Two decades of immunocapture liquid chromatography tandem mass spectrometry for pharmaceutical discovery and development: reflections and recommendations for optimal deployment

1. Background

The use of immunocapture LC-MS/MS (IC-MS) methods to measure protein biomarkers is now over 20 years old. What started as an alternative to ligand binding assays (LBA) for protein biomarkers has now been used for multiple applications across the pharmaceutical industry [1–4]. Although initial applications focused on biomarkers, the use of IC-MS has expanded to include the bioanalysis of novel large molecule drug modalities [5–7].

Simply stated, IC-MS methods combine the power of immunoaffinity capture methods for sample preparation with advantages introduced by LC/MS/MS detection. As described by Neubert, et al., multiple formats exist for IC-MS [8]. Readers interested in the practical aspects of IC-MS analysis are also referred to a useful book chapter by Liu [9]. The present communication is not intended as a review of IC-MS methodology. Rather, it seeks to share lessons learned using IC-MS for pharmaceutical research and development and to put forth some basic principles to guide the application of this extremely powerful, albeit specialized, technology.

To understand the role and deployment of IC-MS, it is important to first review its strengths and weaknesses relative to LBA. A comparison of method attributes appears in Table 1 and is offered as a framework to discuss optimum IC-MS deployment. The comparison in Table 1 clearly illustrates the complementary nature of the two methods, underscoring the importance for successful interplay between the techniques either within or across bioanalytical laboratories.

Table 1.

Attribute comparison of ligand binding assays and immunocapture liquid chromatography tandem mass spectrometry methods.

Attribute	LBA	IC-LC-MS/MS
Sensitivity	High – an advantage to LBA which has improved with bead-based detection methods	Moderate – has steadily improved, but often requires larger sample volumes
Dynamic range	Low – assay dependent; sample dilution(s) often required	High – linear dynamic range is often 3 orders
Specificity	Low – depends on antibody fidelity and is even an issue for sandwich methods	High – an essential attribute of MS which allows detection at the proteoform level
Precision	Moderate – acceptable precision	High – owing to the use of SIL internal standards
Matrix Effects	Moderate – typically addressed by establishing an MRD	Low – immunoaffinity and internal standards combine to limit matrix effects. Can be used for tissues or non-serum matrices.
Quantification of mixtures	Poor – a known limitation	Good – use of SILs and chromatography results in a key strength
Reagent burden	High – method development limited by development and testing of antibodies pairs	Low – single antibody required which can have lower specificity
Throughput	High – multiplexed detection coupled with finite read time	Low – limited by LC run time and serial sample detection
Cost of deployment	Low – instrumental simplicity that is widely available	High – cost and required expertise currently limits wider implementation

IC-LC/MS/MS: Immunocapture liquid chromatography tandem mass spectrometry; LBA: Ligand binding assay; MRD: Minimum required dilution; SIL: Stable isotope-labeled.

2. IC-MS for protein biomarkers

A Special Report in Clinical Chemistry by Neubert et al. (2020) gave an update on the status of IC-MS for protein biomarker analysis [8]. In this report, applications were tabulated for two categories: clinical diagnostics and pharmaceutical/academic applications. While substantial growth has occurred in both application areas, significant barriers regarding clinical diagnostic adoption were acknowledged.

Mass spectrometry (MS) analysis offers two key advantages for all protein biomarker applications including clinical diagnostics. The first is the ability to simultaneously quantify multiple protein biomarkers in a single sample with reliable precision owing to the use of stable isotope-labeled (SIL) internal standards. Given that almost all clinical LBA assays involve a single marker, the ability to simultaneously measure related proteins allows for greater insights and certainty to be gleaned about a biological response or pathway. Most reported IC-MS assays involve small mixtures (<5 proteins) often connected to a common pathway. Representative examples include applications to incretins [10–12], amyloid-beta [2,13,14] and apolipoproteins [15,16].

Assays have also been developed for larger protein panels using mixtures of antibodies to capture either target proteins or their tryptic peptide surrogates after enzymatic digestion. Using the former approach, Krastins and co-workers showed excellent agreement for 16 clinically relevant proteins analyzed in human plasma both by IC-MS and a standard clinical analyzer [16]. The second approach, pioneered by Leigh Anderson and known as SISCAPA [17], was used to measure a panel of 89 proteins in human plasma [18]. More recently, SISCAPA was applied to measure 22 clinically relevant proteins in dried blood spots (DBS) in a study designed to show the ability of multiplexed analysis to allow longitudinal assessment of accepted clinical end points in individual patients [19].

A second, more powerful, advantage is the unique ability of mass spectrometry to discern protein specificity, a profound limitation that continues to negatively impact LBA methods. Although MS detection is not essential for viable clinical biomarker analysis, it is highly recommended that MS be employed, not only for biomarker discovery, but to understand the fidelity of LBA methods used in clinical decision making. Ultimately, this unique and powerful advantage is the key to sustained growth of IC-MS in the protein biomarker space.

