SUMMARY

• LEGEND-2 was a first-in-human, phase 1, single-arm, open-label study evaluating the safety and efficacy of LCAR-B38M chimeric antigen receptor T (CAR-T) cells in patients with relapsed/refractory multiple myeloma (RRMM).^1,2
  • A total of 74 patients with RRMM were treated across 4 centers in China (57 patients at the Xi’an site; 17 patients each at the Ruijin, Changzheng and Jiangsu sites).²
  • ≥5-Year Follow-Up
    • At a median follow-up of 65.4 months (≥5 years after the last patient was dosed), treatment-emergent adverse events (TEAEs) were reported by all patients during the first month of treatment. Cytokine release syndrome (CRS) was the most common TEAE, occurring in 91.9% of patients.³
    • Median progression free survival (PFS) and median overall survival (OS) were 18.0 and 55.8 months, respectively. The 5-year PFS and 5-year OS rates across all treated patients were 21.0% (95% confidence interval [CI], 12.2-31.4) and 49.1% (95% CI , 37.2-60.0), respectively.³
    • The median duration of response (DOR) was 23 months.³
  • 4-Year Follow-Up
    • At a median follow-up of 47.8 months, grade ≥3 adverse events (AEs) occurred in 60.8% of patients: most commonly, leukopenia (25.7%), aspartate aminotransferase (AST) elevation (20.3%), and thrombocytopenia (18.9%). CRS occurred in 91.9% of patients, of which most were of grade 1 (47.3%) and grade 2 (35.1%).²
    • Median PFS and median OS across all treated patients were 18.04 months and not estimable (NE), respectively. The 24-month OS across all treated patients was 63.4-months.²
    • The median DOR was 23.26 months.²
    • The efficacy of LCAR-B38M was evaluated in a subgroup of 22 patients with RRMM and extramedullary disease (EMD) at a median follow-up of 47.8 months. In patients with EMD vs without EMD, ORR was 81.8% vs 90.4% and complete response (CR) was 54.5% vs 80.8%.²

CLINICAL DATA

LEGEND-2 (clinicaltrials.gov identifier: NCT03090659) was an investigator-initiated, first-in-human, phase 1, single-arm, open-label study evaluating the safety and efficacy of LCAR-B38M CAR-T cells in 74 patients with RRMM. The study was conducted at 4 sites in China (57 patients at the Xi’an site; 17 patients at the Ruijin [n=5], Changzheng [n=3] and Jiangsu [n=9] sites).^2,4

Study Design/Methods

• Key inclusion criteria: patients aged 18-80 years, International Myeloma Working Group (IMWG) diagnosis of multiple myeloma (MM) with documented disease progression ≤12 months of most recent anti-MM medication treatment or autologous hematopoietic stem-cell transplant, ≥3 prior lines of therapy.^2,4
• Lymphodepletion and dosing: each study site had its own protocol for lymphodepletion and variable CAR-T infusion methods (split vs single infusion).^2,4
  • Lymphodepletion was performed using cyclophosphamide 300 mg/m² or cyclophosphamide 250 mg/m² plus fludarabine 25 mg/m². LCAR-B38M CAR-T cells were infused over 3 split infusions (n=65) or as a single infusion (n=9).^2,4
    • Xi’an site: patients received cyclophosphamide 300 mg/m² daily for 3 days on days -5 to -3. Patients received split infusions of LCAR-B38M CAR-T cells on day 1, 3, and 7 (20%, 30%, and 50% of the total dose, respectively) administered over 7 days.^2,4
    • Ruijin and Changzheng sites: patients received cyclophosphamide 250 mg/m² plus fludarabine 25 mg/m² daily for 3 days on days -5 to -3. Patients received split infusions of LCAR-B38M CAR-T cells on day 0, 3, and 6 (20%, 30%, and 50% of the total dose, respectively).^2,5
    • Jiangsu site: patients received cyclophosphamide 300 mg/m² day daily for 3 days on days -5 to -3. Patients received single dose infusions of LCAR-B38M CAR-T cells on day 0.^2,5
  • CAR-T cells were successfully manufactured for all patients and infused at a median dose of 0.513 × 10⁶ cells/kg (range, 0.07-2.10x10⁶). Some patients received a below-target dose of LCAR-B38M.²
• Primary objective: safety of LCAR-B38M CAR-T cells^2,4
• Secondary objective: antimyeloma activity based on IMWG response criteria.^2,4

Patient Characteristics

• Overall, 74 patients (Xi’an site, n=57; Rujin site, n=5; Changzhen site, n=3; and Jiangsu site, n=9) who underwent leukapheresis were included.²
• At the clinical cut-off date of May 25, 2021, 39 patients were continuing the study, 34 patients had died, and 1 patient was lost to follow-up.²
• Baseline patient demographics and disease characteristics of 74 patients enrolled in the study have been presented in Table: LEGEND-2: Baseline Patient Demographics and Clinical Relevance.

