Skin Biopsies Enhance Prediction of Clinical Trajectory in Diffuse Cutaneous Systemic Sclerosis

Abstract

Objective: To evaluate the relationship of skin fibroblast CD34 and α-smooth muscle actin (α-SMA) and immune cell infiltration with disease duration in diffuse cutaneous systemic sclerosis (dcSSc) and identify predictors of improvement.

Methods: Skin biopsies and clinical data were analyzed from patients with dcSSc enrolled in lenabasum (n = 79), belimumab (n = 18), or nilotinib (n = 8) trials. CD34 and α-SMA were scored semiquantitatively. Immune cells (CD20+, CD3+, and CD123+) were counted. Clinical and histologic features were compared between those with early (<18 months) versus later disease (≥18 months). Clinical improvement was defined as >5-point decrease in modified Rodnan skin score (mRSS) at 52 weeks. An Akaike's information criterion-optimal multiple logistic regression model for clinical improvement among 68 mycophenolate mofetil (MMF) users was developed. Fibroblast spatial organization was visualized using imaging mass cytometry (IMC) in a representative sample.

Results: Despite similar baseline mRSS, early disease was associated with lower CD34, higher α-SMA, increased B cells, and greater mRSS improvement compared with later SSc. IMC demonstrated regions of CD34+ fibroblasts in superficial dermis and α-SMA+ fibroblasts in deeper, collagen-dense regions. Among MMF users, high CD34 and α-SMA predicted improvement in early SSc; however, high α-SMA predicted lower odds of improvement in later SSc. The probability of improvement was lowest in early SSc with α-SMA^low/CD34^low immunophenotype and later SSc with α-SMA^high. In later SSc, B cells were higher in α-SMA^high versus α-SMA^low skin.

Conclusion: Fibroblast immunophenotype varied by baseline disease duration and improved model performance for identifying those with improvement on MMF. Skin biopsies may be useful for refining prognosis and guiding patient management decisions.