A 5-Year Follow-up Clinical Study of the B-cell Maturation Antigen Chimeric Antigen Receptor T-cell Therapy HDS269B in Patients with Relapsed or Refractory Multiple Myeloma

B-cell maturation antigen (BCMA)-directed chimeric antigen receptor (CAR) T-cell therapy was found to have remarkable activity and manageable safety in early-phase clinical trials involving patients with relapsed/refractory multiple myeloma (RRMM; refs. 1–5). The BCMA-targeted CAR T-cell products ciltacabtagene autoleucel (2) and idecabtagene vicleucel (3) have been approved by the FDA for the treatment of patients with RRMM. Nevertheless, reports on the long-term efficacy of anti–BCMA CAR T-cell therapies are limited; thus, further investigations into their continued safety and effectiveness are warranted. Additionally, patients with poor physical status, indicated by an Eastern Cooperative Oncology Group (ECOG) score of 3 to 4, who are often excluded from clinical trials, urgently require medical assistance because of their increased propensity for relapse and refractory disease. Our preliminary research (registered at ClinicalTrials.gov under NCT03093168) demonstrated the 77% overall response rate (ORR) with 47% complete response (CR) achieved for the antitumor activity of anti–BCMA CAR T-cell (HDS269B) therapy with favorable safety profiles in patients with RRMM, even in those with poor physical performance status (6). Here, we performed an updated follow-up of the cohort of 49 patients, with a median follow-up of 59.0 months (range, 1.0–81.0 months), and investigated the factors associated with long-term outcomes.

This study was an open-label, multicenter, single-arm, nonrandomized, phase I/II trial of anti–BCMA CAR T cells in patients with RRMM. The criteria for participant inclusion and exclusion were applied as described previously (6). Eligible adults received ≥3 prior lines of therapy, including proteasome inhibitors and immunomodulatory agents, and had adequate organ function. All patients provided written informed consent to participate in the study, which was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines and approved by the institutional review board.

Prior to the HDS269B infusion, all patients received lymphodepletion therapy for 2 to 4 days (cyclophosphamide 300 mg/m²/day, ×3 days + fludarabine 30 mg/m²/day, ×3 days). All 49 patients received an infusion of HDS269B at a dose of 9 × 10⁶ CAR-positive viable T cells per kg (±15% was allowed) weight, produced by HRAIN Biotechnology (Shanghai, China). Details of this process are presented in Supplementary Fig. S1. The structure of HDS269B CAR has been reported previously (7). This murine-derived anti–BCMA CAR utilizes the 4-1BB costimulatory signaling domain and the CD3ζ cytoplasmic region. The retroviral vectors encoding HDS269B CAR were constructed based on a modified Moloney Murine Leukemia Virus–based vector. After T-cell modification with the CAR, the CAR T cells were stimulated and expanded over a 10-day period using anti-CD3 and anti-CD28 antibodies (Thermo Fisher Scientific Cat# 11161D, RRID: AB_2916088; ref. 8).

According to the International Myeloma Working Group criteria (9), treatment efficacy was assessed using the ORR, which is defined as the proportion of patients achieving a partial response (PR) or better. Minimal residual disease (MRD) was assessed in bone marrow samples posttreatment, using a cutoff of 10⁻⁴ to define MRD-positive status (6). Cytokine release syndrome, neurologic events, and other adverse events (AE) were also evaluated as previously described (6). The pharmacokinetic parameters of HDS269B cells were determined by quantifying CAR transgene copies per microgram of DNA in plasma at various timepoints using qPCR). We defined a threshold of 20 copies/μg DNA as the criterion for CAR T-cell presence.

