In addition to the monthly themes, content on the following topics will be published throughout the year:
• Supply chain• Analytics• Business and finance• Manufacturing• Regulatory• Translational R&D
February
Preclinical & clinical development
Rethinking preclinical models: are large-animal toxicity studies still giving the data the field needs, or should more weight shift to human organoids and other new approach methodologies the US FDA is encouraging?
Matching the editing tool to the job: base editing, prime editing, and large-insertion tools such as CRISPR-associated transposases are extending reach from single bases to gene-sized cargo, for example the mutation-bearing exons of the DMD gene. Where does each platform earn its place?
Reading the first in vivo editing results: the Baby KJ CPS1 case offers a template, and positive Phase 3 data points toward a possible first in vivo gene-editing approval. Which elements of the n-of-1 model can scale, and which stay bespoke?
Designing trials where a control arm is contested: accelerated approval, single-arm designs, biomarkers, and natural history comparators are all in play for rare disease, and real-world evidence with a natural history comparator has supported a positive opinion. How should developers build a package regulators will accept, and where does a placebo arm remain unavoidable?
Building the evidence for a lifelong therapeutic: as in vivo trial readouts arrive and long-term follow-up accumulates, durability and modified-cell persistence become central to how CGT products are valued. What follow-up design and endpoints substantiate one-and-done claims over years rather than months?
Baking safety in from the first level: as deployment scales, rare adverse events are surfacing, from AAV immunogenicity to cytokine release syndrome and questions about double-strand-break editing. How should safety assessment be designed in early, and what best practice should be shared across the field?
Compressing timelines to first-in-human: manufacturing days are being cut but IND-enabling work is not, still running around two years when some patient populations do not have that time. What would it take to reach six months, and where are the scientific limits versus process inertia?
Connecting CMC, non-clinical, and clinical into one narrative: regulators want a cohesive story linking manufacture, preclinical data, and defined clinical endpoints. How should small developers structure regulatory strategy when CMC and preclinical teams each think narrowly?
Anchoring stem cell and iPSC-derived programs in credible preclinical evidence: with sources shifting and embryonic lines under policy pressure, what preclinical package supports lineage identity, tumorigenicity risk, and engraftment potential before these products reach patients?
Getting beyond the liver: systemically administered vectors still default to the liver, which is rarely where the target cells sit. How is the field redirecting tropism across viral and non-viral platforms alike, and is complete hepatic detargeting even desirable, given the role of liver exposure in tolerizing the construct?
Engineering capsids for the tissues that matter: what is the state of next-generation AAV capsid engineering for muscle, heart, retina, and the CNS, and how do rationally designed capsids compare with those from directed evolution campaigns in NHPs?
Confronting the AAV packaging ceiling: with large payloads such as Cas nucleases pushing past the 4.7 kb limit, how can developers avoid the costly late-stage redesign that follows oversizing, and where do dual and split-AAV strategies fit for large-gene indications such as Stargardt disease and Usher syndrome type 1B?
Reading AAV sentiment after the adverse events: confidence has rebounded despite prior safety setbacks, with committed developers staying in, though much of the negativity comes from those holding non-viral or biologics positions. Where does AAV hold its niche, and where should developers look elsewhere?
Extending non-viral delivery past the liver: no LNP formulation yet performs well in the brain, retina, or muscle. What delivery chemistries and targeting strategies will extend mRNA and LNP reach into the extrahepatic tissues currently accessible only to viral vectors?
Delivering genome-editing payloads: how is the field delivering transient editor formats such as mRNA and ribonucleoprotein rather than DNA to limit persistent expression, and carrying the DNA donor templates that larger insertions require?
Assessing emerging delivery vehicles: what is the translational potential of virus-like particles, exosomes, R2 retrotransposons, circular-RNA-generating AAV constructs, extracellular vesicle and AAV hybrids, and ultrasound-delivered plasmid DNA?
Compressing vein-to-vein time toward 24--48 hours: with direct-culture CAR-T protocols moving from 14 days to five to seven days, and one-day, non-activated processes now emerging, what process intensification and in-line control tools make these timelines reproducible at GMP standard rather than in a single academic setting?
