Allogeneic stem cell transplantation in children : identification and prevention of complications : adoptive transfer of EBV-immunity

Abstract

Allogeneic stem cell transplantation (SCT) can cure leukaemia, aplastic anaemia, and some inborn errors of metabolism and immunodeficiencies in children. As only 113 of patients who need SCT have an HLA-identical sibling, the use of alternative donors has increased, to over 50% of all paediatric SCTs at our centre. The immunological interactions and the clinical course are much more complex in such transplants. The aim of this study is to investigate and find means to manage some of the problems in paediatric SCT, particularly those inherent in matched unrelated donor (MUD) or partially mismatched donor transplants.

In a single centre study, we compared the outcome of all 59 children, who underwent MUD SCT, with case controls receiving sibling donor grafts. The 5-year probability of survival was 52 % for MUD vs. 77% for sibling recipients (p= 0.014). In ALL, the survival of the MUD (77%) and sibling group was equal. In SAA, survival was 43% vs. 86% (p=0.09) and in metabolic disorders 63% vs. 89% (p=0.025). The incidence of acute graft-versushost-disease (GVHD and the transplant related mortality (TRM) were higher in the MUD group, while eight relapses occurred in each group. The relapse rate was lower in children with chronic GVHD. These results support the use of MUDs when a sibling donor is not available. They also prompted us to examine further whether GVHD is associated with a graft- versus-leukaemia (GVL) effect, also in childhood acute leukaemia.

In all 169 children, who had SCT for ALL and AML at our centre, median time to relapse was 24 months in patients with chronic GVHD and 6 months in those without. The 5-year probabilities of relapse were 30 and 45% (p=0.01). Patients with chronic GVHD had a better survival, 77 vs. 51% (p=0.01). In a Cox regression model, chronic, but not acute GVHD decreased the risk of relapse (RR 0.44) and was predictive of an increased relapsefree survival (RFS) (RR 1.7), most apparent in late-stage disease and in ALL. This is in support of a GVL-effect in childhood leukaemia related to chronic GVHD.

To investigate further which factors impact on survival, relapse, TRM and GVHD we included all 181 children transplanted due to leukaemia at our unit. At the end of follow up 54% of the patients were alive, 27% had died in relapse, while 19% had died of other causes. The 5-year probabilities of acute GVHD grade II-IV, TRM, and RFS were 21%, 18% and 49%. Survival was similar in recipients of related (55%) and unrelated grafts (48%). In multivariate analysis, stage of disease (>CR 2) was an independent predictor for relapse and death. The risk of relapse increased significantly after 1992. A donor positive for 3-4 herpes viruses, increased the risk of acute GVHD, TRM and death, while a female to male transplant increased the risk of TRM, particularly in combination with a mismatch. HLA- matching independently improved survival, RFS and TRM. In children with leukaemia, an unrelated donor was not a risk factor for any of the five endpoints analysed. Instead, other donor characteristics such as HLA- matching, herpes virus serology, immunisation and sex were more important for outcome.

In mismatched donor SCT, the graft is T-cell depleted (TCD) to prevent life-threatening GVHD. TCD, however, increases the risk of Epstein-Barr virus (EBV)-associated lymphoma (PTLD), rejection and relapse. For six recipients at risk to develop PTLD, EBV-specific cytotoxic donor T lymphocyte (CTL) lines were generated by stimulation with EBV-transformed lymphoblasts, several weeks before SCT. Monitoring of the EBV-DNA load with semiquantitative PCR demonstrated that 4 of 5 recipients of TCD grafts and one Wiskott-Aldrich patient had a 4 to 5-log increase of EBV-DNA within 1-3 months after SCT, predicting a high risk of PTLD. In the other three recipients of unmanipulated grafts, the increases were moderate. Two to 4 infusions Of 1x107of the EBVCTLs / m2 resulted in a 2- to 3-log decrease of EBV-genomes in four patients and in stabilisation of the virus load on a moderate level in one case. One child, who received a T-cell culture lacking in EBV-specificity, progressed to fatal PTLD. The results suggest that a rapid increase of the EBV-load occurs in the absence of EBV-specific T-cell precursors, after TCD or in the presence of immunodeficiency. Infusion of EBV-CTLs early after SCT appears to prevent EBV-PTLD.

To investigate the increased EBV-DNA load and the state of viral latency in these patients, we analysed EBVDNA in serum, applied limiting dilution analysis of EBV-DNA in mononuclear cells and RT-PCRs for viral gene expression . The increased virus load is suggested to be due to an expansion of a latently infected B-cell compartment that contains less than 10 EBV genome copies per cell and expresses EBERs, LMP-2A and occasionally LMP 1 and EBNA 1. This is compatible with latency forms I-II in non-proliferating B-cells, but not latency Ill. Most likely this is due to proliferation of EBV+ B cells somewhere in the lymphatic compartment, outside peripheral blood, as we did not find evidence of active replication of virus.