This analysis included 49 patients, focusing on long-term outcomes. Baseline characteristics were summarized using descriptive statistics. ORR, progression-free survival (PFS), and overall survival (OS) were calculated, and the ORR comparison between the two ECOG groups was conducted using the Fisher exact test. The median follow-up time was determined using the reverse Kaplan–Meier method (10). Survival probabilities were illustrated using Kaplan–Meier curves and analyzed using the log-rank test. To assess the differences in survival rates at predetermined timepoints (1, 3, and 5 years) between the two ECOG groups, we used the Fixpoint test, utilizing the R package ComparisonSurv. To identify independent predictors of PFS and OS, we incorporated variables such as age, sex, ECOG performance status, types of myeloma, presence of extramedullary disease (EMD), prior transplantation, number of prior lines of therapy, and CAR T-cell persistence into the Cox regression analysis. Analyses were performed using SPSS version 22 (SPSS, RRID: SCR_002865), GraphPad Prism 9 (GraphPad Prism, RRID: SCR_002798), and R software version 4.1.1 (R Project for Statistical Computing, RRID: SCR_001905). P values < 0.05 were considered statistically significant.

The clinical trial patient data used in this study have been strictly anonymized. Due to the privacy of the patients involved and the sensitivity of the data, the data generated in this study are not publicly available but can be made available upon reasonable request from the corresponding author. We are committed to promoting the sharing and dissemination of scientific knowledge while ensuring data security and protecting individual privacy.

In this study, 49 patients with RRMM received lymphodepletion followed by the administration of 9 × 10⁶ CAR T cells/kg. Of these patients, 40.82% exhibited a poor performance status (ECOG 3–4) at the time of enrollment. At baseline, 42.86% of patients had a high-risk cytogenetic profile, and 63.27% had received ≥4 prior lines of therapy. At the time of infusion, 79.59% showed disease progression. Among those with poor performance status, 30% had EMD, 45% had high-risk cytogenetic aberrations, 70% received ≥4 prior lines of therapy, and 80% showed progression after recent therapy. The representativeness of the study participants and baseline characteristics are detailed in Supplementary Tables S1 and S2.

After a median follow-up of 59.0 months (range, 1.0–81.0 months), 77.55% (38 patients) of the 49 patients included in the trial responded to HDS269B. Overall, 24.49% (12 patients) achieved a stringent complete remission (sCR), 24.49% (12 patients) achieved a CR, 16.33% (8 patients) showed a very good partial response (VGPR), and 12.24% (6 patients) achieved a PR (Table 1). The ORR was similar across patients with different ECOG scores (79.31% for ECOG scores 0–2 vs. 75% for ECOG scores 3–4, P = 0.72).

The median PFS and OS for all patients were 9.5 months [95% confidence interval (CI), 5.01–13.99] and 20.0 months (95% CI, 11.26–28.74), respectively (Fig. 1A and B). The 5-year PFS and OS were 21.3% (95% CI, 12.3%–36.7%) and 34.1% (95% CI, 22.7%–51.3%), respectively (Table 1). The median PFS for patients with ECOG scores 0 to 2 was 11.0 months (95% CI, 3.48–18.53), whereas for those with ECOG scores 3 to 4, the median PFS was 4.0 months (95% CI, 0.00–11.67, P = 0.18; Fig. 1C). The median OS for patients with ECOG scores 0 to 2 was 41.8 months (95% CI, 2.22–81.39), whereas for patients with ECOG scores 3 to 4, the median OS was 10.5 months (95% CI, 6.12–14.88, P = 0.015; Fig. 1D). Patients who were exposed to ≥4 prior lines of therapy exhibited a shorter median PFS and OS than those exposed to <4 prior lines (PFS: P = 0.012; OS: P = 0.0049; Fig. 2A and B). EMD was particularly aggressive (11), with 11 of 49 patients presenting EMD at enrollment. Although the ORRs were similar for patients with and without EMD (64% and 82%, respectively), patients with EMD had a median PFS and OS of 3.0 months (95% CI, 1.65–4.36) and 5.0 months (95% CI, 0.00–16.33) compared with 10.5 months (95% CI, 6.87–14.13) and 24.0 months (95% CI, 13.43–34.57) for patients without EMD (PFS: P = 0.06; OS: P = 0.03, Fig. 2C and D), respectively.

