Point of care diagnosis of gastrointestinal disease using laser spectroscopy

Medical diagnostics is moving from laboratory to bedside. There is a strong trend for complex laboratory analyses to be supplemented, or even replaced, by tests that can be performed at the point-of-care (PoC) by personnel with little or no specialist training. An important feature of PoC tests is their short time-to-result. Provision of diagnostic information at or near real-time supports clinical decisionmaking by enabling rapid and targeted intervention, which improves patient outcome and promotes efficient use of limited healthcare resources. To this end, development of novel instrumentation capable of rapid and accurate measurement of chemical indicators of disease (biomarkers) is a strategic priority.

Clostridium difficile infection (CDI) is an example of an unmet clinical need for PoC diagnostics and is the main focus of this study. CDI is a hospital-acquired infection which produces catastrophic diarrhoea, prolongs hospital stays and can prove fatal in vulnerable individuals. It is highly contagious - current UK NHS intervention policy requires that patients with unexplained diarrhoea must be isolated and treated for the disease before a positive diagnosis is available.

Current tests for CDI use traditional laboratory "wet chemistry" enzymatic and nucleic acid assays, with limited diagnostic performance and a lead time measured in hours. Misdiagnosis can lead to patients being unnecessarily isolated from wards or treated unnecessarily with antibiotics, which contributes to the development of antimicrobial resistance.

Variation in the levels of volatile organic compounds (VOCs) emitted from a range of human samples (e.g. breath, blood, urine) are known to be associated with metabolic status and have been linked to particular diseases. The gastrointestinal tract offers a particularly rich source of information, since many disease states are associated with changes in the bacterial population of the gut (the microbiome) resulting in changes to the VOCs produced, which can be measured using optical spectroscopy.

Our vision is to develop a novel approach based on optical measurement of these volatile biomarkers in the gas phase. By measuring the level of specific biomarker chemicals produced by samples of patients' faeces, we aim to provide early warning of the development of gastric disease. Important benefits of this approach are:

- Samples of faeces are taken using standard clinical procedures, as is normal practice today when symptoms develop.
- Volatile markers may be measured using laser spectroscopy, with a short time-to-result (1-2 minutes). The measurement is highly selective to individual VOCs, which importantly will allow identification of biomarkers against a complex background matrix of over 300 species.
- The method requires minimal sample preparation, avoiding the use of reagents and making it suitable for point-of-care diagnosis.
- Because the measurement system is not in physical contact with the sample, there is no interference or fouling of the sensor.
- It is clinically non-invasive, so is suitable for repeated use in disease monitoring, unlike techniques such as colonoscopy and sigmoidoscopy which are widely used in chronic disease diagnosis but cannot be used on a daily basis.
- The method will allow active disease to be distinguished from mere carriage of Clostridium difficile (the latter being present in a significant percentage of the UK population without ill effect).

We will develop a flexible diagnostic platform targeted at diagnosis of CDI. Disturbances in the gut microbiome are also associated with a range of other gastrointestinal conditions including inflammatory bowel disease and colorectal cancer, and with other diseases such as diabetes. This technique therefore has wide potential application for medical diagnosis and monitoring of a range of diseases at point of care.

Enhanced prediction and diagnosis of diseases in real time and at the point-of-care, the theme of this proposal, is recognised by UK government funding agencies (e.g. InnovateUK, EPSRC, MRC) as a means of addressing national and global health challenges, with the development of sensors to detect and measure biomedical markers being a strategic priority. The instrument developed in this project would allow faster, non-invasive and reliable diagnosis of disease. It is recognised that such a development would impact on improved patient outcomes, which will have a positive impact on economic activity and productivity, and optimise the use of limited healthcare resources. For this project specifically, there would be improved resilience of the UK NHS against outbreaks of infectious disease. This would be mediated in the case of C diff by a change in NHS policy and procedure for dealing with patients suffering diarrhoea of unexplained origin.

For gut infections, earlier and better informed diagnosis could lead to better targeting of antibiotics. Early and precisely targeted therapy is known to benefit patient outcomes, and in the case of infection, a difference can be made within hours. Rapid diagnosis of infection also addresses the growing problem of antibiotic resistance, in that it may help avoid the use of antibiotics where they are not needed. Managing antibiotic resistance is a UK strategic priority and a major cross-sector theme of the UK Research Councils. Being non-invasive, rapid and with a low cost per test, our diagnostic approach is suitable for frequent and repeated use with individual patients, so that an ability to non-invasively monitor the progression of disease at the point of care may also lead to improved control of antibiotic therapy (e.g. personalised dose titration, change or cease).

Diagnosis of many gastrointestinal conditions is complicated by the presence of multiple species of bacteria within the gut microbiota of both healthy and unhealthy patients, thus diagnostic tests based on the presence of bacteria alone can be misleading and differential diagnosis between gastrointestinal diseases is an unmet challenge. In contrast, our technique measures bacterial activity as this is fundamentally linked to the rate of production of volatile bacterial metabolites. Therefore we believe its use will lead to more accurate and meaningful clinical diagnoses. Ultimately, our vision is to develop a multifunctional instrument for diagnosis of a range of gastrointestinal conditions including infection (C diff, C jejuni and E coli), the inflammatory bowel diseases (IBD - Crohn's and ulcerative colitis) and colorectal cancer (CRC). These diseases may all have a similar presentation including abdominal pain and diarrhoea, as may the distressing but pathologically benign condition of irritable bowel syndrome (IBS). The treatments and prognoses for these conditions are very different. The diagnosis and monitoring of CRC and IBD involves the use of expensive and invasive procedures, notably colonoscopy. A rapid method that would reduce the need for such procedures would be beneficial, especially if a diagnosis of cancer could be excluded at an early stage.

Key to impact of this technology is its suitability for use in the clinical environment and acceptance by clinicians and by patients. To realise the societal impacts of this sensing platform, the project ensures that the design, practical implementation and interface of such an instrument can be tailored to the requirements of a clinical environment through effective collaboration and engagement with clinicial staff. The clinical pull at the heart of the project ensures that there is a strong link between technology development and translation to the clinical environment, areas identified by EPSRC as being often disconnected [1].

[1] https://www.epsrc.ac.uk/research/ourportfolio/researchareas/clinicaltechnologiesexcludingimaging/