DISCUSSION
In this retrospective study we have shown that a significant inflammatory reaction at airway level is present in children with airway malacia due to mediastinal vascular anomalies and recurrent respiratory symptoms. A significant neutrophilic alveolitis was detectable by bronchoalveolar lavage thought the procedure was performed when children were in well-being conditions, at least four months after the last respiratory infection. Additionally, microbiological analysis of BAL fluid demonstrated bacterial growth in more than one third of them, but with a pathogen load > 105 CFU/mL only in a small proportion. Haemophilus influenzae was the pathogen most frequently isolated. In a total of 512 bronchoscopies, airway malacia was diagnosed in 160 children (94 males) at a median age of 4.0 years (range, 0 to 17 years). Airway malacia is not a rare condition in the general paediatric population, as reported by Boogaard and co-workers [12]. Out of total of 512 bronchoscopies performed in children with a median age of 4.0 years, 160 cases of airway malacia were detected and classified as primary in 136 cases and secondary in 24 cases [12]. Airway malacia, inducing a dynamic collapse of the airway lumen, reduces cough effectiveness and interferes with mucociliary clearance, favoring airway plugging by bronchial secretions and possibly bronchiectasis formation [13]. Moreover, in a study in which bronchoscopy was performed to exclude other wheezing diseases in children with “severe chronic asthma” (likely including a group with airway malacia), Fayon M and co-workers, showed lower airway neutrophilia and infection (>104 CFU/mL) with Haemophilus influenzae detectable in 39.5% of BAL cultures [14]. Chronic wet cough, evidence of bronchitis at bronchoscopy and airway neutrophilic inflammation are the hallmark of protracted bacterial bronchitis (PBB) [15]. In the first study defining this disorder, Chang AB and co-workers reported that trachea-bronchomalacia was found in 33% of children with chronic cough and BAL cultures in PBB group were positive in 47% of samples for Haemophilus influenzae , in 35% for Streptococcus pneumoniae and in 26% forMoraxella catarrhalis [16]. More than one organism grew in a significant number of PBB patients and 48% of the children had a positive clinical response to 14 days of amoxicillin clavulanate (22.5 mg/kg/dose, b.i.d.) [16,17]. In a follow-up study, Chang AB group evaluated the 2-year outcomes of 161 PBB children [18]. They found that a great proportion of them (43.5%) had more than three episodes of bacterial bronchitis per year, and 8.1% were diagnosed with bronchiectasis (vs the 13% of our patient group) [18].Haemophilus influenzae infection was demonstrated in BAL fluid of 85% of children with bronchiectasis and of 49% of children without bronchiectasis. Recurrent PBB episodes and high Haemophilus influenzae titers in BAL fluid were major risk factors for bronchiectasis at 2 years follow-up, suggesting a role of this specific infection in bronchiectasis pathogenesis [18]. In our patient group, 72.7% of the isolates were Haemophilus influenzae positive but only one child with bronchiectasis had a positive culture forHaemophilus influenzae. Through its ability to form biofilms,Haemophilus influenzae is a chief bacterial pathogen associated with chronic respiratory disorders [19] and subinhibitory concentrations of beta-lactam antibiotics may promote its transformation into the biofilm phenotype, favoring antibiotic resistance and increased susceptibility to reinfection [20]. In randomised controlled trial in children with chronic wet cough it was reported that many of the children whose symptom was not cured by 2 weeks of amoxicillin clavulanate had underlying tracheo-bronchomalacia and needed a longer treatment duration before their cough disappeared [15,21]. Therefore, in view of high frequency of positive isolates forHaemophilus influenzae but also of the general low bacterial load and the significat coesistence of bronchiectasis, caution should be used in inappropriate antibiotic prescription in patients with secondary airway malacia and recurrent respiratory infections. This approach may be chiefly important in children with bronchiectasis, in which chest physiotherapy can be of great benefit [22]. When possible, antibiotics should be chosen on the basis of sputum or cough-swab culture and sensitivity, adherence to treatment should be monitored and the opportunity to switch from twice to thrice daily administration or to increase the dosage should be taken in consideration in specific cases to lengthen the time at higher minimum inhibitory concentration levels [15].