Liquid biopsies are emerging as powerful minimally invasive approaches that have the potential to solve several long-standing problems spanning the continuum of cancer care: early detection of cancer, minimal residual disease tracking, and refinement of the heterogeneity of clinical responses together with therapeutic response monitoring in the metastatic setting. Existing challenges driven by technical limitations and establishment of the clinical value of liquid biopsies represent fields of active research that call for convergence science approaches to bridge scientific discovery with clinical care.

Liquid biopsies are evolving to provide insights in cancer early detection and diagnosis, characterization, treatment monitoring, and management (Fig. 1). The inception of this perspective is marked by a discussion of critical questions that we believe are poised to significantly enhance the role of liquid biopsies in the landscape of cancer care management as this field continues to evolve.

Figure 1.

Implementation of minimally invasive liquid biopsies in the continuum of clinical cancer care. The clinical utility of liquid biopsies is supported by an ever-growing number of studies showing that ctDNA can be implemented in early detection of cancer, minimal residual disease (MRD) tracking, and monitoring of therapeutic response in the context of precision oncology.

Figure 1.

Implementation of minimally invasive liquid biopsies in the continuum of clinical cancer care. The clinical utility of liquid biopsies is supported by an ever-growing number of studies showing that ctDNA can be implemented in early detection of cancer, minimal residual disease (MRD) tracking, and monitoring of therapeutic response in the context of precision oncology.

Close modal

A critical challenge with liquid biopsies is inherent to the scarcity of circulating tumor DNA (ctDNA), as the mutated tumor sequences are present in very small quantities relative to normal cell-free DNA (cfDNA) fragments. As such, conventional approaches to identify ctDNA through PCR and next-generation sequencing (NGS) have been limited by error rates in the range of the levels of ctDNA. These technological bottlenecks have been in-part circumvented by deep and redundant sequencing together with implementation of nucleotide barcodes (to distinguish tumor variants from sequencing errors), cfDNA size selection, and sequencing error suppression algorithms (that boost assay sensitivity by improving signal-to-noise ratio bioinformaticaly), multimutation tracking, and identification and removal of mutations resulting from clonal hematopoiesis. Tumor-informed approaches in particular, in which personalized liquid biopsy assays are designed guided by comprehensive genomic analyses of tumor tissue, have allowed for even higher sensitivity by lowering the limit of detection below 0.01% mutant allele frequency, while at the same time alleviating issues related to noncancer origin of plasma variants in the context of clonal hematopoiesis. In tandem, development of methods to assess methylation of cfDNA using tissue-specific targeted probes has provided a detection method for ctDNA together with tumor tissue of origin (1).

Nevertheless, targeted approaches for detection of ctDNA have reached a biological limit as a result of the several thousand cfDNA molecules present per milliliter of plasma. In part, these may be alleviated in the future by priming agents that increase ctDNA recovery, but at the expense of increasing overall cfDNA levels and through an improved understanding of cfDNA clearance and the natural variation from immune and other physiologic states. More promising in the near term is the development of a new generation of technologies aimed at detecting an increased number of tumor-derived alterations. These have included genome-wide detection of methylation using immunoprecipitation of methylated cfDNA (2), genome-wide detection of cfDNA chromosomal changes and fragmentation profiles (3), genome-wide mutational signatures (4), multianalyte analyses (5), and mutational (6) approaches for ctDNA detection. The clinical implementation of several of these methods is now underway and opens new avenues for improved outcomes for patients with cancer through an interplay of early cancer detection and capture of minimal residual disease (MRD) that can enable early therapeutic interception.

One of the greatest future health benefits for reducing cancer mortality may come through early cancer detection. As discussed above, this could be enabled through a new generation of liquid biopsies that identify individuals with cancer and guide them to the appropriate diagnostic interventions. We believe that these approaches would initially provide the most benefit as minimally invasive screening methods in asymptomatic high-risk populations in which screening is already known to improve clinical outcomes. One of the best examples is in individuals with lung cancer, in which early detection has been shown to reduce mortality. This survival benefit will likely increase further through the clinical implementation of neoadjuvant immunotherapy for patients with early stage disease, in which tumors are less complex and emergence of resistance is less likely. Given the low current screening rates for individuals at risk of lung cancer, highly sensitive and accessible tests open the possibility for broad implementation. Similar approaches could benefit population screening for early detection of colon, breast, and cervical cancer as recommended by the United States Preventive Services Task Force, as well as other cancers. The characteristics required for these methods are distinct from those needed in diagnostic applications for detection of MRD or therapeutic response and require highly sensitive and accessible tests for individuals that could be broadly incorporated prior to or instead of current screening approaches. A number of trials are underway to validate such tests, and at least one multicancer and one lung cancer screening test have been initially validated for clinical use. Broad screening in the average risk population for multiple cancers not currently recommended for screening may ultimately prove beneficial, but demonstration of utility in this setting will likely require much larger clinical trials demonstrating survival benefit.

