NEW YORK (GenomeWeb) – University of Michigan researchers have found that for pediatric cancer patients, a comprehensive sequencing protocol was able to identify an actionable mutation in about 46 percent of patients and led to a clinical action in 54 percent of the group.
Reporting today in the Journal of the American Medical Association, Arul Chinnaiyan's lab described implementing the team's MiOncoSeq protocol, which it launched for adult cancer patients in 2011 and pediatric cancer patients in 2012. The pediatric version is a prospective, observational clinical case series with the goal of studying the feasibility and utility of integrating clinical sequencing in the management of kids' cancer.
Of the 104 patients initially screened, 91 who had sufficient tumor tissue for sequencing enrolled. All patients and families received genetic counseling throughout the study. The 91 children all had relapsed, refractory, or rare cancer, 28 with hematological malignancies and 63 with solid tumors.
The researchers performed tumor and germline exome sequencing as well as tumor transcriptome sequencing.
Because of the possibility of secondary findings in the germline DNA, the researchers implemented a "flexible default" consent model, where all findings that would influence the current cancer management would be disclosed, but parents could choose whether to receive secondary findings associated with hereditary cancer syndromes. Secondary findings not related to cancer were not disclosed. Families from 80 of the 91 patients agreed to receive secondary findings.
A multidisciplinary precision medicine tumor board that included both pediatric and adult oncologists, geneticists, pathologists, biologists, bioinformaticists, bioethicists, genetic counselors, study coordinators, and ad hoc experts met weekly to deliberate sequencing results. Select findings were validated with a CLIA test and returned to the oncologist.
Median turnaround time from study enrollment to case presentation at the tumor board was 53 days, longer than the group's anticipated three to four weeks. The authors attributed the longer than expected turnaround time to bioinformatics analysis and wait times between tumor board meetings.
Potentially actionable findings were identified in 42 out of the 91 patients, or 46 percent, and secondary findings in nine patients, or 10 percent, of the group. Both exome sequencing and transcriptome sequencing contributed to actionable findings.
Of the 28 patients with hematological malignancies, actionable findings were identified in 15 patients.
In one patient, the team found a cryptic gene fusion involving ETV6 and ABL1 that was identified by transcriptome sequencing, but "had not been detected by other standard diagnostic tests including cytogenetics and fluorescence in situ hybridization for the BCR-ABL fused genes," the authors wrote. In fact, transcriptome sequencing contributed to about 20 percent of the total actionable findings, which would have otherwise been missed, the authors said.
The ETV6-ABL1 fusion indicated sensitivity to the tyrosine kinase inhibitor, imatinib. The patient had previously not responded to standard therapies and after the sequencing was placed on imatinib and chemotherapy. She did not tolerate chemotherapy, however, so was treated with imatinib alone and went into remission.
Other actionable findings in the patients with hematological malignancies included a cryptic EBF1-PDGFRB gene fusion in a patient with refractory pre-B cell acute lymphoblastic leukemia and FLT3 kinase mutations in three patients.
Of the 63 patients with solid tumors, actionable findings were identified in 27. In one patient, sequencing even clarified a diagnosis. The patient had originally been diagnosed with infantile myofibromatosis, but sequencing clarified the diagnosis as high-grade spindle cell sarcoma. The patient was negative for the ETV6-NTRK3 gene fusion, which is found in nearly 90 percent of patients with infantile fibrosarcoma, but transcriptome sequencing found a novel in-frame fusion of the LMNA-NTRK1 genes, which is "functionally analogous" to the ETV6-NTRK3 fusion, the authors wrote.
The finding of the fusion also suggested that the patient's treatment should be changed, as NTRK1 fusions are typically sensitive to crizotinib.
"Within six weeks of starting therapy, she achieved a partial remission and has since maintained a favorable response to crizotinib for more than eight months without experiencing major toxic effect," the authors wrote.
Overall, the researchers acted on the findings of 23 patients, including changing treatments for 14 patients and providing genetic counseling about future cancer risk and familial risk for nine patients.
Nine of the 14 patients whose treatment was altered experienced partial or complete clinical remission. For five patients, the treatment change was unsuccessful.
The authors said that the main reasons for not being able to act on a patient's findings were that either the patient was already in remission and the genomically informed treatment was poorly defined, or the treating physician thought no additional therapy was necessary. Other challenges included limited access to drugs or results coming too late.
Although patients and the families were not charged, the group estimated that costs were around $6,000 per patient, which included sequencing, analysis, and associated labor costs.
Going forward, the authors said that more research, including a multi-institution study involving a control group, would be needed to truly assess "whether the treatment changes based on the study actually improved clinical outcome compared to standard of care."
In an editorial published alongside the JAMA study, researchers from the Children's Hospital of Philadelphia wrote that the "data strongly suggest that precision medicine enhanced by genetic evaluation may improve the outcomes of children with cancer." However, they added that the "most troubling reality highlighted" in the study was the fact that a significant number of patients with refractory cancer and potentially actionable findings could not be treated based on the finding because there was a lack of data on pediatric dosage or no available clinical trials for the pediatric population.