POINT of CARE: PrOviding an Innovative New Triage and Treatment strategy for Cervical cAncer scReening Efficiency

PROJECT SUMMARY Although preventable, cervical cancer is a leading cause of mortality in low- and middle-income countries (LMIC). Infection with high-risk human papillomavirus (HPV) can lead to cervical precancer (cervical intraepithelial neoplasia grade 2 or higher, or CIN2+). The World Health Organization (WHO) endorses both

ablation and excision treatment of CIN2+. Excision requires specialized equipment and high-level providers and is generally not used in LMIC. Thermal ablation (TA) is commonly used to treat CIN2+ in LMIC. Conventional TA is performed with a non-portable desktop device that requires electricity. The WHO endorsed

TA in 2019 based largely on data from the desktop device, which uses an endocervical probe for every treatment. Cure rates with the desktop device were 97% after 18-years of follow-up, likely due to using an endocervical probe being used. Endocervical probes can treat the entire squamocolumnar junction (SCJ),

where cervical cancer arises, versus a flat probe alone. Handheld, inexpensive, portable, battery-operated TA devices were recently created for use in LMIC. None of the current handheld devices have an endocervical probe, and cure rates in LMIC are significantly lower at 83%. Liger Medical (Lehi, UT) manufactures a

prototype that is a combination handheld TA device and digital colposcope, the IRISTM, for use in LMIC. The IRISTM will be modified to include an endocervical probe. In addition, the IRISTM will integrate a machine learning (ML) algorithm, automated visual evaluation (AVE). When AVE is combined with HPV genotyping,

preliminary data has shown an area under the Receiver Operating Characteristics (ROC) curve (AUC) of 0.89 with a limited total extreme misclassification rate of 3.4% (CIN2+ as normal or normal as CIN2+).The purpose of this proposal is to modify the current IRISTM device and clinically validate the changes through the following

aims: Specific Aim 1: To modify the IRISTM prototype device for optimal use in LMIC by a) modifying the existing probes and b) adapting it to have a machine learning component, automated visual evaluation (AVE); Specific Aim 2: To clinically validate the IRISTM endocervical probes by estimating the cure rates of CIN2+ one year

after treatment; and Specific Aim 3: To perform a retrospective validation of the AVE algorithm on IRISTM images utilizing HPV genotyping and histopathologic disease status. The IRISTM device has the potential to be a global game changer in the field of cervical cancer prevention. Successful modification of the IRISTM will result in more effective precancer treatment, less

over-treatment, and increased efficiency of screen-and-treat approaches. This project will bring effective point-of-care triage and treatment to cervical cancer elimination programs in LMIC.