As momentum grows for implementing liquid biopsy in clinical practice, the technology for detecting cancer DNA circulating throughout the body is advancing on another front: saliva as the biopsy medium. Saliva offers a noninvasive alternative to blood draws, with testing potentially performed at the point of care and as sensitive and specific as other platforms. But challenges remain in finding a clinical niche for saliva liquid biopsy in an increasingly competitive field of more mature blood-based technologies.

Considerable research exploring saliva liquid biopsy has taken place at the University of California, Los Angeles (UCLA) School of Dentistry, under the watchful eye of David Wong, DMD, DMS, Felix and Mildred Yip professor and associate dean for research. From early funding in 2002 by the National Institute of Dental and Craniofacial Research, Wong’s lab has been developing iterations of saliva-based diagnostics, most recently with a self-contained desktop device called Electric Field Induced Released and Measurement (EFIRM).

Wong recently presented clinical trial data involving EFIRM’s use to detect in patients’ saliva the presence of mutations in the epidermal growth factor receptor (EGFR) gene, a predictive biomarker for metastatic lung cancer drug treatment (J Clin Oncol 2016, 34 suppl; 8520). The UCLA group reported almost the same results as achieved with tumor tissue DNA, the gold standard. “We are seeing a performance, a concordance, that is too good to be true,” said Wong.

Moving Beyond Proof of Concept

Though the trial produced encouraging results, it included just 37 lung cancer patients, too few to reach any definitive conclusions, especially since only six had EGFR mutations. “The authors’ results are very promising, but are only proof-of-concept promising,” said Cloud Paweletz, PhD, head of the Translational Research Laboratory at the Dana-Farber Cancer Institute’s Belfer Center for Applied Cancer Science in Boston. “Prospective validation based on the authors’ initial results are needed.”

The UCLA group envisions a broad use for saliva-based liquid biopsies in the early detection of cancer, including pancreatic cancer, but chose the lung cancer predictive biomarker mutations to establish proof of principle for EFIRM. Thoracic oncologists typically identify these lung cancer mutations from tumor tissue obtained from a lung biopsy, but because a biopsy isn’t always an option, mutations can be identified from plasma. EFIRM has the potential to be a point-of-care alternative to laboratory-based testing, since it can generate results within 15 minutes, according to Wong.

EFIRM in Detail

EFIRM, invented by UCLA bioengineer and adjunct assistant professor Fang Wei, PhD, while working on her doctoral dissertation at Peking University, uses oscillating electric fields to bring macromolecules into contact with an array of conducting polymer electrodes decorated with capture probes. The charged surface draws down macromolecules with a net positive charge, Wong explained, and reversing the field polarity repulses nonspecific sequences. The capture probes can be designed for nucleic acids or proteins.

For tumor DNA, separate detector probes, labeled with fluorescein isothiocyanate, mix with the saliva sample after the latter has been centrifuged to remove cells and stabilized with RNAse. Then, anti-fluorescein antibody conjugated to horseradish peroxidase binds the detector probes, horseradish peroxidase substrate is added, and the amperometric signal generated by the subsequent redox reaction is measured. EFIRM requires just a 40 µL sample, according to Wong. Unlike polymerase chain reaction (PCR)-based assays, he stressed, “This is direct detection.”

EFIRM is not the only device being developed for saliva diagnostics for systemic diseases. Investigators at the University of Kentucky College of Dentistry’s Center for Oral Health Research have used a device that contains a modular and miniaturized sensor system and universal analyzer, now being further refined by John McDevitt, PhD, chair of biomaterials and biomimetics at New York University College of Dentistry. They also are experimenting with a point-of-care lateral flow device. Jeffrey Ebersole, PhD, professor and associate dean for research and graduate studies at Kentucky, thinks UCLA’s EFIRM could also be broadly useful. “There’s no reason that it could not be used to look at some of the same molecules that we’re looking at in other diseases,” he said.

Challenges Ahead

Wong hopes ultimately to gain Food and Drug Administration approval for EFIRM. In two proof-of-principle studies in lung cancer—the aforementioned investigation involving 37 patients, and another with 40—Wong and his team compared in a blinded fashion EFIRM-analyzed saliva and plasma samples with tumor tissue DNA analyzed via PCR, and found excellent concordance. In both trials, there were a few cases in which the electric current values for wild-type tumors overlapped with the electric current generated by mutant tumor DNA, so EFIRM wasn’t perfect. Wong indicated that an improved EFIRM is now “robust for clinical performance.”

Strictly speaking, the researchers couldn’t report a false positive or false negative rate, because they hadn’t established a cutoff to define the presence of these mutations—the mutation signal is treated as a continuous variable, a challenge that will need to be remedied in order for EFIRM to advance. “In the diagnostic setting you can’t look at a whole population of people and create a new cutoff; you have to have the cutoff in advance for an individual sample,” said University of Pittsburgh thoracic oncologist James Herman, MD. Wong acknowledged this. The cutoff “needs to be in place by the time EFIRM turns into a clinical assay, and hopefully that will be in the next six months,” he indicated.

Wong is now preparing to analyze the serum of 300 lung cancer patients archived at NYU Langone Medical Center in New York City. But this trial will analyze blood plasma, not saliva, and will be a retrospective analysis, introducing possible selection bias. Wong said his former postdoctoral scholar, Wei Liao, PhD, will launch a large trial in China next year, sponsored by the company Liao founded, EZLifeBio. This new China trial could lead to regulatory approval there, according to Wong. Meanwhile, he is actively seeking a sponsor for future trials in North America.

Even as Wong and his colleagues work to advance EFIRM towards clinical practice, Herman noted that as a liquid biopsy EFIRM’s current design does not detect the most clinically useful EGFR mutations in lung cancer. As now configured, EFIRM tests the exon 19 and L858R mutations, which predict response to the EGFR inhibitors, gefitinib and erlotinib, in metastatic non-small cell lung carcinoma. But virtually all these patients already get a lung biopsy—it’s required for diagnosis—and the tumor tissue gets tested for EGFR mutations. Occasionally a patient presents with metastatic disease so obvious that a tissue biopsy isn’t necessary, and a liquid biopsy might be useful. “That specific application would be rather small,” said Herman. “The real place where people are looking at liquid biopsies is in the progressive setting, when therapeutic resistance occurs.” One EGFR mutation, T790M, accounts for most of these drug-resistant cases. UCLA is now optimizing EFIRM for the T790M mutation, according to Wong. “It will be important to take the device to the next level,” he said.

As Wong and his colleagues continue developing EFIRM as a tool in lung cancer diagnostics, they are also working on a separate line of investigation involving pancreatic ductal adenocarcinoma (PDAC). The team recently reported creating EFIRM probes for the top four KRAS mutations (G12D, G12V, G12R, and G12C with > 90% frequency in total) associated with PDAC (J Clin Oncol 2016, 34 suppl; e15702).

Still, Wong acknowledged that EFIRM has a long way to go to make its mark in clinical practice. As he put it, “Hopefully these efforts…could reach a point that could catalyze a momentum.”

Ken Garber is a freelance writer in Ann Arbor, Michigan. +Email: [email protected]