Unlocking the potential of ctDNA for predicting tumor prognosis

Dear Editor,

We read with interest the meta-analysis by Zhang et al.¹, which demonstrated that plasma circulating tumor DNA (ctDNA) could serve as a robust prognostic marker for esophageal cancer patients. Recently, ctDNA has emerged as a promising noninvasive biomarker for cancer detection. We commend the authors for their extensive efforts on this topic. Here, we highlight the potential of ctDNA-related indicators in predicting tumor prognosis and identify current technical limitations for future research focus.

1. ctDNA growth rate: The postoperative detection of ctDNA serves as a molecular confirmation of subclinical residual disease, with the level of ctDNA providing a proxy for tumor burden. A key advantage of ctDNA analysis lies in its capacity to serially assess ctDNA concentration, thereby enabling continuous monitoring for molecular recurrence and changes in tumor burden, reflecting treatment response. Henriksen et al.² were the first to report that 47% of recurrent colorectal cancer (CRC) patients exhibit a rapid ctDNA growth pattern, suggesting that ctDNA growth rates have the potential to identify patients who may benefit from immediate therapeutic intervention instead of waiting for clinical recurrence. These findings indicate that the quantification of ctDNA changes not only correlates with tumor activity but also holds predictive value for the prognosis of CRC patients, analogous to the dynamics of the protein biomarker PSA used for prognostic assessment in prostate cancer patients. Furthermore, the evaluation of ctDNA velocity or doubling time permits the further stratification of ctDNA-MRD-positive patients into distinct prognostic subgroups, offering a unique opportunity to guide clinical decision-making. Taken together, the ctDNA growth pattern may provide valuable guidance to clinicians regarding the urgency of intervention. Future research needs to address how to more rapidly and robustly determine ctDNA growth patterns, which would further enhance its clinical utility.

2. ctDNA fraction (ctDNA%): ctDNA%, representing the proportion of tumor-derived cell-free DNA, emerges as a prognostic factor across diverse cancer types and clinical settings. Its incorporation into commercial and laboratory tests is increasingly common, but its utility for risk stratification remains inadequately supported by evidence. Fonseca et al.³ discovered that ctDNA% strongly predicts overall survival, progression-free survival, and treatment response in metastatic castration-resistant prostate cancer (mCRPC), independent of therapeutic context, and outperforms established prognostic clinical factors. Furthermore, ctDNA% exhibited associations with multiple clinical indicators of tumor burden and disease aggressiveness, suggesting its potential as an actionable tool for risk stratification in mCRPC. However, it should be noted that optimal ctDNA% thresholds for risk stratification are likely to vary by cancer type. Therefore, the thresholds derived by Fonseca et al. for mCRPC should not be extrapolated in a tumor-agnostic manner. Future research is needed to establish optimal ctDNA% thresholds for risk stratification in different cancer types.

3. ctDNA methylation: ctDNA harbors a wealth of genetic information from cancer cells, among which methylation mutation partially reflects the presence and malignancy of in situ cancer. Epigenetic markers, such as methylated cytosines, have long been recognized as robust cancer biomarkers for tumor patients. A longitudinal analysis⁴ suggested that the presence of molecular residual disease could be effectively detected by ctDNA methylation as early as one month postoperatively, indicating a high risk of recurrence. These findings imply that the longitudinal assessment of ctDNA methylation may potentially accelerate risk stratification, molecular residual disease assessment, and recurrence detection. Due to its noninvasive nature and the high-quality cancer-related information obtained, ctDNA methylation holds significant promise and clinical value as a liquid biopsy approach. Future research could focus on exploring whether the combination of carcinoembryonic antigen and ctDNA methylation tests is linked to enhanced efficiency in risk stratification.

4. The combination of ctDNA with biomarkers reflecting the intrinsic characteristics of tumors: The classification of Consensus Molecular Subtypes (CMSs) is considered a comprehensive biomarker that reflects the intrinsic biological characteristics of CRC itself. Li et al.⁵ were the first to explore the prognostic value of the combination of ctDNA status, CMSs classification, and clinical risk. They found that the combination of these three factors is a much better predictive factor than each factor alone. These findings suggest that postoperative ctDNA indicating residual disease, combined with classification and clinical risk reflecting the intrinsic characteristics of tumors, may redefine the risk stratification of tumors and better predict relapse. Therefore, despite the significant prognostic value of ctDNA, the role of clinical and molecular features reflecting the intrinsic characteristics of tumors cannot be ignored in individualized treatment and monitoring. Future research should explore the important role of these combination biomarkers in optimizing individualized treatment and monitoring across different cancer types.

Despite the unique advantages of ctDNA detection as a liquid biopsy technique, it faces certain challenges, such as long detection cycles, high costs, and sensitivity issues (false positives and negatives). (i) Mutated ctDNA comprises only 0.01–1% of cfDNA, reaching about 10% in advanced cancers. The ‘Matthew effect’ in PCR limits the detection of low-frequency mutations, as dominant normal DNA fragments may overshadow mutated ones. (ii) Detection of genes with clonal hematopoietic mutations commonly causes false positives, necessitating analysis of both plasma DNA and leukocyte DNA to exclude these effects. (iii) ctDNA preservation and extraction face challenges like unstable yields due to freezing and hemolysis, even with specialized tubes. Thus, developing advanced methods to improve accuracy, reduce costs, and enhance accessibility remains crucial.

In conclusion, while ctDNA shows great promise for noninvasive cancer prognosis and monitoring, further research is needed to overcome technical limitations and enhance clinical application. Advancing detection methods and integrating ctDNA with other biomarkers will improve its predictive accuracy and utility in personalized cancer treatment.

Ethical approval

Not applicable.

Consent

Not applicable.

Source of funding

Not applicable.

Author contribution

Z.-Q.Z.: original draft conception and writing; Z.J.C.: writing – review and editing. All authors reviewed the manuscript.

Conflicts of interest disclosure

The authors declare no conflict of interest.

Research registration unique identifying number (UIN)

Not applicable.

Guarantor

Zhangkai J. Cheng.

Data availability statement

Not applicable for correspondence.

Provenance and peer review

Not applicable.