Bringing CAR-T manufacturing to the bedside: can turnkey, standardized point-of-care workflows deliver a product in around seven days at roughly $20,000 all-in, and is decentralized production the realistic route to the eligible patients centralized models are not reaching?
Solving cryopreservation for the cells a process cannot afford to lose: freezing single cells is routine, but freezing iPSC-derived islet aggregates remains much more difficult and forces fresh-delivery models, while in CAR-T the open concern is whether freeze-thaw preferentially costs the most therapeutically valuable cells. What formulation and freezing tools close this gap?
Controlling starting material variability: apheresis material differs by patient, disease, and prior treatment, and healthy-donor material for allogeneic products brings its own variability. How can developers and their suppliers secure consistent cellular starting material and critical raw materials across autologous and allogeneic workflows?
Manufacturing beyond the CAR-T template: closed systems were built around autologous CAR-T, but TILs need feeder-driven expansion to far greater cell numbers, NK cells resist lentiviral transduction, and iPSC-derived products require differentiation control. Where can platforms flex across modalities, and where are dedicated processes unavoidable?
Balancing speed, potency, and safety in process design: shorter culture preserves less differentiated, more potent T cells, but compressed processes leave less room for in-process control and graft-versus-host risk management in allogeneic products. How should developers make that trade-off deliberately rather than by default?
Meeting the manufacturing demands of a different patient population: as CAR-T moves into autoimmune and solid tumor indications, treatment moves out of specialist cancer centers and toward outpatient rheumatology settings. What does that shift require of manufacturing scheduling, batch size, and turnaround?
Building tomorrow's cell therapy workforce: knowledge is lost at every handover, while hiring, training, and retaining skilled operators remains a top constraint. What training models and documentation practices keep processes reproducible as teams turn over?
Making the automation case: as closed, automated platforms promise to strip labor and cost out of autologous CAR-T manufacturing, how should developers scrutinize headline throughput claims, and where does automation improve unit economics rather than move the bottleneck?
Building data architecture before the science: some companies now spend months defining data formats and flows so everything is captured machine-learning-ready from day one. What does a data-first model look like in practice, and can established developers retrofit decades of legacy data?
Accelerating capsid discovery with machine learning: next-generation sequencing, DNA synthesis, and machine learning are cutting the number of selection rounds needed in AAV directed evolution campaigns in non-human primates. How far can in silico design replace physical rounds, and what is the equivalent route for LNPs?
Moving process development from wet lab to dry lab: which viral vector and cell therapy process development and CMC decisions can now be made computationally, and what does that shift mean for how CDMOs and tool providers structure their services?
Compressing release with real-time analytics: where can at-line monitoring, automated data capture, and digital batch records shorten multi-week release testing for CAR-T and other autologous products, and what evidence do regulators expect before automated methods replace manual release?
Digitizing chain of identity and custody: as CAR-T vein-to-vein times compress and production decentralizes, what orchestration systems keep every step traceable across scheduling, apheresis, manufacture, and return?
Automating the harder cell therapies: TILs, CAR-NK cells, and iPSC-derived products each bring process demands that CAR-T platforms were not designed for. Which automated systems can flex across modalities, and where are dedicated workflows unavoidable?
Assessing the next wave of tools: what is the translational potential of emerging approaches such as R2 retrotransposons, circular-RNA-generating AAV constructs, ultrasound-delivered plasmid DNA, and inducible gene switches?
Validating AI in regulated workflows: with machine learning now touching discovery, process control, and release, what documentation, validation, and explainability will regulators require, and how should developers build that in from the outset?
Making AAV economics work at population scale: systemic, CNS, and broad-population indications need far more material, so a tenfold gain in yield per liter can be wiped out by a tenfold rise in demand. Which upstream intensification and cost-of-goods levers move programs from serving a handful of patients to serving hundreds of thousands?
Moving AAV and lentiviral vectors onto producer and stable cell lines: can a shift away from transient triple transfection (for AAV) toward antibody-style stable-producer formats cut carried-over host cell and plasmid DNA, improve batch-to-batch consistency, and lower cost?
