Developing a prototype device for point-of-care non-alcoholic fatty liver disease diagnostics

Non-alcoholic fatty liver disease (NAFLD) is caused by fat accumulation in the liver and affects 20-30% of adults globally, ˜20% of whom progress to a more serious condition called non-alcoholic steatohepatitis (NASH). NASH-related liver damage is reversible, but only if diagnosed early. However, NASH diagnosis rates are extremely low, and for most patients diagnosis occurs too late for intervention. New tests are needed that allow more patients to be tested for NASH, resulting in earlier diagnosis and treatment. Given that NASH is common, these tests ideally need to be provided by a GP and, therefore, must be low-cost and require little equipment. Existing tests and those in development do not meet these requirements, and there is a need for innovative approaches that fill this gap.

This proposal addresses this need using Rosa Biotech's sensing platform, Pandra, which can detect NASH in patient blood samples with high accuracy. Pandra detects disease using the principle of olfaction (the sense of smell). The olfactory system uses arrays of protein receptors that bind to odorants as they enter the nose. The brain recognises patterns of receptor stimulation and interprets the pattern as a smell. When applied to biosensing, this principle is called differential sensing. It allows small changes in a sample's total contents to be analysed, without relying on pre-identified markers of disease.

Rather than olfactory receptors, which are challenging to work with, Rosa's sensing technology, Pandra, uses arrays of designed proteins that bind coloured dyes. Exposing the array to samples containing fats displaces the dyes from the proteins to different extents, resulting in colour changes across the array. This produces a unique 'fingerprint' for the sample. An artificial intelligence (AI) algorithm is trained to recognise differences between fingerprints associated with disease in known samples. The algorithm is then used to detect disease in unknown samples. Pandra is cheap to produce, usable by non-experts, and sensitive to small changes in fat composition. This makes it well-suited to NASH diagnostics.

This project aims to produce an early prototype Pandra sensor that could be deployed at the point-of-care in a GPs surgery. This would provide clinicians with an invaluable new tool for initially assessing high-risk patients for NASH and deciding the best referral route. Ultimately, it will result in a new test that will enable more patients to be tested for NASH, reducing the number progressing to more severe liver disease, cancer, and death.