Tanzania Journal of Health Research Health User's Trust Fund (HRUTF) ISSN: 1821-6404 Vol. 11, Num. 4, 2009, pp. 163-169

Tanzania Journal of Health Research, Vol. 11, No. 4, October, 2009, pp. 163-169

Prevalence and severity of cervical squamous intraepithelial lesion in a tertiary hospital in northern Tanzania

J. Obure¹*, O. Olola¹, B. Swai¹, P. Mlay¹, G. Masenga¹ And D. Walmer²

¹Kilimanjaro Christian Medical Centre, P. O. Box 3010, Moshi, Tanzania
²Duke University Medical Centre, Durham NC 27713, USA

Received March 3, 2009
Revised 10 September 2009
Accepted 12 September 2009 

Code Number: th09029

Abstract

Cervical cancer is the second most common cancer in women worldwide and the leading cause of cancer deaths in Tanzanian women.  Prevention of cervical cancer relies on the detection and treatment of Squamous Intraepithelial Lesion (SIL), a premalignant disease stage. Worldwide there are overwhelming reports associating SIL and HIV infection, however in Tanzania such reports are limited. A cross-sectional hospital-based descriptive study was conducted to determine the prevalence and severity of SIL in 234 HIV seropositive and seronegative women aged 18-68 years old at Kilimanjaro Christian Medical Centre in northern Tanzania. A structured questionnaire was used to collect sociodemographic data. In addition, blood was collected for rapid HIV antibody testing and CD4+ T-lymphocyte counts to associate with prevalence and severity of SIL from the cervical smear collections. A total of 214 subjects had smear results regarded as valid for interpretation, of which 46.3% were HIV seropositive. Overall rate of SIL was 17%. Proportion of SIL among HIV seropositive subjects was 32% versus 4% in seronegative subjects (OR=13.3, 95% CI=4.2-46.4). Low CD4+ T-lymphocyte cell count was associated with higher prevalence of SIL (P=0.001). The relationship between CD4+ T-lymphocyte cell counts and the severity of cervical SIL was significant (P=0.007). Marital status and number of lifetime sex partners were risk factors significantly associated with SIL (P=0.004 and 0.005, respectively). SIL was not associated with age, education level, parity or age at sex debut. The prevalence and severity of cervical SIL was significantly associated with HIV infection and immunologic disease progression. These findings underscore the need for HIV screening among women with SIL, and the need for cervical cancer screening in HIV-infected women. Marital status and number of lifetime sex partners were significant risk factors associated with SIL.

Key words: HIV, cervical cancer, squamous intraepithelial, lesion, risk factors, Tanzania

Introduction

Cervical cancer remains the second most common cancer in women worldwide with 493,000 new diagnoses and 274,000 deaths occurring each year. Eighty percent of these deaths occur among women from developing countries(Parkin et al., 2002). In Tanzania the annual crude incidence rate of cervical cancer is 40.6 while crude mortality rate is 32.5 (WHO, 2009).   In resource-rich countries, screening and treatment of pre-malignant cervical lesions is done on a regular basis, resulting in an 80% reduction in cervical cancer incidence and related mortality.  In contrast, in resource-limited countries 60-80% of cases are diagnosed at an advanced stage of the disease with an accompanying poor prognosis(Parkin et al., 2002; Sankaranayanan et al., 2001).

Persistent infections with sexually transmitted oncogenic strains of human papilloma virus (HPV) such as types 16, 18 and 45 are central aetiological factors for the development of high grade squamous intraepithelial lesions (SIL) and cervical cancer worldwide(Bosch et al., 1995; Franco et al., 2001). HIV seropositive women are at higher risk for SIL and cervical cancer compared to their seronegative counterparts(Palefsky et al., 2001). Such factors account for higher proportions of preinvasive cervical neoplasia in HIV-infected than HIV-uninfected women worldwide(La Ruche et al., 1998; Mandelblatt et al., 1999; Massad 1999).  Unlike other parts of the Africa, few study reports on the association between HIV infection and cervical SIL have been published in Tanzania(Mandelblatt et al., 1999; Moodley et al., 2006; Parham et al., 2006). In West Africa, a study reported that high grade SIL was strongly associated with HIV-2 infection and not HIV-1 (Hawes et al., 2003). In East Africa, one study in Tanzania reported no association between HIV infection andSIL (ter Muelen et al., 1992); while another study investigating risk factors for SIL in HIV-infected women did not include HIV seronegative women as controls(Kapiga et al., 1999).