To investigate the effect of depth of response on prognosis, we observed that the patients with CR/sCR at 6 months landmark analysis showed significantly better PFS and OS than patients without CR (PFS: P = 0.04; OS: P = 0.02), which was consistent with the trend at the 3 and 12 months landmark analyses, although no statistical difference was found (Supplementary Fig. S2A–S2F).

MRD-negative status has been demonstrated to be prognostic for prolonged PFS and OS in drug-based myeloma therapy (12). Among all the enrolled patients, 22 had available MRD data at day +28 after the CAR T-cell infusion, as these samples met calibration standards, passed stringent quality checks, and had a sufficient cell count for assessment, rendering them suitable for evaluation. Of these, 14 achieved MRD-negative status (10⁻⁴), including 1 with sCR, 2 with CR, 3 with VGPR, and 8 with PR. To delve deeper into the relationship between early MRD status and long-term outcomes, we compared MRD status at day 28 after CAR T-cell infusion with PFS and OS. The results showed that patients who achieved MRD-negative status had significantly prolonged PFS and OS compared with MRD-positive patients (Fig. 3A and B). These findings highlight the crucial role of early MRD status in predicting disease progression and survival outcomes. Similar separation of PFS and OS curves in favor of patients who were MRD-negative was seen at 3 and 6 months (Fig. 3C–3F).

The median time to peak concentration (Tmax) was 11 days (range, 3–29), with a significant decrease in CAR T-cell proliferation after 28 days. To comprehensively elucidate the association between CAR T-cell expansion and clinical outcomes, we calculated the median peak copy numbers of HDS269B meticulously. We identified a pronounced association between CAR T-cell expansion and improved clinical outcomes in patients with multiple myeloma. Specifically, patients achieving PR or better showed a higher peak of CAR T-cell expansion compared with those who did not achieve a PR (Fig. 4A). The patients who remained progression-free for up to 5 years postinfusion also exhibited a significantly elevated median peak of CAR T-cell expansion compared with those with disease progression (Fig. 4B). Furthermore, an assessment of the relationship between OS and peak CAR T-cell copy number revealed that patients with a longer OS demonstrated significantly higher median peak copy numbers (Fig. 4C). Subsequently, we detected CAR T-cell persistence at ≥6 months (n = 21), ≥12 months (n = 14), ≥24 months (n = 8), and ≥36 months (n = 2). Among patients with CAR T-cell persistence, we observed significantly improved PFS and OS compared with those without detectable CAR T cells in the blood (Supplementary Fig. S3), thus demonstrating the potential of CAR T-cell persistence as a predictive marker for long-term therapeutic success.

HDS269B treatment during the extended follow-up period was well tolerated, and the side effects were manageable, with a toxicity profile similar to that reported previously (6). Overall, 100% of patients experienced at least one AE. The most common grades (≥3) of prolonged AE experienced after 28 days were hematologic AEs, including neutropenia (38.8%, 19/49), anemia (40.8%, 20/49), thrombocytopenia (30.6%, 15/49), and lymphopenia (26.5%, 13/49). Neither second primary malignancies nor delayed immune effector cell–associated neurotoxicity syndrome was observed.

HDS269B showed notable efficacy in patients with RRMM, with variability in the long-term outcomes. To understand these variations, we analyzed PFS across different timepoints. We categorized patients based on their PFS: <6 months (n = 20), ≥12 months (n = 20), ≥24 months (n = 15), ≥36 months (n = 10), ≥48 months (n = 9), and ≥60 months (n = 4). Notably, 70% of those with PFS <6 months received ≥4 prior lines of therapy compared with 22% with PFS ≥48 months (Supplementary Table S3).

In univariate Cox regression analysis, ≥4 prior therapy lines and CAR T-cell persistence at ≥6 months were identified as a significant predictor of prognosis in 49 patients (Supplementary Table S4). Those with ≥4 prior lines of therapy had worse PFS (HR = 2.372; 95% CI, 1.164–4.832; P = 0.017). The median PFS was 7.50 months (95% CI, 0.52–14.48) for patients with ≥4 prior lines of therapy, versus 21.00 months (95% CI, 6.45–35.55) for those with <4 prior lines of therapy (Fig. 2A). However, multivariate Cox regression analysis revealed CAR T-cell persistence at ≥6 months as an independent predictor of PFS (Supplementary Table S4).

