Impact of HIV infection on cervical neoplasia in a peri-urban population in Tigoni, Kenya

Abstract

BACKGROUND Cervical cancer is both a preventable and potentially curable disease. It is preventable by vaccination and screening, and curable by treatment at an early stage. The single most important factor that has made a difference in lowering mortality and morbidity from cervical cancer is screening and treatment of early lesions. It remains the commonest cause of cancer deaths among women in low resource settings despite effective screening tools, primarily because mounting successful sustainable cytology-based programmes has not been possible. Emerging data is showing that infection with human immunodeficiency virus (HIV) influences the persistence of high risk human papillomavirus (HPV) infection, now known to be a necessary cause of cervical cancer, and therefore cervical carcinogenesis.

However, the long-term impact of the HIV epidemic on the incidence of cervical neoplasia is still uncertain, especially in countries where the prevalence of both diseases is high. The effect of treatment with highly active antiretroviral therapy (HAART) that is allowing more people to live longer with sub-optimally functioning immune system and how this will therefore influence the incidence of acquired immunodeficiency syndrome (AIDS)-related tumours such as cervical cancer is still unfolding. It has been shown that HIV -infected women are at increased risk of cervical neoplasia, including cervical cancer, and that progression is faster and treatment outcomes poorer. HIV -infected women therefore require special clinical surveillance of their lower genital tract for early detection and eradication of HPV-related neoplasia.

Although HPV infection is a necessary cause, it is not sufficient for the development of cervical neoplasia. Other risk factors such as host genetic and environmental factors may also be contributory. For example, it has been shown that certain human leucocyte antigen (HLA) class II alleles increase the risk for cervical carcinoma, primarily because the carriers of those alleles appear to be more susceptible to oncogenic HPV. Since HLA is involved in cellular immunity, certain HLA alleles make the carriers less efficient in viral clearance. Only about five percent of eligible Kenyan women are screened for cervical cancer, partly because awareness is low, and more importantly Pap smear screening services are largely unavailable. In contrast, HIV/AIDS awareness is very high, with over 90% of the Kenyan population having some information about it. Moderate successes in HIV management programmes present a potential entry point for the screening of cervical cancer, as HIV-positive women regularly access services for investigation and treatment.

By screening for cervical neoplasia, a previously unscreened population that has a high prevalence of HIY, it is expected that the influence of HIV on the prevalence of cervical neoplasia will be elucidated. This and a better understanding of the co-factors that influence progression of cervical neoplasia may leap forward prevention strategies for populations devastated by the dual impact of both infections. Based on the above observations, a study was designed to determine the impact of HIV infection on cervical neoplasia in a peri-urban population in Tigoni, Kenya Broad Objective: To determine the impact of HIV infection on cervical neoplasia in a peri-urban population in Tigoni, Kenya. Specific Objectives: I. To determine the prevalence of HI V in a peri-urban female population in Tigoni, Kenya 2. To determine the influence of HI V on cervical neoplasia in the same population 3. To determine the risk factors of HIV and HPV co-infection and the correlation with cervical lesion grade 4. To describe the risk association between HLA antigens and cervical neoplasia MATERIALS AND METHODS Study site and population: The study was carried out at the Tigoni District Hospital, in Kiambu West District in central Kenya. Although living on the outskirts of Nairobi, with good road transportation and fairly well served by public health services, this peri-urban population has no access to cervical screening. HIV prevalence in this region at the time the study commenced was higher than the national prevalence. Study design: Two study designs were used, a cross-sectional survey to determine the prevalence of cervical neoplasia and HIV infection, as well as to describe the impact of HIV co-infection on HPY genotype distribution and cervical lesion grade; and a case control study design to determine the risk factors for cervical neoplasia, and the association between HLA class I and II antigens and cervical neoplasia.

Sampling and sample size: Women 25-60 years were invited to participate. With an estimated 10 percent prevalence of cervical neoplasia, and in order to obtain at least 500 women with an abnormal Pap smear, approximately 4,500 women were screened. Screening procedures: All eligible consenting women were provided with information about cervical neoplasia, HIV transmission, testing and the meaning of test results before blood was drawn for HIV testing by enzyme-linked immunosorbent assay (ELISA). After a Pap test, women who had an abnormal smear according to the Bethesda system of reporting cervico-vaginal cytology, and those who were HIV positive, were appropriately referred. Abnormal cytology was any atypical squamous cells of undetermined significance (ASCUS) or greater. The HlV testing, CD4 counts, HPV and HLA typing [both by polymerase chain reaction (PCR)] were done and reported using standard protocols.

Ethical considerations: The study and protocol amendments were approved by the Kenyatta National Hospital/UniversityofNairobi - Ethics and Research Committee, and permission granted by the Tigoni Hospital Management Board. All principles of research ethics on human subjects were observed. For those with a cervical abnormality or who were HIV positive, appropriate counseling, referrals and arrangements for clinical support were made. RESULTS A total of 4,491 women were enrolled. The mean age was 36.8 years (±8.9). The HIV prevalence was 14.3 percent. The 30-34 year age group was the most affected. HIV sero-positive women were significantly younger than HIV negative women (p <0.000 I). Early age of sexual debut, more than 2 lifetime sexual partners, and mean parity greater than 3 were significant independent risk factors for HIV infection. Contraception method use (current or ever used) was significantly higher in women who were HIV negative than HIV positive (p <0.0001).

