State of the Industry 2026 Virtual Panel: Advanced Therapies (pt. 1)

Heikki Lanckriet (HL): There are definitely some strong headwinds. In the U.S., the administration’s stance on vaccines and cell therapy has been disruptive, leading to reduced investment in several modalities. Some areas are faring better — AAV in particular appears to be attracting more funding compared to last year — but mRNA has been hit hard. Many programs that relied partly on BARDA support have lost funding and stalled. On the other hand, we see more positivity in Asia, where investment across modalities remains strong, and Europe also continues to look relatively healthy. So, while the U.S. is pulling back, other regions are stepping up, forcing companies to realign their strategies accordingly.

Jean-Simon Diallo (JSD): From what I’m seeing, AAV-focused CDMOs are struggling, with clients pulling back. In contrast, lentiviral vectors are experiencing renewed momentum. At Virica, we’re also seeing a resurgence in animal health applications, with lots of new targets and activity. Human vaccines remain steady, though the pullback in mRNA is puzzling. Interestingly, that shift may benefit viral vectors, as there’s renewed interest in “old-school” approaches, like polyvalent influenza vaccines using established vectors.

HL: I agree with Jean-Simon on lentiviral vectors. Three years ago, lenti was out of favor, but now it’s rebounding, largely driven by renewed strength in cell therapy. The two trends are tightly linked.

Jason Slingsby (JS): We’ve also seen that lenti resurgence. At Oxford BioMedica, we worked for years on lenti for in vivo gene therapy, and now the field is catching up, especially with the potential for in vivo CAR-T. That shift from vectors being seen as tools to being recognized as drug products in their own right is important. It raises the bar for critical quality attributes, which is where our focus on downstream purification and yield improvement comes in. More broadly, affordability and reimbursement remain major challenges for advanced therapies. Oncology has historically been well supported, but rare diseases face steep barriers. Hundreds of patients used to be seen as a viable market; companies now expect thousands. I hope new delivery technologies will reopen commercial pathways for rare disease therapies that would otherwise be abandoned.

Almira Bartolome (AB): From the CDMO perspective, it’s been a difficult investment climate for at least the past two to three years, particularly for early-stage companies. Investors want data before they commit, but it’s very hard to generate that data without funding, especially before clinical trials. At the same time, we see increasing interest in new modalities, like non-viral delivery and exosomes. On the cell therapy side, pharma is taking the space more seriously, with acquisitions and licensing activity picking up. That’s partly because advanced therapies are more difficult to replicate and less vulnerable to biosimilar competition. Despite the tough environment — exacerbated by geopolitical uncertainty and mRNA setbacks — there are still bright spots.

HL: The funding climate is a major factor. Three years ago, capital was flowing into preclinical startups, but that’s no longer the case. Investors and pharma are focusing instead on late-stage assets that can generate revenue as patents expire. This leaves preclinical and early phase I programs stuck in a “graveyard zone,” where it’s extremely difficult to raise capital and advance development.

JS: In Europe, we’ve traditionally looked to the FDA for regulatory leadership. Now, however, there’s room for other competent authorities to step forward. For instance, the MHRA recently enabled bench-side CAR-T treatments, while the FDA has also taken constructive steps, such as dropping the REMS requirement for CAR-T products — a data-driven move that will improve access. So, while there are real headwinds, especially in areas like mRNA and vaccines, this moment also creates opportunities for agencies like the EMA to show more initiative.

JS: It is a tough balance — you want leadership, but you also need harmonization. The U.S. remains the dominant market, so developers naturally gravitate toward its regulatory direction. But there are cycles of disruption and realignment, and perhaps in 6–12 months we’ll see calmer waters. Ultimately, harmonization is what drug developers need, because most of us want to operate in the major territories and avoid duplicating efforts.

