR-T.

We performed a retrospective study that included all MM patients consecutively treated with BCMA CAR-T at the Massachusetts General Hospital between 2016 and 2024. Serological responses were graded according to the IMWG response criteria [6]. Early treatment responses were defined as a ≥VGPR at 1 month and CR at 3 months based on the published response kinetics with ide-cel and cilta-cel [1,2,3,4,5]. PET/CT was available in a subset of patients with pre-treatment and re-staging scans. Re-staging scans were obtained at the discretion of the treating investigator, and the same group of expert nuclear medicine radiologists interpreted all the scans. Two definitions for PET/CT negativity were evaluated based on the evolving imaging response criteria in MM patients: (1) Deauville score (DS) < 3 (current criteria); and (2) DS < 4 (proposed criteria) [6, 9]. Additional details regarding the statistical analysis are included in the Supplemental Methods. This study was approved by the Institutional Review Board of the Massachusetts General Hospital (15-190). All methods were performed in accordance with the Declaration of Helsinki, and all patients provided informed consent per institutional guidelines.

A total of 158 patients with relapsed and refractory MM were included. Baseline patient characteristics before treatment with BCMA CAR-T are shown in Supplementary Table 1. One hundred thirteen patients (72%) had penta-exposed disease, 73 (46%) had revised high-risk disease, and 61 (39%) had EMD. Patients received the following BCMA CAR-T products: ide-cel (n = 77; 49%), cilta-cel (n = 46; 29%), and investigational agent (n = 35; 22%). The median follow-up was 24.8 months (95% CI 22.5–30.3).

Serological treatment response assessment was available at 1 and 3 months following BCMA CAR-T in 155 and 128 patients, respectively; the remaining patients either progressed and/or died before the landmark time point. ≥VGPR attainment occurred in 101 patients (65%) at 1 month, and CR attainment occurred in 52 patients (41%) at 3 months. On multivariable logistic regression analysis, a baseline LDH > 210 U/L (49% vs 75%; OR 0.31, p = 0.002) and EMD (29% vs 48%; OR 0.40, p = 0.02) were independently associated with lower odds of achieving a ≥VGPR and CR at 1 and 3 months, respectively (Supplementary Table 2).

Early serological response attainment and depth significantly correlated with PFS and OS following BCMA CAR-T (Fig. 1; Supplementary Fig. 1). Patients with a ≥VGPR at 1 month (PFS: 24.9 vs 3.6 months; OS: not reached [NR] vs 15.5 months) and CR at 3 months (PFS: 35.0 vs 11.9 months; OS: NR vs 38.1 months) had significantly longer PFS and OS. After adjusting for revised high-risk disease, LDH, ferritin, EMD, and BCMA CAR-T drug product, early response attainment at 1 month (PFS: HR 0.22, p < 0.001; OS: HR 0.34, p < 0.001) and 3 months (PFS: HR 0.35, p < 0.001; OS: HR 0.25, p = 0.001) both remained independently associated with significantly longer survival on multivariable modeling (Supplementary Table 3). Early serological attainment was also associated with prolonged survival in patients with and without EMD (Supplementary Fig. 2).

Fig. 1: Impact of early serological response attainment on survival following BCMA CAR-T.
figure 1

Kaplan–Meier curves for progression-free survival (PFS) and overall survival (OS) according to categorical serological response at 1-month landmark (A, B) and 3-month landmark (C, D) following BCMA CAR-T. VGPR very good partial response, CR complete response.

Full size image

Baseline PET/CT was obtained in 107 patients (68%) before BCMA CAR-T. Baseline PET/CT was positive in 97 out of 107 patients (91%), and the following radiographic findings were identified: diffuse heterogeneous bone involvement (n = 21, 20%), focal bone lesions (n = 80, 80%), EMD (n = 49, 46%), and PMD (n = 33, 31%). Fifty-four patients (50%) had >3 focal bone lesions, and EMD involved the soft tissue and viscera in 23 (21%) and 27 (25%) patients, respectively. The presence of EMD with a SUVmax > 5.9 on baseline PET/CT was associated with a significantly shorter PFS and OS on multivariable modeling (Fig. 2; Supplementary Table 4), expanding upon similar data in transplant-eligible MM patients [11].

Fig. 2: Impact of baseline PET/CT characteristics for extramedullary disease on survival following BCMA CAR-T.
figure 2

Kaplan–Meier curves for progression-free survival (PFS) and overall survival (OS) according to extramedullary status (EMD) as assessed on the baseline PET/CT (A, B). Patients with EMD were stratified into two groups based on the SUVmax measured by PET/CT: low (SUVmax ≤5.9) and high (SUVmax >5.9). Patients without EMD had an estimated 18-month PFS of 54%, as compared to 41% and 12% in patients with low and high SUVmax EMD, respectively. In addition, patients without EMD had an estimated 18-month OS of 81%, as compared to 71% and 39% in patients with low and high SUVmax EMD, respectively. SUVmax, maximum standardized uptake value.

