Cervical intraepithelial  neoplasia  is  still  a  public  health problem among women in less developed countries in the world (Banura et al., 2008; Ferlay et al., 2015). In the first world, like the United States, and Scandinavian countries, the decline in incidence and mortality has been well registered (Parkin et al., 2002, 2003; Forman et al., 2013; Ferlay et al., 2015), but this is not yet the case for developing nations, Sub-Saharan Africa. This is in part due to lack of information about the disease, and lack of screening services for detection of preinvasive and early invasive disease (de Oliveira et al., 2013; Bingham et al., 2003; Banura et al., 2008). Because of lack of efficient prevention and early detection services to complement treatment of the malignancy (Forman et al., 2013; Ferlay et al., 2015), over 200,000 deaths are reported annually in both developing and developed countries in the world (Ferlay et al., 2015). Estimates on cervical cancer incidence in 2012 showed that, Uganda had the highest age-standardized incidence rate in the world of 44.4/100,000, while East Africa had age-standardized rate (ASR) of 42.7/100,000, and the World had ASR of 14.0/100,000.It was further reported by WHO in 2014, that 3915 women in Uganda were positive for cervical cancer and that 2160 had died of the disease (ICO, 2016). Cervical cancer was documentedas the main cause of mortality among economically disadvantaged women of reproductive age in Uganda (Mutyaba et al., 2007).

Various agents in association with the cause of cervical intraepithelial neoplasia and cervical cancer development have been outlined. These include age at first intercourse (Camargo et al., 2011), circumcision (Cook et al.,1994; Deacon et al., 2000), history of sexually transmitted infections (STI) (Szaba and Short, 2000; Mosciki et al., 2001; Muñoz et al., 2004; Tobian et al., 2009), use of oral contraceptives (Muñoz et al., 2004; Baldwin and Jensen et al., 2013), infection with HIV (Stanley, 2010; Denny et al., 2012), and multiple sexual partners (Walker et al., 2002). The commonest risk factor is infection with human papillomavirus (HPV) (Banura et al.,2008).It has been estimated that, 33.6% of women in the Uganda are infected with human papillomavirus, and that 47.5 per 100,000 women every year develop cervical cancer (Bruni et al., 2014; Nakisige et al., 2017).Previous reports by Uganda Cancer registry from2003-2007 showed that, the prevalence of cervical cancer by histological type was as follows: squamous carcinoma, 22.7%, and Adeno carcinoma, 2.3%.Currently, there is no detailed data that outlines the burden of cervical intraepithelial neoplasia and associated factors among women in Eastern Uganda. Cervical cancer related studies that were previously conducted involved selected groups, and small numbers (Mukama et al., 2017), not sufficient enough to provide   data   on  the burden  of  cervical  intraepithelial neoplasia and its clinical presentations in the general population in Uganda. Yet this information is crucial and important to support the implementation of HPV vaccination and cervical cancer screening programs among women in Uganda (Ferlay et al., 2015). Because HPV vaccines do not protect against all types of HPV that cause cervical cancer, vaccinated persons may become infected with HPV types against which vaccines do not offer protection. For this reason, vaccinated persons should be encouraged to be screened for cervical cancer later in life according to national policies. Screening should continue to prevent cancer in both non-vaccinated women and vaccinated women to provide protection from oncogenic HPV types for which HPV vaccines do not offer protection (WHO, 2014). The current study was designed to serve as a national baseline data to support the implementation of HPV vaccination and to help in planning for cervical cancer screening programs in Uganda by estimating the burden of CIN, and associated factors among the studied population in Eastern Uganda, Uganda. Our study adopted the Standard World Health Organization cytologic nomenclature for condyloma and cervical intraepithelial neoplasia (CIN) grades 1-3 and Bethesda system for rating cytology using the nomenclature LSIL and HSIL (Mark et al., 2018; Canepa et al., 2019). According to the CIN histological classification system, lesions were graded primarily according to the proportion of the epithelium which was occupied by basaloid, undifferentiated cells (Mark et al., 2018; Canepa et al., 2019). CIN1or CIN grade 1 was when the upper two thirds of the epithelium showed good differentiation, although some nuclear abnormalities persisted up to the surface. CIN2 or CIN grade 2 was whenthe upper half of the epithelium showed differentiation and maturation, again when nuclear abnormalities (atypia) persisted to the surface. CIN3 or CIN grade 3 was whenthe maturation was confined to the superficial third of the epithelium, or nuclear abnormalities were marked throughout the whole thickness of the epithelium.

