Comparison of Tumor Burden Quantification between Liquid Biopsy and CT

Academic Background

In the era of precision medicine, the validation of non-invasive biomarkers faces numerous challenges. Tumor burden, as a potential prognostic marker, has been demonstrated to hold significant value in various studies. Tumor burden refers to the total amount of cancer cells in the body and can be measured in multiple ways, including molecular markers, serum markers, tumor protein markers, and imaging markers. In recent years, liquid biopsy has emerged as a non-invasive method for monitoring tumor recurrence, assessing residual disease, and evaluating tumor burden. Liquid biopsy relies on circulating tumor DNA (ctDNA) released into the bloodstream from apoptotic or necrotic tumor cells. The tumor fraction (TF), which represents the percentage of ctDNA in cell-free DNA samples, is considered a non-invasive surrogate marker for studying disease burden dynamics.

However, the accuracy of liquid biopsy in assessing tumor burden has not been fully validated. Although ctDNA has shown significant value in molecular profiling and targeted therapy, its relationship with tumor burden requires further investigation. This study aims to evaluate whether TF from liquid biopsy can serve as an effective surrogate for tumor burden, particularly in the context of diverse metastatic cancers, using contrast-enhanced CT (CECT).

Source of the Paper

This paper was authored by Lama Dawi et al., with the research team from Gustave Roussy Institute, Université Paris-Saclay, France. The paper was published in November 2024 in the journal Radiology, titled Liquid Biopsy versus CT: Comparison of Tumor Burden Quantification in 1065 Patients with Metastases.

Research Process

Study Subjects and Data Collection

This retrospective single-center study included 1065 patients with metastatic cancer who underwent liquid biopsy and CECT between January 2021 and January 2023 at Gustave Roussy Institute. All patients provided written consent, and the study was approved by the Gustave Roussy Institutional Review Board.

Imaging Analysis

Contrast-enhanced CT scans with a slice thickness of 5 mm or less were used. All visible tumor lesions were manually annotated on the largest surface of the axial slice using the SPYD 2D annotation tool developed by Owkin. Lymph nodes, bone metastases, and other lesions were also included in the annotation. The volume of each lesion was calculated using a mathematical formula, and the total tumor volume (TTV) was the sum of all lesion volumes.

Liquid Biopsy Analysis

TF values from liquid biopsy were computed using the FoundationOne Liquid CDx assay (Roche), with TF >10% considered high. Liquid biopsies were categorized into three groups based on ctDNA levels: contributory, noncontributory, and failed.

Statistical Analysis

The correlation between TF and TTV was assessed using linear regression models and Spearman correlation coefficients. Receiver operating characteristic (ROC) curve analysis was used to determine the TTV cutoff. Additionally, the study analyzed the predictive ability of metastatic sites, histology type, and TTV on the contributory status of liquid biopsy.

Key Results

Patient Characteristics

The study included 1065 patients (median age 62 years, 537 females), with a total of 56,288 lesions annotated, primarily located in the lung (20,334), lymph nodes (11,651), and liver (10,277). Among the liquid biopsies, 763 were contributory, 254 were noncontributory, and 48 failed.

Correlation between TF and TTV

The linear regression model showed a weak correlation between TF and TTV (R² = 0.17; ρ = 0.41; p < 0.001). ROC curve analysis revealed an area under the curve (AUC) of 0.74 for TTV and TF categories, with an optimal TTV cutoff of 151 cm³ and a TF cutoff of 10%. The sensitivity was 57%, and the specificity was 80%.

Prediction of Liquid Biopsy Contributory Status

The AUC for TTV in predicting the contributory status of liquid biopsy was 0.71, with an optimal cutoff of 37 cm³. Liver lesion volume was significantly associated with contributory liquid biopsy in the validation cohort.

Conclusion

Although there is a correlation between TF from liquid biopsy and TTV from CT, TF does not accurately reflect TTV. The results suggest that relying on TF as a measure of tumor burden should be approached with caution, while imaging data (e.g., TTV and liver tumor volume) can optimize patient selection for liquid biopsy.

Research Highlights

1. Large-Scale Data: The study included 1065 patients and 56,288 lesions, providing robust statistical significance.
2. Multidimensional Analysis: The study not only analyzed the correlation between TF and TTV but also explored factors influencing the contributory status of liquid biopsy.
3. Clinical Application: The study offers valuable insights into the application of liquid biopsy in tumor burden assessment, particularly in optimizing patient selection for metastatic cancer.

Research Significance

This study provides new insights into the application of liquid biopsy in tumor burden assessment. Although the correlation between TF and TTV is weak, combining imaging data can lead to more accurate tumor burden evaluation, enabling more personalized treatment plans. Future research should further investigate the dynamics of tumor DNA shedding to improve the accuracy and reliability of liquid biopsy.

Additional Valuable Information

The study also noted that liver lesion volume showed a slightly stronger correlation with TF than TTV, possibly due to the liver being a common metastatic site or the higher release of ctDNA. Additionally, clonal hematopoiesis, as a confounding factor in liquid biopsy, may mask true tumor-derived mutations, which warrants further investigation.

This study provides important evidence for the combined application of liquid biopsy and imaging in tumor burden assessment, offering significant scientific and clinical potential.