1. Introduction
Hematopoietic cell transplantation (HCT) remains a cornerstone of curative therapy for a broad spectrum of hematologic malignancies, including acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), lymphomas, multiple myeloma (MM), and myelodysplastic syndromes (MDS) [ 1 ]. Over recent decades, advances in conditioning regimens, donor selection, graft engineering, infection prevention, and supportive care have substantially improved transplant outcomes and reduced non-relapse mortality (NRM). Nevertheless, disease relapse remains the leading cause of treatment failure after HCT, affecting approximately 30–50% of patients depending on disease subtype, disease status at transplantation, conditioning intensity, and transplant modality [ 2 , 3 , 4 ]. Once overt relapse occurs following HCT, prognosis is generally poor, underscoring the importance of sensitive post-transplant monitoring strategies capable of identifying residual disease before clinical recurrence develops.
Conventional measurable residual disease (MRD) monitoring methods, including multiparameter flow cytometry (MFC) and allele-specific polymerase chain reaction (PCR)-based assays, have improved the detection of residual malignant cells and refined post-transplant risk stratification. However, these approaches have important limitations. Bone marrow-based monitoring is invasive and susceptible to sampling bias, particularly in diseases characterized by spatial heterogeneity or patchy marrow involvement. In addition, conventional assays may fail to fully capture clonal evolution and emerging resistant subclones during post-transplant disease progression [ 5 , 6 ]. Variability in assay platforms, sensitivity thresholds, and interpretation criteria across centers also limits standardization and broader clinical applicability. These limitations have stimulated increasing interest in minimally invasive liquid biopsy approaches for post-HCT surveillance.
Circulating tumor DNA (ctDNA), which consists of tumor-derived DNA fragments released into the bloodstream through apoptosis, necrosis, or active secretion by malignant cells, has emerged as a promising biomarker for real-time molecular disease monitoring [ 7 ]. Tumor-specific ctDNA assays may detect molecular relapse weeks to months before overt hematologic or radiographic recurrence becomes clinically evident [ 8 , 9 ]. Unlike conventional single-site tissue or marrow sampling, ctDNA potentially provides a more comprehensive representation of tumor heterogeneity and evolving clonal architecture [ 10 ]. Serial ctDNA monitoring has therefore generated substantial interest as a tool for dynamic MRD assessment, relapse prediction, and longitudinal disease surveillance across multiple hematologic malignancies.
Importantly, circulating nucleic acid approaches in the HCT setting are biologically heterogeneous and should not be considered interchangeable methodologies. Tumor-specific ctDNA refers to circulating tumor-derived nucleic acids identified through mutation-based, rearrangement-based, or patient-informed assays. In contrast, total cell-free DNA (cfDNA) quantification reflects overall circulating DNA burden without tumor specificity and may be influenced by tissue injury, inflammation, endothelial damage, infection, or conditioning-related toxicity. Methylation-based or whole-genome bisulfite sequencing (WGBS) approaches evaluate genome-wide cfDNA signatures that may reflect tissue injury and immune activation rather than direct molecular relapse [ 11 , 12 ]. Similarly, cell-free RNA (cfRNA) profiling may provide insights into inflammatory and immune-related processes, whereas chimerism assays assess donor–recipient hematopoietic composition rather than tumor-derived DNA directly [ 13 ]. Distinguishing these methodologies is essential when interpreting the growing liquid biopsy literature in HCT populations.
In the post-transplant setting, ctDNA-guided monitoring may ultimately support risk-adapted management strategies through earlier identification of impending relapse [ 14 , 15 ]. Potential interventions triggered by molecular relapse detection could include tapering of immunosuppression, donor lymphocyte infusion, targeted therapies, bispecific antibodies, or cellular immunotherapies administered at a lower disease burden [ 16 ]. However, despite increasing interest in ctDNA for post-HCT surveillance, the available literature remains fragmented. Published studies vary substantially in underlying diseases, transplant modalities, assay platforms, ctDNA targets, timing of sample collection, and endpoint definitions [ 17 , 18 ]. Moreover, prior reports have frequently combined biologically distinct liquid biopsy methodologies—including total cfDNA, methylation-based profiling, and chimerism monitoring—within the same analytical framework, complicating interpretation of the prognostic significance of tumor-specific ctDNA.
To address these gaps, we conducted a systematic review and meta-analysis evaluating the prognostic significance of tumor-specific ctDNA and tumor-informed cfDNA monitoring in patients undergoing HCT for hematologic malignancies. We specifically aimed to quantify associations between ctDNA positivity and relapse, overall survival (OS), and relapse-free/progression-free survival (RFS/PFS), while separately synthesizing non-tumor-specific circulating nucleic acid approaches narratively because of their biological and methodological heterogeneity. Through this approach, we sought to provide a clearer assessment of the current evidence supporting ctDNA-based MRD monitoring in the transplant setting and to identify key priorities for future clinical translation and research.
2. Materials and Methods
2.1. Study Design and Reporting Standards
This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines [ 19 ]. The review protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO; CRD420261392100).
