Standard-of-care idecabtagene vicleucel for relapsed/refractory multiple myeloma

Key Points

• •Largest real-world study of ide-cel in R/R MM shows favorable safety and efficacy profile that mirrors trial experience.
• •Significant comorbidities present in 77% patients; large sample size allowed identification of prognostic factors for safety and efficacy.

Visual Abstract

Abstract

Idecabtagene vicleucel (ide-cel) was the first US Food and Drug Administration–approved chimeric antigen receptor T-cell (CAR-T) therapy for multiple myeloma (MM). However, because clinical trials are highly selective with stringent eligibility criteria, the objective of this study was to evaluate the safety and effectiveness of standard-of-care (SOC) ide-cel in the real world. Using the Center for International Blood and Marrow Transplant Research registry, we evaluated 821 patients who received SOC ide-cel. Median follow-up was 11.6 months. Median age was 66 years, and the cohort included 31% patients aged ≥70 years, with 15% Black and 7% Hispanic, and 77% of patients with ≥1 significant comorbidity. The median number of prior lines of therapy was 7, 15% patients previously received B-cell maturation antigen–directed therapy, 17% had extramedullary disease, and 27% had high-risk cytogenetics. Overall response rate was 73%, and complete response rate was 25%. Median progression-free survival was 8.8 months. Treatment-related mortality was reported in 6% of patients. Cytokine release syndrome was diagnosed in 80% of patients (grade ≥3, 3%). Immune effector cell–associated neurotoxicity syndrome was observed in 28% (grade ≥3, 5%), with no cases of Parkinsonism reported. Clinically significant infections were seen in 45% of patients. Second primary malignancies were reported in 4%, including 1% myeloid malignancies. This is, to our knowledge, the largest real-world study of ide-cel CAR-T therapy in patients with relapsed/refractory (R/R) MM. We observed a favorable safety and efficacy profile that mirrors trial experience, even in the setting of significant comorbidities in 77% of patients, many of which would have made them ineligible for the registrational KarMMa clinical trial. This trial was registered at www.clinicaltrials.gov as #NCT03361748.

Sidana et al report on the largest real-world analysis of chimeric antigen receptor T-cell therapy in multiple myeloma (MM), studying 821 patients, the majority of whom would have been ineligible for the pivotal trial, who received idecabtagene vicleucel (ide-cel), which targets B-cell maturation antigen (BCMA), in the late relapse setting. The 77% overall response rate, 25% complete response rate, and median progression-free survival of <9 months are consistent with the trial, expanding applicability and indicating that high-risk genetics, high disease burden, and prior anti-BCMA therapy are negative prognostic factors in multivariable analyses. As treatment options for patients with MM expand, these real-world data offer robust estimates of ide-cel efficacy and safety to guide clinical decision-making.

Introduction

Idecabtagene vicleucel (ide-cel) was the first commercially available chimeric antigen receptor (CAR) T-cell (CAR-T) therapy for multiple myeloma (MM).¹ Ide-cel is a murine-derived CAR-T therapy designed to target B-cell maturation antigen (BCMA), which is almost universally expressed on plasma cells, both normal and malignant.^{2, 3, 4} It was initially approved in March 2021 by the US Food and Drug Administration for treatment of patients with relapsed/refractory MM (R/R MM) with 4 prior lines of therapy, including a proteasome inhibitor (PI), immunomodulatory drug (IMiD), and anti-CD38 antibody.⁵

In the pivotal phase 2 KarMMa trial, that included 128 patients treated with ide-cel after a median of 6 prior lines of therapy, the response rate was 73%, including a complete response (CR) rate of 33%.¹ The median progression-free survival (PFS) was 8.8 months. These data were superior to those seen in historical cohorts of similar patients.⁶ However, clinical trials of CAR-T therapy have stringent eligibility criteria resulting in inclusion of fit patients with limited comorbidities as trial participants. All genetically manipulated cellular products are required to have a long-term follow-up study to evaluate safety and efficacy in a broader, real-world setting as a condition of drug approval. The Center for International Blood and Marrow Transplant Research (CIBMTR) registry was used to run this prospective observational study to evaluate the real-world outcomes of ide-cel. We hypothesized that standard-of-care (SOC) ide-cel would demonstrate similar effectiveness and safety profile when administered to a clinically diverse a real-world population of patients, consistent with results from the KarMMa-1 trial. In addition, we used these data to further explore outcomes in different patient populations, including those treated with alternate lymphodepleting agents.

Methods

Data sources

Data were collected using the CIBMTR registry. Details about CIBMTR data collection and validation are described in the supplemental Methods, available on the Blood website.⁷ The National Marrow Donor Program institutional review board reviews CIBMTR research. Patients or guardians gave informed consent to share data for research.

