Abstract
Background Antinuclear antibodies (ANA) play an important role in the diagnosis and classification of Systemic Lupus Erythematosus (SLE). Positivity of ANA by immunofluorescence (IIF) or an equivalent assay is part of the 1997 American College of Rheumatology (ACR) revised classification criteria and the 2012 revised criteria for the diagnosis of SLE from the Systemic Lupus International Collaborating Clinics (SLICC).
The aim of this study was to provide an overview of the sensitivity of automated fluorescence immunoassay (FEIA) versus IIF in SLE patients and consequently the number of false negatives.
Methods MEDLINE, EMBASE and Cochrane database searches (2000-March 2018) were conducted to identify fully-paired, cross-sectional or case-control studies of the diagnostic accuracy of FEIA and IIF in connective tissue diseases.
A meta-analysis of the sensitivity of FEIA versus IIF in SLE patients was conducted using hierarchical, mixed-effect models.
Methodological quality of included studies was assessed using the QUADAS-2 tool.
Results Out of 1798 papers identified, 11 studies (691 SLE patients) met the criteria for inclusion and reported the diagnostic accuracy of both IIF and FEIA. Four of the studies included consecutive patients (single-gate) and seven studies were case-control (two-gate).
Overall, sensitivity was 81% (95% CI 74%–86%) for FEIA and 88% (95% CI 79%–93%) for IIF, the difference being not statistically significant (p=0.14).
In the QUADAS-2 assessment, four studies included a positive ANA test as criteria for the diagnosis/classification of SLE which could introduce a high risk of bias in the overall sensitivity estimate. After excluding these four studies, the sensitivity point estimate for the remaining eight studies (431 patients) was 82% for FEIA and 81% for IIF (p=0.9).
Based on a 1.51% average prevalence of SLE in the single-gate studies, for every 1000 consecutive suspected patients screened, 15 patients will have SLE. Three out of fifteen will have a false negative test with FEIA versus two out of fifteen with IIF when sensitivity estimates from all studies are used. Three out of fifteen will have a false negative test with both tests when sensitivity estimates exclude the four studies that incorporated ANA results into the patient classification/diagnosis.
Conclusions Pooled sensitivity estimates of IIF and the automated FEIA in SLE patients was similar and not statistically different.
Taking into consideration the average prevalence of SLE in the single gate studies, both tests showed similar number of false negatives.
Funding Source(s): Carmen Andalucia and Sigrid Sjölander are Thermo Fisher Scientific employees. Michelle Orme was a Thermo Fisher Scientific consultant.
