Measurement of TNF-Alpha Drug Levels and Free Versus Total Anti-Drug Antibodies Using Three Commercially Available Assays
Introduction: Commercial assays are now available for therapeutic drug monitoring (TDM) of anti-TNF drugs and antibodies (ADAb). Utility of free versus total ADAb assays remains debatable, further complicated by lack of assay standardisation. Here we report analytical comparison of 3 commercially available assays for Infliximab (IFX) and Adalimumab (ADAL) drug levels (DL) and ADAb.
Method: Prospective evaluation of IFX and ADAL DL and ADAb was performed using our local LISA-TRACKER (LT) assay automated on e-Robot in IBD patients. Samples were also analysed by Immundiagnostik (IM, Germany) and Promonitor (PM, Spain) ELISA automated on Grifols Triturus. LT and PM utilises
a specific bridging ELISA to quantitatively measure free-ADAb whereas IM utilises a dissociation step to enable detection of total-ADAb generating semi-quantitative results. IFX assays measure free drug but differ in microtitre plate coating and secondary detection reagents. Data was analysed using Passing Bablok and bias plots. LT and PM kits were provided at no cost.
Results: Samples analysed in different batches showed different kit biases against each other for IFX. Batch 1 showed that LT assay had 43.9% positive bias against ID kit whereas batch 2 demonstrated –26% negative bias. Both LT and ID kits used had different lot numbers. This change in bias was not observed in ADAL assays which showed consistent and systematic bias. PM kit showed concentration dependent bias changes within the same assay. 4 patients tested (n = 79) IFX ADAb positive with undectable DL with one exception where total/free ADAb was negative using ID and PM assays. A further 17 patients tested total ADAb positive using IM with detectable DL (0.5–9.2 ug/mL). 1 patient tested ADAL ADAb positive using LT, PM and ID assays however ID and PM assays produced positive resuts on a further 4 specimens, all with ADAL DL >5 ug/mL.
Conclusion: Although commercial assays are now available, our data highlights the need for assay standardisation. Free ADAbassays were in agreement however ADAb positivity was higher using ID assay. Significance of total ADAb positivity is unknown. Results for both IFX and ADAL confirm that DL are not transferable from one assay to another and as such, common therapeutic cut-offs will not be applicable. Further work is warranted to establish the cause of batch-to-batch variation observed.
