Intraoperative frozen section histopathology diagnosis (IFSH) is a commonly used adjunct to guide ablative head and neck surgery. However, the literature regarding IFSH skull base and sinonasal tumors lacks consensus regarding benefit. IFSH has been reported to be less reliable for sinonasal tumors due to the unique pathologies and anatomy of this location. The aim of this study is to understand the utility of IFSH in skull base and sinonasal tumor surgery and assess what factors are of influence.
Following institutional review board approval, a retrospective review of IFSH was performed of a departmental sinonasal and skull base tumor database at a tertiary care center from 1973 until 2016. Inclusion criteria were patients who underwent surgical resection. Exclusion criteria included no IFSH performed or lack of clinicopathologic information. Demographic information (age, sex), clinical and treatment information (site and approach) as well as histopathologic information (frozen and final) were collated and analysed.
An incorrect diagnosis occurred when tumor pathologic type on IFSH differed to that on formalin. An insufficient quality/quantity sample, as classified by a pathologist, resulted when sample quality or size affected interpretation. While an inaccurate IFSH occurred when there was discordance between the IFSH and final pathology. An IFSH result was considered ambiguous when the read was not definitively positive or negative (i.e. atypical or suspicious for malignancy). Analysis of IFSH were calculated using final pathology as the gold standard.
576 cases were identified and after exclusion criteria there were 230 cases for final analysis. Mean age at time of surgery was 55.2 (SD +/-17.4) years, with 131 males (57.0%) and 99 (43.0%) females. Primary tumor site most commonly found in the nasal cavity in 37.4% (n=82), followed by maxillary sinus 24.8% (n=57). Surgical approach was open in 93.5%, (n=215). Most common pathology was squamous cell carcinoma (SCC) accounting for 35.5% (n=82).
A total of 846 IFSH were performed with a median of 3 (IQR 1-5) per case. An incorrect diagnosis was given in 2 samples (0.2%). 12.2% (n=28) of patients had an inaccurate IFSH on one or more IFSH specimens. An insufficient quality/quantity sample was found in 2.1% (n=18) of samples. An ambiguous result was returned on 34 samples (4.0%) in 27 patients (11.7%).
IFSH analysis was done per specimen for sensitivity, specificity, positive predictive value, and negative predictive value. For the entire cohort it was 89.3%, 98.3%, 95.7%, and 95.6% respectively. For SCC it was 92.7%, 97.9%, 92.7%, and 97.9% respectively. The lowest sensitivity per specimen was for sarcoma (70.8%) and esthesioneuroblastoma (75.0%).
IFSH can be used in sinonasal and skull base resection but the limitations of this technique should be understood by the surgeon. IFSH has a high specificity for sinonasal malignancy. Sensitivity is influenced by tumor type; where IFSH is more sensitive in detecting SCC and salivary malignancies and less so sarcoma and esthesioneuroblastoma. Clinically ambiguous results are difficult to interpret and may not correlate as well with final histopathology.