Abstract Background Biomarkers have become an integral part of drug development and decision-making. The quantification of cytokines, commonly used to gain insights into pathological processes in indications such as Inflammatory Bowel Diseases (IBD), and for monitoring therapeutic intervention, should rely on robust and performant methods. Relevant reference samples are crucial for evaluating the performance of biomarker assays implemented in the context of clinical studies, but oHow to evaluate the parallelism of a method for biomarkers that are non-detectable in healthy donors (HD)? oDoes the selectivity of the recombinant protein in the biological matrix reflect the endogenous biomarker? oHow does the stability of the recombinant protein compare with endogenous biomarker? oCan you use incurred samples (informed consent, sufficient volume) for evaluating parallelism and stability? Methods To address these questions, we generated samples with endogenous biomarkers, named MAQC, by spiking supernatants of stimulated PBMC in the serum or plasma from HD. Three to four levels of MAQC per cytokine were prepared and used to assess parallelism, selectivity and long-term stability (LTS) at -80°C of pro- and anti-inflammatory cytokines and chemokines by use of V-plex, S-plex kits from MSD and ELLA kits. For all these parameters, accuracy was determined, with a target acceptance of 70-130% of the nominal value for selectivity and LTS or upstream dilution for parallelism. Results We were able to generate serum and plasma reference samples with low- to high-range levels of Th1 (IFN-g, IL-12p70, IL-2, TNF-α, CCL-2), Th2 (IL-4, IL-13, IL-10), Th17 (IL-17A) and Proinflammatory (IL-6, IL-8, IL-1β) cytokines. Selectivity was acceptable for most cytokines, except for IL-13, IL-4, IL-17A and IL-6. For IL-6, a major interference was observed in some samples, likely due to the presence of soluble IL-6 receptor in the original serum sample. For the other 3 cytokines, interference, leading to a bias close to -30%, was observed that could not be lifted by diluting samples. For all cytokines but IL-13, endogenous diluted similarly as recombinant protein from calibrator. Twelve months LTS was demonstrated for all cytokines in MAQC and comparable to recombinant proteins spiked in human serum. Conclusion To conclude, reference samples with endogenous cytokines/chemokines can be produced from primary immune cells. These can be used to assess parallelism and long-term stability when sufficient volume of matrices with high enough concentrations are not available. These would be even more valuable if they could be commercially available for use as quality controls in validation and clinical studies.
Read full abstract