Pushing AAV full/empty capsid ratios toward 100%: empty capsids add immunogenic load without delivering payload. How are affinity and charge-based purification platforms, and newer approaches that assemble capsid and genome separately, raising the proportion of genome-containing particles without sacrificing batch yield?
Cleaning up the inputs: with expensive GMP-grade plasmid and transfection reagents driving both cost and variability, what is the role of synthetic and enzymatically produced DNA, and higher-quality starting materials more broadly, in improving consistency and shortening lead times?
Solving lentiviral vector purification and aggregation: large particles aggregate and will not pass a 0.2 μm filter, and purification steps can induce aggregation. What formulation additives and downstream strategies keep material filterable, and how should developers manage the purity versus recovery trade-off?
Taming LNP heterogeneity at scale: with particle-to-particle variability in size and component ratios (e.g. cholesterol content) undermining reproducibility, what mixing, formulation, and in-process control tools deliver homogeneous mRNA-LNP products at commercial volumes?
Formulating for the real world: most AAV still sits in off-the-shelf PBS inherited from academia, which constrains stability, storage, and reach. What formulation development is needed as gene therapies move toward common and ophthalmic indications?
Keeping process knowledge through handover: know-how leaks when programs move from academia to biotech, from preclinical to commercial, and through acquisition, while the trained workforce stays scarce. How can structured technology transfer and documentation keep a vector process reproducible when the people who built it move on?
"95% full by which method?": empty/full capsid ratio is both a moving target and method-dependent, with analytical ultracentrifugation, ion-exchange chromatography, and charge-detection mass spectrometry each returning different numbers for the same AAV material. How can the field align orthogonal methods so results are comparable across developers, sites, and regulators?
Cracking the potency assay: potency is the most product-specific test a developer will run and the top recurring topic at US FDA Office of Therapeutic Products town halls, yet existing guidance is least prescriptive here. How are teams building functional potency assays for CAR-T and viral vector products without a template to follow?
Linking vector integrity to potency: truncated and partial genomes lack efficacy but still register in many assays, driving potency variation run to run and manufacturer to manufacturer. What higher-resolution integrity and genome-content methods separate intact, functional AAV from material that only looks like it?
Proving the vector did what you think it did: quantifying transduction and on-target activity to a regulatory standard remains difficult for viral vectors, and n-of-1 and ultra-rare programs sharpen the need for molecular tools sensitive enough to read efficacy endpoints in very small patient numbers.
Release testing as the new bottleneck: one-day and seven-day CAR-T processes are facing multi-week release. Rapid methods, mycoplasma and endotoxin testing, next-generation sequencing for sterility, and fast phenotype and functionality readouts must keep pace.
Characterizing heterogeneous cell products: autologous CAR-T, TIL, and iPSC-derived products vary by patient, batch, and site. What multiplexed identity, phenotype, and functionality panels can demonstrate comparability across lots and manufacturing sites, when a batch cannot always be reproduced from the same starting material?
Defining phase-appropriate CQAs: critical quality attributes remain undefined and stage-dependent, shifting between first-in-human and late-stage development. How should analytical packages evolve across phases, and what are the risks of deferring analytical strategy until late?
Building the missing standards and harmonizing expectations: unlike monoclonal antibodies, CGTs have few reference materials or recognized standard tests, and US and European requirements remain unaligned. What shared standards would move the field toward a systematic analytical base?
Paying for one-and-done therapies: how can developers and payers move beyond a fee-for-service model built for chronic dosing, using outcomes-based approaches such as the US sickle cell models, and what price satisfies both risk-benefit and cost-benefit analysis when appetite for $3 million gene therapies is understandably low?
Bridging the gap between approval and access: with only around 25% of eligible US CAR-T patients treated and Europe's Joint Clinical Assessment stalling single-arm-approved products, what will it take for approval to translate into treated patients?
Rescuing transformative but unviable assets: Strimvelis, priced at €594,000 per patient, treated only about 45 ADA-SCID patients in total, leading to its transfer to a not-for-profit distribution model. Can marketplace models such as the CGT Exchange, not-for-profit and academic stewardship, and the European hospital exemption rehome scientifically sound ultra-rare programs?
