Comparison of intraoperative imprint cytology with frozen section for lymph node metastasis in patients with head and neck cancer.

Abstract

e18548 Background: Intraoperative evaluation of lymph nodal metastasis in head and neck carcinoma assumes importance in deciding the extent of lymph node dissection. Frozen section is the most commonly employed technique. But it requires significant investment in terms of resources, time and personnel. Intraoperative imprint cytology is a rapid, reliable and an inexpensive alternative. So we conducted a prospective diagnostic test accuracy study to assess the diagnostic accuracy of intraoperative imprint cytology and frozen section for lymph node metastasis in Head and neck Cancer when compared to Gold standard, which is the final Histopathology. Methods: All patients presenting with squamous cell carcinoma of head and neck with clinically N0 neck undergoing surgery as primary treatment were included in the study. For patients undergoing surgery, neck nodal tissue were sent for intraoperative assessment of clinically suspicious nodes. Intraoperative assessment was done by imprint cytology and frozen section and was reviewed and reported by two independent pathologist. The sensitivity, specificity and accuracy of imprint cytology and frozen section were calculated with reference to the final histopathology report. Time duration for reporting were also calcuted for both. Results: Total 34 patients with clinically N0 neck were included in the study and 85 slides were examined for frozen section and imprint cytology. Occult nodal metastasis was found in 16 patients (47.06%). The mean time duration for reporting for frozen was 41.18 ±3.62 minutes where as for imprint cytology it was 18.12 ±2.01 minutes. The concordance rate between the frozen section and imprint cytology was 94.12%. The sensitivity of frozen section was 100 % where as for imprint cytology it was 85.7 %( p = 0.00). The specificity was 98.6 % for frozen section and 95.8 % for imprint cytology (p = 0. 27). The positive predictive value was 93.3 % and 80 % for frozen section and imprint cytology respectively (p = 0.01). The negative predictive value was 100% for frozen section and 97.1% for imprint cytology (p = 0.12). The accuracy for frozen section was 98.8 % where as for imprint cytology it was 94.1%(p = 0.09). Conclusions: Imprint cytology provides a cheaper, accurate, reliable and a rapid alternative for frozen section for intraoperative assessment of neck nodes in squamous cell carcinoma of head and neck and it assumes importance in resource driven countries like India.