Table 1. Selected studies by tumor type, the gene(s) used for ctDNA analysis, and outcomes.
| Year, First
Author [Ref] |
Study Type | Tumor Type | Number
of Patients |
Biomarker
Method |
Tested Genes | Outcome |
|---|---|---|---|---|---|---|
| Lung cancer | ||||||
| 2016, Adrian G.
Sacher [8] |
Prospective | Advanced NSCLC | 180 | ddPCR | KRAS, EGFR | Detection of KRAS and EGFR mutations, lower turnaround time compared to tissue |
| 2016, Jeffrey C.
Thompson [10] |
Cohort | Advanced NSCLC | 102 | Hybrid capture NGS | 70 cancer-related genes | Detection of potentially actionable variants in 75% of patients; concordance with tissue, 79% |
| 2018, Nicolas
Guibert [13] |
Cohort, blinded
to tissue |
Advanced NSCLC | 168 specimens
from 46 patients |
Amplicon-based NGS | 36 cancer-related genes | Detection of EGFR mutations, rare variants and fusions with high specificity. Early detection of resistance mechanisms in serial samples. |
| 2018, Charu
Aggarwal [20] |
Prospective | Advanced NSCLC | 323 | Hybrid capture | 73 cancer-related genes | Detection of actionable alterations in 20% of stage IV M1b patients in plasma but not tissue. Complementarity of tissue and plasma |
| 2014, Aaron M.
Newman [25] |
Cohort | Early-stage lung
cancer |
103 | Hybrid capture | 16 cancer-related genes + 8 proteins | Detection of ctDNA in early stages (stage I sensitivity, 50%) |
| 2017, Christopher
Abbosh [26] |
Cohorts | Early-stage lung
cancer |
96 | Patient-specific
multiplex PCR |
10–22 SNVs | MRD and subclonal evolution |
| Gastro-intestinal tumors | ||||||
| 2016, Honglei
Luo [32] |
Cohort | Esophageal, SCC | 8 | Illumina TruSight
sequencing |
90 cancer-related genes | Multigene panel has a role in detection and monitoring response to treatment |
| 2010, Hiroki
Takeshita [33] |
Case-control | Esophageal, SCC | 96 patients,
40 controls |
PCR-applied
biosystems |
CCND1 amplification | Poor prognostic value of CCND1 amplification |
| 2016, Masami
Ueda [34] |
Cohort | Esophageal, SCC | 13 | HiSeq2000 | 53 cancer-related genes | Multigene panel is associated with a greater accuracy of tumor recurrence compared to imaging methods (post-op) |
| 2015, Katsutoshi
Shoda [36] |
Case-control | Gastric | 52 patients,
40 controls |
PCR-applied
biosystems |
HER-2 | HER2 amplification can be used for therapeutic monitoring |
| 2017, Katsutoshi
Shoda [37] |
Case-control | Gastric | 60 patients,
30 controls |
PCR-applied
biosystems |
HER-2 | HER2 amplification can be used for therapeutic monitoring |
| 2015, Hideaki
Kinugasa [38] |
Cohort | Gastric | 25 | PCR - QX200, Bio-Rad | HER-2 | High concordance in detection of HER-2 between ddPCR and tissue IHC/FISH |
| 2016, Wen-Liang
Fang [39] |
Cohort | Gastric | 277 | TaqMan qPCR | 68 mutations (8 genes) | High ctDNA levels are associated with peritoneal recurrence and poor prognosis |
| 2017, Jeanne
Tie [43] |
Prospective,
cohort, multicenter |
Colon | 230 (1046
plasma samples) |
Safe-SeqS PCR | 15 cancer-related genes | ctDNA detection after stage II colon cancer resection provides direct evidence of residual disease and identifies patients at very high risk of recurrence. |
| 2018, Jeanne
Tie [44] |
Prospective,
cohort, multicenter |
Colon | 95 | Safe-SeqS PCR | 15 cancer-related genes | ctDNA detection after adjuvant chemotherapy for stage III colon cancer resection can identify patients at very high risk of recurrence. |
| 2017, Lone V.
Schøler [45] |
Longitudinal
cohort |
Colon | 45 (371
plasma samples) |
QX200 PCR | Somatic structural variants, KRAS | Postoperative ctDNA analysis detects residual disease and identifies patients at very high risk of relapse. Longitudinal surveillance allows
early detection of relapse and response to intervention. |
| 2015, Oliver A.