Patient eligibility

Patients with R/R MM were included in this study if they had received SOC ide-cel in the United States between May 2021 and June 2023 after at least 4 prior lines of therapy. Patients needed to have at least 1 follow-up evaluation completed. Data cutoff for follow-up was October 2023. The first evaluation occurred on day 100 (or earlier in case of patient death). Patients who received treatment with nonconforming ide-cel or as part of a clinical trial were excluded.

Definitions and outcomes

High-risk cytogenetics were defined by the presence of del 17p/monosomy 17, t(4;14), t(14;20), and t(14;16) by fluorescence in situ hybridization or cytogenetics at any time after MM diagnosis. Triple-class exposure was defined as exposure to a PI, IMiD, and anti-CD38 antibody. Penta-drug exposure was defined as exposure to 2 PIs (bortezomib, ixazomib, carfilzomib, or oprozomib), 2 IMiDs (lenalidomide and pomalidomide), and an anti-CD38 antibody (daratumumab and/or isatuximab). Bridging therapy was defined as the last line of therapy before CAR-T infusion initiated or completed in the interval between leukapheresis and lymphodepleting chemotherapy.

The primary outcomes were efficacy as defined as the overall response rate (ORR) according to International Myeloma Working Group criteria⁸ and PFS from the time of CAR-T infusion. PFS was defined as the time from ide-cel infusion to progression or death, whichever occurred first. Secondary end points included overall survival (OS) and the following measures of safety: incidence of cytokine release syndrome (CRS), immune effector cell–associated neurotoxicity syndrome (ICANS), delayed neurotoxicities, immune effector cell–associated hemophagocytic lymphohistiocytosis–like syndrome (IEC-HS), cytopenia, infections, second primary malignancies (SPMs), and treatment-related mortality (TRM). OS was defined as time from CAR-T infusion to death or last follow-up. CRS and ICANS were graded per the American Society of Transplantation and Cellular Therapy consensus criteria.⁹ Neutrophil count recovery was defined as achieving a neutrophil count of >0.5 × 10³/μL for 3 consecutive days with the first day counted as the day of recovery. Platelet recovery was defined as a platelet count of >20 × 10³/μL without platelet transfusion within the last 7 days of the follow-up time point. Prolonged cytopenia was defined as absence of neutrophil and/or platelet recovery by day 30 after CAR-T infusion. Clinically significant infections were defined as any infection requiring treatment as defined in the CIBMTR manual. Note, culture-negative neutropenic fever without clear source or upper respiratory infections, which are presumed viral but in which no virus was identified, are not considered clinically significant. TRM was defined as death in absence of MM disease progression. Disease progression was defined by the treating physician and reported according to specifications defined in the CIBMTR forms manual.

Statistical analysis

Descriptive statistics were used to summarize patient and treatment characteristics and outcomes. Survival probabilities were calculated using the Kaplan-Meier method, and survival curves were compared using the log-rank test. P values for survival analysis reported throughout the article are log-rank P values. Cox proportional hazards models were used to identify risk factors associated with the following outcomes: overall response, CR, PFS, OS, CRS grade ≥2, and any grade ICANS. Variables considered for inclusion into the models were: age category (aged <60, 60-69, and >70 years), sex, race (White, Black, African-American, others, and not reported), Eastern Cooperative Oncology Group (ECOG) performance status score at infusion (ECOG performance status score of 0-1, ECOG performance status score of >2, or missing), number of prior lines of therapy (4-5, 6-10, 11-15, and 16 -21), prior BCMA-directed therapy excluding prior BCMA CAR-Ts (no prior BCMA therapy, BCMA therapy <6 months prior, and BCMA >6 months prior), International Staging System (ISS) stage at diagnosis (I, II, III, and unknown), plasma cells in the bone marrow before infusion (<50%, >50%, and unknown), lactate dehydrogenase before infusion (normal, elevated, and unknown), presence of extramedullary disease (EMD), disease status at infusion (better than very good partial response [VGPR], partial response [PR], stable disease, progressive disease, and unknown status), absolute neutrophil count before ide-cel infusion (>1 × 10³ vs <1 × 10³/μL), platelet count before ide-cel infusion (>50 × 10³ vs <50 × 10³/μL), type of lymphodepleting chemotherapy (fludarabine and cyclophosphamide vs bendamustine), and ide-cel dosing (>400 million cells vs <400 million cells). A stepwise variable selection was used to identify the covariates for inclusion in the final model with P value of .05 for selected variables. Proportionality assumption and cross-variable interactions were tested on all models. Hazard ratios (HRs) with 95% confidence intervals (CIs) are reported. All P values were 2 sided, and the difference between 2 variables was considered significant at P < .05.

Results

The study cohort included 821 patients with R/R MM who received SOC ide-cel. The median follow-up of the cohort was 11.6 months (range, 1.1-26.7; CONSORT flow diagram; supplemental Figure 1).

