Phase-Appropriate CMC Requires a New Phase-Appropriate Manufacturing Model

Key Takeaways
Operation TrialBlazer and new FDA guidance provide more explicit support for phase-appropriate CMC strategies in first-in-human biologics development.
The FDA now states that a stable clonal cell pool may be sufficient for appropriate phase I INDs, potentially allowing conventional clone selection and other CMC activities to be deferred.
Published stable-pool programs demonstrate that accelerated early clinical manufacturing and later transition to clonally derived production processes can be feasible when comparability is appropriately managed.
Low clinical-material requirements for some early programs, high-potency biologics, and increasingly diverse antibody formats are challenging manufacturing models built around larger conventional bioreactor scales.
GBI Biomanufacturing's SmartScale platform combines phase-appropriate CMC, stable pools, small-scale GMP production, and integrated drug substance and drug product capabilities to create a right-sized pathway from development candidate to clinical material.

Rethinking the Path to First-in-Human

The United States has a speed problem in early drug development, and regulators are increasingly treating it as a strategic one. Operation TrialBlazer, launched by the U.S. Department of Health and Human Services (HHS) in 2026, frames the movement of early clinical research outside the United States as a threat to the country's leadership in biomedical innovation. The roadmap points to faster pathways in China and Australia, as well as delays that accumulate before and after an Investigational New Drug (IND) application becomes effective. HHS reports an average of 380 days between a pre-IND meeting request and IND submission, while Institutional Review Board review, contracting, and other activities can add substantial time after submission.¹

While chemistry, manufacturing, and controls (CMC) represent just one component of this broader challenge, they are an area where manufacturers can have a direct and meaningful impact.

TrialBlazer argues that drug developers sometimes generate more CMC and non-clinical information than is needed to support a first-in-human (FIH) study and that uncertainty around phase-specific expectations encourages additional work before an IND is filed.¹

I have long argued that drug developers are often asked to commit too much capital to CMC before a molecule has generated enough evidence to warrant that investment. That concern remains, but the greatest opportunity is not simply to move CMC farther upstream but to shorten the path from a sufficiently mature candidate to GMP material that can generate meaningful human data.

The same phase-appropriate thinking now being applied more explicitly to the IND package should also be applied to the manufacturing process. If a phase I study needs a small amount of a potent biologic, the drug developer should not automatically be forced into a manufacturing platform designed for hundreds or thousands of liters. CMC scope, manufacturing scale, and capital commitment should reflect the question the program is trying to answer at that stage.

A Longstanding Principle with New Specificity

The U.S. Food and Drug Administration (FDA) has been telling the industry for decades that early development does not require the same information package as later development. Its 2006 guidance on exploratory IND studies noted that existing regulations provide considerable flexibility in the information needed for an IND, depending on the purpose of the investigation, proposed human exposure, and associated risks. The FDA also observed that sponsors had not taken full advantage of that flexibility and often supplied more supporting information than regulations required.²

The exploratory IND pathway itself was deliberately narrow. These studies were intended to occur early in phase I, involve very limited human exposure, and have no therapeutic or diagnostic intent. They could be used to investigate pharmacokinetics, pharmacodynamics, target interactions, or biodistribution, or to help select among candidate molecules before committing additional development resources.²

That model did not translate especially well to many therapeutic proteins. For some biologics, very low exposure may not answer the pharmacologic questions needed to inform development. Manufacturing economics can also undermine the rationale for a deliberately small exploratory study. If the GMP campaign still resembles a conventional biologics program, there may be little advantage in building it merely to expose a handful of participants or generate a narrow clinical data set.

The opportunity today is broader. Phase-appropriate principles can be applied to conventional FIH phase I development itself, allowing drug developers to achieve clinically meaningful exposure without first completing activities whose greatest value may not be realized until later stages of development.

What the FDA Has Clarified in 2026

The FDA states that some sponsors provide more CMC information than is necessary for FIH phase I INDs and has clarified both the minimum information expected at this stage and information that can generally be submitted later. Analytical method validation data are not expected, process controls unrelated directly to product safety need not yet be specified, stability data from a specific clinical lot may not always be necessary at submission, and the initial stability package may not need to cover the full anticipated duration of the clinical study. The agency estimates that focusing on minimum, phase-appropriate CMC information could reduce application-development time by up to 12 months.³

For recombinant biological products regulated by the Center for Drug Evaluation and Research (CDER), the implications are particularly important. The FDA's current FIH CMC framework states that a stable clonal cell pool may be sufficient for a phase I IND. Derivation of a clonally derived cell bank is not expected at this stage, nor is a conventional two-tier cell-banking system. Certain process controls can be deferred, and product-specific viral-clearance work may not be expected where appropriate prior knowledge provides an adequate scientific basis.⁴

This level of specificity changes practical decision-making. Development teams can examine individual CMC activities to distinguish those that are essential for protecting study participants and ensuring the quality of clinical material from those that can responsibly be deferred until the program generates additional data and evidence.

