Non-invasive multi-cancer detection using DNA hypomethylation of LINE-1 retrotransposons

Marc Michel,Maryam Heidary,Anissa Mechri,Kevin Da Silva,Marine Gorse,Victoria Dixon,Klaus von Grafenstein,Charline Bianchi,Caroline Hego,Aurore Rampanou,Constance Lamy,Maud Kamal,Christophe Le Tourneau,Mathieu Sene,Ivan Bieche,Cecile Reyes,David Gentien,Marc-Henri Stern,Olivier Lantz,Luc Cabel,Jean-Yves Pierga,Francois-Clement Bidard,Chloe-Agathe Azencott,Charlotte Proudhon

DOI: https://doi.org/10.1101/2024.01.20.23288905

2024-10-30

Abstract:Purpose: The detection of circulating tumor DNA, which allows non-invasive tumor molecular profiling and disease follow-up, promises optimal and individualized management of patients with cancer. However, detecting small fractions of tumor DNA released when the tumor burden is reduced remains a challenge. Experimental Design: We implemented a new highly sensitive strategy to detect base-pair resolution methylation patterns from plasma DNA and assessed the potential of hypomethylation of LINE-1 retrotransposons as a non-invasive multi-cancer detection biomarker. The DIAMOND (Detection of Long Interspersed Nuclear Element Altered Methylation ON plasma DNA) method targets 30-40,000 young L1 scattered throughout the genome, covering about 100,000 CpG sites and is based on a reference-free analysis pipeline. Results: Resulting machine learning-based classifiers showed powerful correct classification rates discriminating healthy and tumor plasmas from 6 types of cancers (colorectal, breast, lung, ovarian, gastric cancers and uveal melanoma including localized stages) in two independent cohorts (AUC = 88% to 100%, N = 747). DIAMOND can also be used to perform copy number alterations (CNA) analysis which improves cancer detection. Conclusions: This should lead to the development of more efficient non-invasive diagnostic tests adapted to all cancer patients, based on the universality of these factors.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to detect circulating tumor DNA (ctDNA) by non - invasive methods in the early stage of cancer or when the tumor burden is low, so as to achieve the early diagnosis of multiple types of cancers. Specifically, the researchers have developed a new high - sensitivity strategy - the DIAMOND (Detection of Long Interspersed Nuclear Element Altered Methylation ON plasma DNA) method, which is used to detect the hypomethylation pattern of LINE - 1 retrotransposons in plasma DNA as a biomarker for non - invasive multi - cancer detection.

### Research Background and Objectives
- **Background**: Circulating tumor DNA (ctDNA) exists in the plasma of cancer patients and can be used for non - invasive tumor molecular analysis and disease monitoring. However, when the tumor burden is low, it is still challenging to detect the trace amounts of tumor DNA in plasma.
- **Objective**: To develop a highly sensitive method, using the hypomethylation pattern of LINE - 1 retrotransposons as a biomarker, to achieve non - invasive detection of multiple types of cancers.

### Experimental Design
- **Method**: The researchers developed the DIAMOND method. This method targets 30 - 40,000 young LINE - 1 retrotransposons, covering about 100,000 CpG sites, and is analyzed based on a reference - genome - free analysis pipeline.
- **Sample**: The study used two independent cohorts, with a total of 747 samples, covering 6 types of cancers (colorectal cancer, breast cancer, lung cancer, ovarian cancer, gastric cancer, and choroidal melanoma).

### Main Results
- **Classification Performance**: The machine - learning - based classifier shows strong classification ability and can distinguish healthy and tumor plasma samples in two independent cohorts, with an AUC value of 88% to 100%.
- **Copy Number Variation Analysis**: DIAMOND data can also be used for copy number variation (CNA) analysis, further improving the accuracy of cancer detection.
- **Detection of Different Cancer Types**: This method performs well in multiple cancer types, especially in metastatic colorectal cancer, breast cancer, stage III ovarian cancer, and local gastric cancer, with an AUC value close to 100%.

### Conclusions
- **Significance**: The DIAMOND method provides an efficient and universal non - invasive diagnostic test for all cancer patients. Based on the universality of these factors, it is expected to play an important role in future clinical applications.

### Formula and Technical Details
- **Methylation Detection**: The DIAMOND method achieves methylation detection at single - base resolution through PCR amplification and deep sequencing, combined with a reference - genome - free analysis pipeline.
- **Data Analysis**: Machine - learning algorithms are used to train classification models, integrating individual CpG sites and molecular - level methylation patterns, improving the sensitivity and specificity of detection.

In conclusion, this paper aims to solve the existing technical barriers in current non - invasive cancer detection, especially for early - stage cancers and cases with low tumor burden, by developing and validating the DIAMOND method.