1
Overview of the Development of Biotherapeutics and the Role of Bioanalytical Sciences

Robert Dodge

The Novartis Institutes for BioMedical Research

1.1 Introduction

Bioanalysis, specifically the measurement of the concentration of biotherapeutics in biological matrices as well as the measurement of anti-therapeutic antibodies, has been critical to the development of biotherapeutics. In this chapter, a review of the critical role bioanalysis has played in advancing the development of biotherapeutics is outlined. While bioanalysis of small molecule or low molecular weight drugs has also played a critical role in development, this chapter is focused on biotherapeutics, defined here as protein-based therapeutics as well as novel modalities such as cell and gene therapies.

1.2 Pharmacokinetic Measurements

1.2.1 Initial Role of Bioanalysis in Pharmacokinetic Measurements

Starting with the introduction of the first biotherapeutics such as insulin, which involved extraction and purification of porcine or beef insulin for use in treating diabetes, bioanalysis has played a critical role in drug development. Some of the earliest and most critical aspects of the use of insulin as a biotherapeutic, such as formulation, were able to be addressed with accurate and precise bioanalytical measurements to drive understanding of pharmacokinetic (PK) properties. For example, initial extracts of insulin had extremely short half-lives, requiring multiple injections per day. The need to develop formulations with longer half-lives was apparent, and methodologies to measure the PK properties were understood to be crucial in development (Yalow & Berson, 1960). The first robust insulin bioanalytical immunoassay used to measure insulin concentration was reported in 1960 (Yalow & Berson, 1960). In fact, a major advance in the use of extracted insulin as a useful biotherapeutic was brought about by formulations with protamine and zinc resulting in an increased half-life and a reduction in the number of injections a patient needed per day (Quianzon & Cheikh, 2012).

While animal-extracted biotherapeutics such as insulin and growth hormone were important drugs, the biotherapeutic field began to progress at a rapid rate with the advent of molecules produced using recombinant DNA methodologies. Recombinant production of biotherapeutics involves the production of a protein therapeutic from DNA exogenously added to a host, such as Escherichia coli or a mammalian cell line, resulting in production of a protein not constitutively produced by the host organism. Along with the increase in the number of biotherapeutics developed using recombinant technology, this was also the start of the need for very accurate, precise, and robust bioanalytical measurements. The resulting bioanalytical data played a determining role in allowing the assessment of safety and efficacy during the biotherapeutic development process.

For Humulin, the first insulin produced by recombinant DNA methodologies, bioanalysis was critical in assessing key aspects of development such as comparability, safety, and efficacy and dosing. Since the development of Humulin, the bioanalysis of biotherapeutics and the resulting concentration and immunogenicity data focused on understanding exposure, PK parameters such as half-life, and comparability (e.g., bioavailability and biocomparability) have consistently provided data to allow critical decisions and understanding during the drug development process. This has been true since the first recombinant DNA biotherapeutics were approved over three decades ago, and bioanalysis remains a small but crucial component in biotherapeutic drug development today.

The first recombinantly produced protein biotherapeutics designed for use in humans were insulin (Humulin) in 1982 and growth hormone (Protropin) in 1985 (Drugs@FDA, 2022). Prior to that time, all protein therapeutics, as well as vaccines, had been obtained exclusively via protein extraction from animals or human blood. Both of these recombinantly produced biotherapeutics produced by recombinant DNA methodology had historical counterpart protein drugs extracted from animal (insulin) or human (somatotropin) counterparts, although somatotropin had been removed from the market due to concerns about the transmission of Creuzfeldt-Jacob disease (Underwood et al., 1985).

Insulin as a purified extract had been used as a therapeutic for diabetes since 1922, and many aspects of the dosing and PK were well understood (Quianzon & Cheikh, 2012). Thus there was the need for accurate and precise bioanalytical measurements to determine and compare in vivo concentrations of the recombinant insulin to understand any differences between recombinant insulin produced in E. coli bacteria compared to the previous products extracted from porcine pancreas. While today we understand and can characterize posttranslation modification, folding state, and correlate in vivo and in vitro activity of proteins, the understanding of any possible differences between recombinant and naturally produced proteins was not fully understood initially. Therefore, in order to ensure efficacy, dosing, and safety, bioanalytical measurements were a key component in verifying that recombinant insulin could be used in place of the extracted insulin products.

The bioanalytical assays put in place to understand insulin PK and immunogenicity were not so different to the concepts and methodologies that are still used today. Also since this first biotechnology-derived biotherapeutic insulin had been used in humans for decades as an animal-extracted product, research and a desire for accurate, sensitive, and selective methods for measuring and comparing the concentration of insulin were critical.

The procedure for developing the assay and the challenges to obtaining accurate, precise, specific measurements were not significantly different to those that the bioanalytical community faces today. Challenges such as the minimum required dilution, matrix interferences, and stability were all active areas of investigation in obtaining accurate and precise concentration measurements. Certainly examining the earliest development of bioanalytical assays for biotherapeutics concentration determination and contrasting this with current methodologies and challenges is an interesting and instructive exercise. It points to key areas for long-term solutions in biotherapeutic bioanalytical assay development, such that areas that were challenging in the 1960s and are still challenging today, six decades later, should lead to long-term changes in our approach to assay development today.

Yalow and Berson (1960) began development of their ligand binding concentration assay for insulin as would be done today for a ligand binding assay (LBA) for generation of antibodies for use as reagent or tools. They knew that high-quality reagents would be key to developing an effective LBA. They immunized guinea pigs with multiple rounds of beef insulin and isolated high-titer serum in responsive animals. Then, as today, this step of boosting animals with multiple doses of antigen in order to drive affinity maturation and high titer (concentration) of antibodies specific to the dosed antigen was well understood. The actual procedures followed to make polyclonal antibodies for LBAs today may vary in animal species (goat, monkey, rabbit, and even chicken), but the fundamental process and time required to generate polyclonal antibodies for assay reagents used by most bioanalytical scientists differs little from the original work. There are more modern methodologies for obtaining antibodies to analytes, for example, phage technologies such as HuCal (Knappik et al., 2000), that may significantly reduce the time needed to identify high-affinity antibodies, but this is not typically a default or widely used platform for reagent generation in the bioanalytical community and, as in many processes, biological generation of a tool reagent in animals is often more efficient than can be done ex vivo, or by human design.

The assay developed initially for insulin was a radioimmunoassay, with the same concept as a competition assay that may be used now with colorimetric or luminescence detection (Soeldner & Slone, 1965). However, some of the more sensitive methodologies such as fluorescence or luminescence probes were not available early in the development of sensitive LBAs, and so the use of radioisotopes, which are typically avoided in most bioanalytical laboratories today, was common in the late 20th century.

As with most LBA formats (ex sandwich format, competition format) in use in the modern bioanalytical laboratory, the next requirement in assay development is reagent labeling. This may involve antibody labeling with an enzyme, biotin, or, in the case of a competition assay, labeling the biotherapeutics. This is exactly the step taken in 1960, where the analyte, insulin, was labeled with 125I. Not unlike issues encountered in labeling today, reagents in a modern assay, and adjustment of label level, were challenges faced in obtaining maximum sensitivity and specificity (Nowatzke et al., 2017). There are some technologies today that allow for site-specific or highly efficient labeling of reagents (e.g., azide-alkyne coupling [Breinbauer & Köhn, 2003]), but again these methods are not in general use. Despite advancement in ease of use through labeling kits with prepackaged chemicals and spin columns, novel chemistries for precise labeling of proteins, and for bioanalytical LBAs, the fundamental process and strategies and methods for developing LBAs are fundamentally...