This series represents one of the largest single-institution experiences of AIFRS published to date. Timely diagnosis is necessary to improve patient outcomes and limit morbidity. Maintaining a high index of suspicion in at-risk patient populations, followed by prompt evaluation and management, is crucial in suspected AIFRS. The presence or absence of certain findings appear to correlate with biopsy results and may aid in appropriately gauging clinical suspicion for the presence of AIFRS.

Objective Identify methods to improve the frozen-section diagnosis of acute invasive fungal rhinosinusitis. Study Design Biopsies with frozen section for suspected acute invasive fungal rhinosinusitis were reviewed to identify causes for missed diagnoses and evaluate methods for potential improvement. Setting All aspects of the study were performed at the Penn State Milton S. Hershey Medical Center. Subjects and Methods All frozen sections performed for suspected acute invasive fungal rhinosinusitis between 2006 through 2017 were reviewed with their diagnoses compared to the final diagnoses. Sensitivity and specificity were determined for each biopsy specimen to evaluate the diagnostic method and for each patient for its effectiveness on outcome. Causes for frozen-section failures in diagnosis were identified. A periodic acid-Schiff stain for fungus (PASF) was modified for use on frozen tissue (PASF-fs) and applied both retrospectively and prospectively to frozen sections to determine its ability to identify undetected fungus and improve diagnostic sensitivity. Results Of 63 biopsies positive for acute invasive fungal rhinosinusitis, 51 were diagnosed on frozen section, while 61 were identified by including the novel PASF-fs stain, reducing the failure rate from 19% to 3%. Of 41 cases that were positive, 34 were diagnosed on frozen section. Of the 7 that were not, 5 were identified by including the PASF-fs, reducing the failure rate from 17% to 5%. Conclusions Frozen section interpretation of biopsies for suspected acute invasive fungal rhinosinusitis using a PASF-fs stain should enable a rapid and accurate diagnosis with improved outcomes by shortening the time to surgery.

ABSTRACT

Acute invasive fungal rhinosinusitis (AIFRS) is a fulminant infection in immunocompromised patients requiring rapid diagnosis (DX), frequently made on frozen section (FS) of sinonasal biopsies, followed by prompt surgical debridement. However, FS interpretation is often difficult and DX sometimes not possible. In this study we sought to characterize reasons for misinterpretation and methods to improve diagnostic accuracy. The FS slides from 271 biopsies of suspected AIFRS in a 16-year period were reviewed and the morphologic features evaluated for their utility in DX. Recurring specific patterns of necrosis were identified, which to our knowledge have not been described in the literature. Although they provide strong evidence for AIFRS, identifying fungus consistently in necrotic tissue is essential for DX. Clues to identifying fungus and pitfalls in misidentification were identified, but even with expert knowledge of these, a gap in accurate DX remained. The key to FS DX of AIFRS is to improve fungus identification in necrotic tissues. Methods had been sought in the past to stain fungus at FS without consistent success. The Periodic Acid Schiff's Reaction for Fungi was modified by our histopathology department for use on frozen tissue (PASF-fs) resulting in effective staining of the fungus. It stained fungus on all 62 positive slides when applied retrospectively over hematoxylin and eosin (H&E) stained FSs and used prospectively at FS for DX. Although knowledge of histologic morphology on FS is important, the crucial value of this study is the novel use of PASF-fs to identify fungus in the DX of AIFRS.