FDA cell and gene therapy approvals: evidence patterns behind recent licenses

Cell and gene therapies are licensed by FDA's Center for Biologics Evaluation and Research (CBER) as biologics under section 351 of the PHS Act. A defining feature of the class is that most products reached the market on the strength of small, single-arm trials, often with the patient's own pre-treatment history or an external natural-history cohort serving as the comparator. This overview groups the approved products by modality and sets out the pivotal evidence FDA relied on for each: the trial, its design, the number of patients, and the primary endpoint result.

Engineered T-cell therapies (CAR-T, TCR, and TIL)

These autologous products are manufactured by collecting a patient's T cells, genetically modifying them ex vivo, and reinfusing them. Every approval in this group rested on a single-arm trial with response rate as the primary endpoint.

Kymriah (tisagenlecleucel, Novartis) was the first CAR-T approved. Its pediatric and young-adult relapsed/refractory B-cell precursor ALL indication was supported by ELIANA (NCT02435849), an open-label, multicenter, single-arm trial with 63 patients evaluable for efficacy; the primary endpoint of complete remission or CR with incomplete count recovery (CR/CRi) within 3 months was 83% (52/63; 95% CI 71-91), all MRD-negative ⁸⁷⁸²⁷⁸. The subsequent DLBCL indication drew on JULIET (NCT02445248), an open-label single-arm trial in adults with relapsed/refractory DLBCL after two or more lines including rituximab and anthracycline, with a 68-patient efficacy-evaluable subgroup and ORR by independent review as the primary endpoint ²⁷⁸. More recently, FDA granted accelerated approval (2025-03-21) for relapsed/refractory follicular lymphoma after two or more lines, based on the single-arm Phase 2 ELARA study (E2202); complete remission rate by independent review was 69.1% (95% CI 58.8-78.3) ^{671672673676677669670}.

Yescarta (axicabtagene ciloleucel, Kite) received regular (not accelerated) approval on 2017-11-13 for adults with relapsed/refractory large B-cell lymphoma after two or more lines ^655661657663. The pivotal ZUMA-1 (KTE-C19-101) was a single-arm, open-label, multicenter Phase 1/2 trial; 108 patients were treated and 101 were efficacy-evaluable, with ORR as the primary endpoint. Independent-review ORR in the Phase 2 population was 72% (73/101; 95% CI 62-81) and investigator-assessed ORR was 83% (84/101) ^655660663664.

Tecartus (brexucabtagene autoleucel, Kite) received accelerated approval on 2020-08-12 for relapsed/refractory mantle cell lymphoma, based on the single-arm Phase 2 ZUMA-2 (KTE-102-C19) ³⁷⁷³⁹⁶. In the 60-patient inferential analysis set, FDA re-adjudicated ORR was 86.7% (95% CI 75.4-94.1) with a complete response rate of 61.7% ^{377387388389390}.

Breyanzi (lisocabtagene maraleucel, Bristol-Myers Squibb) gained accelerated approval (2025-03-21) for relapsed/refractory follicular lymphoma after two or more lines on the strength of the single-arm, multicohort Phase 2 study JCARO17-FOL-001; in the 94-patient efficacy population, independent-review ORR was 96% (95% CI 90-99) with 73% complete responses and a median duration of response not reached ^274645646.

Abecma (idecabtagene vicleucel, Bristol-Myers Squibb) was approved 2021-04-16 (traditional approval) for relapsed/refractory multiple myeloma after at least four prior lines including a proteasome inhibitor, an IMiD, and an anti-CD38 antibody. The pivotal single-arm study MM-001 used ORR (and stringent complete response) as the primary endpoint; both exceeded their pre-specified null rates of 50% and 10% respectively ⁶³¹.

Carvykti (ciltacabtagene autoleucel, Janssen) received regular approval on 2022-02-28 for relapsed/refractory multiple myeloma after four or more prior lines. CARTITUDE-1 was a Phase 1b/2, single-arm, open-label study; 97 of 113 enrolled patients received the product, and ORR by independent review was 97.9% (95% CI 92.7-99.7) ^{833848849834832}. A 2024 supplemental approval (2024-05-17) moved the product to an earlier line, again citing an ORR of 97.9% ³⁹⁹.

Aucatzyl (obecabtagene autoleucel, Autolus) was approved 2024-12-06 for adults with relapsed/refractory B-cell precursor ALL. Evidence came from the FELIX study (Phase 1b/2), with FDA basing its decision on complete remission within 3 months of infusion together with durability as the measure of clinical benefit ^593594595596.

