Mature CD3+CD8+ and CD3+CD4+ T cells can be classified into CD45RA+CD62L+ naïve (TN) and CD45RO+ memory (TM) subsets, the latter of which includes effector memory (TEM) and central memory (TCM) T cells. Murine studies in MHC-matched and –mismatched models demonstrated that transplanting TN cells causes severe GVHD, purified TCM causes mild GVHD, and TEM do not cause GVHD. In vitro studies have similarly demonstrated that human donor CD8+ T cells specific for recipient minor H antigens are found predominantly within the TN cell subset. In sum, these data suggest that selective TNcell depletion may alter the incidence or severity of GVHD in human HCT.
We developed an effective process for engineering human peripheral blood stem cell (PBSC) grafts that depletes CD45RA+ TN and retains CD34+ stem cells and functional CD45RO+ TM specific for diverse opportunistic pathogens. We initiated a clinical trial to evaluate selective depletion of TN cells from allogeneic PBSC for the prevention of GVHD in patients with acute leukemia. Each of the first 22 patients has engrafted (median day 12), regimen-related toxicity has been acceptable, T-cell numbers recover faster than reported for TCD HCT (median 412 CD3+ T cells/μl on day 28), and there is no increase in the rate of relapse, opportunistic infections, or EBV reactivation compared to patients treated with T cell replete PBSC grafts. Early onset gastrointestinal symptoms that are compatible with mild acute GVHD occur frequently, but rapidly respond to therapy, and most patients have successfully tapered immunosuppression. T cell responses to pathogens recover early after HCT. At a median of 14 months follow-up, overall and disease free survival are 80% and 75% respectively, and the frequency and severity of chronic GVHD is substantially reduced compared to recipients of T cell replete PBSC grafts.