Eligible studies were evaluated according to the PICO framework as follows: Population: Patients of any age with hematologic malignancies undergoing allogeneic or autologous HCT, including AML, ALL, malignant lymphoma, MDS, MM, and related disorders. Index prognostic factor: Detection of tumor-specific ctDNA or tumor-informed/tumor-associated cfDNA for MRD assessment at any time point before or after HCT. Comparator: Patients with undetectable ctDNA/cfDNA at the same assessment time point. Outcomes: Primary outcomes included relapse, cumulative incidence of relapse (CIR), RFS, and PFS. Secondary outcomes included OS, graft-versus-host disease (GVHD), graft failure, transplant-related mortality (TRM), and other post-transplant complications. Study design: Prospective or retrospective observational cohort studies reporting hazard ratios (HRs) with 95% confidence intervals (CIs), or providing sufficient data for HR reconstruction. Reviews, editorials, conference abstracts without sufficient data, case reports, and non-English studies were excluded.
2.2. Information Sources and Search Strategy
A comprehensive literature search was conducted from database inception through 30 November 2025, using the following electronic databases: PubMed/MEDLINE; Embase; Web of Science; EBSCO; Cochrane Central Register of Controlled Trials (CENTRAL).
Additional supplementary searches were performed using Google Scholar and manual screening of reference lists from relevant articles and reviews. Clinical trial registries, including ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform (ICTRP), were also searched for ongoing or unpublished studies.
The search strategy combined controlled vocabulary (MeSH and Emtree terms) with free-text keywords related to: Circulating tumor DNA and liquid biopsy; Measurable residual disease; Hematopoietic cell transplantation; Relapse and survival outcomes.
Representative search terms included: (“circulating tumor DNA” OR ctDNA OR “cell-free DNA” OR cfDNA OR “liquid biopsy” OR “minimal residual disease” OR MRD) AND (“hematopoietic cell transplantation” OR HSCT OR allo-HCT OR ASCT OR “bone marrow transplantation”) AND (relapse OR survival OR prognosis).
Search strategies were adapted appropriately for individual databases.
2.3. Study Selection and Data Extraction
Two independent reviewers (D.T.D. and J.L.E.) screened titles and abstracts according to predefined eligibility criteria. Potentially relevant articles subsequently underwent full-text review. Disagreements regarding study eligibility were resolved through discussion and consensus or adjudication by a third reviewer (A.T.). Data extraction was performed using a standardized pilot-tested form. Extracted variables included: Study characteristics (author, year, country, design); Patient demographics and disease subtype; Transplant characteristics; ctDNA/cfDNA assay methodology; Sampling time points; Outcome measures; HRs and corresponding 95% CIs.
When multiple time points were reported, clinically relevant landmark assessments (e.g., pre-HCT, day +30, day +90/+100 post-HCT) were preferentially extracted.
When available, multivariable-adjusted HRs were preferentially extracted over univariable estimates. If HRs were not directly reported, they were reconstructed from Kaplan–Meier curves using the method described by Tierney et al. [ 13 ].
Our systematic search identified 849 records across electronic databases and supplementary searches. After removal of 241 duplicates, 608 unique studies underwent title and abstract screening. Forty-eight full-text articles were assessed for eligibility, of which 10 observational cohort studies (6 retrospective and 4 prospective) met inclusion criteria for the final systematic review ( Figure 1 ). Studies eligible for quantitative meta-analysis were further restricted to those reporting extractable HRs for tumor-specific ctDNA/cfDNA outcomes.
2.4. Risk of Bias Assessment
Risk of bias was independently assessed by two reviewers using the Quality In Prognosis Studies (QUIPS) tool [ 19 ]. This instrument evaluates six domains: Study participation; Study attrition; Prognostic factor measurement; Outcome measurement; Study confounding; Statistical analysis and reporting.
Each domain was classified as low, moderate, or high risk of bias. Discrepancies were resolved through consensus discussion.
2.5. Data Synthesis and Statistical Analysis
All statistical analyses were performed using R software (version 4.3.2) with the meta and metafor packages.
Only studies evaluating tumor-specific ctDNA or tumor-informed/tumor-associated cfDNA MRD assays with extractable HRs for relapse, OS, RFS, or PFS were eligible for quantitative synthesis. Studies evaluating total cfDNA quantity, methylation-based cfDNA profiling, cfRNA, or chimerism-only monitoring were synthesized narratively because these assays do not directly measure tumor-specific ctDNA and demonstrated substantial biological and methodological heterogeneity.
Pooled HRs and corresponding 95% CIs were calculated using a generic inverse-variance random-effects model (DerSimonian and Laird method) to account for anticipated heterogeneity in disease populations, transplant modalities, assay methodologies, and sampling schedules.
Statistical heterogeneity was assessed using Cochran’s Q test and the I 2 statistic. An I 2 value > 50% was considered indicative of substantial heterogeneity. Because relatively few studies contributed to several pooled analyses, pooled estimates and heterogeneity statistics were interpreted cautiously.
Pre-specified subgroup analyses according to disease subtype, HCT modality, timing of ctDNA assessment, and assay platform were planned. However, formal subgroup analyses were limited by the small number of eligible studies contributing to each pooled endpoint. Sensitivity analyses were planned by sequential exclusion of studies judged to have high overall risk of bias.
Publication bias was evaluated visually using funnel plots and statistically using Egger’s regression test when sufficient studies were available for pooled analysis. Formal publication bias assessment was limited by the small number of studies contributing to several pooled outcomes.
