Determining which biomarkers are ready for evaluation in primary care for use in early detection and diagnosis of gastro-intestinal cancers: a systematic review

Start Date Oct 2019

Code C/M6-C

Status Ongoing

Project Lead
Dr Natalia Calanzani and Dr Claudia Snudden
Senior Lead
Prof Jon Emery, Ms Paige Druce, Dr Kristi Milley, Dr Javiera Martinez Gutierrez, Jasmeen Oberoi (all Melbourne), Dr Garth Funston, Dawnya Behiyat, Rachel Boscott, Dr Smiji Saji (all Cambridge), Dr Mike Messenger (Leeds)


General practitioners need to be able to differentiate between potentially life-threatening conditions such as gastro-intestinal (GI) cancers from more common, less serious GI diseases. While GI symptoms are common in primary care, a cancer diagnosis is much less common. Furthermore, initial cancer symptoms are often non-specific, and most patients presenting with them will not have cancer. There is the constant need to balance the risk of missing a cancer against unnecessarily referring patients to undertake investigations that can be invasive and result in complications. Biomarkers are diagnostic tools that can be used alongside other tests and/or decision support tools in order to improve early cancer detection.

Substantial investment has been made into the development of biomarkers to aid cancer detection, but research is often carried out in specialist settings. Very few biomarkers have been evaluated, translated, or implemented in primary care. This translation is important so biomarkers are appropriate for a low prevalence population. Therefore, we are investigating which novel biomarkers are ready for evaluation in primary care for use in the early detection and diagnosis of GI cancers.

Aims & objectives

This review aims to determine whether there are any single or groups/panels of novel biomarkers that have the potential for use in primary care to aid early detection of GI cancers. We are interested in oesophageal, gastric, pancreatic, gallbladder, colon, rectal and anal cancers, and are focusing on sample sources which are feasible to be implemented in primary care: blood (serum and plasma), urine, faeces, saliva or breath.

We aim to describe the identified biomarkers either singly or in a panel, the characteristics of the included populations (both cases and controls), and the available measures of diagnostic accuracy (sensitivity, specificity, positive predictive value, negative predictive value and area under the curve).

Methodology & activity

We are following the PRISMA statement for reporting systematic reviews. Using a pre-defined search strategy, we have searched MEDLINE, Embase, Emcare, and Web of Science for studies published in peer-reviewed journals. We have also searched reference lists of included studies.

Title and abstract screening were independently completed by at least two reviewers, followed by independent screening (also two or more reviewers) of full-text articles. All disagreements were solved by consensus. Data extraction was piloted to ensure consistency and is being carried out by trained reviewers. Wide heterogeneity in results is expected; narrative synthesis is being used to report results (in text, diagrams and summary tables).

Outputs & impact

This review seeks to identify much needed evidence on biomarkers that can be potentially used in primary care to aid early detection of GI cancers. Some of these biomarkers are also used for prognosis and surveillance and to investigate patients in secondary care (where cancer prevalence is much higher).

We also expect that this review will be a stepping stone to other studies aiming to assess the feasibility of implementing biomarkers in primary care, considering issues such as acceptability to patients and healthcare professionals, safety and cost-effectiveness. Review results will be particularly relevant to countries where primary care can play a prominent role in cancer detection.

Next steps

We are planning to report the review results in two peer-reviewed publications: one focusing on upper GI and the second on lower GI cancers.

Back to Research