August 1, 2012 (Vol. 32, No. 14)
Angelo DePalma Ph.D. Writer GEN
The diversity of development tools, an abundance of manufacturing options, and standardization of many unit operations has ushered in a golden age for monoclonal antibodies (mAbs). mAb process development and GMP manufacturing increasingly rely on platform processes involving streamlined production (usually a two to three week fed-batch cell culture), protein A capture, ion-exchange intermediate and polishing columns, and standard filtration steps.
Because of these developments, both innovator companies and contract manufacturers have achieved a high degree of familiarity with mAb manufacture, process development, and clinical development.
Platform manufacturing and quality initiatives are mutually supportive and enabling. As manufacturing platforms become more widely adopted, developers of mAbs have a somewhat easier time implementing quality by design and process analytics, which further aid in establishing standardized unit operations and support CMC regulatory submissions.
While platform processes have become common both upstream and downstream during mAb manufacturing, purification remains something of a bottleneck and an increasing cost contributor, particularly in chromatography-intensive processes. Rising titers are an often-cited contributor to upstream-downstream mismatches, while finding replacements for protein A has met with spotty success.
And while platforming is desirable in the manufacture of conventional mAbs, all bets are off for nonstandard antibody-like molecules, particularly antibody fragments, due to higher variation in titers and the need to alter purification from standard mAb operations. “Antibody fragments can also vary in terms of stability and solubility, presenting significant challenges to formulation scientists,” notes Gregory Zarbis-Papastoitsis, Ph.D., senior director for protein production at Eleven Biotherapeutics. Eleven develops antibodies, antibody fragments, and nonantibody proteins to treat eye diseases.
Chinese hamster ovary (CHO) cells remain the most popular mAb-producing cells due to regulatory familiarity and the large number of successful clinical- and commercial-stage products expressed in these cells. “However, as second-generation expression systems enter clinical trials and commercialization, CHO may experience a significant drop in popularity,” says Richard Hetrick, director of business development at Cytovance Biologics.
Single-use bioreactors are now mainstream and capable of handling clinical and even production-scale batches for many mAbs. The high doses at which antibodies are administered guarantees, however, that large stainless steel bioreactors will not disappear.
Despite vendors regularly breaking the size barrier for bioreactor bags, the point will eventually be reached where the bags become too large to handle. Disposal issues may also be problematic for very large bags. Regardless, upper size limits become moot as cellular and culture productivity continues to rise.
“With increase in yields it is debatable whether there will be much demand for 10,000-liter bioreactors in the future, as was projected a number of years ago,” Hetrick observes.
Improvements in volumetric productivity for mAb processes do not appear to be slowing down. To this point most were a result of media and feed strategies, but high-expressing cell lines and transfection strategies have played a role as well.
In June, ProBioGen announced a modification that improves volumetric yield in certain cells up to 2.5-fold for certain products produced in CHO cells. The technique involves co-expressing an enzyme along with the therapeutic mAb. The enzyme, according to ProBioGen, acts “on several cellular pathways and results in substantially enhanced volumetric productivities of protein drugs.”
Although CHO cells are currently the most popular mAb-producing cells, as second-generation expression systems enter clinical trials and commercialization, Cytovance scientists believe that CHO cells may experience a significant drop in popularity.
Modifications
Over the last 15 years approximately 30 therapeutic monoclonal antibodies have received regulatory approval. The majority of these proteins are full-length, unmodified IgG1 molecules. “Some reasons for the success of this molecular class are that full-length IgG1s are structurally stable, possess a long serum half-life, and their Fc regions can also confer secondary immune functions or effector functions,” observes Walter Low, Ph.D., director for antibody engineering at PX’Therapeutics. PX specializes in developing therapeutic antibodies and other recombinant proteins.
But “canonical” IgG1 molecules possess inherent disadvantages, according to Low. They target mainly cell surface antigens—a therapeutic limitation—and their large size and complexity make them difficult to produce to homogeneity, hence their high manufacturing costs. Many of these drawbacks are addressed through advanced manufacturing and cell-line engineering technologies, which have improved volumetric productivity by as much as 10-fold over the last dozen years.
But biopharmaceutical companies are also devoting substantial resources to developing “next-generation” antibody-based therapeutics that include mainly antibody-drug conjugates (ADCs), bi-specific antibodies, antibody fragments, and engineered antibodies.
ADCs are full mAbs or antibody fragments chemically linked to a cytotoxic drug. The combination target specificity and cytotoxicity represents a powerful alternative for treating cancer. Several of these ADCs are currently in advanced clinical trials and many more are in early development.
Antibody conjugates are not a new idea. Early implementations involved chelation, through addition of EDTA-like chemical side-arms, with radioisotopes. Some of these agents are still used for imaging and therapy. The modern version of these “stealth” molecules are composed of antibodies and cytotoxic drugs.
In Adcertis (Seattle Genetics), a lymphoma drug, the antibody targets CD30-expressing cells while the toxin induces cell death. Similarly T-DM1 (Roche) combines the breast tumor-seeking antibody Herceptin with Immunogen’s TAP cytotoxic agent. Adcertis is approved, T-DM1 is in Phase III, and other related molecules are in various stages of development.
For example Merck recently announced a collaboration with Ambrx for “smart bomb” antibody compounds consisting of an IgG chemically bound to a cytotoxic compound. This is Ambrx’ third foray in immunoconjugates. It is also working with Pfizer on a conjugate for an undisclosed indication, and has an antibody-peptide conjugate in preclinical development.
