(M1230-06-38) Evaluation of Free vs Total Measurement of Monoclonal Antibody-Based Therapeutics with Low Binding Affinity to Soluble Target

Purpose: The decision to use a free or total PK assay for full length monoclonal antibody (mAb) drugs specific for soluble targets remains a challenge and continues to be a debated topic. The measurement of free mAb may be considered more relevant if it is expected to correlate with pharmacologic effects. However, a strong understanding of key factors such as endogenous soluble target level in circulation, expected drug levels and available reagents for assay development need to be considered. Oftentimes, what mAb species could be measured rely on selection of reagents, sample dilution scheme, and binding affinity of mAb to soluble target. Therefore, it is imperative that bioanalytical scientists investigate assay conditions that ensure there is a clear understanding of the desired form of analyte to be measured. Standard approaches for a free PK assay include capturing with recombinant target or a blocking anti-idiotypic antibody (anti-Id-Ab). For a total PK assay, non-blocking anti-Id-Abs can be employed. Here, we present a case study for a low-affinity mAb (mAb A) with a relatively high dissociation constant (K_D = 9.6 E-08 M) and fast off-rate (half-life = 3 min). Due to the low affinity interactions of mAb A with target Y protein, a free PK assay may not be feasible as a potential dissociation shift of free mAb:bound mAb equilibrium prior to the measurement. Therefore, two PK assays measuring free or total mAb were developed and evaluated to select the most appropriate format.

Methods: Free PK assay: The PK assay was performed by ELISA. Briefly, 96-well Nunc Plates were coated with recombinant target Y protein. Next, standards, QCs and samples were diluted in assay diluent buffer before incubated in the plates. The captured mAb was detected with biotinylated anti-human IgG Fc, followed by NeutrAvidin-HRP and Pico substrate. To assess free mAb measurement in the presence of recombinant target Y protein, the mAb recovery of spiked pre-formed mAb A:target complexes in human serum was evaluated and compared to complexes pre-formed with other two higher-affinity mAbs and their relevant target proteins on the same assay format. Total PK assay: The PK assay was performed by ELISA. Briefly, 96-well Nunc Plates were coated with a blocking anti-Id-Ab. Next, standards, QCs and samples were pretreated with acid and neutralized to a mild basic pH before incubated in the plates. Direct dilution in assay diluent buffer served as control condition. The captured mAb was detected with a biotinylated non-blocking anti-Id-Ab that is non-competing with the capture anti-Id-Ab, followed by NeutrAvidin-HRP and Pico substrate. Effects of acid pretreatment, the final pH after neutralization, and acid incubation time on total mAb measurement were assessed. Target tolerance levels at LLOQ and MQC were determined in the final assay condition.

Results: The PK assay measuring free mAb (i.e. unbound mAb and partially bound) was developed. The mAb recovery of complexes at various molar ratios of mAb A to recombinant target Y exhibited target-ratio dependent decline and was similar in trend compared to complexes formed with other two high-affinity mAbs (Figure 1). However, at a target:mAb molar ratio of 20, 40% of free mAb was still detected in mAb A PK assay but not in mAb B or mAb C PK assays. It could be that the low affinity of mAb A shifted dissociation equilibrium during sample processing, leading to an artificial increase in the amount of measured free mAb, which could distort PK profile and subsequent exposure-response analysis. Therefore, development of a total PK assay measuring total mAb containing both free and bound mAb was explored. Acid pretreatment, titration of final pH after neutralizing acidified samples, and acid incubation time were evaluated (Figure 2). mAb complexes pretreated with 300 mM acetic acid increased recovery by 3 to 10% compared to no acid treatment. Extending acid incubation duration (e.g. 20 ± 10 mins) increased complexes recovery by 3 to 5% compared to shorter incubation time. We also evaluated impact of the final pH after neutralization on the mAb recovery of complexes. These results demonstrated that raising to a moderately basic pH helped reduce reformation of mAb complexes without affecting binding to the capture antibody. Based on the results, the total PK assay was validated and could tolerate at least 0.5 µg/mL target, which is 10-fold greater than endogenous target concentration in healthy human serum samples (Figure 3).

Conclusion: This study highlighted the unsuitability of free PK assay for measuring free mAb A due to its low affinity to target Y protein. However, the measurement of total mAb in the total PK assay was still impaired with the presence of target Y protein, due to the use of a blocking anti-Id-Ab as the only available capture antibody. Here, we showed that careful optimization, and appropriate reagents and assay conditions testing are required for free or total assay format determination.

Figure 1. (A) Target interference was tested for mAbs A (red), B (blue), and C (green) in free PK assay formats at lower QC levels (mAbs A and B at LLOQ level; mAb C at LQC level) using complexes prepared at the indicated molar ratios of recombinant target to the corresponding mAb. mAb A represents analyte of interest with weak binding affinity; mAbs B and C represent control analytes with intermediate and strong affinity, respectively. mAb recovery data of pre-formed complexes was normalized to a percentage of free mAb alone. (B) Biacore characterization of half-life constant and the binding kinetics of mAbs A, B, C to the corresponding recombinant targets at 37˚C.

Figure 2. mAb recovery of HQC (left) and LQC (right) complexes prepared at indicated molar ratios of recombinant target to mAb A using the total PK assay format. (A). Comparison of the mAb recovery of complexes with or without acid pretreatment. 3 mM Tris in no acid pretreatment condition was used to match the final neutralization pH in acid pretreatment condition. (B). Comparison of the mAb recovery of complexes after sample exposure to 300 mM acetic acid and neutralization with different concentrations of Tris to obtain different final pH conditions. (C) Comparison of the mAb recovery of multi-level QC complexes prepared at target:mAb molar ratio of 2:1 (left) and 5:1 (right) during the indicated acid incubation times between 10 to 40 mins using the total PK assay format. mAb recovery data of pre-formed complexes was normalized to a percentage of free mAb alone.

Figure 3. mAb recovery of MQC (A) and LLOQ (B) complexes with recombinant target protein at indicated concentrations. Samples were neutralized with 200 mM Tris after acid incubation in 300 mM acetic acid for 20 ±10 mins. Dashed line represents ±20 or 25% acceptance level. mAb recovery data of pre-formed complexes was normalized to a percentage of free mAb alone.