A breakthrough novel method to resolve the drug and target interference problem in immunogenicity assays

Abstract

Biological matrix interference in detection and quantitation immunoassays remains a major challenge in the field of bioanalysis. For example, circulating drug may interfere with the detection of anti-drug antibodies (ADA) and drug target, or ADA may interfere with quantitation of drug levels in PK/TK analysis. Monoclonal antibody drug interference, especially for human IgG4 drugs, presents an additional challenge for ADA analysis due to its longer half-life and higher dose. Assay tolerance to such interference may depend on assay platform and reagents. Various approaches have been used to improve drug tolerance in ADA analysis but limited success was observed. We have developed a breakthrough novel method that uses Precipitation and Acid dissociation (PandA) to overcome drug interference in the ADA assay. The method principle is based on four components for detection of total ADA (free ADA and drug bound ADA) in the presence of drug in patient samples: (1) use excess drug to saturate free ADA to form drug bound ADA as drug:ADA complexes, (2) precipitate the complex using an agent such as PEG, (3) acid dissociate ADA from drug and immobilize (capture) free ADA (and free drug) under acidic conditions (without neutralization) onto a large capacity surface, and (4) detect free ADA (not the captured drug) using specific anti-human Ig detection reagent.

In this manuscript, we are describing case studies for three humanized monoclonal antibodies (an IgG1 and two IgG4 drugs). The three drug specific PandA ADA assays resulted in complete recovery of ADA in samples containing drug levels in excess of those expected in patients, in contrast to the commonly used acid dissociation approach in ECL bridging assays. This breakthrough novel method shows significant improvement over the current approaches. In fact, the drug interference or under detecting of ADA in all three cases was eliminated. This assay principle could be used not only for ADA assays but also PK and biomarker (drug target) analysis in the presence of interference factors.

Introduction

Drug tolerance remains a major technical challenge for immunogenicity assessment to monitor ADA as part of patient's safety monitoring which is critical to clinicians, manufacturers and regulatory agencies for drug clinical safety and efficacy. Interference in pharmacokinetic and pharmacodynamic assays has increasingly become an important area of interest [1], [2]. Despite some publications indicating that ADA to most protein therapeutics have no significant impact on their safety and efficacy [3], regulatory guidance documents (EMA and FDA) and white papers stress the importance of developing both sensitive and drug tolerant ADA immunoassays [4], [5] to better manage potential serious clinical consequences, adverse events and negative impact on efficacy. Porter et al. describes the ability to detect and measure antibodies to protein therapeutic to be a flawed analytical exercise [6]. The most widely adopted approaches in use currently still have limitations of timing, sensitivity or accuracy due to the presence of interference factors that are the same or resemble the binding partners in the ligand assay. These interferences include but not limited to drug interference in the ADA assay, ADA and/or drug target interference in the PK assay, and drug interference in the drug target biomarker assay. The commonly used ADA method is the bridging assay where a multi-valent ADA bridges between a capture drug (unlabeled or biotin labeled) and a labeled detection drug. This format is susceptible to endogenous drug interference causing false negative ADA results. Drug target interference may cause false positive results in the bridging assay format. Being recognized as one of the major challenges in the analytical method development field, many approaches have been used to mitigate this problem such as acid or base dissociation, competitive inhibition of interference using specific antibodies, removal of the interference factors, solid phase extraction with acid dissociation (SPEAD), affinity capture elution (ACE) and many others. The use of acid dissociation in a bridging assay has shown some improvement in drug tolerance for the detection of ADA [7], [8]. A similar improvement in drug tolerance was seen using the SPEAD [9] and ACE [10] methods. Although success has been described for these methods, each technology has limited utility for a broader application. A neutralization step typically follows the acid or base dissociation to allow binding partners to bridge causing the interference factor if still in solution to re-bind as well, maintaining the issue. Therefore, assay conditions such as pH and time for acid exposure need to be optimized to achieve maximal reduction of interference. Although ligand free method (such as LC–MS/MS) can be used for PK and drug target assays, enrichment of target analyte is often needed to improve assay sensitivity.

Although the above approaches demonstrated some improvement in drug tolerance, sensitivity and relative accuracy is not maintained therefore risking false negative and under-reporting of ADA incidences and titers in treated patients. Despite industry regulatory guidance documents and white papers recommending sensitivity between 250 and 500 ng/mL [4], [5], drug tolerance is sometimes evaluated without any acceptance criteria, and clinical protocols are then written instructing long wash-out periods before antibody measurement to allow for drug clearance and the avoidance of false negative results due to drug interference. This approach is not desired due to risks in missing ADA assessment in early time points especially in the case with a long half-life drug, multi-dosing regimen and the wash out period approach is not feasible.

Increasing the minimum required dilution and/or changing the sample dilution buffer (to have less percent of human serum/plasma) will negatively influence assay sensitivity and relevance to patient's samples. Industry scientists are constantly evaluating emerging technologies in a race to find a suitable solution to the interference problem without the negative impact on assay sensitivity [11].

Polyethylene glycol (PEG) precipitation has been described in the literature for characterization of circulating immune complexes [2], [12] and used in early ADA assays for insulin and insulin derived drug products using the radioimmunoprecipitation assays [13]. PEG precipitation of the target molecule or immune complex is size (or molecular weight, MW) based and PEG concentration dependent. The higher the PEG concentrations, the lower MW targets it will precipitate. To reduce the precipitation of non-specific serum proteins such as albumin and immunoglobulin, a low concentration of PEG is used to precipitate large MW Drug:ADA immune complexes. Using the principle of precipitation, coupled with acid dissociation and capturing on high capacity surface under acidic conditions (preventing the binding partners from re-binding), allows specific detection of ADA or drug or drug target using specific detection reagents. We believe this concept can be applied to broader applications for reduction or elimination of interference in immunoassays for ADA, PK, and biomarker (such as drug target) assays.

We developed a novel approach that is effective in reducing and eliminating the interference problems caused by drug or target in ADA detection. This approach was successfully applied to three monoclonal antibody therapeutics: Drug A is a humanized IgG1. Drug B is a fully human IgG4 and Drug C is a humanized IgG4. In this manuscript, we present three case studies to demonstrate elimination of drug interference in ADA assays for all three drugs.

For Drug A, the PandA method was compared to the traditional ECL bridging assay with and without acid dissociation to improve drug tolerance.

For Drug B, a specific challenge was seen in the ECL bridging assay with acid dissociation since the target for Drug B changes from a monomer to a dimer at low pH causing false positive results. The dimerization effect is seen in 100% of normal serum samples and disease baseline samples in the ECL bridging assay with acid dissociation. This phenomenon was similar to what was reported by Dai et al. where it was found that high apparent incidence of anti-drug antibody (ADA) in phase 1 studies was the result of detection of drug target, a homodimer, due to its ability to bridge drug molecules. Dai et al. found that the acid-dissociation-based pretreatment of samples used for mitigating drug interference dramatically increased drug target interference [14]. For Drug B, the assay sensitivity and drug tolerance was compared between the new PandA method and the existing ECL bridging assay without acid dissociation.

For Drug C, the assay sensitivity and drug tolerance was compared between the new method and the existing ECL bridging assay with acid dissociation where desired drug tolerance was not achieved.