Shiroshita A, Oda Y, Takenouchi S, Hagino N, Kataoka Y: Accuracy of Anti-GBM Antibodies in Diagnosing Anti-Glomerular Basement Membrane Disease: A Systematic Review and Meta-Analysis. American Journal of Nephrology DOI 10.1159/000518362
Laboratory testing for anti-GBM antibodies has been available (for research or clinical care) for many decades. Yet standardization across laboratories is virtually non-existent and uncertainty remains concerning the overall accuracy of the multiple assays that are used. Western blot assays are the “gold standard” but have very limited availability. Turn-around time is very important in clinical utility.
In a very scholarly and comprehensive systematic review and meta-analysis of six studies involving 1,691 patients and 11 different anti-GBM assays, Shiroshita and colleagues have helped to clarify uncertainties concerning the accuracy of anti-GBM assays. Overall, the sensitivity and specificity of these assays were 93% (95% CI 84–97) and 95% (95% CI 94–99). However, an assessment of the overall quality of the reports was low, due to design features and the possibility of bias. The ELISA methods using the COLIV alpha 3 NC domain as the substrate were the most frequently used test (55%). Not surprisingly “false negative” tests were encountered (using immunofluorescence microscopy [IF] of linear IgG deposits as the defining finding of anti-GBM disease) up to about 16%. False positive tests were less common. The lack of standardization of “cut-off” values distinguishing “positive” from “negative” results was apparent. Misclassifications may have occurred since not all of the included studies employed a “gold standard” of IF on kidney biopsy. Nevertheless, the high specificity of ELISA anti-GBM means that a diagnosis of anti-GBM disease can be based on serology alone, especially in cases with a high a priori probability of anti-GBM disease. On the other hand, the lower sensitivity of anti-GBM antibody assays, including ELISA, suggests the need for multiple testing, such as indirect immunofluorescence assays, when ELISA is negative, especially when a high a priori probability of anti-GBM antibody disease is present. The inclusion of cases of “atypical anti-GBM disease” will only accentuate the possibility of “false negative” results for anti-GBM disease. This issue was not specifically addressed in the meta-analysis.