Routine ventilation scans in children with cystic fibrosis : diagnostic usefulness and prognostic value

Krypton ventilation scans (VS) provide an index of peripheral lung function, and may be particularly useful in children unable to perform pulmonary function testing. This communication reports on three linked studies which investigated whether a routine VS in young children with cystic fibrosis (CF) is diagnostically or prognostically useful. Study 1: In a preliminary study in 1991, VS were compared with clinical examination and chest radiography (CXR) in 50 CF children (29 females, 21 males) aged 0.4-5.2 years (median 2.2 years). The chest was divided into six zones, and abnormalities scored from 0 (normal) to 2 (very abnormal). Clinical examination was unhelpful in predicting abnormalities on imaging. In five children (10%) with a normal CXR, VS was abnormal, and in a further eight children (16%), CXR markedly underestimated VS changes. Study 2: In order to determine the long-term prognostic significance of VS abnormalities, we followed up 27 (19 females, 8 males) of the children from study 1, who had had their first VS at presentation at median age 1.6 years (range 0.4-5.2), scoring the same six zones from 0 to 2. Follow-up was for a mean of 11.6 years (range 7.8-14.8). Spirometry at age 7 years showed a mean forced expiratory volume in 1 s (FEV1) of 96% (range 46%-145%) and a mean forced vital capacity (FVC) of 96% (range 46%-145%). A poor VS score at presentation was correlated with percent predicted FEV1 at age 7 (r=0.4, P=0.042, 16% of variance explained). Those with a normal VS at presentation had a mean FEV1 at presentation of 99% (range 80%-129%). Whereas four patients had an abnormal VS, a normal CXR and a low FEV1 at age 7 years, no patient had a normal VS, an abnormal CXR and a low FEV1 at age 7 years. Study 3: Fifty children (29 females, 21 males) aged 0.5-6.0 years (median 3.8) were prospectively studied in 1998, to determine whether the findings in study 1 were stable over time, and to assess whether VS altered clinical management. Symptoms and clinical examination did not predict abnormalities on imaging. Thirty (60%) children had a normal VS while only five (10%) had a normal CXR. There was a significant correlation between the total scores of CXR and VS (P=0.007, 14% of variance explained). Further, VS detected additional abnormalities in seven patients (14%). Sixty-five percent of patients with an abnormal VS had modifications of treatment, including bronchoscopy, compared with 23% of those with a normal VS. We conclude that VS is a sim