We recently sat down with Ashkan Afshin, Founder of Novelna, a proteomics-based liquid biopsy company. He provides his take on leveraging proteomics for early cancer detection, along with his perspectives of the early cancer detection and precision medicine landscapes.  

‍

Key Takeaways:

• Proteomics technologies have advanced significantly in recent years, allowing for greater precision in protein measurement and enabling new cancer early detection approaches

• Although proteomics tests may face logistical challenges as they move into clinical settings, these challenges can largely be addressed by selecting stable proteins and leveraging widely available platforms

• In the long-term, Dr. Afshin expects single-cancer and pan-cancer tests to be complementary, with pan-cancer testing potentially becoming the preferred screening method, and single-cancer tests serving as follow-up diagnostic tests

• Novelna’s approach is focused on building a high sensitivity test (up to 5x Galleri) at a low cost (below $50 and even below $10 in the longer term, cost not final price)

‍

To begin, can you give us an overview of your background so that we can understand how it shaped your founding of Novelna?

Absolutely. My background is in medicine, and I have always been passionate about keeping people healthy. My focus is on extending healthy life expectancy at the population level, not just the individual level, so after medical school, I transitioned from practicing medicine to public health to have a bigger impact. There, I published several high-impact articles demonstrating how cardiovascular disease, diabetes, and cancer can be prevented through lifestyle modifications. However, a key finding was that while lifestyle is important for cancer prevention, it is not always sufficient. Some cancers cannot be fully prevented by lifestyle changes alone.

Early detection is crucial for identifying cancers, particularly solid tumors. Cancer screenings face issues such as compliance, cost, and accuracy. Our current cancer detection methods are mostly focused on symptomatic patients, and there is a need for improved diagnostics for the general population. I focused my academic work on addressing this problem and advancing the next generation of cancer diagnostics.

‍

It’s great to hear that you saw an opportunity to enter this space with something new. Could elaborate more on the unmet needs that you saw, and the unmet needs that you and other companies are still trying to address in the early cancer detection landscape?

To summarize the issues with cancer screening: they are incomplete for patients, inaccurate for providers, and expensive for payers. In other words, we are dealing with three P's: payers, providers, and patients. To create effective innovations in healthcare, we need to address the needs of these three groups.  

Before entering the biotech space, we conducted a systematic review on innovation in healthcare and published it. The results showed that new innovations often lead to increased complexity and higher costs, creating tension with payers and posing scalability challenges. Unlike other industries where automation and technology improvements reduce costs, technological advancements in healthcare tend to increase costs.

To turn innovations into effective solutions in healthcare, we need to break this cycle. We must innovate in ways that not only improve outcomes but also reduce costs. From an economic perspective, we aim for innovations that are not just cost-effective but also cost-saving. While current reimbursement strategies approve cost-effective solutions, we aspire to develop innovations that are both life-saving and cost-saving.

‍

That’s a great point, and that brings us to our next question: how do you build a highly accurate test at a low cost? Specifically, what is the role of proteomics in improving accuracy and cost efficiencies?

There are differences between what we learn in academic settings and what we observe in the real world. From an epidemiological perspective, the most important characteristic for a screening test is sensitivity. I want the lowest possible false negative rate because missing a patient is a significant concern.

However, from a health policy perspective, cost is a major issue, and false positives present a different problem. There's always a trade-off between false negatives and false positives. We aim to reduce both, but false positives lead to undo harm and increased downstream costs for the patient.

One of the problems with existing genomic tests is that they are specific but not sensitive. For instance, early studies of the Galleri test from GRAIL showed a sensitivity of only 18% for stage one cancers, though it performs better for later stages. This test was marketed as an early detection tool, but by the time DNA is detectable in the blood, the cancer is often already advanced.

Our approach shifted to focusing on proteins, which are closer to the biology of the disease. We hypothesized that proteins could provide a more sensitive test. Initially, we worked with high-concentration proteins using mass spectrometry, which provided good signals but were ultimately not the best. Over time, we moved to low-abundance proteins, which showed greater potential for early-stage detection. We found that for stage one cancers, our approach achieved almost five times higher sensitivity compared to tests like Galleri.

By focusing on proteomics and low-abundance proteins, we can achieve higher sensitivity for early detection. Although we haven't published the results yet, our research shows that with the right multiplexing of proteins—using a panel of 10 to 20—we can mitigate the risk of false positives. This approach allows us to achieve high sensitivity while maintaining acceptable specificity in the test’s performance.  

‍

What are the challenges that you anticipate with launching a proteomics assay in the clinic? How would you plan to address those challenges?

There are challenges and opportunities in proteomics compared to genomics. In genomics, the tools and platforms, like those from Illumina, are widely available and well-established, making it easier to use and scale. For proteomics, the choice of platform can present both challenges and opportunities.

For example, using established platforms like Luminex or flow cytometry is relatively straightforward since they are widely used. However, developing tests on newly developed platforms can be more challenging and may complicate scaling. To mitigate this risk, we are considering ensuring our final product can be used on established platforms.

Another challenge is protein stability. Proteins are more prone to degradation compared to DNA, so processing samples within 15 to 30 minutes after collection is crucial. To address this, we focus on using proteins in our panels that are more stable and can withstand room temperature for longer periods. New technologies or methods, such as Streck tube, can preserve proteins from degradation, which helps. Single protein measurements, like PSA, have been in use for a long time, so this isn’t a major hurdle for us.