 

Study area

This study was conducted at selected regional referral hospitals in Eastern Uganda. These hospitals include; Tororo, Mbale, Butalejja, Kumi, and Soroti referral hospitals. Availability of established cancer screening facilities, trained oncologists and technologist in cancer diagnosis and treatment  for both inpatients and outpatients within these hospitals, have turned them into referral centers for cervical cancer screening  for the neighboring districts that lack these facilities.    Districts   like  Moroto,  Katakwi,  Amuria,  Kaberamaido, Ngora, Serere, Namalu, and Nakapiripirit with inadequate cancer screening units and oncologists entirely rely on these selected referral hospitals for all their diagnostic and treatment programs (Figure 1).

 

 

Study design

This is a longitudinal study, with follow-up until there was diagnosis of high-grade cervical intraepithelial neoplasia (CIN). A total of 1,077 women aged 15-55 years and seeking health services at the cancer screening health facilities of Mbale, Tororo, and Butalejja and Soroti referral hospitals in Eastern Uganda were invited to participate in the screening project after qualifying for inclusion criteria. The study protocol  adopted  in  this  study  was  with  minor modifications of the method as previously described by Collins (2010) and all results were reported by using the Bethesda system for reporting cervical cytology and WHO dysplasia reporting system (Mark et al., 2018; Nayar et al., 2015; Singhal et al., 2018; Canepa et al., 2019). Follow-up of the study participants was in two main phases. Firstly, in the absence of cervical cytological abnormality, allow men were asked to return after six months for three consecutive negative smears before they were allowed to leave study, while women with inadequate smear were immediately requested to return for a repeat smear; secondly, in the presence of cervical cytological abnormality, all women who were referred for immediate colposcopy and histological examination. This was maintained for six months until there was no evidence of that abnormality. This was defined as three consecutive negative colposcopic  assessments and three consecutive negative  smears.  

If another cervical cytological abnormality was detected; the same procedure would be applied. Finally, at the end of the follow-up, the study protocol called for treatment to be postponed in women detected with cervical cytological abnormality until there was histological evidence of CIN3, and then recalled later for treatment, and thereafter left the study. Factors associated with cervical Intraepithelial neoplasia were obtained by use of questionnaires. These factors included the study participants’ age at first intercourse, being circumcised, and history of sexually transmitted infections, usage of oral contraceptives, infection with HIV, and their sexual behavior if they had multiple sexual partners. After meeting with the patients, written informed consent was obtained and a brief oral questionnaire was administered.

Retention and follow-up of the studied participants

A total of 1,077 women aged 15-55 years and seeking health services at the cancer screening health facilities of Mbale, Tororo, and Butalejja and Soroti referral hospitals in Eastern Uganda were invited to participate in the screening project after qualifying for inclusion criteria. Follow up schedules were in months. Between June 2017- August 2017, 416/1,077(38.6%) women  were  detected with abnormal cytology, of which 394/1,077(36.6%) had HSIL, and 22/1077(2.0%) had LSIL. All were subjected for immediate colposcopy and histological assessment between September, 2017-Oct 2017; their histological assessment showed that, 180/1,077(16.7%) had moderate CIN2, and 214/1,077(19.9%) severe CIN3, while 22(2.0%) participants had mild CIN1.Between November 2017-December2017, 394 (36.6%) participants who had developed CIN2/CIN3 neoplasia were treated, and the participants who did not respond to treatment were asked to return for further review (Figure 2).

 

 

Inclusion and exclusion criteria

The study included all women aged between 15-55 years, who resided and attended Healthcare services at the hospitals in Eastern Uganda. The study also included all women who consented and were willing to comply with the stipulated follow-up schedule. The study excluded all women who were <15 and ≥ 55years. Also women in menstruation cycle were excluded because blood discharge during the menstrual cycle would interfere with visual activity of the cervix and would give false positive results. The study also excluded all women aged 15-55 years who failed to  give their consent for the study

Sample collections from patients

Cytological samples

Cytological samples were collected by minor modifications of the method as previously described by Collins (2010). Two cytological samples were collected at each visit, using independent Ayres’s wooden spatula for each patient. The first sample was obtained by performing a 360º rotation of the transformation zone; the specimen collected was then applied to a slideto prepare a cervical smear for immediate cytological evaluation and reporting. The second sample was taken by performing a 360º rotation of the transformation zone as in the first sample collection. The spatula end was then broken-off into a 15-mL holding tubes containing 5 mL of PBS (pH 7.2), and then stored for future virological examination. Samples were kept temporarily at 4°C for an average of 6 h and then stored in a freezer at -20°C for further studies. All cervical cytological materials were examined and reported by our laboratory cytologist using the WHO dysplasia reporting system (Mark et al., 2018; Nayar et al., 2015; Singhal et al., 2018; Canepa et al., 2019).