Zill [46] |
Cohort | Pancreato-biliary
carcinomas |
26 | Illumina Hi-Seq
2500 |
54 cancer-related genes | ctDNA sequencing is feasible, accurate, and sensitive in identifying tumor-derived mutations. |
| 2016, Kjersti
Tjensvoll [47] |
Cohort | Pancreatic | 14 (53
plasma samples) |
Mx3000P rtPCR | KRAS | ctDNA can be used as a marker for monitoring treatment efficacy and disease progression. |
| 2016, Naoto
Hadano [48] |
Cohort | Pancreatic | 105 | TaqMan assay PCR | KRAS | ctDNA can predict poor survival |
| 2017, Daniel
Pietrasz [50] |
Prospective,
cohort |
Pancreatic | 135 | Ion AmpliSeq NGS | 112 cancer-related genes | ctDNA is an independent prognostic marker in advanced pancreatic adenocarcinoma |
| 2018, Belinda
Lee [51] |
Cohort | Pancreatic | 42 | SafeSeqS assays
PCR |
KRAS | ctDNA analysis is a promising prognostic marker in early-stage pancreatic cancer and guides risk-adaptive treatment strategies. ctDNA detection
post-operatively helps to identify disease progression after standard adjuvant chemotherapy. |
| 2016, Stine Dam
Henriksen [52] |
Prospective
case-control |
Pancreatic | 95 patients,
183 controls |
Methylation-specific
PCR |
28 cancer-related genes | ctDNA promoter hypermethylation is a diagnostic biomarker that helps distinguish malignant from benign pancreatic disease. |
| 2018, Andreas
Wolfgang Berger [53] |
Cohort | Biliary tract cancer | 24 | 1010× depth Sequencing | 15 cancer-related genes | The molecular landscape is represented in ctDNA. |
| 2015, Atsushi
Ono [54] |
Cohort | Hepatocellular
carcinoma |
46 | Illumina Hi-Seq
2500 |
ctDNA detection post-surgery reflects tumor progression and disease recurrence. | |
| 2016, Wenjun
Liao [55] |
Cohort | Hepatocellular
carcinoma |
41 | Illumina MiSeq™ | Cancer-related
genes TERT, TP53, and CTNNB1 |
ctDNA mutation detection is associated with vascular invasion and predicts a shorter recurrence-free survival time. |
| 2016, Ao
Huang [56] |
Cohort | Hepatocellular
carcinoma |
48 | QX200 PCR | Cancer-related
genes TERT, TP53, and CTNNB1 |
ctDNA analysis can detect intratumoral heterogeneity and may have a promising role in the therapeutic management. |
| Breast cancer | ||||||
| 2017, Heather A.
Parsons [57] |
Prospective
cohort |
Triple-negative
breast cancer |
26 | HiSeq 2500 Illumina | 33 cancer-related genes | High concordance between ctDNA analysis and tumor tissue analysis, allowing monitoring of the therapeutic effect. |
| 2014, Julia A.
Beaver [58] |
Prospective
cohort |
Breast cancer | 29 | ddPCR | PIK3CA mutations | In patients with early-stage breast cancer, mutations can be detected in tumor tissue using ddPCR, and ctDNA can be detected in blood before
and after surgery. |
| 2016, Diana H.
