Emerging Landscape of Targeted Therapy of Breast Cancers With Low Human Epidermal Growth Factor Receptor 2 Protein Expression

The HER2 is a receptor tyrosine-protein kinase encoded by the ERBB2 gene on chromosome 17q12.⁴  The HER2 pathway comprises membrane receptors and their ligands which dimerize and phosphorylate to enable signal transduction from the cell exterior to the nucleus through internal processing facilitated by protein tyrosine kinases. The final output of transcription factors regulates cellular functions, such as proliferation, differentiation, and survival. In cancer, overexpression of HER2 promotes angiogenesis, invasion, and metastasis, which results in more aggressive clinical behavior.

HER2 overexpression/gene amplification, seen in 15% to 20% of BCs, is characterized by high histologic grade, aneuploidy with p53 mutations, and activated PI3K/AKT and Ras/Raf/MEK/MAPK pathways.^5–8  BCs with HER2 overexpression/gene amplification have a high propensity to metastasize to the brain and visceral organs.^9,10 

Despite their aggressive behavior, HER2-positive BCs have successfully responded to targeted drugs, such as monoclonal antibodies (eg, trastuzumab and pertuzumab), tyrosine kinase inhibitors (eg, lapatinib, neratinib, and tucatinib), and ADCs (eg, ado-trastuzumab emtansine [T-DM1]) and fam-trastuzumab deruxtecan [T-DXd]), all of which are effective in metastatic disease.²  Tyrosine kinase inhibitors and monoclonal antibodies are also effective in the (neo)adjuvant setting.² 

Based on the survival benefit of these anti-HER2 targeted agents, it is recommended that every newly diagnosed BCs, recurrences, and metastases be assessed for HER2 expression per the ASCO/CAP guidelines.^2,11  Since their inception in 2007, the ASCO/CAP HER2 guidelines underwent updates in 2013 and 2018 to provide guidance for a more consistent and reproducible evaluation of HER2 for patient treatment. Currently, HER2 status is considered positive when more than 10% of tumor cells show strong complete membranous staining by IHC or when they show weak to moderate complete membranous staining in more than 10% of tumor cells and HER2/CEP17 ratio 2.0 or greater with average HER2 copy number at or above 4.0 signals per cell (ISH-amplified or positive).² 

The 2018 ASCO/CAP HER2 guideline established 5 result categories for HER2 ISH scoring.²  ISH group 1 is positive (HER2/CEP17 ratio ≥2.0 with average HER2 copy number ≥4.0). ISH group 2 (HER2/CEP17 ratio ≥2.0 with average HER2 copy number <4.0) reflects a phenomenon likely due to monosomy 17, which confers no benefit with anti-HER2 treatment with trastuzumab. ISH group 3 (HER2/CEP17 ratio <2.0 with average HER2 copy number ≥6.0) comprises an uncommon phenomenon (0.4%–3%) with limited evidence as to clinical benefit from anti-HER2 treatment with trastuzumab.¹²  ISH group 4 (HER2/CEP17 ratio <2.0 with HER2 copy number ≥4.0 and <6.0) accounts for 5% of cases, and ISH group 5 is negative (HER2/CEP17 ratio <2.0 with average HER2 copy number <4.0). IHC should be evaluated concurrently for dual-probe ISH groups 2 through 4 to assess HER2 status.² 

No formal definition of BC with low HER2 protein expression exists. However, most published studies and clinical trials define HER2-low BC as invasive BC with HER2 IHC expression of score 1+ or 2+ without HER2 gene amplification by ISH.^4,13–20  Based on this definition, many patients would be classified as having HER2-low BC, with an estimated overall frequency of 59% in BC.²¹  HER2-low BC encompasses a heterogeneous group of tumors with a broad spectrum of clinical and histomorphologic features, including subtype, grade, immunophenotypic, and molecular subtypes. HER2 protein expression can range within HER2-low BC. An extensive study in HER2-negative tumors showed that HER2-low BC represented more than half (59.7%) of the tumors in the series, with a higher frequency of HER2 IHC score 1+ (40.4%) compared with score 2+ (19.3%)²¹ . HER2-low BC can be hormone receptor (HR) positive or negative. However, studies indicate that HER2-low BC is more likely to be HR positive.^21–25 

Schettini et al²¹  reported HR positivity in 88% of HER2-low BC and observed a higher incidence of HER2-low BC in HR-positive tumors (65.4%) compared with triple-negative BC (TNBC; 36.5%). In addition, results indicated that 74.3% of HER2-low BC corresponded to the ductal histologic subtype and 19.2% was lobular; most were Nottingham grade 3 (50.3%), or less frequently graded 2 (39.1%). Compared with BC with HER2 IHC score 0, HER2-low BC had no statistical difference in histologic subtype, Nottingham grade, or Ki-67 proliferation index (with 14% cutoff) but showed significantly older patient age at the time of diagnosis and comparatively higher pathologic stage, with a larger tumor size and greater nodal involvement.²¹ 