AB: I’m not a regulatory expert, but I can share what I’ve observed. At a conference earlier this year as the new U.S. administration was coming in, there were discussions between EMA and FDA stakeholders about pilot programs in genetic medicines. So, collaboration is happening. But resources are also a concern, especially with layoffs at the FDA. Reputation also plays a big role. Take the Sarepta situation: the handling of it, including leadership resignations and returns, created confusion and uncertainty. Ultimately, regulatory agencies exist to protect patients and ensure therapies are safe and effective. But when uncertainty and reputational issues surface, it undermines confidence across the field.

HL: Our client conversations haven’t really been affected by regulatory or funding changes. What we are seeing, though, is a much stronger interest in U.S.-based manufacturing. Right now, our facilities are in the UK, but more clients — both pharma and biotech — are asking about onshoring supply chains to the U.S. as a hedge against future tariffs or trade restrictions. That trend is definitely shaping how we think about growing the business.

JSD: Tariffs are also a real concern. In some cases, we’re looking at potential tariff rates of 60% or more, which can be prohibitive. The uncertainty makes it worse — you don’t know from one day to the next how things will shift. For customers, it raises the uncomfortable question of who absorbs those costs. It’s not that working with U.S. clients is difficult in itself, but the unpredictability around tariffs creates a lot of hesitation and some very tough conversations.

AB: I wouldn’t call myself a regulatory expert, but I’ve followed the Sarepta situation closely. The approvals were controversial from the start, with disagreements among stakeholders and regulators. That said, accelerated approvals exist for a reason: to address unmet needs where patients simply don’t have time to wait. I don’t think one case should call that pathway into question. Nonetheless, it’s important to re-examine safety, particularly off-target effects. New technologies, such as synthetic capsids and organ-targeted delivery platforms, could help address these issues by improving specificity, lowering doses, and reducing risks. So, while this case is concerning, it may also accelerate innovation in safer delivery methods.

JS: Lower dosing is the long-term goal for gene therapies. With AAV vectors in particular, understanding the product’s nature is critical. There are other Duchenne programs in the pipeline with promising phase III data, and the field will benefit from a “design–build–test–review” cycle that steadily improves vectors, purity, and trial design. Gene therapy has faced safety concerns before, whether insertional mutagenesis risks with lenti in CAR-T, or going back decades to the Jesse Gelsinger case. None of these diminish the seriousness of Sarepta, but the field advances incrementally, with engineering-minded improvements that ultimately enhance safety and outcomes.

JSD: The Jesse Gelsinger incident happened before gene therapy was established and had a chilling effect for years. Today, the context is different: we already have successful commercial products with clean safety records. To me, this is a blip. Investors may overreact in the short term, but the science will recover. One new success story could restore confidence quickly.

HL: However, I did find it strange that Sarepta so quickly announced a pivot from AAV to siRNA after the setback. It felt like a knee-jerk reaction, and it risks feeding negative sentiment toward viral vectors.

JSD: Meanwhile, Big Pharma isn’t retreating — Eli Lilly and AstraZeneca are actively investing in gene therapy. Sarepta’s pivot may reflect internal issues more than an industry-wide signal.

AB: For AAV, I don’t think we’re facing a capacity crunch. If anything, there’s excess capacity following several years of consolidation. Larger organizations absorbed many smaller players, and now some have divested again. That’s not necessarily bad: advanced therapies require specialization, not just equipment but the drug-specific expertise and know-how to handle complex modalities. Specialization is becoming more important than simply having more capacity.

Overcapacity may also have been overestimated because forecasts were based on imperfect assumptions. Many gene therapies, even when they reach commercial stage, face reimbursement barriers that limit adoption. At the same time, the manufacturing methods those forecasts relied on were not always robust or efficient, so capacity was likely overstated.

One area that remains a real pain point is plasmid production. There are plenty of providers, but reliable, high-quality plasmids are still hard to secure. I recently came across data suggesting that around 31% of clinical materials tested contained inaccurately designed plasmids, so there is clearly a need for higher standards and better plasmid manufacturing.