Liang [60] |
Retrospective
chart review |
Breast cancer | 100 | Illumina Hi-Seq 2500 | TP53, PIK3CA,
ERBB2, and EGFR genomic alterations |
Robust concordance between tissue and blood for detection of PIK3CA mutation and ERBB2 amplification, but not for TP53 mutation and
EGFR amplification. Directional changes of TP53 and PIK3CA mutant allele are associated with response to therapy and PFS. |
| 2016, Marion
Rudolph [61] |
Cohort | Breast cancer | 600 | HiSeq 2500 Illumina | 306 cancer-related genes | AKT1E17K is the most likely disease driver in selected breast cancer patients and its detection in blood is achievable in advanced-stage disease. |
| 2017, Mary Ellen
Moynahan [62] |
Prospective | Breast cancer | 724 | QX200 PCR | PIK3CA | Improvement in PFS was maintained using everolimus, irrespective of PIK3CA genotypes (detected by ctDNA), and it was consistent with
previous analysis of archival tumor DNA using NGS. |
| 2016, Sarat
Chandarlapaty [63] |
Prospective | Breast cancer | 541 | QX200 PCR | ESR1 | ESR1 mutations are prevalent in ER-positive metastatic breast cancer treated with aromatase inhibitors. Both Y537S and D538G mutations
are associated with aggressive disease biology. |
| 2018, Rosaria
Condorelli [65] |
Cohort | Breast cancer | 3 | QX200 PCR | RB1 | Somatic RB1 mutations can emerge after exposure to CDK4/6 inhibitors. |
| 2016, Fei
Ma [66] |
Prospective,
cohort |
Breast cancer | 18 | HiSeq 2500 Illumina | 368 cancer-related genes | ctDNA analysis provides information regarding resistance to treatment and guides administration of anti-HER2 targeted therapy in the metastatic setting. |
| 2017, Francesca
Riva [67] |
Cohort | Triple-negative
breast cancer |
46 | Illumina MiSeq | TP53 | ctDNA levels decreased quickly during neoadjuvant chemotherapy (NCT) and helped identify minimal residual disease after surgery. Slow
decrease of ctDNA levels during NCT was strongly associated with shorter survival. |
| 2015, Isaac
Garcia-Murillas [68] |
Prospective,
cohort |
Breast cancer | 55 | HiSeq 2500 Illumina | Cancer-related genes | In patients with early stage breast cancer, ctDNA analysis can identify patients at high risk for relapse and guide adjuvant therapy. |
| 2017, Kala
Visvanathan [69] |
Cohort | Breast cancer | 141 | cMethDNA assay | 10 cancer-related genes | ctDNA gene methylation is a strong predictor of survival outcomes. |
| 2017, Hiroyo
Takahashi [70] |
Cohort | Breast cancer | 87 | Methylation-specific PCR | RASSF1A | Met-ctDNA is a more sensitive marker than CEA and CA15-3 and it can be used to monitor clinical tumor response to neoadjuvant chemotherapy. |
| 2016, Ming
Shan [71] |
Cohort | Breast cancer | 749 | MethyLight | SFN, P16,
hMLH1, HOXD13, PCDHGB7 and RASSF1a |
Epigenetic markers in serum have potential for diagnosis of breast cancer. |
| 2018, Charlotte
Fribbens [72] |
Prospective,
cohort |
Breast cancer | 83 | Enhanced
tagged-amplicon sequencing (eTAm-Seq) |
ESR1, KRAS, NRAS and HRAS | In patients with progressive disease after first-line aromatase inhibitors, ctDNA analysis demonstrated high levels of genetic heterogeneity and
frequent sub-clonal mutations. Sub-clonal KRAS mutations were found at a high frequency. |
| 2016, Peilu
Wang [73] |
Cohort | Breast cancer | 126 | Bio-Rad QX100 dd PCR | ESR1 |
ESR1 mutations were detected at very low allele frequencies in some primary breast cancers, and at high allele frequency in patients with
metastatic breast cancer. ESR1-mutant clones are enriched by endocrine therapy. |
| Gynecological cancers | ||||||
| 2014, Karina Dahl
Steffensen [74] |
Cohort | Ovarian cancer | 144 | QiaSymphony, multiplex qPCR | ctDNA detection | In patients treated with bevacizumab, high ctDNA levels were associated with poor PFS and OS. |
| 2016, Christine A.
Parkinson [75] |
Retrospective
analysis |
Ovarian cancer | 40 | ddPCR | TP53 | ctDNA is correlated with volume of disease at the start of treatment. |
| 2015, Elena
Pereira [76] |
Cohort | Ovarian/
endometrial cancer |
44 | qPCR using TaqMan®, ddPCR | ctDNA detection | ctDNA is an independent predictor of survival in patients with ovarian and endometrial cancers. |
| 2017, Adriaan
Vanderstichele [77] |
Prospective,
cohort |
Ovarian cancer | 68 | HiSeq 2500 Illumina | Chromosome instability | ctDNA analysis demonstrated that chromosomal instability can help detect ovarian cancer. |
| 2012, Maura
Campitelli [78] |
Cohort | Cervical cancer | 16 | RT-qPCR | ctDNA detection | ctDNA analysis demonstrated that the HPV mutational insertion is a highly specific molecular marker and it is detected in most patients with
stage 2-4 cervical cancer. |
| 2017, Zhigang
Kang [79] |
Retrospective
analysis |
Cervical cancer | 19 | ddPCR | HPV genetic components | HPV genetic insertion in ctDNA represents a promising tumor marker. |
| Urological cancers | ||||||
| 2017, Matti
Annala [80] |
Prospective,
cohort |
Prostate cancer | 319 | NimbleGenSeqCap, Illumina | 73 cancer-related genes | Biallelic gene loss detected in ctDNA can help prioritize therapy. |
| 2016, Alexander W.
