Background/Purpose: Myeloablative autologous hematopoietic stem-cell transplantation (HSCT) was recently demonstrated to provide benefit over monthly cyclophosphamide (CYC) in the treatment of diffuse cutaneous systemic sclerosis (dcSSc) (1).  As the dysregulation of the B cell compartment is implicated in the pathogenesis of SSc (2), we used antibody repertoire sequencing to characterize B cell repertoire dynamics over the course of HSCT and CYC treatment with the goal of identifying characteristics that differentiate responders from non-responders.

Methods: Immunoglobulin heavy-chain (IGH) sequencing from peripheral blood RNA was performed on HiSeq 2500 using an approach adapted from previously described methods (3). Fastq files de-multiplexed by sample were processed with MIGEC and MIXCR software to establish consensus sequences and identify unique clonotypes within each patient’s repertoire. Further analysis was performed using VDJtools, Immcantation, and IMGT-HighVQuest software. Samples yielding fewer than 3,500 sequences aligned to IGH genes were excluded from analysis. Event-free survival (EFS) at 54 months post treatment, characterized by no significant organ damage, was used to define treatment responders. The cohort included 13 HSCT recipients (8 responders), 14 CYC treated patients (7 responders) and 15 healthy controls. Responders were assessed at 3-4 time points (baseline and at 26 and/or 36 and/or 48 months post-treatment), non-responders at 2-3 time points (baseline and any time points leading up to and including EFS failure) and healthy controls at a single time point.

Results: Mean IGH repertoire diversity as assessed by the Inverse Simpson Diversity Index was equivalent in both dcSSc treatment groups and healthy controls at baseline (HSCT 1037±604.7, CYC 919±735.5, and healthy controls 1232±600.6). At last study time point, HSCT recipient responders exhibit an increase in mean IGH repertoire diversity compared to baseline (1100±578.9 vs. 1638±420.5, p=0.037, paired t-test) which was not observed in HSCT non-responders. Although in comparing last study point to baseline, there was no difference observed in CYC treated patients, CYC treated non-responders exhibit lower mean IGH repertoire diversity at last study time point than healthy controls (1232±600.6 vs. 551.2±365, p=0.012, unpaired t-test).

Conclusion: Increased mean IGH repertoire diversity from baseline is seen in HSCT responders.  CYC treated non-responders have decreased mean IGH repertoire diversity as compared to healthy controls.

References: (1) Sullivan KM et al. N Engl J Med 2018;378:35-47. (2) Yoshizaki A. Immunol Lett 2018;195:76-82. (3) Turchaninova MA et al. Nat Prot 2016;11:1599-1616.