Finally, new modalities like exosomes are another challenge. There aren’t many players with real capabilities, and much of the technology is now concentrated in the hands of a single large CDMO. That creates barriers for smaller companies that can’t access or afford those services.

AB: In terms of raw numbers, there’s capacity, but much of it is reserved for internal use. At 3PBIOVIAN, our plasmid platform was originally built to support our own AAV work, though we do take on some external projects. The real issue is reliability. We’ve had many clients come to us after poor experiences elsewhere. Cost is another challenge: plasmids can account for 30% of the cost of an AAV batch. We’re seeing growing interest in synthetic DNA as a way to reduce costs, even if CDMOs aren’t producing it themselves. Many are starting to test it in their platforms.

HL: We see strong momentum toward moving away from plasmids altogether. Synthetic DNA avoids bacterial sequences, improves consistency and safety, and the data supporting it are becoming increasingly compelling for both developers and CDMOs. More broadly, though, I’d say there’s overcapacity across the industry — not just viral vectors but also mRNA. Many facilities were built on pandemic-level demand assumptions that don’t reflect today’s reality. Outside of COVID-scale vaccines, most programs are preclinical or early phase I, where material needs are far smaller. The mismatch means high operating costs, which inflate batch prices. I expect we’ll see consolidation, closures, and opportunities for private equity to acquire undervalued assets and rebuild more sustainable platforms.

JSD: Cost is the biggest driver. Plasmids are expensive, and clients come to us looking for ways to get more out of each batch or reduce the amount of plasmid needed. Standard plasmids also carry elements that may trigger antiviral responses, which creates further inefficiencies. From my perspective, overcapacity is widespread and one of the root causes of the industry’s current turbulence.

JS: On the downstream side, we’re seeing increasing demand for solutions tailored to synthetic capsids and retargeted vectors. Lenti was dominated by VSV-G for years, but now there’s far more diversity — Nipah and other envelopes, for example. The problem is that traditional downstream methods, like ion exchange chromatography, don’t always work well with these new designs. If upstream titers improve dramatically, as with Virica’s enhancers, but downstream yields remain only 30%, the cost savings from plasmid efficiency are lost. The challenge is that manufacturing is risk-averse: everyone wants improvements, but few want to be the first to adopt them.

JSD: Another way companies are tackling plasmid costs is through stable producer cell lines. It’s already common on the lenti side and is starting to gain traction for AAV, though it’s more challenging there. Still, the trend is clear — many are looking to avoid plasmids entirely.

HL: I believe we’ll get there, but it won’t be because of a single breakthrough. Progress will come from steady, iterative improvements across multiple fronts. On the upstream side, synthetic DNA and novel capsids are already helping. Downstream, affinity ligands and other purification advances are making a difference. Host cell engineering is another area where productivity is improving. Each of these developments sharpens the overall process. Taken together, they’ll lead to purer, more potent, and more cost-effective products, making smaller patient populations a more viable target again over time.

JSD: At the end of the day, need drives adoption. If a client really needs a solution, that’s when they’ll take the risk. Some biotechs have that innovator mindset, and those make great partners to break through first. Over time, reputation lowers the barriers. CDMOs can be both a gateway and a challenge: once you have proven success, they can spread adoption across their platforms, but they usually require client buy-in before trying anything new. With today’s overcapacity, CDMOs are under pressure to differentiate, creating more willingness to adopt.

HL: CDMOs like to evaluate new tools, but they rarely act as champions. The push has to come from therapy developers. Strong data packages that show you’ve solved a real pain point are essential. Once a client pushes a CDMO to adopt your technology, then the CDMO takes it seriously. For us, that meant starting directly with drug developers and letting their demand drive adoption.

JS: One effective strategy is to seed technologies early in academic and R&D settings. If innovators already use your tool, they’ll go to a CDMO and say, “This is how we’ve always done it.” That makes adoption much easier. But you need those early champions. And getting into the clinic is pivotal: once you’ve proven yourself in one modality or jurisdiction, others will follow, though you often need separate “firsts” in the US, Europe, Japan, and so on. Some segments, like gene editing, tend to be more open to pushing boundaries, which creates earlier opportunities.

