NEW YORK (GenomeWeb) – Liquid biopsy is continuing to attract new entrants as existing attention in the field has turned away from tumor profiling in advanced cancers to applications like therapy monitoring, residual disease detection, and early cancer screening.
One new name, Lexent Bio, is planning to advance a technology for measuring and monitoring drug response in later-stage cancer patients, which it believes it can also translate to detection of minimal residual disease in patients who have had a tumor surgically removed.
The firm has released little information about its approach so far, but Ken Nesmith, one of several cofounders of Lexent and its CEO, said this week that the company aims to commercialize a platform for monitoring patient-specific signals in cell-free DNA.
The company cofounders initially developed the approach as employees of hereditary testing firm Counsyl, where Nesmith was senior director of product strategy. The company then separated out as Lexent.
Haluk Tezcan, the firm's chief medical officer, said that the technique the company is commercializing uses initial pre-treatment whole-genome sequencing of a patient's cell-free DNA to develop a bespoke signature that can then be used to track how they are responding to a particular drug.
Based on Tezcan's description, the approach shares certain aspects with techniques used in academic studies over the last few years, and with the methods described by some of the other companies working to commercialize liquid biopsy monitoring tests — namely the use of a biomarker signal specific to an individual patient's cancer to yield a readout of whether their tumor is being effectively subdued or is resisting treatment.
But in limiting its analysis to blood-based DNA — without the need to start with a patient's tumor tissue sequence — Lexent appears to have set itself apart from other approaches described by companies like Natera, which rely on genomic information from a patient's tissue sample to develop the signature that then enables blood-based monitoring.
Tezcan declined to detail what the content of the signatures Lexent gleans from WGS of cell-free DNA look like, but he suggested that they go beyond a simple set of circulating tumor mutations.
Research in the field has shown that various signals in the blood can indicate the presence of cancer, including copy number variation, and aneuploidy signals, as discovered and reported by some prenatal testing firms that have competed with Counsyl in that market.
Lexent hasn't published any data yet, but Nesmith said that the firm has been studying patients prospectively, and that the evidence from this effort has shown that its method can accurately identify signals in patients' blood that indicate they are either responding or not responding to treatment more quickly and accurately than currently used imaging techniques or clinical signs of disease.
Tezcan said that the company's trial has been recruiting patients with any solid tumor, and has included patients treated with a variety of drugs, including chemotherapy, hormonal treatments, and immunotherapies.
"It looks like [the efficacy of the method] is relatively universal, and not disease specific … and it looked pretty consistent across [these various] therapies as well," he added.
Initially, Lexent hopes to market its approach as a tool for monitoring treatment in advanced cancers, but according to Tezcan, the approach could also potentially be applied in the adjuvant setting as a way to assess residual disease in patients who have had a tumor surgically removed.
The firm already has a second clinical trial running in that vein — launched last year and described as an effort to examine the potential of circulating tumor DNA as a surrogate for disease burden pre- and post-surgery.
"That will require longer clinical development,' said Tezcan. "The technology may not be that different, but the question takes a lot longer to answer because in this setting [of recurrence prediction or monitoring] the timelines can be up to five or 10 years."
For late-stage cancer patients, the proposition of blood-based monitoring is that by offering a more accurate, more convenient, more rapid readout on how a tumor is responding to treatment, it could help get patients off ineffective drugs, and on to potentially effective alternatives sooner.
As the use of immunotherapies has grown, new monitoring needs have also emerged — specifically, methods to resolve cases of pseudoprogression: when clinical signs or radiologic imaging might suggest a patient has failed to respond to a drug, although his/her disease is actually demonstrating an initial surge ahead of what will be a strong response.
"Today imaging is the standard," Tezcan said, "so if we can come out with a test that complements or improves on imaging, that's the goal."
The cost of the company's approach is substantially lower than radiologic imaging, Nesmith added.
But the question of whether the ability to outperform current tools will translate into improved patient outcomes is still open, and something Lexent, and other companies targeting this space will ultimately have to address to prove their utility.
"Our goal is to be able to reduce delays that lead to unnecessary toxicity on ineffective therapies, and prevent patients from getting [other drugs] that could help them ... but to be clear, unless there is an effective therapy available [to switch to] what we do won't change outcomes," Tezcan agreed.
As the field continues to evolve, he added, tools like Lexent's can also help to accelerate clinical trials of new therapies that will help realize its promise.
Lexent hopes to publish the first data from its prospective analyses during the next few quarters, and is working on setting up clinical validation studies, which would begin next year, and ideally yield results within a year and half.
In the background, the company is working on developing a commercial architecture, with the intention to launch the test first as an LDT, though Nesmith said that it will likely be attractive to labs to be able to preform monitoring in house, so there may be a demand for translation to a more disseminated model.