Baseline characteristics and patient disposition

As shown in Table 1, median age of the cohort was 66 years (range, 29-90); 31% of patients were aged ≥70 years, and 59% of patients were male. Most patients were White (79%), with Black and Hispanic patients comprising 15% and 7% of the cohort, respectively. In this cohort, 89% of patients had an ECOG performance status of 0 or 1 at baseline. At least 1 clinically significant comorbidity was present in 77% of patients before ide-cel. Common comorbidities included pulmonary disease (33%); prior malignancies (19%); cardiovascular including cardiac disease (14%), arrythmias (12%), and heart valve disease (3%); hepatic disease (8%); and renal (6%) and cerebrovascular (5%) disease. Types of prior malignancies are described in supplemental Table 1. High-risk cytogenetics were present in 27% of patients. When including 1q abnormalities, high-risk cytogenetics were noted in 52% of patients with available data. EMD on latest scan before infusion was noted in 17% of patients with baseline imaging data. High bone marrow plasma cell (BMPC) percentage (>50%) was present in 14% of patients.

Table 1.

Baseline characteristics and treatment of patients who underwent SOC ide-cel

Characteristic	N = 821
Age, median (range), y	66 (29-90)
≥70, n (%)	251 (31)
Sex, n (%)
Male	486 (59)
Ethnicity, n (%)
Hispanic (any race)	55 (7)
Race, n (%)
White	652 (79)
Black or African American	120 (15)
Asian/Pacific Islander	24 (3)
American Indian/Alaskan Native	2 (0.2)
Other	23 (3)
ECOG performance status score, n (%)
0-1	728/770 (94)
2-4	42/770 (6)
Not reported	51
Myeloma subtype, n (%)
Oligo/nonsecretory	17 (2)
ISS disease stage, n (%)
I	210/420 (24)
II	142/420 (34)
III	68/420 (16)
Not reported	401
EMD, n (%)
Yes	85/488 (17)
Not reported	333
High marrow burden, BMPCs of >50%, n (%)
Yes	71/508 (14)
Not reported	313
Cytogenetic abnormality, n (%)
Any high-risk cytogenetics∗	196/727 (27)
High-risk cytogenetics including 1q	381/727 (52)
Not reported	94
Plasma cell leukemia (active or history), n (%)	13 (1.6)
Prior therapies
Number of prior antimyeloma therapies, median (range)	7 (4-21)
Prior autologous SCT, n (%)	683 (83)
Prior allogeneic SCT, n (%)	22 (3)
Prior anti-BCMA therapy, n (%)	121 (15)
Prior belantamab mafodotin, n (%)	113 (14)
Prior CAR-T therapy, n (%)	2 (0.2)
Prior BsAbs, n (%)	3 (0.4)
>1 BCMA-targeting therapies, n (%)†	3 (0.4)
Triple-exposed, n (%)	776 (95)
Penta-exposed, n (%)	490 (60)
Clinically significant comorbidity, n (%)	631 (77)
Cardiac disease	118 (14)
Arrythmia	99 (12)
Heart valve disease	22 (3)
Cerebrovascular disease	41 (5)
Hepatic	63 (8)
Pulmonary	269 (33)
Renal	48 (6)
Prior malignancy	152 (19)
Baseline cytopenia, n (%)
Hemoglobin, <8 g/dL	109 (13)
Platelets, <50 × 103/μL	96 (12)
Absolute neutrophil count, <1 × 103/μL	50 (6)
Bridging therapy	442 (54)
Lymphodepletion chemotherapy, n (%)
Fludarabine + cyclophosphamide	741 (90)
Bendamustine	51 (6)
Cyclophosphamide + other	19 (2)
Cyclophosphamide only	8 (1)
Not reported	1
CAR-T dose of >400 million, n (%)	331 (40)
Days from leukapheresis to infusion, median (range)	46 (24-501)

Note: when ≥5% data were missing for a given variable, denominator changed to patients with data available. Percentage totals may exceed 100% because of rounding.

Cytopenias were defined as hemoglobin of <8 g/dL, absolute neutrophil count of <0.75 × 10³/μL, and platelets of <50 × 10³/μL; high marrow burden was defined as >50% plasma cells before CAR-T therapy.

Triple-exposed disease: exposed to an IMiD, PI, and an anti-CD38 monoclonal antibody; penta-exposed disease: exposed to lenalidomide, pomalidomide, bortezomib, carfilzomib, and daratumumab or isatuximab.

SCT, stem cell transplantation.

^∗High-risk cytogenetics: includes del(17p), t(4;14), t(14;16), and t(14;20).

^†Includes n = 3 patients who received both BCMA CAR-T and belantamab.

