The Truth Is in the Blood: The Evolution of Liquid Biopsies in Breast Cancer Management

The Truth Is in the Blood: The Evolution of Liquid Biopsies in Breast Cancer Management


Dr. Marcelo Rocha Cruz

Dr. Ricardo Costa

Dr. Massimo Cristofanilli, credit Northwestern Medicine

By Marcelo Rocha Cruz, MD; Ricardo Costa, MD, MSc; and Massimo Cristofanilli, MD

Article Highlights

  • Liquid biopsy may improve therapies and achieve better clinical outcomes in patients with cancer, including patients with breast cancer.
  • CTCs have been investigated as a faster, easier, and more reliable method for detection of progressive disease and tumor resistance.
  • The prognostic value of CTC detection presents important variation according to breast cancer molecular subtypes. Although the prognostic value of CTC count in metastatic breast cancer has been well established, its utility in defining the benefit of changing versus continuing initial therapy early in treatment has been challenged.
  • CTCs and ctDNA represent two faces of the same spectrum and advance our capabilities to evaluate the continuum of breast cancer diagnostics from early detection to treatment planning and monitoring of resistance.

Tissue has been the issue for cancer diagnosis, molecular evaluation, and prognostic and predictive factors for initial therapy response over the past few decades. However, it has not been clearly demonstrated that increasing our understanding of a tumor’s molecular fingerprint is affecting patients’ symptoms, improving response to treatment, or, ultimately, improving survival. In fact, recent data on tumor heterogeneity and cancer cells’ dynamic mutations throughout time have shown that one biopsy is insufficient to fully explore the complexity of a tumor.1,2 Performing multiple organ biopsies is also not feasible or practical. Circulating tumor cells (CTCs) and cell-free circulating tumor DNA (ctDNA) analysis—collectively termed “liquid biopsies”—are candidate surrogates that unveil tumor heterogeneity, describe real-time tumor genomic evolution, detect developing resistance mechanisms, and measure disease burden. The answers to improve therapies and achieve better clinical outcomes may be hidden in the bloodstream.

The first report on tumor cells in the peripheral circulation was attributed to Ashworth in 1869.3 Since then, the existence, origin, and clinical significance of CTCs have been debated. The introduction of immunohistochemical techniques in the late 1970s led to renewed interest in the detection of CTCs.4,5 CTC concentration in the bloodstream is very low, and their life span is usually short. Technology has evolved to introduce tools with higher sensitivity and specificity to detect CTCs, raising the bar to initiate studies that could elucidate their possible association not just with early detection of treatment-resistant disease in the metastatic setting but also with minimal residual disease in early-stage solid malignancies. Furthermore, recent CTC molecular profiling techniques have begun a new era in the understanding of the pathways of invasion and metastasis, as well as real-time tumor resistance mechanisms and the future of personalized medicine.

The Prognostic Value of CTCs in Breast Cancer

It is widely recognized that the detection of disease progression with anatomical analysis using standard and functional imaging methods represents a late picture of cancer development and resistance. In many cases, the burden of progressive disease at the time of imaging evaluation (usually with intervals varying between 9 and 12 months) is advanced to the point that patients are already experiencing reduced clinical performance and organ function precluding the choice of the next lines of therapy. CTCs have been investigated as a faster, easier, and more reliable method for detection of progressive disease and tumor resistance (Table).

In 2004, the first prospective multicenter clinical trial with the objective to determine clinical significance of levels of CTCs in metastatic breast cancer was published.6 This study was designed to prospectively determine the prognostic and predictive value of CTCs in patients with metastatic breast cancer who were about to start a new systemic treatment. The 177 enrolled patients underwent peripheral blood collection at monthly intervals for up to 6 months after enrollment. CTC detection was made using the CellSearch system, which is approved by the U.S. Food and Drug Administration. A cutoff of five CTCs/7.5 mL was used to stratify patients into positive and negative groups (positive, ≥ 5 CTCs/7.5 mL; negative, < 5 CTCs/7.5 mL). The study showed that patients in the positive group had shorter progression-free survival (PFS) and overall survival (OS) than patients in the negative group. Furthermore, the CTC status at first follow-up after initiation of therapy had an even greater association with PFS and OS. On multivariate Cox hazards regression analysis, CTC levels were the most significant predictors of PFS and OS at baseline and first follow-up.