To further clarify the predictors of long-term outcomes, we categorized patients based on their OS: <6 months (n = 11), ≥12 months (n = 30), ≥24 months (n = 21), ≥36 months (n = 18), ≥48 months (n = 15), and ≥60 months (n = 7). Patients with OS <6 months had a higher incidence of EMD compared with those with OS ≥60 months (55% vs. 0%, P = 0.017). Notably, 91% of patients with OS <6 months had been exposed to ≥4 prior lines of therapy compared with 29% with OS ≥60 months (Supplementary Table S5).

Univariate Cox regression analysis showed that the ECOG scores 3 to 4 (HR = 2.35; 95% CI, 1.15–4.78; P = 0.02), EMD (HR = 2.32; 95% CI, 1.06–5.05; P = 0.04), and having received ≥4 prior lines of therapy (HR = 3.17; 95% CI, 1.36–7.41; P = 0.01) were associated with shorter OS (Fig. 5A). Conversely, CAR T-cell persistence at ≥6 months (HR = 0.18; 95% CI, 0.08–0.41; P < 0.01) was significantly associated with longer OS. The statistically significant predictors for OS in multivariate analysis were ECOG scores 3 to 4 (HR = 3.17; 95% CI, 1.38–7.26; P = 0.01) and having received ≥4 prior lines of therapy (HR = 2.53; 95% CI, 1.07–6.01; P = 0.04) as independently shorter OS predictors, whereas CAR T-cell persistence at ≥6 months (HR = 0.15; 95% CI, 0.06–0.36; P < 0.01) was significantly associated with longer OS (Fig. 5B).

Anti–BCMA CAR T-cell therapy has heralded a paradigm shift in the treatment of patients with RRMM for whom the prognosis following salvage chemoimmunotherapy was previously dismal. Despite the reports of promising initial response rates (4, 6, 13, 14), the effect of this treatment on long-term prognosis remains to be elucidated. Here, we report detailed updated long-term follow-up results from 49 patients with RRMM treated with HDS269B including analysis of PFS and OS predictors, even in 20 patients with poor performance status. Our findings highlight the sustained efficacy of HDS269B, particularly in patients with ECOG scores 0 to 2, and that early MRD-negative status, higher and persistent CAR-T-cell expansion, and the absence of EMD correlate with improved prognosis. More importantly, better ECOG scores, fewer prior therapy lines, and CAR T-cell persistence at ≥6 months were identified as key factors for prolonged OS.

With a median follow-up of 59.0 months, the median PFS and OS for all patients were 9.5 and 20.0 months, respectively, a discrepancy with other anti–BCMA CAR-T clinical trials that can be attributed to the high disease burden and poor performance status (ECOG scores 3–4) of the 20 patients enrolled in this study (14, 15). Notably, there was no new treatment-related mortality or toxicities including cytokine release syndrome or immune effector cell–associated neurotoxicity syndrome during the extended follow-up period. Furthermore, it is noteworthy that these patients with poor performance status achieved a 75% ORR, 45% CR rate, and a median OS of 10.5 months. These data suggest the potential for extending access to CAR-T clinical trials to patients with similar performance status, although better outcomes may be achieved by implementing CAR-T therapy earlier in the disease course.

The recent update on the 5-year results from the LEGEND-2 phase I clinical trial of LCAR-B38M in 74 patients with RRMM with ECOG scores 0 to 2 indicated an ORR of 87.8%, a median PFS of 18 months, and a median OS of 55.8 months (15). In contrast to the LEGEND-2 trial, the patients in our study had a higher disease burden, including a higher proportion of patients with ECOG scores more than 2 (41.38% vs. 16.20%), a greater number of prior lines of therapy (median of 4 prior lines vs. 3), and had more high-risk cytogenetic aberrations (41.38% vs. 35.70%). In addition, nearly 80% of the patients experienced disease progression after their most recent prior therapy. However, in patients with ECOG scores 0 to 2, we observed similar survival outcomes to those reported in the LEGEND-2 cohort.