Wyatt [81] |
Cohort | Prostate cancer | 65 | Illumina MiSeq, Ion Ampliseq | 19 cancer-related genes | Genomic profiling of ctDNA is feasible in mCRPC patients and provides important insights into enzalutamide response and resistance. |
| 2015, Arun A.
Azad [82] |
Cohort | Prostate cancer | 62 | PCR-based BEAMing | AR | AR gene aberrations in ctDNA are associated with resistance to enzalutamide and abiraterone in mCRPC. |
| 2015, Alessandro
Romanel [83] |
Cohort | Prostate cancer | 97 | Ion Torrent Sequencing | AR | Plasma AR sequencing can identify primary resistance to abiraterone. |
| 2016, Samanta
Salvi [84] |
Cohort | Prostate cancer | 59 | RT-PCR, ddPCR | AR | Detection of circulating AR copy number gain is a non-invasive biomarker for outcome of patients with CRPC treated with enzalutamide. |
| 2017, Sumanta K.
Pal [87] |
Prospective,
cohort |
Renal cell
carcinoma |
220 | HiSeq 2500 Illumina | 73 cancer-related genes | Higher rates of detection by ctDNA analysis after systemic therapy compared with baseline was noted for NF1, TP53, and VHL, indicating
clonal evolution of genomic alterations. |
| 2018, Neeraj
Agarwal [89] |
Cohort | Urothelial
carcinoma |
369 | HiSeq 2500 Illumina | 73 cancer-related genes | ctDNA NGS identified similar genomic alterations with tumor tissue. The genomic landscape was similar between lower tract and upper
tract urothelial carcinoma. |
| 2017, Emil
Christensen [91] |
Cohort | Urothelial
carcinoma |
831 | ddPCR | FGFR3 and PIK3CA | ctDNA levels in the urine and plasma were positively correlated and indicated that higher levels of FGFR3- and PIK3CA-mutated DNA
can predict disease progression. |
| Skin cancer | ||||||
| 2016, Gregory A.
Chang [92] |
Cohort | Melanoma | 43 | ddPCR | NRAS, BRAF | ctDNA had a higher sensitivity than LDH to detect disease progression. |
| 2016, Anne C.
Knol [93] |
Cohort | Melanoma | 38 | RT-PCR | BRAF | ctDNA BRAF mutation is a prognostic factor of OS and it is correlated with tumor burden. |
| 2015, Elin S.
Gray [94] |
Cohort | Melanoma | 48 | ddPCR | NRAS, BRAF | ctDNA is a biomarker of response to kinase inhibitor therapy and it can be used to monitor resistance to treatment. |
| 2015, Maria
Gonzalez-Cao [95] |
Cohort | Melanoma | 22 | TaqMan assay PCR | BRAF | Detection and accurate quantification of low- BRAF V600E in ctDNA can predict treatment outcome. |
| 2016, Ademi
Santiago-Walker [96] |
Cohort | Melanoma | 732 | PCR-based BEAMing | BRAF | BRAF mutation using ctDNA analysis can be detected in >75% of patients and is a prognostic marker. |
| 2016, Max
Schreuer [97] |
Cohort | Melanoma | 36 | qPCR | BRAF | Quantitative analysis of BRAF mutation in ctDNA is a monitoring tool during treatment with BRAF/MEK inhibitors. |
| 2017, Stephen Q.
Wong [98] |
Cohort | Melanoma | 52 | Amplicon sequencing, ddPCR | NRAS, BRAF | ctDNA is a powerful complementary modality to functional imaging for real-time monitoring of tumor burden and genomic changes
throughout therapy. |
| 2018, R.
