Performance of polymerase chain reaction techniques detecting granzyme B in the diagnosis of acute renal rejection: a meta-analysis.

BACKGROUND:Studies have shown that granzyme B expression was up-regulated during acute renal rejection, which provides hopes for a noninvasive and reliable diagnostic method. However, the diagnostic performance has not been assessed systematically so far. METHODS:Relevant materials were searched from electronic databases before February 1, 2012. Clinical studies that reported the diagnostic performance of detecting granzyme B mRNA using polymerase chain reaction techniques in acute renal rejection patients were included. The data of true-positive, true-negative, false-positive, and false-negative cases identified by granzyme B detection were extracted from each data set of the relevant materials. The publication year, sample origin, mRNA quantification method, and housekeeping gene were extracted as potential confounding variables. RESULTS:Fourteen studies with a total of 514 cases were included. The overall diagnostic performance of granzyme B was pooled sensitivity of 0.76 (95% confidence interval [CI], 0.69-0.81), pooled specificity of 0.87 (95% CI, 0.83-0.90), diagnostic odds ratio of 19.27 (95% CI, 1.20-195.00), and area under the summary receiver operating characteristic curve value of 0.8900±0.0270. The univariate analysis of potential variables showed no statistically significant variations. CONCLUSIONS:Despite interstudy variability, granzyme B mRNA demonstrated both sensitivity and specificity in the diagnosis of acute rejection and showed consistency under circumstances of methodologic changes. However, the clinical practice requires careful consideration and other supplementary tests.