An analysis restricted to patients who were about to start first-line systemic therapy after the diagnosis of metastatic breast cancer had similar findings.7 The cutoff of five CTCs/7.5 mL became standard. Moreover, the DETECT trial showed that serial measure of CTCs before and after chemotherapy was prognostic and could be a useful tool to measure treatment efficacy.8 Finally, a pooled analysis of individual patient data from 20 international studies confirmed the independent prognostic effect of CTC count on PFS and OS in patients with metastatic breast cancer.9

The prognostic value of CTC detection presents important variation according to breast cancer molecular subtypes. In HR-positive metastatic breast cancer, detection of higher levels of CTCs and more aggressive disease should suggest the use of combination treatments, particularly with the availability of new CDK4/6 inhibitors and PI3K/mTOR-targeted agents rather than single-agent endocrine treatment even though all the clinical parameters seem to indicate an endocrine-sensitive disease.10

For patients with HER2-positive tumors who received the anti–HER2–targeted therapy, the prognostic value of CTCs was not sustained. The same was true for patients with triple-negative metastatic breast cancer who received bevacizumab as part of their chemotherapy regimen.11,12 These targeted therapies may have some direct action against circulating epithelial cells, reducing the prognostic value of CTC enumeration. A large retrospective study clarified the biologic and clinical implications of CTC enumeration. This study demonstrated that patients with five or more CTCs tend to develop a pattern of progressive disease characterized primarily by onset of new sites and visceral metastasis compared with patients with lower or undetectable cells, who tend to develop growth of individual established lesions. The difference in the type of disease progression appears to explain the significantly shorter survival in patients with higher CTC levels.13

The prognostic value of CTC detection has also been investigated in non-metastatic breast cancer. In the neoadjuvant setting, a meta-analysis of published data could not demonstrate a correlation between the decrease of CTCs and pathologic complete response—an important surrogate endpoint in neoadjuvant therapy trials.13,14 However, the persistence of CTCs after neoadjuvant therapy is associated with an increased risk of relapse and worse survival,15 suggesting that those cells can more effectively be used as evidence of micrometastatic residual disease and predictive of metastatic spread than a marker of chemosensitivity.

A future benefit of CTC detection in this setting would be to select post-neoadjuvant therapy for patients with residual disease according to CTC enumeration and molecular characteristics. An analysis of CTC count in 2,026 patients with early-stage breast cancer before adjuvant chemotherapy and in 1,942 patients with early-stage breast cancer after chemotherapy was recently published.16 This study used the CellSearch system to detect CTCs. After chemotherapy, 22.1% of patients were CTC positive. The presence of CTCs was associated with poor disease-free survival (DFS), distant DFS, breast cancer–specific survival, and OS. CTCs were confirmed as independent prognostic markers in multivariable analysis for DFS and OS without any association with tumor size, grading, or hormone receptor status. The prognosis was the worst in patients with at least five CTCs per 30 mL of blood. The presence of persisting CTCs after chemotherapy showed a negative influence on DFS and OS, suggesting the independent prognostic relevance of CTCs in patients with primary breast cancer.

Another recent study did not confirm these results.17 Using a manually performed immunocytochemistry (MICC) in peripheral blood at primary diagnosis in 1,221 patients, this study could not demonstrate prognostic relevance regarding CTCs that were quantified using the MICC method at the time of primary diagnosis in a cohort of patients with early-stage breast cancer. Pre-analytical factors, as well as different technologies, may have contributed to the discrepant results. In the adjuvant context, the lower frequency of CTC detection in all trials (20% to 30%) would limit its use and suggest exploring other liquid biopsy methods, such as ctDNA, in this setting.

Although the prognostic value of CTC count in metastatic breast cancer has been well established, its utility in defining the benefit of changing versus continuing initial therapy early in treatment has been challenged. SWOG completed a prospective trial (SO500) to answer this question.18 Patients with metastatic breast cancer who presented with increased CTC count after one cycle of first-line therapy were randomly assigned to continue initial treatment or change to an alternative chemotherapy. This study confirmed the prognostic significance of CTCs in patients with metastatic breast cancer receiving first-line chemotherapy. However, for patients with persistently increased CTCs after one cycle of first-line chemotherapy, switching to another chemotherapy regimen early in treatment was not effective in prolonging OS.

There were many limitations in the study design that may have affected the results and contributed to the negative interpretation. There have been no reported data on the chemotherapy regimens used as first-line therapy and at progression. The study allowed “physicians choice” without specification or restrictions. Furthermore, the study was open to enroll patients with HR-positive or triple-negative breast cancer, and the former was the most represented cohort. Despite these major limitations, there were expectations that a simple CTC enumeration could provide indication of clinical utility.

In summary, CTC enumeration remains a strong prognostic biomarker in advanced breast cancer, and its predictive value can only be explored in the context of a study focusing on a therapeutic question addressing single intervention (e.g., targeted therapy) and possible association with biomarker analysis by polymerase chain reaction next-generation sequencing (NGS) of CTCs or using a complementary technology such as ctDNA.

CTC Molecular Profiling

The next step in the CTC landscape is extensive research on CTC molecular profiling, protein expression, and functional evaluation. Studies have shown discrepancies in hormone receptors and HER2 expression in CTCs when compared with primary breast tumors, which are correlated with worse outcomes and could explain tumor cell dynamics in acquiring resistance to some therapies.19,20 The development of NGS and the possibility to perform multigene profiling as well as the perspective to undergo real-time treatment monitoring, which uncovers the intrinsic tumor resistance mechanisms and provides information for targeted therapy selection, have expanded the value of CTCs.

