NEW YORK – It has become apparent that combining DNA and RNA sequencing is beneficial for researchers seeking to understand more about the biology of pediatric cancers and for clinicians looking to pair patients with targeted therapies. Now, a case study involving a child with juvenile myelomonocytic leukemia (JMML) with an exceedingly rare FLT3 fusion may help clinicians to decide on a set of best practices that could lead to more precise treatment of pediatric cancer patients.
In a paper published recently in the Nature journal Leukemia, University of California, San Francisco assistant professor and pediatric oncologist Elliot Stieglitz and his colleagues documented the case of an infant who presented with a diagnosis of leukemia, including an enlarged liver and spleen, and elevated white blood cell counts. The child had a rash that was biopsied, and he was subsequently diagnosed with JMML, an aggressive type of cancer that occurs in about 1.2 children per million annually.
But a traditional JMML-specific DNA-based mutation screen didn't turn up any RAS mutations, a hallmark of the disease. "Virtually all patients with this disease have a mutation in one of a handful of RAS family genes, and to our surprise this patient had no RAS mutations," Stieglitz said.
The physicians started treating the child with chemotherapy, gradually increasing the dose in a bid to wipe out the leukemic cells and ready the patient for a stem cell transplant. But as the chemotherapy wasn't working, making a stem cell transplant impossible, the UCSF researchers decided to look more broadly in order to suss out whether there were any treatable mutations underlying the patient's disease.
They decided to use a more exhaustive DNA-based panel called the UCSF500, a tumor-normal analysis of 500 genes related to cancer. Sequencing for the UCSF500 test, which was launched in 2015, is conducted at UCSF's Clinical Cancer Genomics Laboratory. They also combined the panel with RNA sequencing in order to get a full molecular profile of the patient's disease.
What they found surprised them: no RAS mutations, but instead a fusion involving FLT3 in a rare gene called CCDC88C. FLT3 is expressed on the surface of many hematopoietic progenitor cells, and signaling of FLT3 is important for the normal development of hematopoietic stem cells and progenitor cells.
"To the best of our knowledge, there's never been a reported fusion involving FLT3 in any child with any cancer," Stieglitz said. "But FLT3 is upstream of the RAS pathway, and when we saw the fusion, we immediately knew that this was driving the patient's cancer, and it finally explained why he had JMML but did not have a regular RAS mutation."
By this point, Stieglitz noted, the patient was critically ill and had already had his spleen surgically removed as a palliative measure. The clinicians had no treatments left to offer him, and he wasn't a candidate for a stem cell transplant because he was so critically ill. So, the team decided to try treating him with an oral FLT3 inhibitor called sorafenib.
Sorafenib — which is co-marketed by Bayer and Onyx Pharmaceuticals as Nexavar — is a kinase inhibitor approved for the treatment of advanced renal cell carcinoma, advanced primary hepatocellular carcinoma, FLT3-ITD positive acute myeloid leukemia, and radioactive iodine-resistant advanced thyroid carcinoma.
"We didn't really know what dose to use because it was an infant and sorafenib has never been formally tested in infants," Stieglitz said. "So, we started at half of the adult dose, adjusted for weight, and within 72 hours his white count had returned back to normal. Two weeks later, we performed a bone marrow analysis and the patient was in a morphologic remission for the first time since we had met him."
The researchers eventually added azacytidine — a chemotherapeutic agent used to treat myelodysplastic syndromes — to the boy's treatment regimen, and they were able to perform a stem cell transplant using one of his parents as a donor after 10 weeks of treatment. Sorafenib was stopped after nearly two years, and the patient is still in remission.
"The DNA sequencing that we did using UCSF500 was because he had no response to traditional chemotherapy and because he had no known mutations on a disease-specific panel, so we wanted to look a little bit broader," Stieglitz said. "We used the RNA-seq as a confirmatory assay because DNA-seq is not ideal for fusions — it doesn't really tell you if the fusion is in frame or out of frame. We wanted to confirm that on an RNA-seq based assay."
Importantly, however, the success the researchers had with this case may point to some best practices for pediatric oncology clinicians and researchers. While most pediatric oncologists are forming a consensus on the utility of DNA and RNA sequencing for kids with cancer, the necessary breadth of sequencing is something they haven't yet agreed on: While some say that whole-genome and whole-transcriptome sequencing are necessary, others argue that whole-exome sequencing or targeted RNA panels may suffice, at least in certain circumstances.
However, some prominent oncologists are following a similar route Stieglitz and his team did, using broad cancer gene panels to determine the possible underlying causes of various patients' cancers and find a targeted treatment.
