3X FLAG Peptide Exposure levels are often high in preclinica
Exposure levels are often high in preclinical safety studies, but if exposure is low or absent, a positive signal will potentially be lost in the CIC assay. The intended use of the CIC assays established by Stubenrauch et al. and Carrasco-Triguero et al. [8,12,19] were measurement of total ADAs, thus additional drug added to the study samples in an incubation step helped monomeric ADAs to form binding with the drug. In contrast, the aim of present assay were measurement of in vivo formed CICs and thus, no additional drug were added to the samples.
Determination of accurate sensitivity in these kinds of assays is challenging because of the heterogenic nature of a control that 3X FLAG Peptide contains different sizes of CICs. Therefore, the sensitivity given in ng/ml rather gives an estimate of the total mass of the CICs and might give the wrong estimate on the number of complexes present. Thus, the sensitivity given in nM can be more valuable. The sensitivity in the assay by Stubenrauch et al.  tended to be superior to the sensitivity of the present assay; however, assay sensitivity is relative to the positive control used  and comparing across assay types is not appropriate. As previously described, the assay developed by Stubenrauch et al.  was based on in-house produced chemicals and a different positive control. For non-clinical samples, no current guidelines provides recommendations on sensitivity for CIC assays, however, recommended sensitivity of standard ADA assays is 1000 ng/ml or better [4,20], which correspond to a sensitivity of 7 nM if the antibody has a molar weight of 150 kDa. Assay sensitivity in the present set-up was determined to 0.21–15.51 nM, which was considered acceptable.
When the samples from mice administered with hIgG1 were screened for the presence of ADAs using a standard bridging format, four out of 12 mice were identified to have a monomeric fraction of ADAs. ADAs were only detected in mice having low or absent hIgG exposure concentration in plasma, which suggest that the bridging ADA assay were less drug tolerant compared to the developed CIC assay. This is in line with a previous publication . In the present study, more mice were found to be CIC positive than ADA positive even though the calculated assay sensitivity for the CIC assay was inferior compared to the ADA assay; however, we were able to measure CICs below assay sensitivity in the analysed study samples. Taken together, these findings emphasize the challenges of determining accurate sensitivity in these types of assays.
CP estimation has been thoroughly described in various papers [4,15,21]. Traditionally, a floating CP is applicable if the means display homogeneity across the assay runs  and, hereby a plate CP is set by multiplying a normalization factor with the QCnegative. A refined approach reported by Devanarayan et al. , and also the method used in the present paper, describes a cutpoint set by S/N ratio, which equates to the normalization factor. This leads to a more robust estimation of a CP.
A limitation of the assay format is that ELISAs are not sufficiently sensitive for detection of low affinity antibodies, since multiple wash and incubation steps may cause the complexes to dissociate [22,23].
Conclusion The present study describes the detection of CICs formed by ADAs in complex with hIgG in a simple and easy to use generic ELISA set-up. Contrary to other methodologies, our set-up is based on commercially available reagents, including antibodies that are suitable for analysis of plasma samples from mice exposed to a hIgG antibody, independent on target specificity. We successfully showed the assays ability to detect CICs in mice administered with hIgG1 and hzIgG4. The described method is suitable for CIC screening in mice from safety studies.
Conflicts of interest
Acknowledgement This work was supported in part by Novo Nordisk A/S and the LifePharm Centre for In Vivo Pharmacology at University of Copenhagen.