3. Results
3.1. Characteristics of Included Studies
The 10 included studies encompassed a total of 883 patients, with sample sizes ranging from 10 to 177 participants per study. Studies were conducted internationally, with most originating from the United States (n = 5) [ 20 , 21 , 22 , 23 , 24 ]. Additional studies were conducted in Japan [ 25 ], Denmark [ 26 ], Turkey [ 27 ], South Korea [ 28 ], and Germany [ 29 ]. The spectrum of underlying hematologic malignancies was broad and included AML, ALL, MDS, MM, and various malignant lymphomas. Overall, six studies employed retrospective observational cohort designs [ 20 , 21 , 22 , 25 , 26 , 27 ], while four were prospective cohort or pilot studies [ 23 , 24 , 28 , 29 ]. Eight studies focused on allo-HCT [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 29 ], whereas two evaluated autologous HCT populations [ 20 , 28 ].
The methodologies used for circulating nucleic acid assessment were heterogeneous and included targeted next-generation sequencing (NGS), droplet digital PCR (ddPCR), WGBS, spectrophotometric total cfDNA quantification, and real-time quantitative PCR (RQ-PCR) chimerism assays. Because these approaches capture biologically distinct processes, studies were categorized into: (1) tumor-specific ctDNA studies, (2) tumor-associated cfDNA/cfDNA-MRD studies, and (3) non-tumor-specific approaches, including total cfDNA quantification, methylation-based cfDNA profiling, and chimerism-based monitoring. The main characteristics of included studies are summarized in Table 1 .
3.2. Survival Outcomes
Eight studies evaluated associations between circulating nucleic acid biomarkers and OS, RFS, or PFS. Because included studies assessed biologically distinct biomarker approaches, pooled quantitative analyses were restricted to tumor-specific ctDNA/cfDNA studies with extractable hazard ratios, whereas total cfDNA, methylation-based profiling, and chimerism studies were interpreted qualitatively.
Among tumor-specific ctDNA studies, Dhakal et al. (2022) demonstrated that ctDNA positivity at 3 months following autologous HCT in MM independently predicted inferior PFS (HR = 5.6; 95% CI: 1.8–17.0; p = 0.0003) [ 20 ]. Similarly, Nakamura et al. (2019) reported that ctDNA positivity at 3 months post-allogeneic HCT was strongly associated with inferior OS in AML/MDS (HR = 8.04; 95% CI: 1.66–38.88; p < 0.01) [ 19 ]. Herrera et al. (2016) found that post-transplant ctDNA detection in lymphoma patients was associated with significantly inferior PFS (HR = 3.9; 95% CI: 1.6–9.5; p = 0.003) [ 15 ].
Among tumor-associated cfDNA studies, Pasca et al. (2023) demonstrated that tumor-informed cfDNA-MRD positivity at day +90 after allo-HCT in myeloid malignancies was strongly associated with inferior RFS and OS, with a 2-year cumulative incidence of relapse of 76% among cfDNA-positive patients compared with 21% among negative patients ( p < 0.001) [ 24 ]. Patel et al. (2025) likewise reported that tumor-associated cfDNA positivity was associated with increased mortality and relapse risk following allogeneic HCT in AML (HR for OS = 5.4; 95% CI: 2.5–11.8; p < 0.0001) [ 14 ].
Waterhouse et al. (2022) demonstrated that longitudinal patient-specific cfDNA monitoring enabled early relapse detection, including extramedullary relapse, and was associated with inferior survival outcomes [ 29 ]. In contrast, Yegin et al. (2020), who evaluated total cfDNA quantity rather than tumor-specific ctDNA, did not identify a statistically significant association between cfDNA levels and OS [ 27 ].
Across studies eligible for quantitative synthesis, pooled meta-analysis demonstrated that ctDNA/cfDNA positivity was significantly associated with inferior OS (pooled HR = 7.45; 95% CI: 4.11–13.48) and inferior RFS/PFS (pooled HR = 4.46; 95% CI: 2.22–8.97). Although statistical heterogeneity was low, interpretation was limited by the relatively small number of studies contributing to each pooled endpoint. A summary of survival outcome data is presented in Table 2 .
3.3. Prediction of Graft-Versus-Host Disease and Other Complications
Beyond relapse surveillance, emerging evidence suggests that broader cell-free nucleic acid profiling approaches may also provide clinically relevant information regarding non-relapse complications following HCT. Cheng et al. (2022) utilized low-coverage whole-genome bisulfite sequencing of plasma cfDNA to simultaneously evaluate multiple post-transplant complications, including GVHD, infection, relapse, and graft failure [ 23 ]. Their findings demonstrated that methylation-based cfDNA profiling could capture tissue injury and immune-related signals associated with post-HCT complications, supporting the potential role of broader liquid biopsy approaches beyond tumor-specific relapse monitoring.
3.4. Engraftment and Transplant-Related Mortality
Several studies additionally explored associations between circulating nucleic acid biomarkers and transplant-related complications, including TRM, NRM, and graft-related outcomes. Yegin et al. (2020) reported an inverse relationship between pre-transplant total cfDNA levels and transplant-related complications, with lower total cfDNA levels associated with increased complications and mortality [ 27 ]. Because this study measured total cfDNA quantity by spectrophotometry rather than tumor-specific ctDNA, the biological interpretation of these findings likely differs from mutation-informed MRD approaches.
Haugaard et al. (2019) demonstrated that highly sensitive RQ-PCR chimerism monitoring predicted relapse approximately seven months before morphological recurrence in pediatric ALL/AML patients, while also showing associations with transplant-related outcomes [ 26 ]. Similarly, Herrera et al. (2016) reported associations between post-transplant ctDNA positivity and increased non-relapse mortality among lymphoma patients undergoing allo-HCT [ 21 ] ( Table 3 ).
