Background/Purpose: Systemic lupus is a heterogeneous disease with diverse presentations. Studies have shown that dsDNA antibodies associate with renal disease. However, less is known about comorbidities in SLE patients without dsDNA antibodies. Using a large, novel electronic health record (EHR) cohort of SLE patients with a long duration of follow-up, we sought to identify not only ACR SLE criteria that associate with dsDNA antibody status but also other comorbidities that might not be assessed in cohort studies. We used a technique that scans across EHR billing codes called phenome-wide association study (PheWAS) to compare comorbidities in SLE patients with and without dsDNA antibodies.

Methods: We used our validated SLE algorithm of ≥ 4 counts of the SLE ICD-9 code (710.0) and ANA positive > 1:160 while excluding dermatomyositis and systemic sclerosis ICD-9 codes with an internally validated positive predictive value of 94% and a sensitivity of 86%. We identified SLE cases in a de-identified EHR called the Synthetic Derivative (SD) that contains over 2.8 million subjects with longitudinal data. SLE subjects have on average 9 years of follow-up. dsDNA status was defined as positive if ever positive, negative if there was at least 1 assay and all were negative, and measured via enzyme-linked immunosorbent assays with manufacturer values to determine positivity. Demographics of dsDNA positive vs. negative subjects were compared using chi-square and Mann-Whitney U tests. PheWAS was performed in dsDNA positive vs. negative SLE patients using logistic regression adjusting for current age and race and correcting for multiple testing using Bonferroni (p < 1.35 x 10^-4).

Results: Of 1097 SLE subjects, 521 had a positive dsDNA, 503 negative dsDNA, and 73 with missing data. dsDNA positive subjects were more likely to be African American vs. Caucasian (61% vs. 45%, p < 0.001) and younger at age of first SLE ICD-9 code (37 ± 17 vs. 43 ± 15, p < 0.001) with no difference in sex (female 51% vs. male 52%, p = 0.58). As expected, dsDNA positive subjects, compared to dsDNA negative, were more likely to have renal codes including nephritis, renal failure, and end stage renal disease (Table 1). dsDNA positive subjects were also more likely to have codes for hematologic and serositis criteria. dsDNA negative subjects were more likely to have codes for sleep, pain, and mood disorders.

Conclusion: Using a novel EHR technique in a large SLE cohort with longitudinal follow-up, dsDNA positive subjects were more likely to have codes for renal, serositis, and hematologic involvement. In contrast, dsDNA negative subjects were more likely to have codes related to neuropsychiatric symptoms. Our results demonstrate that PheWAS can expand our understanding of SLE disease heterogeneity by uncovering important clinical differences in subgroups of SLE patients.

Table 1.