Abstract
Typically MDS affects older adults, with a median age of diagnosis at 71 years. The incidence increases from 2.5 /105 in the 6th decade of life to 30/105 in the 8th to as high as 60/105 in patients >80 years of age. Younger patients are sporadically affected; reflecting either an extreme pole of age distribution or perhaps constituting a distinct disease entity or sub-entity. We hypothesized that younger patients have discrete molecular and clinical features and identification of the corresponding differences between older and younger patients will contribute to the understanding of the pathogenesis of this disease.
To investigate the differences between older and younger patients with MDS we assembled a cohort of patients with MDS, MDS/MPN and sAML, stratifying the cohort based on age of presentation and various clinical parameters including outcome. We then characterized molecular defects though both SNP karyotyping and next generation deep sequencing of a subset of the top 60 most commonly mutated genes in myeloid disorders.
In total, 618 patients with primary MDS (treatment related MDS excluded) were analyzed with the median age of presentation of 68 ([12-90]) years, with the 5th percentile at 46 years. To assure the best statistical power, the cohort, ranked according to age increments of 5 years, was split into two groups, those 50 years of age and under (n=68 [12-50], mean=41), and those over 50 (n=550, [51-90], mean=70). Women were over-represented among younger patients (60% vs. 33%, p<.0001). High risk MDS (35%) and sAML (19%) were more common in patients 50 and under. When the presence of family history of cancer or leukemia was investigated, 62% of younger patients were positive for a family history, compared to 72% in older patients (p=.108).
We then proceeded to analyze molecular and genetic defects. Cytogenetic findings showed no significant differences between younger and older patients, with normal cytogenetics frequently found in both groups (37% vs. 47%, respectively). Although the percentage of older patients with a complex karyotype was higher (4% vs. 13%), it was not found to be statistically significant. ASXL1 (12%), RUNX1 (10%), TET2 (10%), and TP53 (9%) mutations were the most prevalent mutations in younger patients. SRSF2 was significantly mutated in patients over 50 (p=.014). When grouped based on functional properties, genes of the spliceosomal complex were significantly more mutated in patients over 50 (p=.021), while those of the PRC2 (p=.084) and Ras family (p=.084) tended to be found in a higher proportion of older patients. No young patients had NF1 (p=.096) mutations. In an attempt to explain the early occurrence of MDS, we also analyzed the cohort for the presence of congenital mutations known to be associated with the risk of leukemia, but an increase incidence of germ line mutations of TP53, RUNX1, the telomerase complex, heterozygous Fanconi Anemia genes (e.g. FANCA, FANCB, FANCC, FANCD2), and other DNA repair genes (e.g. MSH2, MSH6, and PMS2) was not detected.
While there are several marked differences in the mutational spectrum between younger and older MDS patients, overall our results suggest that younger patients with MDS constitute an extreme pole of a demographic continuum rather than a distinct group associated with a distinct signature of molecular pathogenesis.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.