3.5. Relapse Prediction
All included studies evaluated relapse-related outcomes, although substantial heterogeneity existed in biomarker methodology and study design ( Table 4 ). To minimize biological heterogeneity, quantitative pooling was restricted to tumor-specific ctDNA/cfDNA studies with extractable hazard ratios, while total cfDNA, methylation-based profiling, and chimerism studies were interpreted qualitatively.
Among tumor-specific ctDNA studies, Dhakal et al. (2022) reported that ctDNA positivity at 3 months post-autologous HCT strongly predicted relapse in multiple myeloma and outperformed multiparameter flow cytometry MRD assessment [ 20 ]. Nakamura et al. (2019) demonstrated that serial ctDNA measurements at 1 and 3 months after allo-HCT were highly predictive of relapse in AML/MDS, with increasing ctDNA kinetics providing particularly high sensitivity for impending recurrence [ 25 ]. Herrera et al. (2016) found that post-transplant ctDNA detection was associated with significantly increased disease progression risk in lymphoma patients [ 21 ]. Kim et al. (2024) further demonstrated that ctDNA monitoring identified high-risk lymphoma patients who may benefit from intensified upfront transplant strategies [ 28 ].
Among tumor-associated cfDNA studies, Pasca et al. (2023) showed that day +90 cfDNA-MRD positivity was associated with a markedly increased cumulative incidence of relapse in myeloid malignancies [ 24 ]. Patel et al. (2025) similarly reported strong associations between tumor-associated cfDNA positivity and relapse following allogeneic HCT in AML [ 22 ]. Waterhouse et al. (2022) demonstrated that longitudinal patient-specific cfDNA monitoring enabled early detection of relapse, including extramedullary disease [ 29 ].
In contrast, Yegin et al. (2020) evaluated total cfDNA concentration rather than tumor-specific ctDNA and reported that lower total cfDNA levels were associated with increased relapse risk in AML [ 27 ]. Cheng et al. (2022), whose primary focus was post-transplant complication profiling rather than MRD surveillance, reported that methylation-based cfDNA profiling could also capture relapse-associated signals [ 23 ]. Haugaard et al. (2019) demonstrated that chimerism-based MRD monitoring using highly sensitive RQ-PCR predicted relapse approximately seven months before morphological recurrence in pediatric leukemia [ 26 ].
Across studies eligible for quantitative synthesis, the pooled HR for relapse or disease progression associated with ctDNA/cfDNA positivity was 12.57 (95% CI: 4.59–34.46), supporting a strong association between molecular positivity and adverse post-transplant oncologic outcomes.
3.6. Risk of Bias Assessment
Risk of bias assessment using the QUIPS tool demonstrated an overall moderate risk of bias across included studies. Most studies showed low risk of bias in outcome measurement and statistical reporting domains, reflecting the use of objective clinical endpoints and generally transparent statistical analyses.
The most frequent sources of potential bias involved the prognostic factor measurement and study confounding domains. Variability in ctDNA assay methodologies, timing of sample collection, ctDNA thresholds, and plasma processing protocols contributed to moderate methodological heterogeneity across studies. In addition, several retrospective studies lacked comprehensive multivariable adjustment for established prognostic factors, including disease risk, cytogenetic abnormalities, conditioning intensity, donor type, or GVHD status.
Studies evaluating total cfDNA quantity or non-tumor-specific approaches were generally judged to have a higher risk of bias due to limitations in biological specificity and assay standardization. Retrospective studies additionally demonstrated increased susceptibility to attrition and participation bias compared with prospective cohort studies. A detailed summary of the QUIPS assessment is presented in Table 5 .
Risk of Bias Domains (QUIPS): Study Participation; Study Attrition; Prognostic Factor Measurement; Outcome Measurement; Study Confounding; Statistical Analysis and Reporting.
Interpretation: Most studies demonstrated low risk of bias in outcome measurement and statistical reporting domains. Moderate or high-risk judgments were primarily driven by variability in ctDNA assay methodologies, inconsistent sampling schedules, retrospective study design, incomplete adjustment for established prognostic variables, and limited sample sizes.
3.7. Meta-Analysis of ctDNA Positivity and Survival Outcomes
Only studies evaluating tumor-specific ctDNA or tumor-informed/tumor-associated cfDNA with extractable hazard ratios were included in the quantitative meta-analysis. Studies evaluating total cfDNA quantity, methylation-based cfDNA profiling, cfRNA, or chimerism-only monitoring were excluded from pooled analyses because of substantial biological and methodological heterogeneity.
The summarized HR data used in the pooled analyses are presented in Table 6 . Meta-analysis demonstrated statistically significant associations between ctDNA/cfDNA positivity and adverse clinical outcomes across all primary endpoints.
For overall survival, ctDNA/cfDNA positivity was associated with significantly inferior OS (pooled HR = 7.45; 95% CI: 4.11–13.48; p < 0.001) ( Figure 2 ). Similarly, pooled analysis demonstrated significantly inferior RFS/PFS among ctDNA/cfDNA-positive patients (pooled HR = 4.46; 95% CI: 2.22–8.97; p < 0.001) ( Figure 3 A). Finally, ctDNA/cfDNA positivity was strongly associated with relapse or disease progression (pooled HR = 12.57; 95% CI: 4.59–34.46; p < 0.001) ( Figure 3 B).
