Precision Medicine at Work: Genetic Profiling to Guide Stem-Cell Transplantation in Myelodysplastic Syndromes

The myelodysplastic syndromes (MDS) comprise a clinically and pathologically heterogeneous group of hematologic malignancies that are collectively characterized by clonal hematopoiesis, abnormal myeloid differentiation, ineffective hematopoiesis, and risk of progression to acute myeloid leukemia. This phenotypic heterogeneity reflects a variety of genetic lesions that contribute to disease. Somatic mutations in numerous genes are known predictors of both clinical phenotype and overall survival.^1,2  Allogeneic transplantation remains the only curative therapy for MDS, and mortality after transplantation remains high owing to relapsed disease and transplant-related complications. While the impact of somatic mutations in MDS on post-transplant outcomes has emerged,^3,4  how to integrate these genetic details into patient management to identify those patients for whom a transplant is beneficial and tailor conditioning regimen intensity to a given person and their disease remains unknown. Recent work by Dr. R. Coleman Lindsley and colleagues begins to address these clinical challenges.

Dr. Lindsley and colleagues sequenced a set of 129 genes with suspected roles in MDS or inherited or acquired bone marrow failure syndromes in archived pre-transplant blood samples from 1,514 individuals who underwent hematopoietic stem cell transplantation (HSCT) for MDS. They identified at least one mutation in 79 percent of patients, with a median of two driver mutations per patient (range, 0-15 mutations). The genes most commonly mutated were ASXL1 (19.6%), TP53 (19.1%), DNMT3A (15.0%), TET2 (12.2%), and RUNX1 (11.5%). Associations with patient outcomes (e.g., overall survival, relapse, and death without relapse) were analyzed among the 32 most commonly mutated genes in more than 20 patients. In this analysis, no mutations were associated with a prolonged survival. Mutations in genes significantly associated with poorer overall survival compared with the absence of the mutation included TP53 (hazard ratio for death [HR], 1.96; adjusted p<0.001); the TP53 regulator, PPM1D (5.8% of patients; HR, 1.64; adjusted p=0.002); and JAK2 (2.4% of patients; HR, 1.77; adjusted p=0.03). Nongenetic factors associated with shorter survival included those factors incorporated into standard prognostic risk models (such as age, performance-status, degree of HLA matching, elevated blast percentage, and others). TP53 mutations were the most powerful predictor of survival after transplantation, independent of clinical factors. In multivariate analyses, only mutations in TP53 were independently associated with shorter survival. Of note, the negative impact on overall survival of TP53 mutations was independent of the size of the mutant clone (defined by a variant allele fraction of >10% or <10%) or the presence of more than one TP53 mutation. Mutations associated with an increased risk of relapse included TP53 (HR, 2.01; p<0.001) and RAS pathway mutations (NRAS, KRAS, PTPN11, CBL, NF1, RIT1, FLT3, and KIT; HR vs. no RAS pathway mutation, 1.56, p=0.002). JAK2 V617F mutations were associated with a greater risk of nonrelapse mortality (HR, 2.10; p<0.001). The authors then generated a hierarchical prognostic model based on recursive partitioning analysis for overall survival, including clinical and genetic variables and effect of conditioning intensity.

Patients with TP53 mutations were unified by poor survival and a high risk of relapse. Notably, the median survival was similar whether such patients received a myeloablative or reduced-intensity conditioning regimen (7.5 and 9.2 months, respectively; p=0.19), and the incidence of relapse was also similar despite the intensity of conditioning. Among patients 40 years or older who lacked a TP53 mutation, the presence of RAS pathway mutations was associated with shorter survival than the absence of RAS pathway mutations (median, 0.9 vs. 2.2 years; p=0.004), owing to a higher risk of relapse. This higher risk of relapse was restricted to those who received a reduced-intensity conditioning regimen. Patients older than 40 years with JAK2 mutations had a shorter median survival than those without JAK2 mutations (0.5 vs. 2.3 years; p=0.001), associated with a higher non-relapse mortality, regardless of the intensity of the conditioning regimen. These data suggest that among older MDS patients undergoing HSCT, high-intensity conditioning regimens may benefit patients with RAS pathway mutations but not those with TP53 or JAK2 mutations.

The study also yielded important insights into the biology of MDS in specific subsets of patients. When comparing younger (<40 years) and older patients with MDS, mutations in TET2, DNMT3A, SRSF2, SF3B1, and PPM1D were more common in older patients, and mutations in GATA2, PIGA, and compound heterozygous mutations in the Shwachman-Diamond syndrome–associated SBDS gene were more common in younger patients, suggesting a difference in disease pathophysiology. Indeed, in young adults, 4 percent of the patients had compound heterozygous mutations in SBDS that were purported to be germline, indicating an underlying inherited predisposition to MDS. Of note, all seven of the patients with biallelic SBDS mutations had somatic TP53 mutations, which may provide insight into mechanisms of clonal evolution in Shwachman-Diamond syndrome. Mutations in the TP53 regulator PPM1D were more common among patients with therapy-related MDS than those with primary MDS (15% vs. 3%; p<0.001).