NEW YORK (GenomeWeb) – With its goal of streamlining the process of collecting patient data, Cambridge, Massachusetts-based start-up Novaseek is tackling one of the more challenging portions of the diagnostics and drug-development processes.
Assembly of suitable patient cohorts is among the more expensive and time-consuming parts of such research, and has proven a significant hurdle for many companies, particularly in the molecular diagnostics space where, even after developing a marker or test, firms have struggled to stay afloat financially while generating the large amounts of clinical data required to prove clinical utility and drive adoption.
Novaseek CEO Kate Torchilin suggested that these challenges present an opening for her firm, which offers a cloud-based platform to aid researchers in assembling and tracking clinical study cohorts. She added that her previous experience in the clinical diagnostics space has given her a firsthand look at the issue.
Prior to joining Novaseek, Torchilin was vice president and general manager of connected health at Alere. From 2003 to 2008 she was the director of global clinical and toxicology markets at Thermo Fisher Scientific where she helped establish the company's clinical and toxicology mass spec business.
"My team was actually launching clinical applications of mass spec into diagnostic labs, so both the discovery and validation of [clinical] assays is something that I spent many years being directly involved in," she said.
"The challenge that I repeatedly observed throughout my life in diagnostics and research tools and pharma R&D is that disconnect between scientists and clinical researchers in the lab and their ability to formulate experiments and do the discovery that reflects the clinical significance of the disease in the real world," she noted. "That gap was really very, very obvious, both in my groups and other groups around me trying to develop new biomarkers and then launch them."
Novaseek is addressing this problem through its cloud-based Clinical Data Network for Research (CDNR) platform, which allows researchers to define patient cohorts, request specimens, and track, manage, and analyze study data.
The company aims to bring together what Torchilin said have traditionally been two separate portions of the patient data collection workflow, providing what she described as an "end-to-end solution."
Companies like 12B2 (funded by NIH) or Explorys (part of IBM), help researchers develop specific patient cohorts for clinical trials based on de-identified data gathered from sources around the world, Torchilin said.
"On the other hand," she added, "there are specimen vendors like BioIVT, Asterand, and iSpecimen that acquire specimens for research from labs."
"[Our] value proposition is the ability to integrate the two… providing real world, data-enriched biospecimens for translational medicine, clinical feasibility, and observational studies," Torchilin said.
Biospecimens procured using the company's CDNR platform come with fully-consented patient health data from electronic health records, she noted, adding that researchers are also able to follow patients prospectively and collect specimens and clinical data over the course of a treatment or disease.
Additionally, she said, the platform provides researchers with historical data on what types of patients present at what clinical sites and how often, allowing them to better understand the subject pools available to them as they work to build their study cohorts.
"As a researcher you are usually in the lab, not in the hospital, so you're sometimes several steps separated from the patient," Torchilin said. "You're trying to identify the group of patients, the sub-population, that will give you the maximum chance to discover the biomarker or to validate the marker, or the new test that you're developing. You really need to understand what kind of patients there are, how you define that cohort, and what kind of specimens — from which kind of patients, actually — you will need to collect."
She gave the example of a diabetes researchers looking for early prognostic markers for the disease.
"Say you're trying to find patients who may have been diagnosed with diabetes less than a year ago, who haven't started metformin, and maybe have [hemoglobin A1c] below a certain level," she said. "Now, as a researcher in the lab, you're thinking, 'Okay, so I think now to validate my marker, I need 1,000 patients like this. And at this point, a huge inefficiency happens. Being a scientist myself, you really don't quite know, is that a realistic request? How many patients, realistically, can you find out there who are at that stage of the disease, with that level of HA1C, who have never taken metformin before?"
Novaseek launched in 2014 and began offering services through the CDNR platform last year. According to Torchilin, it has, to date, worked on more than two dozen projects across a variety of categories including oncology, immunoncology, neurology, and cardiovascular disease, and is currently working with four of the top ten global pharma firms as well as several biotech firms and academic teams.
The company works with hospitals who are either currently involved in recruitment for clinical studies or are interested in becoming involved, Torchilin said. Its platform can be integrated with the records of the healthcare systems it works with, which, she said, allows it to search for suitable patients in an automated fashion.
"Once we have a request from the researcher, we start immediately to look among the patients that we see coming through hospitals, and we very quickly, in an automated way, can identify matches," she said. "The alternative route would be to go and maybe partner with a doctor or with some type of a principal investigator, who will ask his nurse, or [other staff] to find the patients. But once you start asking hospital personnel to do some manual work, it becomes [time consuming] and often inefficient."
Currently, Novaseek operates in roughly a dozen hospitals across the US. The company also negotiates access to hospitals outside this group, for instance in cases where customers want to work with a specific key opinion leader or patient population, Torchilin said. She said it takes around two months for Novaseek to get the CDNR platform up and running in a new clinical site.
The company has ten employees and is currently funded by customer revenue and private investment, Torchilin said. She declined to say how much private investment it has taken on to date.