Phase I clinical study of humanized BCMA-single-domain antibodies targeting CAR-T in patients with relapsed or refractory multiple myeloma

Clinical characteristics and disposition

As of May 13, 2024, 15 patients with RRMM received BCMA CAR-T cell infusion therapy at 2.5 × 10⁶ (n = 3), 5.0 × 10⁶ (n = 11), or 7.5 × 10⁶ CAR-positive T cells/kg (n = 1) (Figure S1), according to the dose at enrollment. The median age was 63.0 years (range, 36.0–73.0; Table S1). Overall, 73.3% of the participants (n = 11) had previously received ≥4 lines of MM therapy, and the median was 5.0 lines (range, 3.0–9.0). After CAR-T therapy, no patients in this study received subsequent maintenance therapy or transplantation.

Safety

All 15 patients experienced grade 3–4 treatment-emergent adverse events (AEs) during the study period. The most common grade ≥3 AEs were hematological toxicity, including neutropenia (n = 14, 93.3%), leukopenia (n = 12, 80.0%), lymphopenia (n = 11, 73.3%), anemia (n = 9, 60.0%), and thrombocytopenia (n = 4, 26.7%). One patient in the 7.5 ×10⁶ CAR-positive T cells/kg dose group had a dose-limiting toxicity (DLT) event of grade 4 thrombocytopenia. The most common non-hematologic grade ≥3 AEs were infections (n = 3, 20.0%), but no infectious events occurred in the 0.25 × 10⁶ CAR-positive T cells/kg dose group. Renal calculus was the second most common non-hematologic grade ≥3 AE (n = 1, 6.7%; Table 1).

All 15 patients (100.0%) developed grade 1–2 cytokine release syndrome (CRS), whereas no grade ≥ 3 CRS occurred (Table S2). The median time to CRS was 2.0 days (range, 1.0–13.0 days), and the median duration was 6.0 days (range, 2.0–17.0 days). All CRS recovered without sequelae after supportive treatment (tocilizumab, n = 6; glucocorticoids, n = 4). No immune effector cell-associated neurotoxicity (ICANS) or death was observed.

Efficacy

At the data cut-off, the median follow-up time for all patients was 22.5 months. The median time to first response was 1.0 month (range, 0.9–2.2 months). The overall remission rate (ORR) was 100.0% (95% confidence interval, 78.2%–100.0%; Figure S2); 10 (66.7%) patients achieved strict complete remission (sCR), 3 (20.0%) patients achieved very good partial response (VGPR), and 2 (13.3%) patients achieved PR as the best response. All patients with sCR achieved minimal residual disease negativity. The ORR of 6 patients with extramedullary plasmacytoma was 100.0%: 2 patients achieved sCR, and 4 achieved VGPR. Notably, a gradual deepening of remission was observed in 11 patients (73.3%) during 60–90 days after infusion, and 7 patients (46.7%) achieved sCR at day 360. One year after CAR-T cell infusion, 7 patients achieved sCR, and 4 patients maintained sCR after 2 years.

CAR-T cell pharmacokinetic characteristics

As of May 7, 2024, the CAR transgene in peripheral blood increased rapidly post-infusion and peaked at a median of 14.0 days (range, 8.0–28.0 days). The median persistence of the CAR transgene was 65.0 days (range, 15.0–457.0 days), and BCMA CAR-T was detectable in 26.7% (4/15) of patients at 6 months after infusion (Figure S3).

Preliminary exploratory case observation

A 64-year-old female patient with RRMM received BCMA CAR-T cell infusion therapy with 7.5 × 10⁶ CAR-positive T cells/kg. At baseline, her ECOG performance status was 1. She experienced a DLT event of grade 4 thrombocytopenia after BCMA CAR-T cell infusion therapy. Other grade ≥3 AEs included leukopenia, neutropenia, lymphopenia, anemia, and infection. She developed grade 1–2 CRS, including fever, anoxia, and hypotension. A PR was observed on day 28 post-infusion, and she achieved a VGPR by day 90. One year after infusion, she had reached a sCR, which was sustained for at least 1 year.

