Skip to main content
Premium Trial:

Request an Annual Quote

Deep Lens Integrates Genetic Test Results Into Clinical Trial Prescreening


CHICAGO – Startup bioinformatics vendor Deep Lens is now integrating molecular data into its core platform for clinical trial screening and enrollment, thanks to an exclusive deal to license technology developed at the University of Miami.

Last month, Columbus, Ohio-based Deep Lens licensed an integration engine developed by the Sylvester Comprehensive Cancer Center and UHealth Information Technology, both part of the University of Miami.

The technology harmonizes and normalizes genetic test results from laboratories run by companies including Caris Life Sciences, Foundation Medicine, Guardant Health, NeoGenomics, and Tempus, enabling Deep Lens to deliver a simplified, standardized picture of what is happening to any given patient at the molecular level and providing physicians with knowledge about specific mutations that may qualify patients for specific clinical trials, according to Deep Lens Cofounder and Chief Scientist TJ Bowen.

"What we provided them was a program and a coding language that allows the user to extract structured data from patients' genomic reports," said Jared Cotta, project manager at the University of Miami Sylvester Cancer Center. Cotta coordinates the precision oncology program there.

Deep Lens has already integrated the Miami technology into its flagship product, VIPER, which stands for Virtual Imaging for Pathology Education and Research and combines artificial intelligence with advanced pathology workflows to enable peer-to-peer collaboration and patient identification for clinical trials.

Trial coordinators historically have not had access to genetic test results or if they did, they have lacked the tools to extract information from reports that might help in identifying potential participants. By adding the Miami integration engine to VIPER, cancer care teams and trial coordinators can search test results to streamline the matching process, Deep Lens executives said.

By identifying eligible patients at the time of their diagnoses, VIPER can help fast-track trial enrollment and potentially shorten the duration of trials, according to the firm.

Bowen said that the advent of precision medicine has exacerbated the challenges of clinical trials recruitment.

"The technology [to practice precision medicine] is amazing, and it's really helping patients, but it really causes a burden on the care team to be able to find the right patients and the right studies to get them together," he said. "Our technology really helps bridge that gap and get the patients connected to the right studies."

Deep Lens Cofounder and President Simon Arkell noted that there are few prescreening technologies available for trial coordinators and principal investigators. He noted that some larger and better-funded institutions have internally developed technologies and processes, but those have limitations, and "many doctors don't even know which clinical trials are running at their own institution."

He estimated that 95 percent of cancer centers he talks to do not have any technology more sophisticated than spreadsheets and email for improving trial enrollment.

Bowen said that a tracking system like VIPER is growing in importance as cancer trials become more complex, with multi-arm and basket studies. "You have a really dynamic, changing landscape of clinical trials that is really hard to stay abreast of," he said.

"Without some sort of technological advantage, it's really hard to keep track of what's going on and when trials are opening, when they're closing, when they're pausing, when a patient during their journey is changing from not qualifying to qualifying, so there's just a lot of moving parts," Bowen added.

"We believe it's a travesty of justice that patients are never even finding out about trials that they want to be on or that they qualify for," Arkell said. "There is just a huge disparity between the need and the reality, and we're looking to fill that."

Bowen said that the company also is interested in compressing the timeline of clinical trials. "One of our missions is to get the patients on the right therapy at the right time, and if we can help accelerate the time that those therapies are able to get to the patient, it helps all patients," he said.

Deep Lens was founded in 2017 to "democratize" patient screening and enrollment into oncology clinical trials, the executives said.

The firm emerged from stealth mode in October 2018 by announcing a seed funding round. Deep Lens closed a $14 million Series A round in April 2019 and has raised $17.5 million to date. 

To start the company, the founders purchased a digital pathology imaging workflow and collaboration platform from Nationwide Children's Hospital in Columbus that had been in development for nearly a decade and evolved it into something that supports clinical trial recruitment and enrollment. Deep Lens Cofounder and CEO Dave Billiter created the technology when he was director of informatics at what is now called Wexner Research Institute at Nationwide Children's.

Because of this heritage, VIPER integrates with pathology and laboratory information systems as well as electronic medical records. With the Miami technology, the system now can pull in genetic testing results to aid in the identification of mutations, Arkell said.

Bowen said that the Deep Lens initially added the EMR integration to the digital pathology platform because the search for patients who meet criteria for clinical trials starts in the clinical record. In precision medicine, the record necessarily includes genetic test reports, but those reports typically come back from the lab as PDF files that are not machine-readable. Deep Lens attempts to extract bits of information as discrete data, but molecular tests are complex and vary so much from lab to lab.

The Miami technology parses next-generation sequencing data from molecular testing labs on genes, mutations, cancer type, and tumor location. "Anything that is reported back on the NGS report is something that we gather as well," Cotta said.

