Cytomegalovirus Reactivation Promotes CD8+ T Cell Subset Recovery after Unmodified Allogeneic Hematopoietic Cell Transplantation

Introduction T cell reconstitution after alloHCT can be impaired and is associated with relapse as well as infections, especially viral reactivation. Our study is aimed at characterizing patterns of T cell subset recovery in patients undergoing alloHCT with an unmodified peripheral blood stem cell graft (PBSCT), using an expanded flow cytometric panel, and correlating T cell recovery with clinical outcomes, including cytomegalovirus (CMV) reactivation and acute graft-versus-host disease (GVHD). Methods We included patients who received first allogeneic PBSCT at our center between April 2012 and May 2016. Total CD3, CD4, and CD8 T cells and naive, central memory, effector memory (EM), and effector subsets were measured by 8 color flow cytometry. Flow data were censored after donor lymphocyte infusion, stem cell boost, cytotoxic T lymphocytes, or second alloHCT. Overall trends in recovery were highlighted using LOESS smoothed fits. Linear mixed effect models containing linear and cubic terms for time, random intercepts, and GVHD or CMV recovery were used to test for a difference across these groups. Results One hundred and twelve patients (median age 58 years, range 19-78 years) were included (Figure 1).  Figure 2 shows the recovery of CD3, CD4 and CD8 subsets for up to 24 months post-alloHCT. EM were the predominant population in CD4+ subset [median 46.4% (range, 1.6–83.9%)] while effector cells were the most prominent in CD8+ subset [median 41.8% (range, 7.0–94.1%)]. The incidence of CMV reactivation was 38% in seropositive patients (total 65 patients) with a median time to reactivation of 40 days (range 7-82 days). Patients with CMV reactivation by day +60 had a significantly higher expansion of CD8+ effector (p=0.004) and CD8+ EM (p=0.003) cells compared with patients without CMV reactivation by day +60 (Figure 3). There was no significant difference in the expansion of CD4+ T cells based on CMV reactivation. The cumulative incidence of grade 2-4 acute GVHD was 46% at day +100.  No significant difference was noted in CD4+ or CD8+ T cell expansion based on development of grade 2-4 acute GVHD by day +100 (data not shown). Conclusions Similar to data observed with other stem cell sources, CMV reactivation has a significant effect on post-alloHCT T cell recovery. This association may also provide an explanation for the observation in some studies that CMV reactivation is associated with lower rates of relapse and is being explored further in our dataset. Detailed flow-based analysis of T cell subset recovery after alloHCT can potentially provide additional biological insights to post HCT adverse events and guide strategies to decrease post alloHCT complications and non-relapse mortality.