Thrombophilie-Screening und (semi-) automatisierte Diagnostik von Antiphospholipid-Antikörpern im ambulanten gynäkologischen Bereich

Anti-cardiolipin und β2-glycoprotein I Antikorper prasentieren wichtige Parameter in der Routineimmundiagnostik. Wir haben erstmals funf aktuelle verschiedene Analysegerate einander gegenuber stellen konnen. Methoden: Es wurden 162 Proben von Abortpatientinnen und 43 Proben von Patienten mit diagnostiziertem systemischen Lupus erythematodes untersucht. Ergebnisse: Fur Anti-cardiolipin und β2-glycoprotein I zeigte sich anhand der positiven Ergebnissanzahl der verschiedenen Gerate ein zu vermutender Unterschied. Bei Betrachtung der einzelnen Proben und Cohens Kappa erwiesen sich deutliche Unterschiede. Schlussfolgerung: Wir fanden eine masige bis schlechte Ubereinstimmung zwischen den funf aktuellen Analysegeraten fur Anti-cardiolipin und β2-glycoprotein I Antikorper. Ein solcher Unterschied der Ergebnisse kann zu Fehlinterprtationen als auch zu Fehldiagnosen fuhren. Es sollte eine Standardisierung der Antiphospholipid-Antikorperbestimmung angestrebt werden. Objective: To determine the sensitivity, specificity and accuracy of automated, semi-automated and manual methods for the detection of anticardiolipin and β2-glycoprotein I in women with miscarriage. Design: Prospective clinical trial evaluating diagnostic concordance of the test methods. Setting: Samples were recruited from 99 different G&O outpatient centres located in Germany and rheumatology centers. For laboratory analysis, peripheral blood was shipped to the AescuLabor hemostaseology laboratory at Hamburg, Germany. Sample: A total of 162 samples were analyzed obtained from women at with a history of miscarriage and 42 samples of lupus erythematodes patients. Methods: Manual, modern automated and semi-automated enzyme immunoassay systems were applied for the detection of antiphospholipid and anticardiolipin antibodies. Main outcome measures: Accordance of test results obtained by analytical methods. Results: Amongst the five different methods applied for the measurement of IgG/IgM anticardiolipin and IgG/IgM β2-glycoprotein I antibodies percentual differences were seen for positive results in all four parameters. Itemization of the positive results showed a poor concordance for the positive patient’s samples between the methods tested. These findings were confirmed by Cohen’s kappa coefficients. Conclusion: Comparison of manual and current (semi-) automated methods show a moderate to poor comparability of IgG/IgM anticardiolipin and β2-glycoprotein I antibody results in samples obtained from women with miscarriage. These deviations impede clear clinical and therapeutic decisions in patients.