1. Introduction
Breast cancer (BC) is the most common cancer among women worldwide and one of the major causes of cancer-related mortality in women
[1]. With peak incidence between 35 and 75 years, most cases develop sporadically, and less than 10% of them are hereditary due to germline mutations
[2]. The clinical management of BC has improved remarkably in the last decades by advancements in surgery and (neo)adjuvant therapies in early stages. These have made it possible to achieve a 5 year overall survival rate above 90%
[3]. Nonetheless, BC can spread through distant metastases, which could worsen the 5-year survival rate to as low as 28%
[4], and therapeutic strategies in these cases are mostly palliative in nature. The main obstacle for designing effective treatment approaches in BC is the complex heterogeneity of these tumours
[5].
BC can be classified into four molecular subtypes according to the specific gene expression profiles of 50 genes or the so-called PAM50 assay: luminal A, luminal B, basal-like, and human epidermal growth factor receptor 2 (HER2)-enriched
[6][7]. Among luminal tumours, the luminal A subtype has a higher expression of hormone-activated genes, low levels of proliferation markers and good survival rates, while Luminal B tumours have poorer prognosis, are characterised by higher histological grade, increased expression of tumour proliferation-associated genes and can display overexpression of HER2
[8]. On the other hand, the basal-like subtype presents low levels of luminal-related genes, low HER2 expression and high levels of proliferation genes being the intrinsic subtype with the poorest rates of survival
[7].
As mentioned above, the HER2-enriched tumours are one of the subtypes defined by the PAM50 BC tumour subtyping and are characterised with the highest
ERBB2 gene expression amongst all subtypes both at RNA and protein level. Additionally, HER2-enriched tumours also show an increased expression of tumour proliferation-related genes
[9][10]. The HER2 protein, encoded by the
ERBB2 gene, is a receptor belonging to the EGFR family
[2]. Notably, the HER2 overexpression is not limited to this subtype as HER2-positive tumours (overexpressing the HER2 receptor) with luminal A, luminal B, or even basal-like subtypes can be found. Additionally, it is also possible to find HER2-enriched subtypes in histologically HER2-negative tumours, which can give rise to significant differences in response to treatment and biological evolution
[7][11][12].
In clinical practice, HER2 positivity is determined by immunohistochemistry (IHC) and/or by in situ hybridisation (ISH) in order to tailor targeted therapeutic approaches
[2]. The HER2 status assessment was originally standardised by The American Society of Clinical Oncology and the College of American Pathologists (ASCO/CAP) that published guidelines recommending to test HER2 protein overexpression by IHC and
ERBB2 gene amplification by ISH using FFPE tumour tissue. The ASCO/CAP guideline was updated in 2018, when the criteria were refined to systematise the testing algorithm for controversial HER2 categorisation. For these cases, it is required to have more strict interpretation for ISH together with a concomitant IHC
[13]. Approximately 15–20% BC tumours have HER2 overexpression and/or amplification, and over 50% of these also coexpress hormone receptors
[11].
The HER2 protein is a biomarker with a reported heterogeneity in BC
[14]. In general, critical genomic changes in BC occur mostly during cancer progression, which can create significant variability between primary tumours and its metastases
[5][15]. In addition, systemic treatments create dynamic molecular pressure, which can lead to more biological heterogeneity
[16]. Ideally, the collection of serial tumour biopsies and an extensive longitudinal monitoring could capture the abovementioned heterogeneity and depict the disease changes caused by treatments and progression. HER2 status is a clear example of cancer heterogeneity
[11] since it has been demonstrated that HER2 expression can change from primary disease to metastases by 9–60%
[17][18][19][20]. This phenomenon can lead to an inappropriate treatment selection if HER2 status is not reassessed throughout different disease stages. Considering all the above, it is paramount to find new methodologies permitting a close disease monitoring and HER2 status evaluation to provide a proper clinical management.
