ShanghaiAnd the June 15, 2022 /PRNewswire/– Burning Rock (NASDAQ:BNR) presents the final results from the THUNDER multicenter case-control study (Non-overlapping Detection of EaRly-stage Cancer, NCT04820868) in a poster titled “Non-overlapping detection of multiple cancers by generalized sequencing of null DNA methylation” From Cells (THUNDER): Studies of Development and Independent Verification” at the 2022 American Society of Clinical Oncology (ASCO) Annual Meeting. This is a comprehensive review of clinical performance following the four-year THUNDER study.
Through a rigorous “three-stage” design: tag detection and plate validation, model training and validation (retrospective), and independent validation (prospective), the THUNDER study evaluated the performance of the previously described cfDNA methylation-based technology ELSA-seq in early detection And localization of six types of cancers in the lung, colon, liver, esophagus, pancreas and ovary.
In the study, a custom panel of 161,984 CpG sites was generated and validated by public and in-house (cancer: n = 249; non-cancerous: n = 288) methylome data, respectively. The study used cancer-specific methylation regions to detect cancer where it can distinguish cancer signals from non-cancerous ones, and identify the source (tissue of origin) of the detected cancer signal.
In the training and model validation phase (retrospective), cfDNA samples were collected from 1,693 participants (cancer: n = 735; non-cancerous: n = 958) and approximately three-quarters of the samples were used to create two of the multiple cancer screening blood test models. For blood (MCDBT-1/2) with different cutoffs (cancer: n = 399; non-cancerous: n = 626). About a quarter of the samples were used to build a verification set for a typical blood test to detect multiple cancers (cancer: n = 301; non-cancer: n = 123). In the independent (prospective) validation stage, both models were blindly validated on a prospectively registered independent validation set (cancer: n = 473; non-cancerous: n = 473) to more rigorously test the stability of the high-dimensional modeling.
The study showed that sensitivity MCDBT-1 was 69.1% with a specificity of 98.9% in the independent validation set. MCDBT-2 produced a higher sensitivity (75.1%) and slightly lower specificity (95.1%) compared to MCDBT-1.
The study also showed that ELSA-seq can accurately trace the original tissue (identifying the location of the cancer in the body). For MCDBT-1, the accuracy of the highest prediction origin (TPO) was 89.7% in the training set, and 82.8% in the validation set, 83.2% in the independent verification group, respectively. The MCDBT-2 model also showed a high accuracy of TPO, with the accuracy of TPO reaching 79.4% in the independent validation set.
Finally, an intercept model was applied using indicators such as cancer incidence, cancer stage distribution at diagnosis, five-year survival rate and tumor progression rate at China, to infer phase transformation and survival advantages and to demonstrate the potential public health benefits of ELSA-seq in real-world application. Based on the intercept model computation, these two models achieved sensitivities of 70.6% and 77.5% in detecting the six cancers, respectively. MCDBT-1 reduced the incidence of late stage by 38.7%-46.4% (shift from stage III-IV towards stage I-II), and increased the 5-year survival rate by 33.1%-40.4%.
The THUNDER study laid the groundwork for extending non-invasive cfDNA methylation detection to more cancer types in the future and once again demonstrated the excellent performance and broad application prospects of ELSA-seq in the early detection of multiple cancers. Burning Rock looks forward to collaborating with medical experts to advance clinical validation and development of multiple cancer screening products, thereby discovering more early-stage cancers to improve cancer patient survival, and reduce the social and economic burden stemming from cancer.
about burning rocks
Burning Rock Biotech Limited (NASDAQ:BNR), whose mission is to protect life through science, is focused on the application of Next Generation Sequencing (NGS) technology to the field of precision oncology. Its business consists of 1) NGS-based treatment selection testing for late-stage cancer patients, 2) Global Pharmaceutical Services for biomarker detection and development of concomitant diagnoses, and 3) NGS-based early cancer detection, which has gone beyond proof-of-concept R&D into stage Clinical verification.
For more information please contact [email protected]
source of burning rock