Article Text


Oesophageal II
PWE-008 Do statins prevent the histological subtypes of oesophageal cancer? Prospective data from the UK general practice research database (GPRD)
  1. H Y Bhutta1,
  2. L Alexandre2,3,
  3. A Clark3,
  4. S Holt4,
  5. M Lewis1,3,
  6. A Hart2,3
  1. 1Department of General Surgery, Norfolk and Norwich Hospital, Norwich, UK
  2. 2Department of Gastroenterology, Norfolk and Norwich Hospital, Norwich, UK
  3. 3Norwich Medical School, University of East Anglia, Norwich, UK
  4. 4Roundwell Medical Centre, Norwich, UK


Introduction The incidence of oesophageal adenocarcinoma (OAC) has risen dramatically in the Western world and is associated with a poor prognosis. Statins show anti-cancer properties in experimental work with OAC cell lines for example reduced cell proliferation, increased apoptosis. This study aimed to investigate if statins are negatively associated with the development of two different histological subtypes of oesophageal cancer, OAC and oesophageal squamous cell cancer (OSCC), in a prospective cohort study.

Methods The cohort was over 4 million people in the General Practice Research Database (GPRD), a UK database of 488 nationwide general practices. Information is recorded on medication use prior to development of other illnesses, including cancers. Statin use was defined as a prescription for a minimum of 10 months preceding diagnosis of oesophageal cancer. Approximately half the GP practices in the GPRD are linked to the NHS cancer registry, allowing identification and sub-classification of histologically confirmed cases of OAC and OSCC. Each case was matched with four controls and conditional logistic regression estimated the OR plus 95% CIs for the development of each type of cancer, adjusted for diabetes, BMI, smoking, aspirin, PPIs and drugs that relax the lower oesophageal sphincter.

Results 581 histologically confirmed cases of OAC (77.8% men, mean age 70.7 years, SD=11.4) and 332 cases of OSCC (38.6% men, 71.9 years, SD=12.3) were identified between 2000 and 2008. The median length of statin use in cases and controls prior to diagnosis or index date was 3.9 years (IQR 2.3–6.1 years). Analysis by histological subtype showed an inverse association for OAC (OR 0.61, 95% CI 0.39 to 0.94) and for OSCC (OR 0.41, 95% CI 0.21 to 0.80). Categorisation into lipophilic and hydrophilic statins demonstrated significant inverse associations for only lipophilic statins for both OAC (OR 0.60, 95% CI 0.38 to 0.95) and OSCC (OR 0.46, 95% CI 0.23 to 0.91). There was a dose-response effect for statins and both OAC and OSCC (p<0.05 for linear trend) comparing high and low-dose statin use.

Conclusion The data supports lipophilic statins having a protective effect against the development of both OAC and OSCC, with evidence of a dose-response trend. Confirmation in other populations is required. Detailed statin use, according to class and dose, should be measured in future aetiological studies of both OAC and OSCC. The information supports assessment of these drugs in chemoprevention of OAC in the general population and those with Barrett's oesophagus in clinical trials.

Competing interests None declared.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.