At Memorial Sloan Kettering Cancer Center, pediatrics department Chair Andrew Kung uses the Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) assay — a hybridization capture-based NGS assay for targeted deep sequencing of all exons and selected introns of nearly 500 cancer genes — to sequence pediatric cancer patients that come to the center for treatment.
Meanwhile, Tim Triche, co-director for the Center of Personalized Medicine at Children's Hospital Los Angeles (CHLA), has gotten the most useful clinical results from a targeted panel that he and his team at CHLA have developed in partnership with Thermo Fisher Scientific called the Oncomine Childhood Cancer Research Assay, or OncoKids. The test can assess the full coding regions of 44 cancer predisposition loci, tumor suppressor genes, and oncogenes; hotspots for mutations in 82 genes; and amplification events in 24 genes; as well as 1,421 gene fusions that have been shown to be clinically relevant in a variety of childhood cancers.
At UCSF, Stieglitz and his team test about 50 percent of the patients they see with the UCSF500 panel. It's been helpful from a few different perspectives, he noted. In patients such as the little boy with JMML described in this case, the panel can help pinpoint a target for precision therapy.
"[UCSF500] also has reclassified diagnoses for a number of patients where morphologically we assume it's one type of malignancy, but then when you look at the sequencing level you realize that it's actually a mutation that belongs to a different disease, and switch that type of therapy that's offered," he added. "And because it's one of the few broad cancer panels that actually uses a tumor and normal approach from each patient, we also see that about 10 to 30 percent of patients, depending on the type of cancer, actually have an inherited predisposition to develop a malignancy in the first place. So that helps in terms of counseling other family members, and it can also help because sometimes having an inherited predisposition to cancer actually changes your approach to treatment."
Stieglitz also said that about 10 percent of the patients who get tested with the UCSF500 panel also received RNA-seq. This is mostly because while UCSF500 is CLIA-approved, most RNA-seq is still done at a research level.
"There are some targeted RNA-based platforms that are in a CLIA environment. I think RNA-seq is certainly better for fusion detection, [and it's] also really helpful to look for expression of genes," he added. "But [there are] some practicalities around CLIA environments and around financial resources and reimbursement. Right now, I'm not really familiar with many RNA-seq approaches that are reimbursable by insurance."
Those considerations aside, however, Stieglitz does believe that targeted DNA panels and RNA sequencing — rather than whole-genome sequencing — will become the norm for testing children with cancer and matching them to treatments.
"I personally have not found whole-genome sequencing to be particularly informative in a clinical setting. The number [of genes tested] doesn't have to be exactly 500, but whatever that panel is that has the most common cancer related genes, it's pretty rare to see something that is actionable or targetable outside of [that]," he said.
More broadly, this case has implications for a number of ways children with cancer are treated. For one thing, Stieglitz said, it proves the utility of DNA and RNA sequencing, but moving beyond n-of-1 anecdotes will require the testing to get more comprehensively covered by insurance companies.
Further, this case shows the potential good that can be done by appropriately treating pediatric patients with targeted therapies that were originally developed for adults, and doing it in a timely fashion rather than as a last resort.
An interim analysis of the National Cancer Institute-Children's Oncology Group Pediatric Molecular Analysis for Therapy Choice (MATCH) study — a precision medicine clinical trial for pediatric cancer —found in May that 24 percent of participants are eligible to receive treatment with a targeted therapy, a match rate significantly higher than the 10 percent the researchers expected at the outset of the trial. A similar trial, PrEscription of Intra-Dialytic Exercise to Improve quAlity of Life in Patients With Chronic Kidney Disease (PEDAL), is being conducted in pediatric leukemia patients.
"Assuming there is a drug that's effective and there is a target [in the patient], we should use it. And I think one of the reasons that this [JMML] case was so successful is that we didn't wait for the patient to relapse three or four times before starting a novel treatment," Stieglitz said. "The patient did not respond to one or two different types of chemotherapy, and we used this targeted inhibitor relatively quickly in the patient's disease course. That's when you're most likely to see a response."
That doesn't mean doctors should abandon conventional chemotherapy or immediately look for adult drugs to use on pediatric patients, but "this constellation of having a good target and having a good targeted therapy is still relatively rare, [so] when it happens we should take advantage of it," he added.
Importantly, this case shows the need for clinicians to think outside the box. "When a patient doesn't respond to conventional therapy and you don't find a mutation that's supposed to be common in that disease, start looking more broadly," Stieglitz advised. "This case highlights that when patients are not responding the way you'd expect, use DNA- or RNA-based panels, and when possible, implement precision medicine."