2. Circulating Tumour Cells in HER2-Positive Early Breast Cancer
CTCs can be found in the circulation of cancer patients and serve as an important source of tumour-related information and disease monitorisation
[21]. CTCs detection has been convincingly validated as a biomarker of the worse outcome in both early and metastatic BC
[22]. The BEVERLY-2 study
[23] is a relevant clinical trial focusing on the non-metastatic HER2-positive invasive BC, which encompassed CTC detection and pCR and provided validating evidence of their combined prognostic impact on DFS assessment. This is an open-label, single-arm, multicentre phase II study that adds neoadjuvant and adjuvant bevacizumab to the standard-of-care regimen. Of note, the prognostic value of CTCs and circulating endothelial cells in a survival analysis after 3 years of follow-up was described therein. In this study, CTC positivity was set to ≥1 CTC per 7.5 mL of blood. The detection of CTCs was not associated with patients’ clinical or pathological characteristics. Of note, 46% of the patients were CTC-positive at least once between baseline sampling to presurgery. A high percentage of patients (35%) had detectable CTCs at baseline and the positivity decreased to 7% before surgery due to NAC. CTCs detection during NAC was associated with significant reduction in DFS (CTC-positive patients had DFS of 54% compared to DFS of 83% in patients with no CTCs), but there was no association with pCR. Moreover, patients with no CTCs throughout neoadjuvant stages had a 96% OS compared to 83% for those with ≥1 CTC. When CTC detection was combined with pCR for prognostic stratification, the patients with baseline CTC counts of <1/7.5 mL (CTC negative patients), who achieved pCR, showed excellent prognosis, while those with a baseline CTC of ≥1/7.5 mL and no pCR were at higher risk of relapse. In detail, 3–4 weeks after surgery, 13.2% of included patients were CTC positive. Additionally, in an exploratory analysis, the authors observed that 3–4 weeks after surgery, CTCs were detected in 4% of patients achieving pCR and in 31% that had not. At the end of adjuvant treatments, ≥1 CTC/7.5 mL were detected in 20.7% of the patients, but the mean CTC count in the cohort was 0.3 CTC/7.5 mL. No further differences were discovered after 1 year of adjuvant therapy in the studied patients. Studying the impact of CTCs detection on DFS, the researchers observed that the 3-year DFS in patients with ≥ 1 CTC/7.5 mL at baseline was significantly lower than for patients with no CTC detected in pretreatment samples. However, they did not observe the association between CTCs and DFS in other blood time points. Additionally, patients with CTCs detected at any point during NAC presented a significant reduction in DFS (
Table 1).
Table 1.
Reviewed liquid biopsy studies in early and metastatic breast cancer.
| Study |
Analyte |
Methodology |
HER2-Positive (n) |
Findings |
References |
| ctDNA in HER2-positive early BC |
| García-Murillas et al., 2015 |
ctDNA |
ddPCR |
21 |
- -
-
Correlation between risk of relapse and ctDNA detection.
- -
-
ctDNA sequencing could serve to characterise the genetic landscape of MRD.
|
[26] |
| García-Murillas et al., 2019 |
ddPCR |
41 |
- -
-
Improved relapse lead time over clinical identification of 10.7 months by ctDNA detection.
- -
-
HER2-positive tumours show intermediate ctDNA abundance (0.81 copies/mL).
|
[28] |
| Coombes et al., 2019 |
Patient-specific panel and ultradeep sequencing |
8 |
- -
-
ctDNA detection rate of 89%.
- -
-
ctDNA detection in HER2-positive patients show 100% sensitivity.
|
[29] |
| McDonald et al., 2019 |
TARDIS |
7 |
- -
-
Ten years relapse-free survival is 95% for HER2-positive patients.
|
[30] |
| Parsons et al., 2020 |
NGS with UMIs |
25 |
- -
-
No correlation observed between MRD detection and tumour subtype, including HER2-positive.
|
[31] |
| Rothé et al., 2019 |
ddPCR |
455 |
- -
-
Highest pCR rates for patients with HER2-enriched subtypes and undetectable ctDNA.
|
[32] |
| Zhang et al., 2019 |
Large NGS panels |
27 |
- -
-
ctDNA detection rate of 74.2% in early BC.
- -
-
Patients with ctDNA detected before chemotherapy that becomes negative after that are mainly basal-like or HER2-enriched tumours.
- -
-
Basal-like and HER2-enriched tumour subtypes are more sensitive to chemotherapy.
|
[33] |
| Magbanua et al., 2021 |
Patient-specific panel and ultradeep sequencing |
19 |
- -
-
ctDNA positivity is higher in HER2-enriched tumours, TNBC and larger tumours.