For 96.1 percent of women, this study provided their first opportunity for a Pap smear. Of the 4,363 women who had a satisfactory Pap smear, 91.4 percent had normal cytology and 8.6% were abnormal. Both low and high grade cervical lesions were more common in younger women, 25-39 year age group with a slight peak in the 30-34 year age group. Women older than 40 years tended to have higher grade lesions; low grade lesions in 45-49 years were 5.0 percent and 3.9 percent in the 50-54 years compared to 10.5 percent and 7.4 percent of high grade lesions respectively. Cervical lesions were uncommon in the age group older than 55 years (0.03 percent).

However, these differences were not statistically significant (p=0.138). Of those who had normal cytology (n=3,984), 11.6 percent (461) tested HIV positive, compared to 38.7 percent of women with LSIL and 52.6 percent of women with a high grade squamous intraepithelial lesion (p<O.OOOI). The median CD+4 lymphocyte counts were significantly lower in HIV -positive women with a high grade lesion (p<O.OOO 1). Women with abnormal cytology were more than five times more likely to be HIV positive than those with normal cytology (OR=5.60 [4.47-7.02])(p=0.000), while women younger than 35 years were almost twice as likely to be HIV positive than women older than 35 years (OR=1.881 [95 percent Cl 1.099-3 .220])(p<0.000 1).

A subset of HIV positive women with matched HIV negative women with normal cytology (n=438) were tested for HPV, 140 (31.9 percent) were positive for at least one type of HPV. The 5 most frequent HPV types were 16, 56, 53, 35, 39 with 15 percent, 10.5 percent, 9.3 percent, 7.9 percent, and 7.9 percent respectively. A total of 91 (20.5 percent) had a high-risk type, 44 (10 percent) had an intermediate type, and low risk types had a frequency of 1.1 percent), (p<O.OOO 1). When compared with HIV as a risk factor for HPV infection, 32.5 percent were HPVnegative/HIV positive, 19.4 percent were HPV positive/HIV -negative, 67.5 percent were both HPY positive and HlV positive (p<O.OOOI); HR-HPV positive/HIV positive were 37.7 percent, lRHPY positivelHIV positive 26.3 percent, and LR-HPV positive/HIV positive 3.5 percent, (p<O.OOOI). High risk (HR)-types were the more frequent across all positive samples.

Type 16 was the most frequent type in both normal and abnormal cytology samples, but it had the highest frequency in high-grade cervical lesions (p<O.OOO 1). The rates of HPV infection and abnormal cytology increased as C04+ cell count decreased as did the number of HPV types involved in the infection. A total of 66 women were enrolled for a nested case-control study (43 cases, 23 controls). The ages ranged from 25 to 60 years with a mean age of 36.98 (SO ± 8.7 years). HIV prevalence in this group was 40.9 % (27/66). A total of forty three (43) or 62.2 % of women had a cervical abnormality by Pap smear or cervical biopsy that was graded either low grade or high grade squamous intra-epithelial lesion.

The HLA class I allele B35 and Class II allele DR] (OR=11.5 [1.32-99.46]) and (OR=3.7 [1.01-13.65] respectively) (p=0.018 and p=0.04 respectively) appeared to increase the risk for cervical neoplasia while HLA class I allele BW6 and CW7 appeared to reduce the risk (OR=0.2 [0.06-0.97] and OR=0.2 [0.08-0.71] respectively) (p=0.048 and p=0.012 respectively). CONCLUSIONS The HIV prevalence in Tigoni, Kiambu West District was about 1.6 times the national prevalence among women. The exact reason for this is uncertain but some possible explanations include social susceptibility as well as biological factors that make women more vulnerable to HlV infection. There is need to explore these further in order to design effective intervention programmes. The prevalence of cervical neoplasia was high, up to 8 percent in HIV negative women and significantly higher in HIV positive women. More than 50 percent of those who had a high grade lesion were HIV positive.

Women who were HIV positive were up to 4 times more likely to have an abnormal pap smear compared to those who were negative. Median and mean C04 counts also progressively decreased with the higher grade lesion suggesting that HIV -attenuated HPV -specific immune responses play an important role in HPV -associated disease. HPV prevalence in this sample population was high at 32.7 percent. There was some variation in HPV type distribution in this group from that reported for high-grade cervical lesions and invasive cancer for the African continent. HPV types found in this study group may have important implications for diagnosis and for the development of the next generation of vaccines for the region. Some HLA class I and class II alleles appear to increase the risk for cervical neoplasia, while some alleles appear to be protective. The size of the study cohort was small, however, and the genetic frequency of these alleles in the Kenyan gene pool is unknown. 4. RECOMMENDATIONS l. Majority of women in Tigoni, Kiambu West District, do not have access to cervical cancer prevention services. Screening programmes should include older women, since they tend to have higher grade lesions. Women in Tigoni remain at risk for cervical neoplasia, particularly if they are HIV positive and cervical cancer screening and treatment should be mainstreamed into HIV standard of care. 2. Further exploration of the significance of molecular markers such as HLA phenotyping that may identify those at risk for persistent HPV infection and therefore at increased risk for cervical neoplasia need to be done as this may potentially impact treatment and prevention. 3. The HIV prevalence in Tigoni, Kiambu West District, remains high compared to the national prevalence that has also been showing a downward trend. The reasons remain unclear, and warrant further study in order to develop relevant interventions. 4.· HIV screening and up-scaling of HAART to provide opportunities for prevention of infection to partners and HIV care and treatment in order to stem the epidemic.