ABSTRACT Colorectal cancer (CRC) is diagnosed at advanced stages in many low- and middle-income countries (LMICs). The lack of knowledge of CRC signs and symptoms by patients and community health practitioners frequently leads to delayed presentation with Stage 3-4 disease. This initial delay, paired with limited colonoscopy facilities, leads to prolonged diagnostic delays. The result is a 5-year mortality rate in LMIC up to 5 times higher than that in USA. An innovative solution to this problem could be an affordable, easily deployable, “point of care” molecular test to identify and prioritize patients likely to have a malignancy for expedited colonoscopy and pathology review leading to better outcomes. Continuing our established collaboration with our industrial partner, Cepheid, we propose to build on our strong published data on hypermethylated markers in CRC to develop an affordable, <3- hour, automated CRC-methylation detection blood test that analyzes a panel of five hypermethylated genes in cell free DNA from 1 ml of plasma. The proposed innovations could lead to a single-cartridge assay for quick CRC detection with a 3-fold reduction in cost. In Aim 1a, we will optimize a cartridge-less bisulfite DNA conversion method for plasma and test its efficiency in Patient Set 1 plasma (N= 20 malignant, 20 normal). In Aim 1b we will select one optimal 5-marker panel out of 20 CRC markers using DNA from FFPE samples from the U.S and Nigeria (N= 30 malignant, 30 benign), and one optimal “pan” set will be confirmed in plasma using U.S Patient Set 2 and Nigeria Set 3 (N=35 malignant, 35 benign). In Aim 1c, we will evaluate analytical performance of the CRC-MD assay. Intra-assay reproducibility will be assessed on multiple aliquots of U.S Patient Set 4 plasma (N=35 malignant, 35 benign). Inter-operator reproducibility will be determined using replicate aliquots of plasma from Patient Set 4 (N= 35 malignant, 35 benign). The goal of Aim 2a is to technically validate the CRC-MD assay using prospectively collected samples in Nigeria. We will first select a threshold in a Training set of plasma from Patient Set 4 (N=90 malignant, 90 benign) to optimally balance sensitivity and specificity, and validate performance of the selected threshold in a Test set of plasma from Patient Set 5 (N= 90 malignant, 90 benign). Accuracy (sensitivity, specificity, and positive- and negative-predictive value) of CRC- MD-based diagnosis to distinguish benign versus malignant disease will be measured using histopathological diagnosis of the lesion as the gold standard. Lastly, in Aim 2b, to determine whether the performance of the CRC-MD assay is altered by select patient characteristics, we will test its clinical accuracy among specific patient subgroups classified by age, sex, BMI, and tumor characteristics. Our prior success in developing automated cell-based/liquid biopsy assays with Cepheid has established the path ensuring an accurate and reliable test. This intervention could be cost saving by hastening colonoscopy for those who need it urgently, thus expediting detection and treatment of CRC in LMICs. This will save thousands of lives yearly. This study will also facilitate further development of the CRC-MD assay moving toward future commercialization and access globally.

PROJECT NARRATIVE For use in low- and middle-income countries, we propose to develop an affordable, automated colorectal cancer detection (GeneXpert® CRC-MD) assay that analyzes our published, hypermethylated DNA panel in plasma to rapidly detect malignancy. In collaboration with our industry partner, Cepheid, and our clinical team in Nigeria and Memorial Sloan Kettering Cancer Center, we will analytically and technically validate the assay for future product development. Innovative approaches to increase early detection and decrease treatment delays of CRC can save thousands of lives in the many resource-poor areas of the world.