We also recorded the number of patients who quit itraconazole therapy secondary to adverse reactions. The sample size for the study was calculated (StatsDirect 2.7.2, www.statsdirect.com) assuming a 60% improvement (and 40% worsening) in the itraconazole group and 10% improvement (and 90% worsening) in the control group. With this calculation, 14 subjects were required in each group to detect these differences [confidence level (1 − α) of 95%, power level (1 − β) of 80%]. Data are presented as median

(interquartile range) or number (percentage) as appropriate. Differences between categorical variables at baseline were analysed using R788 order Chi-square or Fisher exact test as applicable. The difference between categorical variables with ordering was analysed using Cochran–Armitage test for trend. The difference between quantitative variables was assessed MEK inhibitor using the Mann–Whitney U test. We first searched the literature for existing systematic reviews on the role of antifungal agents in CPA. No reviews were found. Two authors (RA, GV) then searched the PubMed and EmBase databases, without any limits, to identify the relevant studies published from 1952 onwards describing the role of antifungal agents in CPA. The following search

terms were used: (‘aspergilloma’ OR ‘CNPA’ OR ‘CCPA’ OR ‘CNPA’ OR ‘chronic necrotizing pulmonary aspergillosis’ OR ‘CPA’ OR ‘CCPA’ OR ‘CFPA’ OR ‘CPA’) AND (‘itraconazole’ OR ‘azole’ OR ‘voriconazole’ OR ‘posaconazole’ OR ‘micafungin’ OR ‘antifungal’ OR ‘amphotericin’ OR ‘caspofungin’). In addition, we reviewed our personal files. We included studies reporting on the efficacy of antifungal agents in CPA. We excluded single patient case reports or studies involving <10 patients. Data were recorded on a standard data extraction form. The following items were extracted: publication details (title, authors and other citation details); type of study (prospective or retrospective); antifungal agent, dose and duration of treatment; duration of follow-up;

definitions for overall response used in the individual studies and the overall response rates. During the study period, 34 patients qualified for inclusion in the study of which three patients were excluded (two patients refused consent and one patient Atazanavir was diagnosed as CNPA). Finally, 31 patients (18 men) with a median (IQR) age of 35 (26–44) years were included in the study. Seventeen patients were randomised to the itraconazole group and 14 to the control group (Fig. 1). Majority of the patients (90%) had past history of pulmonary tuberculosis. Aspergillus precipitins were positive in 21 patients. Sputum or BAL fluid culture grew Aspergillus fumigatus in 13 patients. Immediate cutaneous hyperreactivity to Aspergillus antigen was demonstrated in 13 patients but in none, the IgE level exceeded 500 IU ml−1 and A. fumigatus-specific IgE was <0.35 kUA l−1.