Identification of biomarkers to drive treatment using cell-free DNA allows dynamic and safe assessment of tumor biology representing an important advance adding to tissue-based genotyping. Caution must be exercised interpreting commercial matching information as validation is often lacking, and true matching is feasible in a minority of patients.

See related article by Vidula et al., p. 3404

In this issue of Clinical Cancer Research, Vidula and colleagues describe results from a retrospective single-center analysis assessing the utility of tissue-based next-generation sequencing (NGS; n = 118) and plasma-based genotyping [Guardant 360 cell-free DNA (cfDNA; n = 252) for identification of actionable mutations in patients with metastatic breast cancer (1). Both methods of testing identified high rates of mutations, but cfDNA by NGS using the commercial assay identified a higher rate of actionable mutations in 78% of cases (196/252), resulting in matched therapy in 34% of cases. In contrast, tissue genotyping using an institutional assay identified actionable mutations in 50% of cases (59/118), resulting in matched therapy in 11%. Interestingly, tissue genotyping failed in 12 cases—highlighting the difficulty in obtaining adequate tissue in metastatic tumor biopsies. Of note, patients in this study were relatively treatment-naïve; those undergoing cfDNA testing had a median of 0 prior chemotherapy agents and 1 prior hormone therapy as treatment for advanced disease, and patients who underwent tissue testing had a median of 0 prior chemotherapy or hormone therapy lines of treatment.

Comparison of plasma versus tumor mutation testing is complicated by a number of factors, including the site of metastatic tissue, processing issues that exist in bone biopsies, small tumor samples that may not represent tumor heterogeneity or even a dominant mutation, and lack of concordance in timing of tissue acquisition. In a subset analysis, Vidula and colleagues (1) evaluated 30 patients who had both plasma and tumor testing, and 25 of those patients had concordant testing. In this small cohort, more patients were noted to have actionable mutations with plasma compared with tumor tissue (80% vs. 63%). Finally, patient outcome, as defined by overall survival, was improved in those receiving matched therapy in multivariable analysis.

There are a number of important messages from this study, and directions for the future. First, cfDNA offers an important and critical advance in understanding tumor biology, and is clearly complementary to tissue-based NGS, providing important information to direct use of targeted therapies based on clinical efficacy. A number of technologies exist, which vary in sensitivity and extent of genetic interrogation. Real-time or digital PCR is usually targeted to a small number or single variants, whereas NGS represents a broad-based approach looking at a large panel of genes or genetic variations. Indeed, the American Society of Clinical Oncology and College of American Pathologists Joint Review (2) notes that there is significant uncertainty between assays with variation in sensitivity (limit of detection), and care must be taken with collection of blood, processing of plasma, transport, and storage. The limit of detection may also impact the comparison between cfDNA and tissue genotyping, depending on the assay used for tumor tissue analysis. It is critical and sometimes challenging to rule out germline versus somatic acquired mutations. Subclonal variants may be present in low allele fractions, representing tumor heterogeneity that may not predict therapeutic efficacy. Both this joint review and the recent FDA approval of cfDNA for mutations in phosphatidyl inositol kinase (PIK3CA) recommend reflex tumor testing when a specific mutation is not detected in blood.

Another question is the definition of matching in commercial assays. As shown in Fig. 1, few mutations specific to breast cancer have been associated with therapeutic efficacy from matched targeted therapy, and only one mutation (PIK3CA) has regulatory approval based on a prospective phase III trial evaluating the PI3K inhibitor alpelisib in combination with fulvestrant, compared with fulvestrant alone (3). Matching for a number of mutations shown in the Vidula article (1) have not been associated with improved efficacy, for example, the mTOR or AKT, and these were not excluded in the analysis of survival. There is no marker other than expression of the estrogen receptor that has predicted benefit from cyclin-dependent kinase 4/6 inhibitors. In addition, a significant number of patients were enrolled in clinical trials with experimental agents based on cfDNA results, and patients who are able to enroll on trials have been noted to have longer survival, at least in part due to more favorable disease characteristics. Therefore, the conclusion that cfDNA results in greater “matching” to targeted therapies and the retrospective survival data is highly nuanced, and must be interpreted with caution.

Figure 1.

Therapeutic matching with cfDNA specific for advanced breast cancer. PIK3CA, phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha; ESR1, estrogen receptor 1; ERBB2, receptor tyrosine-protein kinase erbB-2; FGFR, fibroblast growth factor receptor; mTOR, mammalian target of rapamycin; PTEN, phosphatase and tensin homolog; SERD, selective estrogen receptor degrader; AI, aromatase inhibitor.

Figure 1.

