Human Papillomavirus types prevalence and their association with cervical dysplasia among HIV and non-HIV infected women attending reproductive health clinics in Eastern Kenya

Background Human Papillomavirus (HPV) causes over 99% of all cervical cancer globally. In 2019; it was responsible for 3,286 deaths in Kenya. Understanding the epidemiological distribution of HPV genotypes by cervical dysplasia and HIV infection is important in designing prevention strategy and management of cervical cancer. Objective To determine HPV genotypes prevalence and their distribution by cervical dysplasia, social-demographic and risk factors associated with cervical cancer among HIV-infected women aged 18–48 years seeking reproductive healthcare in Eastern Kenya. Methods Cervical specimens were obtained for cytology, HPV-genotyping, histology while social-demographic factors were collected using a questionnaire and analysed using Pearson chi-square test. Results 317 womencases: 161(50.8%); control 156(49.2%), mean age: 34.3, range 18–46 years were recruited. Thirteen HPV genotypes associated with cervical dysplasia were: CIN1{cases: HPV81[12(3.8%), HPV11[2(0.6%); control: HPV53 and HPV66[1(0.3%)}, CIN2 {cases: HPV11, HPV16, HPV661(0.3%), HPV816(1.9%) and single case1(0.3%) of HPV9, HPV11, HPV16, HPV44, HPV66, HPV81 HPV88, HPV53 and HPV58; control: HPV81[2(0.6%)} and invasive cancer {cases: HPV16[1(0.3%) and HPV81[3(0.9%); control: HPV16 and HPV66[1(0.3%). Conclusions Cervical dysplasia was associated with more mixed-lr/hrHPV genotypes among HIV-infected than HIV-uninfected women. The finding adds to the pool of knowledge the epidemiological data required in determining the population at risk for cervical cancer.


Introduction
Human Papillomavirus (HPV) is primarily responsible for 99.7% of cervical cancers globally. It is sexually-transmitted and causes cervical neoplastic changes leading to cervical cancer; the second most common type of cancer in women aged 15-44 years in Kenya. It was responsible for 311,365(8.2%) annual global mortality and 3,250(12.8%) in Kenya in 2019 1 . Cervical-screening rate in Kenya is only 3.2% for women aged over 18 years 1,2 . Human Papillomaviruses are grouped based on their oncogenecity as Group I: Carcinogenic: HPV16, 18 Health clinics of eastern Kenya in 2019. A sample size was calculated by infection rate of 2.8% 1 and distributed by 2017 clinic's attendance. Stratified sampling by county of residence and simple-random sampling per county were used to recruit participants.

Social-demographic data collection
Social-demographic data on residence, age, education level, parity, sexual-orientation, and family-planning method were collected by a nurse using a translated (Swahili) questionnaire.

Collection and storage of cervical exfoliated cell samples
External genitalia and cervical opening (os) were examined with a speculum while the participant lay in a lithotomic position. Cervical cytological specimens were collected using a cervical brush (Dacron™ cervical brush; Digene Corporation, Maryland), spread and fixed immediately on a clean glass slide. The brush was stored and transported at 1-4ºC in Digene Specimen Transport Medium for HPV-genotyping.

HPV DNA genotyping
The following procedures were used for HPV genotyping: a.
A mastermix containing PCR buffer (1X), 2.0mM MgCl2, 100µM dNTPs, 0.13 parts Taq polymerase-enzyme, 500nM of respective forward and reverse primary and secondary primers. 5µl of the DNA extract was used in primary PCR while 5µl of primary PCR product was used in nested PCR. First reaction: 4minutes at 95ºC (initial denaturation) then 30cycles of 20 seconds at 95ºC, 40 seconds at 56ºC then 2 minutes at 72ºC. Nested reaction: 4minutes at 95ºC (initial denaturation) then 30cycles of 20seconds at 95ºC and 40 seconds at 60ºC then extension (7 minutes) at 72ºC. Positive control of known CIN2+ and negative control (distilled water) were incorporated in both reactions. c.

HPV genotyping and phylogenetic analysis
Sequences were edited with CHROMAS software Version 2.4.3 then blasted in NCBI http://blast.ncbi.nlm. gov/blast.cgi. HPV type-sequences with unique divergence were phylogenically analysed and referenced from GenBank. Representative sequences and their references in input file underwent multiple alignments with CLUSTAL W in MEGA X software 14 . The Maximum Likelihood method and the Tamura-Nei model were used to infer evolutionary history. Neighbor-Join and BioNJ algorithms were used to construct initial trees for the heuristic search of the matrix of pair wise distances by the Maximum Composite Likelihood (MCL) method by selecting the topology with superior log likelihood value. Eighty-six nucleotide sequences were involved while codon positions were 1 st +2 nd +3 rd .

Ethics approval
KEMRI Scientific Ethical Review Unit approved the study (KEMRI/SERU/CVR/004/3342). Participants were consented orally and data collected was confidentially stored by the Principle Investigator. Mentally-incompetent participants were excluded.

g.
Distribution of HPV genotypes among HIV and non-HIV infected women with cervical dysplasia Phylogenetic tree of HPV samples marked in red aligned against the representation of the different HPV genotypes distributed worldwide. Most HPV81 clustered with those cases detected in Bangkok, Morocco and Thailand while HPV66 clustered with cases reported in Tunisia, Morocco, Iran and India (Figure 3).

Discussion
This study established an overall HPV infection-rate of 27.12% among HIV-infected (23.03%) and HIV-non infected (4.1%) women. It disagrees with the overall HPV infection rate of 40.0%1 in Kenya 11 13.7% in Ethiopia 12 . Embu and Meru Counties had the highest overall HPV infection rate among HIV-infected women. This is indicator of the expected high burden of cervical neoplasia in the region. Women aged below 35 years had a high rate of mixed HPV genotypes and a significant association between HIV infection and abnormal cytology outcome which agrees with published observations 1, 5,7 . A possible explanation is that HIV infection may facil itate the progression of HPV infection to cancer in young women, and an inverse relationship of highrisk HPV prevalence and age has been described 4,11,13 .
Single HPV type's infection in CIN1+ showed diversity compared with multiple HPV type's infection by HIV-infection. HPV44(α10), HPV58, HPV81, and HPV88 do not feature in many studies as potential oncogenic types, and their dominance in cervical dysplasia increases with HIV infection 1,8,18,19 as seen in this study. Members of (α7) and (α9) dominate malignant tissues coding for a hydrophobic E5, hence considered oncogenic. Furthermore, following HPV infection, an

Funding
This study was funded by NACOSTI. The funding institution had no role whatsoever in designing the study, sample and data analysis, or writing of the manuscript.

Availability of data and materials
The datasets are available from the corresponding author on reasonable request.