Loosening the cold-chain constraint on reach: −80 °C storage works for gene therapies for small rare-disease populations but breaks down for common and ophthalmic indications, and for regions with little ultra-cold capacity. How far can better formulation widen geographic reach?
Reaching patients the centralized model cannot: can decentralized and point-of-care CAR-T production, and local manufacturing paired with wide distribution, extend quality-assured supply into markets central manufacturing has never served without weakening the regulatory oversight?
Building access across APAC: with the region now running more CGT trials than North America and in vivo CAR-T dealmaking concentrated in China, how do infrastructure, reimbursement, and regulatory readiness compare across APAC markets, and what does that mean for where products launch?
Making the marketing authorization fit the setting: as regulators pilot more flexible, hospital-based licensing routes for advanced therapies, supply is edging toward the point of care, though marketing authorization holder capabilities are not native to hospitals. Can managed access models widen reach?
Building patient trust: as advocacy grows into a force of its own, from families making high-stakes decisions to graduate-entrepreneurs running their own trials, and vaccine hesitancy bleeds into fears about altering DNA, how can the field engage patients as partners without talking down to them?
Rotating from rare to common: the field was built on orphan disease, where payloads were de-risked and payers willing to cover, but the largest healthcare-system value sits in common indications such as heart disease, wet age-related macular degeneration, and Parkinson's. What must change for developers to take on drug-target and disease-biology risk on top of delivery risk?
Rolling CAR-T out into autoimmune disease: with strong early data and near-term approvals expected, the rollout looks nothing like oncology, since rheumatology patients are not accustomed to inpatient regimens. How should developers plan for a different patient pathway, and how much further out is in vivo CAR-T for autoimmunity?
Cracking the antigens and tumors CAR-T has not: with satri-cel's Claudin18.2 approval in China opening the solid tumor door, why has success beyond CD19 and BCMA proved so hard, and what will it take in aggressive tumors such as ovarian and breast cancer, and in other unmet needs?
Positioning TILs and TCR-T therapies in solid tumors: as approvals establish a foothold, where do TIL and engineered TCR-T approaches offer advantages over CAR-T in solid disease, and what manufacturing and patient-selection realities limit how far they can extend?
Delivering to the tissues that gate new indications: cardiac, muscle, retinal, and CNS targets remain hard to reach, and CNS work is growing, including epigenetic approaches in preclinical models of prion disease. Which capsid and delivery advances are opening new indications?
Proving durable function from iPSC-derived replacement therapies: in Type 1 diabetes and Parkinson's, can hypoimmune engineering let allogeneic iPSC-derived products evade rejection without chronic immunosuppression, and can lineage control deliver cells that engraft and function long term?
Justifying the modality against simpler alternatives: families and clinicians often want durable standard-of-care options, not the newest tool, and CRISPR can be the wrong instrument where a straightforward AAV approach works. How should developers define the minimum viable intervention for a given indication?
Choosing indications where permanent intervention is warranted: where is an irreversible gene edit justified, given that a risk factor knocked out today could prove protective later, and what does that mean for indication selection as editing platforms broaden?
Scaling allogeneic and iPSC-derived products in the bioreactor: iPSCs scale far beyond primary cells, but aggregate culture brings shear, aggregate size, oxygenation, and CO~2~ removal problems. What bioreactor design and process control moves these platforms forward, and how much of the allogeneic scale-up question is clinical, given how few programs have cleared Phase 1?
Scaling out autologous workflows: CAR-T scale-out multiplies cost, facility footprint, and donor variability rather than diluting them. Which parallelization, batch-scheduling, and facility design strategies make hundreds or thousands of individual runs viable?
Defining what a platform is: the term covers everything from modular CAR-T production frameworks to standardized AAV vector backbones. What constitutes a platform, and how much development time and regulatory burden does standardization really remove?
Choosing between building and buying capacity: some developers have built in-house manufacturing citing an immature CDMO ecosystem, while external networks add variability and management burden. How should companies weigh that decision, and at what stage does it change?
