CCR 5 gene polymorphism affects the risk of Gv HD after haematopoietic stem cell transplantation from an unrelated donor

The main barrier to successful haematopoietic stem cell transplantation (HSCT) is the development of post-transplant complications. Although human leucocyte antigen (HLA)-matching is critical in both HLA-matched familial and matched unrelated donor transplants to deter acute graft-versus-host disease (aGvHD) and rejection, recently functional non-HLA immune associated genes have also been considered in attempts to evaluate their potential prediction values and uncover novel factors that may optimize donor selection processes. Mutations and polymorphisms within these non-HLA encoded genes affect, for example, the amount of cytokine/chemokine produced in response to allo-antigen or infection. Knowledge of both patient and donor non-HLA genotype may therefore aid the development of new preventative and therapeutic strategies by taking the degree of ‘risk-associated’ genotype into account. The results of our present work contribute to these studies and strongly suggest that the 32-nucleotide deletion within the CCR5 gene (CCR5Δ32 polymorphism; rs333) is of prognostic value for the outcome of HSCT from unrelated donors. The present study investigated the CCR5 polymorphism in relation to transplant outcome in 360 patients (Table 1) transplanted in seven Polish institutions and their unrelated donors. The CCR5Δ32 polymorphism was analysed by polymerase chain reaction as described previously (Bogunia-Kubik et al, 2006, 2007). Written informed consent was obtained from each patient. The study was approved by the ethics committee of the Medical University in Wroclaw. In univariate analyses, recipients homozygous for the 32 bp deletion suffered more frequently from severe aGvHD than patients lacking this mutation (5/10 vs. 34/313, P = 0·001 and 2/10 vs. 15/313, P = 0·034, for grade III–IV and IV aGvHD, respectively). Also patients grafted from CCR5Δ32 homozygous donors were more likely to develop aGvHD (3/8 vs. 37/318, P = 0·028 and 3/8 vs. 15/318, P = 0·001 for grade III–IV and grade IV aGvHD, respectively). The CCR5Δ32 polymorphism was not found to affect the development of other complications except for observed worse overall survival of patients homozygous for this deletion (12% vs. 52%, P = 0·145). However this latter relationship did not reach statistical significance. CCR5 heterozygosity was not associated with the risk of aGvHD. The incidence of aGvHD was similar in patients with different conditioning regimens, stem cell sources, GvHD prophylaxis or various loci HLA mismatches (data not shown). Logistic regression analysis (Table 2) considering recipient age, donor-recipient gender, 10/10 HLA match and the recipient and donor CCR5 polymorphism confirmed the role of recipient CCR5Δ32 homozygosity as an independent risk factor for grade III–IV aGvHD [odds ratio (OR) = 8·621, P = 0·002]. The results of this analysis indicated that donor-recipient histocompatibility may play a protective role (OR = 0·412, P = 0·012). Moreover, this analysis implicated that a homozygous CCR5Δ32 donor increased the recipient risk of grade IV aGvHD (OR = 10·580, P = 0·005). Our previous work suggested the role of CCR5 gene expression and polymorphism as potential prognostic markers of the outcome of allogeneic HSCT. We showed that lower CCR5 expression (number of CCR5-specific mRNA copies) and the presence of CCR5Δ32 [associated with the loss of the functional CCR5 receptor in subjects homozygous for this mutation and decreased expression in heterozygous individuals (Venkatesan et al, 2002)] is associated with the lower risk of Epstein-Barr Virus reactivation in patients after allogeneic HSCT (Bogunia-Kubik et al, 2007). The presence of CCR5Δ32 has been also found to be inversely correlated with risk of aGvHD in a mixed cohort of sibling and alternative (unrelated and haploidentical) transplant recipients (109 sibling, 25 haploidentical and 65 unrelated; Bogunia-Kubik et al, 2006). Analysis performed in the group of recipients and donors of unrelated transplants (n = 65) showed that the CCR5Δ32 homozygosity is very rare and was present