Endoscopy 2019; 51(04): S113
DOI: 10.1055/s-0039-1681503
ESGE Days 2019 oral presentations
Saturday, April 6, 2019 11:00 – 13:00: Best abstract awards Congress Hall
Georg Thieme Verlag KG Stuttgart · New York

DETECTION OF BARRETT'S ESOPHAGUS THROUGH EXHALED BREATH USING A NON-INVASIVE SCREENING TOOL

Y Peters
1   Radboudumc, Department of Gastroenterology and Hepatology, Nijmegen, Netherlands
,
RWM Schrauwen
2   Bernhoven Hospital, Department of Gastroenterology and Hepatology, Uden, Netherlands
,
AC Tan
3   Canisius Wilhelmina Hospital, Department of Gastroenterology and Hepatology, Nijmegen, Netherlands
,
SK Bogers
2   Bernhoven Hospital, Department of Gastroenterology and Hepatology, Uden, Netherlands
,
B de Jong
1   Radboudumc, Department of Gastroenterology and Hepatology, Nijmegen, Netherlands
,
PD Siersema
1   Radboudumc, Department of Gastroenterology and Hepatology, Nijmegen, Netherlands
› Author Affiliations
Further Information

Publication History

Publication Date:
18 March 2019 (online)

Also available at
 

Aims:

Timely detection of esophageal adenocarcinoma and its precursor Barrett's esophagus (BE) may decrease both cancer mortality and incidence. Currently, an accurate, minimally-invasive screening method for BE for widespread use not available. Our objective was to establish the accuracy with which breath analysis could discriminate patients with BE from controls without BE.

Methods:

In this multicenter cross-sectional proof-of-principle study, patients undergoing a clinically indicated upper endoscopy were invited to provide a 5-minute breath sample prior to scheduled endoscopy. Patients were allocated in three subgroups: BE (defined as ≥1 cm of columnar mucosa with histopathologic confirmation of intestinal metaplasia), gastroesophageal reflux disease (GERD) (defined as GerdQ-score ≥8 or presence of reflux esophagitis), and controls without BE or GERD.

The Aeonose is an olfactory system that analyses volatile organic compounds (VOC). Three metal-oxide sensors interact with VOCs in breath samples to create a digital breath print specific to the VOCs. Data was analyzed by an artificial neural network to identify data classifiers to extract breath-print differences between of patients with BE and GERD or controls. Optimal models were cross-validated using a leave-10%-out approach. Main outcomes were sensitivity and specificity for detecting BE compared with upper endoscopy as reference standard.

Results:

Breath samples were obtained from 153 individuals (60 BE, 53 GERD, 40 controls). Recruitment rates were 97%. Diagnostic accuracy was high for discrimination of BE from GERD and controls (area under the curve [AUC] 0.91, sensitivity 90%[95% CI:79%-96%], specificity 81%[95% CI:71%-88%]). Similarly, breath prints of BE patients could be differentiated from GERD patients (AUC 0.85, sensitivity 72%[95% CI:58%-82%], specificity 89%[95% CI:76%-95%]).

Conclusions:

This portable electronic nose is able to detect the presence or absence of BE in patients with and without GERD. Given the high tolerability, high acceptability and low costs, breath testing may be a promising approach to be used for non-invasive BE screening in a primary care setting.