≥5-Year Follow-Up

Xu et al (2024)³ reported the results from the entire population of the LEGEND-2 study at a longer median follow-up of 65.4 months; ≥5 years after the last patient was dosed.

Results

Safety

• All patients experienced TEAEs during the first month of treatment.
• Grade 3/4 neutropenia and thrombocytopenia were observed in 85.3% and 58.8% of patients.
  • Six months after the infusion, 47.3%, 44.6%, and 59.5% of patients reached normal leukocyte, hemoglobin, and platelet levels, with median values of 4.7×10⁹/L, 125.3 g/L, and 155.0×10⁹/L, respectively.
• Grade 3/4 hepatic disorder was observed in 38.3% of patients.
  • The median levels of AST and alanine aminotransferase (ALT) were 21.0 U/L and 18.5 U/L, respectively.
• There were no significant differences in hematologic and hepatic function recovery between the non- progressive disease (PD) and PD/death groups.
• CRS was the most common TEAE, occurring in 91.9% of patients (n=68); resolving within 30 days post CAR-T cell infusion for most patients.
  • Extended CRS was reported in 3 patients which lasted 31, 37, and 51 days, respectively, with no second wave observed.
• Transient Grade 1 immune effector cell-associated neurotoxicity syndrome (ICANS) was reported by 1 patient.
• One patient died by suicide after a psychiatric disorder developed following myeloma progression, with no signs of movement and neurocognitive TEAEs.
• Patients in CR with a normal Ig level did not report serious infections.
• Virus reactivation or infection was noted beyond 6 months of follow-up without severe complications.
  • Herpes zoster virus was reported in 3 patients and hepatitis B virus in 1 patient; all recovered with antiviral treatment and Ig administration.
  • Of 19 surviving patients with a history of coronavirus disease 2019 (COVID-19), 6 tested negative, and 13 were infected between 2020 and 2022.
    • All 13 infected patients had mild symptoms like fever, cough, and fatigue lasting 1-3 days, with no COVID-19 related pneumonia or deaths reported.
• By the cutoff date, 4 patients developed secondary primary malignancies (SPM) of the lung, esophagus, and cervix 8-32 months after LCAR-B38M treatment.
  • All patients achieved myeloma remission with no detectable concurrent hematologic malignancies; 3 patients died of SPM.
  • One patient, cured of cervical cancer, had a multiple myeloma relapse 50 months after the infusion and achieved CR after bortezomib and lenalidomide treatment.

Efficacy

• As of November 30, 2022, a total of 33 patients (44.6%) were alive; of which, 12 patients (16.2%) were in remission without relapse.
• The longest observed remission was 6.4 years.
• A total of 62 patients experienced PD (n=53 [after partial response or better (≥PR)]) or death.
• A total of 32 patients died due to PD, while 9 patients died due to non-relapse-related events.
• A total of 21 patients with PD had survived after receiving subsequent treatments.
• Median PFS and median OS were 18.0 and 55.8 months, respectively.
  • For patients who achieved CR and minimal residual disease (MRD)-negative CR, the median PFS was prolonged to 28.2 and 30.6 months, respectively. Median OS was not reached for those patients.
  • For patients who did not achieve ≥CR, the median PFS was 4.4 months and median OS was 7.9 months.
  • Median PFS, OS, and DOR in patients achieving CR and MRD-negative CR is presented in Tables: LEGEND-2: PFS, OS, and DOR in Patients With or Without CR and MRD-Negative CR.
• The 5-year PFS and 5-year OS rates across all treated patients were 21.0% (95% CI, 12.2-31.4) and 49.1% (95% CI , 37.2-60.0), respectively.
• The 5-year PFS and 5-year OS rate among the 65 responders were 23.6% and 55.9%, respectively.
• The 5-year PFS and 5-year OS rates of the 54 CR patients were 28.4% and 65.7%, both being significantly higher than those without achieving CR (P<0.0001).
• A total of 24.0% of the MRD-negative responders (12/50) remained in deep remission with sustained responses at data cut off.
• Median DOR for all patients was 23 months, and for MRD-negative responders vs other responders, it was 32.7 months vs 7.5 months (P<0.0001).
• Patients with longer time to best response (≥3.3 months) had longer PFS (P=0.006) and OS (P=0.001) compared with those achieving best response within 3.3 months.

