NEW YORK – Extensive sequencing within a precision oncology program at the Hospital for Sick Children (SickKids) in Toronto has unearthed not only potentially targetable somatic mutations in these kids' tumors, but also germline alterations that elucidate the roots of their cancers and the chances of relapse.
Within the ongoing "SickKids Cancer Sequencing," or KiCS, program, only a small proportion of patients were able to get on a treatment informed by their tumor makers, reflecting the need for more drug trials, a common problem in precision oncology. Even so, the integrated genomic analyses performed by SickKids' researchers highlighted additional diagnostic and management insights that may be drawn from precision oncology programs.
The researchers leading KiCS are conducting panel sequencing, whole-genome sequencing, and transcriptome sequencing on samples from children, adolescents, and young adults with rare, relapsed, and otherwise complex cancer types. Using these tools, they are aiming to identify clinically meaningful mutations, copy number changes, fusions, possible expression-based classifiers, and more.
Patients who are referred to the program because they are suspected to be at increased inherited risk for cancer typically receive germline sequencing using a clinically validated custom panel of 886 genes implicated in pediatric malignancies, explained Anita Villani, co-director of the KiCS program and an oncologist at SickKids.
Children with poor cancer prognosis comprise the primary group of participants in the study, however. "These are pediatric and young adult patients who either have a poor prognosis based on their upfront diagnosis, or have a poor prognosis based on having relapsed or refractory, recurrent disease," Villani said. The study also includes patients who have been referred for genomic testing because they have a rare, poorly characterized, or difficult-to-treat tumor type, she noted, as well as cases where conventional genomic profiling methods do not fully address clinicians' questions.
For these patients, the team is using the 886-gene pediatric cancer panel to do deep sequencing on matched tumor and normal samples, along with whole-genome sequencing on the tumor-normal pairs and tumor RNA sequencing.
The multi-pronged sequencing approach "allows us to carry out much more integrative analysis," said Villani. "We use this data to help corroborate our tumor mutational burden calculations, in addition to looking at … tumor mutational signatures."
The same genomic testing platforms are also being used to assess a subset of cancer cases enrolled through a nation-wide prospective pediatric genomics study, called "Pediatric Oncology for Young People" (PROFYLE), though KiCS itself is focused on cases recruited at SickKids.
Patients are referred to the KiCS program by their treating oncologist or geneticist, and these caregivers subsequently take part in weekly molecular tumor board meetings with oncologists, pathologists, geneticists, genetic counselors, bioinformaticians, and other experts to discuss findings that may inform treatment or management strategies for each case.
The program has recruited more than 400 patients since it started five years ago. "It is a referral-based program but, having said that, we've acquired quite a broad range … of patients, ranging from some of the more common pediatric diagnoses to very uncommon pediatric diagnoses," Villani noted.
At the American Society of Human Genetics (ASHG) virtual meeting on Tuesday, Villani presented data from 200 pediatric or adolescent/young adult cases evaluated through KiCS. The genomic testing strategies, she explained, uncovered targetable alterations in more than half of the patients, while other detected somatic and germline features helped flag patients at risk of relapse.
These results are generally in line with results reported by other pediatric sequencing efforts, according to Donald Parsons, deputy director of Baylor College of Medicine's Texas Children's Cancer and Hematology Center, who leads a similar Texas effort known as KidsCanSeq but was not involved in KiCS.
Though he noted that the proportion of pediatric cancer cases with clinically actionable tumor alterations does tend to vary widely depending on the sequencing strategies used and how "actionability" is defined, Parsons also credited the Toronto team for the breadth of testing approaches applied in KiCS.
"My general impression of this study is that they're doing a really rigorous and excellent, integrated analysis of patients' tumor and blood samples," Parsons said.
The mutations found in the study were some of the same genetic changes identified with conventional methods such as cytogenetic testing, Villani reported, though researchers also identified alterations that would not have been expected initially based on patients' tumor type or age at diagnosis.
Roughly one in 10 clinically actionable mutations turned up in the germline, for example, and at least 12 percent of the patients carried a pathogenic or likely pathogenic variant in a cancer predisposition gene. In contrast, outside of the KiCS program, a relatively small number of patients treated at SickKids are found to harbor germline mutations.