Although statistical heterogeneity across pooled analyses was low, interpretation should remain cautious because only a limited number of studies contributed to each endpoint, thereby reducing the power of heterogeneity statistics to detect true between-study variability.
4. Discussion
This systematic review and meta-analysis demonstrate that tumor-specific ctDNA positivity is consistently associated with increased relapse risk and inferior survival outcomes following HCT across multiple hematologic malignancies. Quantitative synthesis revealed statistically significant associations between ctDNA/cfDNA positivity and relapse or disease progression, OS, and RFS/PFS. Collectively, these findings support the potential role of ctDNA-based MRD monitoring as a promising non-invasive biomarker for post-transplant molecular surveillance and risk stratification.
Our findings are consistent with prior studies supporting the clinical utility of ctDNA-based MRD assessment in hematologic malignancies and extend these observations specifically to the HCT setting [ 17 , 30 ]. One of the most clinically relevant features of ctDNA monitoring is the ability to detect molecular relapse weeks to months before overt hematologic or radiographic recurrence becomes clinically apparent. Several included studies demonstrated that serial ctDNA kinetics may provide particularly informative prognostic signals, with rising ctDNA levels preceding clinical relapse by substantial intervals [ 31 ]. This capability is especially relevant after allo-HCT, where early molecular relapse detection could theoretically permit pre-emptive interventions prior to overt disease recurrence [ 32 , 33 , 34 ].
Importantly, substantial biological and methodological heterogeneity exists among circulating nucleic acid biomarkers evaluated in the transplant literature. Tumor-specific ctDNA assays, tumor-informed cfDNA-MRD approaches, total cfDNA quantification, methylation-based cfDNA profiling, cfRNA analysis, and chimerism-based monitoring are biologically distinct methodologies that should not be interpreted interchangeably. Tumor-specific ctDNA primarily reflects residual malignant clones and disease kinetics, whereas total cfDNA levels may additionally reflect tissue injury, inflammation, endothelial damage, infection, or conditioning-related toxicity. Likewise, methylation-based cfDNA and cfRNA approaches may capture immune activation, GVHD, or organ injury rather than direct molecular relapse. To address this heterogeneity, the present meta-analysis was restricted to tumor-specific ctDNA and tumor-informed/tumor-associated cfDNA studies with extractable hazard ratios, whereas non-tumor-specific approaches were synthesized narratively. Nevertheless, conceptual overlap among these biomarker categories remains an important limitation when interpreting the broader liquid biopsy literature in HCT.
Beyond relapse surveillance, several studies suggested that broader circulating nucleic acid profiling strategies may provide clinically relevant information regarding post-transplant complications. Cheng et al. demonstrated that methylation-based cfDNA profiling using whole-genome bisulfite sequencing could simultaneously inform GVHD, infection, relapse, and graft failure [ 23 ]. Similarly, highly sensitive chimerism monitoring enabled early identification of relapse risk in pediatric leukemia populations [ 26 ]. These findings suggest that future post-HCT surveillance strategies may ultimately integrate tumor-specific ctDNA with immune- and graft-related biomarkers to achieve more comprehensive multidimensional monitoring of transplant recipients [ 35 ].
Although statistical heterogeneity across pooled analyses was low, these findings should be interpreted cautiously. Only a limited number of studies contributed to each pooled endpoint, thereby reducing the statistical power of heterogeneity metrics such as the I 2 statistic to detect true between-study variability. Moreover, substantial clinical heterogeneity remained across studies, including differences in underlying malignancies, transplant modalities, assay platforms, ctDNA targets, sampling schedules, and definitions of molecular positivity. Accordingly, the apparent consistency of pooled estimates does not necessarily imply methodological uniformity, and the magnitude of pooled hazard ratios should be interpreted within the context of relatively small observational datasets.
Another major challenge limiting broader clinical implementation is the absence of assay standardization. Included studies used heterogeneous technologies, including targeted NGS, ddPCR, patient-informed assays, multiplex PCR-NGS, methylation-based profiling, and total cfDNA quantification. Considerable variability also existed in plasma processing protocols, sequencing depth, ctDNA thresholds, timing of sample collection, and definitions of ctDNA positivity. These differences complicate cross-study comparisons and currently preclude the establishment of universally accepted thresholds for clinical decision-making. International collaborative efforts aimed at harmonizing ctDNA collection, processing, analysis, and reporting standards will therefore be essential for future clinical translation.
The potential clinical implications of ctDNA-guided monitoring after HCT are considerable. Earlier molecular relapse detection could support risk-adapted interventions such as tapering of immunosuppression, donor lymphocyte infusion, targeted therapies, bispecific antibodies, or cellular immunotherapies administered at lower disease burden [ 6 , 7 , 8 , 9 , 10 , 32 ]. In lymphoma and multiple myeloma, ctDNA-guided surveillance may also help identify patients who could benefit from earlier intervention prior to overt clinical relapse. However, despite these promising observations, the current evidence base remains predominantly retrospective, single-center, and observational. Consequently, ctDNA should presently be considered an emerging investigational biomarker rather than a fully established standard-of-care monitoring tool. Prospective interventional studies demonstrating improved clinical outcomes through ctDNA-guided therapeutic strategies remain critically needed.
