Clonal Hematopoiesis Almost “Inevitable” in Older Adults

Using whole-genome sequencing (WGS) data, Florian Zink, from deCODE genetics/Amgen Inc. in Reykjavik, Iceland, and co-authors found that clonal hematopoiesis (CH) is highly prevalent in older adults, “trending toward inevitability,” according to their report published in Blood. Although certain somatic mutations were strongly associated with CH, in most cases, no candidate driver (CD) mutations were present.

“Clonal hematopoiesis is much more prevalent than previously reported and appears to be a fate that awaits anybody who becomes old enough,” Kári Stefánsson, MD, the study’s senior author, told ASH Clinical News. “On one hand, it may be a normal part of aging. On the other hand, we show that it carries clearly increased age-adjusted risks of all-cause mortality and hematologic malignancy.”

The researchers analyzed WGS data from 11,262 Icelanders who participated in various disease projects at deCODE genetics (a biopharmaceutical research firm). Patients were excluded from analyses if they had a diagnosis of hematologic malignancy before or within 6 months of blood sampling. By analyzing the mutational spectrum of hematopoietic stem cells and their clonal descendants, the researchers were able to diagnose CH, “irrespective of whether the predominant clone carries a hematologic malignancy–associated driver mutation,” Dr. Stefánsson noted.

More than 3.3 million singleton single nucleotide polymorphisms (SNPs) were identified, 146,389 of which were classified as mosaic somatic mutations. The mean variant allele fraction of the mutations was 0.17 (range = 0.11-0.20).

Younger participants (<35 years) had a median of three mosaic somatic mutations, and the number of mutations climbed rapidly with age. After applying a cutoff of >20 mosaic somatic mutations to classify WGS “outliers” (who were assumed to have CH), the researchers found that 1,403 participants had CH, for a prevalence of 12.5 percent across all age groups. The frequency of WGS outliers increased from 0.5 percent in those <35 years of age to >50 percent in those >85 years of age.

To determine the proportion of patients with CH who carried detectable mosaic somatic mutations, the authors analyzed 18 CD genes known to be mutated in myeloid neoplasia. They observed 286 CD mutations in 16 of the genes in 246 of the participants. Most of those participants (n=196) had CD mutations in either DNMT3A (n=93), TET2 (n=76), ASXL1 (n=25), or PPM1D (n=18) genes.

The probability of detection of these mutations was “strongly dependent” on age, as those patients were on average 18 years older than those who did not have a CD mutation (p<0.001). However, only 12.6 percent (n=177/1,403) of patients had mutations in the 18 known CD genes.

“Targeting known pre-leukemic driver mutations may address only a fraction of the at-risk individuals,” the authors wrote, noting that a deeper understanding of the nature and risk factors of CH is necessary.

Of the 246 patients with CD mutations, 177 (72%) were also WGS outliers and that proportion increased significantly with age (p<0.001), the authors wrote.

WGS outliers had significantly higher rates of all-cause mortality (hazard ratio [HR] = 1.18; p=2.7×10-4), and participants with CD mutations were also at increased risk, regardless of WGS outlier status (HR=1.30; p=0.0042). To place these results in context, the authors determined that the effect of CH on mortality rate is similar to the effect of having ever smoked (HR=1.19; p<0.001).

CH, per outlier status, was also associated with a “substantially increased risk” of developing a hematologic malignancy 6 months after sampling (HR=2.43; p=9.0×10-5), and the risk increased with the number of mosaic somatic mutations, to an HR of 42.2 (p=1.3×10-9) in those with >250 mutations.

Compared with a control cohort of 1,482 phenotypes from the deCODE database, the authors found significant associations between WGS outlier status and smoking, treatment for addiction, psychiatric disease, smoking-related diseases, and chronic pulmonary disease (p<0.001 for all).

Shorter telomeres and loss of the Y chromosome also showed significant associations with CH (p = 1.0×10-3 and p=5.02×10-110, respectively), suggesting that “telomerase activity might have a role in CH,” the researchers wrote. On multivariate regression analysis, both telomere length (p=0.008) and rs34002450 genotype (p=0.003) were significantly independent predictors of WGS outlier status.

“The ability to identify CH cases presents opportunities for monitoring and intervention. … Clearly a deeper understanding of the nature and associated risks of CH would be valuable,” the authors concluded.

“Because so many of our cases have no detectable driver mutation, we speculated whether clonal hematopoiesis might in some cases arise spontaneously through neutral drift in the hematopoietic stem cell population,” Dr. Stefánsson said. “We used modelling to show that this is plausible, particularly if the active stem cell population is small or has poor fitness.”

The study findings are potentially limited because the detection methods used to identify CD mutations may not have had high enough sensitivity levels. In addition, the WGS outlier method the researchers used was “likely to under-report CH” in younger individuals because they had not accumulated enough mosaic somatic mutations to qualify as outliers.

Drs. Zink, Stacey, and Norddahl are employees of deCODE Genetics, a subsidiary of Amgen.


Zink F, Stacey SN, Norddahl GL, et al. Clonal hematopoiesis, with and without candidate driver mutations, is common in the elderly. Blood. 2017 May 8. [Epub ahead of print]