Predictive factors for identifying macrolide responder in treating chronic rhinosinusitis

Abstract

Background: A low-dose macrolide (LDM) has antineutrophilic activity, so they should not work for eosinophilic inflammation. Clinical predictors are required to select favorable patients for LDM therapy appropriately. This study aims to assess individual predictive factors and propose suitable multiple predictive factors for identifying a macrolide responder in treating CRS.

Methodology: Prospective cohort study was done in adult CRS patients. Clinical data collection, Lund-Mackay CT score, visual analog scale (VAS), and sino-nasal outcome test 22 (SNOT-22) were assessed. Patients received 150 mg of roxithromycin once daily plus saline irrigation for 12 weeks. VAS was evaluated at every visit. If the patients had total nasal symptoms VAS > 7 at any visit, they were defined as macrolides non-responders. Nine predictors for macrolide responders were assessed. At the end of treatment, the patients were defined as either macrolide responders or non-responders.

Results: 100 patients (mean–±SD age 47.35 ± 14.13 years, 45% male) with CRS were included. 29 patients met the criteria of macrolide responders. Only local total IgE < 4.76 kU/l (OR: 3.06, 95%CI: 1.16 - 8.06) and serum eosinophils < 3.7% (OR:2.45, 95%, CI 1.01 – 5.93) showed a statistically significant association with macrolide response. Moreover, in a multivariate regression model, local total IgE was the only variable that maintained an independent association with macrolide response (OR: 3.06, 95%CI: 1.16 - 8.06).

Conclusions: Low local total IgE (< 4.76 kU/l) from nasal secretion may be a suitable predictor for identifying macrolides responders in treating chronic rhinosinusitis with LDM.