AB: From a CDMO perspective, adoption usually starts with a specific client need. Smaller, more collaborative CDMOs can sometimes take initiative, especially through grant partnerships where we can test promising technologies that address real platform gaps. The real challenge is the regulatory environment — no one wants to touch a tool that hasn’t been GMP tested. Personally, I see enormous room for improvement in our processes and welcome those conversations, but the path is always slow and careful.

JSD: Regulators could play a bigger role in lowering barriers. Even a preliminary signal from them — a kind of “in principle” greenlight — would help companies and CDMOs feel more confident about adopting new technologies. The challenge is you can’t usually get regulators’ attention unless you already have a sponsor, which creates that chicken-and-egg dynamic. The FDA’s CATT program, which allows applications for platform technology designation, is a good step, but it’s difficult to access, and the guidance is limited. I’d like to see more regulators take leadership here. Otherwise, innovators keep running into the same roadblock when asked, “Has this been in the clinic yet?” and the answer at the beginning is inevitably “no.”

HL: The MHRA process has one upside: while it can be painful to get a GMP license because of the scrutiny and upfront cost, once you have it, it’s a strong stamp of approval. Clients in other jurisdictions see that as meaningful validation, even if it’s not quite the same as having a clinical sponsor. For us, getting that license this year has already eased regulatory objections and built confidence with clients.

JSD: I think it’s not only crisis; it’s also about opportunity. Non-dilutive funding is key here. Grants can incentivize adoption by linking product development with innovative technologies. For example, if a developer applies for funding to support their manufacturing, they can increase their chances by incorporating a novel technology into the plan. That way, when they reach their milestones, the product and the technology are validated together. This is a simple but powerful way governments and nonprofits can help derisk adoption and make innovators more willing to take the plunge.

AB: In Europe, there are multiple grant pathways that can play this role. EU-wide grants are highly competitive, but individual countries also offer strong programs. For example, Innovate UK has very industry-oriented funding, and in Finland there’s the Business Finland grant program. These often tie adoption of new technologies to regional economic development like job creation. That’s especially important now, since early-stage VC funding is much harder to secure than it was a few years ago.

JS: I’d also highlight the role of strategic partnerships with Big Pharma. At Oxford BioMedica, I saw firsthand how a collaboration with Novartis around Kymriah transformed lenti manufacturing. Novartis invested in process development and regulatory strategy, enabling the shift from adherent cell factories to serum-free suspension bioreactors and massively debottlenecking vector supply. Those kinds of partnerships, where a pharma partner is willing to co-invest deeply in process and quality, can move the whole field forward.

We also saw this with AstraZeneca during COVID-19. Their work with the University of Oxford enabled over 20 manufacturers worldwide to produce the SARS-CoV-2 vaccine. That was a remarkable technical and organizational feat. It shows what’s possible when the right level of commitment and scale comes from pharma.

AB: In the last couple of years, we’ve seen more out-of-the-box models. At 3PBIOVIAN, being midsize gives us flexibility in contracting, like smaller, modular work packages. We’re also seeing more risk-sharing in production agreements. In some cases, I’ve even seen equity-based arrangements between CDMOs and biotechs, though those are tricky given our expertise lies in manufacturing, not drug mechanism evaluation. For viral vectors specifically, feasibility studies have become the standard entry point: often priced at cost or below, to derisk things for sponsors and help them evaluate both the technology and the CDMO relationship before committing further. Overall, there’s more negotiation around pricing than we saw before, and sponsors are taking advantage of a more competitive, overcapacity market.

HL: In the synthetic DNA space, we haven’t seen much change. Synthetic DNA is still a highly specialized space with very few global suppliers, especially at GMP scale. That insulates us somewhat from the pricing pressure that CDMOs downstream are facing.