Despite these advantages, clinical adoption of IC-MS remains limited by two factors: low throughput and high cost of deployment (Table 1). It is important for MS advocates to acknowledge the practical implications of these disadvantages and the fact that they are not likely to be overcome in the near future. Sustained growth in CLIA (Clinical Laboratory Improvement Amendments) [20] applications will still occur in areas deemed not viable by LBA methods and should be considered whenever justified by a specifically articulated clinical context-of-use. Ultimately, expanded growth awaits an understanding of the biological significance of individual proteoforms [21] to specific disease states. Because such understanding can only be derived or confirmed using MS tools, the role of mass spectrometry in future biomarker development is clear and foundational. Readers are strongly encouraged to follow developments in proteoform research and its relevant implications to both biomarkers and drug targets [21–23].

A final consideration for protein biomarker applications is the regulatory environment for CLIA-based protein biomarkers measured by IC-MS. Consistent with small molecule examples (e.g., vitamin D, testosterone), protein assays are likely to be introduced as Laboratory Developed Tests (LDTs), which are currently undergoing a transition to regulation by the FDA [24].

3. IC-MS for novel modality bioanalysis

A current, pervasive trend in the pharmaceutical industry is the move away from small organic molecules to large drugs including monoclonal antibodies, bispecific antibodies, Fc-fusion proteins, antibody-drug conjugates (ADC) and antibody-RNA conjugates (ARC). These modalities have been advanced for previously refractory drug targets and to limit toxicities often encountered with small molecules. Due to the size of these drugs, LBA methods are frequently employed, but often lack the specificity required to perform viable pharmacokinetic and biotransformation assessment [25]. IC-MS has dramatically abetted the development of new modalities by offering the capability for rapid method development alongside its acknowledged advantage of specificity. Although this trend may appear recent, it is worthy to cite a landmark paper by Kaur, et al. [5] which incorporated IC-MS (referred to as “hybrid” methods) for ADC bioanalysis as early as 2013. Indeed, this trend has proliferated over the past decade and encompasses several applications [26–28].

Concomitant with the advancement of protein-based therapeutics is the need for routine, reliable assessment of immunogenicity risk. While antibody drug antibody (ADA) assessment is immensely driven by LBA methods, the emergence of immunopeptidomics to identify antigenic peptide sequences is essential for designing safe, efficacious biotherapeutics and is an application that currently can only be performed by IC-MS [29,30].

4. Guiding principles for IC-MS

In response to the previous points made about the power and limitations of IC-MS for protein quantification, two rules are advanced to guide optimal deployment:

4.1. Rule 1 mass spectrometry is an important option, not an imperative

It is simply unrealistic to think that MS can be deployed universally for protein biomarkers or the support of large molecule therapeutic drugs. At the same time, access to MS is essential in both cases to understand biomarker specificity or to provide accurate assessment of drug pharmacokinetics and biotransformation. Regardless of the application, continuous interplay between IC-MS and LBA is strongly encouraged to forge the most cogent and nimble bioanalytical strategy possible. Ultimately, a transition from IC-MS to LBA using an appropriate cross-validation is typically required to supply acceptable throughput and cost for clinical applications.

4.2. Rule 2 measure what you must, not what you can

By far, the single most significant attribute that MS brings to protein quantification in biological matrices is specificity. While this advantage is routinely deployed in IC-MS biotherapeutic analysis, uptake for clinical protein biomarkers continues to lag [8]. Unfortunately, LBA methods face significant issues (e.g., antibody cross-reactivity, incomplete characterization of target protein standards, matrix effects) which can adversely affect the quality and certainty of the results obtained. Indeed, the impact of such factors needs to be understood in a fit-for-purpose manner; however, decisions impacting diagnosis, safety, efficacy and other critical decisions made using protein biomarker data require a higher level of certitude only available by incorporating MS tools in assay development. The degree to which historic clinical protein biomarker results have been impacted by these issues is assay specific and cannot be reliably ascertained. With that said, the scientific community must recognize that such problems are avoidable and can only be properly understood through application of MS technology. Clearly, investment in MS technology is essential in the quest to understand and harness the potential of specific proteoforms as biomarkers and drug targets. Although this does not imply that MS must be utilized for all subsequent clinical applications, proper cross-validation should be regarded as a clear imperative prior to the application of LBA for clinical decision making. Indeed, it is important to proceed with a firm understanding of what we are measuring.

5. Conclusion

The use of IC-MS for pharmaceutical and clinical applications will certainly grow in the years ahead. As mass spectrometry becomes more accessible in the future, practitioners need to focus on best practices for exploiting the core strengths of this technology to achieve maximum return on investment.

Financial disclosure

The author has no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.