A diagnostic algorithm for early diagnosis and management of acute invasive fungal sinusitis

Abstract

PurposeTo define different risk groups of patients suspected of having acute invasive fungal sinusitis (AIFS) and develop a goal-directed diagnostic approach. Materials and methodsForty patients with suspected AIFS biopsied from 2010 to 2020 were included in this study. Patients diagnosed with chronic invasive fungal sinusitis or without biopsy results were excluded. A recursive partitioning analysis (RPA) model was performed to define patient cohorts with the highest risk of having a positive biopsy for AIFS. ResultsThere were a total of 26 patients with biopsy-proven AIFS. Patient characteristics significantly associated with an increased likelihood of a positive biopsy for AIFS on bivariate analysis included facial pain (p = 0.047), platelet count <50,000 cells/mm3 (p = 0.028), and abnormal CT findings, most commonly, bilateral sinus opacification (p = 0.003). The RPA model identified three risk factors for predicting a patient's probability of having a positive biopsy for AIFS, resulting in four-terminal nodes. In the twenty-six patients who had biopsy-proven AIFS, the post-operative 30-day all-cause mortality was 50 % (13/26) and overall mortality was 88.5 % (23/26). Predictors of 30-day all-cause mortality included prolonged interval between biopsy and operative start time (p = 0.042) and earlier initiation of antifungals prior to the operative start time (p = 0.042). ConclusionOur findings indicate that patients with a fever of unknown origin, low platelet count, and/or ANC are at an increased risk of being diagnosed with biopsy-proven AIFS. Using these risk factors, we propose a diagnostic approach that may expedite the treatment of patients with AIFS; however, future prospective studies are needed for validation.