Background: Our laboratory has recently identified prominent gene expression signatures in the blood cells of Systemic Lupus Erythematosus (SLE) patients, including an interferon signature that marks a subset of patients with more severe disease. In October 2001, we initiated a prospective study to enroll 300 patients from the Hopkins Lupus Cohort, with blood and urine sampling at frequent intervals over a one-year period of time. Extensive clinical data is available for each visit and will extend into the future. Initial enrollment is complete, and we currently have 1,354 lupus patient visits collected, and over 23,000 sample aliquots in our freezers at the University of Minnesota. Preliminary microarray studies on 112 Hopkins lupus samples suggest that the IFN signature and a plasma cell/immunoglobulin gene signature correlate with disease activity. We believe that these signatures, and perhaps others, will provide clinically relevant information about current visit disease activity, and may be able to predict future disease activity.

Hypotheses: The hypotheses guiding the experiments proposed here are that (1) gene expression signatures in the blood cells of lupus patients will be able to predict current visit disease activity better than the available measures, and (2) gene signatures will be able to predict future increases in disease activity.

Specific Aims and Study Design: The aims of the current proposal are as follows:

(1) From our primary microarray database in human SLE, we will identify the genes that best discriminate current visit disease activity as well as future development of active lupus. We will accomplish this by building models based on the available microarray data, and identify gene signatures that correlate with these outcomes. We will then develop and validate real-time quantitative reverse transcription (RT)-PCR (TaqMan) assays, and identify those genes and groups of genes that best represent the relevant global gene expression signatures.

(2) We will perform a Phase 1 study to measure the expression levels of ~20 genes (representing the relevant signatures) by TaqMan in archived, longitudinal samples from ~100 patients (~500 total samples). We will correlate data from the TaqMan signatures with clinical variables and outcomes, and use these data to confirm, and if necessary further refine, predictive algorithms.

(3) We will test the predictive algorithms developed in a Phase II validation study using archived, longitudinal samples from an additional 200 patients, for whom concurrent clinical data is available. Finally, we will modify the algorithms as required, and apply and test the next generation of algorithms to an additional 600 patients from the Hopkins Lupus Cohort who will be collected over the next 2 years.

Relevance: The current study proposes to identify gene expression signatures in blood cells of lupus patients that can serve as biomarkers for disease activity. Thus, the proposed study is very relevant to the lupus biomarkers topic area.

Systemic lupus is one of a variety of diseases affecting the immune system and connective tissues. It can be difficult to accurately assess disease activity in SLE, and most of the current laboratory tests lack sensitivity and/or specificity. Patients would benefit from more accurate and timely assessments of underlying disease activity, since therapies could be initiated faster. Thus, the work proposed here has the potential to fill an important gap in current management of this disease.