The history and potential future of monoclonal antibody therapeutics development and manufacturing in four eras

ABSTRACT

Therapeutic monoclonal antibody (mAb) development and the processes for manufacturing drug substance have evolved since the first approval of the mAb in 1986. As the past is often the prologue to the future, the history of these technologies has been classified here into three eras, leading to speculation about what the next era may hold with regard to development and manufacturing strategies, as well as the potential impacts to patients. The substantial increase in production culture titers and bioreactor production volumes and the availability of large-scale contract manufacturing facilities could translate into improved global access for these therapies and an expansion of indications for therapeutic antibodies.

KEYWORDS:

• Cost of goods
• global access
• manufacturing

Introduction

The first era includes the first forays into monoclonal antibody (mAb) therapeutic development following the discovery of technology to produce mAbs. Subsequent eras have brought substantial advances in an array of technologies which have accelerated product development timelines, reduced the cost of manufacturing commercial mAbs, and enabled product supply to millions of patients. Five key elements of each era are as follows: 1) Timeframe, 2) Median range of production culture titers, 3) Estimate of maximum product demand in tons/year, 4) Number of mAbs licensed or approved in that period,¹ and 5) Estimate of the total commercial sales of all mAbs in that period. Note that one ton of a mAb = 1,000 kg = 1 million × 1 gm doses (thus for an annual seasonal prophylaxis treatment this would be 1 million patients). Data for these eras are summarized in Table 1.

Table 1. Profiling the four eras of mAb development and manufacturing.

	Era 1	Era 2	Era 3	Era 4
Years	1975–1998	1999–2019	2020–2023	2024+
Typical titers (g/L)*	<1–2	2–5	5–8	8 +
Maximum single product volume (tons/yr)	<1	1–3	3–8	10 +
# of mAbs licensed or authorized (US/EU)	5 (0.2/yr)	79 (3.4/yr)	44 (11.0/yr)	TBD
Market size ($B/yr, estimated)	<1	1–175	175–225	TBD

*Titers are representative of full-length IgG1 mAbs.

The origin story (1975–1998)

Kohler and Milstein first created mAbs in 1975, and their subsequent Nobel Prize together with Jerne in 1984 acknowledged the potential value these agents could bring to human healthcare, diagnostics, and research. This discovery had a profound effect on biopharmaceutical research and development, ultimately enabling nearly 200 mAbs to be approved for the treatment or prevention of diseases in at least one country, as of mid-2024 (https://www.antibodysociety.org/antibody-therapeutics-product-data/).

The first mAb was approved for human use in 1986. Orthoclone OKT3 (muromonab-CD3) was an anti-CD3 murine antibody used to treat renal transplant patients and reduce transplant kidney rejection. Due to the limitations of the technology available at the time, OKT3 was manufactured using hybridoma cells that were cultivated in mice in their peritoneum; the cells grew as tumors and secreted the mAb, which was harvested for subsequent purification and formulation. The development of recombinant DNA techniques enabled the use of mammalian cell lines for the production of subsequent antibody therapeutic products.

The early history of therapeutic mAbs reflects complications arising from the limitations of the research methods available at the time. MAbs derived from murine sources have failed in human clinical studies due in part to the development of anti-drug antibodies (ADA) in patients. Subsequent research led to the use of chimeric sequences, which were more successful, and later to humanized and fully human sequences, which reduced this barrier further. These advances ultimately enabled many biopharmaceutical companies to have product pipelines based largely on mAb therapies.

The first licensed recombinant full-length mAb therapy was Rituxan (rituximab), developed by IDEC Pharmaceutical Corporation and Genentech. Rituxan is a chimeric antibody manufactured in a Chinese hamster ovary (CHO) cell line, which is the host cell used for most therapeutic recombinant proteins. Rituxan was launched in 1997 from Genentech’s commercial manufacturing facility in South San Francisco, which was also used for the launch and production of tissue plasminogen activator (Activase) and DNAse (Pulmozyme) derived from recombinant CHO cells. With a bioreactor volume of 12,000 L, the production scale was unprecedented for mAb manufacturing at the time.

In this early era of recombinant mAb manufacturing, there was rich debate about the options to be used for mAb purificationin particular, regarding whether Protein A resins should be used in the production process. Concerns were raised about Protein A leached from the resin, which, if present in the product administered to the patient, could cause potential adverse effects. The demonstration of very low residual levels of Protein A in purified product and the precedent established by Rituxan’s approval led to the widespread adoption of Protein A capture, which become the dominant capture step in nearly all mAb purification processes.

