Background/Purpose: We have developed an antigen microarray technology to study antibody profiles to elucidate and diagnose clinical states of SLE patients – the iCHIP^® SLE-key^® assay¹. Our first product is the SLE-Key^® Rule-Out test with 94% sensitivity, 75% specificity, and a negative predictive value (NPV) of 93%². Here our goal was to determine whether particular antibody reactivities are associated with SLE disease activity index (SLEDAI) and can be used to monitor SLE disease activity.

Methods: We analyzed data from serum samples of 232 SLE patients (SLEDAI 0-25) tested on the iCHIP^® containing ~200 antigens. Intensities were measured for both IgG and IgM autoantibodies. Two types of analyses were performed. For a dichotomous analysis, we defined two groups of patients: low SLEDAI (SLEDAI ≤2 [n= 123]) compared to moderate-high SLEDAI (SLEDAI >2 [n=109]). For each of the 382 features, we fit a univariate logistic regression model and estimated the odds ratio (OR), indicating the relationship between binding to a specific antigen and the subject’s SLEDAI score. An OR significantly above or below 1 indicates a strong association between the feature and the dichotomous SLEDAI variable. We calculated the FDR-adjusted p-values for each univariate test, and constructed a FDR-adjusted confidence interval for the features with an adjusted p-value < 0.05 using a confidence level of 0.995. We also calculated the Pearson correlation coefficient between the intensity score for each feature and the continuous SLEDAI score.

Results: Thirty six individual antibody reactivities successfully separated between low and moderate-high SLEDAI groups in the dichotomous analysis. Nineteen of these features displayed OR>1 (mean 1.39±0.18), while the remaining 17 features displayed OR<1 (mean 0.53 ±0.08). IgG isotype was most prominent in both groups, while IgM isotype frequency increased in the latter group. Among the serologic reactivities with the largest OR were connective tissue antigens such as collagen III (IgG), vitronectin (IgG), laminin (IgG) and collagen IV (IgG) (OR 1.64-1.78); the reactivities with the smallest odds ratios were TNF receptor (IgG; 0.39) and GLP1 (IgM; 0.42). Autoantibodies to ssDNA (IgG), dsDNA (IgG and IgM), U1snRNP (IgG), Sm (IgG), and histones (IgG), which all showed significant OR, also significantly correlated with the SLEDAI score (correlation coefficients 0.46-0.57). In addition, 13 proprietary oligonucleotide sequences correlated with disease activity, with a median SLEDAI correlation coefficient of 0.44.

Conclusion: This initial proof of concept study shows that the SLE-key^® microarray can detect individual autoantibody reactivities associated with high or low SLEDAI scores. Based on these results we are using the iCHIP^®microarray technology to screen hundreds of antigens from relevant molecular pathways to establish a multivariate classifier for disease activity monitoring.   References: ⁽¹⁾Fattal et al; Immunology, 2010 ⁽²⁾Putterman et al., Journal of Immunological Methods, 2016   Acknowledgements: The authors wish to acknowledge the Innovative Medicines Initiative Joint Undertaking under grant agreement n° 115308 BIOVACSAFE.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/autoantibody-reactivities-correlated-with-sle-disease-activity-identified-by-the-sle-key-ichip-platform/