Correlation of human papilloma virus presence with precancerous and cancerous lesions of uterine cervix by immunohistochemistry

Background : Cancer of cervix is the most common form of cancer in females of developing countries. Cervical cancer is the best example of common human malignancy with a proven infectious etiology. The data linking human papilloma virus (HPV) infection with the epidemiology and pathogenesis of cervical neoplasia is convincing. There are various methods for detection of HPV like immunohistochemistry, polymerase chain reaction, liquid phase hybridization (hybrid capture test), in situ hybridization etc. Materials and Methods: We studied the profile of precancerous and cancerous lesions of uterine cervix and correlated human papilloma virus (HPV) presence with precancerous and cancerous lesions of uterine cervix by immunohistochemistry. Total 50 cases were first studied, analyzed, and classified histologically and then immunohistochemistry was done. Results : Majority of the cases 36/50 (72%) under study are squamous cell carcinomas, which is the most common carcinoma found in the uterine cervix. HPV positivity for all cases of cervical carcinoma was found to be 34.7% (16/46 cases), for carcinoma in situ, it was 50% (2/4 cases). It can be thus hypothesized that with decreasing differentiation, there is loss of expression of HPV in the cervical epithelial cells. Conclusion: To conclude, though this study confirms the correlation of HPV presence with precancerous and cancerous lesions of uterine cervix, it also suggests that there must be other co-factors involved in cervical carcinogenesis as well.

Keywords: Carcinoma cervix, human papilloma virus, precancerous lesions of cervix

Cancer of cervix is the most common form of cancer in females of developing countries and is the second most common cancer, ranking after breast cancer in females of developed world. ^[1] Cervical cancer is the most common cancer of female genital tract in India with approximately 1,00,000 (one lakh) new cases occurring each year. This accounts for almost 20% of all new cases diagnosed in the world annually. ^[2]

Cervical cancer is the best example of common human malignancy with a proven infectious etiology. ^[3] The data linking human papilloma virus (HPV) infection with the epidemiology and pathogenesis of cervical neoplasia is convincing. In humans, 85 types of papilloma viruses have been characterized and fully sequenced and more than 120 putative novel types have been partially characterized. ^[4]

Approximately 40 different types infect the epithelium of anogenital tract. Papilloma virus are epitheliotropic viruses that predominantly infect skin and mucous membrane and produce epithelial proliferation at the site of infection. ^[5] HPV infection of cervix falls into two groups. HPV types 6, 11, 31, and 35 are associated with condyloma acuminata and low grade CIN, which does not progress and is only rarely associated with invasive cancers. This group has been termed BENIGN HPV. HPV types 16, 18, and 33 are associated with flat condylomata, low grade CIN that progresses, high grade CIN, and invasive carcinoma, the so called MALIGNANT HPV.

Various methods available for detection of HPV include immunohistochemistry, polymerase chain reaction, liquid phase hybridization (hybrid capture test), in situ hybridization etc.

We studied the profile of precancerous and cancerous lesions of uterine cervix and correlated human papilloma virus (HPV) presence with precancerous and cancerous lesions of uterine cervix by immunohistochemistry.

The biopsy material for this study was obtained from patients of cervical cancer/squamous intraepithelial lesion admitted in Department of Gynecology and Obstetrics and Department of Radiotherapy, S.N Medical College, Agra. Tissue sections were obtained from these biopsy specimens in the histopathology laboratory of our department. A total of 50 cases were included in the study. The distribution is shown in [Table 1].

3-5 micrometer sections of 10% neutral buffered formalin fixed and paraffin-embedded human cervical biopsy tissue samples were obtained. These were mounted on silanized/ polylysine coated slides. The sections were dried for 16 hrs at 37 ⁰ C, followed by 1 hour at 60 ⁰ C. Heat-induced epitope retrieval using microwave oven was used for antigen retrieval. HPV immunostaining was applied using ready-to-use polyclonal rabbit anti-papilloma virus antibody by DAKO. Primary antibody was Polyclonal Rabbit Anti-Papillomavirus (ready to use preparation – No dilution required) by DAKO. Detection system used was Dakocytomation Envision ^TM + System- HRP Labelled Polymer. It uses a mouse/rabbit/murine-derived anti-Papilloma-virus antibody to react with major capsid antigen of HPV, particularly L _1. The gene L ₁ is predominantly expressed in nuclei of koilocytes because its expression is limited to differentiated cells. ^[6]

The antibody was directed against L ₁ capsid protein; therefore, nuclear positivity was taken to be positive. Cytoplasmic positivity, however, intense was considered negative.

Total 50 cases were studied, analyzed, and classified histologically. Majority of the cases 36/50 (72%) under study were squamous cell carcinomas, which is the most common carcinoma found in the uterine cervix. 4/50 (8%) of high grade dysplastia (HSIL/ CINIII/Ca-in-situ) were also taken to cover the spectrum of lesions from precancerous to cancerous end. Apart from this, 7 cases of adenocarcinoma, 2 cases of adenosquamous carcinoma, and 1 rare case of transitional cell carcinoma were also included. Majority of the cases of cervical cancer, both squamous and adenocarcinomous, were in 41-60 years age group interval.