The biology and molecular features of HER2-low BC are not fully understood. The potential molecular mechanisms by which these tumors express increased HER2 protein levels without HER2 gene amplification have been explored. In HER2 nonamplified tumors, upregulation of HER2 in response to activation of the NF-κB pathway due to multiple types of external stimulation, including the tumor microenvironment, therapy exposure (chemotherapy, radiation therapy, and antiendocrine therapy), and crosstalk between estrogen receptor (ER) and HER2 pathways, has been described as a potential mechanism to increase HER2 protein expression as a means for tumor adaptation, survival, and treatment resistance.^26–32 

The genomic features of HER2-low BC are still not fully characterized. Recently published studies using next-generation sequencing identified HER2 somatic mutations in approximately 2% to 5% of primary BCs.^33–35  These mutations have been associated with resistance to anti-HER2 therapy, decreased relapse-free survival, and worse clinical outcomes.^36–39  Most BCs with HER2 mutations lack co-occurring HER2 gene amplification (<1%).^33,39–41  In HER2-low BC, the exact frequency of activating HER2 mutations is unknown; however, several studies have identified somatic mutations in cancers lacking HER2 gene amplification.^34,42  Connell and Doherty⁴³  identified higher frequencies of HER2 mutations in BC with HER2 IHC score 1+ (2.9%) and score 2+ (2.0%), compared with score 0 (1.6%) and score 3+ (1.3%) tumors.

A gene expression profiling analysis (PAM50) identified the biologic subtypes of HER2-low BC,²¹  including a majority of Luminal A (50.8%) molecular subtype compared with the less prevalent Luminal B (28.8%), HER2 enriched (3.5%), and basal-like (13.3%). Clinical HR status significantly influenced the distribution of intrinsic subtypes in HER2-low BC.²¹  HR-positive, Luminal A subtype tumors were more frequent in HER2-low BC (58.9%) versus HER2 IHC score 0 (2.8%; P < .001). In contrast, the Luminal B subtype showed lower frequency in HER2-low BC (8.0%) versus HER2 IHC score 0 tumors (34.9%; P < .001). The frequency of HER2-low BC was lowest in the basal-like subtype (1.9% versus 33.4%; P < .001).²¹  Additionally, an individual gene expression analysis identified 34 differentially expressed genes in HER2-low BC versus HER2 IHC score 0 tumors, a pattern observed in hormone-positive cancers but not in TNBC. Also, downregulation of many genes associated with proliferation (CCNB1, CCNE1, MELK, MKI67, MYBL2) and basal-like features (KRT14, KRT17, KRT5, FOXC1, MYC), and tyrosine kinase receptors (EGFR, FGFR4) was observed in HER2-low BC.²¹  Conversely, HER2-low BC showed differential upregulation of several genes related to the Luminal subtype (BCL2, BAG1, FOXA1, ESR1, PGR, AR).²¹ 

The HER2-low BC category encompasses a heterogeneous group of cancers. Comprehensive studies of the prognosis, treatment, and outcome data on patients with HER2-low BC are limited and show conflicting results. HER2-low BC displays a range of HER2 IHC reactivity. At the higher end of the spectrum, BC with HER2 IHC equivocal (score 2+) expression shows higher histologic grades and proliferation rates than ER-positive/HER2-negative (scores 0 to 1+) BC.^22–24,44  Multiple studies have reported poorer prognosis in HER2 IHC score 2+ cancer compared with HER2-negative (score 0 or 1+) BC, an observation that holds even when these tumors are categorized by HR status.^44–46  In contrast, in an exploratory overall survival analysis based on 2 separate data sets in 1304 patients with a median follow-up of 90.3 months, Schettini et al²¹  reported no statistically significant difference in overall survival between HER2-low BC and HER2 IHC score 0 groups. Furthermore, the NSABP-47 trial showed no clinical benefit from adding trastuzumab to chemotherapy in patients with early-stage HER2-low BC.⁴⁷ 

In HER2-positive BC, intratumoral heterogeneity (ITH) for HER2 protein expression and HER2 gene amplification has been well documented. Published studies report a frequency of HER2 genetic heterogeneity between 2.7% and 13% of HER2 positive BC.^48–57  Although precise diagnostic criteria to define the presence of HER2 ITH have not been established for use in clinical practice, its presence is inherently acknowledged in the 2018 ASCO/CAP HER2 guidelines, which adopt a greater than 10% positivity threshold for IHC and ISH to classify tumors as HER2 positive.²  Multiple patterns of HER2 genomic ITH have been described, including tumors with discrete regional variations in expression in a “clustered pattern” and more diffusely intermingled variations.⁵⁸  The frequency of ITH for HER2 in HER2-low BC is not known. However, studies demonstrating a high frequency of HER2 genetic heterogeneity in HER2 equivocal BC provide indirect evidence that ITH may be prominent in HER2-low BC.^50,57,59,60  In HER2-low BC, the presence of ITH could potentially have predictive implications for targeted treatment response. A published study in HER2-positive cancers treated neoadjuvantly with T-DM1/pertuzumab without chemotherapy showed lower pathologic complete response (pCR) rates in tumors displaying HER2 ITH.⁶¹ 

