(T1430-01-04) A Sensitive Electrochemiluminescence-Based Method for the Detection of Anti-drug Antibodies against Anti-PD-1 Antibody Drug Pembrolizumab in Human Serum

Purpose: Targeting programmed cell death protein 1 (PD-1) with therapeutic antibodies such as pembrolizumab is a promising strategy for treating multiple cancer types either as a monotherapy or as part of a combinatorial approach. The therapeutic-specific anti-drug antibodies (ADA) may impact safety and efficacy of the therapeutics, so we have developed and validated a method to detect anti-pembrolizumab antibodies in human serum.

Methods: Quality control and test samples are incubated overnight in a 96-well plate with biotinylated- (Bt-) and ruthenylated (sTag-) pembrolizumab conjugates. The samples are then plated in duplicate wells of a streptavidin-coated Meso-Scale Discovery (MSD) plate. Any ADA that binds to both reagents will form a complex that is captured by the streptavidin-coated MSD plate and generate an ECL signal.  The ECL signal intensity is proportional to the amount of ADA present in the PC and test sample.

The bridging assay involves a 3-tiered approach:
a) A Screening Assay is used to determine signal intensity relative to pooled normal human serum (NC). Samples at or above the screening cutpoint (CP:1.09) are potentially positives for ADAs, and these samples require further testing in the confirmation assay.
b) A Confirmation Assay is used to determine the specificity of screening assay positive samples, as confirmed with incubation in the presence and absence of drug. Samples with a percent inhibition (%Inhibition) by the drug at or above the confirmation CP of 28.09% are confirmed positive for ADAs. These samples require further testing in the titration assay.
c) A Titration Assay is used to determine the relative amount of ADAs in the sample. Confirmed positive samples are diluted in pooled normal human serum to bring the response level across the titration CP of 1.40.

Results: Precision and accuracy of PC samples were evaluated and the result summary is presented in Table 1. The %CV of the ratio to NC of each PC was ≤20% at each concentration level in the screening and confirmation format. The assay is sensitive to 7 ng/mL PC. The selectivity of the method was evaluated by analyzing 10 lots of human cancer patient serum spiked with PC at a nominal concentration of 7 ng/mL. The result of the evaluation demonstrated an acceptable selectivity for the PD-1 antibody drug in human serum (Table 2). The screening false positivity rate was elevated in the absence of PC, but only one of the false positive samples tested positive in the confirmation assay format. Biotin present at 800 ng/mL in the samples does not interfere with PC at 7 ng/mL PC. Hemolysis and lipemia do not interfere with the bridging assay and no prozone effect was observed (Table 3).

Conclusion: A sensitive ECL method for detection of anti-pembrolizumab antibodies was developed and validated.


Table 1. Precision and Accuracy


Table 2. Selectivity


Table 3. Hemolysis, Lipemia and Biotin Demonstrate no Assay Interference