- -
-
47.4% of HER2-positive cases achieved pCR.
|
[34] |
| ctDNA in HER2-positive metastatic BC |
| Kingston et al., 2021 |
|
Guardant360 cfDNA assay-ddPCR |
72 |
- -
-
More mutations are found in the ERBB2 gene in patients with more anti-HER2 targeted lines of treatment.
- -
-
Exclusive ERBB2 mutations in HER2-positive tumours.
|
[35] |
| Davis et al., 2020 |
Guardant360 |
75 |
- -
-
TP53, PIK3CA and ERBB2 were the most common mutated genes in HER2-positive tumours.
- -
-
ERBB2 was the most frequent CNV in the HER2-positive cohort.
- -
-
CNVs in ERBB2 were detected in 96.4% of HER2-positive patients.
|
[36] |
| Allegretti et al., 2021 |
|
ddPCR and NGS |
20 |
- -
-
Molecular subtype switch as consequence of treatment.
- -
-
Detection of new targetable alterations (ESR1, MYC and FGFR1).
|
[37] |
| CTCs in HER2-positive early BC |
| Pierga et al., 2015 |
CTCs |
CellSearch® |
52 |
- -
-
Detectable CTCs are associated with shorter 3-years DFS and OS.
- -
-
In the neoadjuvant setting:
- ⚬
-
Combining CTC detection and pCR those patients with baseline CTCs counts of <1/7.5 mL and pCR show excellent prognosis yet those with baseline CTC of ≥1/7.5 mL and no pCR are at higher risk of relapse.
- ⚬
-
Patients presenting ≥1 CTC have a DFS of 54% and OS of 96% compared with a DFS and OS of 83% in patients without CTCs detected.
|
[38] |
| Riethdorf et al., 2017 |
CellSearch® |
59/63 |
- -
-
≥1 CTC and ≥2 CTCs before NAC is associated with reduced DFS and OS.
- -
-
CTCs-negative patients with pCR show the best prognosis.
- -
-
CTC-positive patients with decreased tumour response correlate with high risk of relapse.
- -
-
For HER2-positive BC, the detection of ≥2 CTCs/7.5 mL show reduced DFS.
|
[39] |
| Kasimir-Bauer et al., 2016 |
AdnaTest ® |
56 |
- -
-
CTCs are detected in 24% and 8% of BC patients before and after NAC, respectively.
- -
-
There was no association between CTCs detection and PFS or OS.
- -
-
Therapy-resistant CTCs detected after NAC could indicate worse outcome.
|
[40] |
| Azim et al., 2013 |
CellSearch® |
51 |
- -
-
No significant CTC counts decrease during NAC is associated with lower pCR rates
|
[41] |
| Trapp et al., 2019 |
CellSearch® |
279 |
- -
-
CTCs detection have prognostic features in patients with luminal A-like, luminal B-like or TNBC tumours but not in the HER2-positive subtype.
|
[42] |
| Janni et al., 2016 |
CellSearch® |
688 |
- -
-
A decreased OS is associated with CTCs detection two years after adjuvant chemotherapy in Luminal A-like, luminal B-like and TNBC tumours.
- -
-
CTC status has prognostic relevance in HR-negative/HER2-positive tumours but not in HR-positive/HER2-positive.
- -
-
A high prognostic value when HER2-positive patients are combined independently of the HR status.
- -
-
No demonstration of a prognostic value for CTCs detection during follow-up in patients with HER2-type tumours.
|
[43] |
| Georgoulias, et al., 2012 |
|
Cytocentrifugation + IF |
- |
- -
-
CK+/HER2+ CTCs detected in 90% of the patients.
- -
-
Anti-HER2 therapy decrease the number of CTCs, the risk of relapse and increase DFS
|
[44] |
| CTCs in HER2-positive metastatic BC |
| Munzone et al., 2012 |
CTCs |
CellSearch® |
53 |
- -
-
No demonstrated prognostic value for CTCs detection during follow-up in patients with HER2-type tumours.
|
[45] |
| Giordano, et al., 2012 |
CellSearch® |
101 |
- -
-
CTCs detection is strongly associated with OS prediction in all MBC subtypes excluding HER2-positive patients treated with anti-HER2 therapy.
|
[46] |
| Meng, et al., 2004 |
Ferrofluids + FISH |
- |
- -
-
HER2 expression in ten CTCs was enough to diagnose the ERBB2 gene status.