“We expect this area of mAb research and development to intensify in the coming years,” says Dr. Zarbis-Papastoitsis.
According to PX’Therapeutics, the flexibility and capabilities of its mAb technology platform enables it to address a myriad of antibody development related issues.
Large biotech firms are scurrying to exploit bi-specific antibodies as well. Novartis recently paid GenMab $2 million to tap into the latter’s DuoBody™ technology for creating bi-specific mAbs. The full value of the deal could reach $175 million for GenMab, which has its own development pipeline as well.
Bi-specific antibodies combine the affinities and activities of two different mAb fragments and target two distinct antigens simultaneously, for example a tumor target and a cytotoxic immune system cell.
“Bi-specific antibodies pose unique challenges in development and manufacturing and can require added effort and time in bringing new therapeutics to commercialization,” says Kevin Bailey, Ph.D., vp for preclinical manufacturing at Regeneron Pharmaceuticals. Regeneron develops monoclonal antibodies to block individual therapeutic targets in oncology, cardiovascular diseases, infectious diseases, and others.
A somewhat related approach is to administer two mAbs at once with the idea of delivering an orthogonal one-two punch to the therapeutic target. Such treatments would be prohibitively expensive unless the innovator companies collaborated on a designated combination product, an unlikely occurrence. An innovative way around this idea is to produce two or more mAbs simultaneously in the same cells.
That is the idea behind Oligoclonics® from Merus, which creates mixtures of mAbs with different specificities through co-transfection of cells for multiple antibodies. This approach essentially creates a combination therapy in one manufacturing process. Merus received the first European patent for the technology in June.
Engineered antibodies seek to improve the safety or efficacy of conventional antibodies or fragments through modification of glycosylation. In particular glyco-engineered antibodies containing low or no core fucose residues on the Fc N-glycan show improved affinity for Fc gamma receptor IIIa, and thus improve antibody-dependent cellular cytotoxicity—a critical factor in antitumor activity.
BioWa is one firm at the forefront of eliminating fucose from antibodies. The company’s Potelligent® technology enhances mAb activity, increases Fc binding, lowers the effective dose of an antibody therapeutic, and requires no change in the manufacturing process. In June, BioWa signed an agreement with Lonza to investigate the potential for reducing fucose residues in mAbs produced through Lonza’s GS Gene Expression™ system.
Improving mAbs through genetic engineering, manufacturing innovations, formulation, or some combination of the three is, of course, the idea behind “biobetters.” Formulation and preformulation are established strategies for improving prospects of development-stage small molecule drugs, for example by improving solubility.
The same is true for proteins and antibodies, notes Indu S. Javeri, Ph.D., CEO of CuriRx. “Preformulation is essential for mAbs, as it allows us to identify and understand the molecule’s strengths and weaknesses. This information, in turn, helps manufacturers develop high-yield purification processes and resolve manufacturing-related issues.”
Some enhancement strategies involve non-mammalian expression systems, others improve on CHO cells. All pose both advantages and disadvantages in terms of economics, productivity, and regulatory risk.
Some larger CMOs offer proprietary cell lines and expression systems for mAbs and other therapeutic proteins. But, according to Hetrick, proprietary technology “seems to add little value to the service offerings of smaller to mid-sized CMOs.”
He cites the limited reactor volumes and number of campaigns possible at such organizations, which would not support the infrastructure of proprietary expression systems. Since smaller CMOs operate mostly in Phase II and earlier, proprietary technologies would “not in most cases translate into a significant decision criterion for selecting that CMO. Moreover, CMOs tend to employ whatever cell line a client specifies.”
Time and Markets
So despite platforming and the familiarity it brings, the cutting edges of mAb development and manufacturing continue to evolve as this molecular category adapts to medical and market needs.
“Innovator companies are positioning themselves as developers of second- and third-generation therapeutics that address shortcomings of first-generation mAbs,” says Hetrick. “The claims for these molecules are impressive, yet it remains to be seen how they perform in the clinic and what additional developmental or manufacturing challenges they will pose.”
These challenges will be met, as they usually are. A more significant hurdle to innovation in mAbs, according to Hetrick, is early-stage funding. “Although there is an abundance of capacity, and pricing has not increased over the past couple of years, the number of underfunded early-stage clients remains constant. There is pressure for CMOs to do more for less and at some point you reach a wall in terms of acceptable margins.” Despite the “platforming” and mainstreaming of single-use process equipment, process development leading to a first clinical batch still takes 12 to 18 months, and success relies more than ever on resolving scientific and engineering issues.
Hetrick cautions developers to achieve phase-appropriate regulatory conformance and acquire sufficient clinical materials rather than pursuing the ultimate process during very early development. Organizations often spend too much effort attaining high-expression levels for Phase I and avoiding future process royalties, while the real value lies in the product and its performance in the clinic.
“The faster you can get there, the more value you build in a shorter period of time. I have literally seen projects get totally derailed for this reason alone. Sponsors need to approach risk, efficiency, cost, and regulatory compliance pragmatically.”
Regeneron’s Dr. Bailey concurs: “A key challenge continues to be achieving the right balance between speed-to-clinic versus having an optimized manufacturing process during clinical development, as well as the timing for making that transition.”