Histological samples

Histological samples were collected by the method previously described by Collins (2010). Two histological specimens were taken. The first specimen was taken upon colposcopic evidence of abnormal epithelium by performing a punch biopsy for immediate histological examination; and the second histological specimen was taken from an area of abnormal epithelium, and stored for further histological studies. All cervical histological samples were examined and reported by our laboratory pathologist using original CIN terminology (Nayar et al., 2015; Canepa et al., 2019; Barut et al., 2015).

Quality assurance/quality control

There was strict monitoring of the quality and quantity of specimens, turnaround time, sample preparation and testing, decontamination, reagents and equipment, and reviewing test results and control using standard laboratory operation procedures of microbiologyby Cheesbroughet (2006).

Sample destruction plan

National Standard Operating Procedures (SOP) on health care waste management (Sapkota et al., 2014; Bessesen et al., 2015) was adopted and used for sample destruction and management after the study. All materials used in the assay, including reagents and specimens, were disposed of in a manner that would inactivate infectious agents. Solid wastes were autoclaved. Liquid wastes were treated with sodium hypochlorite to a final concentration of 1.0% (1:5 dilution of household bleach) and allowed 30 min for decontamination before disposal.

Ethical consideration

The ethical approval of the study was sought from Kampala International University (KIU), Mbarara University of Science and Technology (MUST) Institutional Research and Ethics Committee (IREC) on Human Research (Approval No. 06/01-17) and Uganda National   Council   for Science   and   Technology   (Approval   No. HS2246). All research protocols were performed in accordance with the ethical standards of the committees on human experimentation laid down in the Helsinki declaration of 1975 as revised in 2000. The participants were requested to sign the informed consent form before participating in the study. No participant was forced to participate in the study. Participants were coded instead of reflecting their real names. The participants were respected in relation to their right of their cultural beliefs and rights and were allowed to withdraw from the study without any condition.  All research protocols were performed in accordance with the ethical standards of committees on human experimentation laid down in the Helsinki declaration of 1964 revised in WMA (2000).

Data management and statistical analyses      

Data were analyzed by the use of descriptive statistics and regression using IBM SPSS version 20. Descriptive statistics was used to obtain socio-demographic characteristics of the study participants, histologic prevalence of cervical cancer and age-specific histologic prevalence of cervical cancer and disease stages during the screening period. The outcome of high grade cervical histological transformation was dichotomized as presence or absence of the cervical cancer and tested against suspected factors associated with high grade cervical histological transformation to assess for associations. Univariable analysis was used, and all the variables with a p ≤ 0.2 were entered into stepwise forward multiple logistic regression model. Interaction and confounding were assessed, and values of p ≤ 0.05 were regarded as statistically significant relationships.

 

 

 

 

Socio-demographic characteristics of participants

Study participants were classified in relation to the hospital type, age, educational level, and marital status. In relation to hospital types, Soroti Regional Hospital registered the least number of participants 126(11.7%), while Butalejja Hospital had 227(21.1%), Mbale Regional Hospital 302 (28.0%) and Tororo Hospital had the highest 422(39.2%).In relation to age, the participants with 45-55 years represented the lowest number,160(14.9%) compared with those having 25-35yearswith 344(31.9%) respectively. The median age of the study participants was 33 years with the median range of 15-55 years and interquartile range (IQR) = 14-36 years. In relation to the educational level, graduates had the lowest number 129(12.0%) compared to primary scholars with the highest 415(38%). In relation to marital status, the widowed participants were the least registered 118(11.0%), followed by divorced participants183 (17.0%), then singles 326(30.3%); the highest were married 450 (41.8%) (Table 1).

