Introduction
Ciltacabtagene autoleucel (cilta-cel; CARVYKTI), a BCMA-directed autologous CAR T-cell therapy with a distinctive biepitope targeting domain, has substantially altered the treatment landscape for relapsed/refractory multiple myeloma (RRMM). Since its initial FDA approval in 2022, its indication has expanded to adult patients who have received at least one prior line of therapy (including a proteasome inhibitor and an immunomodulatory agent) and are refractory to lenalidomide—a considerably broader population than originally approved 1. This review synthesizes pivotal trial data, updated follow-up analyses, postmarketing safety communications, and available real-world evidence up to April 2026 to provide a balanced assessment of cilta-cel's benefit–risk profile and practical considerations for clinical adoption.
Efficacy: Depth, Durability, and Survival Benefit
The efficacy of cilta-cel in heavily pretreated RRMM is exceptional by historical benchmarks. In CARTITUDE-1, a phase Ib/II single-arm study of 97 patients with a median of 6 prior lines of therapy, the overall response rate (ORR) reached 97.9% (95% CI, 92.7–99.7), with 82.5% achieving stringent complete response (sCR) at 27.7-month median follow-up. Median time to first response was approximately one month, with best response attained by 2.6 months. MRD negativity at the 10⁻⁵ threshold was demonstrated in 91.8% of evaluable patients 2. At approximately 5 years of median follow-up, about one-third of treated patients remained alive and progression-free for five or more years without additional myeloma therapy, and median OS was approximately 60.6–60.7 months, representing highly notable outcomes in this heavily pretreated setting 2. Sustained MRD negativity for ≥6 months correlated with a 27-month PFS rate of 73.0% and OS rate of 93.5%, underscoring the clinical significance of response depth 2.
In CARTITUDE-4, a randomized phase III trial of 419 lenalidomide-refractory patients after 1–3 prior lines, cilta-cel demonstrated superior progression-free survival versus standard therapy (PVd or DPd): median PFS was not reached versus 12 months (HR 0.41; 95% CI, 0.30–0.56; p < 0.0001), and CR/sCR rates were 74% versus 22% 3. Although early Kaplan-Meier OS curves showed an early imbalance, a prespecified second interim analysis with 33.6-month median follow-up showed a statistically significant OS benefit, and the FDA determined that the overall benefit of CARVYKTI continues to outweigh its potential risks for the approved use 56. Adjusted comparisons versus the prospective LocoMMotion real-world cohort further confirmed an 85% reduction in PFS risk (HR 0.15) and 80% reduction in mortality risk (HR 0.20), with clinically meaningful quality-of-life improvements 4.
Safety Profile: Acute and Delayed Toxicities
The safety profile of cilta-cel is substantial and requires structured management within experienced centers. The major toxicities are summarized in the table below.
| Toxicity | Incidence (Any Grade) | Grade ≥3 | Timing | Reversibility |
|---|---|---|---|---|
| Cytokine Release Syndrome (CRS) | 78% in CARTITUDE-4 and 95% in CARTITUDE-1; 84% across CARTITUDE-1 and CARTITUDE-4 | Grade 3–4: 3% in CARTITUDE-4 and 4% in CARTITUDE-1; 4% across both studies | Median onset day 7 | Usually resolves, but fatal or life-threatening CRS has been reported |
| ICANS | ~16–23% | ~2–3% | Median onset day 8 | Most recover; 2 fatal cases |
| Parkinsonism / movement and neurocognitive toxicities | arkinsonism occurred in 3% across CARTITUDE-1 and CARTITUDE-4; earlier CARTITUDE reports described a low but clinically important risk of movement and neurocognitive toxicity | Grade ≥3 parkinsonism: 2% | Median onset day 27–43 | Often persistent and potentially severe; fatal or life-threatening complications have been reported |
| Grade 3–4 neutrophil count decreased / neutropenia | Not applicable as an “any-grade” toxicity category when reported as Grade 3–4 laboratory abnormality | Grade 3–4 neutrophil count decreased was reported in 95% of CARTITUDE-4-treated patients | Post-infusion; prolonged or recurrent cytopenias may occur | Many patients recover, but recurrent Grade 3–4 cytopenias after initial recovery and cytopenias at time of death have been reported |
| Grade 3–4 platelet count decreased / thrombocytopenia | ~60–80% | Grade 3–4 platelet count decreased was reported in 47% of CARTITUDE-4-treated patients in the current label | Post-infusion; prolonged or recurrent thrombocytopenia may occur | Recovery is variable, and recurrent or persistent cytopenias require monitoring and supportive care |
| Infections | 57% across CARTITUDE-1 and CARTITUDE-4 | Grade ≥3: 24% | Variable | Severe, life-threatening, or fatal infections occurred; Grade 5 infections were reported in 5% of patients across CARTITUDE-1 and CARTITUDE-4 |
| Hypogammaglobulinemia | IgG <500 mg/dL in >90% | ~12% as AE | Post-infusion | Requires IVIG monitoring |
| IEC-EC (immune effector cell-associated enterocolitis) | Postmarketing (frequency unquantified) | Potentially fatal | Weeks–months post-infusion | Fatal cases (perforation, sepsis) |
| Secondary Primary Malignancies | 20 events in 16/97 patients | Hematologic and solid tumors have been reported; current boxed warning focuses on secondary hematological malignancies | Long-term | Variable; lifelong monitoring is recommended |
CRS occurred in 95% of patients in CARTITUDE-1 and 78% in CARTITUDE-4, was predominantly grade 1–2, and should be managed with supportive care, tocilizumab, or tocilizumab plus corticosteroids as indicated; prophylactic systemic corticosteroids should be avoided unless clinically necessary 1. Neurologic toxicities deserve particular attention. ICANS affects 16–23% of patients, with rare fatal events. A distinct movement and neurocognitive syndrome (IEC-PKS/parkinsonism-like) was observed in 5% of CARTITUDE-1 patients, with autopsy findings suggesting on-target BCMA expression in the basal ganglia as a potential mechanism 11. Critically, implementation of risk mitigation strategies—aggressive pre-infusion tumor debulking, early CRS/ICANS control, handwriting assessments for early detection, and monitoring beyond 100 days—reduced MNT incidence to below 0.5% in subsequent studies 7. A Mayo Clinic real-world cohort (n=235) identified post-infusion absolute lymphocyte count peak (ALCpeak) ≥3 × 10⁹/L as a practical predictor of delayed neurotoxicity (AUC = 0.838), with a negative predictive value of 98% below this threshold, enabling risk-stratified surveillance 16.
