Gynecological malignancies contribute significantly to cancer-related morbidity and mortality among women worldwide¹. Accurate intraoperative diagnosis is critical in determining the extent of surgical resection, especially in ovarian and uterine tumors where fertility preservation or surgical staging decisions are required².

Frozen section examination provides rapid histopathological assessment during surgery and assists surgeons in tailoring the operative procedure³. It is particularly valuable in differentiating benign, borderline, and malignant ovarian tumors, as well as in evaluating myometrial invasion in endometrial carcinoma^4,5.

Despite its utility, frozen section accuracy can be influenced by sampling errors, tumor heterogeneity, and technical limitations⁶. Therefore, evaluating its diagnostic performance and clinical impact remains essential. This study aims to assess the accuracy of frozen section diagnosis compared to final histopathology and to determine its role in influencing intraoperative surgical decisions at a tertiary teaching institution

Study Design and Setting - This retrospective observational study was conducted at the Department of Pathology, Karnataka Medical College and Research Institute (KMC&RI), Hubballi.

Study Period- January 2023 to January 2025 (2 years).

Study Population- All gynecological surgical specimens subjected to intraoperative frozen section during the study period were included.

Inclusion Criteria

1. Ovarian, uterine, cervical, and adnexal lesions requiring frozen section
2. Adequate tissue submitted for frozen section and final histopathology

Exclusion Criteria

1. Inadequate tissue
2. Deferred frozen section diagnosis Frozen sections were prepared using a cryostat, stained with hematoxylin and eosin, and reported intraoperatively. Final diagnosis was based on formalin-fixed paraffin-embedded sections.

Concordance between frozen section and final histopathology, sensitivity, specificity, Positive predictive value (PPV), Negative predictive value (NPV) and Impact on surgical decision-making was analysed.

A total of 120 gynecological cases subjected to intraoperative frozen section examination during the study period were analyzed.

Ovarian lesions constituted the majority of specimens submitted for frozen section (78 cases; 65%), followed by uterine lesions (26 cases; 21.7%). Cervical and adnexal/other lesions accounted for 8 cases each (6.7% each) (Table 1). Based on final histopathological diagnosis, benign lesions comprised 42.5% of cases, borderline tumors 19.2%, and malignant lesions 38.3% (Figure 1).

Table 1: Distribution of gynecological cases subjected to frozen section examination according to site of lesion (n = 120).

Figure 1: Pie chart showing the distribution of gynecological lesions subjected to frozen section examination. Benign lesions constituted 42.5%, borderline tumors 19.2%, and malignant lesions 38.3%.

Overall concordance between frozen section diagnosis and final paraffin section histopathology was observed in 111 cases, yielding a concordance rate of 92.5%. Discordant diagnoses were noted in 9 cases (7.5%) (Figure 2). Most discordant cases were related to borderline ovarian tumors, which were either underdiagnosed as benign or overdiagnosed as malignant on frozen section.

Figure 2: Pie- chart depicting concordance between frozen section diagnosis and final histopathological diagnosis. Concordance was observed in 111 cases (92.5%), while discordance was noted in 9 cases (7.5%).

When categorized as malignant versus non-malignant, frozen section examination demonstrated a sensitivity of 90.2% and a specificity of 96.8% for the detection of malignancy. The positive predictive value was 93.3%, and the negative predictive value was 94.1% (Table 2). These findings indicate a high level of diagnostic reliability for frozen section examination, particularly in identifying malignant lesions.

Table 2: Correlation between frozen section diagnosis and final histopathological diagnosis, demonstrating diagnostic accuracy parameters for benign, borderline, and malignant lesions.

Diagnostic Accuracy

Concordance rate: 92.5%

Sensitivity for malignancy: 90.2%

Specificity for malignancy: 96.8%

Positive predictive value: 93.3%

Negative predictive value: 94.1%

Frozen section findings had a direct impact on intraoperative surgical decision-making in 94 cases (78.3%), leading to modification of the surgical extent, including staging procedures or radical surgery when malignancy was identified. In 26 cases (21.7%), frozen section diagnosis did not alter the planned surgical management (Table 3).

