Molecular stratification by gene expression as a paradigm for precision medicine in systemic sclerosis

Abstract

Clinical presentation in patients with systemic sclerosis (SSc) is heterogeneous in terms of organ involvement and disease progression. Measurement of serum autoantibodies has provided a useful method for identifying patients at the highest risk for SSc complications including progressive interstitial lung disease, scleroderma renal crisis, and pulmonary arterial hypertension. Genome-wide transcriptional profiling of SSc skin has been used to measure and quantify SSc heterogeneity on the molecular level. This genomic approach has shown that at least four distinct gene expression subsets can be identified in SSc that permit a molecular subclassification beyond the well-described clinical subsets of limited and diffuse cutaneous SSc. Gene expression between clinically involved (lesional) and clinically uninvolved (non-lesional) skin samples show remarkable consistency demonstrating the systemic nature of disease. Each subset has its own distinct and consistent underlying deregulated molecular pathways, implicating different pathophysiological processes in different patient subsets. These data suggest that clinical trials will benefit by targeting therapy to the appropriate patient subset. This hypothesis has been tested in investigator-initiated trials of mycophenolate mofetil (CellCept) that targets lymphocyte proliferation and abatacept (Orencia) that prevents T-lymphocyte activation. In both cases, patients in the inflammatory subset showed clinical improvement, whereas patients in other subsets did not. Preliminary results with other therapeutic agents similarly suggest that only a subset of patients will improve during treatment. Molecular subset identification using a companion diagnostic will increase study validity and the likelihood of therapeutic benefit. Determination of the specific deregulated molecular pathways underlying SSc in individual patients will allow personalized precision medicine in SSc and will likely improve therapeutic responses.