Genomic prediction of relapse in recipients of allogeneic haematopoietic stem cell transplantation

Abstract

Allogeneic haematopoietic stem cell transplantation currently represents the primary potentially curative treatment for cancers of the blood and bone marrow. While relapse occurs in approximately 30% of patients, few risk-modifying genetic variants have been identified. The present study evaluates the predictive potential of patient genetics on relapse risk in a genome-wide manner. We studied 151 graft recipients with HLA-matched sibling donors by sequencing the whole-exome, active immunoregulatory regions, and the full MHC region. To assess the predictive capability and contributions of SNPs and INDELs, we employed machine learning and a feature selection approach in a cross-validation framework to discover the most informative variants while controlling against overfitting. Our results show that germline genetic polymorphisms in patients entail a significant contribution to relapse risk, as judged by the predictive performance of the model (AUC = 0.72 [95% CI: 0.63-0.81]). Furthermore, the top contributing variants were predictive in two independent replication cohorts (n = 258 and n = 125) from the same population. The results can help elucidate relapse mechanisms and suggest novel therapeutic targets. A computational genomic model could provide a step toward individualized prognostic risk assessment, particularly when accompanied by other data modalities.

Fig. 1

Schematic representation of the study setup. a Leave-one-out cross-validation (LOOCV) for feature selection and classification model fitting. Each sample is systematically left out in each fold. Prediction error estimates are based on left out samples (blue). b The analysis procedure within each LOOCV fold includes a first round of feature selection with a logistic regression association test followed by fitting a Random forest classification model on variants below an initial association p-value threshold

Fig. 2

Estimated predictive performance of the model. The results from a the discovery dataset, and b–c the replication datasets. The left-hand side panels show the prediction value distributions over the LOOCV folds for the actual relapsed and non-relapsed groups by the Random forest classification model. The middle panels show the prediction ROC curves and AUC values. In a, the solid black ROC curve indicates the genetic model, the dashed gray curve indicates the model with principal components, and clinical and genetic variables, and the dotted purple curve shows the result using principal components and clinical data only. In b, the dashed green curve and the dotted blue curve show the results for allowing variants with <11 and <81 missing values, respectively. In c, the black curve and the dotted green curve show the results for higher (<0.3) and lower (<0.2) imputed genotype quality filtering stringencies, respectively. The right-hand side panels in a–c show the odds ratio for the correct prediction (y-axis) along the prediction model output values (x-axis). The p-values are calculated with one-sided Mann–Whitney test. The statistical power of the AUC is calculated at alpha level 0.01