- -
-
37.5% of cases with HER2-negative primary tumours have ERBB2 amplified in their CTCs during progression.
|
[47] |
| Flores et al., 2010 |
CellSearch®/CellSearch Profile Kit (PFK) + FISH |
45 |
- -
-
33% of metastatic BC patients with HER2-negative disease presented HER2-amplified CTCs
|
[48] |
| Jacot et al., 2019 |
CellSearch® |
- |
- -
-
9.1% of HER2-negative MBC patients with ≥1 CTC/7.5 mL presented ≥1 HER2-amplified CTC.
|
[49] |
| Wang et al., 2020 |
CellSearch® |
- |
- -
-
High-risk HER2+ CTC-patients had shorter survival and higher progression risk.
- -
-
Anti-HER2 therapy increased PFS.
- -
-
PFS was higher in patients switching from high- to low-risk HER2+ CTCs during treatment.
|
[50] |
| Pestrin et al., 2012 |
CellSearch® + IF |
- |
- -
-
No benefit was observed when patients with HER2-negative tumours and HER2-positive CTCs were treated with anti-HER2 therapy.
|
[51] |
| Agelaki et al., 2015 |
IF |
2 |
- -
-
Lapatinib substantially decreases the number of HER2-positive CTCs in patients with HER2-negative tumours only in those with disease stabilisation.
|
[52] |
| Fehm et al., 2021 |
CellSearch®+IHC or FISH |
- |
- -
-
Patients with HER2-negative tumours presented HER2-positive CTCs.
- -
-
CTCs detection is not associated with OS.
- -
-
Patients with no CTCs detected at first follow-up had better OS.
- -
-
Lapatinib treatment is effective to decrease HER2-positive CTC independently of HER2 status of the primary tumour.
|
[53] |
| Müller et al., 2021 |
|
CellSearch® |
- |
- -
-
The presence of ≥1 strong-stained HER2 CTCs is associated with shorter OS but not with better PFS.
- -
-
The association with shorter OS was not observed in patients with moderate-stained HER2 CTCs.
|
[54] |
Liquid biopsy studies in early and metastatic breast cancer (BC) including or focusing on HER2-positive BC patients (n). Studies employing both circulating-tumour DNA (ctDNA) or circulating tumour cells (CTCs) were examined. HER2, human epidermal growth factor receptor 2; CTCs, circulating-tumour cells; ctDNA, circulating-tumour DNA; ddPCR, droplet-digital PCR; MRD, minimal residual disease; pCR, pathological complete response; TNBC, triple-negative breast cancer; IF, immunofluorescence; BC, breast cancer; MBC, metastatic breast cancer; NGS, next-generation sequencing; UMIS, unique molecular identifiers; cfDNA, circulating-free DNA; SNV, single nucleotide variant; CNV, copy number variant; DFS, disease-free survival; OS, overall survival; NAC, neoadjuvant chemotherapy; PFS, progression-free survival; IHC, immunohistochemistry; and FISH, fluorescence in situ hybridisation.
An additional publication describing results from the GeparQuattro study
[24], which included a mixture of BC subtypes, again postulated that CTC-negative patients achieving pCR showed the best prognosis in contrast to those CTC-positive with a suboptimal response to therapy, who had higher risk of recurrence. Moreover, patients with <2 CTCs without pCR had an intermediate risk of metastatic disease. Specifically, for HER2-positive BC, the GeparQuattro trial defined that the detection of ≥2 CTCs/7.5 mL was associated with a reduced DFS in a multivariate analysis, suggesting that some of these preoperatively detected CTCs are resistant to therapies and lead to subsequent relapse (
Table 1). Indeed, the presence of subclones composed of epithelial-to-mesenchymal transition (EMT)-like CTCs and stem cell-like CTCs was described to provoke resistance to conventional oncologic treatments
[25][26][27], including NAC
[28][29] (
Table 1). Finally, the NeoALTTO phase III trial
[30] also studied CTC fluctuation during the therapy of HER2-positive patients. While no significant decrease in CTC counts was achieved during NAC, a tendency towards lower pCR rates was observed in patients with detectable CTCs. This study was likely underpowered, but a meta-analysis could help to uncover interesting associations (
Table 1).