Immunogenicity

• Among the 26 patients achieving CR with available Ig data, 21 had serologically normal Ig levels; the Ig levels of 1 patient normalized at 5.4 years and those of the remaining 20 patients normalized within 5 years.
• The median time to full Ig restoration was 16.7 months, with the earliest recovery observed at 9.1 months post infusion. All patients who remained relapse-free had serum Ig recovery.
• The 5-year OS rate for the 21 CR patients with full Ig recovery was 100%, which was not significantly different from the five patients with abnormal Ig levels (P=0.14).
• Disease relapse occurred at 68.3-, 69.5-, and 62.4-months post infusion in 3 patients who had achieved MRD-negative CR, respectively.
  • One patient developed extramedullary lesion without detectable tumor in bone marrow, and 2 patients had an increased level of monoclonal paraprotein. All proceeded with subsequent therapies.

Pharmacokinetics/Pharmacodynamics

• During the observation period up to 180 days post CAR-T cell infusion, the average circulating lymphocyte counts peaked between 14 and 28 days and gradually declined after 1 month.
• Although the non-PD cohort exhibited higher average lymphocyte count peaks compared with PD/death patients during CAR-T cell expansion, a high peak level of LCAR-B38M did not correlate with CR or sustained remission in the 16 patients with available transgene data.
• The prognostic relevance of long duration of the transgene persistence (T_last) is presented in Table: LEGEND-2: Prognostic Relevance of T_last.
• As of the cut-off date, CAR-T cells remained detectable in 1 patient who had maintained MRD-negative CR for 5.6 years at the last follow-up.

4-Year Follow-Up

Zhao et al (2022)² reported the results from the entire population of the LEGEND-2 study at a median follow-up of 47.8 months (range, 0.4-60.7).

Results

Safety

• All patients experienced ≥1 acute AEs within 100 days after infusion and grade ≥3 AEs were reported in 45 patients (60.8%). See Table: LEGEND-2: Summary of AEs Occurring in ≥15% of Patients.
• CRS occurred in 68 patients (91.9%). The median time to CRS onset was 9 days (range, 1-19) and median duration of CRS was 9 days (range, 3-57).
  • The maximum CRS grade observed in patients was: grade 1: 47.3%, grade 2: 35.1%, grade 3: 8.1%, grade 4: no events, grade 5: 1 event.
  • CRS was managed using tocilizumab (44.6%), oxygen supply (36.5%), and vasopressor therapy (9.5%).
  • Symptom resolution occurred in 66 patients but persisted in 2 patients.
• Reversible grade 1 central nervous system (CNS) toxicity manifested by aphasia, agitation, and seizure-like activity was reported in 1 patient. No patient reported signs or symptoms of parkinsonism.
• The most common chronic AE observed was B cell aplasia.
• Viral activation was reported in 4 patients (herpes zoster virus infection [n=3; at 7, 26, and 46 months after infusion, respectively] and hepatitis B virus activation [n=1; at 7 months after infusion). All viral infections resolved with antiviral drug therapy and supporting treatment.
• Second primary non-hematologic tumors with undetectable myeloma cells were reported in 4 patients. All cases were considered unrelated to LCAR-B38M therapy.
  • Lung cancer was diagnosed at 8 months after infusion in 1 patient who underwent surgery followed by chemotherapy, but eventually died of pneumonia.
  • Lung cancer was detected at 32 months after infusion in 1 patient who survived.
  • Esophageal cancer developed at 15 months after infusion in 1 patient who eventually died.
  • Cervical cancer was reported at 8 months after infusion in 1 patient who responded to standard anti-tumor treatment.
• In total, 34 deaths were reported. The causes of deaths were as follows:
  • Disease progression (n=28)
  • CRS and tumor lysis syndrome (n=1, day 13)
  • Pulmonary embolism and potential acute coronary syndrome (n=1, day 22)
  • Respiratory failure associated with subsequent therapy (n=1, day 436)
  • Esophageal carcinoma (n=1, day 493)
  • Infection (n=2, days 157 and 529, respectively)

