Probing saliva for clues about oral cavity and oropharynx cancer

From the Chen and Raftery Groups, Public Health Sciences Division

Oral cavity squamous cell carcinoma (OCC) and oropharyngeal squamous cell carcinoma (OPC) are common forms of oral cancer, with an estimated 51,540 new cases expected to be diagnosed in the US in 2018. Major risk factors for the development of these cancers include tobacco use, heavy alcohol consumption, and infection with a specific type of human papillomavirus, HPV16. Survival rates for OCC and OPC vary widely dependent on stage at diagnosis, with the 5-year relative survival rate as low as 20% for patients whose cancers have spread to distant sites. The ability to correctly diagnose metastasis to the neck lymph nodes in OCC and OPC patients has been challenging, as it relies on imaging techniques that have low sensitivity to detect nodal metastasis. Thus, the identification of more sensitive biomarkers that could detect OCC and OPC at earlier stages or distinguish between nodal and non-nodal cases would be highly beneficial. Researchers in Dr. Chu Chen’s group in the Public Health Sciences Division investigated whether differences in metabolite levels in saliva could distinguish between OCC/OPC cases and controls. Their results were recently published in the journal PLoS One.

Previous metabolomic studies of saliva, serum, or plasma samples collected from OCC or OPC patients have been successful in distinguishing the cases from controls but none assessed whether metabolite profiling could identify between nodal positive and nodal negative patients. “Metabolomics is a powerful approach to simultaneously examine aqueous and lipid metabolites in the sample,” said Dr. Pawadee Lohavanichbutr, Fred Hutch Staff Scientist and first author of the study. To investigate metabolite profiles, the authors conducted an extensive metabolomics-based study by analyzing saliva samples using four different metabolomic approaches: nuclear magnetic resonance (NMR), targeted liquid chromatography-mass spectrometry (LC-MS) of aqueous metabolites, global LC-MS of aqueous metabolites, and global LC-MS lipidomics. Two sets of one hundred samples each were included in the study. The first set was comprised of twenty controls and eighty OCC cases with known nodal status. The second set was comprised of twenty controls and a mix of eighty OCC and OPC patients. Of all metabolites detected across the various metabolomics platforms, 453 overlapped between the two sample sets and all downstream analyses focused only on these metabolites.

Graphical representation of pathway analysis results based on metabolite levels in saliva samples of cases and controls.
Pathway analysis results based on metabolite levels in saliva samples of cases and controls. Node color and size reflect strength of p-value (color) and impact on pathway (size). Image from publication

First, the authors assessed whether the salivary metabolite profile of OCC patients differed from that of controls. In the first set of samples, three metabolites were significantly different across three of the four platforms, while one platform (lipidomics) did not detect any significant differences. Amino acids glycine and proline were significantly lower in OCC cases as compared to controls, a finding that was validated in the second set of samples. Both metabolites remained significantly different even after controlling for smoking and drinking status. In addition, four metabolites were significantly different between early stage OCC patients and controls in both sets of samples. However, among the 79 OCC patients with known nodal status, levels of 21 metabolites differed between the node positive and node negative groups but none were statistically significantly different. Additionally, there were no significantly different metabolites in OPC patients as compared to controls.

The authors then conducted pathway analyses to investigate which metabolic pathways might be most prominent in differentiating OCC cases from controls. The analysis identified five pathways that demonstrated both high significance and high impact (see Figure). Four of the pathways were specific to amino acid metabolism while the citric acid cycle was the fifth pathway. Proline and glycine metabolic pathways were among those identified in this pathway analysis. Glycine availability is important in nucleotide synthesis which is essential for cell proliferation and tumor growth. The results from this metabolomics study may be consistent with what is known regarding cancer cell metabolism, as cell culture studies have shown that in cancer cell lines, glycine is rapidly consumed and subsequently used for synthesis of nucleotides. Thus, it is hypothesized that the lower levels of glycine in saliva may be due to the OCC tumor cells taking it up to support tumor growth.

Although there were no differences detected between node positive and node negative OCC patients, other results from this study are encouraging. Dr. Lohavanichbutr summarized, “Our result showed that salivary metabolites can distinguish patients with oral cancer from patients without oral cancer. Collecting saliva for testing is non-invasive and has minimal burden to the patients. If confirmed by other studies, our finding could become a convenient test for early detection of oral cancer.”

Cancer Consortium authors include Chu Chen and Daniel Raftery.

This research was supported by the National Cancer Institute.

Lohavanichbutr P, Zhang Y, Wang P, Gu H, Nagana Gowda GA, Djukovic D, Buas MF, Raftery D, Chen C. 2018. Salivary metabolite profiling distinguishes patients with oral cavity squamous cell carcinoma from normal controls. PLoS One. doi: 10.1371/journal.pone.0204249