Indian Journal of Pathology and Microbiology

: 2011  |  Volume : 54  |  Issue : 1  |  Page : 21--24

A cytohistological study of Ki-67 expression in ovarian tumors

Monisha Choudhury, Seema Goyal, Mukta Pujani 
 Department of Pathology, Lady Hardinge Medical College and Smt. Sucheta Kriplani Hospital, New Delhi, India

Correspondence Address:
Mukta Pujani
Department of Pathology, Lady Hardinge Medical College and Smt. Sucheta Kriplani Hospital, New Delhi - 110 001


Background and Objective: The study was conducted to evaluate the diagnostic accuracy of imprint cytology in ovarian neoplasms, and to investigate the biological significance of Ki-67 antigen expression in benign and malignant ovarian tumors and correlate it with histological type, grade, and stage of malignant tumor. Materials and Methods: A total of 50 cases including 25 prospective and 25 retrospective cases were studied. Imprint cytology was performed on 25 ovarian tumors and compared with histopathological diagnosis. Ki-67 immunohistochemistry was performed on all 50 cases. Results: On immunohistochemistry, benign tumors had a mean Ki-67 index of 3.2 ± 3.7 while malignant tumors had a mean Ki-67 index of 33.1 ± 16.7, the difference being statistically significant. Significant correlation was observed between the Ki-67 index and stage of the tumor; however, there was no correlation between the grade of differentiation and histological type of tumor with the Ki-67 index. Conclusions: In the present study, the Ki-67 index was higher in advanced stage tumors; hence a higher Ki-67 index points toward the aggressive behavior and poorer clinical outcomes.

How to cite this article:
Choudhury M, Goyal S, Pujani M. A cytohistological study of Ki-67 expression in ovarian tumors.Indian J Pathol Microbiol 2011;54:21-24

How to cite this URL:
Choudhury M, Goyal S, Pujani M. A cytohistological study of Ki-67 expression in ovarian tumors. Indian J Pathol Microbiol [serial online] 2011 [cited 2022 Dec 5 ];54:21-24
Available from:

Full Text


Rapid peroperative diagnosis of ovarian masses is important for the surgeon to conserve the other ovary in young patients with benign tumors and for the early institution of chemotherapy in malignant tumors. Imprint cytology is a cost-effective technique which may be used to augment the frozen section analysis. It provides a rapid diagnosis with higher accuracy rates as compared to the frozen section when used alone. [1]

The monoclonal antibody Ki-67 reacts with a nuclear antigen that is expressed only by the proliferating cells. The high proliferation rate has been associated with tumor aggressiveness and correlates with the prognosis and other known clinicopathological features of the tumor. [2]

In the present study, the diagnostic accuracy of imprint cytology has been evaluated in diagnosing benign and malignant tumors of the ovary and correlated with histological diagnosis. Immunochemistry with Ki-67 has been done on the imprint smears and the tissue sections of benign and malignant ovarian tumors. Ki-67 antigen expression was correlated with known prognostic factors like histological type, grade of differentiation, and stage of tumor for ovarian tumors.

 Materials and Methods

A total of 50 cases were taken up in the study which comprised 25 prospective cases and 25 retrospective cases, consisting of both benign and malignant ovarian tumors. The study was approved by the institutional ethical committee.

All the cases were reviewed and representative sections from the tumor were selected. The corresponding blocks were cut into 3- to 5-μm-thick sections on poly-l-lysine-coated slides. Immunostaining was performed on these sections using LSAB+ technique with the monoclonal MIB-1 antibody Ki-67 (DAKO; Denmark).

Twenty-five cases of ovarian tumors which were operated during the study period were included. Imprint smears were made directly from the fresh unfixed specimen on plain and poly-l-lysine-coated slides. Two slides were immediately fixed in 95% ethanol for Papanicoloau stain and hematoxylin and eosin (H and E) stain. Rest of the smears were air dried and stained with Giemsa. Smears on poly-l-lysine-coated slides were wrapped in an aluminum foil and stored below -80°C for Ki-67 immunocytochemistry using peroxidase antiperoxidase technique.

Cytological diagnosis was made on imprint smears and compared with the final tissue diagnosis. Immunohistochemistry was performed for Ki-67 on representative tissue sections. Positive Ki-67 staining was observed as brown, granular nuclear staining [Figure 1]b and [Figure 2]b. For Ki-67 scoring, the most positive area of the tumor was selected avoiding foci of inflammation. The number of positive nuclei was counted in 500 tumor cells in a high-power field (×400 magnification). The average of three counts over the same slide was taken and expressed as the percentage of Ki-67-positive cells in the tumor. [2]{Figure 1}{Figure 2}

Statistical Analysis

Histopathology of the tumors was used as the gold standard for final diagnosis. Statistical evaluation was done using the chi-square test, Student's t-test, and ANOVA F-test. A P-value ≤ 0.05 was considered as significant and a P-value ≤ 0.01 as highly significant.