JSD: For us, it depends on whether we’re providing products or services. On the services side, especially for larger projects, clients increasingly expect us to share risk upfront, tying outcomes to payment milestones. That wasn’t the norm a few years ago. On the product side, we’ve leaned into the RUO model, “kitifying” our technology so it’s easier to evaluate quickly and at low cost. That grassroots approach helps seed adoption and build credibility before moving into more expensive GMP contexts.

JS: For new manufacturing technologies, adoption hinges on proving you can save time, reduce cost, or mitigate risk. The last one is the toughest because it requires customers to trust something unproven. We’re focused on demonstrating time and cost savings with strong data. We also see value in making our technology accessible at small scale, whether in research or process development labs. If innovators can get comfortable with it early, they’re more likely to carry it forward into GMP manufacturing. Keeping the barrier low for creative people is the best way to build momentum.

HL: It really depends on the application. For vaccines, it’s already proven. The same goes for some therapeutic proteins where location isn’t critical, or in blood cells like T cells and NK cells where there’s more control. Those are the low-hanging fruit where non-viral approaches are gaining traction. But when it comes to targeting specific organs, things get much harder. Progress is being made, but it’s still a long road. Overall, I see meaningful opportunities in non-viral delivery, particularly in areas where precision targeting isn’t essential.

JS: I’m a fan of simplifying and taking a reductionist, chemical-first approach. Lipid nanoparticles, for instance, are improving in their ability to retarget specific cell types and organs. But viral vectors still have built-in advantages, like the integration machinery in lenti and gamma-retro systems. For some therapies, like CAR-T, you want long-lived engineered cells, so integration is valuable. Many in the field would love to move away from viral vectors, but evolution has had hundreds of millions of years to perfect viruses. We shouldn’t underestimate that head start.

JSD: Viruses are clever, and there are limits to how far we can go by trying to engineer around them. For vaccines, mRNA is a great tool, but it’s not always sufficient. Some pathogens, like HIV, demand a robust T cell response that we still struggle to achieve. For systemic therapies, you face saturation problems: you need to get gene cargo everywhere, which is tough to do better than evolution has managed. There’s definitely promise, but I think we need to stay humble about the challenges.

JS: Advanced therapies today are tiny today compared to where they’ll be in 20 or 30 years. The market could be 10, 20, even 40 times bigger, and when that happens, there will be space for many different modalities. Viral vectors will dominate in some areas, but there will be niches where non-viral methods carve out meaningful roles. This won’t be a winner-takes-all scenario — it’ll be multiple winners, each serving different patient needs.

AB: Viruses are expensive, but they’re also effective, especially for monogenic diseases. There’s no single solution that will work for everything. Different technologies will coexist, and CDMOs like ours will adapt by building expertise in both established and emerging platforms.

JSD: AI is the big one. In our case, high-throughput screening means we generate a massive amount of data, which is perfect for AI to mine for insights. The challenge is that as you move into larger-scale manufacturing, the number of iterations drops dramatically. Fewer runs make it much harder to apply AI meaningfully. Collaboration and data-sharing would solve this, but in practice, most groups still work in silos. Without pooling that data, AI’s potential in process development will remain limited.

JS: I’d highlight the value of process analytical technologies (PAT). Right now, most viral vector manufacturers rely heavily on static batch records. Operators follow instructions, but the critical decisions — like the exact harvest time — are often based on small-scale studies done months earlier. If we had real-time, inline analytics monitoring vector production, we could tailor processes much more precisely, improving yields and consistency. It requires investment and integration, but the payoff could be transformative.

JS: Not yet. People in the monoclonal space are already applying inline mass spectrometry and other real-time analytics. Viral vector manufacturing is starting to borrow from that playbook, with perfusion technologies now making their way in. We still lack a standardized downstream equivalent to monoclonal Protein A, which limits scalability. But the lesson from antibodies is clear: standardization plus economies of scale drive costs down. With vectors, we’re not there yet, but as products mature and demand grows, the same transition is possible. The Covid response already showed that the biggest lever for cost reduction was facility utilization — running clean rooms continuously, at full tilt. That model could apply again once there are enough commercial products to justify it.

