Published in:
Open Access 01-12-2022 | Hypersensitivity Pneumonitis | Research
Bronchoalveolar lavage lymphocytosis in hypersensitivity pneumonitis: a retrospective cohort analysis with elimination of incorporation bias
Published in: BMC Pulmonary Medicine | Issue 1/2022
Login to get accessAbstract
Background
Recent studies support the diagnostic role of bronchoalveolar lavage lymphocytosis (BALL) in patients with suspected hypersensitivity pneumonitis (HP). Our study aim was to determine the spectrum of BALL findings with elimination of incorporation bias in non-fibrotic and fibrotic patients and assess correlates of positive BALL cut-off and BALL association with long-term outcomes in those with fibrotic disease (f-HP).
Methods
A single-center retrospective cohort study was pursued of patients undergoing diagnostic bronchoscopy for interstitial lung disease. Strict study enrollment was based on recent ATS/JRS/ALAT diagnostic guidance meeting ‘moderate’ or higher diagnostic confidence. BALL findings were assessed in both fibrotic and non-fibrotic HP patients with regression and survival analysis pursued for correlates of positive BALL cut-off and long-term outcome.
Results
A total of 148 patients (88 fibrotic and 60 non-fibrotic) meeting moderate or higher diagnostic confidence were included. Median BALL in f-HP was 15% compared to 19% in non-fibrotic patients, with only 28% of f-HP meeting diagnostic cut-off (≥ 30%) compared to 41% of non-fibrotic. For f-HP, centrilobular nodules on computed tomography was positively correlated with a diagnostic BALL (OR 4.07; p = 0.018) while honeycombing was negatively correlated (OR 6.9 × e−8; p = 0.001). Higher BALL was also associated with lower all-cause mortality (HR 0.98; p = 0.015).
Conclusion
With elimination of incorporation bias, most patients with well-described HP did not meet diagnostic BALL thresholds. Higher BALL was associated with better long-term survival in those with fibrosis, but its diagnostic role may be more additive than characteristic or distinguishing.