CMV-Specific T Cells Restricted By Shared and Donor, but Not By Host HLA Molecules Reconstitute in the First 180 Days after Allogeneic HSCT and Protect from CMV Reactivation: Results of a Prospective Observational Study

Introduction: Cytomegalovirus (CMV) reactivation and disease are important risk factors after allogeneic hematopoietic stem cell transplantation (allo-HSCT), and strongly affect morbidity and mortality after transplant. CMV-specific T cell reconstitution controls CMV reactivation and protects against serious adverse events but a protective level of CMV-specific T cell response or standardized method for its monitoring have not been yet determined.

Methods: We designed a prospective, single-center observational study to assess if the kinetic and quality of CMV specific T-cell reconstitution impact the incidence and severity of CMV reactivations. We have enrolled 84 consecutive patients affected by hematological malignancies receiving allo-HSCT followed by Cyclophosphamide and Rapamycin between December 2017 and February 2019. Here we report preliminary data on the first 61 patients. Patients received allo-HSCT from family (siblings=10; HLA haploidentical=24), unrelated HLA-matched (n= 24) donors or cord blood (CB, n=3). The CMV serostatus of host (H) and donor (D) pairs was: H⁺/D⁺(n=40, 65%), H⁺/D^-(n=20, 33%) and H^-/D⁺ (n=1, 2%); H^-/D^-(7% of the overall transplanted population at our center) were excluded. CMV DNAemia was assessed weekly in whole blood (WB).

Absolute numbers of polyclonal and CMV-specific T cells were quantified by flow cytometry using Troucount™Tubes (BD) and Dextramer®CMV-Kit (Immudex), respectively, in the graft and fresh WB at days -7, +30, +45, +60, +90, +120, +150, +180 and +360. Dextramer CMV kit includes reagents for the identification of CMV-specific lymphocytes restricted for several HLA class I molecules: A*01:01/*02:01/*03:01/*24:02 and B*07:02/*08:01/*35:01. These alleles allowed the longitudinal evaluation of 54 out of 61 (89%) patients.

Results: At a median follow-up of 226 days post-HSCT, 31 (57%) patients experienced a CMV-related clinically relevant event (CRE, median +63 days), including 8 patients (15%) with CMV disease (median +59 days). Univariate analyses showed that the incidence of CMV clinically-relevant reactivation (CRE) was influenced by H/D CMV serostatus (0.90 in H⁺/D^- versus 0.44 in H⁺/D⁺pairs, p=0.015) and by previous acute Graft-versus-Host Disease (aGvHD) requiring systemic immunosuppression (0.82 in aGvHD grade II-IV versus 0.52 in aGvHD grade 0-I, p=0.051). The disease status at transplant, the donor type (HLA-matched versus HLA-haploidentical/CB donors), donor's or host's age did not significantly affect the probability to develop CRE.

For each time-point, we compared the absolute number of CMV-specific lymphocytes in patients experiencing or not a subsequent CRE. Our data demonstrate that higher levels of CMV-specific CD8⁺T cells in the donor apheresis and at +45 days after allo-HSCT are associated with reduced risk of subsequent CRE (median CMV-specific CD8⁺cells/kg in the apheresis=5x10³in CRE-positive patients (CRE⁺) and 5x10⁵in CRE-negative patients (CRE^-), p=0.012; median CMV-specific CD8⁺at +45 days=0.14 cells/μL in CRE⁺and 1.21 cells/μL in CRE^-, p=0.034). Furthermore, patients with any Dextramer positivity at +45 days displayed a lower incidence of CRE compared with subjects who were negative (CRE probability: 0.5 vs 1.0, p=0.003). Conversely, the absolute number of neither polyclonal CD3⁺CD8⁺T lymphocytes nor total CD3⁺T cells correlate with subsequent CRE. Taking advantage of the HLA mismatched-HSCT setting, we then dissected CMV-specific T-cell response according to HLA restriction elements (H/D=shared n=45, D-restricted n=14, H-restricted n=11). In H⁺/D⁺pairs, we observed a fast and similar kinetic of reconstitution of CMV-specific lymphocytes restricted by H/D and D HLAs. Conversely, in H⁺/D^-pairs, we detected only CMV-specific CD8⁺lymphocytes restricted for H/D haplotypes. Host-restricted cells remained undetectable for the first 180 days after HSCT.

Conclusion: Early after allo-HSCT and in the donor apheresis, the level of CMV-specific CD8⁺T cells measured by Dextramer staining differs in patients experiencing or not subsequent CRE. Furthermore, our findings indicate that CMV reactivations can prime H/D-restricted T cells presumably educated in the donor thymus; conversely, D- and H-restricted donor-derived lymphocytes have not yet undergone neither cross-priming nor thymic education respectively.