SAN FRANCISCO (GenomeWeb) – As part of an ongoing precision oncology trial, Wake Forest Baptist Comprehensive Cancer Center has profiled the mutational spectrums of two patient populations who have been historically understudied: smokers and patients of African-American ancestry.
The team recently published results from more than 400 patients who enrolled in the trial, and has so far sequenced around 1,000 total patient tumor samples, according to Wei Zhang, a professor of cancer biology and a senior author of the study. The Theranostics study focused primarily on patients with advanced disease, but Zhang said that the team is also testing Guardant Health's liquid biopsy assay on both early- and late-stage patients as part of Wake Forest's Precision Oncology trial.
In the recent study, published last month in Theranostics, both smokers and nonsmokers, as well as patients from a variety of ethnicities enrolled, but the team focused on identifying mutations that were unique to both the smoker population and African-American population. Zhang said that the patient population at Wake Forest has a higher-than-average percentage of both smokers and patients of African-American ancestry, enabling the team to figure out whether those populations have unique mutational signatures.
African-Americans make up more than 13 percent of Wake Forest's patient population. And, Zhang said that due in part to the tobacco industry's strong presence in North Carolina, more than 60 percent of patient samples analyzed in the study were from current or former smokers.
The researchers enrolled 431 patients in the Precision Oncology trial between May 2015 and May 2016. Foundation Medicine sequenced patients' tumor samples and the Wake Forest researchers analyzed the results to try and determine whether specific genetic alterations were more prevalent among smokers and/or African-American patients. Patients could enroll regardless of the type of cancer, although the predominant cancer type was lung followed by colorectal.
The team found that patients with a history of smoking tended to have tumors with a higher mutational load characterized by mutations to DNA damage repair genes, chromatin remodeling genes, and the tumor suppressor TP53. In addition, these tumors were more heterogeneous than tumors from non-smokers. As a result, "treatment can be more challenging," Zhang said. The researchers also validated their findings in data from the Cancer Genome Atlas.
With regards to the African-American population, the researchers found that those patients were more likely to have mutations in the TP53 gene than Caucasian patients, a finding that the team validated from the TCGA data.
In addition, Zhang said, the researchers found five genes that were more likely to be amplified in the African-American patients versus Caucasian patients: MCL1, RUNX1T1, CDK8, KAT6A, and RAD21. All five genes were verified to be differentially amplified in African-American samples from the TCGA.
One reason the group wanted to focus on studying the mutational spectrum of tumors from African-American patients is that they tend to have poorer prognoses than Caucasian patients. The identification of a higher proportion of TP53 mutations in the African-American cohort could potentially point to one cause of why there are outcome disparities between populations, Zhang said, although he emphasized that further research studying larger numbers of patients would be necessary to confirm that.
With regards to the differences in gene amplifications, he said there are a number of potential causes that would need to be further investigated. For instance, he said, tobacco companies tended to market menthol cigarettes to African-Americans, which could potentially cause a different mutational spectrum than non-menthol cigarettes. In addition, the study did not examine differences in other environmental factors, which could also play a role. "We don't really know," he said. "That's not in the clinical information or the questionnaires, but we've been discussing modifying our protocols and surveys to capture that information for future study."
Zhang added that the group is continuing to validate the results of this study as well as the results from a second group of more than 400 patients to see if those findings match the initial ones. "It looks like they do, but we're still working on the analysis," he said.
In addition, he said, one ongoing challenge has been to use the genomic results to modify a patient's treatment. In the first set of patients, only around 10 percent had their treatment changed due to the genomic findings, Zhang said. Increasing that proportion is "a major push for our precision oncology program," he said.
One reason for the low percentage is that because most patients who receive genomic profiling have very advanced disease, many die before the results can be used to change management. Going forward, he said, one major push would be to streamline the whole process of testing and clinical decision making.
"We have to make this process more efficient as some cancers are very aggressive and there is not a lot of time to wait around," Zhang said. Oncologists are "often stressed for time and there is a lack of a system for automating the decision-making process," he said. Now, the Wake Forest precision oncology and informatics teams are working on making the processing smoother and more efficient, he said.
Other ongoing challenges are figuring out which patients do benefit from genomic information, "expanding the patient population who can benefit from this information, and identifying more patients who can be treated by targeted therapy," Zhang said.
Nonetheless, the study authors pointed to several examples of where Foundation Medicine sequencing enabled physicians to change the patient's management. In one case, a male smoker with metastatic colorectal cancer was found to have a very high tumor mutation burden and was put on nivolumab, a checkpoint inhibitor. Four months later, a CT scan revealed that the patient had stable disease. Nivolumab was also used to successfully treat an African-American nonsmoker with metastatic colorectal cancer. She did not have a hypermutated tumor, but sequencing revealed amplification of PD-L1 and PD-L2. Six months after starting nivolumab the metastatic nodules that had developed in her chest, peritoneal space, and skin began to "melt away," the authors wrote, and the patient was tumor-free after eight months.
The tumor profiling study using Foundation Medicine's FoundationOne test is just one part of Wake Forest's precision oncology program. It also has a liquid biopsy component, using Guardant Health's Guardant 360 assay. In May, the researchers reported in the Journal of Hematology and Oncology results from 177 patients who had been analyzed with the Guardant 360 assay. Zhang said the importance of that study was that the team found that mutations could be found in circulating tumor DNA from early-stage patients.
Nonetheless, Zhang said that profiling tumors of early-stage patients poses its own challenges. For one, he said, many patients respond well to current standard treatment regimens, so there is not always a case to be made to perform tumor profiling. If tumor profiling identifies a potential targeted therapy, but the patient is already responding well to the current standard treatment, then you have to justify why you'd do something different, he said. "There will need to be a different strategy for early-stage patients [than] for late-stage patients where there's more urgency."
In total, around 1,000 patients have had their tumors analyzed via Foundation One and 400 patients have had Guardant's liquid biopsy assay as part of Wake Forest's Precision Oncology trial. Zhang said that the goal is to double these numbers over the next year and to eventually make tumor sequencing standard of care.
However, there are still challenges, particularly with regards to how testing is paid for. Currently, testing is partially offset by the companies. Wake Forest is a participant in the Foundation Medicine-led Precision Medicine Exchange Consortium, whereby member organizations share de-identified data, including outcomes data.
Going forward though, Zhang said figuring out how these tests are paid for will be a priority, adding that ideally, insurance companies would begin to cover more of the tests' costs.