Notably, at least some of the pediatric or young adult cases had mutations in BRCA1/2, genes that are well known to increase the risk of breast, ovarian, pancreatic, prostate, and other cancers in adults, but have been more difficult to conclusively link to pediatric cancers.
Some of these patients with germline mutations would not have been predicted to harbor them based on family history alone. However, it remains to be seen how these mutations in BRCA1/2 affect pediatric cancer features and susceptibility, Parsons said. "We're still trying to figure out what it means, but it's definitely something that is potentially clinically important for that family," he said.
Mutations in BRCA1/2 and other genes are known to hobble a DNA repair mechanism, called homologous recombination. Some of the kids and young adults in KiCS had such homologous recombination defects or a mutational signature dubbed a "BRCAness" signature. These features are used to guide PARP inhibitor use in some adult cancer types.
"There seems to be a subgroup of pediatric cancers for which defects in homologous recombination repair seem to be a significant driver of the cancers themselves," Villani said, adding that the data available so far has allowed researchers to explore "whether these germline variants were passengers or whether they, in fact, could be drivers of the pediatric tumors."
The team also saw a rise in tumor mutational burden (TMB), a proposed marker for checkpoint blockade immunotherapy response, in around 9 percent of the KiCS participants profiled. In at least some patients who were tested after treatment or relapse, TMB appeared to increase with time, consistent with a possible role for cancer treatment in new mutation events or secondary, treatment-related tumors.
More research is needed to understand the clinical implications of this rising TMB in kids and young adults, but Villani posited that it may reveal an additional therapeutic opportunity in some cases. "That is certainly something we're actively continuing to work on," she said.
The researchers are also looking into sub-clonal dynamics within the tumors of a subset of patients on whom they have collected samples longitudinally and deeply sequenced using panel testing. Insights from these investigations may inform the sampling strategies used by pediatric precision oncology programs in the future.
"At the level of a precision oncology program," Villani said, "it becomes very relevant to the question of, 'Is re-biopsying at relapse and at subsequent relapses something that should be actively pursued? What is the frequency with which we expect to see the occurrence of new drivers and new targets?'"
With these and other analytical approaches, KiCS researchers are "doing an interesting job of trying to squeeze every bit of information they can out of the data," and go "beyond simply looking at mutations that might guide a drug or germline mutations that might indicate susceptibility," Parsons said.
Pairing targets to treatment
Despite the significant number of clinically actionable alterations detected among KiCS patients presented at ASHG, Villani noted that a relatively small proportion of those individuals ultimately received targeted treatments corresponding to the mutations found. She did not provide the specific number of KiCS participants who received biomarker-guided treatment in her ASHG presentation.
"The goal, of course, is to be able to offer these patients therapy based on the molecular findings, and we do return these results to referring clinicians and do the best we can to support them in identifying mechanisms through which they might obtain drugs," she explained, noting that the KiCS program is not currently paired with specific clinical trials of targeted agents, though such strategies are being pursued through the national PROFYLE program.
In some cases, data generated through KiCS has helped clinicians enroll patients on clinical trials being done at the institutional level, while other investigators have gotten creative in securing access to targeted treatments through single patient protocols or compassionate access routes with drug companies.
But access to investigational drugs is not the only thing limiting the ability of a patient to be treated according to an actionable mutation, Villani explained. In some cases, patients may not have exhausted standard-of-care options and are eligible for another effective, approved treatments, or their cancer may have progressed to a point that necessitates palliative care.
Even so, "Our goal, particularly within the Canadian context, is to increase our capacity to have clinical trials available for pediatric and young adult patients, such that they may leverage the opportunities that might be suggested from their molecular data," she said, adding investigators will continue homing in on the alterations that are most meaningful for patients, including those tied to treatment response, as more data comes out of KiCS and other pediatric sequencing studies.
The team is continuing to sequence new cases, while updating the analytical tools used to interrogate and interpret the resulting targeted sequence, DNA sequence, and transcriptome sequence data. In addition to generating data that is expected to demonstrate the utility of genomic profiling in pediatric cancer, the study may also help to untangle the suite of genomic tests that are most informative for pediatric cancer patients in the clinic.
"The question is still, for clinical purposes … what are the tests that are most useful and necessary?" Parsons said.