An accurate, affordable, and transferable assay for minimally invasive detection of esophageal squamous cell carcinoma (ESCC) in Africa and the United States

Esophageal squamous cell carcinoma (ESCC), comprising 90% of esophageal cancer (EC) worldwide, has a 5-year fatality rate of 90% in low- to moderate-income countries, or LMICs, primarily due to an inability to diagnose the disease before it becomes fatal. Indeed, the value of low-cost, accurate, and assessable diagnostics is central to the current Notice of Special Interest (NOSI) we are responding to, which highlights EC as one of the four “highest-priority cancers” (NCI SBIR/STTR Program Announcement Regarding Cancer Prevention, Diagnosis, and Treatment Technologies for Low- Resource Settings, NOT-CA-21-062). To address this pressing need, we developed an affordable, accessible, safe, and accurate diagnostic strategy applying our novel ESCC-specific methylated DNA biomarkers to esophageal cells retrieved using our noninvasive, swallowable sponge-capsule device. In our Phase I studies (n = 94), we developed, optimized, and validated a multivariate biomarker panel, which accurately diagnosed ESCC in sponge-capsule samples (sensitivity = 92.3%, specificity = 86.7%; AUC = 0.94). Building on this success, our Phase II project will perform a robust independent validation of this assay and achieve commercialization in the USA while implementing assessments and feedback toward early commercialization in Uganda, an LMIC. In Aim 1, we will transfer the EsoFind assay from Johns Hopkins University to Capsulomics and perform extensive analytical validation, simplifying workflow to improve throughput, efficiency, and accuracy while reducing cost and complexity. In Aim 2, a cohort of 300 new patients from the US and Uganda will be studied to develop rigorous models using advanced biostatistical methods, including earlier ESCC cases. We will first generate data in a training set (n=200 patients), building new models based on our SBIR Phase I biomarker panel (cg20655070, SLC35F1, TAC1, ZNF132, and ZNF542) but also including clinical features (e.g., age and gender). We will then independently validate this new algorithm in a test set (n=100 patients) for validation. Finally, Aim 3 will focus on establishing the commercialization of EsoFind in the USA, including submission for FDA clearance of our sponge-based Cell Extraction Device (CED) and preparation for its launch as a Laboratory Developed Test (LDT). Additionally, we will garner user feedback in Uganda, while developing a plan for transfer and marketing as a model for use in other LMICs. Thus, this Phase II SBIR project will streamline, expand, validate, and achieve early commercialization of this safe, affordable, accurate, noninvasive streamlined assay to diagnose ESCC in the USA and Uganda. If successful, our strategy has the potential to be applied in larger, longer-term asymptomatic at-risk population-based screening studies to improve ESCC detection and survival.