HPV positivity for all cases of cervical carcinoma was found to be 34.7% (16/46 cases), for carcinoma in situ, it was 50% (2/4 cases) [Table 2]. Overall, HPV positivity for all 50 cases of precancerous and cancerous lesions of uterine cervix was found out to be 32% (16/50 cases). Chi-square test was applied on the above-mentioned observations. It was found that P value > 0.05 and degrees of freedom = 4.

Overall, about 30% of all squamous cell carcinomas were found to be positive, out of which, large cell non-keratinizing type contributes nearly half of the cases (46.15%); adenocarninoma cervix cases have shown over 70% positivity for HPV by immunostaining.

In our study, overall 34.7% of all cases of cervical carcinomas were positive for HPV immunostaining, 30% of all squamous cell carcinomas showed positivity for HPV. This compares fairly well with the positivity rate of 25% for squamous cell carcinoma reported by Sheperd et al^[7] who detected types 6, 16, and 18 using type-restricted monoclonal antibodies raised against fusion protein representing L ₁ major capsid protein. Our results are also comparable with the study conducted by Ravazoula et al^[8] for immunohistochemical detection of HPV protein in cervical cancer who reported a positivity of 31%.

In the present study, HPV immunostaining showed a positive result for only 34.7% cases of cervical cancer. Hence, not all cases of cervical cancer show HPV expression, suggesting that though HPV plays a pivotal role in cervical carcinogenesis, there must be many other co-factors involved as well. We observed that half of the cases belonging to small cell non-keratinizing group of squamous cell carcinoma were positive for HPV (50%). About 36% of large cell non-keratinizing squamous cell carcinoma and 14% of large cell keratinizing squamous cell carcinoma were positive for HPV.

In our study, 50% of the cases of high grade squamous intra-epithelial lesion (Ca-in-situ) were found to be positive for HPV by immunohistochemistry. This can be compared with the results of the study conducted by the Heopfner et al^[9] for immunohistochemical detection of HPV, in which he reported a positivity of 50% for cervical dysplasia and 69% for Ca-in-situ. Another study conducted by Gupta et al^[10] on paired papanicolaou-stained cervicovaginal smear and histologic sections of cases of genital condylomas (using immunoperoxidase methods) revealed 67% nuclear positivity in cervical smear and 62% nuclear positivity in tissue sections of same cases.

The higher positivity in pre-invasive lesions of cervix (50%) in comparison to diagnosed cases of carcinoma cervix (34%) could be explained by the fact that L ₁ gene is predominantly expressed in the nuclei of koilocytes, and its expression is limited to more differentiated cells. ^[6] So, it can be hypothesized that with progression from precancerous to cancerous state (i.e. loss of differentiation), there is relative decrease of L ₁ capsid protein expression.

In a case-control study of 436 women with cervical cancer and 387 control women enrolled in Columbia and Spain, Bosch et al, (1992) found that HPV DNA was the single strongest risk factor for the development of cervical cancer (odds ratio-23.8). ^[11] Bosch et al^[12] in an international study of over 1000 invasive cancers collected in 22 countries showed that HPV DNA was detectable in 93% cases using PCR-based method to examine frozen tissue specimens. HPV 16 was positive in 54% cases, HPV 18 in 15% cases, and HPV 31, 46 in 24% cases.

In a case-control study by Chichareon et al,^[13] from Thailand, the fraction of risk attributable to HPV in one population was 95% for invasive squamous cancers of cervix. In a study of colposcopically-negative high risk, women enrolled from STD clinics in New York city.

Fiedler et al,^[14] studied the expression of high risk HPV types 18 and 45 E ₇ oncoprotein in cervical carcinoma biopsies by immunohistochemistry. Anti-HPV 18 E ₇ protein polyclonal antibodies stained all tumor cells but not adjacent connective tissues. These findings suggest that high risk E ₇ oncoprotein of HPV 16, 18, and 45 are expressed continuously in cervical cancers and are linked in 80% of all cervical cancers worldwide.

Ravaozoula et al,^[8] studied immunohistochemical detection of HPV protein and c erb B2 receptors in cervical lesion specimens from females. 75 cases of CIN and invasive cancer were examined and stained. C erb B2 receptors were found to be positive in 11.7% cases. HPV protein was detected in 31% cases. Koilocytes were seen in 64% cases.

It can be thus hypothesized that with decreasing differentiation, there is loss of expression of HPV in the cervical epithelial cells. To conclude, though this study confirms the correlation of HPV presence with precancerous and cancerous lesions of uterine cervix, it also suggests that there must be other co-factors involved in cervical carcinogenesis as well.

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/1755-6783.102033

[Table 1], [Table 2]