Abstract 13788: Systematic Review: Utility of Human Immune Responses to Group A Streptococcus Antigens as a Diagnostic Indicator for Acute Rheumatic Fever

Abstract

Introduction: Previous studies have established that group A streptococcus (GAS) antibody titre is correlated with a diagnosis of acute rheumatic fever (ARF). However, results vary in the usefulness of GAS antibodies, particularly anti-Streptolysin-O (ASO) and anti-DNase B, in confirming a recent GAS infection. Therefore, we sought to provide, from published studies, an evidence-based synthesis of the correlation of streptococcal serology to establish the usefulness of immunological data in aiding the diagnosis of ARF. These findings are anticipated to have implications where echocardiography is not freely available, especially where ARF is rampant. Methods: We conducted a comprehensive search across a number of databases. Applying a priori criteria, we selected articles reporting on studies, regardless of study design, that evaluate the levels of antibodies against GAS-specific antigens in ARF subjects against control values or a published standard. Data were extracted onto data extraction forms, captured electronically and analysed using STATA software. Risk of bias was assessed in included studies using the Newcastle-Ottawa Scale . Results: The search strategy yielded 534 studies, from which 24 met the inclusion criteria, reporting on evaluation of titres for SLO (n=10), DNase B (n=9), anti-streptokinase (ASK) (n=3) amongst others. Elevation in titres was determined by comparison with controls and upper limit of normal (ULN) antibody values as determined in healthy individuals. Meta-analysis of case-controlled studies revealed moderate odds ratio (OR) correlations between ARF diagnosis and elevated titres for SLO (OR=10.57; 95% CI, 3.36 to 33.29; 10 studies) and DNase B (OR=6.97; 95% CI, 2.99 to 16.27; 7 studies). Conclusions: While providing support for incorporating SLO and DNase B in the diagnosis of ARF, we present the following reflections: an elevation in SLO and DNase B levels are not consistently associated with an ARF diagnosis; increasing the number of GAS proteins in the test is warranted to improve sensitivity; paired (acute and convalescent) samples could provide a more accurate indication of a rising titre. Use of community-based controls as a standard is not a reliable marker by which to gauge recent GAS infection