Patterns of Lymphocyte Subsets and Index of Bone Marrow Output (KRECs) Correlate Differently with Graft-Versus-Host Disease and Relapse

Introduction. Allogeneic haematopoietic stem cell transplantation (HSCT) is a potentially curative option for several haematological diseases. Its efficacy relies primarily on the Graft-versus-tumor (GVT) effect, which is promoted by donor immune cells. However, GVT partially overlaps with Graft versus Host disease (GvHD), the most common cause of morbidity and mortality in HSCT, since they may share immune effector cells and antigen targets, such as minor histocompatibility antigens. The development of a functional immune system is one of the main factors influencing the clinical outcome of HSCT. Immune deficiency as well as the effect of GVT/GvHD imbalance can expose patients to a high risk of opportunistic infections and disease relapse. Many studies analyzed immune reconstitution after HSCT both retrospectively and prospectively. However, immune parameters that univocally associate with GVHD or GVT have not been identified yet. In this study, lymphocyte subsets together with index of thymic and bone marrow output were evaluated at different time points, in order to identify possible indicators/predictors of GVHD and ineffective GVT.

Methods. Prospective evaluations of immune reconstitution were performed in 37 patients before SCT, at 30, 90, 180 days and 1 year after SCT. CD4+/CD8+ naïve, central memory, effector memory, terminally differentiated effector memory cells, subsets of regulatory T-lymphocytes, immature B cells, naïve, switched and unswitched memory B cells, memory double negative (IgD-CD27-) B cells were analysed by flow cytometry. Analysis of thymic and bone marrow output was performed by detection of T cell receptor excision circles (TRECs) and kappa-deleting recombination circles (KRECs). TRECs and KRECs were simultaneously quantified by a duplex quantitative Real-Time PCR. A 2-step multivariate analysis was performed using principal component analysis (PCA) and Cox regression analysis, to solve the problem of the high number of variables (immunological, patients- and transplant related) in comparison with the relatively limited and heterogeneous pool of patients.

Results. Twelve patients developed acute GVHD (median time: 28 days, range 17-120). Chronic GVHD was observed in 8 patients (median time: 6 months, range 4-8). In multivariate analysis, acute GVHD correlated positively with values of regulatory central memory T-cells (HR 1, 24; p=0,0006) and negatively with values of regulatory naïve T-cells (HR 0,52; p=0,006) at day +30. CD4+ effector memory T-cells at day +90 were positively associated with chronic GVHD (HR 1,3; p=0,04). The relapse rate (27%; median time: 5,5 months, range 3-12) was used as clinical index of ineffective GVT. The following cluster of immunological parameters at day +90 correlated positively with relapse: CD8+ effector memory cells, immature B cells, mature, naïve, switched memory B cells, memory double negative (IgD-CD27-) B cells, and KRECs (HR 2,4; p=0,006).

Discussion Different clusters of immunological parameters were evidenced as predictors of GVHD and ineffective GVT, allowing a clear-cut distinction between these immunological reactions. The non-univocal association of regulatory T-cells with GVHD may be explained by the relative efficiency of different subsets of regulatory T-cells (naïve>central memory regulatory), as shown in some experimental models. Increased values of CD8+ effector memory cells could be an early sign of ineffective GVL. Imbalance toward a lymphocyte B-response, and especially toward “senescent” memory (IgD-CD27-) B cells, could promote tolerance to tumor cells. The validation of these clusters of immunological parameters as specific early indicators of GVHD or GVT could potentially allow the development of pre-emptive and targeted therapies.