AB: Four or five years ago, platforms were all the rage. Then people stopped talking about them, because every AAV project needed significant customization anyway: different serotypes, cell types, tweaks at every step. More recently, platforms are back in discussion, but with a more pragmatic framing: they’re starting points, not complete solutions. That’s why feasibility studies are so important. Sponsors want to test compatibility before making big commitments. Another complication is licensing. HEK293 is standard, but depending on which derivative you use, you may need to pay multiple license fees. That fragmentation in the supply chain feeds into the high cost of these therapies.

JSD: I’d add that the licensing issue is actually pushing people away from standardization. We hear constant calls for “industry-standard platforms,” but when market leaders charge steep fees for their platforms, VPC2.0 in AAV MAX for example, companies look for workarounds. Many are developing their own HEK lines specifically to avoid licensing costs. That behavior fragments the field even more. It’s ironic — we know standardization would help the industry scale, but the economics incentivize diversification instead.

JS: Technology could help in unexpected ways. For example, moving from ex vivo CAR-T to in vivo CAR-T would massively reduce supply chain, logistics, and manufacturing burdens, with cost and environmental benefits. On the other hand, the industry has embraced single-use disposables to the point where anything else seems naïve. But once you throw them away, they’re gone. In antibody manufacturing, you see reuse, like regenerating Protein A columns, so maybe similar innovations could help us reuse equipment in advanced therapies. It would take regulators and clients coming on board, but there’s precedent elsewhere in biopharma to follow.

JSD: The sheer volume of plastic waste in this industry is staggering. We brought in a sustainability consultant, and the statistics were shocking. I’d love to see us move away from single-use systems in a way that’s safe, but right now it feels entrenched. It’s hard to reconcile that with wanting a greener future for our kids.

AB: One area where we’ve made progress is energy use. At 3PBIOVIAN, our new facility is fully green-certified, and we’re generating much of our own energy. Process intensification also helps reduce the footprint. But advanced therapies are far from green today, especially with how heavily we depend on disposables.

JS: Though, if you take the broader view, one-time cures are inherently more sustainable than chronic therapies. A single treatment is a lot better than monthly injections in terms of lifetime environmental cost.

HL: I’d add that technology improvements — higher yields, better purity with the same inputs — are the quickest way to reduce waste. That’s where innovation really pays off.

JS: My biggest hope is that CAR-T and TCR-T therapies begin to show real efficacy in solid tumors. We’ve already seen remarkable results in liquid tumors, sometimes in patients I know personally, and it would be transformative to see that extend into the far more common solid cancers. What keeps me awake, though, is simply navigating the business cycles. This sector is cyclical by nature, and right now it’s a tough one. But cycles pass, and I look forward to the next upswing.

JSD: I’m hopeful about the potential to replicate in viral vectors what’s already happening with mRNA: building technology ecosystems that can support ultra-small batch manufacturing, even down to therapies for a single patient. We absolutely have the tools to make that possible if we can integrate them. On the pessimistic side, my worry is geopolitical. Progress in this field requires global collaboration, but funding streams are increasingly tied to national interests. Science can buffer against that, but ultimately the money dictates direction. I hope we find our way back to a more cooperative global approach.

AB: I’d love to see us achieve more affordable, more robust manufacturing processes. Maybe even breakthroughs like producing AAV from plants — there are companies already exploring that, and it’s exciting. This this industry is still small and fragile, and if it fragments along geopolitical lines instead of remaining a global enterprise, we all lose.

This discussion is part of the qualitative foundation for Nice Insight’s 2026 State of the Industry report — download the full report for a cross-modality view of how these pressures are reshaping the global CDMO landscape.