A physician takes one small biopsy of a patient’s tumor, analyzes it in their own lab, and receives automatic therapeutic recommendations customized for the patient. This is not quite yet reality, but it’s the future that Navignostics is working towards.
Jana Fischer is the CEO and co-founder of Navignostics, a University of Zurich spin-off in the field of cancer diagnostics. Together with her three co-founders, Jana’s academic research focused on analyzing tumors with advanced imaging and computational approaches. Her team’s method turned out to be extremely good at distinguishing between patient groups with different molecular properties in retrospective cancer cohorts. Encouraged by these results, Jana and her colleagues decided to bring their work into a clinical setting.
The same in simpler terms: Jana was able to explain clearly why certain cancer patients had reacted differently to their treatments. Which was great! Yet merely understanding what had happened in the past wasn’t enough for the four co-founders: they wanted their work to help cancer patients receive the right treatment now and get better in the future.
In addition to helping clinicians diagnose and treat cancer, the Navignostics solution can also help drug developers create new therapies for cancer. When developers are able to identify the patient groups that are most likely to benefit from the new drug, the clinical trials become faster and more efficient.
Navignostics is currently backed by multiple investors, with Bruker in the lead. Today an eight-person team, the start-up is working on bringing a finalized product into the market. Jana says they expect to have the Swiss regulatory approval in 2025, and the US market with the FDA process will be next.
“More data, with better predictive power, from fewer samples, in a shorter time”
The research done by Jana and her colleagues at the University of Zurich shows that Navignostics is already able to make detailed analyses from very small amounts of tissue in a short amount of time, and they can make reliable predictions based on these analyses. Their multiplex solution images tissue sections with subcellular resolution while quantifying the proteins on the single-cell level; the company is also able to measure more than fifty protein markers simultaneously in a single tissue section. Until now, pathologists have only been able to measure one protein at a time and per tissue section, which takes a lot of time and a lot of tissue. Also, even though clinicians already have the ability to test hundreds of genomic markers from small samples, these are not always as accurate as protein markers when making predictions on therapy response; this is especially the case for immunotherapy responses. To summarize, the Navignostics solution promises more data with better predictive power from fewer samples in a shorter time.
Navignostics will begin its operations as a diagnostic service laboratory, while the ultimate goal of quick analysis and therapeutic recommendations done by clinicians in their own office is a guiding star for future development and collaborations. The company is already channeling locally acquired data through a cloud-based analysis pipeline, which means an important part of the process is already independent of a physical location.
“Healthcare will become much more data-driven in the next ten years”
One crucial piece of the diagnostic dream puzzle is access to good data. Jana is a data scientist by education (and heart!), and Navignostics has already worked with great amounts of retrospective cancer data from their clinical partnerships to build their analysis and prediction machine. However, to keep improving the algorithms and patient outcomes, Navignostics needs frequent access to new data, especially follow-up data that reports how patients have responded to given treatment.
In academia, there are quite a few collaboration projects that aim at sharing data, collecting data, and integrating different modes of data, Jana describes. However, on the commercial side it’s much harder to gain access to high-quality data sets. Academic research data is not always standardized enough to be suitable, and even if it was, commercial use is typically not allowed; furthermore, the different commercial players are well aware of the value of their data, Jana states. A company like Navignostics needs to have deep pockets or something to offer in return.
One way to gain clinical data is to buy tumor samples from biobanks, which typically costs a fortune. Instead, Navignostics has collected their data mostly from clinics through collaborations. This kind of data is typically not handed out or even sold to just anyone. You need to have a well-defined project with a good purpose as well as ethics approval, Jana elaborates. It also helps if you can bring them further in their own research.
Gaining access to data is not the only bottleneck; collecting data in the first place is also a challenge. The most valuable data sets include follow-up on patient responses, and this is something that needs to come from clinicians – who are notoriously busy in their daily work. There have been examples where data has been collected directly from volunteering patients, but such data doesn’t come in an easily usable, standardized format.
We need to find ways to incentivize data collection and to make it as easy as possible, Jana states. After all, like a true data scientist, Jana believes that healthcare is becoming significantly more data-driven. Ten years from now, Jana’s company will have led the way to other important companies that provide medical predictions based on molecular data. As a result, the path from examining a potential cancer patient to successful recovery will be shorter and easier.
This article is partially based on this podcast episode.
Comments