Scoring System and Management Algorithm Assessing the Role of Survivin Expression in Predicting Progressivity of HPV Infections in Precancerous Cervical Lesions

Cervical cancer is the second most frequent cancer in women throughout the world, and contributes to most deaths caused by gynecological cancers. The agestandardized incidence rate of cervical cancer in Indonesia is 12.7/100.000 women per year and is the second most frequently encountered cancer in women following breast cancer (WHO, 2008). Cervical cancer is still a major issue in Indonesia as most (62%) of the patients present in an advanced stage. Although screening for lesions by the Papanicolau test (Pap test) has dramatically reduced the incidence of cervical cancer, this test has various limitations, such as low sensitivity for detecting cervical intraepithelial neoplasia 2-3. Other modalities are needed to enhance the accuracy of screening for cervical cancer (Boone, 2012). Human papilloma virus (HPV) is a well-known etiology of cervical cancer. Recent findings in the molecular carcinogenesis by HPV has expanded new areas


Introduction
Cervical cancer is the second most frequent cancer in women throughout the world, and contributes to most deaths caused by gynecological cancers. The agestandardized incidence rate of cervical cancer in Indonesia is 12.7/100.000 women per year and is the second most frequently encountered cancer in women following breast cancer (WHO, 2008).
Cervical cancer is still a major issue in Indonesia as most (62%) of the patients present in an advanced stage. Although screening for lesions by the Papanicolau test (Pap test) has dramatically reduced the incidence of cervical cancer, this test has various limitations, such as low sensitivity for detecting cervical intraepithelial neoplasia 2-3. Other modalities are needed to enhance the accuracy of screening for cervical cancer (Boone, 2012).
Human papilloma virus (HPV) is a well-known etiology of cervical cancer. Recent findings in the molecular carcinogenesis by HPV has expanded new areas 1 RESEARCH ARTICLE

Scoring System and Management Algorithm Assessing the Role of Survivin Expression in Predicting Progressivity of HPV Infections in Precancerous Cervical Lesions
Junita Indarti 1 *, M Farid Aziz 1 , Bethy Suryawati 2 , Darrell Fernando 3 of study, such as potential biomarkers. These biomarkers may be used to detect precancerous lesions, to enhance diagnostic sensitivity, to predict the prognosis, and/or to be considered when choosing the mode of therapy (Tan, 2010).
Survivin is an inhibitor of apoptosis expressed in the embryonic period; it is not expressed in differentiated adult tissues. Survivin is expressed during the G2/M phase of the cell cycle, and supressed when the cell cycle stops. Consequently, exfoliated epithelial cells will not express survivin. Malignancy is related to disruptions in the cell cycle and may therefore express survivin (Li, 2005). Another mechanism for survivin upregulation in cervical carcinogenesi is its normal transcriptional repression by wild-type p53 being eliminated by oncoprotein E6 in high-risk HPV (Branca, 2005).

Materials and Methods
This case-control study was conducted in Gynecologic Oncology Division, Department of Obstetrics and Gynecology, Cipto Mangunkusumo National Referral Hospiral, Jakarta, from October 2009 until May 2010. Samples were selected from women visiting the Colposcopy Clinic. Women eligible for this study were aged 18-50 years, or above 50 years without cytological findings of atrophy, and were not pregnant nor menstruating during the visit. Seventy-four women were included in this study and were divided into case and control groups. All samples agreed to participate in this study and an ethical clearance was issued by the Committee of Medical Research Ethics, Faculty of Medicine University of Indonesia. The case group was further classified as CIN 1, CIN 2, or CIN 3. We obtained 19 samples for CIN 1, 18 samples for CIN 2, 18 samples for CIN 3, and 19 samples as controls.
Polyclonal rabbit anti-survivin antibody (Abcam ® ab8228) was used for ICC and IHC staining. Stainings were conducted in Pathologic Anatomy Laboratorium, Hasan Sadikin Hospital, Bandung, and the slides were interpreted by us and a pathologist. Expression of survivin was measured with immunocytochemistry (ICC) and immunohistochemistry (IHC) staining. A sample is positive if the nucleus with/without cytoplasm is stained dark brown. Intensity is graded on a scale of 0-3: 0: no staining, 1: low intensity, 2: moderate intensity, and 3: high intensity. Distribution is the amount of stained cells. Expression is calculated by multiplying intensity and distribution. Expression of this marker was either low (ICC <70, IHC <110) or high (ICC ≥70, IHC ≥110).
Demographic profiles, risk factors, and test results were analyzed with Stata 9.2, using bivariate analysis and multivariate analysis. Data were also fit to a prediction model and probability scoring to obtain a scoring system.