only in one patient and one unrelated donor. In both cases, the patient with the CCR5Δ32 homozygous genotype and recipient transplanted from the CCR5Δ32 homozygous donor developed grade II aGvHD after transplantation. These observations concur with our present results, however a low frequency of the CCR5Δ32 homozygosity constitutes an important limitation of our analyses. Therefore further studies are needed in diverse ethnic populations with larger samples to prove general relevance of this polymorphism with the clinical outcome of HSCT. In contrast, Ma et al (2010), who also investigated the effect of the CCR5 deletion on the outcome of unrelated HSCT, did not observe significant correlations with GvHD incidence. However McDermott et al (2010) suggested the association between the presence of the H1/H1 CCR5 genotype (that includes homozygous wild type genotype lacking the 32 bp deletion) and better disease-free and overall survival, which is in agreement with the present results. There are some conflicting results in murine models regarding CCR5 expression and risk of GvHD. Murai et al (2003) reported that disrupting the CCR5 gene interrupts recruitment of T cells into Peyer's patches in the gut and, as a result, prevents aGvHD in non-irradiated haploidentical recipients. However, transferring CCR5-deficient (CCR5−/−) T cells to lethally irradiated mice had been associated with earlier onset and a worsening of GvHD (Wysocki et al, 2004) and the overall effect of absent CCR5 expression on donor cells resulted in greater GvHD and donor T-cell expansion in a murine transplant model with intensive conditioning (Welniak et al, 2004). It was documented that CD4+ CD25+ T regulatory cells, whose presence is associated with a lower risk of aGvHD development, need CCR5 to function (Wysocki et al, 2005). These data concur with our observations from the present study. Interestingly, recently Reshef et al (2012) tested the in vitro effect of the CCR5 antagonist maraviroc on lymphocyte function and chemotaxis and further studied its effect in vivo on the incidence and severity of aGvHD in patients undergoing reduced intensity conditioned HSCT that combined maraviroc with standard GvHD prophylaxis. They observed that maraviroc inhibited CCR5 internalization and lymphocyte chemotaxis in vitro without impairing T-cell function of normal donor cells. Of note, the CCR5Δ32 cells were unresponsive to CCL5- and CCL3-induced chemotaxis and were unaffected by maraviroc. Moreover, the addition of maraviroc to standard GvHD prophylaxis resulted in a low incidence of GvHD in high-risk patients with haematological malignancies after allogeneic HSCT. The present results and those of Reshef et al (2012) strongly suggest that typing for CCR5 alleles and knowledge of patient and donor CCR5Δ32 genotype may aid in the further development of new GvHD therapeutic strategies involving the use of maraviroc as an effective drug for homozygous CCR5Δ32-negative transplants to prevent aGvHD. Authors would like to thank Małgorzata Polak from the Lower Silesian Centre for Cellular Transplantation (Wroclaw), Dr Barbara Wysoczańska from the Institute of Immunology and Experimental Therapy (Wroclaw) and Dr Kazimierz Hałaburda, Dr Agnieszka Tomaszewska and Agnieszka Witkowska from the Institute of Haematology and Transfusion Medicine (Warsaw) for their assistance. This study was supported by a grant from The National Centre for Research and Development: N R13 0082 06. KBK performed the CCR5 genotyping, analysed the data and wrote the manuscript; SM, SW did statistical analyses; MP, AG, JN, KK, BK, RMW (Diagnosticians, Laboratory staff) – performed HLA typing, prepared and provided DNA samples for the CCR5 polymorphism study; SKK, MM, MDM, AK, MS, KS, DD, JL, MMD, WWJ, AH, AK, JK, KD, KW, JG (Physicians) analysed, collected and provided clinical data for the patients and donors; AL coordinated the clinical and laboratory work, designed the study, supervised statistical studies and critically revised the manuscript. All the authors significantly contributed to this work and agreed to its submission to the British Journal of Haematology. The authors declare no conflict of interest.