Atypical cervical cytomorphologic predictors: a descriptive study of pre-cervical cancer patients of low education in Kenya

Introduction high risk HPV is the perpetrator of cervical cancer disease, however screening and vaccination is not included in cervical cancer prevention program within public hospitals in Kenya. This descriptive study assessed the association of specific microbial STI and socio-demographic characteristics and practices with cervical cytomorphologic presentations in regards to pre cervical cancer grades amongst health seeking patients attending the reproductive health clinic of Nakuru County referral hospital, a public hospital under newly devolved health services governance. Methods a total of 142 patients (AGC/AIS, n=8; HSIL, n=59; LSIL, n=35; controls, n=40) whose median age ranged between 20-70 years were purposively sampled. A structured questionnaire with closed and open ended entries was administered and STI screening including Pap smear examination for cytomorphological profiling done according to revised 2014 Bethesda classification. Associations were established using chi-square and multivariate logistic regression model to determine prediction of cervical atypia manifestations. Results a majority of the study participants had only primary education or no education in AGC/AIS (63%) and HSIL (73%) relative to LSIL (49%) and controls (53%) (P=0.017). Koilocyte rates were higher in AGC/AIS (25%), HSIL (52%) and LSIL (77%) compared controls (12.5%) (P<0.0001). ASCUS predominated in HSIL (61%) and LSIL (86%), while almost all AGC/AIS had AGCUS (88%). HR HPV 16/18 infection rates were higher in AGC/AIS (100%), HSIL (80%) and LSIL (83%) relative to controls (10%) (P<0.0001), and was associated with higher risk of having AGC/AIS (OR, 2.0; 95% CI, 1.940-1.947; P<0.0001); HSIL, (OR, 36.3; 95% CI, 9.5-139.5; P<0.0001); and LSIL (OR, 50.1; 95% CI, 12.0-209.0; P<0.0001). Conclusion altogether, pre-cervical cancer in Kenyan women is characterized by koilocytosis and ASCUS probably from the high rates of HPV 16/18 infections. Promoting cancer education and screening for high risk HPV infections and pre-cancerous lesions will improve women's reproductive health.


Introduction
Cervical cancer disease causes reproductive ill health and is among leading global women cancers accounting for at least 68% annual women mortality in the developing world [1,2]. The disease is preceded by pre cancer status identified by detection of abnormal cells in cervical smears. The Bethesda system (TBS) [3], is employed to classify cells as per their specific atypia such as koilocytosis  [4][5][6][7][8]. In Kenya, screening for HR HPV subtypes is not included in the cervical cancer prevention policy and neither is vaccination. This study examined HR HPV prediction to cervical cancer development, in addition to other cervical cell atypia manifestation determinants, with the aim of preventing progression of the disease.

Methods
We assessed the association of HR 16 Cervical screening by visual cervical inspection (VIA/VILI test) was conducted. Endocervical scrape smear examination for cytomorphological profiling was conducted using the revised 2014 Bethesda classification [3] into four pre cervical cancer study groups of: 1) LSIL; n=35; 2) HSIL; n=59; 3) AGC/AIS; n=8 and 4) Control-No evidence of pre cervical cancer; n=40. Those above 20 years and were VIA/VILI positive specifically where sexual debutation had begun were included in the study. Pregnant women were not included in the survey as well as those who were not willing to participate in the study. Blood samples were collected and screened for Treponema pallidum and HIV1/2 antibodies [9,10]. Endocervical specimens were collected for Neisseria gonorrhea, Chlamydia trachomatis and HPV antigen detection [11,12], as well as for scrape smears preparations for Pap smear processing and staining. Stained and mounted smears were microscopically examined for cyto-morphological profiling and categorization using the Bethesda system into distinct pre cancer grades [3]. All Pap smear microscopic examinations were conducted at low power X10 for evaluation of cellular sufficiency and for atypia at high power X40. Smear findings were confirmed by a clinical cytologist. Smear microscopic feature images were captured using the Venus 2.0 ® programme mounted on computer CPU (specific for Venus camera imagery) launched on a Leitz compound microscope. Various photomicrograph images were captured in different pre cancer grades.
Data management and analysis: data capture was done using Microsoft office Excel software [13]. After clean up to remove outliers data was exported into SPPS statistical software [14]. Descriptive statistics for frequencies and proportions were generated. Chi-square tests were used to determine associations between dependent variables (pre cervical cancer grades) and independent variables (predictors). Further analysis using multivariate logistic regression models were used to determine the magnitudes of associations between study groups and likelihood tendency (OR) of predicting pre cervical cancer sign manifestation at 95% confidence interval).
Ethics approval and consent to participate: ethical approval for this study was obtained from the Kenyatta University Ethics Review Committee (KUERC-KU/R/COMM/51/228) and Nakuru County Referral Hospital Research and Ethics Committee (RII/VOL.I/08).
Written informed consent was obtained from all the study participants prior to administering the questionnaire and collection of blood specimens and endocervical swabs and scrapes specimens.