Identification of early stage cancers unlocks new opportunities for intervention and may overcome emergence of resistance that characterizes later stage disease. As ctDNA may better capture disease burden and staging, liquid biopsies could refine clinical and pathologic staging by classifying patients with early stage cancers by risk for recurrence at the time of diagnosis (Fig. 1). Furthermore, ctDNA can be tracked longitudinally for MRD detection, informing therapeutic strategies at different timepoints before and after curative intent surgery. Liquid biopsy MRD approaches have predominantly used mutation-based NGS, and clinical sensitivity of ctDNA MRD (percentage of patients who relapsed and who were ctDNA positive) is typically higher when longitudinal MRD assessment is implemented compared with single MRD analysis. As such, longitudinal monitoring may increase the probability of ctDNA detection and distinguish true ctDNA MRD–positive cases. Landmark ctDNA MRD has been shown to capture patients at risk for clinical relapse several months prior to radiographic disease recurrence (7). Conversely, ctDNA MRD can be used to identify patients at very low risk of recurrence who may not need immediate additional systemic therapy and as such can be spared treatment-related adverse events and financial toxicity (8). Despite these promising studies, the predictive value of ctDNA MRD remains to be definitively shown and a fraction of MRD ctDNA–negative patients still experience disease recurrence despite the use of tumor-informed more sensitive approaches, likely because the presence of MRD is below the limit of detection of current ctDNA approaches. Taken together, in thinking about the clinical utility of ctDNA as an enrichment tool for perioperative therapy after curative intent surgery, we are facing the following challenge: although ctDNA detection has been shown to be highly specific for MRD detection, the clinical sensitivity (which reflects the number of patients with detectable ctDNA that also have disease recurrence) remains modest with a false negative rate that is not currently in the desired level for clinical implementation.

As immunotherapy is being established as a standard-of-care neoadjuvant or adjuvant therapy for patients with early stage cancers, emerging data supports the role of ctDNA in predicting pathologic complete response (pCR). In looking at the value of ctDNA MRD at different timepoints during neoadjuvant immune checkpoint blockade, ctDNA clearance post-immunotherapy, as well as preoperatively, may be linked with longer recurrence-free survival (9). As data continue to emerge, there may be patients who are cured with neoadjuvant immunotherapy-containing regimens and may not require surgery or additional therapy, and these patients may be identified by sustained ctDNA suppression. Notably, ctDNA MRD may be very informative in the heterogeneous non-pCR subset (tumors that do not completely regress at the time of resection by pathologic assessment), as patients in the non-pCR group that have ctDNA clearance may attain longer recurrence-free survival compared with patients with detectable ctDNA during neoadjuvant immune checkpoint blockade (9). These findings suggest that ctDNA clearance even in patients that do not attain a pathologic complete response may predict improved survival and that ctDNA MRD may be more accurate than pathologic response in predicting recurrence.

Although tumor-informed personalized ctDNA assays have shown a higher analytical sensitivity to date, these approaches may be less applicable to routine clinical practice as they require the development of patient-specific liquid biopsy assays, which remain labor intensive and costly. We envision that these challenges may be alleviated by new liquid biopsy methodologies that leverage genome-wide feature integration initially in a tumor-informed manner, followed by whole genome analyses of cfDNA, and integrating millions of cell-free DNA features that may increase assay sensitivity. Although somatic mutations (10), mutational signatures (4), chromosomal copy-number changes (3), and methylation (2) are more commonly interrogated, a number of other cfDNA or RNA features have begun to be leveraged, including fragmentation profiles resulting from altered chromatin accessibility (3), fragment end motifs and positioning, nucleosomal footprints, and inferred differential gene expression. Importantly, the incorporation of machine learning in multi-feature genome-wide liquid biopsy analyses has been shown to significantly improve the detection and interpretation of cancer “signals” in the bloodstream (3). Future studies will evaluate the added value of fragmentation changes and methylation on genomic cfDNA profiling, under the premise of developing tumor-agnostic multi-feature integrative ctDNA assays and machine learning methodologies that achieve the sensitivity of detection of newer-generation tumor-informed MRD assays.