In a phase II trial of combined anti–BCMA and anti–CD19 CAR T-cell therapy, a high rate of MRD-negativity (with a threshold of 10⁻⁵) was observed, predominantly among patients achieving CR or better following the combined treatment (16). Conversely, another study showed that achieving MRD-negative status at the first assessment 1 month after the idecabtagene vicleucel CAR T-cell infusion was independent of the depth of response in patients with RRMM (17). Among the assessable patients in our study, 14 (63.6%) achieved MRD-negative status at day +28, including 3 with CR or better, 3 with VGPR, and 8 with PR, demonstrating the need to evaluate that MRD status of all patients who achieve at least a PR in CAR-T therapy. Moreover, research across various hematologic malignancies (4, 18, 19) has demonstrated that a deeper initial response is strongly predictive of response durability. Therefore, this evidence shows that early MRD-negative responses are associated with superior survival in patients with RRMM following CAR T-cell therapy. The strong association of MRD-negative status at 6 months with superior survival outcomes highlights the value of MRD status as a critical predictor for disease progression and overall patient survival post–CAR T-cell therapy.

CAR T-cell engraftment and expansion have been shown to be crucial for tumor eradication (20). This was confirmed in this 5-year follow-up study showing that higher peak CAR T-cell levels are associated with improved outcomes and aligned with research linking CAR T-cell expansion with leukemia remission (20, 21). Additionally, in a phase II clinical trial, a correlation was observed between the amplification of anti–BCMA and anti–CD19 CAR T cells over the initial 28 days postinfusion and the response in patients with RRMM. Although not statistically significant, patients with a higher peak level of CAR-BCMA tended to experience longer PFS (16), which indicated the importance of CAR T-cell functional fitness for therapeutic efficacy. The pharmacokinetic parameters of HDS269B cells observed at later timepoints revealed a marked correlation between the duration of CAR T-cell expansion and enhanced clinical prognoses in patients with multiple myeloma, highlighting the potential of CAR T-cell persistence as a predictive marker for long-term therapeutic success.

Anti–BCMA CAR T-cell therapy has yielded promising clinical outcomes in patients with RRMM, although survival outcomes vary significantly. With a follow-up period of almost 5 years, our current study showed that patients with longer survival had typically been exposed to fewer prior lines of therapy, which aligns with previous studies (22). A more recent study showed improved PFS in patients with multiple myeloma receiving ≤3 prior therapies (16, 23). This could relate to the condition of the T cells, which has a crucial influence on the success of autologous CAR-T therapy. After extensive cytotoxic treatments, T cells often become exhausted, exhibiting reduced expansion, functionality, and elevated expression of inhibitory receptors, all of which affect CAR T-cell efficacy (24). Similarly, in patients with chronic lymphocytic leukemia treated with tisagenlecleucel, T cells from nonresponders showed heightened exhaustion and effector differentiation pathway activation (25). T cells collected early in the course of the disease demonstrated greater viability and proliferative capacity compared with those collected after relapse (26). The ZUMA-12 study on axicabtagene ciloleucel as a first-line treatment for high-risk large B-cell lymphoma showed that a higher proportion of naïve-like T cells led to greater CAR T-cell expansion compared with ZUMA-1 (27). This evidence suggests that outcomes for patients with multiple myeloma may be improved by earlier CAR-T therapy. Several clinical trials, such as CARTITUDE-5 (NCT04923893; ref. 28) and KarMMa-2 (NCT03601078; ref. 29), are being conducted to evaluate early CAR-T therapy in multiple myeloma. In our phase I study, BCMA–CD19 dual-targeting FasTCAR-T GC012F resulted in an ORR of 100%, with MRD-negativity achieved by all transplant-eligible newly-diagnosed high-risk patients and a very favorable safety profile (1). Further research with more patients and longer follow-up will help fulfill the unmet medical need in this currently incurable disease.

