Prevalence and factors associated with high-risk human papillomavirus infection and cervical lesions among women with HIV in Addis Ababa, Ethiopia | BMC Cancer

This study assessed the prevalence of hrHPV infection, the hrHPV genotype distribution, cervical lesions and associated factors in WWH living in Addis Ababa. Using a facility-based retrospective record review of women screened for cervical cancer by HPV-DNA testing and VIA, the study found that the prevalence of hrHPV infection among WWH in Addis Ababa was 27.8%. HPV16 and HPV 18 accounted for 15.3% and 4% respectively. The proportion of women infected with the 12 high-risk genotypes (31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68) other than 16 and 18, termed in the study as ‘others’, was considerably high at 73.3%. VIA screening was conducted in women who were tested positive for hrHPV to detect cervical lesion. The prevalence of VIA positive result in women who were tested positive for hrHPV was 15.5%.

The prevalence of hrHPV infection is said to differ by geographic location, socioeconomic conditions and associated illnesses, such as HIV [16]. This study found the prevalence of hrHPV infection to be 27.8%, which is higher as compared to other studies in Ethiopia. This result might reflect the fact that hrHPV infection is more prevalent among WWH than in the general population because of the inability of the host to clear the infection and reactivation of latent HPV infections due to immune suppression. For instance, a study performed in South Central Ethiopia among the rural general population found the prevalence of HPV to be 23.2% and that of hrHPV to be 20.5% [17]. Another study from Attat Hospital found the prevalence of HPV to be 17.3% and that of hrHPV to be 16% [9]. As compared to studies mentioned above, the prevalence in this study was higher, which might be due to the earlier stated higher rate of infection with HPV in WWH. Additionally, differences in test performance can be considered. However, the prevalence in this study is consistent with the 26.5% hrHPV prevalence identified in Rwanda [18]. With regard to genotype distribution, globally, the most prevalent hrHPV genotypes are HPV 16 and 18, but a study from South Central Ethiopia found that the most common hrHPV genotype was 16, followed by 35, 52, 31, 45 and 18 [17]. Whereas the most common genotypes were 16 and 18, while the remaining larger proportion were listed as others in this study.

The study showed that age group 25–29 years was significantly associated with higher odds of hrHPV-positive result than being between 40 and 49 years. This finding is consistent with multiple studies showing a higher prevalence of hrHPV infection in younger age groups, but most women will clear the infection during subsequent years. For instance, a study from South Africa showed that the prevalence of hrHPV in both HIV-positive and -negative women was the highest in the youngest age group, which was 18–25 years [19]. Another study from Kenya found that WWH under the age of 30 had the highest prevalence of hrHPV [20]. This could be attributed to risky sexual behaviors commonly observed among young adults [21, 22]. Additionally, individuals in this age group may have been recently exposed to HPV, giving their immune systems less time to clear the infection [23]. The study also showed married and widowed women were significantly associated with lower odds of hrHPV infection as compared to divorced women. This finding is also seen in other studies. For instance a study out of Italy found that never married or divorced women were factors associated with hrHPV infection [24]. Similar finding was also seen in Ghana where being single or divorced was highly associated with hrHPV infection [25].

In follow up screening, a considerable proportion (83%) of the hrHPV-positive women came back for VIA screening, and 15.5% of these women had VIA positive result. This result is similar to those of a study conducted in Amhara region, which had a prevalence of cervical lesions in 13.1% [26]. The results for cervical lesions are higher than those described in other studies, which had not initially screened for hrHPV infection before assessing for cervical lesions. For instance, in a study done in North-Western Ethiopia found the prevalence of cervical lesions among WWH to be 9.3% which is lower than the findings in this study [27]. Similarly, a study done in Amhara regional state (Woldia and Dessie Hospitals) the prevalence of cervical lesion among WWH was 9.9% [28]. A study performed in two health centres in Kenya found VIA positivity in these women to be 7%, which was lower than the results of our study [29]. The women in these studies were not initially screened for hrHPV infection which may have contributed to the lower prevalence as compared to this study.

The study also showed that a history of STI was significantly associated with having a cervical lesion diagnosed by VIA. The odds of women who had history of STI were 1.69 times more likely to have cervical lesions diagnosed by VIA than women with no history of STI. This may be because cervical precancerous lesions from hrHPV, a sexually transmitted infection, commonly co-occur with other STIs due to similar transmission methods. This finding was seen with other studies performed both in Ethiopia and in other African countries evaluating cervical lesions in WWH [15, 30] but it is worth mentioning that this study did not find a direct association between hrHPV and a history of STI. This finding may also be as a result of increased cervical inflammation that is seen with STI, studies have shown than cervical inflammation seems to contribute to an increase in false positive VIA test [31, 32].

The strength of this study was that the collected data included all women who had HPV-DNA testing done at 16 pilot health facilities during the study period and therefore enrolled a higher sample size compared to other studies conducted in Ethiopia. In addition, the pilot facilities were located in different sub-cities across Addis Ababa, which increased the representativeness of the study sample.

However, this study had some limitations, like the inability of the Abbott Real-Time hrHPV test, which was used to analyse the hrHPV genotypes, to report specific HPV genotypes other than HPV 16 and 18. The machine reports HPV 31, 33, 35, 39, 45, 51, 52 56, 58, 59, 66 and 68 as “other hrHPV genotypes” and a further differentiation of these positive tested samples in specific HPV genotypes was not possible. In addition, the findings of the study would have been better characterized if they were analysed with other independent factors such as age at first intercourse, number of sexual partners, CD4 count and viral load, but these data were not recorded in the registration logbooks used for this study.