Background: In about 20% of patients with malignancies with acute respiratory failure (ARF), no etiology can be determined, whatever the diagnostic strategy used. Lung biopsy could then be a precious diagnostic tool leading to therapeutic adaptations and increasing chances for cure. The aim of this study was to assess the diagnostic contribution of lung biopsy in patients for whom a complete diagnosis strategy failed to identify ARF etiology.
Methods: All hematology patients admitted for ARF to our ICU between 1995 and 2011, and for whom lung biopsy was performed were included in the study. Lung biopsies were surgical, CT guided, or post-mortem. Histological findings were compared to prebiopsy diagnosis and classified into specific or non-specific diagnosis. Therapeutic impact (or Goldman-class in post-mortem biopsies) was also recorded.
Results: Among the 1440 hematology patients with ARF managed during the study period, 21 (1%) biopsies were performed, including 10 post-mortem biopsies. Histological diagnoses were specific in 10 biopsies, non specific in 8 biopsies and lung parenchyma was normal in three patients. In 8/11 (72.7%) alive patients, lung biopsy had lead to therapeutic modifications, including treatment implementation in 5 patients and treatment withdrawal in 3 patients. One out of 10 (10%) patients had minor complications. For the 10 dead patients, only one Goldman-type 1 error was found.
Conclusion: Diagnostic lung biopsy is rarely needed in hematology patients with ARF. But, it has a 73% therapeutic impact and has overall no major complications. Contribution from post-mortem biopsies seems less relevant.