Cervical cancer is the fourth most common cancer case and mortality among women, accounting for about 662,000 new cases and 349,000 deaths in 2022 globally (1). About 291 million women around the world carry Human papillomavirus (HPV) DNA, of whom 32% are infected either with HPV-16, HPV-18, or both (2). Effective screening and early treatment can effectively reduce morbidity and mortality associated with this disease (3).

The association between high-risk human papillomavirus (HR-HPV) infection and invasive cervical cancer (ICC) is well established. HPV genotypes 16 and 18 are responsible for more than 70% of ICC cases (4). ICC is the most frequently diagnosed cancer in 23 countries and the leading cause of cancer death in 36 countries, mostly found in Sub-Saharan Africa (5). In Ethiopia, the prevalence of ICC is 22 per 100,000, with 7,445 new cases and 5,338 deaths reported in 2020 (6). A study at Tikur Anbesa Specialized Hospital found that cervical cancer is the most common type of cancer among women (39.7%), followed by breast (18.3%) and ovarian cancer (7.1%) (7).

To eliminate cervical cancer, the World Health Organization (WHO) set a 2030 target of screening 70% of women with a high-performance test twice by the ages of 35 and 45 (1). Available screening methods include liquid-based cytology, HPV DNA PCR, Visual Inspection with Acetic Acid (VIA), and Visual Inspection with Lugol's Iodine (VILI), which are currently used in Ethiopia (6). HPV-DNA-based screening is a more effective tool for cervical cancer screening than the other alternatives (8), and WHO recommends HPV-based screening where resources are available. However, in low-resource settings, VIA remains the primary option, despite its inconsistent accuracy and high false-positive rates (20–50%), which lead to unnecessary referrals, increased costs, and overtreatment (7, 9).

HPV and cytology-based screening methods are rarely feasible in low- and middle-income countries (LMICs) due to cost, infrastructure needs, and the requirement for well-trained staff (8). This underscores the importance of affordable and effective screening options. In this context, the OncoE6™ Cervical Test presents a possible alternative, as it is a qualitative assay that detects elevated levels of the E6 oncoprotein expressed in cells infected by HPV-16 and HPV-18. High levels of E6, combined with clinical evaluation, suggest the presence or increased risk of precancerous or cancerous lesions (10). This study evaluated the diagnostic performance of the OncoE6™ test using histology as the reference standard. Additionally, the study compared the performance of the VIA screening method alongside OncoE6™ in detecting cancerous or precancerous lesions.

All women aged 25 to 75 years who visited the selected health facilities during the study period, met the inclusion criteria, and gave consent to participate were included. Women who agreed to provide samples for OncoE6 and histology but not for VIA were also included. Eligible participants had no prior diagnosis of cervical cancer/precancerous changes, no congenital abnormalities preventing cervical access, and were physically able to undergo routine screening. Pregnant women, those with hysterectomies and those treated for precancerous cervical lesions within the last six months were excluded.

Six health facilities were purposively selected and included in the study based on the availability of cervical biopsy services and their caseload. Furthermore, these health facilities had histology testing setups for pathology services and also had a higher number of cases identified as cervical cancer by histology testing for the last six months. The minimum sample size required for the sensitivity and specificity analysis was calculated by using PASS software (PASS 11, NCSS, LLC). The sample size calculation considered the disease prevalence of 20% (11), with 80% statistical power and a 95% confidence level for sensitivity and specificity. Based on the above assumptions, the minimum sample size required for this study was determined to be 245. In our study, the final sample size was 303.

In this study, each participant provided two cervical samples: the first for the OncoE6™ Cervical Test (Arbor Vita Corporation, Fremont, CA, USA) and the second for histological examination. The VIA technique was performed by experienced nurses or midwives, and results were documented prior to collecting the biopsy sample. This sampling order was maintained consistently throughout collection, storage, and transportation. Samples for the OncoE6™ test were stored at -20°C until testing. Histology specimens were processed according to established protocols of the respective health facility, ensuring proper fixation, sectioning, and staining to facilitate accurate microscopic evaluation. Histopathological examinations were performed at the collection facility based on their standards. Pathologists were blinded to the OncoE6 and VIA results. Slides were reviewed by two pathologists, and in cases of disagreement, a third pathologist was consulted, and the final diagnosis was reached by discussion. The OncoE6™ samples were transported to the Ethiopian Public Health Institute (EPHI) national HIV reference laboratory using a triple packaging system with dry ice and temperature monitoring in place. OncoE6™ test results were kept confidential and not disclosed to participants. They were recorded in study data collection forms with anonymous identification numbers, age, date, and site, without names and addresses.

Samples were lysed with the kit reagent and centrifuged at 10,000g for 10 minutes. A 200µL aliquot of the supernatant was incubated again with the lysis and conditioning solutions. After five minutes of incubation with solubilized detector monoclonal antibody (mAb) solution, samples migrated over the strip for 55 minutes, then underwent color development with Alkaline Phosphatase (AP) substrate for another 20 minutes, followed by visual inspection. The tests were performed by trained personnel experienced in serological assays, with biopsy results blinded from the tester. All procedures and interpretations followed the manufacturer's instructions (10).

In the current study, the histology test was used as the reference standard to evaluate the OncoE6™ rapid testing method. After tissue sample collection from the cervix, the sample underwent fixation, was embedded in paraffin wax and stained in the health facility's pathology centre. The Pathologists examine these stained sections under a microscope to identify any histopathological changes (12). All histology results were given back to each participant for further treatment and patient management.

The questionnaire was developed specifically for this study based on a review of previous literature and expert opinion, and used to collect information on socio-demographic, behavioral and gynecological characteristics of the study participants (Supplementary file 1). The information was completed by the health care provider and transferred to the study data manager every week.

This study evaluated the diagnostic performance of the OncoE6™ lateral flow test against histology, assessed VIA performance and examined factors associated with test positivity among women over 25 years old in Ethiopia. The study found that the OncoE6 test had moderate sensitivity (59.4%) and high specificity (98%), while VIA demonstrated high sensitivity (95.8%) but low specificity (26.4%). When used in parallel (OncoE6 OR VIA positive), sensitivity improved to 83.7% with a specificity of 71.4%. Age, education, and age at first sexual intercourse were significantly associated with test positivity.

In this study, the diagnostic performance of OncoE6 (sensitivity 59.4%, specificity 98%, PPV 94.0%, NPV 81.8%) was consistent with findings from Bangladesh (52% sensitivity, 97% specificity) (9) and Burundi (58.3% sensitivity, 97.9% specificity among CIN3 + HIV-positive women) (14). Our results also showed that OncoE6 positivity increased with disease severity, from 2.5% in CIN1 to 54.5% in CIN2/3 and 64.7% in invasive cancer or carcinoma in situ (Fig. 3). This pattern has been documented in other studies, where OncoE6 sensitivity rises for higher-grade lesions, particularly CIN3+ (9, 13–16). This increase is due to its detection of HPV16/18 E6 oncoproteins, which are more strongly associated with advanced lesions (17, 18).

Despite its specificity, OncoE6 may miss cases caused by non-HPV16/18 genotypes, which are prevalent in Ethiopia, where HPV16/18 account for 64–90% of cervical cancers, but non-16/18 types remain significant (18, 19). Lower sensitivity has also been reported elsewhere, including Zambia (31%) (16), Burundi (42.1% for CIN2+) (14), and China (42.4% for CIN2 + and 53.5% for CIN3+) (15), although specificity consistently exceeded 95% (14, 15). This variability may reflect differences in genotype distribution, regional E6 gene mutations (20), and test methodology (21). For example, Yang et al. (20) reported E6 gene mutations in HPV16/18 among cervical cancer patients in Southwest China. This suggests that genetic variation may affect assay performance.

In our study, VIA showed high sensitivity (95.8%) and low specificity (26.4%), which aligned with reports from Pakistan (93.5% sensitivity) (22) and the Multicentric Study for Cervical Cancer Screening and Triage with Human Papillomavirus Tests (ESTAMPA) study (85.9% sensitivity for CIN3+) (23). In Ethiopia, VIA remains the main screening tool where cytology and molecular testing are not accessible (24). However, a systematic review by Sauvaget et al. (25) reported a higher specificity of 92%. Studies in Congo Kinshasa (sensitivity 55.5%, specificity 64.6%) (26) and Nigeria (sensitivity 60%, specificity 94.4%) (27) showed considerable variation. These variations reflect provider training and experience (e.g., physicians versus nurses), patient age (identifying low-grade lesions can be challenging in elderly patients), and benign conditions that resemble aceto-white changes, such as atrophy, immature metaplasia, and HPV infection (26, 28–30).

When OncoE6 and VIA were combined in a parallel strategy, sensitivity and specificity reached 84.0% and 72.1%, respectively. This finding suggests that using both methods simultaneously could significantly improve the overall diagnostic performance, specifically by enhancing the sensitivity of OncoE6 and the specificity of VIA. Similar improvements with dual testing have been reported in other studies (31, 32). This approach could improve screening effectiveness in low-resource settings where advanced tests are not available.

In this study, age was significantly associated with disease severity. Women over 40 were more likely to develop higher-grade lesions. Married women were more likely to have CIN1, while a lower educational background, particularly women with no formal education, was linked to increased positivity with both VIA and OncoE6. Early sexual debut (15–19 years old) was also associated with VIA positivity. Comparable findings have been reported in Addis Ababa, Ethiopia (33), and older age, early sexual activity, long marital duration, and low education were associated with CIN (34–36). The risk of high-grade CIN and cancer is especially elevated in women over 55 due to persistent HR-HPV infections (37). Limited education may contribute to higher risk by reducing awareness of HPV transmission, reducing access to screening and vaccination, and restricting healthcare access, particularly in rural areas (28, 38).

Despite the moderate sensitivity of OncoE6 and the low specificity of VIA, our results showed a high specificity for the OncoE6 test and a high sensitivity for the VIA. In this study, we observed that using a parallel method (OncoE6 OR VIA positive) improved screening performance by increasing sensitivity. Integrating the OncoE6 lateral flow test with the existing VIA method offers a practical and effective way for early, on-site detection of cervical lesions. This is particularly valuable in settings where HPV DNA tests or liquid-based cytology are not available. However, this combined method needs more evaluation in larger, community-based studies in resource-limited settings.