Jeffrey Lee [99] |
Cohort | Melanoma | 161 | QX200 ddPCR | NRAS, BRAF | ctDNA predicts relapse and survival in high-risk resected stage II/III melanoma and can help select patients for adjuvant therapy. |
| Sarcoma | ||||||
| 2016, Manuela
Krumbholz [101] |
Cohort | Ewing | 20 | AccuPrime Taq DNA PCR | EWSR1-FLI1 fusion | Detection of EWSR1 fusion sequence in plasma is a promising noninvasive biomarker for improved therapeutic monitoring. |
| 2016, Masanori
Hayashi [102] |
Cohort | Ewing | 3 | ddPCR | EWS-ETS | Tumor specific EWS-ETS translocation breakpoints in plasma DNA is a highly personalized biomarker for relapsed disease. |
| 2018, David S.
Shulman [103] |
Cohort | Ewing,
osteosarcoma |
166 | Illumina HiSeq 2500 | EWSR1, FUS,
CIC, CCNB3, TP53, STAG2 |
Detectable ctDNA in patients with localized disease is associated with inferior event-free survival and OS at 3 years compared to patients
with undetectable ctDNA levels. |
| 2013, Jacqueline
Maier [104] |
Prospective
cohort |
Gastrointestinal
stromal tumor |
38 | RT-PCR | CKIT, PDGFRA | ctDNA harboring CKIT or PDGFRA was correlated with disease course. |
| 2014, Changhoon
Yoo [105] |
Cohort | Gastrointestinal
stromal tumor |
30 | PCR-based BEAMing | CKIT, PDGFRA, BRAF | Genotyping of the KIT gene in exon 17 of serum ctDNA identified mutations associated with resistance to dovitinib. |
| 2016, Noriko
Wada [106] |
Cohort | Gastrointestinal
stromal tumor |
4 | Sanger sequencing,
PCR |
C-KIT | Detection of secondary C-KIT mutations in ctDNA is useful to select targeted agents and to predict antitumor effects. |
| 2018, Nicholas C.
Eastley [107] |
Cohort | Soft tissue
sarcoma |
11 | Ion AmpliSeq | 57 cancer-related genes | ctDNA analysis detected TP53/PIK3CA mutations concordant with the primary tumor in 2 of 4 cases. |
| Brain tumors | ||||||
| 2013, Mohamad A.
Salkeni [108] |
Prospective | Glioblastoma | 13 | Illumina HiSeq | EGFR (vIII deletion) | ctDNA analysis identified the EGFRvIII deletion in 3 of 13 patients, which was correlated with tumor tissue analysis and may help select
patients for targeted therapy. ctDNA levels were correlated with the extent of tumor. |
| 2016, Shigeki
Yagyu [109] |
Retrospective | Neuroblastoma | 151 | Real-time quantitative PCR | MYCN | Serum MYCN amplification (sensitivity 86%, specificity 95% compared with tissue analysis) was associated with OS. It may help select
treatment prior to tumor biopsy, particularly for patients < 18 months (risk assessment and treatment depend on MYCN amplification status). |
| 2012, Blandine
Boisselier [110] |
Prospective | Glioma | 80 patients,
31 controls |
Digital PCR,
Agilent technologies |
IDH1 | The IDH1 R132H mutation was identified in plasma (specificity, 100%; sensitivity related to the tumor volume and contrast enhancement).
It may help in the diagnosis of patients not amenable to biopsy. |
| Lymphoma | ||||||
| 2016, Vincent
Camus [111] |
Cohort | Hodgkin | 94 | TaqMan assay PCR | XPO1 | The XPO1 E571K mutation in ctDNA can be used as a novel biomarker in diagnosis and detection of minimal residual disease. |
| 2016, Sarit E.
Assouline [112] |
Phase 2 trial | Diffuse large
B-cell |
40 | ddPCR | CREBBP, EP300,
MLL2, FAS, STAT6, TP53 |
Increase in ctDNA levels at 15 days after treatment initiation was associated with resistance to treatment. |
| 2016, Florian
Scherer [113] |
Case control | Diffuse large
B-cell |
92 patients,
24 controls |
CAPP-Seq | BCL2, BCL6, MYC, IGH | ctDNA levels at diagnosis were strongly correlated with clinical indices and were independently predictive of patient outcomes. |
| 2015, David M.