However, CMC still must be rigorous. Clinical material must be produced within an appropriate quality system, and manufacturers need enough process understanding to assess risk, investigate deviations, and release batches with confidence. Phase appropriateness should concentrate early investment on the knowledge and controls necessary for the stage of the asset, with a clear path to the additional work required later.

The Science Has Advanced Faster Than Routine Practice

Industry experience over the past several years shows that faster development is technically possible, although routine practice has changed more slowly.

COVID-19 created an extreme test of conventional timelines. A BioPhorum industry benchmarking effort documented programs that compressed activities traditionally performed in sequence, relied more heavily on platform knowledge, and accepted greater business risk in exchange for earlier clinical evaluation. Thirty percent of the COVID programs included in the analysis reached an IND or Investigational Medicinal Product Dossier in less than six months, while most conventional programs required 13–18 months. The analysis also warned that poorly chosen acceleration strategies can shift work downstream and create restorative development later.⁵

Stable pools illustrate both the opportunity and the gap between technical feasibility and routine practice. A 2024 BioPhorum survey of industry CLD practices identified cell line development (CLD) as a continuing critical-path activity and found that most respondents required 24–28 weeks from gene to lead clone, with a much wider overall range. The variation depended in part on how much information organizations required before selecting a clone and the level of risk they were willing to accept.⁶

Pools were already widely accepted for process, analytical, and formulation development. However, their use remained limited when it came to clinical manufacturing. While 16% of respondents reported using purified pool-derived protein in preclinical toxicology studies, none incorporated pooled material into their standard approach for generating phase I clinical supplies.

The scientific evidence supporting stable pools and deferred cloning was already substantial before the FDA issued its current clarification. The more consequential change is that drug developers now have explicit regulatory language supporting a strategy that previously required them to defend a less familiar interpretation of phase appropriateness.

Manufacturing Scale Remains Part of the Problem

A streamlined CMC package offers limited benefit if drug developers are still required to utilize conventional phase I manufacturing platforms originally designed to support the scale and economics of large pharmaceutical companies.

In practice, much of the biologics manufacturing infrastructure available to early programs has become organized around efficient production at comparatively large scales. The widespread adoption of 2,000-L single-use bioreactors created flexible capacity while avoiding many of the cleaning and changeover requirements associated with traditional stainless-steel systems. Facilities, disposable assemblies, downstream equipment, staffing models, and commercial structures naturally developed around those scales.

For many early programs, actual material demand is much smaller. A published economic model developed by researchers at University College London and AstraZeneca assumed an adjusted phase I monoclonal antibody (mAb) requirement of only 0.1 kg while modeling manufacturing in a 2,000-L GMP facility. The analysis was not intended to demonstrate that every phase I campaign is oversized, but it illustrates the potential disconnect between early clinical demand and conventional manufacturing infrastructure.⁷

The antibody pipeline is also becoming less uniform. Non-canonical formats, including antibody–drug conjugates (ADCs), bispecific and multispecific antibodies, and other immunoconjugates, have accounted for approximately half of antibody therapeutics entering clinical studies annually since 2021.⁸ Increasing molecular complexity means fewer programs can be assumed to behave like a standard IgG platform asset. In some cases, these newer formats also bring characteristics that further reshape early clinical material needs, including very high potency or clinical strategies that incorporate imaging.

Potency, in particular, creates a separate scale consideration. Some T cell–engaging biologics are clinically dosed in microgram quantities, illustrating just how low per-patient dose requirements can be for some highly potent biologics.⁹ Rare-disease programs can produce another version of the scale mismatch because patient populations may remain small well beyond FIH.

We are increasingly seeing CDMOs respond with smaller bioreactors, including 500-L and 1,000-L options. Those are useful additions, but some programs may need 20, 50, or 100 L rather than hundreds of liters. At that point, simply scaling down a conventional architecture may not go far enough.

The industry's manufacturing infrastructure was largely built for an era of high-volume production. Today's emerging biologics often require a fundamentally different manufacturing scale, creating a mismatch between available capacity and actual program needs.

