Biology’s new supply chains: Why the future of biotech may be manufacturing, not discovery

The story starts the same way every time. A small team finds a promising molecule. In early experiments it behaves, reliably, even beautifully. Investors lean in. A paper is drafted. Then the project meets the part of biotech that does not fit neatly into a press release: the factory.

The awkward truth is that modern biology is increasingly constrained by production. Discovery still matters, obviously. But the next decade will be shaped as much by biomanufacturing, bioprocess engineering, and quality systems as by new targets and new modalities. Not because scientists have run out of ideas, but because the hardest questions are now logistical. Can you make it at scale? Can you make it consistently? Can you prove, batch after batch, that you made the same thing?

This is not a niche concern. In advanced therapies, the “product” is often a living system, or a complex protein with delicate chemistry. In a world of gene therapies, cell therapies, ribonucleic acid (RNA) medicines and engineered microbes, manufacturing is increasingly the science that decides whether the science reaches anyone.

The case study thread: the therapy that worked, until it had to be made

Imagine a generic, plausible programme. A lab develops a therapeutic protein that dampens inflammation. Early batches are made in tiny flasks. The protein looks clean. In animal models the signal is encouraging. A first-in-human study is planned.

Then scale-up begins. The team moves from millilitres to hundreds of litres. The cells grow differently. The protein’s sugar decorations, glycosylation, shift slightly. The potency assay, the test meant to show the medicine is active, becomes noisy. A batch is discarded. A delay becomes a quarter. A quarter becomes a year.

Nothing “failed” in discovery. Manufacturing simply revealed that the product was not yet a product.

Regulators have a term for this uncomfortable zone, chemistry, manufacturing, and controls (CMC). It is where biotechs learn that a medicine is not only a sequence or a mechanism, it is a controlled process. FDA guidance for gene therapy CMC, for example, emphasises describing the manufacturing process, controls, and how risks to patients are managed.

Scale-up: why biology does not behave like a recipe

Scaling a chemical process is hard, but scaling living systems adds layers of fragility.

Cells respond to oxygen, temperature, mixing, shear forces, nutrients, and waste build-up. Move from a shake flask to a bioreactor and you change the physical environment, even if the ingredients look the same on paper. That physical shift can alter yield, stability and quality attributes.

This is why “scale-up” has become “scale-out” in some areas, particularly cell therapies, where distributed, smaller-scale manufacturing may fit the product better than a single giant tank. It is also why process development teams can spend years mapping the relationship between process conditions and product quality.

Cell culture versus fermentation: two manufacturing worlds, two risk profiles

Biomanufacturing splits broadly into two production modes.

Mammalian cell culture is widely used for complex proteins such as monoclonal antibodies, because mammalian cells can perform the post-translational modifications, including glycosylation, that often matter for function. It is slower, more expensive, and more sensitive to contamination.

Microbial fermentation, using bacteria or yeast, is often faster and cheaper for many proteins and enzymes, and can tolerate more aggressive mixing and aeration. It is not automatically simpler, but it is a different industrial logic. In many cases, the deciding factor is what the protein needs to look like, not what the finance director prefers.

The manufacturing method shapes the product pipeline. Some therapies are effectively ruled in or ruled out by what can be produced with acceptable quality, cost and throughput.

Continuous manufacturing: a shift from batches to flows

Traditional bioprocessing is a sequence of batches, discrete runs with set starts and stops. Continuous manufacturing and process intensification aim to keep production running and stable for longer, improving productivity and potentially shrinking facility footprints. Reviews and industry analyses describe continuous biomanufacturing as promising but operationally complex, requiring tighter control strategies and integrated analytics.

Regulators have been encouraging modernisation for years. FDA’s Process Analytical Technology (PAT) framework was designed to support more innovative manufacturing and quality assurance, and it explicitly discusses “real time release” concepts based on process measurements rather than only end-product testing. The European Medicines Agency (EMA) has also published guidance on real-time release testing (RTRT), describing release based on manufacturing process information, supported by process understanding and control.

Continuous does not magically remove risk. It changes it. Instead of a single bad batch, you worry about a slow drift that contaminates an extended run. Instead of sampling at the end, you need measurements during the process, and the data systems to prove what happened.

Quality-by-design: building quality into the process, not inspecting it in

Quality-by-design (QbD) is the philosophy that quality should be designed and demonstrated through scientific understanding and risk management, rather than assured only by testing the final product. EMA describes QbD as using statistical, analytical and risk-management methods to design and manufacture medicines that consistently meet quality.

The wider international framework is embedded in the International Council for Harmonisation (ICH) guidelines, particularly Q8, Q9 and Q10, which connect pharmaceutical development, risk management, and a pharmaceutical quality system.

In practice, QbD means identifying critical quality attributes and critical process parameters, then building a control strategy that keeps them within acceptable ranges. It also means documenting change. A well-run process anticipates that change will happen, and sets out how it can happen without altering safety and efficacy.

Analytics and batch release: the rise of measurement as infrastructure

Manufacturing is increasingly an analytics problem.

Advanced therapies often require complex assays for identity, purity, potency and safety. For cell and gene therapies, potency is particularly difficult, since it may reflect multi-step biological cascades rather than a single measurable property. FDA guidance on potency tests for cellular and gene therapy products stresses that potency assay acceptance criteria should reflect clinical experience and manufacturing data.