Selecting a CDMO in a crowded market: the explosion of vector CDMOs has created a trust problem, with few commercial track records, strong providers oversubscribed while weaker ones sit on overcapacity, and consolidation underway. What diligence separates a partner that can deliver at commercial scale?
Integrating upstream and downstream for commercial readiness: what does end-to-end integration require, connecting upstream production, downstream purification, and real-time analytics and controls into a single validated flow rather than a chain of hand-offs?
Funding the process development that scale-up depends on: clients want every novelty in the CMC package but rarely fund the exploratory work needed to know what works, while modular operations serving many small clients strain cleanroom economics. How should the cost of getting a process right be shared?
Sustaining supply for low-volume, high-cost therapies: where patient numbers are small, conventional scale-up logic fails and dedicated capacity is hard to justify. What manufacturing and facility models keep ultra-rare programs supplied without the economics collapsing?
Proving it in vivo or ex vivo first? Momentum runs toward in vivo administration (drug-like logistics, no cell-processing lab, the patient as the site), yet many argue for ex vivo proof of concept first, since every cell can be characterized and once material is administered there is no retrieving it. Where does each approach fit?
Testing the in vivo CAR-T oncology case: does the rationale hold when the T cells being modified are already exhausted, and is autoimmune disease the more plausible near-term in vivo indication? No one expects in vivo CAR-T to displace autologous therapy soon, so where do the two coexist?
Retargeting lentiviral vectors for in vivo use: what is needed to direct lentiviral vectors to target cells in the body, and how is the field resolving delivery-specific liabilities such as CAR incorporated into the envelope redirecting vectors to tumor cells, producer-derived MHC driving alloreactive clearance, and complement inactivation in serum?
Targeting T cells with LNPs: can targeted lipid nanoparticles redirect the mature mRNA-LNP playbook away from the liver and toward T cells to enable in vivo CAR-T, and how far can that tropism extend to other cell types?
Raising the analytical bar for administered products: once a vector is given in vivo it becomes a parenteral drug product, demanding higher stringency and CQAs that are not tested today. How should analytical and CMC packages change as material moves from ex vivo use to direct administration?
Manufacturing for two different products:in vivo lentiviral vector production faces filterability, aggregation, and purity demands that ex vivo material does not, and changing components to enable targeting also changes the manufacturing profile. What does that mean for developers moving a platform across the divide?
Weighing immunogenicity and redosing:in vivo delivery exposes vectors and bacterially-derived editing machinery to the immune system, and immune privilege in the eye and brain is only partial. With transient immunosuppression and IgG-cleaving enzymes now being tested to enable AAV redosing, how much does that change the calculus?
Comparing the patient and health-system case:in vivo administration promises off-the-shelf logistics and broader reach, while ex vivo offers full product characterization before infusion. How do the two compare on cost of goods, site requirements, and realistic access?
Clinical readouts: what did the year's in vivo CAR-T and gene editing results establish, which durability and persistence signals held up in longer follow-up, and where did data fall short of expectations set at the start of 2027?
Approvals and regulatory decisions: which products cleared, which were rejected or delayed, and what did the year reveal about how far the US FDA, EMA, MHRA, and other regulators will flex the evidence bar for rare disease, including progress on platform and prior-knowledge pathways?
Manufacturing and analytics milestones: which advances moved from conference slide to routine practice, in AAV yield and purity, CAR-T vein-to-vein times, automation, and rapid release testing, and which did not?
Delivery and vector progress: how much closer did the field get to reliable extrahepatic targeting with engineered capsids and targeted LNPs, and which emerging vehicles earned attention over the year?
Financing, dealmaking, and business models: how did the funding climate turn, where did capital concentrate across modalities, did big pharma move any closer to the field, and did anyone answer what happens to revenue once a small rare-disease population has been treated?
Access and reimbursement: which coverage and pricing decisions set precedents, what changed in APAC's position as a development and launch market, and did the gap between approval and treated patients narrow at all?
The 2028 agenda: on the year's evidence, which questions now matter most, and where should developers, regulators, and solution providers focus next?