Evaluation of matrix effects and prolonged storage on Trypanosoma cruzi serology in blood donor specimens

Blood products appropriately stored for research protocols provide an invaluable resource for amassing large numbers of specimens for clinical research, especially for low-prevalence diseases, such as Chagas disease.

We evaluated serologic results of 500 blood donation plasma component (PC) specimens confirmed as Trypanosoma cruzi seropositive by Food and Drug Administration-recommended algorithms. Subsets were retested using the T. cruzi enzyme-linked immunosorbent assay (ELISA; Ortho Clinical Diagnostics) and PRISM Chagas assay (Abbott Laboratories). Initial results for vacutainer-derived venous serum (VS) and PC specimens with matching results were also compared.

On initial testing, matrix effects between VS and PC were observed with ELISA demonstrating a mean change in the PC of -0.39 signal/cutoff ratio (S/CO) (p < 0.0001) and PRISM of +0.35 S/CO (p = 0.008). In matched PC specimens between current (retest) versus initial test results, both ELISA and PRISM had a decrease in mean S/COs of -0.76 (p < 0.0001) and - 0.90 (p < 0.0001), respectively. When the change in S/CO for matched PC specimens was analyzed as a function of time, PRISM showed no significant S/CO decrease (Y = -0.002941*X - 0.6250; p = 0.20; R² = 0.005), whereas the ELISA showed a significant S/CO decrease in more recently collected specimens (Y = 0.007183*X-1.516; p < 0.0001; R² = 0.06).

While T. cruzi serology results showed minor but significant differences in matrix effects between initial VS and PC testing values, and minor changes in PC test values over time, our data validate the use of PC specimens for head-to-head test performance comparison studies with the caveat that these limitations are assessed for appropriate study design.