In addition to the previously mentioned, CTCs detection was also employed in the adjuvant setting to correlate CTCs positivity with DFS and OS. In this context, the SUCCESS A clinical trial is a multicenter, open-label, phase III study comparing three adjuvant chemotherapy regimens and two schemes of adjuvant treatment with bisphosphonate (2 vs. 5 years of zoledronate) for early-stage high-risk BC patients
[31]. CTC presence was assessed before and 2 years after chemotherapy
[32]. The investigators observed notable differences in the correlation analyses between CTCs presence and DFS and OS when considering the different BC subtypes. In luminal A-like, luminal B-like and TNBC tumours, the detection of CTCs 2 years after chemotherapy was associated with decreased OS. By contrast, there was no prognostic value of CTCs detection during follow-up in patients with HER2-type tumours, contradictory to other studies (
Table 1)
[33][34][35]. The authors pointed to possible carryover effects, provoked by anti-HER2 targeted treatments, which can explain this phenomenon
[36]. Additional information will likely be obtained through the SUCCESS B trial, an open-label, multicentre and randomised phase III study which only includes HER2-positive BC patients.
Finally, it has been demonstrated that the presence of HER2-positive CTCs in early BC patients and showed that this positivity is an independent factor from HER2-status in the primary tumour
[37][38][39][40]. In an interesting study
[41], it was demonstrated that patients with HER2-negative tumours but positive CTCs treated with trastuzumab showed reduced probability of relapse and improved disease-free interval. These findings could indicate that targeting HER2-positive CTCs is a suitable strategy to avoid relapses in early BC patients. In this regard, it is important to adapt isolating antibody-based platforms to specifically detect HER2-positive CTCs population. Instead, methodologies based on CTCs cell-size couple with approaches to identify HER2-positivity at protein, RNA or DNA level could represent an excellent alternative.
3. Circulating Tumour Cells in HER2-Positive Metastatic Breast Cancer
In the metastatic BC setting, it is crucial to decipher disease’s evolution especially related to resistance to treatments or the appearance of novel tumour clones. These processes are well documented in metastatic BC patients treated with hormone therapy, where estrogen receptor 1 (
ESR1) mutations are responsible for treatment resistance
[42]. CTCs detection can offer not only information regarding their quantity but also their genomic/transcriptomic characterisation
[21][43]. The molecular interrogation of these cells can provide a molecular snapshot of the whole disease, even when metastatic site is not easily accessible by conventional biopsies
[44].
In metastatic BC, it has been demonstrated that HER2 status may change between primary tumours and CTCs
[20][45]. It has been also demonstrated that patients with HER2-positive tumours can present HER2-negative CTCs in their blood. Moreover, the detection of HER2-positive CTCs is not infrequent in HER2-negative tumours, as specified bellow, and could indicate worse clinical outcome. A small-series study, led by Flores and colleagues
[46], reported that 33% of metastatic BC patients with HER2-negative disease had HER2-amplified CTCs. Similar CTC findings were observed in the CirCE T-DM1 trial
[47] and in the plasmaMATCH study focused on ctDNA (
Table 1)
[48]
In an interesting study, Wang and colleagues investigated the impact of anti-HER2 therapy in patients with HER2-negative tumours but positive CTCs
[49]. They observed that these patients with ≥2 HER2-positive CTCs showed less survival and better benefits from anti-HER2 therapies. Moreover, in a follow-up analysis, those patients changing between ≥2 to <2 HER2-positive CTCs presented better survival. Similarly, another study explored the effect of lapatinib in patients with HER2-positive CTCs but HER2-negative tumours
[50]. The authors reported that 7 out of 96 patients with detected CTCs presented HER2-positivity. However, they did not observe tumour response to lapatinib in this patient population but only one patient with disease stabilisation. By contrast, Agelaki and colleagues
[51] also treated patients with HER2-negative advanced tumours but HER2-positive CTCs with lapatinib. They observed a decrease in the HER2-positive CTCs number in patients with disease stabilisation but also in the median number of detected CTCs per patient. However, no objective responses were observed, probably because of the low number of patients included in the study. Moreover, the DETECT III study compared standard therapy alone or in combination with anti-HER2 targeted therapy (lapatinib) in patients with HER2-negative metastatic BC and HER2-positive CTCs
