Introduction: Although previous studies have reported etiologies, diagnostic strategies, and outcomes of acute respiratory failure (ARF) in cancer patients, few studies investigated ARF in cancer patients presenting with diffuse pulmonary infiltrates.
Methods: This was a retrospective observational study of 214 consecutive cancer patients with diffuse pulmonary infiltrates on chest radiography admitted to the oncology medical intensive care unit for acute respiratory failure between July 2009 and June 2011.
Results: After diagnostic investigations including bronchoalveolar lavage in 160 (75%) patients, transbronchial lung biopsy in 75 (35%), and surgical lung biopsy in 6 (3%), the etiologies of diffuse pulmonary infiltrates causing ARF were identified in 187 (87%) patients. The most common etiology was infection (138, 64%), followed by drug-induced pneumonitis (13, 6%) and metastasis (12, 6%). Based on the etiologic diagnoses, therapies for diffuse pulmonary infiltrates were subsequently modified in 99 (46%) patients. Diagnostic yield (46%, 62%, 85%, and 100%; P for trend < 0.001) and frequency of therapeutic modifications (14%, 37%, 52%, and 100%; P for trend < 0.001) were significantly increased with additional invasive tests. Patients with therapeutic modification had a 34% lower in-hospital mortality rate than patients without therapeutic modification (38% versus 58%, P = 0.004) and a similar difference in mortality rate was observed up to 90 days (55% versus 73%, Log-rank P = 0.004). After adjusting for potential confounding factors, therapeutic modification was still significantly associated with reduced in-hospital mortality (adjusted OR 0.509, 95% CI 0.281-0.920).
Conclusions: Invasive diagnostic tests, including lung biopsy, increased diagnostic yield and caused therapeutic modification that was significantly associated with better outcomes for diffuse pulmonary infiltrates causing ARF in cancer patients.