Several limitations of this review should be acknowledged. First, only a limited number of studies contributed to each pooled endpoint, reducing the precision of pooled estimates and limiting the reliability of heterogeneity assessments. Second, most included studies were observational cohorts conducted at single institutions, increasing susceptibility to selection bias and residual confounding. Third, substantial methodological variability existed in assay platforms, timing of sampling, ctDNA targets, and endpoint definitions. Fourth, despite separating tumor-specific ctDNA studies from total cfDNA, methylation-based profiling, and chimerism-based approaches in the revised analyses, conceptual overlap among circulating nucleic acid biomarkers remains a source of complexity within the literature. Finally, although the review protocol was prospectively registered, publication bias could not be robustly assessed because relatively few studies contributed to several pooled analyses.
Despite these limitations, this study provides one of the first focused quantitative syntheses specifically evaluating tumor-specific ctDNA monitoring following HCT across multiple hematologic malignancies. The consistent associations observed between ctDNA positivity and adverse oncologic outcomes support continued investigation of ctDNA as a clinically relevant prognostic biomarker in the transplant setting. Future priorities should include multicenter prospective validation studies, assay harmonization initiatives, standardized sampling strategies, and randomized ctDNA-guided interventional trials to determine whether earlier molecular relapse detection can ultimately improve patient outcomes after HCT.
5. Conclusions
In conclusion, tumor-specific ctDNA positivity is consistently associated with increased relapse risk and inferior survival outcomes following hematopoietic cell transplantation. These findings support the potential role of ctDNA-based MRD monitoring as a promising non-invasive biomarker for post-transplant risk stratification and molecular surveillance. However, further assay standardization, prospective multicenter validation, and ctDNA-guided interventional trials are required before routine clinical implementation can be recommended.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Author Contributions
D.T.D.: data curation, analysis, and writing the original draft. H.T. and A.T.: data curation and analysis. J.L.E.: Conceptualization, analysis, writing—review and editing. All authors have read and agreed to the published version of the manuscript.
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
All data generated or analyzed during this study are included in this published article.
Conflicts of Interest
The authors declare no conflicts of interest.
Abbreviations
The following abbreviations are used in this manuscript: Disease and Clinical Terms ALL Acute lymphoblastic leukemia AML Acute myeloid leukemia CLL Chronic lymphocytic leukemia CMML Chronic myelomonocytic leukemia DLBCL Diffuse large B-cell lymphoma HL Hodgkin lymphoma MDS Myelodysplastic syndromes MM Multiple myeloma MPN Myeloproliferative neoplasm NHL non-Hodgkin lymphoma Transplant-Related Terms Allo-HCT Allogeneic hematopoietic cell transplantation GVHD Graft-versus-host disease GVL Graft-versus-leukemia HCT Hematopoietic cell transplantation NRM non-relapse mortality TRM Transplant-related mortality Molecular and Laboratory Terms cfDNA Cell-free DNA cfRNA Cell-free RNA ctDNA Circulating tumor DNA ddPCR Droplet digital polymerase chain reaction MFC Multiparameter flow cytometry MRD Minimal residual disease NGS Next-generation sequencing PCR Polymerase chain reaction RQ-PCR Real-time quantitative polymerase chain reaction TA-cfDNA Tumor-associated cell-free DNA WGBS Whole-genome bisulfite sequencing Statistical and Methodological Terms AUC Area under the curve CI Confidence interval CIR Cumulative incidence of relapse HR Hazard ratio NR Not reported OS Overall survival PFS Progression-free survival PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses QUIPS Quality in Prognosis Studies RFS Relapse-free survival Other APC Article processing charges CAR-T Chimeric antigen receptor T-cell
References
1. Greco R. Ruggeri A. McLornan D.P. Snowden J.A. Alexander T. Angelucci E. Averbuch D. Bazarbachi A. Hazenberg M.D. Kalwak K. Indications for haematopoietic cell transplantation and CAR-T for haematological diseases, solid tumours and immune disorders: 2025 EBMT practice recommendations Bone Marrow Transplant. 2025 60 1499 1525 10.1038/s41409-025-02701-3 40926035 PMC12583170