Therapeutic matching with cfDNA specific for advanced breast cancer. PIK3CA, phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha; ESR1, estrogen receptor 1; ERBB2, receptor tyrosine-protein kinase erbB-2; FGFR, fibroblast growth factor receptor; mTOR, mammalian target of rapamycin; PTEN, phosphatase and tensin homolog; SERD, selective estrogen receptor degrader; AI, aromatase inhibitor.

Close modal

Real-time testing for targetable mutations in blood is appealing for a number of reasons. First, tumor biopsies are associated with risks for patients and may be difficult to obtain depending on the location of the tumor to be biopsied. A majority of hormone receptor–positive breast cancers, for example, present with bone metastases which may be the only site of disease through a number of lines of therapy. Biopsies of bone are decalcified, which markedly limits the ability to detect mutations. Obtaining sufficient tissue from tumors in lung or necrotic tumors in liver can be extremely challenging. Risks include pain and bleeding, severely limiting the feasibility of serial testing. Often archival tumor tissue from initial diagnosis is used for NGS, which may not accurately reflect mutations acquired with cancer progression and under the pressure of treatment. Serial testing of plasma cfDNA has enabled detection of differential clonal evolution in phase III trial comparing standard therapy with or without targeted agents (4), which has already improved our understanding of changes in tumor biology and mechanisms associated with development of resistance during therapy and could perhaps identify next step effective treatments.

Plasma cfDNA has a number of defined and potential applications to cancer treatment and early detection. As described in this article, matched therapies may be identified on the basis of the finding of specific mutations. Indeed, Vidula and colleagues (1) have published data identifying somatic BRCA mutations in plasma, and has an ongoing clinical trial evaluating PARP inhibition in this setting (NCT03990896). The authors emphasize the importance of understanding the functional impact of specific mutations to identify which are pathogenic, versus mutations that are functionally silent, and present an algorithm to aid in this assessment (5). cfDNA has also identified reversion mutations in BRCA that may predict resistance to PARP inhibition. Clearance of cfDNA has been associated with improved progression-free survival and may indicate efficacy of specific therapeutic strategies. In early-stage disease, several studies have associated detection of cfDNA with a higher risk of metastatic recurrence. In this setting, cfDNA may be personalized to mutations found in early-stage tumor tissue, also allowing identification of acquired mutations over time. One area of active research is to design treatment based on the detection of minimal residual disease (MRD) after neoadjuvant or adjuvant therapy, or the appearance of MRD on or after standard adjuvant therapy. This intriguing experimental approach is based on the hypothesis that starting new treatment to target MRD could prevent or delay the development of metastatic disease.

Further prospective testing is clearly required to understand the true therapeutic value of broader matching using mutations detected by cfDNA. However, plasma cfDNA offers significant advantages over tissue-based NGS, and these two approaches are clearly complementary. Serial testing requires additional evaluation to understand its true value in determining effective sequential matched therapies.

H.S. Rugo reports personal fees from PUMA, Samsung, and Mylan; grants from Roche, Pfizer, Lilly, Novartis, Merck, Sermonix, Daiichi, AstraZeneca, Odonate, and Immunomedics outside the submitted work. No disclosures were reported by the other author.

1.
Vidula
N
,
Niemierko
A
,
Malvarosa
G
,
Yuen
M
,
Lennerz
JK
,
Iafrate
AJ
, et al
Tumor tissue- versus plasma-based genotyping for selection of matched therapy and impact on clinical outcomes in patients with metastatic breast cancer
.
Clin Cancer Res
2021
;
27
:
3404
13
.
2.
Merker
JD
,
Oxnard
GR
,
Compton
C
,
Diehn
M
,
Hurley
P
,
Lazar
AJ
, et al
Circulating tumor DNA analysis in patients with cancer: american society of clinical oncology and college of american pathologists joint review
.
J Clin Oncol
2018
;
36
:
1631
41
.
3.
Andre
F
,
Ciruelos
E
,
Rubovszky
G
,
Campone
M
,
Loibl
S
,
Rugo
HS
, et al
Alpelisib for PIK3CA-mutated, hormone receptor-positive advanced breast cancer
.
N Engl J Med
2019
;
380
:
1929
40
.
4.
O'Leary
B
,
Cutts
RJ
,
Liu
Y
,
Hrebien
S
,
Huang
X
,
Fenwick
K
, et al
The genetic landscape and clonal evolution of breast cancer resistance to palbociclib plus fulvestrant in the PALOMA-3 trial
.
Cancer Discov
2018
;
8
:
1390
03
.
5.
Vidula
N
,
Dubash
T
,
Lawrence
MS
,
Simoneau
A
,
Niemierko
A
,
Blouch
E
, et al
Identification of somatically acquired BRCA1/2 mutations by cfDNA analysis in patients with metastatic breast cancer
.
Clin Cancer Res
2020
;
26
:
4852
62
.