The median prior lines of therapy was 7 (range, 4-21), 95% of patients were triple-class exposed and 60% of patients were penta-drug exposed. Prior anti-BCMA therapy exposure was noted in 15% of patients, most commonly with the antibody drug conjugate (ADC) belantamab mafodotin (14%, n = 113), whereas prior anti-BCMA CAR-T exposure (0.2%; n = 2), prior bispecific antibody (BsAb) exposure (0.4%; n = 3), or exposure to >1 type of anti-BCMA therapy (0.4%; n = 3) was less common. Median time from prior ADC or BsAb exposure was 4 months (range, 0-28), with 57% of patients receiving the last dose within 6 months of CAR-T therapy. Overall, 6 patients had received prior MM-directed CAR-T therapy, all with investigational products (target: BCMA = 5, CD138 = 1; none of the patients responded to the prior CAR-Ts). Median time from prior BCMA-directed CAR-T therapy was 20 months (range, 17-31) with all patients receiving CAR-Ts >6 months before SOC ide-cel. Prior engineered T cells against NY-ESO-1 or MAGE-A4 were administered in 2 patients, respectively.

Bridging therapy was administered in 54% of patients, and the regimens differed considerably (supplemental Table 2). Patients needing bridging therapy were more likely to have ISS stage II or III disease, EMD, penta-drug exposure, shorter time from diagnosis and more baseline cytopenias, and a trend toward more high-risk cytogenetics (supplemental Table 3).

Disease status at the time of CAR-T infusion was available for 817 of 821 patients in the study cohort. Overall, 62% had progressive disease, 17% had stable disease, and 21% had a PR or better. Patients who were in VGPR or better before ide-cel had a shorter time from diagnosis to CAR-Ts, lower baseline cytopenias, and less need for bridging therapy (possibly indicating response from prior line of therapy). There was a trend toward less patients with EMD and baseline BMPCs of >50% (supplemental Table 4).

The most common lymphodepletion regimen was fludarabine and cyclophosphamide (90%), followed by bendamustine (6%), cyclophosphamide + other (2%), and cyclophosphamide alone (1%). In general, patients in the bendamustine group were not enriched in high-risk features and there was no difference in prior myeloma-directed therapy and baseline comorbidities including moderate to severe renal dysfunction compared with fludarabine and cyclophosphamide. A higher proportion of patients in the bendamustine group were aged ≥70 years and with ECOG performance status of ≥2 (supplemental Table 5).

Response to therapy

The ORR in the cohort was 73%, including a VGPR or better rate of 56% and a CR or better rate of 26% (Figure 1A). Supplemental Tables 6 and 7 show response rates in different subgroups of patients. ORR and CR rates in patients aged ≥70 years 70% and 23%, respectively, and in patients with high-risk cytogenetics were 70% and 25%, respectively. ORR and CR rates in other clinically relevant subgroups were as follows: patients with ISS stage III disease (62% and 13%), ECOG performance status of ≥2 (60% and 10%), patients who had received prior anti-BCMA ADC or BsAb therapy (58% and 16%), prior anti-BCMA ADC (59% and 18%), and those receiving bendamustine as lymphodepletion therapy (43% and 10%), respectively.

Figure 1.

Efficacy with SOC ide-cel. (A) Response rate, (B) PFS, and (C) OS. sCR, stringent complete response.

On multivariable analysis, cell dose of ≤400 × 10⁶, bendamustine lymphodepletion, <6 months from prior anti-BCMA therapy, and stable disease or progressive disease status at the time of infusion were associated with inferior response rate. Presence of EMD was associated with inferior CR rate (supplemental Tables 8 and 9).

Survival outcomes

The median PFS of the entire cohort (N = 801) was 8.8 months (95% CI, 7.8-10.5). Estimated OS at 1 year was 67% (95% CI, 63-71). (Figure 1B-C). On subgroup analysis, as shown in Figure 2, patients with high-risk cytogenetics had inferior PFS (median, 7.6 vs 9.7 months; P = .007) and inferior OS (P = .02).

Figure 2.

Survival outcomes with SOC ide-cel in specific subgroups of patients. (A) PFS in patients with high-risk vs standard risk cytogenetics. (B) PFS in patients based on disease status at the time of ide-cel infusion. (C) PFS in patients based on fludarabine/cyclophosphamide vs bendamustine lymphodepletion. (D) OS in patients with high-risk vs standard risk cytogenetics. (E) OS in patients based on disease status at the time of ide-cel infusion. (F) OS in patients based on fludarabine/cyclophosphamide vs bendamustine lymphodepletion. BEND, bendamustine; FLU/CY, fludarabine and cyclophosphamide; PD, progressive disease; SD, stable disease.

