Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement
The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.
Ancestral Differences in Associations Between Disease Manifestations and Serum IFN-
Weckerle1, Corinna E., Franek1, Beverly S., Kelly2, Jennifer A., Bruner2, Gail R., James2, Judith A., Harley2, John B., Niewold1, Timothy B.
Interferon alpha (IFN-a) is a primary pathogenic factor in systemic lupus erythematosus (SLE), and high IFN-a levels may be associated with particular clinical manifestations. SLE disease manifestations are highly variable between patients, and the prevalence of individual clinical features differs significantly by ancestry. Additionally, different clinical features in SLE may demonstrate significant associations with each other that may also differ by ancestry. We used logistic regression modeling to establish the network of associations between different clinical manifestations in large SLE cohorts of several different ancestries, and examined the correlation between ACR criteria and serum IFN-a to detect cytokine-phenotype associations in each background.
We analyzed data reporting presence or absence of ACR criteria as well as IFN-a levels from 724 SLE patients from the LFRR registry at OMRF. IFN-a levels were binned as a high vs. low categorical variable using a cut-off value of 2 standard deviations above the mean of healthy controls. The cohort was first stratified by self-reported ancestral background into 224 African-Americans, 105 Hispanic-Americans and 395 European-Americans. Iterative logistic regression was performed in each background using each of the ACR criteria and the IFN-a variable as an outcome variable serially, with the other variables used as predictor variables. Variables from this initial analysis with a p-value below 0.20 were then used in a repeat logistic regression, and results with p<0.05 in this analysis were considered significant. Positive and inverse correlations between different clinical variables were represented visually in a network diagram.
Of over 100 million possible associations between ACR criteria in our background stratified populations, we found 39 unique associations, forming network maps of relatively sparse density in each background. Of those, only 11 associations were shared by more than one different ancestral background. Using Fisher's Exact test, we found that this differed significantly from a model in which associations between clinical manifestations were shared by at least 2 of the 3 ancestral backgrounds (p<10-7). The network maps of interactions between clinical features were strikingly different in different ancestral backgrounds. IFN-a showed no common associations with clinical features in different ancestral backgrounds. In European-Americans, high IFN-a was linked to malar rash and hematologic manifestations of SLE. In African-Americans, it was linked to immunological manifestations. In Hispanic-Americans it was linked to oral ulcers, photosensitivity, and the absence of arthritis.
We found strikingly different associations between ACR criteria in different ancestral backgrounds, and IFN-a was associated with different clinical manifestations in each background. These data suggest that both the patterns of clinical manifestations and the underlying molecular pathogenesis of SLE differ significantly by ancestry.
To cite this abstract, please use the following information:
Weckerle, Corinna E., Franek, Beverly S., Kelly, Jennifer A., Bruner, Gail R., James, Judith A., Harley, John B., et al; Ancestral Differences in Associations Between Disease Manifestations and Serum IFN- [abstract]. Arthritis Rheum 2009;60 Suppl 10 :306