This is not paperwork. It is the bridge between “we think it works” and “we can prove this vial is what we said it is”.

PAT and RTRT are part of the same trend. As measurement improves, release can become more data-driven. The prize is faster release, fewer failed batches, and more confidence in supply. The cost is the creation of a new engineering layer: sensors, models, calibration, data integrity, and an audit trail that can survive inspection.

Contamination: the threat that ends programmes quietly

Contamination is the biomanufacturing nightmare because it destroys trust in the entire run, and sometimes the facility.

An industry-wide study published in Nature Biotechnology in 2020 analysed viral contamination events in biologic manufacturing and reported that such incidents have occurred repeatedly across companies and cell lines, with significant operational impact. MIT’s summary of the study reported 18 viral contamination incidents since 1985 among companies that shared data, often involving Chinese hamster ovary (CHO) cells, a standard workhorse in protein production.

Viral contamination is not the only threat. Mycoplasma, bacteria and fungi can also compromise cultures. The practical consequence is that manufacturing strategy is often shaped as much by containment and detection as by yield.

This is where single-use technologies have made a mark. Disposable bags and tubing can reduce cleaning burdens and some cross-contamination risks, while creating new dependencies on suppliers and plastics.

Supply chain fragility: the hidden constraint on “scalable” biotech

Biology’s supply chains are not only vials and labels. They include filters, single-use bags, chromatography resins, media ingredients, and sterile connectors. During the COVID-19 pandemic, shortages exposed how fragile these supply chains can be.

An ISPE article described how pandemic-related shortages constrained supplies of essential filters and chromatography resins, arguing for more agile pathways to implement alternatives to maintain supply of approved biologics. Biophorum has also highlighted the lack of a standardised approach for changing Protein A resins under constrained supply conditions, despite the regulatory and technical work required to qualify alternatives.

The implication is blunt. A therapy may be scientifically “scalable” but practically throttled by bottlenecks in specialist consumables.

Supply fragility also complicates resilience planning. It pushes manufacturers towards dual sourcing, platform processes, and designs that tolerate component substitution without drifting out of specification.

Sustainability: the environmental cost of making biology

Biomanufacturing has an environmental footprint, and the trade-offs are not always intuitive.

Single-use systems generate visible plastic waste, but stainless-steel systems require significant water, energy, and cleaning and sterilisation operations. ISPE and academic analyses have discussed life-cycle approaches where single-use can, in some contexts, have lower overall environmental impact than traditional systems, while also raising legitimate concerns about waste management.

Sustainability is not only ethics, it is capacity. Water and energy constraints, disposal rules, and net-zero targets shape facility design and operating costs. Process intensification and continuous approaches are partly about efficiency, not only speed.

How manufacturing constraints decide which therapies reach patients

Manufacturing constraints are not an afterthought. They shape the frontier of what counts as feasible medicine.

• Autologous cell therapies can be clinically powerful, but they are operationally complex, with vein-to-vein logistics, cold chain, and patient-specific variability. Manufacturing capacity becomes clinical capacity.
• Gene therapies raise challenges in vector production, potency testing, and long-term consistency, with CMC documentation and controls central to risk management.
• Biologics such as antibodies can be scaled, but purification steps rely on specialised resins and filters, and are sensitive to supply constraints.

The net effect is that “what we can treat” increasingly depends on “what we can reliably make”.

What could speed this up?

What could speed this up?

• Better analytics and automation, including PAT-enabled control strategies and more mature RTRT approaches, reducing delays and failures.
• Standardised platform processes, especially for common modalities, making scale-up less bespoke and more repeatable.
• Flexible facilities, including modular and single-use-enabled plants that can switch products faster, reducing capital barriers and improving resilience.
• Regulatory clarity on change management, supporting faster substitution of constrained components without compromising quality.
• Workforce investment, particularly in bioprocess engineering, quality, and data systems, the roles that translate molecules into medicines.

What could derail it?

What could derail it?

• Contamination events, especially viral contamination in mammalian cell culture, shutting down runs and eroding confidence.
• Supply chain shocks, including shortages of resins, filters, and single-use components, delaying production and limiting scale.
• Overpromising on “continuous”, deploying complex systems without the controls, models and governance needed to manage drift and data integrity.
• Public trust failures, particularly if quality incidents or opaque manufacturing changes undermine confidence in advanced therapies.
• Sustainability backlash, if waste and energy impacts are treated as public relations rather than engineering constraints.

The grounded conclusion: investment, workforce, and trust

Biotech’s next decade will not be decided only by clever biology. It will be decided by whether societies build and maintain the industrial capability to produce complex therapies safely, consistently and affordably.

That requires investment in plants and equipment, but also in people. Bioprocess engineers, quality professionals, assay developers, data specialists, and operators who can run controlled systems under Good Manufacturing Practice are not supporting characters. They are the plot.

It also requires a more honest public narrative. Manufacturing is where medicine earns trust, batch by batch, through controls that regulators can inspect and patients can rely on. The future of biotech may well be manufacturing, not because discovery is finished, but because the next breakthroughs are increasingly defined by whether we can make biology behave on an industrial schedule.