Startup Develops Potential Milestone in Early Cancer Detection
Quantgene Achieves Breakthrough in Liquid Biopsy Early Cancer Detection by Combining Machine Learning with Next-Generation PCR
The Quantgene researchers analyzed large sets of sequenced tumor data across thousands of patients to identify specific patterns of mutations to detect multiple types of cancer at early stages in regular blood samples.
Preliminary internal results showed a 90 percent sensitivity and 100 percent specificity across a first small batch of colorectal cancer patients, indicating strong performance of Quantgene's approach in early cancer diagnosis. Quantgene achieves a leap in detection accuracy by combining its data with PCR technology, which offers significantly higher cost-adjusted accuracy than sequencing platforms.
"Our intelligent data matrix enables us for the first time to use next-generation PCR for agnostic cancer detection," says Jo Bhakdi, CEO and founder of Quantgene. "By unlocking PCR as an option, we outperform sequencing-based approaches in cost-adjusted accuracy by over 20,000 times."
PCR has long been recognized as the most accurate and cost-effective genomic diagnostics method, but the technology has remained ineffective in early cancer detection due to its requirement to specifically define each target mutation. At over a million different mutations across the 10 deadliest cancer types, PCR can't be used for detecting cancer without a way to significantly reduce the number of target mutations required. As a consequence, less accurate and more complex sequencing approaches came to dominate early-stage cancer detection attempts.
"We love sequencing - it's the technology we rely on to generate our initial data," said Bhakdi. "But it is an imperfect tool to achieve the extreme accuracy required for early cancer detection. PCR is a better match and we have finally found the computational key to use it effectively."
Bhakdi estimates a market price of $500 to screen for multiple cancer types at early stages using Quantgene's data-driven PCR method, compared to over $5,000 for less accurate sequencing based approaches.
"The implementation of machine learning algorithms across petabytes of genomic cancer data was a challenging task but has now delivered a game-changing paradigm in early cancer detection," said Bhakdi. "We couldn't be more excited for the next 12 months and am now looking for the right partners to turn the platform into the first viable blood based cancer screening product."
For more information on Quantgene visit http://www.quantgene.com.