Patients who needed bridging therapy had inferior PFS compared with those who did not (Figure 2). Regardless of the need for bridging therapy, there was a significant impact of disease status at the time of infusion on PFS and OS. Patients with VGPR or better at infusion had the best outcomes (median PFS not reached), followed by patients in PR (median PFS, 10.8 months), and, lastly, those with stable disease (median PFS, 8.8 months) or progressive disease (median PFS, 8 months; P = .009). A similar trend was noted for OS based on disease status, with patients in VGPR or better having superior OS (P = .01). Patients treated with bendamustine lymphodepletion had inferior PFS (median, 3.9 vs 9.1 months; P < .001) and OS (1-year OS estimate, 49% vs 68%) than patients receiving fludarabine and cyclophosphamide lymphodepletion (Figure 2).

Patients receiving prior continuous BCMA-directed therapy, primarily ADC (only 3 patients with BsAb), were observed to have inferior PFS and OS compared with patients who did not receive prior BCMA-directed therapy. The median PFS in patients receiving BCMA-directed ADC (and n = 3 BsAb) within 6 months of CAR-Ts, ≥6 months before CAR-Ts, and no prior BCMA-directed therapy were: 4.9 months, 5.9 months, and 9.7 months, respectively; P < .001 (Figure 3A). One-year estimates for OS in the 3 groups were 50%, 41%, and 70%, respectively (P < .001; Figure 3B). Amongst 5 patients who received prior BCMA CAR-T therapy, 2 patients had stable disease and 3 had progressive disease, all except 1 were alive at last follow-up visit. Median PFS in the 116 patients receiving belantamab mafodotin was 5.3 months (95% CI, 3.7-7.4) and estimated 12-month OS was 52% (95% CI, 41-63).

Figure 3.

PFS and OS based on receipt of prior BCMA-directed therapy, predominantly belantamab mafodotin (excluding prior CAR-Ts) before SOC ide-cel infusion. (A) PFS curve. (B) OS curve. Of the patients who had received prior BCMA treatment, 116 patients received belantamab, and only 3 patients received BsAbs.

Patients with ≥2 high-risk cytogenetics (defined as ≥2 of the following: deletion 17p, t(4;14), t(14;16), t(14;20), and chromosome 1 abnormalities) had inferior PFS and OS compared with those with 1 or no high-risk abnormalities (supplemental Table 10; Figure 3). PFS for each individual high-risk abnormality is described in supplemental Figure 4.

Factors having a significant association with PFS or OS on multivariable analysis are shown in the forest plot in Figure 4. Older age (aged ≥70 years) was a favorable prognostic factor for both PFS and OS. Adverse prognostic factors for PFS included baseline platelet count of <50 × 10³/μL, <6 months from prior continuous BCMA therapy, high-risk cytogenetics, ISS stage III disease, EMD, BMPC of >50%, ECOG performance status of ≥2, bendamustine lymphodepletion, and not achieving at least VGPR with bridging therapy (categories of PR, no response, or progression all were associated with inferior outcomes). Adverse prognostic factors for OS included baseline platelet count of <50 × 10³/μL, ECOG performance status of ≥2, prior BCMA therapy (both <6 months and ≥6 months prior), high-risk cytogenetics, ISS stage II or III disease, and elevated lactate dehydrogenase before CAR-T therapy.

Figure 4.

Multivariable analysis: forest plots showing factors with significant association. Forest plot showing multivariable analysis for factors having a significant association with (A) PFS and (B) OS after SOC ide-cel.

Safety outcomes

Table 2 describes safety outcomes. CRS was seen in 80% (grade ≥3: 3%, including 4 grade 5 events), with median time to onset of 2 days (range, 1-26). ICANS was seen in 28% (grade ≥3: 5%, including 1 grade 5 event) of patients, with median time to onset of 3 days (range, 1-33). Patients with grade ≥3 CRS (n = 24) and ICANS (n = 39) and had higher baseline tumor burden as evidenced by higher proportion of patients having BMPCs of >50% and ISS stage III disease, and for grade ≥3 CRS higher prevalence of plasma cell leukemia. They were more likely to have baseline severe cytopenias and poor performance status, and a higher proportion were in progressive disease at infusion, although this was not statistically significant. Patients with high-grade CRS were numerically enriched in certain comorbidities such as cardiac disease, renal dysfunction, and moderate to severe pulmonary disease (supplemental Tables 11 and 12).

Table 2.

