Advancing Education, Research, and Quality of Care for the Head and Neck oncology patient.
BACKGROUND: Current recommendations by the American Society of Clinical Oncology for the management of squamous cell carcinoma of unknown primary (SCCUP) include submission of suspicious primary sites for frozen section analysis (FSA). However, a recent study found the sensitivity of FSA for SCCUP to be less than 50%, calling into question the utility of this approach. We aim to further investigate the utility of frozen section analysis for identification of the HPV-positive SCCUP.
METHODS: A multi-institutional retrospective review of patients with p16-positive SCCUP who underwent diagnostic laryngoscopy with biopsy (DLB), diagnostic transoral robotic surgery (TORS), or definitive TORS and neck dissection between May 2012 and September 2024 was conducted. Patients were considered to have SCCUP if they presented with a neck mass with no definitive evidence of primary tumor on clinical or radiographic examination (computed tomography [CT] with contrast, magnetic resonance imaging [MRI], positron emission tomography [PET]). Patients with prior head and neck cancer and p16-negative or indeterminate p16 status were excluded. Patients who did not have intraoperative FSA were also excluded. Results of FSA and subsequent permanent histopathologic evaluation of each specimen were collected in addition to basic demographic and clinical information. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for FSA were determined and further analyzed by specimen type. Mann-Whitney U test was performed to assess difference in tumor size between primaries identified on FSA and on permanent section only.
RESULTS: Seventy-eight patients were identified and included for analysis. Sixty-six patients underwent definitive TORS, 7 patients underwent diagnostic TORS, and 5 patients underwent DLB. Primary tumors were identified in 63 patients (80.8%). The primary was identified in 80.8% of patients who underwent TORS and in 80% of the patients who underwent DLB. Of the primary tumors identified, mean tumor size was 1.1 cm (range 0.2-2.5 cm). There was no significant difference in size between patients whose tumor was identified intraoperatively (mean 1.1 cm) and on permanent sections only (mean 0.87 cm) (p = 0.21). The sensitivity, specificity, PPV, and NPV of FSA was 87.7% (95% CI 77.9%-94.2%), 100% (95% CI 94.4%-100%), 100% (95% CI 94.4%-100%), and 87.7% (95% CI 77.9%-94.2%), respectively. False negatives (frozen section negative for carcinoma, subsequent permanent section positive for carcinoma) occurred due to sampling error 6 times and due to misinterpretation of the frozen section 3 times. Frozen specimens included 54 biopsies (39.1%) and 84 complete excisions (60.9%). In the biopsy specimens, sensitivity was 100% (95% CI 86.3%-100%) and NPV was 100% (95.5% CI 88.1%-100%), whereas in the excised specimens, sensitivity was 81.2% (95% CI 67.4%-91.1%) and NPV was 79.5% (95% CI 64.7%-91.1%).
CONCLUSION AND RELEVANCE: In this case series of 78 patients with SCCUP, the sensitivity and NPV of frozen section analysis for identification of the primary tumor was over 85%. Frozen section analysis is a valuable adjunct tool to diagnostic and definitive operation for SCCUP but use should be considered on an institutional basis.