Background Inability to distinguish between infection from increased SLE disease activity based on clinical judgment often compromises timely and effective treatment. Gene expression profiling of circulating leukocytes has been used to differentiate between malignancy subtypes and between sepsis and sterile inflammation from injuries. The IFN signature in SLE has been linked to disease pathogenesis; associations with disease activity remain controversial. A biomarker that precisely and rapidly distinguishes between active disease and infection would allow for more directed therapy, resulting in improved clinical outcomes.
Objectives RNA microarray analysis of peripheral blood leukocytes was used to identify gene expression profiles that differentiate between the host response to infection and increased disease activity in acutely ill, hospitalized SLE patients.
Methods 37 SLE patients with suspected infection or active SLE were recruited from 3 centers in New York, Mexico and the Philippines. Subjects were excluded for pregnancy, a history of infection with Hepatitis B/C, HIV or if they had received treatment for the acute illness. Whole blood was collected in Tempus™ Blood RNA Tubes and disease activity (SLEDAI) was measured at the time of enrollment. Infection determination was based on results of cultures or molecular tests. Others were defined as having active disease (flare). The Illumina HT12v4 platform was used for microarray analyses. Gene expression data were grouped using a modular analysis framework for blood genomics that has been applied to SLE previously (Immunity. 2008; 29:150-64). Clinical characteristics were summarized using appropriate descriptive statistics with correction for multiple comparisons. Statistical significance for microarray modules was determined using a hypergeometric test.
Results 31 subjects had adequate RNA for analysis; 19 of these met criteria for infection and subjects were grouped as either infection or flare. There were no significant group differences in age, disease duration, ethnicity, co-morbid states, history of prior CNS or renal disease or current medication use. Presence of low serum complement or high anti-dsDNA antibody titers did not distinguish between groups. SLEDAI scores were significantly higher with flare (75% with flare had SLEDAI≥12 compared to 15% with infection; p=.002). There was significant over expression of genes encoding immunoglobulin chains and CD38 (plasma cell module; p=.001) and IFN inducible genes (p<.000) in the flare subjects. Subjects with infection demonstrated significant over expression of genes encoding molecules expressed by cells of myeloid lineage (p=.018) and genes encoding molecules inducible by or inducing inflammation (inflammation II module; p=.012).
Conclusions We have identified gene expression signatures that associate significantly with either infection or disease activity. Not surprisingly, genes over expressed in flare are those associated with plasma cells or are IFN inducible. Infection related genes encode molecules such as FcγRIIA, CD86, CD163 and others associated with pattern recognition such as CD14, TLR2, MYD88, TNFR2 and BAFF. While SLEDAI scores correlate with disease activity, identification of a “sepsis signature” provides a more objective and reliable premise for treatment.
Disclosure of Interest None Declared