2. Rieger M.J. Stolz S.M. Müller A.M. Schwotzer R. Nair G. Schneidawind D. Manz M.G. Schanz U. Haploidentical transplant with posttransplant cyclophosphamide vs matched related and unrelated donor transplant in acute myeloid leukemia and myelodysplastic neoplasm Bone Marrow Transplant. 2023 58 1121 1129 10.1038/s41409-023-02042-z 37479752 PMC10555825
3. Akhmedov M. Espinoza J.L. Addressing the surge of infections by multidrug-resistant Enterobacterales in hematopoietic cell transplantation Blood Rev. 2024 68 101229 10.1016/j.blre.2024.101229 39217051
4. Sauerer T. Velázquez G.F. Schmid C. Relapse of acute myeloid leukemia after allogeneic stem cell transplantation: Immune escape mechanisms and current implications for therapy Mol. Cancer 2023 22 180 10.1186/s12943-023-01889-6 37951964 PMC10640763
5. Heuser M. Freeman S.D. Ossenkoppele G.J. Buccisano F. Hourigan C.S. Ngai L.L. Tettero J.M. Bachas C. Baer C. Béné M.-C. 2021 Update on MRD in acute myeloid leukemia: A consensus document from the European LeukemiaNet MRD Working Party Blood 2021 138 2753 2767 10.1182/blood.2021013626 34724563 PMC8718623
6. Sorror M.L. The use of prognostic models in allogeneic transplants: A perspective guide for clinicians and investigators Blood 2023 141 2173 2186 10.1182/blood.2022017999 36800564 PMC10273168
7. Wan J.C.M. Massie C. Garcia-Corbacho J. Mouliere F. Brenton J.D. Caldas C. Pacey S. Baird R. Rosenfeld N. Liquid biopsies come of age: Towards implementation of circulating tumour DNA Nat. Rev. Cancer 2017 17 223 238 10.1038/nrc.2017.7 28233803
8. Martínez-Castedo B. Camblor D. Martín-Arana J. Carbonell-Asins J. García-Micó B. Gambardella V. Huerta M. Roselló S. Roda D. Gimeno-Valiente F. Minimal residual disease in colorectal cancer. Tumor-informed versus tumor-agnostic approaches: Unraveling the optimal strategy Ann. Oncol. 2025 36 263 276 10.1016/j.annonc.2024.12.006 39675560
9. Sorbini M. Carradori T. Togliatto G.M. Vaisitti T. Deaglio S. Technical Advances in Circulating Cell-Free DNA Detection and Analysis for Personalized Medicine in Patients’ Care Biomolecules 2024 14 498 10.3390/biom14040498 38672514 PMC11048502
10. Kurtz D.M. Esfahani M.S. Scherer F. Soo J. Jin M.C. Liu C.L. Newman A.M. Dührsen U. Hüttmann A. Casasnovas O. Dynamic Risk Profiling Using Serial Tumor Biomarkers for Personalized Outcome Prediction Cell 2019 178 699 713.e19 10.1016/j.cell.2019.06.011 31280963 PMC7380118
11. Martello M. Solli V. Mazzocchetti G. Solimando A.G. Bezzi D. Taurisano B. Kanapari A. Poletti A. Borsi E. Armuzzi S. High level of circulating cell-free tumor DNA at diagnosis correlates with disease spreading and defines multiple myeloma patients with poor prognosis Blood Cancer J. 2024 14 208 10.1038/s41408-024-01185-6 39609411 PMC11605000
12. Zou H. Liu W. Wang X. Wang Y. Wang C. Qiu C. Liu H. Shan D. Xie T. Huang W. Dynamic monitoring of circulating tumor DNA reveals outcomes and genomic alterations in patients with relapsed or refractory large B-cell lymphoma undergoing CAR T-cell therapy J. Immunother. Cancer 2024 12 e008450 10.1136/jitc-2023-008450 38443094 PMC11146396
13. Frank M.J. Hossain N.M. Bukhari A. Dean E. Spiegel J.Y. Claire G.K. Kirsch I. Jacob A.P. Mullins C.D. Lee L.W. Monitoring of Circulating Tumor DNA Improves Early Relapse Detection After Axicabtagene Ciloleucel Infusion in Large B-Cell Lymphoma: Results of a Prospective Multi-Institutional Trial J. Clin. Oncol. 2021 39 3034 3043 10.1200/JCO.21.00377 34133196 PMC10166351
14. Zhou X. Lang W. Mei C. Ren Y. Ma L. Jiang L. Ye L. Xu G. Luo Y. Liu L. Serial monitoring of circulating tumour DNA on clinical outcome in myelodysplastic syndromes and acute myeloid leukaemia Clin. Transl. Med. 2023 13 e1349 10.1002/ctm2.1349 37491737 PMC10368807
15. Shastri T. Trabolsi A. Arumov A. Schatz J.H. Bispecific Antibodies in Hematologic Malignancies: Attacking the Frontline BioDrugs 2025 39 793 814 10.1007/s40259-025-00735-z 40751114 PMC12354547
16. Tan X. Yan H. Chen L. Zhang Y. Sun C. Clinical Value of ctDNA in Hematological Malignancies (Lymphomas, Multiple Myeloma, Myelodysplastic Syndrome, and Leukemia): A Meta-Analysis Front. Oncol. 2021 11 632910 10.3389/fonc.2021.632910 33747954 PMC7970179
17. Lee J.S. Cho E.H. Kim B. Hong J. Kim Y.-G. Kim Y. Jang J.-H. Lee S.-T. Kong S.-Y. Lee W. Clinical Practice Guideline for Blood-based Circulating Tumor DNA Assays Ann. Lab. Med. 2024 44 195 209 10.3343/alm.2023.0389 38221747 PMC10813828
18. Tierney J.F. Stewart L.A. Ghersi D. Burdett S. Sydes M.R. Practical methods for incorporating summary time-to-event data into meta-analysis Trials 2007 8 16 10.1186/1745-6215-8-16 17555582 PMC1920534
19. Hayden J.A. Van Der Windt D.A. Cartwright J.L. Côté P. Bombardier C. Assessing bias in studies of prognostic factors Ann. Intern. Med. 2013 158 280 286 10.7326/0003-4819-158-4-201302190-00009 23420236