Safety outcomes with SOC ide-cel

Event	N = 821
CRS, n (%)
Any grade; grade ≥3	657 (80); 24 (3)
Grade 1	470 (57)
Grade 2	155 (19)
Grade 3	9 (1)
Grade 4	11 (1)
Grade 5	4 (0.5)
Grade unknown	8 (1)
Median time to onset from CAR-T, median (range), d	2 (1-26)
Median duration, median (range), d	3 (1-36)
ICANS, n (%)
Any grade; grade ≥3	231 (28); 39 (5)
Grade 1	99 (12)
Grade 2	43 (5)
Grade 3	26 (3)
Grade 4	12 (1.5)
Grade 5	1 (0.1)
Unknown	45 (5.5)
Median time to onset from CAR-T, median (range), d	3 (1-33)
Median duration, median (range), d	4 (1-61)
IEC-HS, n (%)	9 (1)
Tumor lysis syndrome, n (%)	8 (1)
Clinically significant infection, n (%)	367 (45)
Bacterial	189 (23)
Viral	215 (26)
Fungal	20 (2)
Other, >1 pathogen type	78 (10)
Prolonged cytopenia beyond day 30, n (%)∗	228 (28)
Prolonged thrombocytopenia	197 (24)
Prolonged neutropenia	92 (11)
Stem cell boost, n (%)	10 (1)
TRM, n (%)	50 (6)
Death days 0-30 (all cause), n (%)	17 (2)
SPM, n (%)	33 (4)
SPM excluding nonmelanoma skin cancers	13 (1.5)
Myeloid neoplasms/acute leukemia	8 (1)

^∗Prolonged cytopenias: defined as nonrecovery of counts by day 30 to the following threshold among alive patients. Neutrophil recovery: neutrophil count recovery is defined as achieving a neutrophil count of >0.5 × 10³/μL for 3 consecutive days and the first day is the recovery day. Platelet recovery: defined as platelets of 20 × 10³/μL without platelet transfusion in the last 7 days.

On multivariable analysis, age of ≥70 years, female sex, and baseline platelets of <50 × 10³/μL were associated with higher risk of grade ≥2 CRS. On multivariable analysis, a higher likelihood of developing ICANS was noted for patients with baseline clinically significant comorbidities, older age (aged ≥70 and 60-69 years compared with <60 years), female sex, ECOG performance status of ≥2, and preinfusion platelets of <50 × 10³/μL (supplemental Tables 13 and 14).

There were no cases of ide-cel–related Parkinsonism. There were 3 patients with features that may suggest non-ICANS neurotoxicity, 1 patient each with hemifacial spasm, nystagmus and diplopia, and ataxia and stutter, respectively. IEC-HS was seen in 1% of patients.

Clinically significant infections were seen in 45% of patients. Infections were mostly bacterial (23%) or viral (26%), whereas fungal infections were rare (2%). The rate of prolonged cytopenias was 28%, with neutropenia at day 30 seen in 11% of patients and thrombocytopenia at day 30 seen in 24% of patients. SPMs were seen in 4% of patients, including myeloid neoplasms/acute leukemia in 1% of patients (supplemental Table 15). However, it is to be noted that the follow-up is short, and longer-term follow-up is needed to understand the full risk of SPMs after ide-cel. The most common SPMs were nonmelanoma skin cancers. There were no reported malignancies of T-cell origin. The median time from ide-cel infusion to SPM was 6 months (range, 0-27).

By data cutoff, 227 (28%) patients had died. The most common cause of death was myeloma progression (n = 172; 76% of deaths). TRM was seen in 6% of the cohort (n = 50), including 1 patient with both active myeloma and CRS; cause of death was not reported in 6 patients. The most common cause for TRM was infection (n = 20; including COVID-19, n = 4). Death from immune effector cell–mediated associated toxicity was seen in 5 patients, 4 of which were cases of CRS, including 1 case with both CRS and IEC-HS and 1 case with active myeloma and CRS. The fifth case was due to ICANS. Organ failure was the cause of death in 15 patients including cardiac (n = 5), pulmonary (n = 5), renal (n = 3), and or multi-organ failure (n = 2). Four patients died from SPMs including acute myeloid leukemia (n = 1) and myelodysplasia (n = 3). Other causes of TRM included intracranial hemorrhage (n = 3), tumor lysis syndrome (n = 1), transfusion-associated circulatory overload (n = 1), and gastrointestinal malignancy present before CAR-T infusion (n = 1).

Discussion

This is, to our knowledge, the largest study of patients receiving commercial ide-cel CAR-T therapy for R/R MM in the real world, with 821 patients with a median follow-up of almost 1 year from CAR-T therapy. We observed efficacy and safety comparable with the pivotal KarMMa trial, despite 77% of the real-world cohort having clinically significant comorbidities, many of which would have made these patients ineligible for the KarMMa trial. This was also a very heavily pretreated population (median, 7 prior lines of therapy) and 15% of patients had received prior BCMA-directed therapy, primarily ADC. ORR and CR rates of 73% and 26%, respectively, observed in our cohort are highly comparable with those reported in the KarMMa trial (73% and 33%, respectively) and previously reported real-world data for ide-cel (84% and 42%, respectively).^1,10 The median PFS of 8.8 months in our cohort also mirrored that achieved in the KarMMa trial (8.8 months; 128 patients)¹ and previously reported real-world experience in 159 patients receiving ide-cel (8.5 months).¹⁰ Median OS was 15.6 months.