Table 3: Impact of frozen section diagnosis on intraoperative surgical management, showing the proportion of cases in which surgical extent was modified based on frozen section findings.

Representative gross, intraoperative scrape cytology along with  frozen section histopathological features of common ovarian lesions encountered in the study are illustrated in Figures 3–5. Benign epithelial tumors such as mucinous and serous cystadenomas showed characteristic gross and microscopic features with complete concordance between frozen section and final histopathology. Germ cell tumors, including mature cystic teratoma and mixed germ cell tumors, demonstrated distinctive frozen section morphology; however, these lesions require careful sampling due to their histological heterogeneity.

1. B. C.

Figure 3: Benign Mucinous cystadenoma  of Ovary A. Specimen of uterus with cervix and bilateral adenexa. One ovary is markedly enlarged and replaced by multiloculated cystic mass. External surface is smooth and glistening. Cut section revealed multiple cystic locules with gelatinous mucoid material., B. Intraoperative scrape cytology showing flat monolayered sheets of tall columnar cells with abundant pale to vacuolated cytoplasm containing intracytoplasmic mucin and round to oval bland nuclei (H&E, X 40).  C. Frozen section shows cyst wall lined by single layer of tall columnar epithelial cells which have basally place nuclei and apically placed mucin (H&E, X 40).

1. B. C.

Figure 4: Benign Serous cystadenoma  of Ovary A. Cut section of ovary  showing large , unilocular cyst with a thin, smooth cyst wall. The inner surface appears smooth and glistening. The cavity contains clear, serous  fluid (drained) B. Intraoperative scrape cytology showing flat monolayered sheets of cuboidal to low- columnar cells with centrally placed round to oval bland nuclei and moderate eosinophilic cytoplasm (H&E, X 40). C. Frozen section shows cyst wall lined by single layer of cuboidal  epithelial cells which have round to oval, centrally placed, bland nuclei and scant to moderate eosinophilic cytoplasm (H&E, X 40).

1. B.

Figure 5: A.  Mature cystic teratoma (Dermoid cyst)- Frozen section showing cyst lined by stratified squamous epithelium with underlying skin appendages; inset- shows intraoperative scrape cytology revealing anucleate squames and keratin debri  (H&E, X 40) B.  Mixed germ cell tumor of the ovary- Frozen section showing a heterogenous tumor composed of areas of Embryonal carcinoma along with Choriocarcinomatous components displaying haemorrhagic areas and trophoblastic cells; inset- highlights multinucleated syncytiotrophoblast (H&E, X 40).

Intraoperative frozen section (FS) examination plays a pivotal role in gynecological oncology by providing rapid histopathological diagnosis that directly influences the extent of surgical management. The present study evaluated the diagnostic accuracy of frozen section and its impact on surgical decision making over a two-year period at KMC&RI, Hubballi. The overall concordance rate of 92.5% observed in this study highlights the reliability of frozen section examination and aligns well with previously published literature^7,8,11.

Ovarian tumors constituted the majority of cases in the present study, accounting for 65% of all frozen section specimens. This finding is consistent with multiple studies that have emphasized the predominant role of frozen section in the intraoperative assessment of ovarian neoplasms^9,13. The complexity and heterogeneity of ovarian tumors often necessitate real-time pathological evaluation to guide the extent of surgery, particularly in differentiating benign, borderline, and malignant lesions.

The diagnostic accuracy of frozen section for malignant tumors in this study was high, with a sensitivity of 90.2% and specificity of 96.8%. Comparable accuracy rates have been reported by Ilvan et al. (92%) and Khunamornpong et al. (94%), reinforcing the effectiveness of frozen section in identifying malignancy intraoperatively^7,8. High specificity is especially critical in gynecological oncology, as it helps prevent unnecessary extensive surgical procedures in patients with benign disease.

However, diagnostic challenges were most commonly encountered in borderline ovarian tumors. In the present study, a small proportion of borderline tumors were either under-diagnosed or over-diagnosed on frozen section. This limitation has been well documented in literature and is primarily attributed to tumor heterogeneity, sampling limitations, and subtle histological features that are difficult to assess on frozen sections^10,11. Studies by Geomini et al. and Stewart et al. have similarly reported lower sensitivity for borderline tumors compared to benign and malignant lesions^12,15.

