Investigators Intercept De Novo AML 'Playbook'

Neil Osterweil

May 01, 2013

For such a complex and often treatment-refractory cancer, acute myeloid leukemia (AML) has a comparatively modest genetic makeup, according to investigators in the Cancer Genome Atlas Research Network.

Foraging into the genomes of both normal and cancerous cells in adults with AML, they have revealed what lead investigator Timothy J. Ley, MD, calls the cancer's "genetic playbook," a finding that could help in the development of better risk models and, ideally, therapies better targeted to each patient's disease subtype.

The report is published online May 1 in the New England Journal of Medicine.

By comparing genomes from normal skin with genomes from the malignant cells of 200 adults with de novo AML, Dr. Ley, from the Genome Institute at Washington University in St. Louis, Missouri, and colleagues found that the AML genomes have, on average, only 13 mutations. Other types of adult cancers that have been sequenced have hundreds of mutations.

Nearly all AML samples tested had at least 1 nonsynonymous mutation — that is, a mutation that is likely to be significant in the development of de novo AML.

"This dataset will provide a framework for future studies that pertain to the molecular classification of patients with AML. The identification of many potentially important relationships among recurrently mutated AML genes and pathways provides a comprehensive foundation for an understanding of the genetic rules of pathogenesis," they write.

"Giving Great Insights"

The investigators found nonsynonymous mutations in each of 9 different gene categories, including mutations in signaling genes (found in 59% of all samples), genes related to DNA methylation (44%), chromatin-modifying genes (30%), and NPM1, the gene encoding nucleophosmin, which is overexpressed, underexpressed, rearranged, or deleted in several types of cancer (27%).

Although chromosomal deletions, segment repeats, and other abnormalities found in some patients with AML are important for diagnosis and risk stratification, about half of all patients with AML have normal karyotypes. Some patients in this intermediate-risk cytogenetic category respond well to consolidation chemotherapy, whereas others do poorly. Identifying which patients might benefit from which therapy has been challenging, so deeper analysis of the genetic and epigenetic basis of AML is needed, the investigators write.

"This is the largest genome-wide analysis of AML ever performed, and it gives us great insight into not only what's going on in this intermediate-risk group, but also in the more complex karyotypes," John F. DiPersio, MD, PhD, who is one of the study investigators, told Medscape Medical News.

Uncomplicated, Yet Intricate

The findings suggest that AML is a "curiously uncomplicated malignancy.... It's associated with a very limited number of mutations, compared with things like lung cancer or melanoma, so there's a possibility that we may figure out what the drivers are, how this disease functions," said Dr. DiPersio, who is chief of the division of oncology at the Washington University School of Medicine.

"This lack of complexity is relative, however," writes David P. Steensma, MD, from the division of hematologic malignancies at the Dana-Farber Cancer Institute and Harvard Medical School in Boston, Massachusetts, in an accompanying editorial.

"The clonal architecture of AML is dazzlingly intricate, especially in cases arising from the myelodysplastic syndromes, with some subclones becoming extinct over time, and others achieving dominance, unpredictably," he writes.

The investigators performed whole-genome sequencing of primary tumor and matched skin samples from 50 patients with de novo AML, and exome capture and sequencing of paired samples from another 150 patients. The study involved 108 men and 92 women; mean age at study entry was 55 years.

The most frequently occurring mutations were in NPM1 (in 27% of samples), FLT3 (28%), DNMT3A (26%), IDH1/IDH2 (20%), NRAS or KRAS (12%), and RUNX1 (10%).

Mutations in FLT3, DNMT3A, and NPM1 were more frequently grouped together than would be predicted statistically, and these genes were found less frequently in association with transcription-factor fusions than would otherwise be expected, the investigators report.

"This observation, combined with the strong association between samples having concurrent mutations in NPM1, FLT3, and DNMT3A and distinct clusters in messenger RNA, microRNA, and DNA methylation, suggests that samples with mutations in all 3 genes represent a novel subtype of AML," they write.

Secondary AML a Different Disease

Dr. DiPersio cautioned that the study focused specifically on de novo AML, and should not be used to draw conclusions about treatment-related, or secondary, AML.

"Treatment-related AML is probably a different disease, at least in part; it may be substantially different. The lesson we learned is don't assume that treatment-related AML is exactly like de novo AML," he explained.

Genomic studies of treatment-related AML are in progress, and are likely to reveal "substantial and striking" differences between disease types," he said.

In his editorial, Dr. Steensma contends that most AML secrets have been laid bare.

"It is likely that all the common, recurrent genetic lesions in AML — the molecular equivalent of major causes of death, such as stroke and heart disease — are now described. In individual cases, rare genetic events may occur, akin to uncommon causes of death, such as falling down a well or being struck by space debris," he writes.

The Cancer Genome Atlas Research Network is funded by the National Institutes of Health.

N Engl J Med. Published online May 1, 2013. Abstract, Editorial


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.