A large cohort allowed us to conduct robust multivariable analyses to identify predictors of efficacy. Similar to other reports, prior BCMA therapy was associated with lower likelihood of response and inferior PFS and OS.^{10, 11, 12, 13} Most patients receiving prior BCMA therapy in our cohort had received the ADC belantamab mafodotin, and the use of prior BsAb or CAR-Ts was rare. Patients with prior CAR-T therapy were not included in the prior BCMA therapy group, because CAR-T therapy is a 1-time therapy, whereas others are continuous treatments, and we postulated that this may lead to differential impact on outcomes. Similar to prior reports, we observed that the need for bridging therapy was associated with inferior PFS, likely because of bridging being more commonly used in patients with an aggressive disease biology and those who were more heavily pretreated. Another finding from our study is the impact of disease status at the time of ide-cel infusion. Patients with VGPR or better at the time of ide-cel infusion had the best outcomes, followed by those in PR, and lastly those with stable or progressive disease. This is consistent with emerging data in the field, including from the KarMMa-3 trial, in which patients who had a reduction in the monoclonal protein with bridging therapy had better efficacy outcomes after ide-cel.^{14, 15, 16, 17} This remained consistent in multivariable analysis as well, with patients achieving PR, patients with no response, or those with progression each having worse outcomes compared with those achieving VGPR or better. This likely reflects both disease biology and treatment history, because patients with VGPR or better before CAR-Ts having had a shorter time from diagnosis to CAR-T therapy, less aggressive disease biology, and lower likelihood of baseline cytopenias. Ciltacabtagene autoleucel (cilta-cel) is the other commercially available CAR-T therapy for MM.^{18, 19, 20} Use of cilta-cel in lower disease burden state as an earlier line of therapy has also been reported to be associated with better outcomes, particularly a lower risk of delayed neurotoxicity.²¹ Taken together, these data highlight the importance of effective bridging therapy before CAR-T therapy in select patients to further improve its efficacy and potentially reduce toxicities.

In 2022, a national shortage of fludarabine necessitated the use of alternative lymphodepletion, and bendamustine was adopted by many centers extrapolating data from CD19 CAR-T therapy.²² Bendamustine was the most common (6% of patients) alternative lymphodepletion used in our cohort. We observed that bendamustine lymphodepletion was associated with inferior PFS (3.9 vs 9.1 months) compared with fludarabine/cyclophosphamide lymphodepletion. Prior data on outcomes with bendamustine lymphodepletion with BCMA-directed CAR-T therapy are conflicting. An initial report from 2 centers suggested similar efficacy with bendamustine lymphodepletion in patients receiving BCMA-directed CAR-T therapy, with known confounders having been adjusted using inverse probability of treatment weighting.²³ A larger cohort form a multicenter study reported inferior outcomes with bendamustine on univariable analysis but not on multivariable analysis.²⁴ This study is, to our knowledge, the largest cohort of bendamustine lymphodepletion reported to date, and median PFS of only 3.9 months suggests that fludarabine/cyclophosphamide should remain the SOC lymphodepletion until additional data are available. Unadjusted confounding factors may also account for dismal outcomes with bendamustine lymphodepletion in our study that may be difficult to elucidate in our cohort.

Consistent with previous data from CAR-T therapy and other immunotherapies in myeloma, high-risk cytogenetics, poor performance status, and ISS stage III disease were independently associated with inferior PFS and OS with ide-cel in our cohort.^{10,12,25, 26, 27} Presence of EMD was also associated with inferior PFS but not OS.^10,27 It is to be noted that only half of the cohort had baseline positron emission tomography–computed tomography scan data before CAR-T therapy to document presence of EMD. Interestingly, age of ≥70 years had a favorable association with PFS and OS. This finding is also consistent with prior data,^10,28 and may reflect a selection bias with only fit, older adults proceeding to CAR-T therapy whereas being more liberal in selection of younger patients. Regardless, this highlights that older adults deemed fit to proceed to ide-cel can have excellent outcomes, although risk of CRS or ICANS is higher in this group as discussed hereafter.

