BioPharmaceuticals R&D, AstraZeneca, South San Francisco, CA 94080, USA.
BioPharmaceuticals R&D, AstraZeneca, South San Francisco, CA 94080, USA.
J Immunol Methods. 2020 Feb;477:112688. doi: 10.1016/j.jim.2019.112688. Epub 2019 Oct 30.
Biologics are potentially immunogenic and can elicit immune response. Complex biologics, such as bispecific antibodies or multi-domain molecules can induce anti-drug antibodies (ADA) with specificity to different domains. Domain specific ADAs may differently affect drug efficacy and safety, and thus, characterization of ADA domain specificity has become a regulatory expectation for multi-domain biologics. Unlike well-established methods for screening, confirmation, titer and neutralizing ADA detection, characterization of ADA domain specificity is an emerging field. The conventional approach for determination of ADA domain specificity is a competitive inhibition with domain-containing molecules. When developing a conventional domain specificity assay for moxetumomab pasudotox, a recombinant anti-CD22 immunotoxin, comprised of two functional domains (CD22-binding fragment and truncated Pseudomonas exotoxin A (PE38), we encountered a bioanalytical challenge. The method was able to detect immunodominant anti-PE38 (ADA-PE) but generated false negative results for low abundant CD22-binding domain ADA (ADA-BD) in a polyclonal sample. Troubleshooting experiments using control samples with varying levels of each ADA subtype demonstrated that a major factor for successful ADA identification was the ratio of the ADA signals contributed by each ADA subtype. To overcome this unique bioanalytical challenge, we developed a novel approach, which ensures detection of a domain-specific ADA subtype regardless of its relative level in a polyclonal ADA sample by evaluating signal inhibition by a respective domain-containing molecule at the condition when signals from all other ADAs are fully blocked. The method has been used for characterization of ADA domain specificity in moxetumomab pasudotox clinical trials, including study 1053, the pivotal Phase III study in hairy cell leukemia patients. It allowed for successful detection of ADA-BD in the presence of immunodominant ADA-PE, enabling accurate determination of domain specificity for moxetumomab pasudotox. The results demonstrated that the method was superior than the conventional approach. The method could be applied broadly to other biologics with two or more domains when there is a need to detect a minor ADA subtype in polyclonal samples.
生物制剂具有潜在的免疫原性，并可能引发免疫反应。复杂的生物制剂，如双特异性抗体或多结构域分子，可能会诱导针对不同结构域的抗药物抗体 (ADA)。针对不同结构域的 ADA 可能会以不同的方式影响药物的疗效和安全性，因此，对多结构域生物制剂的 ADA 结构域特异性进行特征描述已成为监管期望。与用于筛选、确证、滴度和中和 ADA 检测的成熟方法不同，ADA 结构域特异性的特征描述是一个新兴领域。确定 ADA 结构域特异性的传统方法是使用含有结构域的分子进行竞争抑制。在开发莫昔单抗帕斯妥昔单抗（一种由两个功能结构域组成的重组抗 CD22 免疫毒素，包含 CD22 结合片段和截短的铜绿假单胞菌外毒素 A (PE38)）的常规结构域特异性测定方法时，我们遇到了生物分析方面的挑战。该方法能够检测免疫显性抗 PE38 (ADA-PE)，但在多克隆样本中，对于低丰度的 CD22 结合结构域 ADA (ADA-BD) 产生假阴性结果。使用具有不同水平每种 ADA 亚型的对照样本进行的故障排除实验表明，成功鉴定 ADA 的一个主要因素是每种 ADA 亚型的 ADA 信号贡献的比例。为了克服这一独特的生物分析挑战，我们开发了一种新方法，该方法通过在所有其他 ADA 信号完全阻断的情况下，评估相应的含有结构域的分子对信号的抑制作用，确保无论多克隆 ADA 样本中特定 ADA 亚型的相对水平如何，都能检测到特定结构域的 ADA 亚型。该方法已用于莫昔单抗帕斯妥昔单抗的临床试验中的 ADA 结构域特异性特征描述，包括在毛细胞白血病患者中进行的关键性 III 期研究 1053。该方法能够在存在免疫显性 ADA-PE 的情况下成功检测到 ADA-BD，从而能够准确确定莫昔单抗帕斯妥昔单抗的结构域特异性。结果表明，该方法优于传统方法。当需要在多克隆样本中检测次要 ADA 亚型时，该方法可广泛应用于具有两个或更多结构域的其他生物制剂。