The October 2025 FDA safety communication added immune effector cell-associated enterocolitis (IEC-EC) as a new boxed warning following postmarketing reports of severe diarrhea, weight loss, gut perforation, sepsis, and death occurring weeks to months post-infusion. Treatment-refractory IEC-EC warrants exclusion of T-cell lymphoma of the gastrointestinal tract 517. Secondary hematologic malignancies (MDS, AML) have been reported across the CAR-T class, though attribution to cilta-cel remains confounded by extensive prior alkylator and IMiD exposure in this population 2.
Benefit–Risk Balance by Patient Subgroup
Net clinical benefit is most favorable in patients who are lenalidomide-refractory after 1–3 prior lines, or triple-class exposed/refractory after multiple lines, with ECOG performance status 0–1 and adequate organ function 115. High-risk cytogenetics, ISS stage III, plasmacytomas, and high tumor burden are associated with attenuated durability (shorter PFS and OS) despite high initial response rates, necessitating closer post-treatment surveillance and earlier intervention planning 2. Subgroup analyses by race in CARTITUDE-1 should be interpreted cautiously because of small patient numbers; available data underscore the need for equitable access, representative enrollment, and attentive follow-up across racial and ethnic groups 2.
Critically, prior BCMA-directed therapy significantly diminishes cilta-cel's efficacy: BCMA-naive patients achieved ORR of 92% versus 70% in BCMA-exposed patients in a multi-institutional real-world analysis, with worse outcomes when prior BCMA therapy was administered within the preceding six months—consistent with antigen escape 18. International Myeloma Working Group guidelines now recommend prioritizing BCMA-directed CAR-T before other BCMA-directed agents whenever feasible 18. Frail patients (ECOG >2), those with uncontrolled infection, active inflammatory disorders, or limited access to specialized centers face substantially higher absolute risks of treatment-related morbidity and mortality, for whom the benefit–risk calculus is less favorable 112.
Clinical Adoption Implications
Successful adoption requires specialized CAR-T centers with trained teams, 24/7 tocilizumab availability, ICU access, and standardized CRS/ICANS grading protocols. Intensive post-infusion monitoring is mandatory: daily for approximately 14 days, with ongoing surveillance for at least four weeks, during which patients must remain in proximity to the treating center 113. A 3–4 week manufacturing window necessitates individualized bridging therapy to maintain disease control and reduce tumor burden prior to infusion 115. Real-world claims data from 242 patients confirm that outpatient-first administration is feasible in selected patients—approximately one-third of outpatient recipients required no hospitalization within three months—without compromising CRS/ICANS rates or 6–12-month outcomes 14. This finding supports expanding access while maintaining safety.
The rapid community adoption of bispecific antibodies (growing from 4.7% of qualifying patients in 2022 to 72.9% by mid-2025 in the US Oncology Network) creates important sequencing pressures 19. As bispecific use increases, the proportion of patients arriving at CAR-T evaluation with prior BCMA exposure grows, potentially diminishing cilta-cel's efficacy. This underscores the urgency of early referral to specialized centers before performance status declines and before bispecific therapy compromises BCMA antigen expression or T-cell fitness 1819.
Current Clinical Interpretation
CARVYKTI represents an important advance in RRMM treatment, offering a single-infusion strategy that can induce deep and durable remissions, including treatment-free remission in a subset of patients with otherwise limited options. The clinical benefit is supported by phase III randomized evidence confirming OS benefit in earlier-line lenalidomide-refractory disease, and by robust real-world comparative data showing dramatic improvements over standard care in heavily pretreated patients 359.
Nevertheless, the toxicity burden is real and multidimensional: acute CRS and neurologic toxicities require institutional expertise; prolonged or recurrent cytopenias and hypogammaglobulinemia require sustained hematologic and immunologic support; and delayed risks—particularly IEC-EC and parkinsonism, both of which may be severe, life-threatening, or fatal—must be clearly communicated to patients and monitored long-term 51617. For appropriately selected patients—particularly those who are BCMA-naive, lenalidomide-refractory, have adequate functional reserve, and have access to qualified centers—the benefit–risk balance is generally favorable, consistent with the approved indication and available trial evidence. The key unresolved questions shaping adoption over the next 12 months include the true real-world incidence and optimal management of IEC-EC, optimal sequencing relative to bispecific antibodies and other novel agents, predictive biomarkers for severe neurotoxicity, and long-term secondary malignancy risk as survival extends 15. For hematology/oncology clinicians, the practical imperatives are early identification and referral of eligible candidates, individualized bridging strategies, structured toxicity mitigation pathways incorporating ALCpeak monitoring, and comprehensive patient counseling encompassing both the transformative potential and the managed risks of this therapy 1516.