NGS data is usually already structured in an XML format, but clinician-friendly reports lose some of the machine readability. "In the reports that are received from the NGS companies, it looks nicer in the report format, but from a structured-data standpoint, it's not useful," Cotta said.

The Miami integration engine addresses that issue by harmonizing data from different labs and structuring the information in a way that allows VIPER to query the records.

According to Bowen, VIPER brings capabilities to community oncology clinics that were previously only available to major comprehensive cancer centers.

"When you have technology that can alert you to potential treatments for patients who are getting reports back that suggest they might qualify, that really helps with [clinics] understanding where their patients are," he said.

"That's where the democratization comes in, because now a smaller network of oncology clinics or smaller centers is now able to compete with the large comprehensive cancer centers even without the personnel," Arkell said.

A poster prepared for the American Society of Clinical Oncology conference that was held online this year due to the COVID-19 pandemic showed that VIPER helped a single clinical research coordinator at a community cancer center manage prescreening for 20 different oncology trials at once, covering 11 different cancer types and 12 biomarkers. Over a four-month period, this research coordinator was able to prescreen 5,700 patients and identify 150 previously undetected patients, according to the presentation.

VIPER helps research institutions overcome issues including the complexity of data needed to support precision medicine, lack of care teams' access to data, narrow enrollment windows, and inadequate tools for managing prescreening data, Deep Lens said. Patient-facing sites including,, eTACTS,, and TrialReach are "helpful," according to the poster, but they are not all that clinician-friendly, much less integrated into EMRs.

While many bioinformatics companies aggregate data, Arkell said that Deep Lens stands out because it focuses exclusively on patient prescreening for clinical trials in oncology.

"A cancer center that has thousands, if not tens of thousands, of patients coming through the doors each year, making sure that you're tracking them in a workflow so that no one falls between the cracks is very, very important," he said. "There are technologies out there that may use some AI … but we have a holistic approach where the technology is actually used by the research coordination team."

VIPER also allows for the identification of qualified patients who are actually available when a trial is recruiting.

According to Bowen, other technologies are query-based and put the burden on trial coordinators to find specific data. "Our tool really is more passive and proactive. It has algorithms and systems running behind the scenes that will actually surface that information to the team as opposed to making them go out and look for it," Bowen said.

"If you're doing this in specific windows of time, you're going to miss patients. But if something is constantly looking in the background for you, we're going to surface those patients when the right time comes," he added.

Aivita Biomedical, an immuno-oncology company based in Irvine, California, has been using VIPER to aid in recruitment since last year.

Jim Langford, vice president of clinical operations at Aivita, has a history in contract research organizations and once owned a company specializing in electronic data capture. When he learned of Deep Lens, he saw potential in VIPER to help Aivita with its development of an autologous dendritic cell-based vaccine for ovarian cancer.

In the study, an open-ended trial at Hoag Hospital in Newport Beach, California, the company blends neoantigens from stem cells of the cancer with the patient's dendritic cells, then injects the serum eight times over six months. The goal is to improve overall survival rates, Langford said.

Hoag had been identifying just one patient a month. "We said, 'Let's see if we can do better,'" according to Langford.

Aivita, which was not part of the study prepared for ASCO, got VIPER installed at the hospital and loaded its trial protocol and inclusion/exclusion criteria into the system. Within a few months, Hoag was able to uncover nine potential subjects who qualified, and three enrolled, Langford said.

"Prior to that, I'm just taking the word of the coordinator, or PI, or whoever that they only saw one patient or they didn't see any patients this month," Langford said. "Now we can look at the same information and say, 'Well, why not this patient?' That's something I've never had before."

Langford said the impact was "dramatic" and almost immediate. "What I found most valuable about it was it gave me a common language to speak to the coordinator and the [principal investigator] regarding patients who could qualify for my clinical study."

This study has slowed a bit due to the pandemic because the debulking surgery is considered elective, so Aivita has not been able to get tumors to make its personalized product, Langford said. However, an Aivita trial of a glioblastoma therapy has continued without interruption.

Langford said that he has requested that other clinical sites consider VIPER, and said that they have been receptive. "It becomes fairly obvious to the clinical sites that this could be something of value," according to Langford.

Deep Lens is focused exclusively on cancer, but the technology could be adapted for other applications and disease states in the future.

"A long-term goal is that this could become a pervasive platform for all clinical trials, but we're not going there yet," Arkell said. "We want to make sure we get it right before we move on to other things," Bowen added.

Billiter told 360Dx in 2018 that the firm had planned to seek clearance from the US Food and Drug Administration and foreign regulatory agencies after validating its technology for primary clinical diagnosis. However, a company spokesperson said this week that FDA review is no longer necessary because Deep Lens has since moved away from clinical applications of its technology.