Determination of variables for a more accurate diagnostic approach in suspected acute invasive fungal rhinosinusitis: A non‐concurrent cohort study

Objective To describe a group of patients with suspected acute invasive fungal rhinosinusitis (AIFRS) diagnosis, and identify factors associated with a greater risk of presenting this disease. Design Non‐concurrent cohort study. Setting A single‐centre non‐concurrent follow‐up of patients with suspected AIFRS between August 2015 and July 2018. Participants 50 inpatients referred due to suspected AIFRS at Hospital Clínico Universidad Católica based on the association of a predisposing factor (neutropenia/immunodeficiency/poorly controlled diabetes) with fever of unknown origin. Main outcome measure The primary outcome was AIFRS diagnosis, defined as a concordant tissue biopsy. Results Acute invasive fungal rhinosinusitis was confirmed in 18% (9/50) of the evaluated patients. AIFRS was significantly associated with a positive galactomannan (P = .04), and a paranasal sinus MRI with lack of contrast enhancement (LoCE) (P = .04) orbit compromise (P = .03) or global extrasinusal extension (P = .04). LoCE and extrasinusal extension in the paranasal sinus/brain MRI were risk factors for AIFRS (OR 16; CI 1.2‐210.6 and OR 12.75; CI 1.3‐128.8, respectively). Conversely, a nasal endoscopy showing healthy mucosa was identified as a protective factor for AIFRS (OR 0.06; CI 0.007‐0.57). Conclusions In patients with suspected AIFRS, we identified laboratory and radiologic variables associated with the disease, which may help for a more accurate diagnostic algorithm and approach in this population.