Pleural effusion, where fluid builds up in the cavity around the lungs, can develop in a range of conditions, from mild heart problems to severe infections or cancer.
Malignant pleural effusions, where cancerous tissue is present in the cavity, affect around 50,000 people in the UK every year.
So far, no biomarker can accurately tell whether the fluid build-up is caused by cancer, which is the cause that clinicians are most concerned with identifying. Invasive and costly biopsies are the only way to find out.
The best existing method for checking whether a biopsy is needed is looking at protein or cells from the pleural area by microscope. This fluid is collected via a minimally invasive ‘tap’ when patients arrive at hospital. However, this is only accurate in about 60% of cases, leading to delays in diagnosis and patients undergoing treatment they might not need.
DNA methylation (DNAm) is a chemical process that regulates how genes are expressed. It is often disrupted when cancer is developing, where it is a hallmark of tumour development and cancer progression.
DNAm-based methods for detecting cancer are starting to see some success. DNAm measured in pleural fluid is likely to be particularly useful in indicating pleural malignancy early.
We want to improve accuracy and time to cancer diagnosis by:
Developing a pleural DNAm signature to identify cancerous pleural effusion when patients come to hospital
Evaluating whether this DNAm signature improves on existing methods of detecting pleural cancer
We will use pleural fluid samples from both malignant and non-cancerous cases, to find DNA methylation across the whole genome. We’ll use machine learning to identify the DNAm biomarker signatures that best differentiate between malignant and non-cancerous cases. We will benchmark against existing processes for diagnosis.
How our results will be used
Our work could help to develop a point-of-care test to rapidly and cost-effectively tell whether a pleural effusion is cancerous or not.