An example of this effort is the investigation of the PIK3CA pathway in breast cancer CTCs. PIK3CA somatic mutations are one of the most frequent mutations in a variety of human cancers. In breast cancer, activation of the PI3K pathway correlates with lack of sensitivity to the HER2-directed antibody trastuzumab and the kinase inhibitor lapatinib, leading to a significantly worse outcome in patients with PIK3CA-mutated tumors compared with wild-type tumors.21 It is also recognized as an important factor in breast epithelial transformation and metastatic progression, as well as in acquired endocrine resistance in HR-positive tumors exposed to tamoxifen, fulvestrant, or aromatase inhibitors. Recent studies have detected and enumerated CTCs from peripheral blood of metastatic breast cancer and performed genomic profiling of these cells. As a result, the presence of PIK3CA mutation in CTCs is correlated with worse clinical outcome.22,23 Wide CTC genomic profiling with NGS can also offer analysis of the full PIK3CA pathway, including PTEN loss, together with other actionable mutations.

ctDNA

Tumor progression leads to increased cancer cell turnover. As a consequence, there is an increase in the number of apoptotic and necrotic cells releasing DNA fragments in the bloodstream. These fragments are the main origin of ctDNA. In metastatic breast cancer, initial data have shown that ctDNA correlates with tumor marker CA 15-3 and CTCs in revealing early changes in tumor burden.24 Also, the detection of somatic mutations in ctDNA using NGS has confirmed that this method is reliable as a surrogate marker of tumor heterogeneity and the development of resistance mechanisms.25,26 Higgins et al. demonstrated the ability to detect PIK3CA mutations in patients with metastatic breast cancer using the BEAMing assay.27 Nakauchi et al. have recently compared somatic mutations of TP53 and PIK3CA found in ctDNA with primary site tissue.26 Peripheral blood samples were obtained from 17 patients with metastatic breast cancer, and genomic sequencing results were compared with the primary tumor. Of note, the interval between the peripheral DNA analysis and the primary site tissue is longer than 50 months in the majority of patients, denoting higher odds to find acquired resistance mutations to previous therapies as well as new circulating clone cells.1,2 As a result, mutations found in ctDNA were not present in the paired primary tumor, suggesting the value of ctDNA in exploring tumor heterogeneity.

Current technology has presented the possibility to detect and perform molecular profiling of both ctDNA and CTCs from blood samples of patients with metastatic breast cancer, which opened perspectives for the comparison of prognostic value between ctDNA and CTCs.28 Shaw et al. reported the results of the CTC count and cell-free DNA quantification (cfDNA) of 112 patients with metastatic breast cancer.29 CTCs were enumerated by CellSearch, and CTC count was compared with the amount of cfDNA, serum CA 15-3, and alkaline phosphatase. Mutation profiles of CTCs, cfDNA, and primary tumors were also analyzed. In the whole cohort, cfDNA levels and cell counts (CTC > 5/7.5 mL) were significantly associated with OS, unlike CA 15-3 and alkaline phosphatase. NGS analysis of CTCs from five patients revealed mutational heterogeneity in PIK3CA, TP53, ESR1, and KRAS genes. Furthermore, cfDNA profiles provided an accurate reflection of mutations seen in CTCs. ESR1 and KRAS gene mutations were absent from primary tumor tissue, which may again reflect either acquired mutations during disease progression or a clone of cell from a heterogeneous tumor. Most recently, detection of PIK3CA and ESR1 mutations—the most common genomic alteration in HR-positive metastatic breast cancer—have been evaluated in the context of large, randomized clinical trials and seem to offer an unprecedented opportunity to advance precision medicine applications in clinical practice.30,31

Liquid biopsy is on the way to impacting breast cancer management from initial detection through resistant metastatic disease (Fig. 1). CTCs and ctDNA represent two faces of the same spectrum and advance our capabilities to evaluate the continuum of breast cancer diagnostics from early detection to treatment planning and monitoring of resistance. In the era of personalized medicine, a simple blood sample can offer the opportunity for a real-time assessment of the complexity of the disease in an individual patient, identifying mechanisms of resistance that can be conquered by selected therapies and translated into a decision-making tool for patients.

About the Authors: Dr. Cruz is with the Developmental Therapeutics Program, Division of Hematology Oncology, Northwestern University Feinberg School of Medicine. Dr. Costa is with the Developmental Therapeutics Program, Division of Hematology Oncology, Northwestern University Feinberg School of Medicine. Dr. Cristofanilli is the associate director, Translational Research and Precision Medicine, Robert H. Lurie Comprehensive Cancer Center of Northwestern University, and professor of medicine with the Feinberg School of Medicine.