20. Dhakal B. Sharma S. Balcioglu M. Shchegrova S. Malhotra M. Zimmermann B. Billings P.R. Harrington A. Sethi H. Aleshin A. Assessment of Molecular Residual Disease Using Circulating Tumor DNA to Identify Multiple Myeloma Patients at High Risk of Relapse Front. Oncol. 2022 12 786451 10.3389/fonc.2022.786451 35186734 PMC8848740
21. Patel V. Kabat M. Ip A. Kaur S. Suh H.C. Cho C. Vesole D. Donato M. Albitar M. Rowley S.D. Peripheral Blood Tumor Associated Cell-Free DNA Testing as a Predictor for Relapse Postallogeneic Stem Cell Transplant for Acute Myelogenous Leukemia Transpl. Cell Ther. 2025 31 1010 1019 10.1016/j.jtct.2025.08.010 40865683
22. Herrera A.F. Kim H.T. Kong K.A. Faham M. Sun H. Sohani A.R. Alyea E.P. Carlton V.E. Chen Y. Cutler C.S. Next-generation sequencing-based detection of circulating tumour DNA After allogeneic stem cell transplantation for lymphoma Br. J. Haematol. 2016 175 841 850 10.1111/bjh.14311 27711974 PMC5123935
23. Cheng A.P. Cheng M.P. Loy C.J. Lenz J.S. Chen K. Smalling S. Burnham P. Timblin K.M. Orejas J.L. Silverman E. Cell-free DNA profiling informs all major complications of hematopoietic cell transplantation Proc. Natl. Acad. Sci. USA 2022 119 e2113476118 10.1073/pnas.2113476118 35058359 PMC8795552
24. Pasca S. Guo M.Z. Wang S. Stokvis K. Shedeck A. Pallavajjala A. Shams C. Pallavajjala R. DeZern A.E. Varadhan R. Cell-free DNA measurable residual disease as a predictor of postallogeneic hematopoietic cell transplant outcomes Blood Adv. 2023 7 4660 4670 10.1182/bloodadvances.2023010416 37276081 PMC10448421
25. Nakamura S. Yokoyama K. Shimizu E. Yusa N. Kondoh K. Ogawa M. Takei T. Kobayashi A. Ito M. Isobe M. Prognostic impact of circulating tumor DNA status post-allogeneic hematopoietic stem cell transplantation in AML and MDS Blood 2019 133 2682 2695 10.1182/blood-2018-10-880690 30936070
26. Haugaard A.K. Madsen H.O. Marquart H.V. Rosthøj S. Masmas T.N. Heilmann C. Müller K.G. Ifversen M. Highly sensitive chimerism detection in blood is associated with increased risk of relapse after allogeneic hematopoietic cell transplantation in childhood leukemia Pediatr. Transpl. 2019 23 e13549 10.1111/petr.13549 31313439
27. Yegin Z.A. Can F. Gökçen S. Sadioğlu R.E. Özkurt Z.N. İlHan Ç. Yağcı M. The Impact of Pre-transplant Cell-free DNA Levels on Leukemia Relapse and Transplant-related Complications in Allogeneic Hematopoietic Stem Cell Transplant Recipients Balk. Med. J. 2020 37 138 143 10.4274/balkanmedj.galenos.2020.2019.8.25 31970974 PMC7161624
28. Kim J. Le T.M. Lee D. Nguyen H.D.T. Cho H.J. Sohn S.K. Kim J.G. Jeong S.-Y. Ham J.Y. Jeong J.Y. Circulating-tumor DNA Assessment in Diffuse Large B-cell Lymphoma to Determine Up-front Stem Cell Transplantation: A Pilot Study In Vivo 2024 38 372 379 10.21873/invivo.13448 38148060 PMC10756469
29. Waterhouse M. Pennisi S. Pfeifer D. Scherer F. Zeiser R. Duyster J. Bertz H. Finke J. Duque-Afonso J. Monitoring of Measurable Residual Disease Using Circulating DNA after Allogeneic Hematopoietic Cell Transplantation Cancers 2022 14 3307 10.3390/cancers14143307 35884368 PMC9323743
30. Abdulmawjood B. Roma-Rodrigues C. Fernandes A.R. Baptista P.V. Liquid biopsies in myeloid malignancies Cancer Drug Resist. 2019 2 1044 1061 10.20517/cdr.2019.88 35582281 PMC9019201
31. Kolb H.-J. Graft-versus-leukemia effects of transplantation and donor lymphocytes Blood 2008 112 4371 4383 10.1182/blood-2008-03-077974 19029455
32. Liang E.C. Dekker S.E. Sabile J.M.G. Torelli S. Zhang A. Miller K. Shiraz P. Hayes-Lattin B. Leonard J.T. Muffly L. Next-generation sequencing–based MRD in adults with ALL undergoing hematopoietic cell transplantation Blood Adv. 2023 7 3395 3402 10.1182/bloodadvances.2023009856 37196642 PMC10345845
33. Muffly L. Liang E.C. Dolan J.G. Pulsipher M.A. How I use next-generation sequencing–MRD to plan approach and prevent relapse after HCT for children and adults with ALL Blood 2024 144 253 261 10.1182/blood.2023023699 38728375 PMC11302453
34. Medina A. Puig N. Flores-Montero J. Jimenez C. Sarasquete M.-E. Garcia-Alvarez M. Prieto-Conde I. Chillon C. Alcoceba M. Gutierrez N.C. Comparison of next-generation sequencing (NGS) and next-generation flow (NGF) for minimal residual disease (MRD) assessment in multiple myeloma Blood Cancer J. 2020 10 108 10.1038/s41408-020-00377-0 33127891 PMC7603393
35. Maffini E. Ursi M. Barbato F. Dicataldo M. Roberto M. Campanini E. Dan E. De Felice F. De Matteis S. Storci G. The prevention of disease relapse after allogeneic hematopoietic cell transplantation in acute myeloid leukemia Front. Oncol. 2022 12 1066285 10.3389/fonc.2022.1066285 36530990 PMC9748414