Use of iFOBT in Patients Presenting With Alarm Symptoms of Colorectal Cancer

Background: Colorectal cancer (CRC) is among the most frequent types of cancer in Denmark with 5,015 incident cases in 2015 (3,455 colon cancer and 1,560 rectal cancer), and an average annual number of deaths of 1900 persons in the period 2010-2014. Around 20% of patients are diagnosed in the most advanced stage of disease, contributing to high mortality, as an advanced stage of cancer is associated with a poor prognosis.

The Danish cancer patient pathway for CRC was implemented in 2008 with the aim of assuring fast and efficient diagnosing and treatment of cancer patients. In accordance with the cancer patient pathway, patients are referred by their general practitioner (GP) to fast-track colonoscopy when presenting with alarm symptoms such as rectal bleeding, change in bowel habits for more than four weeks, iron deficiency anaemia, and/or substantial general symptoms (e.g., weight loss or abdominal pain). Referral to colonoscopy is solely based on a clinical evaluation by the GP of the patient´s symptoms, still, the positive predictive value (PPV) of these criteria (i.e., the proportion of patients with alarm symptoms who have CRC) is only 3%-8%. At the same time, there is a considerable burden on endoscopy units to perform fast-track colonoscopies.

Immunochemical faecal occult blood tests (iFOBT) detect occult haemoglobin in the faeces, which may be caused by precancerous lesions or CRC. In the Danish national screening programme for CRC, the iFOBT employed is quantitative, and analysed by the OC-Sensor DIANA (Eiken Chemical Company, Ltd, Japan). Studies evaluating quantitative iFOBT to detect CRC in patients presenting with alarm symptoms, referred from primary care, report sensitivities ranging from 88% to 100%, specificities from 77% to 94%, PPV from 7% to 35%, and negative predictive values (NPV) from 98% to 100%. Overall, evidence of the accuracy of iFOBT in this population is limited and characterised by heterogeneous results, which is attributable to e.g., use of different cut-off values. Measuring faecal occult haemoglobin in patients with alarm symptoms may be a valuable tool for GPs deciding whether to refer the patient to fast-track colonoscopy. Not least, assessing the diagnostic accuracy of the iFOBT by type of alarm symptom and combinations of symptoms will provide insight into characteristics of patients that are likely to benefit from colonoscopy.

Aim: The aim of this study is to evaluate the diagnostic accuracy of the iFOBT among symptomatic patients who are referred to colonoscopy in the cancer patient pathway for CRC.

Materials and methods: The study will be conducted as a diagnostic accuracy study. Consecutive patients ≥40 years, referred to fast-track colonoscopy at the Regional Hospital of Horsens, will be invited by mail to collect a single faecal sample for haemoglobin measurement. The invitation will include a test kit with instructions, information about the study, and a consent form. Participants will be asked to collect the faecal sample before the laxative administration prior to colonoscopy, and immediately submit the sample for analysis at the Regional Hospital of Randers.

Faecal haemoglobin measurement will be conducted using the OC-Sensor DIANA analyser, and test results will be registered in the clinical laboratory information system. All participants will be examined by colonoscopy at the Regional Hospital of Horsens, and colonoscopy results will be registered in Danish national health registries. To obtain information on alarm symptoms, the patient´s GP will be asked to fill out a questionnaire concerning which type of alarm symptoms the patient presented with at the time of referral.

A positive iFOBT result (i.e., presence of occult blood in the faeces) will be defined as ≥35 µg haemoglobin/L. Patients will be considered to have been diagnosed with CRC by colonoscopy if they are registered by a relevant International Classification of Diseases, 10th edition (ICD-10) code for CRC, and a relevant procedure code for colonoscopy, and/or by a relevant Systematized Nomenclature of Medicine (SNOMED) code for CRC.

Using results of colonoscopy as the reference standard, the sensitivity, specificity, PPV, and NPV will be calculated as measures of the diagnostic accuracy of the iFOBT. The diagnostic accuracy of the iFOBT by sex, age, and type of alarm symptom will also be assessed.

Perspectives: This study will provide knowledge on the diagnostic value of the iFOBT among patients presenting with alarm symptoms of CRC in general practice. The study thus will contribute new and valuable data that may be used to evaluate the existing referral criteria for the cancer patient pathway.