Acute graft-versus-host disease (aGVHD) occurs in approximately 20–50% of matched-related stem cell transplants (HCT). The presence of increased frequencies of CD4+CD25+ regulatory T cells (Treg) in murine donor grafts has been shown to ameliorate aGVHD. In similar models, several groups have demonstrated that the transfer of equivalent numbers of naïve T cells (vs. various memory subsets) confers a relatively greater risk for aGVHD. To test the hypothesis that natural variations in donor graft naïve and Treg content would be associated with the incidence of aGVHD in graft recipients, we conducted a detailed analysis of donor graft T cells within unmanipulated grafts of 70 SCT recipients transplanted for AML/MDS following the administration of a uniform fludarabine/busulfan conditioning regimen. Median follow up time was 38 months. 19% (n=13) of recipients experienced grades II–IV aGVHD (n=3 grades III–IV). Cryopreserved donor grafts samples were analyzed using 9-color flow cytometry (with a panel including a viability dye and MAbs recognizing CD4, CD8, CD25, CD127, CD45RA, CD27, Foxp3, and Ki-67). Tregs were defined as being CD4+CD25+CD127loFoxp3+. CD45RA and CD27 were used to measure naïve (CD45RA+CD27+) and memory conventional CD4+ and CD8+ T cells. In addition, the proliferating fraction of all cells was defined by Ki-67 nuclear-antigen expression. We conducted analyses based on donor graft T cells assessed both continuously and by quartiles, with similar results. No significant associations were found between aGVHD incidence and total Tregs (HR 1.01, 95%CI 0.9–1.2, p 0.8), nor with proliferating Tregs (HR 1.2, 95%CI 0.2–5.9, p 0.8). Since murine models typically fix the dose of Tregs vs. conventional T cells (Tcon), we also analyzed GVHD risk by donor graft Treg:Tcon ratio, finding no association (HR 1.03, 95%CI 0.8–2.1, p 0.3). Although patients in the highest Treg:Tcon quartile received Treg doses known in vitro to induce suppression (>1:21.5), there was no reduction in aGVHD (HR 1.5 Quartile 4 vs others, 95%CI 0.5–4.9, p 0.5). Additionally, no significant association with aGVHD was found with total naïve T cells (HR 1, 95%CI 0.9–1.0, p 0.9), naïve CD4+ (HR 1, 95%CI 0.9–1.03, p 0.6), or naïve CD8+ cells (HR 0.98, 95%CI 0.9–1.03, p 0.6). Because naïve:memory cell ratios were also fixed in murine models demonstrating the importance of naïve T cells in GVHD, we also assessed naïve:memory T cell ratios within donor grafts with respect to recipient aGVHD. No reduction in aGVHD was observed considering naïve:memory ratios analyzed continuously (HR 1.1, 95%CI 0.3–4.2, p 0.8) or in the quartile receiving the lowest naïve:memory ratio (HR 0.5 Quartile 1 vs others, 95%CI 0.1–2.2, p 0.3). Our results demonstrate that donor graft Treg and naïve T cell content, assessed in the context of unmanipulated PBSC transplantation in matched siblings, does not predict recipient aGVHD incidence. These data suggest that the therapeutic efficacy of Treg enrichment and/ or naïve T cell depletion may require grafts to be manipulated to supraphysiologic levels for maximal therapeutic benefit.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution