Cervical cancer screening by visual inspection and HPV testing in Eswatini.

In 2009, visual inspection with acetic acid (VIA) followed by cryotherapy (VIA-and-cryotherapy), was introduced into the Eswatini cervical cancer prevention programme. We present screening results of 654 women attending VIA-and-cryotherapy who participated in a sexually transmitted infections prevalence study, at which samples for HPV DNA testing and liquid-based cytology (LBC) were also collected. VIA positives (VIA+) ineligible for cryotherapy, suspected cancers and women with high-grade squamous intraepithelial or worse lesions (HSIL+) on LBC were referred for diagnosis and treatment. Women with negative VIA who were HPV positive (HPV+) and those VIA+ treated with cryotherapy were recalled for another VIA one-year later. The positivity rates of VIA, HPV, atypical squamous cells of undetermined significance or worse cytology abnormalities (LBC ASCUS+) and low-grade squamous intraepithelial or worse lesions (LBC LSIL+) were 9.7%, 42.6%, 13.2% and 5.3%, respectively. HPV testing detected 29 of 31 LSIL+ (93.6%, 95%CI: 78.6-99.2) while VIA only detected 11 (35.6%, 95%CI: 19.2-54.6). The HIV prevalence was 43% (95%CI: 39.2-46.9). HIV positives were at increased risk of being VIA+ (age-adjusted odds ratio: 2.5, 95%CI: 1.5-4.3), HPV+ (3.7, 2.6-5.3) and having LSIL+ (16.3, 4.9-54.8). The ineligibility rates for cryotherapy were 38% (24 of 63 VIA+), and 46% among HIV positives (18 of 39 VIA+). HPV testing was substantially more sensitive than VIA, thus, HPV followed by ablative treatment may be more effective. However, the high ineligibility for cryotherapy highlights the need for improving the assessment of eligibility for ablative treatment and for strengthening colposcopy, particularly in populations with high HIV prevalence.

Introduction

Cervical cancer remains a significant public health problem, with 604,127 new cases and 341,831 deaths occurring annually worldwide (Sung et al., 2021). Eswatini (formerly Swaziland) has the highest incidence of cervical cancer, where approximately 6.5% of women develop cervical cancer before 75 years of age (Arbyn et al., 2020), and 35% of women aged 15–49 years are affected by the human immunodeficiency virus (HIV) (UNAIDS - ONUSIDA - Swaziland, 2021).

Persistent human papillomavirus infection (HPV) infection is necessary for the development of cervical cancer (Walboomers et al., 1999). Both HPV and HIV are sexually transmitted infectious agents, and infection by one virus may accelerate transmission of the other (Smith-McCune et al., 2010). Results from a recent meta-analysis indicate that the risk of incident HPV infection is doubled among women living with HIV (WLHIV) compared to HIV negative women (any HPV type or high-risk HPV types; adjusted-pooled relative risk, 95%CI: 2.77, 2.05–3.75, 2.18, 1.58–3.01, respectively). These results also show that WLHIV have a decreased ability to clear HPV after treatment for high-grade cervical lesions, and have more remaining/recurrent disease, compared to HIV negative women (Looker et al., 2018). Additionally, HIV infection and HIV-associated immunosuppression can increase a woman's susceptibility to HPV infection and increase HPV persistence, thereby facilitating progression to high-grade lesions and cancer (De Vuyst et al., 2008).

Tools to prevent cervical cancer are available: 1) effective vaccines against HPV infection, and disease caused by HPV infection, and 2) cervical screening to detect and treat precancerous lesions.

In 2018, WHO made a call to action for the elimination of cervical cancer (). Based on modelling estimates, achieving elimination would be feasible for most low- and middle-income countries (LMIC) by 2100 (Brisson et al., 2020), by scaling up HPV vaccination and cervical screening to reach and maintain the WHO 2030 targets of 90% HPV vaccine coverage, 70% screening coverage with high-performance screening tests, and management of 90% of women with cervical disease. In particular, to reach the 70% screening coverage, recent WHO guidelines on cervical screening and treatment strongly recommend using HPV DNA testing with or without triage, followed by ablative treatment of screened positives in whom treatment eligibility is confirmed ().

Unfortunately, HPV vaccines are not yet part of the Extended Programme on Immunization (EPI) in Eswatini. Cytology-based screening was introduced in the national cervical cancer prevention programme in 1983, however, most Pap smears have been done in symptomatic women (Okonda et al., 2009). In 2009, see-and-treat cervical screening by visual inspection of the cervix after the application of acetic acid (VIA) followed by treatment with cryotherapy, “VIA-and-cryotherapy”, was added to the programme. The aim of this was to enhance early detection of cervical lesions, and enable the extension of cervical cancer prevention services in four political regions. Results from VIA tests in two of these regions between 2011 and 2014 confirmed the subjectivity of the test and most importantly highlighted a very high proportion of potential cervical lesions among WLHIV in comparison to HIV negative women (Ginindza et al., 2017). Following this report, a prevalence study of sexually transmitted infections (STIs) nested within the screening programme was carried out between 2015 and 2017. Samples were collected at the time of VIA-and-cryotherapy screening for testing of several STIs, including high-risk HPV. Here we present the screening results of 654 sexually active women who participated in the study.

Materials and methods

The overall methods of the study have been published elsewhere (Ginindza et al., 2017). Briefly, sexually active women aged 15 and older attending one of five health care facilities that had fully implemented the VIA-and-Cryotherapy across the country were invited to participate. Two nurses in each health care facility were trained/re-trained on performing and reporting VIA-and-cryotherapy during a 5-day intensive course that included lectures, review of photos of normal and abnormal cervices and practical sessions observing examinations. After providing informed consent, contact information and basic socio-demographics to nurses; women who agreed to participate were assigned a unique study ID number. Then, the nurse exposed the cervix by inserting an unlubricated bivalve speculum, and collected cervical cells using a cytobrush that was first vigorously shaken inside a vial containing ThinPrep PreservCyt (Hologic-Cytic Company, Marlborough, USA) solution, and then discarded. This sample was used for HPV testing and LBC preparations.

Once the PreservCyt vial was collected, the nurse examined the cervix with the aid of a 100-watts lamp and identified the squamocolumnar junction. The nurse applied 5% acetic acid to the cervix and waited one minute to inspect for possible acetowhite lesions. VIA results were documented according to the Ministry of Health cervical cancer screening guidelines in a specific data collection form (The Kingdom of Swaziland, 2013) as: (1) VIA negative if non-significant or no acetowhite changes were observed, (2) VIA positive if an acetowhite area was noticed in and around the transformation zone, and (3) VIA suspicious for invasive cancer if a growth or ulcerative lesion was observed. Women who were positive for VIA were immediately offered treatment with cryotherapy after checking eligibility for ablative treatment, and those with suspected cancer or not eligible for cryotherapy were referred for biopsy and/or treatment by a gynaecologist at the Mbabane Government Referral Hospital and the Raleigh Fitkin Memorial Hospital (RFM). A woman was considered eligible for ablative treatment if: (i) the transformation zone was type 1 (completely ectocervical), (ii) a lesion did not involve >75% of the ectocervix, (iii) that the lesion did not extend into the cervical canal or into the vaginal walls, and, (iv) no suspicion of cancer; in accordance with the WHO Guidelines for Treatment of Cervical Intraepithelial Neoplasia 2–3 and Adenocarcinoma in situ (). VIA and cryotherapy were performed by ten nurses, two per facility.

During the visit and after HIV pre-counselling, 4 ml of blood were collected in a vacutainer tube from participants consenting to HIV testing. All specimens were transported to the Eswatini Government National Referral Laboratory in Mbabane where the testing was carried out.

HPV testing and LBC results were not immediately available. However, once results were available, HPV results were used to decide clinical management of VIA negative women. Women with LBC HSIL+ were referred for biopsy and/or treatment, independently of any other screening results.

Women who 1) received cryotherapy and 2) were VIA negative but tested positive for HPV at initial screening were recalled for a second screen with VIA at one year. All women were contacted by telephone and in case of no response, a next of kin whose telephone was provided on the baseline consented form was contacted; women who did not attend this visit before the end of the study were considered lost to follow-up. No additional samples were collected for HPV testing or LBC at this follow-up visit. VIA positives treated with cryotherapy at initial screening (group 1) were referred to the hospital for biopsy and/or treatment, while those in the second group were treated with cryotherapy. All VIA negatives returned to regular care.

HPV testing, LBC, and HIV-testing

The collected PreservCyt vial was used to first perform the Xpert HPV assay (Cepheid, Sunnyvale, USA) following the manufacturer's protocol. The Xpert HPV Assay is a qualitative in vitro test for the detection of the E6/E7 region of the viral DNA genome. The Xpert HPV test gives results in six separate channels: (i) sample adequacy control (SAC), (ii) P1: HPV 16, (iii) P2: HPV18/45, (iv) P3: HPV 31/33/35/52/58, (v) P4: HPV 51/59 and (vi) P5: HPV39/68/56/66. A woman could test positive for more than one channel and hence have multiple infections (e.g., P1 and P2, HPV 16 and HPV 18/45). Xpert HPV results were coded as positive if the test was positive for at least one of P1-P5 channels, and negative otherwise.

After HPV testing, PreservCyt vials were kept at room temperature until sent to the National Referral Laboratory at Mbabane Government Referral Hospital for ThinPrep LBC (Hologic-Cytic Company, Marlborough, USA) preparation by a trained technician, and then interpreted by a pathologist. LBC results were reported using the Bethesda nomenclature and grouped as: negative, ASC-US/AGUS (atypical squamous cells of undetermined significance/atypical glandular cells of undetermined significance), LSIL (low-grade squamous intraepithelial lesions) and HSIL+ (high-grade squamous intraepithelial lesions or worse) (Solomon et al., 2002).

The Alere Determine HIV-1/2 Ag/Ab Combo test was used to detect HIV-1/2 antibodies and free HIV-1 p24 antigen. Reactive specimens were confirmed by Trinity Biotech Uni-Gold Recombigen HIV test. All participants who tested HIV positive were post-counselled and referred for adequate care.

Statistical analysis

Data collected via questionnaire included socio-demographics, reproductive and sexual health and cervical screening history (previous Paps or VIAs). Histological results were not available as no biopsies were collected at initial screening or follow-up.

VIA positivity was calculated including and excluding suspected cancers. Two cytology thresholds were considered for LBC: ASCUS+ (ASC-US/AGUS or worse cytology) and LSIL+ (LSIL or worse). Positivity rates of VIA, HPV (overall, HPV16, HPV16/45 and other high-risk types) and LBC (both thresholds) were calculated overall, by age and by HIV status. Prevalence of HPV by cytologic grade within three age groups (20–29, 30–39 and 40–49) in HIV negative and HIV positive women was also estimated.

Logistic regressions were used to calculate odd ratios (OR) crude (COR) and adjusted for age and HIV status (AOR) of associations between screening history (Pap or VIA) and risk of having a positive screen (VIA, HPV or LBC). Likelihood ratio tests were used to explore potential interactions between screening history, age and HIV status.

Ethical considerations

The study was approved by the Eswatini Scientific Ethics Committee (MH599C/FW00015267/IRB0009688) and the Biomedical Research Ethics Committee of the University of KwaZulu-Natal (BE 242/14). Informed consent was obtained from all the participants. Under Eswatini national guidelines, consenting age for HIV testing and counselling is 12 years (Swaziland Integrated HIV Management Guidelines, 2015).

Results

Overall, 654 women aged 15–49 years participated in the study, only one of them refused to be tested for HIV infection. Two women did not have VIA results available: one was positive for HPV, had HSIL diagnosis on LBC and was referred for further evaluation; the other was negative in both HPV and LBC. There were 11 women with no valid HPV results (nine VIA negative, two VIA positive).

Among 652 women with VIA results, 586 (89.9%) had negative VIA, 63 (9.7%) positive and cancer was suspected in three. In the VIA negative group, 354 women were HPV negative and were returned to regular care, 223 were HPV positive and one had LBC HSIL. 39 of the 63 VIA positive were eligible for and received cryotherapy. The woman with LBC HSIL who was VIA negative and HPV positive, as well as the 24 women not eligible for cryotherapy and the three suspected cancers, were referred for biopsy and/or treatment at the local hospital. Only 75 (34%) of 222 VIA negative/HPV positive women attended the one-year follow-up: 60 were VIA negative and returned to regular care and 15 were VIA positive (9 treated with cryotherapy, 6 referred); while 26 (67%) of the 39 treated women attended the visit: 13 were VIA negative and returned to regular care, 13 VIA positives were referred for further diagnosis (Fig. 1).

Fig. 1

Flowchart of the cervical screening process.

VIA: visual inspection after acetic acid; NEG: Negative; POS: Positive; HPV: human papillomavirus testing; LBC: liquid-based cytology; HSIL: high-squamous intraepithelial lesions on LBC; HSIL+: HSIL or worse lesions on LBC.

Screening tests results

The overall positivity rates excluding suspected cancers, inadequate LBC and missing results were 9.7% for VIA, 42.6% for HPV, 13.2% for LBC ASCUS+ and for 5.3% for LSIL+. In total, 273 (43%) tested positive for HPV, 223 (39%) in VIA negatives, 47 (77%) in VIA positives and 2 of 3 women with suspected cancer on VIA. There were 69 women without LBC results (66 women with insufficient sample for LBC preparations, 3 with missing results), 506 (87% of 583) with normal LBC and 77 (13%) women with abnormal LBC. Among those with abnormal results: 46 had ASC-US, 29 LSIL and only two HSIL; 74% (57) were VIA negative, 23% (18) VIA positive and one had suspected cancer on VIA.

HPV positivity was around twice as high in VIA positive women compared to those with VIA negative results (39% vs 77%, p < 0.001), and increased with cytological grade from 33.9% in women with negative LBC to 60% in those with ASC-US, 96.4% in those with LSIL and 100% in those (n = 2) with HSIL (p-for-trend < 0.001) (Table 1). According to Xpert results, 64 (23.4%) of HPV positives were positive for HPV16, 78 (28.6%) for either HPV18 or HPV45 or both (HPV18/45), and 209 (76.6%) for at least one of the 11 high-risk (other HR) HPV types included in the test (Supplementary Table 1).

Table 1

VIA, HPV testing and LBC screening results

	VIA
	Negative			Positive			Suspected cancer		Missing
	586 (89.9%)			63 (9.7%)			3 (0.5%)		2 [0.3%]
LBC Results	HPV negative	HPV positive	HPV missing	HPV negative	HPV positive	HPV missing	HPV negative	HPV positive	HPV negative	HPV positive	All
Unsatisfactory	15	34	3	2	11	1	–	–	–	–	66 (10.2%)
Negative	323	147	2	8	23	–	1	1	1	–	506 (78.0%)
ASC-US	15	22	1	3	4	–	–	1	–	–	46 (7.1%)
LSIL	–	18	–	1	9	1	–	–	–	–	29 (4.5%)
HSIL	–	1	–	–	–	–	–	–	–	1	2 (0.3%)
Missing	1	1	3	–	–	–	–	–	–	–	5 [0.8%]
All	354 (61.4%)	223 (38.6%)	9 [1.5%]	14 (23.0%)	47 (77.0%)	2 [3.2%]	1 (33.3%)	2 (66.7%)	1 [50.0%]	1 [50.0%]	654 (100%)

LBC ASC-US includes ASCUS, AGUS and HPV infection cytologic diagnoses.

Denominators of percentages in parentheses exclude missing values.

Denominators of percentages in brackets “[]” include all women in each category.

Positivity rates did not vary significantly by age, but there was a decreasing trend of HPV positivity which was 58% in those <20 years and 21% in 45–49-years-olds (Supplementary Table 2).

Screening and HIV status

The prevalence of HIV among 653 women with HIV testing results was 43% (95%CI: 39.2–46.9) and increased from 21% in women aged 15–19, to 57% in those aged 35–39, then decreased to 40% in those aged 45–49 (Supplementary Table 3). HIV positive women were substantially more likely to have a positive screen compared to HIV negative women, age-adjusted odds ratios were 2.5 (95%CI: 1.5–4.3) for a positive VIA, 3.7 (2.6–5.3) for a positive HPV, 4.9 (2.9–8.4) for ASCUS+ and 16.3 (4.9–54.8) for LSIL+, though the latest association is only based on 29 women diagnosed with LSIL (three HIV negative and two HIV positive women with HSIL) (Table 2).

Table 2

Age-adjusted odds ratios of HIV status and risk of a positive screen.

Screening test	No. screened positive/ no. HIV negative	%	No. screened positive/ no. HIV positive	%	Age-adjusted odds ratios (95%CI)
VIA positive	24/368	6.5	39/280	13.9	2.5 (1.5–4.3)
HPV positive	117/369	31.7	156/274	56.9	3.7 (2.6–5.3)
LBC ASCUS+	22/341	6.5	55/242	22.7	4.9 (2.9–8.4)
LBC LSIL+	3/341	0.9	28/242	11.6	16.3 (4.9–54.8)

Fig. 2 depicts the prevalence of HPV by age (from 20 to 49 years) and cytologic grade (Negative, ASCUS, LSIL+) among HIV negatives and HIV positives. Despite small numbers, the prevalence of HPV decreased by age and increased with cytologic grade and was higher among HIV positive women compared to HIV negatives.

Fig. 2

Prevalence of high-risk HPV among HIV negative and HIV positive women by cytology grade within different age groups.

ASCUS: includes atypical squamous cells of undetermined significance (ASC-US) and atypical glandular cells of undetermined significance (AGUS) LBC results. LSIL+: low-grade squamous intraepithelial lesions or worse LBC results.

Despite small numbers, the proportion of VIA positive women ineligible for cryotherapy was also higher among HIV positive women compared to HIV negative women (46.2%, 18 of 39 HIV positives, 25%, 6 of 18 HIV negatives) at initial screening, and 18 of 24 ineligible women were HPV positive (14 HIV positives and 4 HIV negative ineligible); no differences were observed by age (Supplementary Table 4).

Screening results and cervical screening history

Among 640 women who recalled having previous Pap or VIA screening, 430 had not been screened before, 100 had at least one Pap and one VIA, 77 had Pap but no VIA, and 33 VIA but no Pap. Of the 177 with Pap, 115 had had only one and 62 two or more. <10% of women aged 15–24 years reported having been screened in the past, while about 40% of those 25–49 years have had at least one Pap or VIA (data not shown).

Women aged 25–49 years with no previous Pap or VIA were at significant increased risk of having a positive HPV result (OR: 1.8, 95%CI: 1.2–2.6 for no previous Pap and OR: 1.6, 1.1–2.4 for no previous VIA) and those with no previous Pap were also at increased risk of having LBC ASCUS+ (OR:1.9, 1.0–3.6); however, significance was lost when odds ratios were adjusted for age and HIV status (Table 3). Further stratified analysis confirmed increased risk for HPV positivity among younger (25–34 years) HIV negative women with no previous Pap or VIA (ORs: 3.5, 1.3–9.3; 3.2, 1.0–10.2, respectively) while no significant results were observed in older women or in HIV positives of any age (Supplementary Table 5).

Table 3

Associations between having screening history and having a screen positive result

	VIA				HPV test			LBC ASCUS+
Previous screens	N	% VIA positive	COR (95%CI)	AOR (95%CI)	N	% HPV positive	COR (95%CI)	AOR (95%CI)	N	% LBC ASCUS+	CO R95%CI)	AOR (95%CI)
Pap
Ever	176	11.4	1.0	1.0	173	32.4	1.0	1.0	156	8.3	1.0	1.0
Never	459	8.5	0.72 (0.41–1.28)	0.62 (0.34–1.13)	457	46.2	1.79 (1.24–2.59)	1.46 (0.98–2.19)	414	15.0	1.94 (1.03–3.63)	1.78 (0.92–3.46)
VIA
Ever	134	12.7	1.0	1.0	133	33.8	1.0	1.0	118	9.3	1.0	1.0
Never	514	9.0	0.68 (0.37–1.22)	0.59 (0.32–1.10)	510	44.7	1.58 (1.06–2.36)	1.26 (0.81–1.95)	465	14.2	1.60 (0.82–3.15)	1.47 (0.72–2.98)

COR = Crude Odds Ratio; AOR = Adjusted Odds Ratio, adjusted for age and HIV status.

Numbers in bold represent statistically significant results and those in bold italics borderline significant results.

Discussion

To our knowledge this is the first study reporting on cervical screening by VIA, HPV and LBC, allowing comparisons of different outcomes among HIV positive and HIV negative women in Eswatini, a country severely affected by the HIV epidemic. The study has been nested within an STIs prevalence study of 655 sexually active women who attended regular cervical screening across the country.

The prevalence of HIV was 43% (39%–47%), showing an increasing trend from 21% in women aged 15–19 years, up to 57% in 35–39 years old (p for tend<0.001), to then reach a plateau of about 40%–45% in women 40 and older. The high HIV prevalence observed is consistent with previous reported estimates for the country: 39% (among 18–49 years old) reported by the 2011 Eswatini HIV Incidence Measurement Survey (SHIMS) (Bicego et al., 2013) and 37% (aged 15–64 years) of self-reported HIV status published by our group (Ginindza et al., 2017).

HIV positive women were at increased risk of having a positive VIA (age-adjusted odds ratio, AOR: 2.5, 95%CI: 1.5–4.3), testing positive for HPV (3.7, 2.6–5.3) or having cytological abnormalities (AOR for ASCUS+: 4.9,2.9–8.4, AOR for LSIL+: 16.3, 4.9–54.8), the strong association of HIV infection with higher HPV persistence and higher prevalence of cervical lesions that if left untreated may eventually progress to cancer (Looker et al., 2018).

We found that women with no previous screening (with Pap or VIA) were at high risk of having a positive HPV or LBC screen, but no association with VIA screening results. After stratification by HIV status, the association with HPV positivity was still strong for younger HIV negative women but was lost for HIV positive women. This suggests that younger women may have been more aware of the benefits of screening than older women, and that the risk conferred by HIV infection for HPV positivity diminishes the role of other potential risk factors.

The main limitation of our study is the lack of histological confirmation of cervical disease. However, the study was embedded in the national screening programme, which is based on VIA followed by cryotherapy, and our results represent those from real-life conditions, particularly from settings where histopathology may be limited. Although HPV and LBC results were not fully used to determine clinical management and no biopsies were collected ahead of cryotherapy treatment, assuming LBC HSIL (n = 2) as true disease, HPV testing detected both lesions while VIA was negative in one and was not done in the other. If the threshold was LSIL+, HPV testing detected 29 of 31 lesions (93.6%, 95%CI: 78.6–99.2, including one without HPV result) while VIA only detected 11 (35.6%, 95%CI: 19.2–54.6 including one without VIA result). This shows the sensitivity of HPV testing was substantially higher than that of VIA (p < 0.001), which is in line with multiple reports demonstrating unequivocally that HPV testing is more sensitive in different resource settings (Almonte et al., 2015), and hence, more effective than VIA in the prevention of cervical cancer ().

The key advantage of screen-and-treat is the possibility of treating all women with cervical lesions as detected by the screening test used. Ablative treatment with cryotherapy or thermal ablation has a favourable trade-off of benefits and harms and can be introduced more widely than excisional treatment in LMICs. However, about 50% of women aged 50 and above are ineligible for ablative procedures and will require excisional procedures (Wentzensen et al., 2021). In a study evaluating treatability with cryotherapy on cervigrams collected in a large screening study in Costa Rica, two expert reviewers agreed that 26% of HPV positive women would have not been eligible for cryotherapy due mostly to inability to visualise the squamocolumnar junction (Gage et al., 2009). In our study, this proportion was 38% overall and 46% among HIV positive women. Unfortunately, hospital diagnosis and treatment data were not available, hence, ineligibility was not confirmed at colposcopy. Nevertheless, these results highlight that assessment of eligibility for ablative treatment should be improved, possibly by using novel visual examination aided-devices at primary care facilities, and that colposcopy services will still be needed even within screen-and-treat approaches, and that this need may be higher in populations with high prevalence of HIV.

In summary, the results of our study support the introduction of HPV testing in primary screening, even in limited resource settings. However, treatment capacity should be strengthened ahead of implementing HPV-and-treat with or without triage, particularly when targeting populations with high HIV prevalence, such as that of Eswatini.

Disclaimer

“Where authors are identified as personnel of the International Agency for Research on Cancer / World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy or views of the International Agency for Research on Cancer / World Health Organization.”

CRediT authorship contribution statement

Themba G. Ginindza: Conceptualization, Methodology, Writing – original draft, Writing – review & editing. Mathilde Forestier: Writing – review & editing. Maribel Almonte: Methodology, Writing – original draft, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.