Genomics research peering into the pathogenesis of adult de novo acute myeloid leukemia (AML) reveals many important relationships in both genomic and exomic mutations, according to a review published May 1 in the New England Journal of Medicine.
Timothy J. Ley, MD, of the Genome Institute at Washington University and Siteman Cancer Center in St. Louis, Mo., and colleagues from the Cancer Atlas Research Network, outlined genetic factors thought to be involved in the initial development of AML.
It is well accepted that acquired genetic abnormalities are at the center of AML pathogenesis, but researchers generally find that even after single-nucleotide polymorphism (SNP) arrays and high-density comparative genomic hybridization, almost half of their samples have a normal karyotype with few structural abnormalities. What’s more, targeting sequencing identifying recurrent mutations have been conducted with promising results and mutations found in FLT3, NPM1, KIT, CEBPA, TET2, DNMT3A and IDH1, many of which have been found to be involved in pathogenesis in animal models, but AML patients tested often elude these mutations. Still, discovered AML mutations account for the smallest percentage of mutations found in all types of cancer in adults.
Chemotherapy has been helpful in the management of patients with cytogenetic profiles on the low end of AML risk stratification. However, complex alterations and monosomy karyotype contribute to high-risk profiles and prompt physicians to conduct allogeneic transplantation during first remission. Intermediate risk classification has required special handling and the development of advanced biomarkers. People with this profile generally have a normal karyotype and have mixed response to chemotherapy. Some in this classification have dismal prognoses. The most modern standard of genetic testing for this group incorporates algorithm classification for FLT3, NPMI1, CEBPA and KIT, and additionally, DNMT31, IDH1/2 and TET2.
For this research, a total of 200 clinical cases of AML were analyzed using one of two methods: whole-genome sequencing of the primary tumor and normal skin sample, which accounted for 50 cases, or whole-exome sequencing of paired skin and tumor samples for the remaining 150 cases, along with RNA and microRNA sequencing and DNA methylation anlaysis.
Significant mutation was found in 23 genes and 237 genes were found mutated in two or more samples. Most samples incurred at least one mutation categorized nonsynonymous for nine gene classifications identified as possible culprits in AML pathogenesis. Patterns of cooperation and mutual exclusivity suggested strong biologic relationships among several of the genes and categories. Among these are signaling genes accounting for 59 percent of cases, DNA-methylations-related genes for 44 percent, chromatin-modifying genes for 30 percent, the gene encoding nucleophosmin (NPM1) for 27 percent, myeloid transcription-factor genes for 22 percent, transcription-factor fusions accounting for 18 percent of cases, tumor suppressor genes for 16 percent, spliceosome-complex genes in 14 percent of cases and cohesin-complex genes for 13 percent.
The investigation led to the revelation of 2,315 somatic single-nucleotide variants and 270 deletions and small insertions in tier 1 genomic coding regions. There were 13 tier 1 mutations per sample on average. Sample stratification was set in 10 categories based on cytogenetic risk profile, fusion events and high-risk mutated TP53. Among these, MLL fusions were discovered in 11 samples that also showed the fewest (2.09) recurrent tier 1 mutations compared to the overall mean of 5.24 for a p value of 0.002 following multiple comparison correction. Another 20 samples of PML-RARA fusions also had diminished numbers of tier 1 mutations revealing a mean of 3.25 (p=0.001). Seven samples with RUNX1-RUNC1T1 fusions had a mean of 7.85 tier 1 mutations for a p value of 0.04.
“This finding suggests that MLL fusions require fewer cooperating mutations than other AML-initiating events,” wrote the authors.
Genetic mutations within biological classes such as cohesin-encoding mutations, signaling proteins and those of the spliceosome and histone modifiers demonstrated a pattern of mutual exclusivity indicating that any of these could be involved in AML pathogenesis.
Analysis of methylation patterns support previous studies of CpG islands for transcription-factor fusions as well as mutations in IDH1/2, but researchers were surprised to find CpG sparse areas of the genomes had the greatest methylation signatures. Expression signatures of mRNA and miRNA were connected to mutations of FLT3, NPM1 and DNMT3A, indicating that this model may shape a new category of AML with its own epigenetic features.
Researchers concluded that at least one potential driver mutation was present in almost every AML sample. A myriad of genetic events comprise an intricate system with many moving parts that could have an impact on the development of AML.
David P. Steensma, M.D., from the division of hemotologic malignancies at Dana-Farber Cancer Institute and Harvard Medical School in Boston, provided an editorial published in conjunction with this review and focused on the current age of broad genetic surveys and particularly those for cancer. Detractors of these large arcs of study have said that it isn’t hypothesis driven and is too expensive.
“But as discoveries from these projects stack up, and as terabytes of observational data yield new insights into disease biology and prompt the development of pathway-driven target therapies, the usefulness of such approaches is becoming undeniable,” wrote Steensma.
The author mentioned the conclusion of the 8-year Cancer Genome Atlas, by which the above review was made possible, estimating that information regarding 10,000 cancer genomes has been unpacked during this project. From the AML research, Steensma counts new knowledge regarding isocitrate dehydrogenase and neomorphic enxyme activity to be especially significant, as well as glioma research and DNMT3A mutations often seen in AML, which could help improve patient prognosis.
“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,” Steensma wrote. “In individual cases, rare genetic events may occur, akin to uncommon causes of heath, such as falling down a well or being struck by space debris. Within two years, the door to major new genetic findings will also close for most other common neoplasms and even for some rarer tumors.”
Steensma projected that investigators just entering the field will still have job security in functional studies, including those for mutant proteins like cohesions and spliceosome components involved in cancer. A major challenge cited for translation into clinical practice was the logistical incongruence between how long genetic testing may take for AML patients and how quickly initiation of therapy is required. Many patients need immediate treatment, which could conclude by the time results of genetic testing are made available.