Due to the contradictory results reported on the relationship between SIL and HIV serostatus in the previous Tanzanian studies(ter Meulen et al., 1992; Kapiga et al., 1999) and its impact on HIV/AIDS-related morbidity, this study was therefore conducted to re-examining this relationship. Specifically this study was carried out to determine the prevalence and severity of SIL and predisposing factors in HIV-infected women in relation to the degree of HIV disease progression. 

Materials and Methods 

Study area

This study was conducted at Kilimanjaro Christian Medical Centre (KCMC) in Moshi Tanzania. KCMC is one of the four teaching and referral hospitals in Tanzania. The hospital has 450 bed capacities and caters to residents of northern parts of Tanzania. Department of Obstetrics and Gynaecology runs Cervical Cancer Clinic (CCC) thrice a week. Most of the clients attending this clinic are women who come voluntarily to seek screening services and few referred patients from various other KCMC in- and out-patient departments. 

Study subjects

A total of 234 women aged 18-68 years, with no prior history of total hysterectomy and willing to take HIV test and participate in the study were conveniently enrolled in the study from September 2006 to March 2007. Of these, 85 were women known to be HIV seropositive referred from KCMC Infectious Diseases Clinic (IDC) and 149 were those with unknown HIV serostatus who came for Pap smear test at the KCMC-CCC.

Study design and data collection

This was a cross-sectional prospective descriptive study. A Kiswahili structured questionnaire was administered to study participants to collect information on socio-demographic characteristics, risk factors for SIL and cervical cancer. After HIV pre-test counselling by a trained nurse to women with unknown HIV serostatus, 10ml of venous blood was collected to determine HIV antibody status using both Capillus and Determine HIV1/2 test kits. Samples from HIV seropositive women were sent for CD4+ T-Lymphocyte count measurement using a FACS Count Flow Cytometer (Becton Dickinson Biosciences). Cervical smears were collected using both wooden Ayres spatula and endocervical brush techniques from all subjects. The cells were transferred to the dry glass slide, fixed using 50% alcohol, and sent to the KCMC Pathology Department for staining and interpretation.  These procedures were performed by a trained cytotechnician, and interpretations were performed by a single Consultant Pathologist who was blinded to the subjects’ HIV serostatus. Pap smear results were graded according to Bethesda system of cervical cytology interpretation (Solomon et al., 2002); with the categories of normal smear, low grade and high grade SIL.

Data analysis

Data collected was entered and analysed using Statistical Package for Social Science (SPSS) version 12. Numerical variables were summarized into tables and charts while categorical variables were compared among women with or without outcome of interest using Chi-square test at 95% confidence level. Two sided p-value was used to assess level of significance. 

Ethical consideration

Ethical clearance was obtained from KCMC Research Ethics Committee before beginning of the study. The purpose of the study, procedures to be taken and any untoward effects were thoroughly explained to the subjects whom were only enrolled post signing informed consent forms. Women who opted not to participate in the study received the equivalent care to study participants according the KCMC therapy guidelines. To insure confidentiality hospital registration number and unique study number were used instead of names while study documents and data were kept in the locker accessible to research team only. Pass word known to research team was used to access all research electronic data. 

Results 

Socio-demographic characteristics

The mean age ± SD of 234 women recruited in the study was 35±9 years. About three quarters (73.1%) were married/co-habiting and only 4.7% divorced. Majority of the women, 72.2% had primary school education and only 2.6% had no formal education. Of 234 women, 65.4% were para 1 to 3 and only 2.1% were nulliparous; mean parity of women was 3 (range, 2-10). More than 45% had their sex debut at the age between 15 and 18 years; mean age at sex debut was 19±3.7 years. Most women (90.6%) had 1 to 3 lifetime sexual partners with only 9.4% having 4 or more partners; mean number of lifetime sexual partners was 2 (range, 1-10). Majority of women, 67.5% mentioned that their partners had only one sexual partner while 2.6% had 4 or more sexual partners (Table 1).  