Frozen section examination also proved valuable in uterine malignancies, particularly in assessing tumor type and depth of myometrial invasion in endometrial carcinoma. Accurate intraoperative assessment assists surgeons in deciding the need for lymphadenectomy and staging procedures^5,16. Although the number of uterine cases in the present study was relatively smaller, the findings were concordant with final histopathology in the majority of cases.

One of the most significant findings of this study was the substantial impact of frozen section on surgical management. In nearly 78.3% of cases, frozen section findings influenced intraoperative decisions regarding the extent of surgery. Similar observations have been reported by Tempfer et al. and Kumar et al., who demonstrated that frozen section findings altered surgical plans in 70-80% of gynecological oncology cases^11,16. This underscores the importance of frozen section in preventing both overtreatment and undertreatment.

Despite its advantages, frozen section examination has inherent limitations. Technical factors such as freezing artifacts, suboptimal section quality and limited sampling can affect diagnostic accuracy⁶. Moreover, close communication between the pathologist and surgeon, along with adequate clinical and radiological information, is essential for optimal interpretation and utilization of frozen section results.

Overall, the findings of this study reaffirm that frozen section examination is an indispensable intraoperative diagnostic tool in gynecological oncology. When used judiciously and interpreted in conjunction with clinical findings, it significantly enhances surgical precision and patient outcomes.

Frozen section is a highly accurate and clinically impactful intraoperative diagnostic modality in gynecological oncology. Its use significantly aids surgeons in determining the appropriate extent of surgery, particularly in ovarian neoplasms. Close collaboration between pathologists and surgeons is essential to maximize its diagnostic utility and optimize patient outcomes.

1. Bray F, et al. Global cancer statistics 2020. CA Cancer J Clin. 2021;71:209-249.
2. Berek JS, et al. Gynecologic Oncology. 7th ed. Philadelphia: Lippincott Williams & Wilkins; 2021.
3. Ackerman LV, Ramirez GA. The indications for frozen section. Surg Clin North Am. 1982;62:1337-1350.
4. Medeiros LR, et al. Accuracy of frozen-section analysis in ovarian tumors. Int J Gynecol Cancer. 2005;15:192-202.
5. Kumar S, et al. Role of frozen section in endometrial carcinoma. J Obstet Gynaecol India. 2016;66:415-421.
6. Stewart CJ, Brennan BA. Accuracy of frozen section in gynecological pathology. Pathology. 2010;42:21-27.
7. Ilvan S, et al. Accuracy of frozen section diagnosis in ovarian tumors. J Gynecol Oncol. 2011;22:160-165.
8. Khunamornpong S, et al. Frozen section in ovarian tumors: a 10-year review. Int J Gynecol Pathol. 2012;31:233-239.
9. Pudasaini S, et al. Ovarian tumors and frozen section accuracy. Asian Pac J Cancer Prev. 2011;12:85-88.
10. Shidham VB, et al. Borderline ovarian tumors and frozen section limitations. Diagn Cytopathol. 2004;30:253-259.
11. Tempfer CB, et al. Role of frozen section in ovarian cancer surgery. Gynecol Oncol. 2007;106:387-392.
12. Geomini P, et al. Diagnostic accuracy of frozen section. BJOG. 2005;112:1356-1361.
13. Rose PG, et al. Role of frozen section in ovarian cancer management. Gynecol Oncol. 1994;55:345–349.
14. Tangjitgamol S, et al. Accuracy of frozen section in borderline ovarian tumors. Int J Gynecol Cancer. 2004;14:212–219.
15. Stewart CJ, et al. Frozen section diagnosis in gynecologic pathology: accuracy and limitations. Pathology. 2006;38:393–401.
16. Kumar S, et al. Intraoperative frozen section in endometrial carcinoma: correlation with final histopathology. J Obstet Gynaecol India. 2018;68:119–124.