ABSTRACT
OBJECTIVE: To compare sirolimus levels measured in whole blood using two analytical techniques: high-resolution liquid chromatography and microparticle enzyme immunoassay, and to evaluate whether hemoglobin, hematocrit, and time from kidney transplantation influence results obtained using the immune-enzymatic technique. METHOD: A retrospective, observational study in which all transplanted patients with at least one measurement of sirolimus levels using high-resolution liquid chromatography or microparticle enzyme immunoassay from October 2004 to May 2005 were consecutively included. For statistical comparisons simple linear regression, ANCOVA, intra-class correlation coefficient, and the method of agreement limits were all used. RESULTS: Ninety-one patients were assessed for a total of 307 measurements (median: 2, inter-quartile range: 1-4, range: 1-15) of sirolimus levels. The straight-line equation using the linear regression analysis was as follows: MEIA = 0.70 (95% CI: 0.39-1.01) + 1.14 (95% CI: 1.10-1.17) x HPLC/UV. The intra-class correlation coefficient between both measurements was 0.955 (95% CI 0.944-0.964). Mean overestimation using enzyme immunoassay was 24.8% +/- 19.4%. Difference in means between both measurements was 1.9 +/- 1.3 ng/mL. Agreement limits were established between -0.8 ng/mL (95% CI: -1.05; -0.55) and +4.6 ng/mL (95% CI: 4.35; 4.85). Factors such as post-transplant time, hemoglobin, and hematocrit did not influence overestimates obtained using enzyme immunoassays. These results were not influenced by non-independence in measurements. CONCLUSIONS: Despite enzyme immunoassay overestimates in establishing sirolimus levels in whole blood, its correlation with chromatography is acceptable. Added to its benefits versus chromatographic techniques, this renders enzyme immunoassay a good alternative for the measurement of sirolimus levels in whole blood.
