A Disease Burden Defined by Small Numbers and High Stakes
Gene and cell therapy has grown from a single FDA approval in 2017 to eleven marketed products by mid-2026, but the addressable populations remain narrow by design. Spinal muscular atrophy affects roughly 1 in 10,000 births. Hemophilia B affects about 5,000 patients in the United States. Sickle cell disease, the largest of the group at over 100,000 US patients, is still a fraction of any cardiometabolic indication. Relapsed or refractory large B-cell lymphoma and multiple myeloma, the CAR-T indications, involve tens of thousands of patients annually, not millions. The commercial logic of this therapeutic area is inverted from primary care pharma: low volume, extreme per-patient value, and a delivery infrastructure that looks more like a specialty surgical procedure than a prescription fill.
Three Distinct Mechanisms, Not One Drug Class
Gene and cell therapy is really three separate technology platforms wearing one regulatory label. In vivo gene replacement uses adeno-associated virus (AAV) vectors to deliver a functional gene directly into a patient's cells. Novartis' Zolgensma (onasemnogene abeparvovec) for SMA and CSL Behring's Hemgenix (etranacogene dezaparvovec) for hemophilia B fall here, along with Sarepta's Elevidys, an AAVrh74-based micro-dystrophin therapy for Duchenne muscular dystrophy.
Ex vivo cell therapy removes a patient's own T-cells, engineers them with a lentiviral or retroviral vector to express a chimeric antigen receptor (CAR), and reinfuses them after lymphodepleting chemotherapy. This is the CAR-T category: Novartis' Kymriah, Gilead/Kite's Yescarta and Tecartus, Bristol Myers Squibb's Breyanzi and Abecma, and Johnson & Johnson/Legend Biotech's Carvykti.
The third platform, gene editing, edits the genome directly rather than adding or replacing a gene. Vertex Pharmaceuticals and CRISPR Therapeutics' Casgevy uses CRISPR-Cas9 ex vivo on a patient's hematopoietic stem cells to reactivate fetal hemoglobin, approved for sickle cell disease and beta thalassemia in December 2023. Bluebird bio's Lyfgenia, a lentiviral (not CRISPR) approach for the same sickle cell indication, was approved the same month, giving physicians a genuine mechanism choice for the first time in this disease.
The Leading Marketed Products and Who Controls Them
Gilead's Kite division and Novartis remain the commercial anchors of CAR-T, with Yescarta and Breyanzi both expanding into second-line lymphoma and driving multibillion-dollar franchise revenue for Gilead and BMS respectively. J&J and Legend's Carvykti has taken meaningful multiple myeloma share from BMS's Abecma since its CARTITUDE-4 data showed superiority over standard triplet regimens, illustrating how fast share shifts even in a category with only a handful of competitors.
On the in vivo side, Sarepta's Elevidys carries the most reimbursement controversy of any gene therapy on the market: its accelerated approval in 2023 was followed by a 2024 label expansion to a broader Duchenne population, then by two patient deaths from acute liver failure in 2025 that triggered a partial clinical hold and intensified payer scrutiny. CSL Behring's Hemgenix and BioMarin's Roctavian, both aimed at hemophilia, have seen slower-than-expected uptake because physicians and patients are wary of one-time, irreversible interventions replacing decades-old, well-tolerated prophylactic regimens.
Why This Sector Doesn't Have a Patent Cliff, It Has a Capacity Cliff
Gene and cell therapies do not face the classic small-molecule patent-cliff dynamic because there is no biosimilar or generic pathway for a one-time, autologous, individually manufactured product. Instead the constraint is production capacity. Autologous CAR-T therapies require dedicated vein-to-vein logistics: a patient's cells are harvested, shipped to a manufacturing site (Kite's facility in El Segundo, Novartis' plant in Morris Plains, New Jersey, or Legend's site in Somerset, New Jersey), engineered over roughly two to three weeks, and shipped back, with turnaround delays directly costing lives in aggressive lymphomas. Manufacturing slot scarcity, not patent protection, has been the real limiter on volume growth since 2021.
AAV vector supply is its own chokepoint. Plasmid and vector production capacity is concentrated among a small number of CDMOs, including Catalent (now part of Novo Holdings), Thermo Fisher's Brammer Bio unit, and Lonza, and any single-supplier disruption can delay launches by quarters. Bluebird bio's near-collapse in 2024, which forced it to sell itself to private equity firms SK Capital and Carlyle for roughly $30 million in cash plus contingent payments, a fraction of its prior valuation, demonstrated that manufacturing overhead and slow commercial ramp, not IP erosion, can bankrupt a gene therapy pioneer even with three approved products on the market.
The competitive threat these platforms actually face is technological substitution rather than biosimilar entry. Allogeneic, off-the-shelf CAR-T candidates from companies including Allogene Therapeutics and Caribou Biosciences aim to eliminate the vein-to-vein bottleneck entirely by using donor-derived, gene-edited cells that don't require individualized manufacturing, which would undercut the economics of every autologous product on the market if clinical durability holds up.
Where the Pipeline Is Heading
The next wave is in vivo gene editing, skipping the ex vivo cell harvest altogether. Intellia Therapeutics and Regeneron's NTLA-2001 for transthyretin amyloidosis, delivered via lipid nanoparticle rather than AAV, showed durable knockdown of the disease-causing protein in mid-stage trials and represents a template for treating systemic diseases without lentiviral manufacturing at all. Beam Therapeutics is advancing base editing, a more precise successor to CRISPR-Cas9 nuclease editing, into sickle cell and alpha-1 antitrypsin deficiency programs.
FDA's Office of Therapeutic Products, spun out of the former Center for Biologics Evaluation and Research's gene therapy division specifically to handle the caseload, has signaled it will lean on regenerative medicine advanced therapy (RMAT) designations and platform-level reviews to speed applications for related constructs, an implicit acknowledgment that the review model built for individual biologics will not scale to the dozens of AAV serotypes and CAR constructs now in clinical trials.
The durable investment thesis in gene and cell therapy is no longer whether the science works, multiple approvals since 2017 have settled that question, but whether manufacturing platforms can be standardized enough to bring cost per patient down from seven figures toward something closer to complex biologics pricing, and whether allogeneic and in vivo approaches can outrun the operational fragility that has already sunk one first-generation pioneer.