Assessment of Impact of Hemolysis in Human and Nonhuman Serum Specimens on Multiplex Cytokine Quantification

Presence of hemolysis, the breakdown of red blood cells, in biological samples such as serum and plasma can negatively affect accurate quantification of many biomarkers including commonly measured clinical chemistry and cytokine parameters. Consequently, hemolyzed samples are frequently rejected as unsuitable for biomarker analysis. Currently, the potential interference of hemolysis on cytokine quantification is incompletely understood. Furthermore, the degree of interference with hemolysis may vary by species, cytokine, and quantification platform. Here, we evaluated the effect of various degrees of hemolysis on cytokine quantification accuracy in animal and human samples.

Frequency of hemolyzed samples in 339 internal animal studies was evaluated and analyzed for various degrees of hemolysis based on visual descriptors converted from H index values. Separately, hemolysis was simulated in vitro by a combination of freeze-thaw, osmotic shock, and shear stress using blood from mice, rats, dogs, cynomolgus macaque monkeys, and humans. Quantification of cytokines interferon gamma, interleukins (IL) IL-10, IL-1β, IL-2, IL-6, interferon gamma-induced protein 10, monocyte chemoattractant protein-1, and tumor necrosis factor-α was evaluated after spiking into hemolyzed and nonhemolyzed serum samples using Meso Scale Discovery and Bio-Rad Bioplex 200 platforms.

Comparison of cytokine levels in hemolyzed samples vs nonhemolyzed controls demonstrated that slight hemolysis (H index = 1 equivalent to 11-130 mg/dL of hemoglobin) did not interfere with the accuracy of cytokine quantification in most cases.

Evidence of slight hemolysis was present in most hemolyzed samples across species. We show that, in general, the cytokines can be accurately quantified in animal samples when slight to moderate hemolysis is present.