Abstract
Allogeneic haematopoietic stem cell transplantation (HSCT) using volunteer unrelated donors (UD) is a life-saving intervention for patients with haematological malignancies. It is recognised that numerous genetic factors in both patient and donor play a role in outcome. TGF beta 1 is a member of a highly pleiotrophic family of growth factors involved in the regulation of numerous immunomodulatory processes and may play a role in carcinogenesis. Several functional polymorphisms have been identified, such as a single nucleotide polymorphism (SNP) at codon 10 (c.29T>C, p.L10P) of exon 1. Conflicting data has been published regarding the impact of the SNP on plasma levels and its role in sibling HSCT. To date there are no published data in UD HSCT. We hypothesised that this polymorphism may influence the outcome of UD HSCT by modulating the immune response. We genotyped, for the presence of a SNP at codon 10, a large group of patient/donor pairs (314) who underwent an UD HSCT using a donor provided by the Anthony Nolan Trust, in a UK transplant centre. The transplant took place between 1997 and 2006 and the median follow up time was 5.8 years (0.8–10.18 years). The diagnoses were chronic myeloid leukaemia 69 (21%), acute lymphoid leukaemia 76 (24%), acute myeloid leukaemia 76 (24%), myelodysplasic syndrome 25 (8%), lymphoproliferatives disorders 37 (11.8%) and others 31 (9.9%). Myeloablative conditioning regimens were used in 69.9% of transplants; T-cell depletion was included in 86.9% of conditioning protocols. Sixty eight percent of the transplants were full matched (10/10), and 12.7% were 12/12. The patients’ observed SNP frequencies were TC 55.7%, TT 32% and CC 12.1%, and for the donor were 51%, 34.7% and 14.3% respectively. There were no significant effects of the presence of a SNP in either the patient or donor groups alone. However, when we analysed the impact of the total number of SNPs present in the pair, we found that multiple SNPs (3–4 SNPs vs 2 or less) were associated with a significantly decreased overall survival (OS) (5 years: 30% vs 42%, log-rank p=0.04), disease free survival (DFS) (5 years: 17% vs 25%, log-rank p=0.02) and a higher treatment related mortality (1 and 3 years: 42% and 45% vs 25% and 30%, respectively, log-rank p=0.03). In multivariate analysis there was a trend to improved OS in the pairs with fewer SNPs (HR: 0.7; 95% CI 0.4,1.0; p=0.07). We speculated that the impact of the donor genotype might differ depending on the patient genotype. In patients with a wild type genotype, the donor genotype did not impact significantly on outcome. Conversely, the patients with a SNP at codon 10 (TC or CC), had significantly better DFS when using a donor with a wild type genotype compared to those with a SNP present (5 years: 34% vs 21%; log-rank p= 0.04). In conclusion, we have shown for the first time in a large number of UD HSCT pairs that increased numbers of SNPs in the TGFB1 gene at codon 10 in patients and donors are associated with a worse outcome following UD HSCT. While an exact functional mechanism remains unclear, these data emphasise the importance of pursuing functional analyses of TGF beta in this setting. In addition, identification of these SNPs pre-transplant will allow for transplant conditioning and immunosuppression regimens to be tailored to the individual patient, as well as assisting in the most appropriate choice of donor.
The impact of HLA genotyping on survival following unrelated donor haematopoietic stem cell transplantation
