Utility of intraoperative frozen sections in surgical decision making for acute invasive fungal rhinosinusitis.

BACKGROUND: Acute invasive fungal rhinosinusitis (AIFRS) represents a fulminant, potentially fatal, disease process in immunocompromised patients. The diagnosis often rests on high index of clinical suspicion, with relative paucity of data on the diagnostic and therapeutic implications of intraoperative frozen sections.

METHODS: Retrospective review was performed for 18 cases undergoing endoscopic sinus surgery for AIFRS. Reliability of intraoperative frozen section diagnosis was evaluated for all patients using final pathology as the gold standard.

RESULTS: A total of 66 frozen sections were performed. Diagnostic accuracy of frozen sections illustrated sensitivity of 72.7% (95% confidence interval [CI], 0.57 to 0.85), specificity of 100% (95% CI, 0.85 to 1.00), positive predictive value (PPV) of 100% (95% CI, 0.89 to 1.00), and negative predictive value (NPV) of 64.7% (95% CI, 0.46 to 0.80). There was no statistically significant difference in sensitivity of frozen sections in cases of Mucor and Aspergillus at 68.8%% and 76.2%, respectively (p = 0.61).

CONCLUSION: This study represents the largest series assessing the diagnostic accuracy of frozen section analysis in AIFRS. Frozen section analysis is an effective tool for guiding intraoperative decision making in patients with AIFRS with a high PPV. A Low NPV underscores the importance of clinical suspicion and intraoperative decision making based on endoscopic findings when negative frozen section results are encountered. Further, frozen section analysis appears to be equally effective in detecting either Mucor or Aspergillus.