Validation of a Novel Cervical cancer screening and triage algorithm - The human papillomavirus-Automated Visual Evaluation (PAVE-MALAWI) study in Blantyre and Chiradzulu districts
Study design
Prospective diagnostic accuracy study
Context
Cervical cancer remains a leading cause of cancer-related mortality among Malawian women. Despite national efforts, incidence and mortality rates are still among the world’s highest. Since 2017, Doctors without Borders (MSF) has partnered with Malawi’s Ministry of Health to implement a comprehensive cervical cancer program. Recent integration of HPV self-sampling, thermal ablation, and decentralized care aims to increase screening and treatment coverage and reduce preventable deaths.
Objectives
Multi-site primary objective
- To assess the diagnostic performance of the PAVE triage approach for HPV-positive women by comparing its sensitivity with that of the Standard of Care (SOC) triage approach (VIA or cytology), at equal specificity, using histopathological findings of CIN2 or higher (CIN2+) as the reference threshold.
Malawi Site-specific objectives
Primary objective
- To assess the difference in sensitivity of the PAVE-triage approach for triaging all HPV-positive women compared to the Standard of Care (SOC) triage approach (VIA) at equal specificity
Secondary Objectives
1. To determine the clinical accuracy of the PAVE approach for triaging
- HPV-positive women compared to the SOC approach.
- HPV-positive WLWH compared to the SOC approach.
- by restricted partial genotyping of HPV without using visual inspection/imaging in all HPV positive women and among HPV positive WLWH
2. To assess the difference in clinical sensitivity of the PAVE-triage approach among women living with HIV (WLWH) for triaging HPV-positive compared to the SOC triage approach (VIA) at equal specificity to classify HPV-positive WLWH
3. To describe the three study population groups (pop 1,2 and 3) screened for cervical cancer (demographic- and clinical characteristics, HIV-status, proportion HPV-positive, proportion precancer)
4. To estimate the sensitivity and specificity of the overall PAVE approach (hrHPV test & then triage by HPV partial genotyping and AVE) among the total screened population (pop 1)
5. To determine one- and two-year treatment outcomes (residual or recurrent CIN2+), HPV reinfection or persistence, as well as performance of tests of control (hrHPV test, Partial genotyping, PAVE) against truth standard histopathology among women who underwent thermoablation, overall and by HIV status.
Our role
- PAVE study site
- Coordination for the Malawi site
Partners
- Consortium PAVE
- Fondation MSF
- Ministry of health, Malawi
Methodology
Prospective diagnostic accuracy study of the PAVE screening approach, with an observational cohort assessing the outcomes of diagnosed precancer and treated with thermoablation.
Study population
- Population 1: 25-49-year-old women screened by MSF cervical cancer screening activities in the selected health centres, and outreach campaigns and meeting the eligibility criteria during the study inclusion period
- Population 2: The women among the population 1 who test positive for human papillomavirus (any of the 13 oncogenic genotypes detected by ScreenFire test)
- Population 3: The women among the population 2 who are treated by thermoablation (nested prospective observational cohort sub study)
Sample size
- Estimated eligible women to be screened (group 1): N=10 000
- Screened women positive for HPV (group 2): estimated N= ~2500 (Needing the screening of an estimated number of women: 10 000)
- HPV positive women who undergo thermoablation (group 3): estimated N= ~500
Study site
Ten health centres and their catchment areas supported by MSF cervical cancer screening activities in Blantyre and Chiradzulu District, Malawi. (MSF Malawi study site is one of the 10 study sites of the international multicentric study)
Interim results
Self-sampled HPV test was done to 6,076 women, 43.3% of the recruited have known HIV infection. The overall prevalence of HPV infection was 40.7%, slightly higher among WLWH (52.5%). 788 pathology diagnosis were confirmed by external review. Of the CIN2+ cases (N=116), 49.1% were HPV16/18/45-positive, else 35.3% were HPV31/33/35/52/58-positive. The sensitivity of the triage methods to detect histologically confirmed CIN2+ was 21.5% for VIA, 53.5% for HPV16/18/45 genotyping, and 60.5% for the combined approach. No significant difference in triage sensitivity was seen by HIV status.
Conclusion
HPV16/18/45 genotyping, alone and in combination with visual evaluation, showed improved detection of precancer among HPV-positives as compared to VIA alone. Better risk stratification is expected when using extended genotyping (including HPV31/33/35/52/58) and/or AI-assisted visual evaluation. This preliminary finding urges the adaptation of WHO and national recommendations to the unique context of High HPV and HIV prevalence, and HPV genotype distribution.
