Diagnostic Accuracy of Intraoperative Imprint Cytology and Frozen Section in Ovarian Tumours: Experience From a Tertiary Care Centre in Montenegro

OBJECTIVE: Intraoperative assessment of ovarian tumours is crucial for surgical decision-making, especially regarding extent of resection and fertility preservation. Frozen section is standard, while imprint cytology is a potential alternative method. The aim was to evaluate and compare the diagnostic accuracy of imprint cytology and frozen section in ovarian tumours using final histopathology as the reference standard in this study. METHODS: This retrospective study included 29 consecutive patients with ovarian masses operated at a tertiary care centre in Montenegro during 2023-2024. All cases were evaluated intraoperatively using both imprint cytology and frozen section. Final histopathological diagnosis on paraffin-embedded sections, classified according to WHO 2020 criteria, served as reference standard. Archival material was reviewed independently by two pathologists blinded to final diagnosis; discrepancies were resolved by consensus. Statistical analysis included sensitivity, specificity, predictive values, accuracy, and correlation testing using chi-square test and Spearman correlation with p < 0.05 significant. RESULTS: Imprint cytology correctly classified 28/29 cases, yielding an overall diagnostic accuracy of 96.6%. Sensitivity was 100% for benign and borderline tumours and 80% for malignant tumours. Frozen section correctly classified 26/29 cases, with an overall accuracy of 89.7%; sensitivity was 100% for benign, 50% for borderline and 60% for malignant tumours. Agreement with final histopathology was higher for imprint cytology (κ = 0.907) than for frozen section (κ = 0.704). The difference in overall accuracy was not statistically significant (McNemar p = 0.500). CONCLUSION: Imprint cytology demonstrated promising diagnostic performance and may represent a useful adjunct or alternative to frozen section, particularly in resource-limited settings. Larger prospective multicentre studies are required to validate these findings.