The clinical implementation of liquid biopsies represents a critical step toward precision immuno-oncology, as optimal patient selection for immunotherapy represents a critical challenge that is intensified by the heterogeneity of clinical responses and inefficiency of existing biomarkers and imaging to capture the timing and magnitude of therapeutic response. We and others have shown that ctDNA kinetics during immunotherapy predict clinical outcomes with differential ctDNA profiles in responding compared with nonresponding tumors (11). Serial liquid biopsies can stratify patients early and accurately by risk of disease progression and can be particularly informative in refining heterogeneous clinical responses in the context of radiographically stable disease (12). Building on these proof-of-concept studies, ctDNA was tested as an early endpoint of response to immunotherapy in the BR.36 ctDNA molecular response adaptive clinical trial (13) that answered the following questions: what is the definition of ctDNA molecular response, when does it occur, and what is its concordance with radiographic RECIST response for patients with treatment-naïve metastatic PD-L1+ non–small cell lung cancer (NSCLC). Using a tumor-agnostic, white blood cell DNA-informed NGS approach, complete clearance of maximal mutant allele fraction of tumor-derived plasma mutations on cycle 3 of pembrolizumab was defined as ctDNA response, and although ctDNA and radiographic responses were overall concordant, there were several notable discordant cases across all categories of radiographic responses, with ctDNA molecular response better predicting progression-free and overall survival compared with radiographic response (13). The question that arises from all proof-of-concept studies supporting the role of ctDNA as an early endpoint of immunotherapy response is whether the earlier detection of response to therapy or the identification of molecular disease progression may impact patients’ outcomes. As such, ctDNA interventional clinical trials that randomize patients to treatment intensification or de-escalation based on ctDNA response are needed to establish ctDNA molecular response in clinical cancer care. Although ctDNA molecular responses are dependent on the limit of detection of the liquid biopsy assay used, we are hopeful that in the future, regulatory approvals may be pursued for generic ctDNA molecular responses, allowing the incorporation of multiple assays with similar analytical performance characteristics.

Comprehensive genomic profiling (CGP) via liquid biopsies is less invasive compared with a tissue biopsy, typically has faster turn-around time, can be informative in cases with insufficient tumor tissue and missed or failed tumor testing, may better capture tumor heterogeneity as well as tumor evolution and emerging resistance in the context of therapy, and has the potential to be a more accessible testing modality in the community. It is important to note the limitations and opportunities for improvement of ctDNA CGP for solid malignancies, including higher false negative rate, differential sensitivity, and reportable range for mutations other than single-base substitutions as well as a higher positive rate because of detection of somatic non–tumor-derived mutations, with a representative example being that of clonal hematopoiesis.

Biopsying compartments more proximally to the primary tumor, such as pleural fluid, peritoneal fluid, cerebrospinal fluid, and urine, may boost the sensitivity of detection for tumors that do not follow a hematogenous spread paradigm, such as mesotheliomas, or those “protected” by cellular barriers, such as primary brain tumors. As cfDNA carries the genetic and epigenetic footprint of the tissue of origin, liquid biopsies have shown promise in capturing cancer lineage (3, 14) and may be leveraged for therapeutic stratification and for capturing nongenetic mechanisms of therapeutic resistance, such as small-cell transformation of EGFR-driven lung adenocarcinomas under the selective pressure of an EGFR tyrosine kinase inhibitor.

What are the critical next steps to bridge scientific discovery with clinical cancer care and integrate liquid biopsies in the decision matrix for individuals with cancer, thus bringing liquid biopsies to the forefront of precision medicine? We believe that ctDNA molecular response (or lack of) is ready to be broadly tested prospectively in clinical trials. As an example, the second stage of the BR.36 trial is a currently enrolling phase II/III randomized study capitalizing on ctDNA disease progression to escalate immunotherapy monotherapy to combination immuno-chemotherapy (NCT04093167). Certainly, ctDNA response can be implemented in treatment de-escalation in the setting of sustained ctDNA clearance as well as early switch in the setting of ctDNA recrudescence or emergence (15). The latter may also have a role in refining drug development strategies and establishing ctDNA as an early endpoint to support drug approvals (15). In thinking about ctDNA adaptive clinical trials for patients with advanced/metastatic cancer, ctDNA molecular response may be utilized as a stratification approach to identify a patient population less likely to attain therapeutic benefit earlier than imaging and randomize those patients to continue their treatment regimen versus escalating or switching therapies. Similarly, in the context of standard of care, patients with molecular progression early on standard-of-care therapy are identified and considered for a clinical trial, which is enriching for this patient population. In the future, newer cost-effective mutation-independent and tumor-agnostic approaches, including those using genome-wide fragmentation, mutation, or methylation signatures, may allow for these applications to be more widely and effectively utilized.