 

 

Histologic prevalence of cervical intraepithelial neoplasia among women with abnormal cytology from the study hospitals in Eastern Uganda, during 2017-2018 screening period

Four hundred and sixteen  women  were  diagnosed  with abnormalcytology; 22/416(5.3%) hadLSIL, and 394/416 (94.7%) had HSIL. Histological assessment was immediately performed and a punch biopsy was removed for histological examination. The results showed that 22/416(5.3%) had CIN1, 180/416(43.3%) moderate CIN2, and 214/416(51.4%) severe CIN3. Table 2 shows the cytological and histologic findings of women with cervical abnormalities within the respective study hospitals in Eastern Uganda.

 

 

Age-specific histologic prevalence of cervical intraepithelial neoplasia during the screening period, 2017–2018

Cervical intraepithelial neoplasia (CIN) were graded according to the CIN histological classification system (modified, 1990).The lesions were graded primarily according to the proportion of the epithelium which was occupied  by  basaloid,  undifferentiated cells (Mark et al., 2018; Canepa et al., 2019). Of the 1,077(100%) women who participated in the screening process, 661(61.4%) were histologically diagnosed with normal epithelium, while 22(2.0%) women were diagnosed with mild CIN1, 180(16.7%) with moderate CIN2, and 214(19.9%) with severe CIN3. Table 3 shows the comprehensive age-specific histologic prevalence of cervical intraepithelial neoplasia among the participants.

 

 

Univariable analysis between social demographic variables and cervical histological transformations in women from four selected hospitals in Eastern Uganda

When the predictor variables for cervical histological transformation were subjected to univariable analysis using a logistic regression analysis, they had the following logistic regression values: Tororo Hospital (OR=0.414; 95% CI: 0.275-0.625; p < 0.05), age group 15-24 (OR=0.485; 95% CI: 0.321-0.731; p < 0.05), age group 25-34 (OR=0.671; 95% CI: 0.460-0.978; p < 0.05), age group 35-44 (OR=0.683; 95% CI: 0.467-0.999; p = 0.05), widowed (OR=1.667; 95% CI: 1.046-2.658; p < 0.05), last menstrual period <1-week (OR=1.888; 95% CI: 1.189-2.997; p < 0.05), last menstrual period 3-4 weeks (OR=1.667;95% CI: 1.165-2.386; p < 0.05 ), use of contraceptives (OR=2.630; 95% CI:  2.029-3.408;  p < 0.05), multiple sex partners (OR=54.920; 95% CI: 29.583-101.54; p < 0.05), history of STI (OR= 3.173; 95% CI: 2.47-4.148;  p < 0.05) and HIV (OR=0.379; 95% CI: 0.294-0.488, p < 0.05) were found to be statistically significant (p  ≤ 0.05) with cervical histological transformation (Tables 4 and 5). When the univariate significant predictor variables for cervical histological transformations were subjected to multivariable regression analysis, they had the following logistic regression values: age 15-24 (OR=0.479; 95% CI: 0.277-0.827; p < 0.05), age group 25-34 (OR=0.531; 95% CI: 0.323-0.875;  p < 0.05) widowed (OR=2.083; 95% CI: 1.110-3.912; p < 0.05), use of contraceptives (OR=2.139; 95% CI: 1.532-2.986; p < 0.05), history of STI (OR=2.866; 95% CI:  2.047-4.012; p < 0.05) and HIV (OR=0.295;   95% CI:0.212-0.409;  p < 0.05)  were  found to have statistically significant relationships (p < 0.05) with high grade cervical histological transformation (Table 6).

 

 

 

 

 

 

 

 

In the current study, cytological and histological findings showed that, women who were diagnosed with HSIL were at an increased risk of developing CIN. Of the 416/1077(38.6%) women who were diagnosed with CIN, 22/1,077(2.0%) had mild CIN1, 180/1,077(16.7%) had moderate CIN2, and 214/1,077(19.9%) had severe CIN3. Older women aged 45-55years were frequently diagnosed with the advanced disease stage; as compared to sexually young adults aged 15-24 who were less commonly diagnosed with the late disease stage. Older women were less often diagnosed at early stage of cervical cancer compared to younger  women.  This  may be due to lack of seeking obstetrics and gynecological medical care in post-menopausal period,  particularly  for  women  who  live  in  rural areas where health care services are difficult to access. Another likely factor could be lack of awareness   about  their  susceptibility  to  cervical cancer (Abotchie and Shokar, 2009; Kharbanda et al., 2015; Mitchell et al., 2017). In line with our study, are qualitative studies  conducted  in  India, South Africa and Uganda which found out that older women who were of low socio economic status and unemployed, were less likely to participate in cervical cancer screening (Bradley et al., 2004; Kaku et al., 2008).

