The global incidence and mortality rates of cancer are experiencing a rapid rise. In the year 2020, cancer-related deaths constituted almost 20% of all deaths. The costly treatments for cancer impose a substantial economic burden on patients, whereas early cancer detection presents an opportunity to enhance control over cancer progression, improve survival rates, and lower healthcare expenses.
Nevertheless, the current majority of early cancer screening methods have notable drawbacks, including invasiveness and a tendency for high false-positive rates. These limitations can easily result in overdiagnosis and overtreatment.
In recent years, there has been a significant surge of interest in investigating the functions of ctDNA alterations concerning early cancer detection and the identification of tumor origins. Although ctDNA holds immense promise in these aspects, the accuracy of detecting tumor occurrence requires further refinement due to several factors. These factors encompass the relatively low abundance of ctDNA in the bloodstream of early-stage cancer patients, the distinct heterogeneity of ctDNA features across various cancer types, and the need for deeper sequencing to enhance detection sensitivity.
The research team has developed a multimodal detection method called SPOT-MAS (Screening for the Presence of Tumor by DNA Methylation and Size). This method comprehensively analyzes epigenomics, fragmentomics, DNA copy number, and cfDNA end sequences, enabling the diagnosis of diseases and localization of tumor lesions in a single screening process.
The final results of the study revealed that SPOT-MAS achieved promising sensitivity (72.4%) and specificity (97.0%) in detecting five common cancer types using shallow-depth sequencing. Furthermore, the method demonstrated a 70% accuracy in predicting the tissue of origin. Notably, utilizing multiple features in the testing model outperformed relying on a single feature alone. Among the cancers tested, SPOT-MAS showed the highest efficacy in detecting liver cancer, accurately identifying approximately 90% of the cases. However, its performance was relatively less effective for breast cancer, where it could only detect about half of the cases. These findings are consistent with previous research, indicating that liver tumors release a substantial amount of ctDNA, while breast cancer sheds a comparatively lower amount of ctDNA.Overall, the study highlights the potential of SPOT-MAS as a valuable tool for early cancer detection and provides insights into the varying ctDNA dynamics among different cancer types.
Extensive case-control studies have demonstrated that SPOT-MAS, with its distinct integration of cfDNA feature multimodal analysis and advanced machine learning algorithms, can effectively and affordably detect and locate various types of cancer. These significant findings offer essential supporting evidence for the integration of SPOT-MAS into clinical practices as an adjunctive cancer screening method for high-risk populations.