Implementing liquid biopsies in clinical decision-making for patients with early stage disease remains challenging. However, as sensitive technologies continue to improve, it is important to consider ctDNA-based MRD as a stratification and enrichment strategy in clinical trials. As an example, ctDNA MRD after definitive therapy can be used as a stratification factor in trials, enrolling patients both negative and positive for ctDNA or in separate trials for the ctDNA MRD-positive and ctDNA MRD-negative cohorts. Ultimately, our clinical decision matrix could be visualized under the prism of liquid biopsies as a tic tac toe board, in which ctDNA detectable/undetectable represents the opening move determining the direction of the game in terms of systemic therapy first versus surgery first, and the player's moves represent strategic therapeutic decisions based on monitoring ctDNA MRD. The vision here is that this sequential alignment of markers in a row over time represents a winning strategy for our patients.

Ultimately, the utilization of liquid biopsies in the continuum of cancer care, from early detection, to detection of residual disease, to therapeutic response and detection of resistance, provides multiple opportunities for reducing the morbidity and mortality of cancer. The implementation of liquid biopsies in the clinic calls for collaborative efforts across industry and academia to develop and test these approaches in the next-generation ctDNA adaptive clinical trials and deliver the earliest and best clinical care with precision.

V. Anagnostou reports grants from Astra Zeneca, Bristol Myers Squibb, Delfi Diagnostics, Personal Genome Diagnostics, honoraria from Neogenomics, and Foundation Medicine and personal fees from Astra Zeneca outside the submitted work; in addition, V. Anagnostou has a patent for cancer genomic analyses, ctDNA therapeutic response monitoring, and immunogenomic features of response to immunotherapy (63/276,525; 17/779,936; 16/312,152; 16/341,862; 17/047,006; 17/598,690) issued. V.E. Velculescu reports grants, personal fees, and other support from Delfi Diagnostics during the conduct of the study; personal fees from Epitope; and personal fees from Viron Therapeutics outside the submitted work; in addition, V.E. Velculescu has a patent for genomic analyses and liquid biopsy patents and applications pending, issued, licensed, and with royalties paid from DELFI, PGDx, LabCorp, Qiagen, Sysmex, Agios, Genzyme, Esoterix, Ventana, and ManaTBio; and in addition V.E. Velculescu is a founder of Delfi Diagnostics, serves on the Board of Directors for this organization, and owns Delfi Diagnostics stock, which is subject to certain restrictions under university policy. In addition, Johns Hopkins University owns equity in Delfi Diagnostics. V.E. Velculescu divested his equity in Personal Genome Diagnostics (PGDx) to LabCorp in February 2022. V.E. Velculescu is an inventor on patent applications submitted by Johns Hopkins University related to cancer genomic analyses and cell-free DNA for cancer detection that have been licensed to one or more entities, including Delfi Diagnostics, LabCorp, Qiagen, Sysmex, Agios, Genzyme, Esoterix, Ventana, and ManaT Bio. Under the terms of these license agreements, the University and inventors are entitled to fees and royalty distributions. V.E. Velculescu is an advisor to Viron Therapeutics and Epitope. These arrangements have been reviewed and approved by the Johns Hopkins University in accordance with its conflict-of-interest policies. No other disclosures were reported.

This work was supported in part by the NIH grants CA121113 (to V. Anagnostou and V.E. Velculescu), CA006973 (to V.E. Velculescu), and CA062924 (to V.E. Velculescu), a Cancer Research Institute Torrey Coast Foundation GEMINI CLIP award (to V. Anagnostou), an ECOG-ACRIN Thoracic Malignancies Integrated Translational Science Center grant UG1CA233259 (to V. Anagnostou and V.E. Velculescu), the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation (to V.E. Velculescu), Stand Up To Cancer-LUNGevity-American Lung Association Lung Cancer Interception Dream Team Grant (grant number: SU2C-AACR-DT23–17 to V.E. Velculescu), the Gray Foundation (to V.E. Velculescu), The Honorable Tina Brozman Foundation (to V.E. Velculescu), the Commonwealth Foundation (to V. Anagnostou and V.E. Velculescu), and the Cole Foundation (to V.E. Velculescu). The indicated Stand Up To Cancer grant is administered by the American Association for Cancer Research, the scientific partner of SU2C. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