Despite a third of patients being aged >70 years and many having baseline cytopenias and clinically significant comorbidities, the safety profile of ide-cel was also comparable with the clinical trial data, and previously reported real-world data.^10,12,29 Grade ≥3 CRS and ICANS were seen in 3% and 5% of patients, respectively, and patients who experienced these high-grade events were more likely to have high tumor burden, more baseline cytopenias and evidence of disease progression at baseline, and worse performance status. Overall, the rate of high-grade CRS and ICANS in this real-world population is comparable with that of the KarMMa-1 clinical trial data (grade ≥3 CRS: 5% and grade ≥3 ICANS: 3%) and previously reported real-world experience (grade ≥3 CRS: 3% and grade ≥3 ICANS: 6%). Importantly, in this large cohort of patients, there were no reported cases of CAR-T–associated Parkinsonism, which has been reported with cilta-cel.^20,30 There were 3 patients with manifestations suggestive of delayed, non-ICANS–like neurotoxicity. TRM of 6% in this cohort is comparable with that of the KarMMa trial (7%, of which 3% attributed to ide-cel) and from the previously reported real-world US CAR-T consortium experience (5%). Clinically significant infections were seen in 45% of patients in our cohort compared with 34% reported with prior real-world experience, whereas grade 3 or 4 infections were seen in 22% of patients in the KarMMa-1 trial. Notably, clinically significant infections are not the same as grade 3 or 4 infections, limiting cross-study comparison. Our large cohort size allowed us to conduct a multivariable analysis to identify factors associated with grade ≥ 2 CRS and any-grade ICANS with ide-cel, a novel contribution of our study. Older age (≥70 years), female sex, and baseline low platelets were independently associated with increased risk of both grade ≥2 CRS and ICANS, whereas poor performance status and presence of clinically significant comorbidities were associated with higher risk of ICANS. As management of patients getting CAR-T therapy moves toward outpatient administration in the future, identifying such factors can help risk stratify patients who may still benefit from inpatient monitoring, as well as benefit from prophylactic strategies to mitigate immune-mediated toxicities with CAR-T therapy.

Strengths of our study include that this is, to our knowledge, the largest cohort of patients treated with ide-cel reported to date. Median follow-up of 1 year allowed us to capture short and intermediate term outcomes, but longer follow-up is needed for identifying risk of long-term toxicities. Our cohort size also allowed us to identify subgroups who fare favorably with CAR-T and identify subgroups in which outcomes are inferior, and such patients may benefit from future consolidative therapy approaches after CAR-T cell therapy. Missing data on date of last therapy and response to last line of therapy is a limitation that underestimated the true rate of bridging therapy. Data on refractoriness to prior therapy (after initial response) are not captured on CIBMTR forms, hence we could not report on proportion of patients who were triple-class refractory or penta-refractory. Data on manufacturing failure rate and patients not receiving ide-cel after apheresis for other reasons are also not available in the CIBMTR database. Data on minimal residual disease negativity were not available in most patients, and we are unable to report minimal residual disease–based outcomes. Prior exposure to BCMA-directed therapy included mainly patients who received BCMA-ADCs and not BsAb; further analysis with patients with prior BCMA BsAb are required to validate these findings. Despite these limitations, our data set provides important data on the real-world applicability of ide-cel.

In conclusion, ide-cel demonstrated a favorable safety and efficacy profile in the real-world setting, comparable with a pivotal registration trial population, highlighting the importance of a standardized approach to collect these data. This was seen despite patients having clinically significant comorbidities and having received extensive prior myeloma-directed therapy, including BCMA-directed therapy in a subset of patients. These results further support ide-cel as a therapeutic option for a broad, real-world application for patients with R/R MM.

Conflict-of-interest disclosure: S.S. reports research funding from Bristol Myers Squibb, Allogene, Janssen, and Novartis and consultancy for Bristol Myers Squibb, Janssen, Sanofi, Oncopeptides, Takeda, Regeneron, AbbVie, Pfizer, BiolineRx, Legend, and Roche/Poseida. A.A. reports research funding from AbbVie, Adaptive Biotech, K36 Therapeutics, Janssen, and Regeneron and an advisory role with Karyopharm, Bristol Myers Squibb, Sanofi, Janssen, and Pfizer. L.D.A. reports consulting/advisory board activity with Janssen, Celgene, Bristol Myers Squibb, Amgen, GlaxoSmithKline, AbbVie, BeiGene, Cellectar, Sanofi, and Prothena and research funding from Bristol Myers Squibb, Janssen, GlaxoSmithKline, and AbbVie. H.H. reports consulting/advisory board activity with Janssen, Amgen, GlaxoSmithKline, and Karyopharm. H.L. reports research funding from Alexion (AstraZeneca), Nexcella, Janssen, AbbVie, Prothena, Protego and consulting/advisory board activity with Alexion (AstraZeneca), Nexcella, Karyopharm, Arcellx, AbbVie, and Pfizer. C.L.F. reports honoraria from/consulting role with Bristol Myers Squibb, Seattle Genetics, Celgene, AbbVie, Sanofi, Incyte, Amgen, and ONK Therapeutics/Janssen and research funding from Bristol Myers Squibb, Janssen, and Roche/Genentech. The remaining authors declare no competing financial interests.