Concordance analysis of cells present in bronchoalveolar lavage in patients with interstitial lung diseases by type of readings and pulmonary location

Abstract

Introduction: The study of the inflammatory profile of bronchoalveolar lavage (BAL) is useful in the diagnosis of interstitial lung diseases (ILD) and may avoid subsequent invasive techniques. Usually, the lavage is performed in the middle lobe, and the results are calculated through a manual cell count. Objectives: To assess if there is any difference between obtaining samples from different lobes or varying the quantity of cells counted. Methods: Seventy patients with ILD undergoing BAL during the diagnostic process were included prospectively between 5.2013-10.2015. Middle-lobe (ML) samples were examined by two different laboratories, giving results through a 100 or 500-cells count, respectively (LabH / LabR). An additional sample from the lower right lobe (LRL), obtained only from patients with acceptable pulmonary function (FVC> 50% and DLCO> 30%), was analyzed the same way. Results: Seventy ML and forty-six LRL samples were obtained. Concordance between 100-cell or 500-cell count was low in all series, except for the lymphocyte count (Tables 2 and 3). On the other hand, concordance between samples obtained from different lobes was high, regardless of the number of cells counted (Tables 4 and 5). Conclusions: The cell count of BAL samples differs depending on the technique used, being probably more reliable the one performed on a higher total number of cells. The analysis of two lung lobes BAL has no benefit in comparison to BAL for a single lobe.