[52]. Their results showed favourable outcomes after combined targeted treatment indicated by early decline of CTC counts. Still, this clearance was not significant compared to the standard arm [57.1% vs. 50.0%;
p = 0.63]. Of note, patients in the lapatinib arm showed a tendency towards a better progression-free survival (PFS) and a significantly improved OS by univariate (HR 0.54; 95% CI, 0.34–0.86;
p = 0.008) and multivariate (HR 0.55; 95% CI, 0.34–0.90;
p = 0.016) cox regression analysis when compared to those in the standard arm. The CTC clearance at the end of treatment was not associated with OS, although better OS was observed in patients with no evidence of CTCs at first follow-up compared to patients with detectable CTCs (HR 0.36; 95% CI, 0.17–0.76;
p = 0.005) (
Table 1). Finally, in a very recent research article also within the DETECT program, the authors analysed the HER2 status of CTCs in patients HER2-negative tumours and its clinical significance
[53]. Herein, CTCs HER2 status was analysed in 1159 CTCs-positive patients. They observed that patients with estrogen and progesterone-positive status were more prone to present strong-stained HER2 CTCs. The authors also demonstrated the association of ≥1 CTCs strong-stained for HER2 with shorter OS but not between the proportion of HER2-positive CTCs and clinical outcome. The association with OS is lost when including moderate-stained HER2 CTCs. Importantly, CTCs HER2 status was not significantly associated to PFS in this cohort.
4. Conclusions and Future Perspectives in HER2-Positive Breast Cancer and Liquid Biopsy
While most of the studies reviewed thus far are based on limited sample sets, it is already clear that liquid biopsy and ctDNA detection in the HER2-positive early BC setting can provide prognostic information regarding both the response to NAC and early relapse. In this regard, ctDNA monitoring could be a tool to modulate NAC schedules and avoid overtreatment in low-risk patients or increase dosage in those with suboptimal responses. In the adjuvant setting, the detection of MRD and recurrence before it is detectable clinically could crucially improve BC patients’ survival by the timely starting of adjuvant therapy as well as by preventing overtreatment in patients with a lower risk of relapse. In this regard, large clinical studies employing ctDNA detection for clinical decision-making and personalisation are needed for the final validation of this methodology. Ideally, these studies employ ultrasensitive methodologies to detect ultralow levels of ctDNA.
Likewise, more studies are needed to determine the role of ctDNA detection in the metastatic HER2-positive BC. The development of novel methodologies to sensitively track ERBB2 amplifications and mutations in blood and characterise disease evolution, resistance and the acquisition of HER2-positivity in tumours previously HER2-negative is paramount to improve the clinical manage of BC patients.
More research also warrants to clarify the possible application of CTCs counts in clinics both in the early and metastatic setting. During NAC, certain associations between CTCs detection and worse outcomes were already observed. However, important discrepancies can be found between CTCs counts and response to NAC. In the same manner, no conclusive results exist to clarify whether CTCs detection in early HER2-positive BC during adjuvant therapy would be beneficial for patients’ stratification. In addition, a shift between HER2-negative to positive disease at advanced disease stages was demonstrated using CTCs detectable in the blood of some patients with HER2-negative tumours. Moreover, changes in CTCs counts and clinical benefits were addressed in these patients when they are treated with anti-HER2 therapies. However, there is a need to develop studies including a higher number of patients to achieve consistent conclusions. It is necessary to develop novel and highly sensitive technologies to detect as well as to characterise HER2-negative or positive CTCs in the metastatic BC setting. Overall, it is important to shed light over the variability aspect between techniques as a factor behind the contradictory observations in the different studies. On top of that, it is crucial to understand how the disease is evolving and detect early resistance to be able to efficiently treat it. The advanced disease can provide us with high CTCs numbers that, unlike ctDNA, potentially offer deep insights about the disease transcriptomics and genomics.