The aim of this study was to determine whether the prevalence of Human PapillomaVirus (HPV) was increased in patients with Barrett esophagus compared with controls in a prospective study. Secondary objective was to identify, if present, the type of Human PapillomaVirus (HPV) associated in Barrett esophagus.

Barrett's esophagus (BE) is defined by the replacement of normal stratified squamous epithelium in the distal third of esophagus by specialized intestinal metaplasia. It is related to gastro-esophageal reflux disease. Diagnosis is suspected during endoscopy and confirmed on biopsy. The classification of CM from Prague is validated to describe BE during endoscopy. The main complication of the BE is adenocarcinoma, according to metaplasia-dysplasia-cancer sequence. The incidence of esophageal adenocarcinoma is variable, ranging from 0.4 to 3 %. BE is found in 100% esophageal adenocarcinoma and in 42% junction adenocarcinoma. Among the unknown risk factors involved in the onset of dysplasia, viruses can't be excluded.

It is well established that Human Papillomavirus (HPV) is strongly associated with squamous cell dysplasia of female uterine cervix and its progression to cervical carcinoma. HPV is also implicated in others invasive carcinomas including uterine cervix, vulvar, vaginal, anal, penile, head and neck squamous cell carcinoma. Several studies showed that HPV could be associated in head and neck cancers and that tumor HPV status in patients with oropharyngeal squamous cell carcinomas was an independent prognostic factor for survival. The association between HPV and esophageal squamous cell carcinomas is still controversed with epidemiological studies reporting prevalence of mucosal HPV DNA ranging from 0 to 70%. Studies that have investigated HPV and adenocarcinoma of esophagus or Barrett's esophagus (BE) are scarce and data are not clear. A recent prospective study showed that HPV was strongly associated with Barrett'dysplasia and esophageal adenocarcinoma.

The aim of this study was to determine whether the prevalence of HPV was increased in patients with BE compared with controls in a prospective study. Secondary objective was to identify, if present, the type of HPV associated in BE.

Inclusion Criteria:

• written consent obtained before gastroscopy,
• patients aged more than 18 years,
• normal coagulation parameters Inclusion criteria for group A "control" and B "BE" were as follows, respectively: patients undergoing gastroscopy for non esophageal reason and presented no clinical, endoscopic or histopathological evidence of GERD or complication and patients undergoing gastroscopy for BE surveillance or suspected BE during gastroscopy.

Exclusion Criteria:

• written consent not obtained before gastroscopy,
• inability to give informed consent,
• pregnant or nursing women,
• major person protected by french law,
• person who is not affiliated to a social security regime,
• patient who is in a exclusion period for another clinical study,
• curative anticoagulation treatment,
• hemostatic disturbances.