Out of the 50 cases, 20 (40%) were benign, 29 (58%) were malignant, and 1 was a borderline ovarian tumor. A total of 41 (82%) out of 50 tumors were histologically surface epithelial tumors, 3 (6%) were sex cord stromal tumors, 4 (8%) were germ cell tumors, 1 (2%) was small cell carcinoma, and 1 (2%) a Krukenberg tumor [Table 1]. {Table 1}

No significant difference was observed in the mean ages of patients in benign (38.1 years) and malignant tumors (38.2 years) (P = 0.97). The difference in the mean diameter of benign (11.5 ± 9.7 cm) and malignant tumors (13.9 ± 6.9 cm) was not significant (P = 0.3). In this study, 68% of benign tumors were predominantly cystic while 35% of malignant tumors were predominantly solid and all the solid cystic tumors were malignant, the difference being statistically significant (P = 0.002). The sensitivity and specificity of cytology for diagnosis of malignant ovarian tumors was found to be 80% and 100%, respectively. The diagnostic accuracy on cytology for malignant tumors was 96%.

Benign tumors had a mean Ki-67 index of 3.2 ± 3.7% while the malignant tumors had a mean Ki-67 index of 33.1 ± 16.7%. The difference was highly significant on statistical analysis (P<0.001). The mean Ki-67 index for stage III/IV tumors was 39.5 ± 17.2% as compared to stage I/II tumors with a Ki-67 index of 25.7 ± 14.0%. The difference was statistically significant (P = 0.03). No significant correlation was observed between the Ki-67 index and grade of differentiation of tumors (P = 0.84).

There was no significant correlation of mean Ki-67 labeling index with serous versus nonserous histology of tumors. A significant correlation was observed between Ki-67 immunocytochemistry [Table 2] and immunohistochemistry [Table 3] of benign and malignant tumors.{Table 2}{Table 3}


Ovarian cancer is the third most common malignancy among women accounting for 5.5% of all cancers. [3] About 80% of all ovarian tumors are benign and occur mostly in young women between 25 and 40 years of age. [4] The importance lies in distinguishing benign and malignant tumors of the ovary in the reproductive age group where the conservation of other ovary is important. Intraoperative diagnosis of ovarian tumors can be obtained using either frozen section or imprint cytology. The main advantages of imprint cytology as compared to frozen section are rapidity of preparation with same accuracy rate, excellent preservation of cellular details, no freezing artifact, and cost effectiveness. The limitation of imprint cytology is its inability to distinguish in situ from invasive carcinomas as it does not allow for the evaluation of the depth of infiltration of tumor. [5]

In our study, the diagnostic accuracy of imprint cytology for malignant lesions was 96% (24 out of 25 cases) which was higher as compared to Nagai et al.'s, [6] who found 83.6% accuracy. Correct specific diagnosis was possible in 88% cases. Cytology could not offer the correct diagnosis in a case of Krukenberg tumor because the smears were hypocellular and showed the presence of only spindle-shaped fibroblast-like cells along with few scattered epithelial cells. However, immunostaining showed Ki-67 positivity in the few scattered cells.

The rate at which a tumor proliferates has long been considered to bear a relationship with its clinical course. The determination of growth fraction using Ki-67 is a simple method and has been shown to have a prognostic value in a variety of malignancies like CNS tumors, lymphoproliferative diseases, connective tissue tumors, and breast tumors. [7] Ki-67 expression in different ovarian tumors has been studied by various authors across the world. However, there is a paucity of such a study in the Indian literature.

We evaluated Ki-67 expression on imprint smears (immunocytochemistry) in 25 cases of ovarian tumors and on histopathological sections (immunohistochemistry) in 50 cases of ovarian tumors [Figure 1], [Figure 2]. The mean Ki-67 index for 20 benign cases was 3.0 ± 3.7% and 4 malignant cases was 48.6 ± 24.2% on immunocytochemistry. The mean Ki-67 index for benign tumors was 3.2 ± 3.7% while malignant tumors had a mean index of 33.1 ± 16.7% on immunohistochemistry. A significant difference was obtained between the mean Ki-67 indices in benign and malignant tumors on both immunocytochemistry and immunohistochemistry. These findings are in close agreement with Garzetti et al.'s. [8]

The mean Ki-67 index in stage III/IV was 39.5 ± 17.2% as compared to 25.7 ± 14.0% in stage I/II cancers, the difference being statistically significant. This is in concordance with findings of Harlonzinska et al.[9] and Viale et al.[10] No significant correlation of the Ki-67 index with the grade of differentiation could be demonstrated as observed by other authors. [8],[11] In contrast, Viale et al.[10] have reported a significantly higher mean Ki-67 index in high-grade tumors. According to Kobel et al.,[12] the median Ki-67 index differed significantly between low-grade serous (2.5% with a 95% confidence interval of 0.5-20.4%) versus high-grade serous tumors (22.4% with a 95% confidence interval of 3.6-69.9%). No correlation was found in the mean Ki-67 index with serous versus nonserous histology in this study as also reported by other studies. [9],[10] A significant correlation was observed in the Ki-67 index of benign and malignant tumors on immunocytochemistry as well as immunohistochemistry (P<0.001).