Tecelra (afamitresgene autoleucel, Adaptimmune) is an engineered TCR T-cell therapy approved 2024-08-21 for adults with unresectable or metastatic synovial sarcoma that is MAGE-A4-positive and HLA-A*02-restricted, after prior chemotherapy ^591592803. The pivotal SPEARHEAD-1 (Cohort 1) was a multicenter, single-arm, open-label study; in 44 treated patients, confirmed ORR by independent review per RECIST v1.1 was 38.6% (17/44, all partial responses; 95% CI 24.4-54.5) ^803805193.

Amtagvi (lifileucel, Iovance) is a tumor-infiltrating lymphocyte (TIL) therapy that received accelerated approval on 2024-03-08 for unresectable or metastatic melanoma previously treated with a PD-1 antibody (and a BRAF inhibitor if BRAF V600-mutant) ⁴⁸⁷. The single-arm Phase 2 C-144-01 had a 156-patient safety set and an 82-patient efficacy set; objective response rate by independent review was 28.0% (23/82; 95% CI 18.7-39.1), with median duration of response not reached ^487889900888.

In vivo gene therapies

These products deliver a functional gene directly to the patient, most often via an adeno-associated virus (AAV) vector. The hemophilia programs share a distinctive endpoint strategy: a within-subject, non-inferiority comparison of annualized bleeding rate (ABR) after treatment versus the patient's own pre-treatment baseline.

Luxturna (voretigene neparvovec, Spark Therapeutics) was approved 2018-01-12 for biallelic RPE65 mutation-associated retinal dystrophy. Uniquely for the class, its pivotal Study 301 was a randomized, controlled, open-label Phase 3 trial (31 randomized: 21 treatment, 10 control). The primary endpoint, one-year change in multi-luminance mobility testing (MLMT), showed a median score change of 2 for treatment versus 0 for control (p=0.001) ³⁶⁹.

Zolgensma (onasemnogene abeparvovec, Novartis Gene Therapies) received regular approval (PDUFA date June 1, 2019) for spinal muscular atrophy in children under 2 with biallelic SMN1 mutations ⁶⁰²⁶¹⁸. The Phase 1, open-label, single-arm, dose-escalation trial AVXS-101-CL-101 treated 15 patients (12 at the therapeutic dose); in the high-dose cohort, 12/12 survived without permanent ventilation at 24 months and CHOP-INTEND motor scores rose from a baseline mean of 28.2 to 55.5 at 24 months ^111163612615.

Hemgenix (etranacogene dezaparvovec, CSL Behring) was approved 2022-12-29 (traditional approval) for adults with hemophilia B on Factor IX prophylaxis or with a history of serious bleeding ^407419416. The pivotal single-arm HGB study CT-AMT-061-02 (54 subjects treated) used ABR as the primary endpoint in an intra-subject non-inferiority design; mean ABR fell to 1.9 bleeds/year (months 7-18) from 4.1 during lead-in, an ABR ratio of 0.46 (95% CI 0.26-0.81) ^{409422425414426}.

Roctavian (valoctocogene roxaparvovec, BioMarin) was approved 2023-07-21 for adults with severe hemophilia A without pre-existing anti-AAV5 antibodies; approval was traditional, following a complete response letter noting the surrogate endpoint was inadequate for accelerated approval ^{519529535528520521}. The Phase 3 GENEr8-1 (Study 270-301) was open-label, single-arm, multinational (134 treated, 112 in the primary efficacy population); mean ABR fell from 5.4 to 2.6 bleeds/year, a change of -2.8 (95% CI -4.3 to -1.2), meeting the non-inferiority margin ^{524531536537528}.

Elevidys (delandistrogene moxeparvovec, Sarepta) first received accelerated approval on 2023-06-22 for Duchenne muscular dystrophy ³⁹⁷³⁹⁸. Notably, the Phase 3 confirmatory EMBARK (SRP-9001-301) study, a randomized, double-blind, placebo-controlled trial in 125 boys, did NOT meet its primary endpoint: the Week 52 difference in North Star Ambulatory Assessment total score was 0.65 (95% CI -0.45 to 1.74; p=0.244) ³⁹⁷³⁹⁸. This is an instructive example of a confirmatory trial failing to demonstrate a statistically significant functional benefit after an accelerated approval based on a surrogate (microdystrophin expression).

Beqvez (fidanacogene elaparvovec, Pfizer) received traditional approval on 2024-05-20 for adults with moderate-to-severe hemophilia B without neutralizing antibodies to the AAVRh74var capsid ^565572587. The Phase 3, open-label, single-arm Study C0371002 treated 45 patients; the ABR non-inferiority endpoint showed a model-derived ABR of 2.5 versus 4.5 bleeds/year at baseline (difference -2.1; 95% CI -4.8 to 0.7), within the pre-specified margin of 3.0 ^{588589587570571576}.