Full size image

Among the patients with a positive baseline PET/CT, 54 and 23 had a re-staging scan at 1 and 3 months, respectively. Baseline EMD was present in 50% and 78% of the patients with re-staging scans at 1 and 3 months, respectively. PET/CT negativity by the current (DS < 3) and proposed (DS < 4) response criteria occurred in 15 (28%) and 18 (33%) patients at 1 month, respectively, and in 10 (43%) and 19 (83%) patients at 3 months, respectively. In both the 1-month and 3-month landmark analyses, PET/CT negativity was not associated with PFS by the current or proposed response criteria (Supplementary Fig. 3). Fourteen patients had a re-staging PET/CT at 12 months, of whom 12 had attained PET/CT negativity (by both imaging response criteria). At 12 months, patients with PET/CT negativity had an estimated 36-month PFS and OS of 70% and 100%, respectively.

We also evaluated the prognostic impact of achieving an early combined serological and PET/CT response. Early combined response attainment at 1 month was not associated with PFS (p > 0.05 for all comparisons). However, there was a trend for longer PFS in patients with a combined CR and PET/CT negativity at 3 months (18-month PFS: 100% vs 52%; p = 0.18; Supplementary Fig. 4).

Overall, we show that early serological response attainment—specifically, ≥VGPR at 1 months or CR at 3 months—is significantly associated with both prolonged PFS and OS following BCMA CAR-T, including in MM patients with EMD. Wong and colleagues reported similar findings in a smaller study (n = 54), which showed that early serum free light normalization on days 15 or 30 following BCMA CAR-T correlated with longer PFS [12]. Collectively, these data establish early serological response attainment as an important treatment goal and a prognostic marker for MM patients treated with BCMA CAR-T. We also show that MM patients who are unlikely to derive long-term benefit from BCMA CAR-T can be identified based on early serological response kinetics, thereby providing the framework for a response-adapted risk stratification following BCMA CAR-T. This observation provides the rationale for early treatment intensification in MM patients with suboptimal early responses to BCMA CAR-T, and should herald clinical trials evaluating this management strategy. Phase 2 clinical trials investigating consolidation therapy after ide-cel with elranatamab (NCT06138275) and talquetamab (NCT06066346) are now actively ongoing. Maintenance therapy after ide-cel is also under active investigation with trials involving lenalidomide (NCT05032820), iberdomide (NCT06179888), and mezigdomide (NCT06048250). Similarly, the presence of EMD with a SUVmax > 5.9 on baseline PET/CT was associated with a significantly shorter PFS and OS on multivariable analysis, providing the rationale for exploring debulking options prior to CAR-T.

An unexpected finding was that early PET/CT negativity was not prognostic for survival after BCMA CAR-T. There are several potential explanations for this finding. Firstly, the lack of standardization regarding the timing and requirements for re-staging scans likely introduced selection bias. Secondly, our PET/CT cohort was highly enriched for patients with EMD. It is possible that delayed or impaired effectiveness of BCMA CAR-T against EMD contributed to the lack of prognostic value for early PET/CT negativity. Consistent with this hypothesis, a recent retrospective study by Born and colleagues showed that PET/CT negativity at 1 month correlated with longer PFS following BCMA CAR-T [13]. However, none of the patients with EMD or PMD attained PET/CT negativity at 1 month, suggesting that patients with a low tumor burden drove the prognostic impact they identified [13]. Re-staging scans at later time points may therefore offer improved prognostic value. Indeed, we observed a prolonged PFS in the subset of patients with a combined serological CR and PET/CT negativity at 3 months (18-month PFS: 100%) and 12 months (36-month PFS: 70%). A similar finding was reported for MRD-negative patients after ide-cel in KarMMa, wherein the survival benefit of a concurrent negative PET/CT became more pronounced at later landmark time points [14]. In the IMAJEM trial, PET/CT negativity also became significantly prognostic later in the treatment course at the pre-maintenance stage [8]. Additional studies are nevertheless needed to optimize the use of PET/CT in MM patients treated with BCMA CAR-T.

A notable limitation of this study is the lack of data on MRD negativity in patients who achieve early deep serological responses. Two previous studies suggested that early attainment of MRD negativity at 1 and 3 months correlated with longer survival following BCMA CAR-T, though persistent EMD was often associated with a transient MRD-negative status without ever attaining a serological CR [14, 15]. Future studies that integrate serological, MRD, and PET/CT responses assessments together may further refine response-adapted risk stratification following BCMA CAR-T.

In summary, our findings show that MM patients who achieve a serological ≥VGPR at 1 month or CR at 3 months after BCMA CAR-T have prolonged PFS and OS, whereas early PET/CT negativity at 1 month was not prognostic. This study provides the framework for a response-adapted risk stratification in MM patients treated with BCMA CAR-T.