Infusion of memory T cell (CD45RA-depleted) DLI improves CMV-specific immune response early after αβT cell-depleted HSCT: first results of a prospective randomized trial
Zhanna B. Shekhovtsova, Maria A. Dunaykina, Larisa N. Shelikhova, Dmitry N. Balashov, Elena E. Kurnikova, Yakov O. Muzalevsky, Alexey S. Kazachenok, Elena Yu. Osipova, Viktoria V. Kiseleva, Dmitry E. Pershin, Daria A. Shasheleva, Rimma D. Khismatullina, Sergey L. Blagov, Andrei B. Abrosimov, Irina P. Shipitsina, Elena I. Gutovskaya, Galina A. Novichkova, Alexey A. Maschan, Michael A. Maschan
Dmitry Rogachev National Medical Centre of Pediatric Hematology, Oncology and Immunology, Moscow, Russia
Contacts: Dr. Zhanna B. Shekhovtsova
E-mail: shekhovtsova@fccho-moscow.ru
Summary
αβ T cell depletion effectively prevents severe GVHD in mismatched HSCT, but in a proportion of cases delayed immune recovery leads to increased infection risk and NRM. We’ve shown in a pilot study that infusion of low-dose memory T cells (CD45-RA depleted) is safe after engraftment among recipients of αβ T cell-depleted grafts (PMID:29269793). We initiated a prospective trial to directly test the efficiency of this approach. We report here an interim result of a prospective, randomized, single-centre trial (NCT02942173).
Patients and methods
A total of 100 pediatric patients with malignant disorders (ALL, n=56; AML, n=30; NHL, n=8; acute mixed-lineage leukemia, n=5 and MPD, n=1) were enrolled between October 2016 and September 2018. Patients were randomly assigned to receive CD45RA-depleted DLIs (experimental arm), n=54, or not (control arm), n=46. Median age at HSCT was 8.9 years, m:f ratio – 42:58. The conditioning consisted of either treosulfan (n= 50) or TBI (n= 50) in combination with fludarabine and thiotepa. GVHD prophylaxis included tocilizumab at 8 mg/kg at day 0, abatacept at 10 mg/kg at day 0, +7, +14 and +28, and bortezomib at 1.3 mg/m2 at days -5, -2, +2, +5. Neither anti-thymocyte globulin nor calcineurin inhibitors were used. Donors were HLA-haploidentical (n=94) or matched (n=6). All donors and 76% of the recipients were CMV seropositive. PMBC grafts were split and TCRαβ/CD19 depletion and CD45RA depletion were performed with CliniMACS Prodigy. The median dose of CD34+ cells was 10x106/kg, αβT cells – 28×103/kg. In the experimental arm memory DLIs were infused on day 0 at 25×103/kg and on days +30, +60, +90, +120 at 50×103/kg. In the control arm 8 patients received DLI after engraftment to prevent relapse (n=6) or treat infections (n=2). Primary endpoints were the cumulative incidence (CI) of CMV viremia (>500 copies/ml) by day +100 and the CI of acute GvHD grade ≥ II.
Results
Median follow-up for survivors was 1 year (0.2–2). Engraftment of WBC and platelets was achieved in 99 pts, one patient died at day +8. WBC and platelets engrafted at a median of 11 days and 14 days, respectively. The incidence of CMV viremia was 45% (36-56) overall, 41% (30-56) in the experimental arm vs 50% (38-67) in the control arm, with no significant difference. The CI of aGvHD ≥ grade II was 10% (6-18) overall, 10% (4-23) in the experimental arm vs 9% (4-24) in the control arm (non-significant difference). Two patients died, one per treatment arm, resulting in 2% (0-14) CI of TRM at 1 year among the whole cohort. Causes of death were pre-engraftment bloodstream infection and disseminated adenovirus infection. Patients randomized to experimental arm acquired anti-CMV reactivity significantly earlier, according to IFN-g ELISPOT assay on day +30 after HSCT (p=0.0001).
Conclusions
Co-infusion of donor-derived memory cells (DLI) with the αβ T cell-depleted graft is safe and improves recovery of virus-specific immune responses. Replacement of ATG with targeted blockade of CD28/CD80 costimulation and IL-6 receptor does not compromise engraftment and GVHD control, and is associated with low rate of non-relapse mortality.
Keywords
Leukemia, children, CD45RA, trial, prospective, randomized.
