Methods
The present study was conducted in the Pediatric Pulmonology Unit of the Attikon University Hospital in Athens, which is one of the main tertiary referral centres for pediatric pulmonary disorders in Greece. We retrospectively analysed data of 54 children with CSLD/bronchiectasis, without an identifiable cause, who had been diagnosed and followed in our department. Investigations in all patients included sweat test and/or CF gene mutation analysis and measurement of serum immunoglobulins. Nasal nitric oxide test and high-speed video microscopy analysis were performed only in patients who fit the clinical phenotype of primary ciliary dyskinesia. All had Pa isolated at least once in their sputum, or cough/throat swabs, or bronchoalveolar lavage (BAL) cultures.
Data regarding the time of chronic wet cough commencement, the time of referral to our department (which coincided with the time of diagnosis of CSLD/bronchiectasis), and the duration of follow up were retrieved from the medical records. Microbiological data and the percent-predicted FEV1 (ppFEV1) values (calculated with the NHANES III reference equations) of all patients with valid spirometries were also retrieved. For the purposes of this study, disease remission was defined as the absence of daily wet cough for four months along with three negative cultures taken during the last three consecutive follow-up visits. All patients received ciprofloxacin per os for at least 3 weeks after the isolation of Pa, and started (if they had not already been on) daily chest physiotherapy. Some of them also received long term treatment with twice-daily inhaled colistin and/or thrice-weekly azithromycin and/or nebulized hypertonic saline. Three-monthly follow-up visits were arranged for all children. In all visits, sputum or cough/throat swabs cultures were obtained and spirometry was performed on all patients ≥5 years old who were able to cooperate.
Radiological records were retrospectively reviewed for the presence of bronchiectasis in chest high-resolution computed tomography (HRCT) scans which were obtained before or at about the same time with the isolation of Pa. The evaluation of the HRCT scans was performed by the pediatric radiologists of our hospital (E.A., S.P., and A.M.) who were aware of the patients’ clinical history. The criteria for the diagnosis of bronchiectasis on HRCT were dilatation of bronchi with a broncho-arterial ratio > 0.8010,11; parallel bronchial walls in a longitudinal section (tram sign); visualization of bronchi within 1 cm of pleura. The modified Bhalla score was used to quantify the severity of bronchiectasis12.
Statistical analysis : Variables are described as medians with interquartile range (IQR). Univariate comparisons were performed with Fisher’s exact test and Wilcoxon rank-sum test. Graphical presentation of the data was performed with Kaplan–Meier survival curves. Multivariate analysis was performed with a Cox proportional hazard model, with age, presence of bronchiectasis, use of inhaled colistin, azithromycin, and inhaled hypertonic saline, as the covariates. The proportionality assumption was checked with the Schoenfeld test.