Building SmartScale Around the Clinical Question

GBI Biomanufacturing’s SmartScale platform starts with the amount of material and CMC work the program actually needs rather than with a predetermined vessel size.

The concept combines several tools that already exist into a single FIH pathway utilizing a proven tech stack. Stable-pool CLD can remove clone selection from the critical path when the molecule and regulatory strategy support doing so. Platform-based development can reduce molecule-specific work before GMP production. Manufacturing scale can then reflect the quantity required for the clinical study.

For appropriate programs, GBI is using the OptimumGrowth^® 7-L flask as one element of that architecture. The OptimumGrowth platform occupies a unique space between conventional laboratory-scale culture and stirred-tank manufacturing and is already being used in manufacturing processes within large pharmaceutical companies. Within SmartScale, multiple flasks can support truly small production volumes without requiring the same facility footprint, disposable assemblies, and operating structure as a conventional stirred-tank campaign. Further efficiencies can be gained when one considers multi-product throughput per cleanroom, further reducing the overhead burden carried by sponsors using CDMOs.

The initial SmartScale platform is centered on Chinese hamster ovary (CHO)-expressed mAbs, where prior knowledge is deepest and process behavior is comparatively predictable. More complex molecules can still fit, but they should be evaluated against the platform rather than assumed to fit it. We are building additional data around bispecifics and related formats so the scope can expand as our experience deepens.

The model also extends beyond upstream production. GBI has long handled small, customized batches, precious drug substance material, bioconjugation programs, and low-volume fills. SmartScale adds a right-sized upstream entry point to those capabilities so that small-scale drug substance (DS) can proceed into drug product (DP), filling, release, and stability within the same organization.

A small upstream campaign does not solve the FIH problem if the client then has to transfer the material and program elsewhere before it reaches the clinic. SmartScale is structured around gene-to-GMP drug product in a vial, with a path into conventional bioreactor manufacturing and later-stage production if the asset succeeds. The program can begin at the scale the clinical question requires without limiting the scale the molecule may eventually need.

Why GBI Is Positioned to Rethink First-in-Human Manufacturing
GBI Biomanufacturing brings a unique combination of small-scale flexibility and life cycle manufacturing experience to the development of SmartScale. The company has operated for more than three decades and supports biologics programs from early development through commercial manufacturing, with DS capacity spanning small volumes through 2,000-L single-use production, as well as integrated DP capabilities.
That breadth matters because GBI also has considerable experience at the opposite end of the scale spectrum. GBI teams routinely work with small, customized batches, limited drug-substance materials, bioconjugation programs, and low-volume fills, including programs in which only limited quantities of clinical material are required. The GBI team saw an opportunity to extend that existing small-batch expertise upstream and connect it with GBI’s established analytical, quality, drug-product, and larger-scale manufacturing infrastructure.
The result is a setting in which a drug developer can explore a genuinely small FIH manufacturing strategy without committing the program to a small-scale dead end. If an asset advances, GBI can support conventional bioreactor manufacturing, additional process and product characterization, later-stage development, and commercial-scale production within the same organization.
That combination is central to the SmartScale premise. GBI is not building a separate experimental pathway disconnected from conventional biologics manufacturing. It is adding a smaller entry point to an existing life cycle CDMO platform, allowing early investment and manufacturing scale to reflect the evidence behind the asset while preserving a credible route forward if the clinical data justify further development.

Deferred Cloning Without Borrowing Trouble from Later Development

Stable pools remain one of the places where drug developers understandably ask whether early acceleration could create a later comparability problem.

Published experience shows that this transition can be managed. In a published Boehringer Ingelheim program, GMP drug substance was manufactured directly from transposase-generated stable pools across eight 2,000-L bioreactor runs, generating clinical material in less than three months and enabling phase I approximately six months after transfection.¹⁰ A production clone was later derived from the same parental pool, and comparisons of pool- and clone-derived manufacturing demonstrated comparable product quality and manufacturability for the attributes evaluated.

A separate program from Merck KGaA and IAVI used a nonclonal stable CHO bank first for 200-L preclinical and formulation production and later at 2,000 L for phase I supply.¹¹ The clinical material met predefined product-quality requirements, and clinically ready material was generated within 4.5 months in that specific program.

These examples establish precedent rather than a guarantee. A later derivative clone must still be assessed against material used earlier in development. Their significance is that cloning can become a deferred activity rather than an automatic prerequisite to dosing the first patient.