Kurtz [114] |
Prospective
cohort |
Diffuse large
B-cell |
75 | RT-PCR | Immunoglobulin high-throughput
sequencing (Ig-HTS) |
ctDNA immunoglobulin high-throughput sequencing preceded radiologic evidence of recurrent disease indicating that it may be a
surrogate marker for monitoring disease after complete remission. |
| 2015, Mark
Roschewski [115] |
Retrospective
analysis |
Diffuse large
B-cell |
126 | LymphoSIGHT™ | VDJ | After first-line treatment, disease progression was evident on imaging studies a median of 3.5 months after detection on ctDNA analysis
of the clonal Ig gene sequence. |
| Across Tumor Types | ||||||
| 2012, Geraldine
Perkins [117] |
Cohort from
multiple phase 1 trials |
Colorectal,
breast, melanoma, prostate, ovarian, and other |
105 | Sequenom
MassArray, OncoCarta PCR |
KRAS, BRAF, PIK3CA | ctDNA analysis has potential clinical multi-purpose utility in patients with advanced cancer. |
| 2015, Filip
Janku [118] |
Cohort from
multiple phase 1 trials |
Colorectal,
melanoma, non-small cell lung, and other |
157 | PCR-based
BEAMing |
BRAF, EGFR, KRAS, PIK3CA | Patients with > 1% of mutant ctDNA had shorter median OS compared to patients with ≤ 1%. |
| 2014, Chetan
Bettegowda [119] |
Cohort | Pancreatic,
ovarian, colorectal, bladder, gastroesophageal, breast, melanoma, hepatocellular, head and neck, and other |
640 | BEAMing,
PCR-Ligation, Safe-SeqS |
187 cancer-related genes | ctDNA is a broadly applicable, sensitive, and specific biomarker that can be used for clinical and research purposes in patients with
various tumor types. |
| 2013, Muhammed
Murtaza [120] |
Cohort | Breast,
ovarian and lung |
6 | HiSeq 2500 Illumina | PIK3CA, RB1, GAS6, EGFR | ctDNA analyses can complement invasive tumor biopsies to identify mutations associated with acquired drug resistance in advanced cancer. |
| 2015, Jean Sebastien
Frenel [121] |
Cohort | Colorectal,
ovarian, breast, bladder, glioblastoma, lung adenocarcinoma, endometrial |
39 | Ion AmpliSeq,
ddPCR |
Cancer-related genes | Targeted sequencing of ctDNA has potential clinical utility to monitor the effect of targeted therapies. |
| 2016, Maria
Schwaederle [122] |
Cohort | Lung,
breast, glioblastoma, genitourinary, gastrointestinal, of unknown primary, and other |
171 | Illumina Hi-Seq
2500 |
54 cancer-related genes | A large proportion of patients had detectable ctDNA aberration (s), among which the majority are targetable by an approved drug. |
| 2016, Maria
Schwaederle [123] |
Cohort | Brain,
lung, breast |
168 | Illumina Hi-Seq
2500 |
54 cancer-related genes | ctDNA tests provide information complementary to the tissue biopsies and may be useful in determining prognosis and treatment. |
| 2017, Yulian
Khagi [124] |
Cohort | Skin,
lung, breast, glioblastoma, genitourinary, gastrointestinal, and other |
69 | Illumina Hi-Seq
2500 |
54–70 cancer-related genes | Hyper-mutated ctDNA is correlated with response to checkpoint inhibitor-based therapy and investigation of hypermutated ctDNA
as a predictive biomarker is warranted. |
Abbreviations: AR: Androgen Receptor; ddPCR: droplet digital PCR; FISH: fluorescence in situ hybridization; IHC: immunohistochemistry; mCRPC = metastatic castration-resistant prostate cancer; Met: methylation; MRD: minimal residual disease; NCT: neoadjuvant chemotherapy; NGS: next-generation sequencing; NSCLC: non-small-cell lung carcinoma; OS = overall survival; PCR: Polymerase chain reaction; PFS: Progression free survival; qPCR: quantitative polymerase chain reaction; rtPCR: real-time Polymerase chain reaction; SCC: Squamous cell carcinoma; SNVs: single nucleotide variants.