According to Muñoz et al. (2004), and Sankaranarayanan et al. (2004), long-term use of oral hormonal contraception has also been associated with an increased risk of cervical cancer and its precursors, but it has not been evident in all studies (Bertram, 2004; Bhatla et al., 2018). Our study demonstrated that there was an increased risk for cervical intraepithelial neoplasia in women who had used oral hormonal contraception method with ORs of 2.139 (p<0.001). The study indicated that usage of oral contraceptives increased the risk of developing cervical intraepithelial neoplasia.

Out of605/1077(56.2%) women who consented to having used oral hormonal contraceptives, 81/605(13.4) were diagnosed with CIN2, and 194/605(32.1) with CIN3 as opposed to 472/1077(43.8) who did not use oral contraceptive method in which 99/472(21.0 women were diagnosed with CIN2 and 20/472(4.2) with CIN3.This trend confirmed previous work in the region (Banura et al.,2008) which showed that prolonged usage of oral hormonal    contraception   method    contributed    to   an increased risk of cervical intraepithelial neoplasia development with an adjusted OR of 3.15 [95%CI;1.62-6.13, p=0.001] among women in central Uganda.

In the current study, there was no significant risk for CIN in women who had circumcised sexual male partners with unadjusted ORs of 1.714(p<0.609). Out 164/1077(15.2) women who had circumcised male partners only 12/164(7.3) women were diagnosed withCIN2, and 13/164(7.9) with CIN3 in contrast to913/1077(84.8) women who had un-circumcised male partners in which 168/913(18.4) were diagnosed with CIN2 and 201/913(22.0) with CIN3 respectively. The study demonstrated that women who had circumcised partners were at a lower risk of developing cervical abnormalities than women with uncircumcised sexual partners. Indeed, our findings are in line with other studies (Cook et al., 1994) which showed that, male circumcision was associated with the prevention of common sexually transmitted diseases. Furthermore, data analysis from International Agency for Research on Cancer (IARC) established that circumcised men not only had substantial lower risk of penile HPV infections than uncircumcised men, but also that their partners had a lower   risk  of   HPV   infections   and   a   lower   risk   of developing cervical cancer (Wideroff and Haderer, 1999).The protective effect was more pronounced among women whose male partners engaged in high-risk sexual behaviors. Further evidence of the protective effect comes from two resent randomized controlled trials. One trial showed that circumcision of adolescent boys and men in a rural Uganda population had reduced HPV prevalence by 35% (Tobian et al., 2009).

The present study also showed that, there was an increased risk for CIN development in women who had history of STI’S with ORs of 2.866 (p<0.001). Out of 635/1077 (60.0) women who had history of STI’S,103/635(16.2) were diagnosed with CIN2, and 196/635 (30.9) with CIN3 respectively. Our findings are in agreement with other studies by Mosciski et al. (2001)who showed that, cervical infections with STI’S such as chlamydia trachomatis and HIV may increase susceptibility to cervical cancer development by inflammation or micro abrasions of the epithelium resulting from sexual intercourse. This may allow direct access to basal epithelium cell. It is possible that STI’S could enhance the oncogenic effect of an already established HPV infection by influencing local response (Schmuz et al., 1989).

Last but not the least, cervical intraepithelial neoplasia has long been associated with HIV infection, and in fact, it was added as an AIDS-defining illness to early AIDS case definitions by the CDC. In the present study, women who were HIV positive had an increased risk of CIN development with ORs of 0.295(p<0.001). As women with HIV have a higher frequency of HPV coinfections, rates of pre-cancer and cancer are typically higher. Previous studies in sub-Saharan Africa have demonstrated a consistent association of cervical cancer and HIV in Senegal (Holmes et al., 2009); South Africa, Zimbabwe (Moodley, 2006); and also in other low-and middle income countries (Forhan et al., 2015).

 

This study showed that the prevalence of cervical intraepithelial neoplasia infections was high in the studied population of women in Eastern Uganda. Age at first intercourse, multiple sexual partners, Infection with HIV, long-term use of oral contraceptives, and cervical infections with STI had statistically significant associations with high grade of cervical intraepithelial neoplasia development. There is need for effective and affordable screening methods for early detection of precancerous lesions and treatment in poor resource setting

 

The authors have not declared any conflict of interests.