Although not a mAb, the approval of Enbrel (etanercept) and the subsequent commercial supply challenges are key elements of the first era. Enbrel is a soluble recombinant protein composed of the tumor necrosis factor (TNF) receptor p75 fused with an antibody Fc manufactured using a process that is highly similar to that of most mAbs. Produced in CHO cells, Enbrel was approved and launched in 1998 for the treatment of rheumatoid arthritis by Immunex, which had a manufacturing agreement with Wyeth, and was later acquired by Amgen. The commercial demand for Enbrel quickly became so great that patients were put on a lottery for access to treatment until the product supply could adequately cover the demand.² Between Amgen and Wyeth, as the two companies supplying the global supply of Enbrel, they deployed six of the largest CHO manufacturing sites to simultaneously manufacture Enbrel.

While these events occurred over 25 years ago, there was an echo during the COVID-19 pandemic of the third era (2020–2023). The lesson learned during the Enbrel episode (i.e., inability to meet demand for a blockbuster product has huge negative impacts) left an indelible imprint on innovator companies developing mAb therapies. As a result, many innovator companies invested substantial capital resources in CHO cell production facilities, which has led to a pendulum effect on the mAb supply/demand curve in favor of innovator companies’ internal capacities.

Maturation and scale-up (1999–2019)

In the second era, platform manufacturing processes used by many innovator companies and contract manufacturing organizations (CMOs) were established as mAb therapeutic development expanded globally. Data shown in Table 1 for this 21-year period reflects a maturing industry. Production titers, facility capacities and product demands all rose significantly.³

Early blockbusters in this era include Enbrel and Rituxan, as well as Herceptin (trastuzumab), Humira (adalimumab), and Avastin (bevacizumab). Sales of these and other blockbusters have been in the range of $1B -$20B per year since 2000 and beyond. This led many biopharmaceutical companies to invest in mAb manufacturing technology.

As a result, the industry converged on a drug substance manufacturing platform: fed-batch CHO cell culture for production using chemically defined media, clarification by centrifugation, Protein A capture followed by one or two polishing chromatography steps and virus filtration, ultrafiltration and diafiltration to provide assurance of the viral safety of the product and concentrate and formulate the drug substance.^4–6

The increase in the production culture titers early in this period led to concern over whether the purification process could handle the increased mass generated upstream. Multiple papers and conference sessions focused on the ‘purification bottleneck’. The benefits of the mAb process platform described above, however, were rapidly apparent. While many early mAb manufacturing facilities were designed to process titers less than 2 g/L, many purification bottlenecks could be addressed by increasing the volume of tanks used to capture the intermediate process pool or cycling the chromatographic steps. New separation technologies were not needed, as relatively simple modifications of a well-established process were sufficient.

As confidence in large-scale CHO culture processes increased, the design basis of new manufacturing plants evolved. Innovator companies generally moved ahead first with manufacturing network expansions in response to the lessons learned from the market’s response to Enbrel. It became common for innovators and CMOs to build CHO manufacturing facilities that had bioreactors ranging from 12 – 25kL. These would frequently be built with 6–12 bioreactors, often feeding two purification trains downstream. This enabled more than one product to be produced simultaneously in the plant and debottlenecks a campaign with a single product produced in all bioreactors.

COVID experiences (2020–2023)

In the third era, product development strategy rapidly pivoted to enable the accelerated launch of mAbs to address the COVID pandemic. Many companies (Lilly, Regeneron, Vir Biotechnology, AstraZeneca) and CMOs/innovator partners (Amgen, FujiFILM, Genentech/Roche, Samsung, Wuxi Biologics) produced mAb drug substance to treat or prevent COVID-19. The total volume of COVID mAbs produced is estimated to be 2–8 tons/yr for any single product, with a total of 30+ tons for all (Table 1). The range of titers shown in Table 1 were reported in publications or conference presentations and reflects high titers resulting from state-of-the-art cell-line development and cell culture production capabilities developed in the 2010s.

An oft-cited article published in February 2020⁷ states that the industry’s entire capacity for mAb production was 25 tons/yr in 2018, but based on increasing production culture titers and continued facility expansions, this estimate can now be revised up to 80–100 tons/yr. COVID-19 mAbs have treated 10s of millions of patients, which may be more than the total number of patients treated with all other commercial mAbs to this point.

MAb development and manufacturing processes were substantially impacted by events related to the pandemic.^8–15 The experience has reinforced key drug substance manufacturing goals, including process portability, process simplicity, and dual-sourcing of drug substance manufacturing and raw material suppliers.¹⁶ These goals were pursued by multiple innovator companies and their partners and achieved in the context of accelerated regulatory review and authorization by virtue of the need to address the COVID-19 pandemic.