Kebilidi (eladocagene exuparvovec, PTC Therapeutics) received accelerated approval on 2024-12-11 for aromatic L-amino acid decarboxylase (AADC) deficiency, administered directly into the CNS ³⁷⁰³⁷⁶. The pivotal Study AADC-002 was single-arm (n=13; 12 analyzed) with an external natural-history control; the surrogate endpoint was a >20% increase in CSF homovanillic acid at Week 8, and 8 of 12 treated children (67%) achieved a new gross motor milestone by Week 48 ^{370376810821372373}.

Adstiladrin (nadofaragene firadenovec, Ferring) is a non-replicating adenoviral gene therapy approved 2023-01-12 (full approval) for BCG-unresponsive, high-risk non-muscle invasive bladder cancer with carcinoma in situ ⁵³⁹. The open-label, single-arm Phase 3 study rAd-IFN-CS-003 enrolled 107 patients with CIS; complete response at any time was the primary endpoint, achieved by 53.4% at the first (3-month) assessment (95% CI 43-63), with a median duration of response of 9.7 months ^{2021192425131415161718}.

Vyjuvek (beremagene geperpavec, Krystal Biotech) is a topical, redosable HSV-1-based gene therapy approved 2023-06-12 for dystrophic epidermolysis bullosa ^460461466484. Its pivotal Study B-VEC-03 used an unusual intra-subject randomized, placebo-controlled, double-blind Phase 3 design in which two matched wounds per patient were randomized to drug or placebo (31 subjects). Complete wound closure occurred in 20/31 (64.5%) treated wounds versus 8/31 (25.8%) placebo wounds, a difference of 38.7% (95% CI 13.9-63.5; p=0.012) ^{478484471483470472}.

Ex vivo gene-modified and gene-edited cell therapies

These products combine gene therapy with cell therapy: the patient's hematopoietic stem cells (or skin cells) are harvested, genetically modified with a lentiviral vector or CRISPR editing, and reinfused after myeloablative conditioning.

Zynteglo (betibeglogene autotemcel, bluebird bio) received traditional approval on 2022-09-08 for beta-thalassemia patients requiring regular red-cell transfusions ^{346826830823828829}. Two Phase 3 single-arm studies supported the BLA: HGB-207 (22 evaluable) and HGB-212 (14 evaluable), with transfusion independence as the primary endpoint. Transfusion independence was achieved by 20/22 (90%) in HGB-207 and 12/14 (85.7%) in HGB-212 ^{35718364365366367368}.

Skysona (elivaldogene autotemcel, bluebird bio) received accelerated approval on 2022-10-06 to slow progression of neurologic dysfunction in boys 4-17 with early, active cerebral adrenoleukodystrophy (CALD) ^{648652653654649650}. The single-arm Phase 2/3 Study ALD-102 treated 32 boys (26 in the primary analysis); Month 24 major functional disability-free survival was 88% (23/26; 95% CI 70-98) ^649650653654.

Lyfgenia (lovotibeglogene autotemcel, bluebird bio) received traditional approval on 2023-12-08 for patients 12 and older with sickle cell disease and a history of vaso-occlusive events ^705703696697. In Group C of the Phase 1/2 study HGB-206 (32 evaluable), the primary endpoint of complete resolution of vaso-occlusive events between months 6 and 18 (VOE-CR) was met by 28/32 (87.5%; 95% CI 71.0-96.5) ^{694700702708687}.

Casgevy (exagamglogene autotemcel, Vertex) is the first CRISPR/Cas9 gene-edited therapy, approved 2024-01-10 (traditional approval) for sickle cell disease in patients 12 and older with recurrent vaso-occlusive crises ^770781792785. The multicenter, open-label, single-arm Study 121 infused 44 patients (31 efficacy-evaluable); the primary endpoint of freedom from severe VOCs for at least 12 consecutive months (VF12) was met by 29/31 (93.5%; one-sided 98% CI 77.9-100) ^{776780781789787}.

Lenmeldy (atidarsagene autotemcel, Orchard Therapeutics) received traditional approval on 2024-04-09 for pre-symptomatic and early-symptomatic metachromatic leukodystrophy ⁷⁴⁵. Evidence integrated two single-center, single-dose pivotal studies (Study 201222 and Study 205756) and an expanded-access program against an external natural-history cohort (37 treated total). The primary endpoint, severe motor impairment-free survival, was dramatically better in treated patients: in the pre-symptomatic late-infantile analysis, 1/20 treated versus 28/28 untreated had severe motor impairment or death (log-rank p<0.001) ^{745768750756757}.