Efficacy

• Efficacy outcomes are summarized in Table: LEGEND-2: Best Responses in All Patients. Survival outcomes are summarized in Table: LEGEND-2: Survival Outcomes. The median times to initial response and best response were 1.02 months (range, 0.43.5) and 3.32 months (range, 0.43-28.52) respectively.
  • CR rates were comparable between patients with high vs standard cytogenetic risk (80% vs 66.7%) and no significant difference was observed in survival outcomes.
  • ORR and CR rates were comparable among various subgroups differentiated by age, time from disease diagnosis to CAR-T cell therapy, disease stage, number of prior therapies, and prior proteasome inhibitor (PI)/immunomodulatory drug exposure.
  • The longest PFS without maintenance treatment was 60 months (observed in a patient with λ light-chain-type MM and extensive EMD).
  • PD developed in 43 of 65 patients who had achieved ≥PR. The median OS of patients with PD was 36.14 months (95% CI, 21.13-NE).
  • Death was reported in 22 patients with PD. The median OS in the 21 surviving patients was 19.68 months after receiving subsequent line of therapy.

Pharmacokinetics/Pharmacodynamics

• The median T_last was 280 days (range, 81466).
  • Patients with T_last >280 days had a longer OS than patients with T_last <280 days (P=0.010); however, no such correlation was observed with PFS or DOR.
• No significant difference was detected in T_last between the cyclophosphamide + fludarabine (n=7) vs cyclophosphamide alone (n=46) groups. Although the number of patients were small and not proportional; overall T_last did not differ between patients with and without PD (287.5 vs 231 days; P=0.921).
• Serum soluble B-cell maturation antigen (sBCMA) was detected in all of the 47 evaluable patients. A mean decrease of 94.2% from baseline was observed in 37 patients.
  • The median time to nadir sBCMA was 155 days (range, 3-1444).
  • sBCMA levels rebounded higher in 11 evaluable patients who relapsed vs 14 evaluable patients who did not relapse (median 38,909 pg/mL vs 17,357.5 pg/mL, P=0.002).
  • Patients with >80% decline in sBCMA from baseline had a longer OS (P<0.001) and PFS (P=0.01) than patients with a mild decrease in sBCMA.
  • No correlation between sBCMA levels and DOR was observed.
• Positive antidrug antibodies (ADAs) were reported in 34 of 55 (62%) evaluable patients with the mean time to ADA positivity being 229.6 days (range, 13-905).
  • The appearance of ADAs was less frequent in the cyclophosphamide + fludarabine group compared to the cyclophosphamide alone group (28.6% vs. 66.7%).
  • No significant difference was observed in ADA positivity between single-dose infusion (66.7%) and split infusion (60.9%).
  • No difference in median T_last was observed between patients with (n=32) vs without (n=18) ADA positivity (339 vs 137 days, P=0.104).
  • The presence of ADAs had no impact on DOR, PFS, or OS.

Clinical Data - Subgroup Analysis: Patients with RRMM and EMD

Zhao et al (2022)² conducted a subgroup analysis of patients with vs without EMD in the LEGEND-2 study at a median follow-up of 47.8 months.

Results

• Of the 74 study patients, 22 patients (Xi’an site, n=17; East China, n=5) had baseline EMD, including EMD not adjacent to bone (n=15; 2 of whom had concomitant plasma cell leukemia) and bone-based EMD (n=7).
• No significant differences were observed in gender, age, or number of prior lines of therapy in patients with EMD vs without EMD. However, in patients with EMD, the light-chain MM subtype was more frequent than the immunoglobulin G (IgG) or immunoglobulin A (IgA) subtypes.
• Efficacy outcomes by EMD status are summarized in Table: Efficacy Outcomes by EMD Status.
• Extramedullary relapse was reported in 18 of 22 patients with baseline EMD, of whom 1 patient died due to infection, 1 patient was nonresponsive to LCAR-B38M, and 2 patients have not relapsed.

Literature Search

A literature search of Ovid MEDLINE^®, Embase^®, BIOSIS Previews^®, and Derwent Drug File (and/or other resources, including internal/external databases) was conducted on 03 July 2024.