We have studied the differential cytology of BAL fluid in 48 children aged 3-16 years (mean age±SD 7.9±3.5 yrs) undergoing elective surgery for nonpulmonary illnesses. A flexible bronchoscope (Pentax 3.5 or 4.9 mm) was wedged in the middle lobe or one of its segments. BAL was performed with 3×1 ml·kg -1 body weight of normal saline warmed to body temperature. The first sample was studied separately; subsequent samples were pooled for analysis.The mean recovery was 58±15%. Significantly more granulocytes and less lymphocytes were found in the first, as compared to the pooled, sample. Total cell counts were highly variable and ranged from 0.5-57.1 × 10 4 ·ml -1 . Macrophages were the predominant cell type, with a mean percentage of 81.2±12.7%. The relative proportion of lymphocytes was higher than that reported in most studies of adult volunteers (16.1±2.4%). No age dependency was observed for either cell type. The mean percentage of granulocytes was 2.5±3.3%. Absolute granulocyte counts were significantly higher in children under 8 yrs of age.This study provides the first reference data on BAL differential cytology in children without pulmonary disease and will be the basis for future investigations of BAL in paediatric lung diseases.

Bronchoalveolar lavage (BAL) is increasingly used in the pediatric age group. However, normal values for BAL fluid (BALF) constituents are lacking. As part of an investigation to define reference values in children, we studied lymphocyte surface markers of BALF in 28 children 3 to 16 yr of age without bronchopulmonary disease. All of them were undergoing elective surgery for nonpulmonary illnesses. BAL was performed under general anaesthesia with tracheal intubation. A flexible bronchoscope (Pentax 3.5 or 4.9 mm) was wedged into the middle lobe or into one of its segments, and 3 x 1 ml/kg body weight normal saline warmed to body temperature were instilled and immediately withdrawn. The first sample was studied separately; subsequent samples were pooled for analysis. Studies on lymphocyte surface markers were performed on the pooled sample only. The distribution of B-cells, pan T-cells, and CD57 positive cells was within the range reported for adult subjects. However, CD4/CD8 ratios were lower than in adults (0.7 +/- 0.4, mean +/- SD). This decrease in the CD4/CD8 ratio was caused by an increase in CD8 cells. Comparative analysis of blood and BALF lymphocytes in a subgroup of children showed that CD4/CD8 ratios in blood were within the normal range reported for this age group and significantly higher when compared with BALF. The lower CD4/CD8 ratios in normal children have to be considered in the interpretation of BALF in children with pulmonary diseases.