Reports of outcomes in older adults and frail patients receiving BCMA-directed CAR T-cell therapy for RRMM are rare as such patients have historically been excluded from these pivotal trials because of concerns about severe adverse effects. Our analysis revealed a clear variation in the prognosis of patients based on their performance status, with a generally more favorable OS among patients with ECOG scores 0 to 2. This aligns with the 5-year follow-up results of the LEGEND-2 study, in which patients who were less heavily pretreated and those with a favorable performance status seemed to derive greater benefit from LCAR-B38M CAR T-cell therapy (15). As the disease progresses and patients undergo more treatments, their physical condition typically worsens, further highlighting the potential benefits of earlier CAR-T intervention.

EMD has been previously recognized as an influential factor in the poor prognosis of multiple myeloma (22, 30); even if patients received therapy with a CAR T-cell approach, the response rate of patients with EMD present at the time of infusion seems to be similar to that observed in patients without EMD (3, 31). However, EMD has shown an inferior long-term outcome with shorter PFS and OS than those with no EMD (32, 33). Patients with EMD in this study had a shorter PFS and OS, which were similar and consistent with previous studies (32, 33). However, EMD was not identified as an independent prognostic factor in multivariate Cox regression analysis, which might be due to the small sample size of EMD patients we included as well as the two types of EMD (34), paraskeletal plasmacytomas and extramedullary plasmacytomas, consisting of soft-tissue masses with no contact with bones as a result of hematogenous spread, not being analyzed separately. The tumor microenvironment in EMD can hinder CAR T-cell movement through physical barriers, as well as the production of soluble factors, and immunosuppressive cells (35, 36). Moreover, the heterogeneity and persistence of EMD may lead to the emergence of clones that are resistant to anti–BCMA CAR-T therapy (37). In general, patients with myeloma as well as EMD have inferior long-term survival compared with those without EMD, even after CAR-T therapy. Thus, EMD remains a challenge to the success of CAR T-cell therapy.

Despite the relatively long-term follow-up after CAR-T therapy in this study, some limitations should be noted, including the lack of differentiation between bone-related EMD and soft tissue–related EMD due to the small cohort size and variability in criteria for MRD-negative status across different studies, which may constrain the widespread applicability of our findings. Larger-scale studies are required to confirm our results and standardize the MRD assessment criteria. Additionally, exploring the long-term effects and potential late complications of CAR-T therapy in a more diverse patient population could provide deeper insights and enhance the generalizability of our results. These avenues of investigation are crucial for refining CAR-T therapy and maximizing its efficacy and safety for broader patient populations. HDS269B phase II clinical trials that are active and recruiting in China are registered with ClinicalTrials.gov (NCT05594797).

In conclusion, our 5-year follow-up data indicate that ECOG performance status, the number of prior lines of therapy, and CAR T-cell persistence ≥6 months are important independent prognostic indicators in patients treated with anti–BCMA CAR T-cell therapy. Therefore, early administration of CAR T-cell therapy during the disease course may improve outcomes, especially in patients with poor ECOG performance status.

F. Xiang was an employee of HRAIN Biotechnology Co., Ltd. during the conduct of the study. X. Sun was an employee of HRAIN Biotechnology Co., Ltd. during the conduct of the study. No disclosures were reported by the other authors.

D. Chen: Data curation, methodology, writing–original draft, writing–review and editing. Y. Zhu: Data curation, methodology, writing–original draft, writing–review and editing. Z. Chen: Conceptualization, supervision, validation, investigation, methodology, writing–review and editing. S. Jiang: Conceptualization, supervision, validation, investigation, methodology, writing–review and editing. H. He: Data curation, methodology, writing–review and editing. W. Qiang: Data curation, methodology, writing–review and editing. F. Xiang: Conceptualization, data curation, methodology. X. Sun: Conceptualization, data curation, methodology. J. Du: Conceptualization, data curation, supervision, funding acquisition, investigation, methodology, writing–original draft, project administration, writing–review and editing.

This study was partially funded by the Shanghai Scientific and Technological Committee (21Y11908900), the Clinical Research Plan of SHDC 2022CRW015, and the Science and Technology Innovation Action Plan of Shanghai 23S11905500.