Validated detection of human anti-chimeric immune responses in serum of neuroblastoma patients treated with ch14.18/CHO

Human/mouse chimeric monoclonal antibody (mAb) ch14.18/CHO is directed against disialoganglioside GD2. Activity and efficacy of this mAb are currently determined in ongoing clinical Phase II and -III studies in high-risk neuroblastoma (NB). Based on the chimeric nature of this mAb, some patients may develop a human anti-chimeric immune response (Mirick et al., 2004) which impacts on pharmacokinetics and may induce anti-anti-idiotype (Id) mAb with a potential survival benefit. Therefore, a validated method of quantitative detection of human anti-chimeric antibodies (HACA) in serum samples of NB patients treated with ch14.18/CHO is an important tool for monitoring of clinical trials. Here, we report a validated sandwich enzyme-linked immunosorbent assay (ELISA) according to the one arm binding principle using ch14.18/CHO as a capture mAb and biotinylated ch14.18/CHO mAb for detection. Ganglidiomab, a monoclonal anti-Id Ab to ch14.18/CHO (Lode et al., 2013), was used as a standard for assay validation and HACA quantification. Systematic evaluation of the established ELISA procedure revealed an optimal serum sample dilution factor of 1:160. Assay validation was accomplished with a set of tailored quality controls (QC) containing distinct concentrations of ganglidiomab (3 and 15μg/ml). The coefficients of variation (CV) for all within-assay and inter-assay measurements using QCs were under 20% and the limit of detection (LOD) was 1.1μg/ml. Three patients (P1, P2, P3) treated with a 10day continuous infusion of 100mg/m(2) of ch14.18/CHO were selected for analysis with this assay. Selection was based on ch14.18/CHO drug level on day 8 in cycle 2 of >10μg/ml (expected) (P1) and of <2μg/ml (unexpected) (P2 and P3). Both patients with unexpected low ch14.18/CHO levels revealed a strong signal in the HACA ELISA. Interestingly, ch14.18/CHO-mediated complement-dependent cytotoxicity (CDC) could not be detected in P2 in contrast to P3 suggesting anti-NB activity even in the presence of HACA. We showed that neither eight freeze-thaw cycles nor storage at room temperature for up to 168h affected HACA stability in serum. In summary, we describe a validated ELISA method suitable for the assessment of HACA in NB patients treated with ch14.18/CHO.