Khouja et al. [13] found a correlation between high Ki-67 expression and the presence of residual disease after primary surgery, ascitis, higher International Federation of Gynecology and Obstetrics substage, poor differentiation, and higher grade. Min et al.[14] reported that Ki-67 overexpression (>50%) correlated with the grade and type of tumor in the ovarian cancers. Ki-67 overexpression (>50%) was absent in the borderline ovarian tumors, whereas ovarian carcinoma showed expression of Ki-67.

In the present study, the Ki-67 index was higher in advanced stage tumors; thus it can be concluded that a higher Ki-67 index points toward the aggressive clinical behavior. FIGO stage and grade are the two most important prognostic factors in ovarian cancers. The growth fraction estimated by the Ki-67 index in this study reflects more closely the clinical stage of tumor than the degree of differentiation of tumors.


1Souka S, Kamel M, Rocca M, El-Assi M, Hebeishy N, Sheir SH. The combined use of cytological and frozen section in the intraoperative diagnosis of ovarian tumors. Int J Gynecol Obstet 1990;31:43-6.
2Wong WS, Tattersall MH. Immunohistochemical determination of tumor growth fraction in human ovarian carcinoma. Br J Obstet Gynaecol 1989;96:720-4.
3Gupta MC, Mahajan BK. Epidemiology of non communicable diseases. In: Gupta MC, Mahajan BK, editors. Textbook of Preventive and Social Medicine. 3 rd ed. New Delhi: Jaypee Brothers; 2003. p. 306-28.
4Crum CP. The female genital tract. In: Kumar V, Abaas AK, Fausto N, editors. Robbins and Cotran Pathologic Basis of Disease. 7 th ed. Philadelphia: WB Saunders Company; 2004. p. 1059-118.
5Scucchi LF, Di Stefano D, Cosentino L, Vecchione A. Value of cytology as adjunctive intraoperative diagnostic method: An audit of 2250 consecutive cases. Acta Cytol 1997;41:1489-96.
6Nagai Y, Tanaka N, Horiuchi F, Ohki S, Seki K, Sekiya S. Diagnostic accuracy of intraoperative imprint cytology in ovarian epithelial tumors. Int J Gynaecol Obstet 2001;72:159-64.
7Brown DC, Gatter KC. Monoclonal antibody Ki-67: Its use in histopathology. Histopathology 1990;17:489-503.
8Garzetti GG, Ciavattini A, Goteri G, De Nictolis M, Stramazzotti D, Lucarini G, et al. Ki-67 antigen immunostaining (MIB 1 monoclonal antibody) in serous ovarian tumors: Index of proliferative activity with prognostic significance. Gynecol Oncol 1995;56:169-74.
9Harlozinska A, Bar JK, Sedlaczek P, Gerber J. Expression of p53 protein and Ki-67 reactivity in ovarian neoplasms: Correlation with histopathology. Am J Clin Pathol 1996;105:334-40.
10Viale G, Maisonneuve P, Bonoldi E, Di Bacco A, Bevilacqua P, Panizzoni GA, et al. The combined evaluation of p53 accumulation and Ki-67 (MIB 1) labeling index provides independent information on overall survival of ovarian carcinoma patients. Ann Oncol 1997;8:469-76.
11Jordan PA, Kerns BJ, Pence JC, Kohler MF, Bast RC Jr, Kinney RB, et al. Determination of proliferation index in advanced ovarian cancer using quantitative image analysis. Am J Clin Pathol 1993;99:736-40.
12Kobel M, Kalloger SE, Boyd N, McKinney S, Mehl E, Palmer C, et al. Ovarian carcinoma subtypes are different diseases: Implications for biomarker studies. PLoS Med 2008;5:e232.
13Khouja MH, Baekelandt M, Nesland JM, Holm R. The Clinical Importance of Ki-67, p16, p14, and p57 Expression in patients with advanced ovarian carcinoma. Int J Gynecol Pathol 2007;26:418-25.
14Min KW, Park MH. The expression of c-erbB-2, EGFR, p53 and Ki-67 in ovarian borderline tumors and carcinomas of the ovary. Korean J Pathol 2007;41:296-306.