Results
After data input, we determined a cut-off point for each variable by using the ROC curve. The cut-off was the point yielding maximum sensitivity and specificity. Cut-off points for each of the variables are: age <41 years, course of education ≥13 years, parity ≥2, amount of sexual partners ≥2, age at first sexual intercourse <22 years, ICC expression of survivin: ≥70, and IHC expression of survivin: ≥110.
Demographic profiles analyzed in this study were age, education, and parity. Risk factors studied were amount of sexual partners, age at first intercourse, use of oral contraceptives, smoking, and presence of sexually transmitted infections (STIs). Bivariate analysis of these factors (Table 1) revealed that age <41 years significantly increases the risk of developing CIN 1, 2, and 3. Women with more than two sexual partners, women who had their first sexual intercourse before 22 years, and women with positive STIs have statistically significant risks of developing CIN 3. Other factors were not statistically significant.
HPV DNA was examined with hybrid capture 2 (HC2) and results were classified according to the degree of CIN. Women infected with high-risk DNA had a significantly  may be subsequently managed with a follow-up Pap test after 4-6 months, HPV DNA examination, or colposcopy examination. The survivin biomarker may have a role in deciding whether to manage the patient conservatively (follow-up) or to directly perform a colposcopy.
In our study, women <41 years had significantly increased risks of developing all grades of CIN. Higher grades of CIN are found in accordance with increasing age. This finding is consistant with the natural history of precancerous cervical lesions, as CIN 1, 2, and 3 needs 5, 3, and 1 years to develop to carcinoma in situ (CIS).  found that age was significantly and inversely associated with HPV clearance. Therefore, women ≥41 years will more likely have CIS or cervical cancer, rather than precancerous lesions.
We found that course of formal education is not a significant factor for developing CIN, although the odds are increased in women with education ≥13 years. Other studies on course of education have been inconsistent. Castle (2008) found that CIN 1-3 and CIS were more frequently found in women with <12 years of formal education, owing to lack of compliance during therapy. Poomtavorn (2011) also supported these findings. On the other hand,  found that women with higher level of education were associated with high-risk HPV infections. Although we found no significancy in bivariate analysis, this factor was included in the multivariate analysis as the p value was <0.25, and it was included in the fit model.
Parity is not a significant factor in our study. This finding is consistent with Belinson (2008), who found no significant correlation between parity and CIN. On the contrary, Jensen (2013) found the risk of developing CIN 3+ in women with persistent HPV infection who had given birth was 1.78 (95%CI 1.07-2.94).  found parity was a significantly associated with impaired HPV clearance. Alterations in the immune system owing to hormonal changes during pregnancy may increase the susceptibility of HPV infections and/or malignant changes (American Cancer Society, 2009).
We found that women with ≥2 sexual partners have a significantly increased risk of developing CIN 3. Our finding was consistant with de Boer (2006), who discovered the OR of cervical cancer in women with >1 sexual partner was 5.83 (95%CI 2.98-11.36). Almonte (2011) found the age-adjusted odds ratio in women with ≥5 sexual partners was 2.1 (95%CI 1.4-3.2). The risk is also increased if the woman had sexual intercouse with a male with multiple partners. This sexual behaviour may introduce infections with different types of HPV; high-risk HPV increases the risk of dysplasia.
Bivariate analysis revealed that women who have had their first sexual intercourse before 22 years have a significantly increased risk of developing CIN 3. Ruiz (2012) found the odds for high-grade lesions was 3.55 higher in women who had a short interval (<3 years) between menarche and first sexual intercourse. Almonte (2011) found the OR in women who had their first sexual intercourse before 18 years compared to ≥20 years was 1.5 (95%CI 1.2-2.0). Infection with high-risk HPV in the maturation period of the genitalia may induce atypical  A high survivin expression significantly increases the risk of developing more severe grades of CIN ( Table 2).
All of the factors with p value less than 0.25 were included into the multivariate analysis. A stepwise backward elimination was performed, factors were excluded if the exclusion did not significantly change the maximum likelihood ratio. Factors from the full model were then fit into a prediction model (fit model). A score was determined by dividing the coefficient by SE coefficient (Table 3).
The probability of developing CIN by survivin ICC examination is shown in Table 4.

