Washington, D.C.-Oral squamous cell cancer, or OSCC, arises in the cells that line the mouth. The disease is often diagnosed at a late stage, which comes with a poor prognosis, and 450,000 people die from it every year, according to the International Agency for Research on Cancer. Many studies have reported microbial alterations associated with the disease, but with inconsistent results.
A meta-analysis published this week in mSystems may help clear up the conflict. The researchers behind the work combined data from 11 previous studies to identify microbial biomarkers in the saliva and associated with OSCC. The alterations were consistent across the studies, and the team also reported that the disease seems to disrupt known microbiome patterns associated with age and gender.
That finding suggests "fundamental changes in how the disease affects host-microbe interactions," said Zhenjiang Zech Xu, Ph.D., a microbiologist at Southern Medical University in Shenzhen, China, and senior author on the work. "We identified consistent patterns that might have been obscured in individual studies."
Xu, whose lab focuses on interactions between the host and microbiome, undertook the study both because oral cancer is a lethal disease with poor prognosis, and because saliva samples are noninvasive and easy to collect. "This is an attractive area for developing potential early detection methods and investigating the role of oral microbes in the cancer progression," Xu said.
Many previous studies have compared the genome of saliva in healthy people to that in people diagnosed with OSCC and reported a connection between changes in the oral microbiome and oral cancer. Those findings suggest microbial changes may promote tumor growth, but they've also unearthed discrepancies. Some have reported that the alpha diversity of oral microbes in people diagnosed with cancer is higher than that of healthy people; other studies have reported the inverse. Similarly, some studies have reported higher abundance of specific taxa, while other studies have reported lower abundance of those same taxa.
After integrating data from 11 studies, Xu said, the researchers identified alterations consistent across all the studies. They reported that taxa including Streptococcus, Lactobacillus and Prevotella were enriched in OSCC samples, for example. Howerver, they also found that OSCC disrupts microbial patterns that are associated with age or gender in healthy people. That observation, Xu said, suggests that OSCC may not only disrupt individual taxa but also change the way the host and microbiome interact in the mouth.
Once they'd identified biomarkers, the researchers trained a machine learning model on those alterations and found that it could accurately predict disease status. In addition, the model's accuracy improved when it trained on more data. "This suggests potential for developing a saliva-based screening tool for OSCC," Xu said.
He also cautioned that the science isn't quite there yet. First, he said, researchers need to validate these biomarkers in prospective studies and establish standards for the best ways to collect and analyze samples.
Xu said the group plans to continue studying demographic differences in the microbial signatures of OSCC. "They could help explain why OSCC risk and progression vary among different populations," he said. They are also planning a large-scale study that follows high-risk individuals over time, tracking how their oral microbiome changes. According to Xu, better understanding of the connection between oral microbiome dysbiosis and oral cancer will help researchers develop better therapeutic and diagnostic tools.