BCMA-targeted CAR-T cells have the potential to induce durable and deep remission in patients with RRMM. Currently, 4 BCMA-targeted CAR-T therapies have been approved: Carvykti, Abecma, Equecabtagene Autoleucel, and Zevalocel^9,10. The therapeutic effects of BCMA-targeted CAR-T therapy are encouraging. In a clinical trial by Carvykti, the response rate was 97%, and 67% of patients achieved a complete response⁵. All these BCMA-targeted CAR-T therapies showed similar results in clinical trials, and Zevalocel achieved an ORR of 100.0%. In our study, we achieved an ORR of 100.0%, among which the best response in 10 patients (66.7%) was sCR, and all patients achieving sCR showed minimal residual disease negativity. Moreover, we observed that patients with extramedullary plasmacytoma also achieved an ORR of 100%. We therefore applied BCMA-targeted CAR-T cells to patients with heavy tumor burden. However, further trials are required to investigate the efficacy of this treatment in this population.

All patients in our study experienced grade 1–2 CRS, whereas no grade ≥3 CRS occurred. No significant differences were observed among dose groups. In agreement with previous reports, CRS was among the most reported AEs in BCMA-targeted CAR-T therapies^5,6. The time to onset of CRS post CAR-T infusion varies across studies and is influenced by multiple factors, such as the design of the CAR construct, the CAR-T cell dose, the cellular composition of the CAR-T cell product, the manufacturing process used to generate the CAR-T cells, and host characteristics (e.g., tumor burden and patient age)¹¹. The onset of CRS toxicity usually occurs within the first week after CAR-T cell therapy¹. In our study, the median time to onset of CRS was 2.0 days, in agreement with this expected pattern and the reported range for other anti-BCMA CAR-T therapies in RRMM (2.0–9.0 days)^5,6,12. No neurotoxicity was observed, and common grade ≥3 AEs were hematological toxicities. One patient in the 7.5 × 10⁶ CAR-positive T cells/kg dose group had a DLT event of grade 4 platelet count decrease, whereas no DLT events occurred in the other low-dose groups. Additional studies are required to explore the relationship between hematologic toxicity and CAR-T therapies. The most common non-hematologic grade ≥3 AEs were infections (20.0%), and all patients recovered. Overall, the BCMA CAR-T cells showed a manageable safety profile.

The CAR structure in this study, used by BCMA CAR-T, is composed of a single-domain antibody targeting BCMA, a CD8 hinge region, transmembrane region, 4-1BB co-stimulatory domain, and CD3-ζT cell activation domain. Heavy-chain antibodies consist of only 2 heavy chains without light chains but retain the function of conventional antibodies, thereby forming a small, stable, single-domain structure with high affinity and specificity, which is easily humanized for therapeutic purposes. In this study, we observed no ICANS in response to BCMA CAR-T cell administration. BCMA CAR-T cells are safe and effective for treating patients with MM and induce sustained remission.

Ill-defined generalizability

This study has several limitations. First, the relatively small sample size limits the generalizability of the findings, and the primary objective was to evaluate the initial safety profile of BCMA CAR-T cells. Efficacy outcomes should therefore be interpreted with caution, and further studies are needed to confirm our findings in patients with RRMM. Second, given the exploratory nature of this study and the limited follow-up duration, survival data are currently immature and therefore are not reported herein. Moreover, the Kaplan-Meier method, although widely used, might overestimate survival probabilities in small cohorts, and the generated curves should therefore be interpreted in light of this limitation. Third, although ORR estimates were accompanied by 95% confidence intervals calculated with the Clopper-Pearson method, these intervals were wide, thereby reflecting the limited sample size, and consequently do not necessarily imply statistical precision. Collectively, these statistical and methodological limitations highlight the need for further investigation in larger clinical studies with longer follow-up.

In conclusion, BCMA CAR-T is well tolerated and highly efficacious in patients with RRMM. Patients with extramedullary plasmacytoma showed comparable responses to those without extramedullary disease.

Author contributions

Conceived and designed the analysis: Gaofeng Zheng, Ruyi Xu, Zhen Cai.

Collected the data: Ruyi Xu.

Contributed data or analysis tools: Heng Mei.

Performed the analysis: Gaofeng Zheng, Xiaoyan Han, Donghua He, Yanling Weng.

Wrote the paper: Ruyi Xu, Cheng Wen, Zhuoxiao Cao.