Discussion
It is a common consensus that women with abnormal Pap smears should undergo colposcopy exams followed transformations.
We found no significant risk of developing CIN in women using oral contraceptives. In a systematic review by Gadducci (2011), oral contraceptive consumption increased the relative risk of cervical cancer, and the risk decreased after pill discontinuation. Nevertheless, WHO does not recommend any changes in oral contraceptive practice as the benefit outweighs the risk. Hellberg (2005) stated the immunosupressive nature of progesterone correlated with HPV infections. Progesterone enhances HPV mRNA and stimulates viral replication. In the transformation zone, estradiol is converted to estrone, which is associated with malignant transformation in estrogen-sensitive cells. Although we found no significancy in bivariate analysis, this factor was included in the multivariate analysis as the p value was <0.25, and it was included in the fit model.
Various studies have reported the role of sexually transmitted infections in increasing the risk of CIN. Roeters (2010) found the OR for Chlamydia and Gardnerella in high-grade smears are 7 and 12, respectively, and the OR for Trichomonas was also significantly high. Engberts (2007) reported the OR of Candida is 1.85 for LSIL and 2.0 for HSIL. We found that STIs significantly increases the risk of developing CIN 3. By-products of these pathogens uch as propionate and butyrate may damage epithelial cells, and alterations in the vaginal environment to a proinflammatory state may contribute to the development of ervical lesions (Gillet, 2012).
Infection by high-risk HPV significantly increases the risk of higher grades of CIN. We found high-risk HPV DNA in 21.21% of CIN 1 (OR: 6.73), 56.1% of CIN 2 (OR: 31.04) and 70% of CIN 3 (OR: 50). Our findings are similar to previous studies. However, the amount of women infected with high-risk HPV in our study are lower than other similar researches. Belinson (2008) found highrisk HPV DNA in 85.2% of CIN 1, 96.5% of CIN 2, and 97.4% of CIN 3. This discrepancy may be due to different reagens used; our study used HC2 reagen detecting 13 types of high-risk HPV.
Bivariate and multivariate analysis proved the significance of survivin expression in CIN; the odds are higher in higher grade lesions. Similar findings were reported by Barbosa (2011) andBranca (2008), in which survivin expression was reported to be increased in accordance with tumor progressivity. Therefore, higher grade lesions have higher levels of survivin expression.
Oncoprotein E6 in high-risk HPVs has a significant role in the regulation of survivin-gene transcription. P53 supresses the expression of survivin, thus in conditions with loss of p53 function, such as cancer, expression of survivin is enhanced. HPV E6 induces the activity of survivin promoter region and increases endogenous survivin mRNA in human embryonic fibroblasts (Borbely, 2006;Mita, 2008) We propose a combined scoring system consisting of survivin expression by ICC and factors proved significant from multivariate analysis (Table 5). A biopsy will first be performed on the subject, classified to CIN 1, CIN 2, or CIN 3. Subjects will then be assessed for risk factors and examined for tissue biopsy and survivin expression by ICC staining. The score will then be interpreted by using probability of lesion progressivity. The score will then be interpreted by using probability of lesion progressivity. For example, a patient with CIN 1 and total score of 42, has a 2% chance of regressing, 57% chance of persisting in CIN 1, and 41% chance of progressing to CIN 2-3. This scoring system may help determine whether the patient   Evaluation precancerous lesions by measuring survivin expression can be used to determine whether the patient needs active or conservative management. The algorithm is found in Figure 1. Patients with LSIL on their Pap test results, or patients with positive acetowhite lesions on visual inspection with acetic acid, should be evaluated for survivin expression. If the results are low, the patient should return for subsequent Pap tests after 4-6 months. If the results are high, the patient should be examined with colposcopy. This evaluation helps reduce the need for colposcopy, because in Indonesia colposcopy is not readily available with very few experts centralized in major cities. Survivin expression is easier to perform, as speciments can be shipped to laboratories and the price is cheaper than colposcopy.
In conclusion, we found the significant risk factors for developing precancerous cervical lesions are (1) age <41 years, (2) use of oral contraceptives, (3) ≥2 sexual partners, (4) use of oral contraceptives, (5) positive high-risk HPV DNA, and (6) high survivin expression by ICC or IHC staining. These risk factors can be combined to create a scoring system. Detection of survivin expression by ICC staining may be incorporated in a scoring system and management algorithm of precancerous cervical lesions. Application of this scoring system and algorithm may be beneficial in determining the action of therapy towards the patient.