Predictive value of clinical features for anti-reflux therapy response in suspected gastroesophageal reflux-induced chronic cough.

ABSTRACT

Background: Not all gastroesophageal reflux-induced cough (GERC) patients respond to anti-reflux treatment. It is not certain whether reflux-related symptoms or other clinical characteristics could indicate a successful response to anti-reflux treatment. In this study, we aimed to investigate the relationship between clinical features and anti-reflux response. Methods: We retrospectively analyzed the clinical characteristics of suspected GERC who had reflux-related symptoms or reflux evidence based on abnormal 24-hour esophageal pH value monitoring, or who had no evidence of other common causes of chronic cough in our chronic cough database with a standard case report form. All patients experienced anti-reflux treatment with proton pump inhibitors (PPIs) plus prokinetic agents for at least 2 weeks and were divided into responders and non-responders based on the treatment response. Results: Among 241 patients with suspected GERC, 146 (60.6%) showed a successful response. There was no significant difference in regard to the proportion of reflux-related symptoms, and results of 24-hour esophageal pH value monitoring between responders and non-responders. Compared with non-responders, responders had higher proportions of nasal itching (21.2% vs. 8.4%; P=0.014), tickle in the throat (51.4% vs. 35.8%; P=0.025) and lower proportion of pharyngeal foreign body sensation (32.9% vs. 54.7%; P=0.001). Multivariate analysis showed that nasal itching [hazard ratio (HR) 1.593, 95% confidence interval (CI) 1.025-2.476, P=0.039], tickle in the throat (HR 1.605, 95% CI 1.152-2.238, P=0.005), pharyngeal foreign body sensation (HR 0.499, 95% CI 0.346-0.720, P<0.001) and sensitivity to at least one cough trigger (HR 0.480, 95% CI 0.237-0.973, P=0.042) were associated with the therapeutic response. Conclusions: Over half of suspected GERC patients benefited from anti-reflux therapy. A few clinical features rather than reflux-related symptoms might indicate a response to anti-reflux treatment. Further study is needed for the predictive value.