Table 1: Social demographic characteristics of women enrolled in the study (N= 234)

Of the 234 women, 103 (44%) were HIV seropositive. The median CD4+ T-lymphocyte count for HIV seropositive women was 304 cells/µL (IQR, 161-423). Age, marital status, age at sex debut, their number of lifetime sexual partners and the number of lifetime sexual partners of their male partners were significantly related to HIV serostatus.  The chi-squared value for trend analysis showed that women aged ≥ 35 years, age at sex debut less than 18 years, widowed and multiple lifetime sexual partners were the characteristics associated with being HIV seropositive. Married women and women with age at sex debut more than 18 years were less likely to be HIV-infected (Table 2).

Table 2: Demographic characteristics and HIV serostatus (n, 234)

Prevalence of SIL in HIV-infected and HIV-uninfected women

Among 234 women, 214 (91.5%) had Pap smear results available for interpretation. Of these 214 subjects, 99 were HIV seropositive (46.3%). Overall 36 women had SIL, of which 32 were HIV seropositive (88.9%). The proportion of SIL among HIV seropositive women was 32.3% (32/99) and 3.5% (4/115) for seronegative women (OR= 13.3; 95% CI 4.2-46.4). Pap smear results according to Bethesda classification indicate that 178 (83%) of the women were normal. Women with low and high grade lesion were 21 (10%) and 15 (7%), respectively. 

Table 3: Relationship between rate of SIL and degree of HIV progression according to CD4+ T- lymphocyte count (cells/μL)

Relationship between prevalence and severity of SIL and HIV progression

The prevalence of SIL was higher (58.1%) in subjects with a CD4+T-lymphocyte count below 200 cells/µL compared to 22.4% with 200-499 cells/ µL and 15.8% in subjects with counts above 500cells/µL. This difference was highly significant (P=0.001) (Table 3). In HIV seropositive women with SIL, 37.5% (12/32) had high grade SIL, and 58.3% (7/12) of these women with high grade SIL had CD4+ T lymphocyte counts below 200cells/µL. In comparison, only 8.3% (1/12) of HIV-infected women with CD4+ T-lymphocyte counts above 500 cells/µL had high grade SIL (P=0.007) (Table 4).

Table 4: Relationship between degree of SIL and degree of HIV progression (N=99) 

Note: PAP= Papanicolous; HGL= High grade squamous intraepithelial lesion; LGL= Low grade squamous intraepithelial lesion

Association between socio-demographic characteristicsand rate of SIL

Among 214 women with valid Pap smear results, marital status and number of life time sexual partners were significant risk factors for the development of SIL (marital status- P=0.004; number of life partners- P=0.0049). Divorced women had more SIL 36.4% (4/11), followed by single women 31.8% (7/22), widowed 29.6% (8/27) and married/co-habiting group 11% (17/154) (P=0.004). Women with multiple life time sexual partners had higher rate of SIL (23.1% vs. 8.6%) than women who had one life sexual partner (P=0.0049).  There was no significant association between development of cervical SIL and age, education level, parity and age at sex debut (Table 5). 

Table 5: Relationship between various social demographic characteristics and rate of cervical dysplasia (N=214)

Discussion

Overall rate of SIL by a single Pap smear test in this study population was 17%. However most of subjects with SIL (88.9%) were significantly detected among HIV seropositive women. These results are consistent with other study reports elsewhere (Massad 1999; Moodley et al., 2006; Parham et al., 2006), however they differ from those reported in the study done in Dar es Salaam, Tanzania which showed that HIV serostatus was not significantly associated with SIL(Meulen et al 1992).This could have happened due to a number of reasons. The difference in the mean age in these two study populations, with the previous study having a lower mean age of HIV seropositive women (27.1 vs. 38.6 years). This may be related to the duration needed for development of SIL from the day of HPV and HIV infections. Enrolling women admitted to the ward due to different reasons such as pelvic inflammatory diseases and infertility, the group which may have a higher prevalence of sexually transmitted infections and cervicitis leading to the under detection of SIL from vagina discharges. In contrary, our study population involved women coming for a baseline Pap smear from IDC and the community without any gynaecological complaints. While the study by Meulen and colleagues included mostly asymptomatic HIV cases, our study enrolled mostly symptomatic HIV-infected women with a median CD4+ T-lymphocyte count of 319 cells/µL, which is likely to have increased the likelihood of SIL.