Additionally, there are challenges related to patients with comorbidities, such as autoimmune diseases. If our signature involves immune responses to cancer, it may affect the signal. We have developed methods to mitigate this risk and ensure the test performs well even in individuals with comorbidities. Like any diagnostic test, our initial use case may focus on the general population, and results should be interpreted cautiously for specific populations with comorbidities.

‍

What is Novelna’s approach to making tests cost-efficient and accessible, and what impact do you think a low-cost $100 pan-cancer test could have on the early detection market?

Compared to genomics, proteomics platforms are less expensive. For example, even though Olink is still costly, it’s about half the price of genomic tests. We are exploring ways to reduce costs further because ideally, we want our test to be as affordable and easy to use as a pregnancy test. While achieving this level of affordability right away might not be possible, we aim to reach this goal in five to six years. This will allow us to scale and address logistical issues.

For our initial product, we may need to use lower-cost platforms that we have identified, which could bring development costs below $50 and eventually under $30. Over time, with higher throughput and advancements, we expect to bring the cost below $10.

Regarding impact, early cancer detection is our primary goal. Detecting cancer at an early stage allows for its surgical removal and significantly improves patient quality of life while reducing overall care costs. This approach benefits patients by providing better outcomes, physicians by enhancing performance, and payers by lowering costs. For example, although CT scans are recommended for lung cancer screening, only 5% of eligible people get screened. By lowering the cost and making the test less invasive, we can improve test coverage and scalability.

‍

‍What do you see as the interplay between pan-cancer versus single cancer approaches for early cancer detection tests? Do you see one approach winning out over the other, or are they complementary in some way?

In the short term, I expect to see increased use of cancer-specific tests. These tests are easier to study clinically, obtain regulatory approval for, and secure reimbursement. Patient navigation is also clearer, making adoption in the healthcare system more straightforward. As evidence builds and the challenges of pan-cancer tests are addressed, pan-cancer tests will gain more momentum.

Looking ahead five to ten years, in an ideal world, cancer screening would include pan-cancer tests as a first-line diagnostic tool during routine checkups. These tests would be highly sensitive and cost-effective, helping to identify individuals at risk of cancer. For those who test positive, further diagnostic tests would be used to determine the cancer's origin. In this scenario, cancer-specific tests would serve as diagnostic or secondary screening tools rather than primary screenings. This approach would allow for broad initial screening with one test, followed by more detailed secondary screening or diagnostics for those who need it.

‍

We’ve seen some highs and lows in the blood-based early cancer detection space, for example with Guardant Shield’s FDA approval for a blood-based test, but also Grail’s NHS trial controversy and recent layoffs. Do you have any thoughts on how these affect the market, and what are some of the lessons learned in a landscape that's so fast moving?

There’s a lot to learn from these companies. They have impressive visions, and while our technologies differ, we all aim to detect diseases early with high quality and accuracy. For example, companies like Guardant Health have demonstrated that FDA approval can create significant momentum in the market. Their approach has shed light on study designs for FDA approval and has made subsequent studies easier, especially for blood-based tests. This paves the way for developing the next generation of tests and addresses many challenges.

On the other hand, companies like Grail have a commendable vision with their multi-cancer early detection. However, there are issues with their technology, study design, and target population. These are important lessons for optimizing our approach. For instance, focusing on proteomics can help address sensitivity issues that other technologies face.

Overall, we can learn valuable lessons from both sides. This helps in raising awareness among stakeholders, creating market interest, and accelerating our progress. We have even brought some experts from these companies on board to help us advance more quickly.

‍

At a high-level, what do you think are relevant trends within the early cancer detection space? And are there any aspects of the broader precision medicine landscape that you’d like to comment on?

As we mentioned at the beginning of our discussion, we are in an era of significant technological advancement. By converging these technologies, there are numerous opportunities for innovation. I’m particularly impressed by the progress in proteomics. The depth and precision of protein measurement have advanced remarkably.

When I started this journey, most of the proteins we worked with were measured in the micro to milligram range. Now, we are measuring in the picogram range, which represents a whole new level of precision. This advancement not only enhances our understanding of biology but also opens up new possibilities for diagnostic development.

The progress in both technology and biology enables us to develop next-generation diagnostics that can detect diseases before symptoms appear. For example, pregnancy tests can detect pregnancy before any symptoms occur, and we aim to achieve a similar capability for cancer detection. Although we haven’t fully reached this point yet, I am very optimistic about our ability to detect diseases at their earliest stages, long before clinical symptoms arise. The field is moving in this direction, which will be crucial for extending healthy life expectancy.

‍

You're focused on proteomics, but do you still see value in the various omics interacting with or complementing each other?

Each of these fields, from genomics to protein modifications like phosphorylation and metabolomics, provides specific insights into the underlying disease pathology and is optimal for addressing different questions. Many companies are moving toward multi-omics approaches for diagnostics, which is interesting but presents challenges in terms of cost, workflow complexity, and scaling. For instance, scaling up a multi-omics test in regions like Africa or parts of Asia over the next five years seems challenging.

However, omics technologies have clear applications in therapy selection and patient stratification. Multi-omics approaches can be valuable for therapy selection and assessing treatment response. While genomics and other omics can aid in diagnosis and treatment decisions, proteomics offers a promising position for diagnostics due to its balance of sensitivity, cost, and practicality. Combining genomics with proteomics can be useful for confirming diagnoses and tailoring treatments.

Thank you for this conversation; we appreciate your optimism in how these different aspects of precision medicine, particularly proteomics, will accelerate and further shape how we detect disease.

Thank you.