1.
Liu
MC
,
Oxnard
GR
,
Klein
EA
,
Swanton
C
,
Seiden
MV
,
Consortium
C
.
Sensitive and specific multi-cancer detection and localization using methylation signatures in cell-free DNA
.
Ann Oncol
2020
;
31
:
745
59
.
2.
Shen
SY
,
Singhania
R
,
Fehringer
G
,
Chakravarthy
A
,
Roehrl
MHA
,
Chadwick
D
, et al
.
Sensitive tumour detection and classification using plasma cell-free DNA methylomes
.
Nature
2018
;
563
:
579
83
.
3.
Cristiano
S
,
Leal
A
,
Phallen
J
,
Fiksel
J
,
Adleff
V
,
Bruhm
DC
, et al
.
Genome-wide cell-free DNA fragmentation in patients with cancer
.
Nature
2019
;
570
:
385
9
.
4.
Bruhm
DC
,
Mathios
D
,
Foda
ZH
,
Annapragada
AV
,
Medina
JE
,
Adleff
V
, et al
.
Single-molecule genome-wide mutation profiles of cell-free DNA for non-invasive detection of cancer
.
Nat Genet
2023
;
55
:
1301
10
.
5.
Cohen
JD
,
Li
L
,
Wang
Y
,
Thoburn
C
,
Afsari
B
,
Danilova
L
, et al
.
Detection and localization of surgically resectable cancers with a multi-analyte blood test
.
Science
2018
;
359
:
926
30
.
6.
Kurtz
DM
,
Soo
J
,
Co Ting Keh
L
,
Alig
S
,
Chabon
JJ
,
Sworder
BJ
, et al
.
Enhan­ced detection of minimal residual disease by targeted sequencing of phased variants in circulating tumor DNA
.
Nat Biotechnol
2021
;
39
:
1537
47
.
7.
Abbosh
C
,
Frankell
AM
,
Harrison
T
,
Kisistok
J
,
Garnett
A
,
Johnson
L
, et al
.
Tracking early lung cancer metastatic dissemination in TRACERx using ctDNA
.
Nature
2023
;
616
:
553
62
.
8.
Tie
J
,
Cohen
JD
,
Lahouel
K
,
Lo
SN
,
Wang
Y
,
Kosmider
S
, et al
.
Circulating tumor DNA analysis guiding adjuvant therapy in stage II colon cancer
.
N Engl J Med
2022
;
386
:
2261
72
.
9.
Kelly
RJ
,
Landon
B
,
Zaidi
AH
,
Singh
D
,
Canzoniero
JV
,
Balan
A
, et al
.
Neoadjuvant nivolumab or nivolumab plus LAG-3 inhibitor relatlimab in resectable esophageal/gastroesophageal junction cancer: a phase 1b trial and ctDNA analyses
.
Nat Med
2024
;
in press
.
10.
Zviran
A
,
Schulman
RC
,
Shah
M
,
Hill
STK
,
Deochand
S
,
Khamnei
CC
, et al
.
Genome-wide cell-free DNA mutational integration enables ultra-sensitive cancer monitoring
.
Nat Med
2020
;
26
:
1114
24
.
11.
Anagnostou
V
,
Forde
PM
,
White
JR
,
Niknafs
N
,
Hruban
C
,
Naidoo
J
, et al
.
Dynamics of tumor and immune responses during immune checkpoint blockade in non-small cell lung cancer
.
Cancer Res
2019
;
79
:
1214
25
.
12.
Murray
JC
,
Sivapalan
L
,
Hummelink
K
,
Balan
A
,
White
JR
,
Niknafs
N
, et al
.
Elucidating the heterogeneity of immunotherapy response and immune-related toxicities by longitudinal ctDNA and immune cell compartment tracking in lung cancer
.
Clin Cancer Res
2024
;
30
:
389
403
.
13.
Anagnostou
V
,
Ho
C
,
Nicholas
G
,
Juergens
RA
,
Sacher
A
,
Fung
AS
, et al
.
ctDNA response after pembrolizumab in non-small cell lung cancer: phase 2 adaptive trial results
.
Nat Med
2023
;
29
:
2559
69
.
14.
Mathios
D
,
Johansen
JS
,
Cristiano
S
,
Medina
JE
,
Phallen
J
,
Larsen
KR
, et al
.
Detection and characterization of lung cancer using cell-free DNA fragmentomes
.
Nat Commun
2021
;
12
:
5060
.
15.
Sivapalan
L
,
Murray
JC
,
Canzoniero
JV
,
Landon
B
,
Jackson
J
,
Scott
S
, et al
.
Liquid biopsy approaches to capture tumor evolution and clinical outcomes during cancer immunotherapy
.
J Immunother Cancer
2023
;
11
:
e005924
.