Colon cancer study to be presented at ESMO 2024 shows excellent performance of the digital pathology biomarker Histotype Px® Colorectal in combination with ctDNA in analysis of over 1000 patients
OSLO – September 13, 2024 – DoMore Diagnostics, a leader in AI and deep-learning for precision medicine for cancer, is pleased to announce that a poster with clinical validation of Histotype Px® Colorectal (Histotype Px) used in combination with circulating tumor DNA (ctDNA) on Japanese colon cancer patients will be presented at the European Society of Medical Oncology (ESMO) Congress on September 16 in Barcelona, Spain. The study was conducted by the National Cancer Center Hospital East (Kashiwa, Japan) and the Institute for Cancer Genetics and Informatics (Oslo, Norway), in collaboration with researchers from Sapporo Medical University, Kyushu University, University College London Medical School, and the University of Oxford.
The study included 1082 R0 resected stage II (n=470) and stage III (n=612) patients from the CIRCULATE-Japan GALAXY trial, a large-scale prospective observational study, who were monitored for ctDNA minimal residual disease (MRD) status using Signatera™ (Natera, Inc.) and risk-stratified using Histotype Px AI-powered image analysis of tumor H&E pathology slides. To adjust for baseline covariates across patients who did or did not receive adjuvant chemotherapy (ACT), inverse probability of treatment weighting (IPTW) was used. The primary endpoint was time to recurrence (TTR).
Histotype Px classified 537 (50%), 302 (28%), and 243 (22%) as low, intermediate, and high-risk groups, respectively. Among MRD-negative patients, Histotype Px successfully stratified into distinct Low (94.5% no recurrence at 24 months after surgery), Intermediate (89.7% no recurrence) and High-risk (82.6% no recurrence) groups. MRD-negative patients classified as Intermediate and High-risk by Histotype Px had significantly shorter TTR relative to patients classified as Low-risk by Histotype Px (HR=1.88, p=0.025 and HR=2.96, p<0.001, respectively).
Importantly, there was no discernible difference in time to recurrence between those who received ACT and those who did not in Histotype Px Low-risk and MRD-negative patients (HR=1.01, p=0.98)). In the MRD-negative and Low-risk group, the 24-month no recurrence rate was 94.1% (95% CI: 89.8%-98.7%) for patients who received ACT vs 94.4% (95% CI: 91.4%-97.5%) for patients who did not receive ACT. Thus, combining Histotype Px with ctDNA MRD status may be predictive of ACT treatment benefit in this patient population.
Torbjørn Furuseth, MD, Co-Founder and CEO at DoMore Diagnostics, commented, "These groundbreaking results clearly demonstrate that Histotype Px is not only a reliable risk classification tool on a stand-alone basis, but also that it can significantly improve the risk classification in combination with ctDNA, and become a new paradigm of truly personalized ACT treatment decisions. It is particularly important to predict ACT treatment benefit in this colon cancer patient population, where there is a large medical need for better tools to select patients who are likely to benefit from ACT and spare the patients who will have excellent outcomes without ACT. Histotype Px is a digital biomarker with short turn-around time, which makes it particularly valuable to enable early clinical decision making."
Poster presentation details: Novel clinical decision support (CDS) system optimizing adjuvant chemotherapy (ACT) for colorectal cancer (CRC) by integrating deep learning and circulating tumor DNA (ctDNA) molecular residual disease (MRD): GALAXY histotyping.
Speaker: Dr. Yoshiaki Nakamura (Kashiwa, Japan, Chiba)
554P @ 16 September, 2024, 12 PM-1 PM
Link to Full Abstract
Poster presented:
About DoMore Diagnostics
DoMore Diagnostics uses artificial intelligence to make personalized treatment decisions simple and accessible for all cancer patients. Its unique digital biomarkers predict patient outcomes from routine tumor tissue slides and can be seamlessly integrated into pathologists’ existing workflow. The lead product Histotype Px® Colorectal is a CE-IVDD marked outcome prediction marker for stage II and III colorectal adenocarcinoma that informs the decision of whether to provide adjuvant chemotherapy following surgical resection of the tumor.
About Colorectal Cancer and Histotype Px® Colorectal
Colorectal cancer is the third most common and second most deadly cancer but has seen little innovation in diagnostics and treatment the last decades. Adjuvant chemotherapy (ACT) may be provided after primary surgery to prevent the disease from recurring, however overtreatment is prevalent with up to 85% of patients with stage II and stage III not benefitting from ACT and only suffering side effects.
Histotype Px® is an advanced deep learning algorithm that analyzes digital histology slides, separating patients into distinct Low, Intermediate and High-risk groups to guide ACT treatment decisions. Data previously published in The Lancet and The Lancet Oncology showed that the test can accurately predict survival outcomes in colorectal cancer patients. Histotype Px® Colorectal was developed to provide clinicians with valuable information to guide personalized treatment decisions and improve patient outcomes.
For more information, please visit:
The Lancet publication.
The Lancet Oncology publication.
Contact:
Torbjørn Furuseth, MD
torbjorn.furuseth@domorediagnostics.com