The evidence is beginning to extend into more complex proteins as well. Recent work from Merck & Co. evaluating a multispecific antibody, a cytokine-Fc fusion protein, and a conventional antibody found high comparability among stable pools, pools of selected clones, and lead clones across the culture-performance, productivity, and product-quality attributes examined.¹² In that work, nonclonal material supported preclinical manufacturing while clinical material still came from the lead clone, appropriately limiting how far those findings should be generalized.

The FDA's explicit clarification that a stable clonal cell pool may be sufficient for an appropriate FIH phase I IND brings regulatory practice closer to this accumulated technical experience.⁴

Learning Where CMC Creates the Most Value

My own thinking about accelerated CMC has evolved through trying to put these ideas into practice.

Earlier, I believed there was substantial value in moving CLD and other CMC activities very far upstream and integrating them closely with drug discovery. The logic seemed straightforward: if manufacturing milestones begin sooner, the program should reach the clinic sooner.

What I underestimated was how nonlinear early biotech development can be. Discovery companies do not move through CMC in a predictable sequence, and generating a cell line for an asset that is still evolving does not necessarily create much immediate value. A cell line becomes much more consequential when it sits within a program capable of producing GMP material.

That experience changed the target. Early CMC remains useful when manufacturability information can influence candidate selection or eliminate downstream problems. The more significant opportunity is to make the transition from a credible development candidate to GMP clinical supply substantially faster and more capital efficient.

SmartScale is being built around that point in the development continuum. We are not trying to extend a CDMO deep into discovery and wait for programs to mature. We are adding a smaller, faster entry point immediately upstream of capabilities that already exist at GBI.

Experience has also changed how I think about portability. Drug developers should retain ownership and optionality, and no client should be trapped with a provider because transferring a process has become too difficult. At the same time, a transfer has no inherent value. If an appropriately designed early process can remain within the same technical and quality organization as the asset advances into later-stage manufacturing, avoiding the transfer may be the more efficient choice.

Drug developers should be able to move a program when doing so makes sense, without being forced to move simply because the initial provider has reached the limit of its infrastructure.

Start Small Without Building a Dead End

Phase-appropriate development is fundamentally about deferral. Work is postponed because the program has not yet earned the investment, not because it will never be required.

If a molecule generates an encouraging clinical signal, the developer will eventually need deeper process knowledge, additional characterization, more mature analytical methods, a later-stage cell-banking strategy, and potentially larger-scale manufacturing. Decisions made before FIH should preserve those options without forcing all that work into the period before human data exist.

The same principle applies to manufacturing scale. Using a small-scale process for an initial clinical study should not mean discovering in phase II that the process must be reinvented for conventional equipment. Platform knowledge, material comparability, raw-material choices, analytical continuity, and later-scale requirements still need to inform early decisions.

The BioPhorum speed-to-clinic analysis reinforces this point. Accelerated programs can create restorative work later when companies optimize for the first milestone without managing development risk across the full program.⁵

GBI's existing infrastructure provides a route forward. SmartScale sits in front of established DS, DP, analytical, quality, characterization, and larger-scale manufacturing capabilities. Early investment can remain proportional to the evidence behind the asset while the technical pathway retains the ability to expand as that evidence strengthens.

How Regulatory Flexibility Becomes Normal Practice

Clearer FDA language will not immediately erase conservative development behavior. Companies have rational reasons to avoid being first: a strategy may be permissible in principle and still feel risky if a drug developer expects an individual reviewer to request additional data, challenge a risk assessment, or interpret a flexibility differently. A recent LifeSciVC commentary on TrialBlazer has argued that some apparent overdevelopment may reflect this dynamic.¹³ Regulatory flexibility creates value only when companies have enough confidence to use it.

Practical precedent should help build that confidence. As more programs use stable pools, prior platform knowledge, deferred studies, and smaller manufacturing strategies, developers, regulators, and CDMOs will gain a clearer understanding of where these approaches work well and where additional evidence remains necessary. That experience may also give drug developers greater confidence to question whether later-stage CMC packages need to be purchased before an asset has generated early clinical data.

A conventional approach will still make more sense for some programs. SmartScale is an option, not a requirement. Molecules that do not fit the platform, require much larger quantities, carry unusual process risks, or come from companies that prefer to complete cloning before GMP manufacturing can follow a more traditional path. The objective is not to minimize CMC work indiscriminately but to build operating models that use regulatory flexibility while preserving rigor and future scalability.

For appropriate biologics, CMC scope, manufacturing scale, and capital investment can therefore be designed around the clinical question immediately in front of the developer. If the molecule earns the right to advance, the program can invest accordingly. If it does not, the company has reached that answer without first building a development package for a future that never arrives.