Premium
Impact of genomic risk factors on survival after haematopoietic stem cell transplantation for patients with acute leukaemia
Author(s) -
Pearce K. F.,
Balavarca Y.,
Norden J.,
Jackson G.,
Holler E.,
Dressel R.,
Greinix H.,
Toubert A.,
Gluckman E.,
Hromadnikova I.,
Sedlacek P.,
Wolff D.,
Holtick U.,
Bickeböller H.,
Dickinson A. M.
Publication year - 2016
Publication title -
international journal of immunogenetics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.41
H-Index - 47
eISSN - 1744-313X
pISSN - 1744-3121
DOI - 10.1111/iji.12295
Subject(s) - medicine , oncology , framingham risk score , transplantation , proportional hazards model , hematopoietic stem cell transplantation , allele , disease , biology , gene , genetics
Summary The EBMT risk score is an established tool successfully used in the prognosis of survival post‐ HSCT and is applicable for a range of haematological disorders. One of its main advantages is that score generation involves summation of clinical parameters that are available pretransplant. However, the EBMT risk score is recognized as not being optimal. Previous analyses, involving patients with various diagnoses, have shown that non‐ HLA gene polymorphisms influence outcome after allogeneic HSCT . This study is novel as it focuses only on patients having acute leukaemia ( N = 458) and attempts to demonstrate how non‐ HLA gene polymorphisms can be added to the EBMT risk score in a Cox regression model to improve prognostic ability for overall survival. The results of the study found that three genetic factors improved EBMT risk score. The presence of MAL (rs8177374) allele T in the patient, absence of glucocorticoid receptor haplotype (consisting of rs6198, rs33389 and rs33388) ACT in the patient and absence of heat‐shock protein 70‐hom (+2437) (rs2227956) allele C in the patient were associated with decreased survival time. When compared to the EBMT risk score, the scores combining EBMT risk score with the genetic factors had an improved correlation with clinical outcome and better separation of risk groups. A bootstrapping technique, involving repeated testing of a model using multiple validation sets, also revealed that the newly proposed model had improved predictive value when compared to the EBMT risk score alone. Results support the view that non‐ HLA polymorphisms could be useful for pretransplant clinical assessment and provide evidence that polymorphisms in the recipient genotype may influence incoming donor cells, suppressing the initiation of the graft versus leukaemia effect and reducing survival.