Results
The prevalence of cervical atypia from Pap smear test protocol [15] which denoted the grade of pre cervical cancer is shown in Figure 1.   P=0.360). Surface marker antigen positivity for N. gonorrhea was only detected in HSIL study group (5.1%) whereas antigen positivity for C. trachomatiswas obtained in both HSIL (6.8%); LSIL (2.9%) study groups, Table 3.
Evaluation of socio-demographic characteristics and practices,  obtained. At least 11% attributed its cause to infections while ~50% responded that they did not know the cause, Table 4.

Discussion
HR HPV exposure may not be sufficient enough to cause abnormal transformation of cervical epithelia. Other factors in addition to HR HPV exposure such as having advanced age in women, low immunity, poor chronic ill-health, HIV/AIDS infection, lifestyle and nutritional factors for example organic diet inadequacies, including multiplicity of sexual partners [4,6,[19][20][21][22][23][24][25] are implicated. Pre cervical cancer development has also been linked to under privileged socio-economic state of affairs such as high poverty mainly due to inequality and limitations to access of assets as well as ownership of the same, which berate women in most third world communities such as in Kenya [26 -28]. This study reports cytomorphologic results illustrating a reducing trend in the frequency of koilocytes (~77%, ~52% and ~25% in LSIL, HSIL and AGC/AIS) accompanied by an increasing trend in the frequency of abnormal epithelial cells of high grade cellular lesion types (ASC-H), ~14% in LSIL and ~39% in HSIL, which suggests that higher pre-cancer grade is characterized by manifestation of transformed koilocytes (Figure 2, Figure 3, Figure 4).
However, in the control group, small proportion of women showed koilocytic cells in their smears (12.5%). This suggests that women is this group are exposed to high risk viral agents that cause cytopathic effects and may potentially develop pre cervical cancer signs should persistence of infection ensue. A study conducted in American women shows similarity in the sense that normal healthy women without cervical epithelial lesions may be candidates of pre cervical cancer development [5]. Among the immune compromised individuals cell transformation rates could also be high [37]. Therefore, identification of koilocytes in cervical smears is an important indicator of LSIL pre-cancer diagnosis. Viral agents have been linked to cervical cancer [16][17][18]34]. While univariate analyses of HIV1/2 infection exposure did not yield statistical difference as displayed in Table 2. HR HPV16/18 significantly associated with LSIL, HSIL and AGC/AIS grades. Further additional multivariate regression modeling analyses confirms the association of HR HPV16/18 with pre-cervical cancer grades Table 3.
This implies that the recorded high cases of cervical cancer in pilot studies in Nakuru, [7] are due to high risk HPV strains of 16 High median age which was significantly associated with the pre cervical cancer grades (P<0.0001) was encountered, Table 4. This result is in harmony with reports in other studies which imply that having advanced age places a woman at risk of developing pre cancer signs within the cervix as a result of diminishing ovarian hormones due to menopause [29,30]. In addition, lack of medical check in routine screening of cervical cancer may also contribute to high proportion of median age groups, who may detected this when at advanced stage of cancer. Similarly, knowledge on cause of cervical cancer assessed from varied responses obtained from study participants was linked to pre cancer grades (P= 0.002), as displayed in Table 4. This association could be attributed to low levels of education (P= 0.017) amongst the study participants drawn from the public referral hospital serving Nakuru County residents. Similar studies conducted in other counties in Kenya show consistency in terms of low education levels, illiteracy and ignorance among cervical cancer patients [27,31]. Eight study participants (5.6%) with a

Competing interests
The authors declare no competing interests.
Esther Muitta, Tom Were, and Ng'ethe Muhoho designed the study.
Esther Muitta performed cytomorphological examination, photomicroscopy and interpretations. Esther Muitta and Anthony Kebira Nyamache performed microbial STI screening. Tom Were and Esther Muitta performed statistical analyses, interpretation of results and codrafted the manuscript. All authors read and approved the manuscript.

Acknowledgements
Authors wish to appreciate and acknowledge the patients who participated in this study. We also thank the Management and staff of Nakuru County referral hospital for allowing us to conduct this study. We appreciate more especially, Mrs. Kabura Kinyari and Mrs.
Winny Migwi for your assistance in sample collection and Dr. Titus Gulungu, senior Pathologist for your technical advice.