Low CD4+ T lymphocyte cell counts significantly increased the likelihood of having cervical SIL, with most of the lesions seen at CD4+ T Lymphocyte cell counts below 200cells/µL. This is in agreement with most reports on rate of cervical SIL and HIV infection(Wright et al., 1994; Massad, 1999), possibly due to reactivation and persistent infection with the higher risky HPV such as type 16 and 18 (Wright et al., 1994; Palefsky et al., 2001). The significant statistical difference between the degree of immunosupression  (by CD4+ T-lymphocyte cell counts) and the degree of cervical SIL shows that the more the immunity is suppressed the higher the likelihood of having high grade cervical SIL. The results are consistent with those reported in other studies(Hawes et al., 2003; Parham et al., 2006).

Cervical SIL was significantly more likely to be observed in single women (not married, divorced, widowed) than women with permanent partners (married/cohabiting). Single women were likely to have multiple partners, a factor which was also seen to be significantly associated with SIL. This group was also seen to have a higher rate of HIV infection compared to married/cohabiting women). Multiple partners and HIV infection are factors highly associated with higher rates HPV infections, persistence and progression to SIL (Palefsky et al., 2001; Massad, 1999). A similar trend, however insignificant was seen with single/widowed developing HGL compared to married women in a West African study (La Ruche et al., 1998). Despite the significant association between SIL and number of life time partners noted in our study, a report in Dar es Salaam, Tanzania study could not find this association (Kapiga et al., 1999). It is possible that, the difference in study populations accounted for this as the Dar es Salaam study dealt only with HIV seropositive women whom might have more or less same sex characteristics compared to this study with both HIV seropositive andseronegative women.  Like in the study in Dar es Salaam (Kapiga et al., 1999), age was not significantly associated with cervical SIL.  However, women within 36-49 years had a higher proportion of SIL (24.4%), in comparison to other age groups. In contrast, a study in USA (Wright et al., 1994) reported cervical SIL to be significantly associated with age, and mainly beyond 35 years. The difference could have been attributed by the smaller sample size in our study population. As with the other studies in Tanzania, education level and age at sex debut was not associated with SIL (Kapiga et al., 1999).

In conclusion SIL is highly prevalent in symptomatic HIV infected women in northern Tanzania and therefore it is of critical importance to initiate regular cervical cancer screening programme services to at risk women in the population and especially HIV seropositive women attending centres for treatment and care in order to alleviate preventable causes of HIV/AIDS associated precancerous and cancer morbidities.

Acknowledgements 

We thank Prof John Bartlett of Duke University (USA)-KCMC collaboration for funding of this study. Dr John Crump and colleagues from KCMC Department of Obstetrics and Gynaecology are thanked for their comments on the earlier version of the manuscript. Mr. Gibson Kapanda is acknowledged for his statistical assistance.  

Declaration of conflict of interest 

The authors hereby declare to have no any competing interest.