Zevaskyn (prademagene zamikeracel, Abeona) is an autologous, gene-corrected cell-sheet product approved 2025-06-10 for recessive dystrophic epidermolysis bullosa ⁷¹³⁷¹⁴. The pivotal Phase 3 EB-101-CL-301 (VITAL) was a randomized, intra-subject controlled study in 11 subjects with 43 matched wound pairs. At least 50% wound healing at Month 6 was achieved in 35/43 (81.4%) treated wounds versus 7/43 (16.3%) with standard of care (p<0.0001), with a co-primary pain endpoint also favoring treatment (p=0.0002) ^865871861863.

Other cellular and tissue-engineered therapies

Provenge (sipuleucel-T, Dendreon) was the first therapeutic cancer vaccine / autologous cellular immunotherapy, approved 2010-06-04 for asymptomatic or minimally symptomatic metastatic castration-resistant prostate cancer ²⁹³. Its approval rested on the randomized, double-blind, placebo-controlled Phase 3 trial D9902B, with overall survival as the primary endpoint (the broader program randomized 488 patients to sipuleucel-T) ^293684685.

Lantidra (donislecel, CellTrans) is the first allogeneic pancreatic islet cell therapy, approved 2023-07-31 (regular approval) for adults with brittle Type 1 diabetes uncontrolled despite intensive insulin therapy ^442443452453. The pivotal Phase 3 study UIH-002 (21 enrolled, 20 analyzed) was a single-arm, single-center study; the composite primary endpoint (HbA1c <6.5% and freedom from severe hypoglycemia at one year) was met by 8/21 (38.1%; 95% CI 18.1-61.6), supported by the Phase 1/2 study UIH-001 ^{434435444449452}.

Ryoncil (remestemcel-L, Mesoblast) is an allogeneic bone-marrow-derived mesenchymal stromal cell therapy approved 2025-01-07 (traditional approval) for steroid-refractory acute graft-versus-host disease in pediatric patients 2 months and older ^458155885151. The single-arm, multicenter Study MSB-GVHD001 enrolled 55 patients (54 treated); Day-28 overall response rate was 70% (95% CI 56.4-82.0) in the treated population ^458459881885.

MACI (autologous cultured chondrocytes on porcine collagen membrane, Vericel) was approved 2017-01-08 (full approval) for symptomatic full-thickness cartilage defects of the knee ^715721722. Unusually for the class, its pivotal SUMMIT study (MACI00206) was a randomized, controlled, open-label Phase 3 trial comparing MACI with microfracture (144 patients, 72 per arm; modified FAS of 128). The co-primary endpoints, change from baseline to Week 104 in KOOS Pain and KOOS Function (Sports and Recreational Activities), were both met ^{720726727728729730731}.

Stratagraft (allogeneic cultured keratinocytes and dermal fibroblasts, Stratatech) received traditional approval on 2021-07-06 for adults with deep partial-thickness thermal burns for which surgery is indicated ^490507494495. The Phase 3 STRATA2016 was an open-label, randomized, intra-subject controlled study (71 treated) in which each subject received Stratagraft on one wound and autograft on another. Only 4.3% of Stratagraft sites required autografting by 3 months, and 83.1% (59/71) achieved durable wound closure at Month 3 ^501512504514.

Cross-cutting themes for reviewers

Several patterns recur across the class. First, single-arm designs dominate: with the exception of Luxturna, Elevidys (confirmatory), Provenge, MACI, and the intra-subject controlled skin and burn products, pivotal efficacy came from uncontrolled trials benchmarked against pre-treatment history or external natural-history data. Second, the primary endpoint is modality-specific: response rate for oncology cell therapies, annualized bleeding rate for hemophilia gene therapies, and disease-specific functional or event-free survival endpoints for the neurometabolic and hematologic gene therapies. Third, accelerated approval is common where a surrogate stood in for clinical benefit (Kymriah-FL, Tecartus, Breyanzi-FL, Amtagvi, Skysona, Elevidys, Kebilidi), and the Elevidys confirmatory EMBARK result, which missed its functional primary endpoint, illustrates the residual uncertainty that accompanies that pathway ³⁹⁷³⁹⁸. For any individual product, the FDA review documents cited here contain the full benefit-risk analysis, the CMC and comparability assessments that are central to these manufacturing-intensive products, and the postmarketing requirements, all of which are natural follow-up questions to pursue.