Comparison of antithrombin activity assays in detection of patients with heparin binding site antithrombin deficiency: systematic review and meta-analysis

Abstract

Antithrombin (AT) deficiency increases the risk for venous thromboembolism, therefore, a highly sensitive assay to identify this condition is crucial. The aim of this paper was to perform a meta-analysis comparing AT activities measured by different AT activity assays in patients with heparin binding site AT deficiency. In addition, the diagnostic sensitivity of selected assays was compared depending on the available data. An extensive literature search was performed considering results with publication date up to July 10, 2021. Seven relevant English-language observational studies, comparing AT activity measured by different AT activity assays in Caucasian Europeans with either the AT Budapest III or AT Padua I mutation were included in meta-analyses. There was no significant difference in AT activity between Labexpert and Innovance in patients with AT Budapest III (P = 0.567) and AT Padua I (P = 0.265), while AT activity determined by HemosIL was significantly higher compared to Innovance for both mutations (AT Budapest III: P < 0.001; AT Padua I: P < 0.001). These results are in line with the results of comparison of diagnostic sensitivity. In patients with AT Budapest III, the AT activity was also higher when measured with Berichrom compared to Innovance (P = 0.002), however, the results of comparison of diagnostic sensitivity across studies were variable. No significant difference (P = 0.117) in AT activity as well as diagnostic sensitivity was observed between Sta-Stachrom and Innovance. The results of our study suggest that Innovance, Labexpert and Sta-Stachrom are the most sensitive activity assays for detection of AT Budapest III and AT Padua I, whereas HemosIL showed considerably lower sensitivity for these two variants. As revealed in our study, the diagnostic sensitivity of AT activity assays to type II heparin binding site AT deficiency is different, and in some assays mutation dependent.

Figure 1

PRISMA flow diagram of included studies.

Figure 2

Meta-analysis comparing AT activities in patients with AT Budapest III measured with different assays and compared to Innovance: (a) Berichrom, (b) Sta-Stachrom, (c) Labexpert and (d) HemosIL. n number, M mean, SD standard deviation, SMD standardised mean difference, CI confidence interval; A. Selection bias; B. performance bias; C. detection bias; D. reporting bias (evaluated according to the Table S8).

Figure 3

Meta-analysis comparing AT activities in patients with AT Padua I measured with different assays and compared to Innovance: (a) Labexpert, (b) HemosIL. n number, M mean, SD standard deviation, SMD standardised mean difference, CI confidence interval; A. Selection bias; B. performance bias; C. detection bias; D. reporting bias (evaluated according to the Table S8).

Figure 4

Mean differences (± SD) in AT activity between individual AT activity assays and Innovance in patients with AT Budapest III and AT Padua I as determined in different studies.

Figure 5

Diagnostic sensitivity of individual AT activity assays to either AT Budapest III and AT Padua I as determined in different studies meta-analysis comparing diagnostic sensitivity between Innovance and Labexpert (a) and HemosIL (b) in patients with AT Padua I.

Figure 6

Synthesis of the results of studies with regard to mutation and AT activity assay.