Multiple mAbs were authorized for treatment and prevention of COVID-19, which was enabled by a massive increase in production capacity. At one point, at least nine very large-scale (VLS) drug substance manufacturing facilities were engaged in producing COVID-19 mAbs. The rapid authorization of mAb therapies for COVID-19 combined with the fact that over 200 mAb therapies have been licensed acknowledges the fact that they are generally a low risk to patients and bodes well for future mAb product development of this class of products. The titers of the second era represent cell culture technologies of that period, different than the COVID-19 panel of mAbs which used more advanced technologies.

There has been a call for reinventing mAb manufacturing based on a lack of innovation in recent years. Is the success of the COVID-19 mAb product development sufficient evidence to confirm that conventional manufacturing technology is not broken? The extremely rapid authorization of eight mAbs and robust supply to tens of millions of patients cannot be underrated. So why are some organizations establishing new manufacturing technologies for mAbs? What problem are they hoping to solve? It is clearly not maximal production capacity or minimal cost of goods (COGs), as economies of scale weigh heavily in favor of VLS cell culture facilities, or process portability, which is favorable for intensified processes.¹⁷

Some have touted the benefits of distributed manufacturing with small-scale facilities, which would inherently be more expensive and complex regarding quality assurance, regulatory management, and balancing the supply chain.

With the production expansions announced by four major mAb CMOs (see below), why would an innovator company seek to increase internal production capacity using innovative manufacturing technology when a low-cost, low-risk alternative is readily available?

Should a future pandemic arise and mAb products be needed quickly and in large volumes, the precedent set by the COVID-19 response should be followed, i.e., deploy conventional state-of-the-art technology and engage the existing manufacturing facilities of multiple innovator companies and contract development and manufacturing organizations.

Expanding the future for mAbs (2024 - ?)

The fourth era looks to the future of mAb therapies, when questions such as ‘could new therapeutic areas become accessible? and ‘might expanding mAb therapies to the developing world be possible?’ can be addressed. The titer of 8+ g/L for this era (Table 1) reflects the current state of the art for cell-line development and a production media platform, which should be possible for similar mAbs in the years to come. This reflects a Phase 3 process optimization investment that would improve upon the third-era COVID-19 mAb titers, which were obtained from platform technology where there was little or no time for optimization. Improving production and feed media, feeding schedule, and bioreactor process control targets are likely to yield a 25% increase or more in titer.

Recently, several CMOs with VLS drug substance manufacturing capabilities have announced expansions. Samsung Biologics is building its fifth plant in South Korea with construction and start-up expected to take just a few years. FujiFILM is building standardized facilities in Denmark and North Carolina, a significant expansion in the United States. Celltrion is building its third plant in South Korea. Roche’s Vacaville, CA facility, which is currently the largest CHO manufacturing facility in the world, has been sold to Lonza. Boehringer Ingelheim is expanding production in Vienna. As global CMOs continue to expand, it is not clear whether innovator companies need to make multi-billion dollar investments for internal mamnufacturing in the future. It could be possible that 10–100 million doses of a blockbuster mAb product could be manufactured in just a year or two at a CMO. The pendulum of mAb supply from innovator or CMO partner appears to have swung again, with CMOs to soon have more capacity overall than innovator companies.

This opens new opportunities for the expansion of patient access to mAb therapies for both developing and developed countries.^17,18

Many innovator companies and CMOs are making a concerted effort to establish continuous drug substance manufacturing processes. It is unlikely that this will expand the future of mAb therapies in the sense that they would open new therapeutic areas and are unlikely to enable a reduction in the COGs offered by VLS manufacturing.¹⁷ As of 2024, mAbs are more like insulins, vaccines and plasma-derived products than other parenteral recombinant protein products regarding manufacturing strategies and COGs. The manufacturing of mAbs uses portable platformed processes and a network of production facilities with a similar design basis. They can enjoy the economy of scale that minimizes their COGs. Several companies have reported that they have produced a 100 kg batch of drug substance; it is likely that some of the largest facilities could produce a 150 kg batch.

How does that change strategies for mAb development and manufacturing? mAbs are now a unique family of biologics that have significant advantages in development and manufacturing compared to many other biologics.

The movement of multiple innovators into treatment of Alzheimer’s was a leading indicator of their confidence in being able to supply very large volumes for a high-dose chronic therapy. This opens opportunities in new therapeutic areas compared to the past, when the combination of a large annual dose requirement and potentially limited manufacturing capacity might have dissuaded innovator companies from considering these targets. Recent license applications for mAbs used for the prevention of migraines and the treatment of hypercholesterolemia are similar examples.

