A Survival Benefit for Reduced Intensity Allogeneic Transplants from Young Unrelated Donors Compared to Older Sibling Donors Depends on the Graft CD8 T-Cell Content

Background: Younger donor age is associated with better survival in unrelated donor bone marrow transplants, but young unrelated donors have not shown an overall survival (OS) advantage over older sibling donors.  We previously reported that a high CD8 T-cell content of peripheral blood (PB) stem-cell grafts was associated with improved OS in reduced intensity alloSCT (Reshef et al, Tandem, 2014).  In this study we examined whether donor age correlated with graft T-cell content, and whether CD8 T-cell content may inform the selection of an optimal donor.

Methods: We studied 200 patients (pts) who underwent PB alloSCT after Flu/Bu2 conditioning (2007 to 2014).  Cumulative incidence and Cox regression analyses were used to evaluate associations between cell doses and outcomes, with adjustment for significant covariates.

Results

Patients: The median age was 62 (range 21-76) and diseases included AML (86), MDS (44), NHL (30), ALL (11), CLL (8) and others (21).  Pts were allografted from HLA-matched (86%) or mismatched (14%); sibling (42.5%) or unrelated (57.5%) donors.

Outcomes: A higher CD8 cell dose was associated with a decreased risk of relapse and improved OS.  The adjusted hazard ratios (aHR) per 1 x 10⁸/kg CD8 cells were: relapse 0.43 (95% CI [0.23 – 0.81]), p = 0.009; OS 0.57 (95% CI [0.33 – 0.97]), p = 0.04.  A cutoff of 0.72 x 10⁸/kg CD8 cells optimally dichotomized pts with differing OS.  Pts who received a CD8hi (>=0.72) graft had significantly improved OS (aHR = 0.53, p = 0.007; Fig 1).  The 1-year survival rates were 77.1% for CD8hi grafts and 50.2% for CD8lo grafts (p = 0.0002).  CD3, CD4 and CD8 doses were not associated with GvHD or NRM.

Donors: Donor age inversely correlated with CD8 cell doses (r = -0.44, p < 0.0001; Fig 2).  All donors over age 60 provided a CD8lo graft while 45% of donors under age 40 provided a CD8hi graft.  In recipients over age 60 (n=118; Fig 3), OS was better for patients allografted from young unrelated donors with a CD8hi graft compared to older sibling donors (p = 0.002) or young unrelated donors with a CD8lo graft (p = 0.002).  The OS benefit was driven by a reduction in relapse risk without significant differences in GVHD or NRM.  We then investigated whether a high CD8 cell dose could be predicted during donor screening.  We found that donors (n=23) whose grafts contained a higher CD8 cell dose had higher %CD8 cells (r = 0.69, p = 0.0004) and lower CD4/8 ratio (r = -0.55, p = 0.007) in a screening blood sample.

Conclusion: A higher CD8 cell dose is associated with improved survival after reduced intensity alloSCT.  Young unrelated donors with CD8hi grafts lead to better survival compared to older sibling donors. Donor selection based on predicted CD8 cell dose should be considered in prospective trials of reduced intensity alloSCT.

Fig. 1. CD8 cell dose and overall survival

Fig. 2. Donor age and CD8 cell dose

Fig. 3. Survival of recipients (age>60) is better with young unrelated donors and a CD8hi graft