Background/Purpose: A serious complication of Systemic Lupus Erythematosus is lupus nephritis (LN). Despite treatment, many patients still develop renal failure. It has been previously shown that serum based IFN-induced chemokine scores are associated with overall lupus activity. However, it is unclear what the association of these scores are to disease activity in LN patients and how they compare to non-LN SLE patients.
Methods: 224 participants from the Autoimmune Biomarkers Collaborative Network (ABCoN), with IFN regulated serum chemokine (MCP-1, CXCL-10, and CCL-19) scores and longitudinal follow-up, were included in this study. MCP-1, CXCL-10 and CCL-19 were measured using chemiluminescence sandwich-based immunoassays. Individual chemokine scores were calculated and made into a composite score (CK Score). The composite CK score was evaluated in its ability to predict disease flare by survival analysis in LN and non-LN patients, by categorizing levels into: high (top quartile), intermediate (2nd and 3rd quartiles combines) and low (bottom quartile). We then performed a subset analysis of 62 patients at times of remission and flare to assess the score's relation to disease activity using the Wilcoxon matched-pairs signed-rank test. Flare was determined as a change in baseline SLEDAI≥3.
Results: Of the 224 participants included in this study, 54.7% were White, 37.5% Black and 7.4% were of other Races. 89.1% were female; the mean age was 43.2 ±11.9y. Median disease duration was 8y (IQR4-13); 52 participants have biopsy-proven LN.
Baseline CK scores were modestly correlated with disease activity as measured by SLEDAI (rho=0.198, p=0.0226). There was no difference at baseline in CK scores for non-LN and LN patients (31.3 v 31.5, p=0.87). Overall, higher CK scores were predictive of flare in non-LN patients (Log-rank p:0.0014, Wilcoxon p:0.0019). In those with LN however, only the highest quartile predicted a flare when compared to low and intermediate values. (Log-rank p=0.021)
Next, for patients in remission who then flared, non-LN patients had significantly higher serum CK scores at flare than at remission (40.21 v 31.66, p=0.01). In patients with LN and renal activity during the flare (renal SLEDAI≥4), flare CK scores were also significantly elevated compared to remission scores (43.47 v 33.66, p=0.047). However median levels for both LN and non-LN patients at flare fell into the range of the intermediate quartile (24.18-46.65) that was not predictive of flare in non-LN patients.
Conclusions: CK scores are correlated with SLEDAI scores, where the highest quartile of baseline score is predictive of increased disease activity in both non-LN and LN patients. Nevertheless at flare median levels are lower than the highest quartile. Therefore, further studies are needed to determine adequate cut-off points of the CK score to have the best sensitivity and specificity to predict flare.
Disclosure: I. Blanco, None; C. Putterman, None; M. Petri, None; E. Baechler Gillespie, None.