The Prognostic Value of Plasma Cell-Free DNA Concentration in the Prostate Cancer: A Systematic Review and Meta-Analysis

Abstract

Objective: By virtue of largely disparate clinical outcomes of prostate cancer (PCA), there is a pressing need to search for useful biomarkers for PCA prognosis. Cell-free DNA (cfDNA) is a promising biomarker for detecting, monitoring, and predicting survival of prostate cancer (PCA). However, the utility of total cfDNA quantitation in PCA in clinical setting remains elusive. Here, we performed a thorough meta-analysis to assess the prognostic value of cfDNA concentration for patients with PCA. In addition, we tested the possibility of the combination of PSA and cfDNA test results to improve the prediction power in PCA prognosis.

Method and materials: More than six databases, including PubMed, Web of Science, Medline, PMC, EMBASE and the Cochrane Library were searched. Results yielded all eligible articles from the date of inception to June 30, 2020. Continuous, diagnostic, and prognostic variables in cfDNA in PCA were included in the meta-analysis by STATA.

Results: A total of 23 articles were enrolled in our meta-analysis: 69.6% (16/23) were related to diagnosis, and 56.5% (13/23) were related to prognosis. The pooled concentration of cfDNA in PCA patients was significantly higher than in the control group (SMD = 0.89, 95%CI = 0.53, 1.26), mirroring results for the prostate-specific antigen (PSA). For the detection test variables, the SROC with 95%CI was 0.87 (0.84-0.90) for cfDNA concentration. In terms of prognostic variables, the concentrations of cfDNA were significantly related with progression-free survival (PFS, logHR = 0.84 (95%CI0.39, 1.28) and overall survival [OS, log HR = 0.60 (95%CI0.29, 0.90)]. Lastly, the test showed no significant publication bias in the present meta-analysis, excluding the diagnostic meta-analysis.

Conclusions: The concentration of cell-free DNA is high in the prostate cancer patients. The present study substantiates the prognostic value of the cfDNA concentration. High concentration cfDNA correlates with poor disease outcome of CRPC. The study cohort with large sample size is needed to evaluate the prognosis value of cfDNA in the future. We also emphasized that combination of PSA and cf DNA quantitation is important in future large individual meta study.

Keywords: cell-free DNA; diagnosis; meta-analysis; prognosis; prostate cancer.

Figure 1

Flowchart of study inclusion and exclusion for meta-analysis.

Figure 2

The heterogeneity of cfDNA and PSA values between the PCA and control group (cfDNA, cell free DNA; PSA, prostate-specific antigen.) (A): Forest plots of the concentration of cfDNA; (B): Forest plots of PSA; (C): Correlation between cfDNA and PSA in Alexander’s and Vandekerkhov’s article.

Figure 3

Prognostic value of cfDNA in prostate cancer. (A): Forest plots for PFS; (B): Forest plots for PFS in CRPC (Castration-Resistant Prostate Cancer); (C): Forest plots for OS; (D): Forest plots for OS in CRPC.

Figure 4

The prognosis value of PSA, cfDNA and the combination of them in the Torquato’s cohort. (A): PSA for PFS; (B): CfDNA for PFS; (C): PSA + cfDNA for PFS; (D): Time-dependent ROC of them for predicting PFS; (E): PSA for OS; (F): CfDNA for OS; (G): PSA + cfDNA for OS; (H): Time-dependent ROC of them for predicting OS.

Figure 5

The assessment of potential publication bias in the meta-analysis. (A): The Begg’s test for the pooled results of difference between two groups. (B) The Deek’s test for the diagnostic meta-analysis. (C): The Begg’s test for the progression free survival meta-analysis.; (D): The Begg’s test for the overall survival meta-analysis; (E): The Begg’s test for the progression free survival meta-analysis in CRPC; (F): The Begg’s test for the overall survival meta-analysis in CRPC.