The combination of production capacity and reduced COGs may also encourage innovator companies to consider indications where the sales price would be lower than most recombinant protein therapeutics. For instance, a blockbuster product that could be produced for approximately $50/gm could be sold for hundreds of dollars per gm, which would be much lower than legacy mAbs but still provide a solid profit margin. Could global healthcare providers (including WHO, IAVI, and GAVI), accept that, while these prices are higher than vaccines, they would still enable distribution of products with significant value to millions of patients?

For example, consider a product that has a market of 20 tons/yr and requires a 500 mg dose. This would yield 40 M doses per year and would cost approximately $1 billion per year for production. These could be mAbs used for prophylaxis of infectious diseases such as malaria, RSV, influenza, and HIV.^19,20 Cocktails of two or more mAbs may also be possible, given this production volume and production cost. There is the potential for a global health organization, disease alliance, or foundation to fund these efforts. In one manufacturing and supply strategy, a VLS facility could be dedicated to a portfolio of mAbs from one or more organizations that share the same platform process and work together as a consortium.²¹ The facility could be staffed and managed by a large CMO, building on the preexisting infrastructure of the site where multiple facilities would be in concurrent operation. The facility would then supply large annual demands for multiple products, including biosimilars of established products, prepare stockpiles for future use during an epidemic (e.g., influenza), and to rapidly switch to produce a novel mAb that might be needed for treatment of a newly emerged pathogen, should that dire situation present itself again. Such a facility would have been very valuable during the first year of the COVID pandemic. The products could be sold on a ‘cost-plus’ basis, forgoing the profit margins imposed by an innovator company.

Experts in global access for biologics agree that, to maximize impact in low- or middle-income countries (LMICs), COGS must be minimized. The strategy proposed above would yield a significant cost savings compared to any other supply model, such as a local, distributed manufacturing system.¹⁷ While such models have made some inroads globally, such as for mRNA vaccines during the pandemic, the manufacturing scale and complexity of the production process, manufacturing scale, raw material procurement and controls for mAbs are much more complicated, thus driving the centralized manufacturing strategy described here. In addition, the establishment of a centralized VLS manufacturing facility using conventional technology is a ‘bird-in-the-hand’ that could be deployed more rapidly and with greater impact than working through the complexities of a distributed manufacturing network in multiple LMICs.

Clearly, the benefits of such large volume production could be enhanced by selecting and engineering mAbs to reduce the dose and/or dosing frequency and thus enable higher volumes of finished goods. Such enhancements could include affinity maturation of the Fv domain to increase potency,²² mutations of the Fc domain to extend half-life (especially valuable for prophylactic indications), and tuning of the effector function through Fc modifications to improve recruitment of T-cells or other immune modulators.²³ For treatments to be broadly available in LMICs, the route of administration cannot be through infusion, but rather by subcutaneous or intramuscular administration. The limited injection volumes of these routes (2–5 mL typically) cannot support high doses without requiring multiple injections. Further, the improved ability to ensure that mAbs in a research pipeline undergo extensive developability analysis enables improved product stability and reduced viscosity,^24–26 which are needed for highly concentrated products delivered in low injection volumes. The dose reduction efforts described above would have three important benefits: increased production volumes, reduced COGs, and enablement of optimal routes of administration.

Despite frequent calls for innovation in mAb manufacturing strategies and technologies, the experiences and success of multiple innovator companies and partners during the COVID-19 pandemic suggests that there is no lack of production capacity in the industry. Conventional mAb manufacturing is capable of rapidly ramping up to very large production volumes, which would drive low COGs because of economies of scale. The innovations that may be needed most are not in manufacturing technology, but in the business models to supply mAbs for novel therapeutic targets and to LMICs as described above.

Conclusions

Monoclonal development and manufacturing have evolved substantially in the nearly 40 years since the first mAb was licensed. Recent advances in production culture titers, continued optimization of upstream and downstream processing platforms, and expanding global manufacturing capacity could enable opportunities for mAb therapies in new therapeutic areas, as well as improved global access to LMICs. The latter opportunity might best be served by a consortium of non-government organizations and global healthcare funding bodies that can bring a portfolio of antibodies forward and leverage a partnership with a large CMO to supply, stockpile, and rapidly develop novel antibodies should another pandemic occur.

Abbreviations

mAb	=	Monoclonal antibody
COGs	=	Cost of goods
TNF	=	Tissue necrosis factor
VLS	=	Very large scale
LMIC	=	Low and middle income countries

Acknowledgments

I want to acknowledge the memory of Michael Kamarck, a generational leader who inspired so many to excel at what we do in the biopharmaceutical industry. This paper is also dedicated to the hundreds of classmates, coworkers, colleagues, and friends who accompanied me on various parts of this journey since 1987. A special thanks to those who helped me with the final manuscript edits and content: Raj Gupta and Kristen Douglas.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported that there is no funding associated with the work featured in this article.