A whole-genome sequencing study from St. Jude Children's Research Hospital and Washington University has identified patterns of mutations differentiating the two clinical subtypes of rhabdomyosarcoma, and uncovered a potential targeted treatment strategy for some of these tumors.
The study, which was performed under the umbrella of the Pediatric Cancer Genome Project and published last month in Cancer Cell, sequenced 16 tumors and matched normal DNA from 13 rhabdomyosarcoma patients using paired-end Illumina sequencing, generating 42x average coverage and 37x average exon coverage.
The group also performed transcriptome sequencing on 15 of the 16 tumors, and validated all the discovered somatic alterations using custom-capture methods as well as sequencing.
The results revealed that the two histological subtypes of rhabdomyosarcoma have significantly distinct genetic features. The embryonal subtype showed many structural and copy number variations, the group reported, including mutations in the RAS/NF1 pathway, which are significantly associated with intermediate- and high-risk embryonal tumors.
In contrast, the second subtype, alveolar tumors, showed fewer genetic lesions overall and no known recurrently mutated cancer consensus genes, supporting the hypothesis that alveolar rhabdomyosarcoma is driven by a single fusion between part of the FOXO1 gene with either the PAX3 or the PAX7 gene.
"The fact that these two subtypes have very different genomic landscapes is important for a few reasons," Michael Dyer, the study's corresponding author, told Clinical Sequencing News this week. "It's been thought for a while that tumors with more chromosomal lesions are more aggressive than those with quieter genomes. But here, we confirmed that the [subtype] with the much quieter genome is much more aggressive."
Also, Dyer said, "we found that the translocation-driven tumors — the alveolar type — really don't have a preponderance, or really any recurrent cancer gene mutations, so it looks like the translocation is driving much of tumorigenesis, which was somewhat surprising."
Overall, the researchers identified and validated 22,120 somatic sequence mutations and 1,275 structural variations across the 16 tumors.
According to Dyer, among the embryonal subtype tumors in the study — which proved to be the more genomically dynamic group — more than half of those considered intermediate or high-risk had mutations in RAS pathway genes such as NRAS, KRAS and HRAS.
This raised hopes that these tumors could be treated with RAS-targeted drugs, a possibility the group then tested by screening RAS pathway inhibitors (as well as a large panel of other potential treatments, including chemotherapeutics and agents that perturb the oxidative stress pathway) in cultures made from samples of human rhabdomyosarcomas grafted into mice.
Unfortunately, the RAS pathway inhibitors showed almost no effect on the xenograft cultures, the authors reported.
"My clinical colleagues and many in the field were eager to immediately start trials, because there is just so little [we can do] and no real leads for this cancer," Dyer said. "But there are 100 different [RAS] drugs we could use … so this motivated us to take these orthotopic xenografts we had made in parallel to the genome project, and to see if we could take those tumors from the mice — which were one or two passages away from the patients — and screen them as primary explants."
He added that his group "picked 100 drugs targeting different components of the RAS pathway and almost none of them did anything."
Luckily, RAS pathway genes were not the only ones the group found mutated in embryonal rhabdomyosarcomas, Dyer said. Analysis of other alterations and some hints from the clinical features of the disease spurred the group to also test drugs that induce or increase oxidative stress, such as histone deacetylase inhibitors, with promising results.
"This group of agents that alter cells' ability to handle oxidative stress has given us some promising leads for translational work, and hopefully eventually for clinical trials," Dyer said.
In addition to looking for potential drug targets, the researchers also used the sequencing study to examine the clonal evolution of rhabdomyosarcoma by comparing tumor samples from two patients with post-treatment recurrences. The analysis suggested that chemotherapy can eliminate the major clone in a tumor, with a minor subclone then seeding recurrent tumors with different genomic profiles.
Dyer said the team is now planning to follow up the study in two directions. On the genomic analysis side, he said the group is planning a more comprehensive epigenetic study of the two rhabdomyosarcoma subtypes.
"We did cursory analysis that showed they are distinct, but now we are going to go back … to look at 11 histone marks and also doing more extensive DNA methylation analysis. The team is also planning proteomic and phosphoproteomic analyses, he added.
In parallel, the researchers also plan to follow up on the oxidative stress drug leads.
The PCGP has previously identified potential treatment targets in medulloblatoma, retinoblastoma and acute lymphoblastic leukemia.
According to the project's website, the effort has now sequenced the complete normal and cancer genomes of 700 pediatric cancer patients since its launch in 2010.