References

• Bosch, X.F., Michelle, M., Nubia, M., Sherman, M., Jansen, A.M., Peto, J., Schiffman, H.M., Moreno, V., Kurman, R. & Shan, K.V. (1995) Prevalence of HPV in cervical cancer; a world wide perspective. Journal of the National Cancer Institute 87, 796-802.
• Franco, E.L., Franco, E.D. & Ferenczy, A. ( 2001) Cervical cancer epidemiology, prevention and the role of HPV infection. Canadian Medical Association Journal 164, 1017.
• Hawes, S., Critchlow, W., Niang, M., Diouf, M., Diop, A., Toure P., Aziz Kasse, A., Dembele, B., Salif Sow, P., Coll-Seck, A.M., Kuvpers, J.M. & Kiviat, N.B. (2003) Increased risk of high grade cervical squamous intraepithelial lesions and invasive cervical cancer among African women with HIV 1 and 2 infections. Journal of Infectious Diseases 188, 555-563.
• Kapiga, S.H., Msamanga. G.I., Spiegelman, D., Mwakyoma, H., Fawzi, W.W. & Hunter, D.J.  (1997) Risk factors for cervical squamous intraepithelial lesions among HIV-1 seropositive women in Dar es Salaam; Tanzania. International Journal of Gynecology & Obstetric 67, 87-94.
• La Ruche, G., Ramon, R., Mensah, I.A., Bergeron, C., Diomande, M., Sylla-K.F., Ehouman, A., Toure, K.C., Welffens, C.E. & Dabis, F. (1998) Squamous intraepithelial lesions of the cervix, invasive cervical carcinoma, and immunosupression induced by HIV in Africa. Journal of Cancer 15, 2401-2408.
• Mandelblatt, J.S., Kanetsky, P., Eggert, L. & Gold, K. (1999 Is HIV infection a cofactor for cervical squamous cell neoplasia? Cancer Epidemiology, Biomarkers & Prevention 8, 97-106.
• Massad, L.S. (1999) Prevalence and Predictors of Squamous cell abnormalities in HIV infected women. Journal of AIDS 21, 33-41.
• Moodley, R.J., Hoffman, M,, Carrara, H., Allan, B.R., Cooper, D.D., Rosenberg, L., Denny, L.E., Shapiro, S. & Williamson, L.A. (2006) HIV and pre-neoplastic and neoplastic lesions of the cervix in South Africa.  BMC Cancer 6, 135.
• Palefsky, J.M., Holly, E.A. Ralstom, M.L., Da Costa, M. & Greenblalt, M.R.  (2001) Prevalence and risk factors for anal HPV infection in HIV positive and high risk HIV negative women. Journal of Infectious Diseases 183, 383-391.
• Parham, G.P., Sahasrabuddhe, V.V., Mwanahamutu, M.H., Shepherd, B.E., Hicks, M.L. & Stringer, E.M. (2006) Prevalence and predictors of squamous intraepithelial lesions of the cervix in HIV infected women in Lusaka, Zambia. Journal of Gynecology Oncology 103, 1017-1022.
• Parkin, M.D., Bray, F., Farley, J. & Pasani, P. (2005) Global cancer statistics. CA: Cancer Journal for Clinicians 55, 74-108.
• Sankaranarayanan, R., Madhukar, A.B. & Rajkumar, R. (2001) Effectiveness of screening programmes for cervical cancer in low and middle income developing countries; Bulletin of the World Health Organization 79, 954-962.
• Solomon, D., Davey, D., Kurman, R., Moriaty, A., O’Connor, D., Prey, M., Raab, S., Sherman, M., Wilbur, D., Wright, T. & Young, N.  (2002) The 2001 Bethesda System: terminology for reporting results of cervical cytological abnormalities. Journal of the American Medical Association 287, 2114-2119.
• Meulen, J., Eberhardt, H.C., Luande, J., Mgaya, H.N., Chang-Claude, J., Mtiro, H., Mhina, M., Kashaija, P., Ockert, S., Yu, X., Meinhardt, G., Gissmann, L. & Pawlita, M. (1992) Human papillomavirus (HPV) infection and cervical cancer in Tanzania, East Africa. International Journal of Cancer 51, 515-521.
• WHO (2009) Information Centre on HPV and Cervical Cancer. World Health Organization, Geneva.
• Wright, T.C., Ellerbrock, T.V., Chiasson, A.M., Devarter, N.V. & Sun, X.W. (1994) Cervical intraepithelial neoplasia in women infected with HIV: Prevalence, risk factors, and validity of Papanicolous smear; American Journal of Obstetrics and Gynaecology 84, 591-597.

Copyright 2009 - Tanzania Journal of Health Research