In addition to the potential negative impact on the therapeutic efficacy of HER2-directed therapies, HER2 ITH may increase the difficulty of identifying HER2-low BC. In routine clinical practice, HER2 initial assessment is performed on the diagnostic core needle biopsy specimen. Per the current 2018 ASCO/CAP HER2 guideline recommendations, in tumors in which the initial HER2 result is negative, repeat testing of HER2 may be ordered on the subsequent resection specimen.²  The risk of sampling error inherent in breast core biopsy specimens combined with the potentially high prevalence of HER2 ITH in HER2-low BC could compromise the optimal identification of HER2-low BC. Additionally, the clone of HER2 antibodies can also influence the identification of patients with low levels of HER2 protein overexpression.^20,62 

The development of anti-HER2 targeted therapy has revolutionized the treatment of patients with HER2-positive BC.^63–66  During the past decades, HER2 testing has continuously evolved according to the cumulative clinical evidence of the anti-HER2 treatment effect. Trastuzumab, a monoclonal antibody that targets domain IV of the external domain of HER2, was the first anti-HER2 targeted agent approved by the US Food and Drug Administration (FDA) in 1998, showing activity in patients with metastatic HER2 scores 2+ and 3+ and weak to strong complete membrane staining observed in more than 10% of the tumor cells.^67,68  The response rates were 18% for HER2 score 3+ and 6% for HER2 score 2+ (P = .06), revealing higher efficacy with higher expression without ruling out potential clinical impact in lower HER2 expression levels. The first ASCO/CAP guideline, published in 2007 for HER2 testing in BC, was based on the available clinical evidence, mainly from adjuvant trastuzumab/chemotherapy trials.^{63–66,69,70}  The second anti-HER2 targeted agent approved by the FDA in 2007 was lapatinib, a small-molecule tyrosine kinase inhibitor that acts by blocking HER1 and HER2, followed by the approvals in 2012 of pertuzumab, a monoclonal antibody that binds to HER2 at a site different than trastuzumab, and in 2013 of T-DM1, an ADC composed of trastuzumab and emtansine.^71–73  During that time, the hypothesis of the potential benefit of trastuzumab in early-stage HER2-low BC was definitively abandoned after the negative results of NSABP B-47.⁴⁷  A timeline of FDA approvals of HER2-targeted BC therapies is shown in Figure 1.

Until the initial clinical trial publications on the effects of T-DXd in patients with metastatic HER2-low BC in 2015, the updating of guidelines and recommendations for HER2 testing were driven by the available clinical evidence essentially from trastuzumab-based therapies.⁷⁴  T-DXd is a second-generation anti-HER2 ADC, engineered with an enzymatically cleavable peptide linker that is stable in plasma and with the topoisomerase I inhibitor payload deruxtecan.⁷⁴  T-DXd has a drug-antibody ratio (DAR) of approximately 8, twice the DAR of T-DM1.⁷⁵  In the study DESTINY-Breast01, T-DXd showed durable antitumor activity in pretreated patients with HER2-positive metastatic BC, with a median of 6 prior treatment regimens.⁷⁶  The overall response rate was 60.9% (95% CI, 53.4%–68.0%). Likewise, preclinical studies indicated that T-DXd could be effective in the HER2-low BC setting.⁷⁷  A subsequent report by Modi et al⁷⁸  provided further evidence supporting the development of T-DXd, specifically in the HER2-low BC population, with an overall response rate of 37.0% (95% CI, 24.3%–51.3%) and a median duration of response of 10.4 months (95% CI, 8.8 months to not evaluable).⁷⁸  More recently, the randomized phase 3 study DESTINY-Breast04 results were presented at the plenary session of the ASCO annual meeting, revealing T-DXd prolonged progression-free and overall survival of patients with unresectable or metastatic previously treated HER2-low BC, when compared with physicians’ choice of therapy.⁷⁹  Among all patients, the median progression-free survival (PFS) was 9.9 months in the T-DXd group and 5.1 months in the physician’s choice group. Overall survival was 23.4 months (T-DXd) versus 16.8 months (physician’s choice).⁷⁹ Figure 2 illustrates the percentage of patients with a confirmed objective response during the DESTINY-Breast04 study.

Potential explanations for the effectiveness of T-DXd in HER2-low BC include the different payload mechanism of action and cell permeability properties, the higher DAR, and the unique characteristics of the linker, all contributing to a bystander effect and ultimately requiring less HER2 expressed in the tissue for a relevant clinical effect. Interestingly, other second-generation ADCs may also present bystander efficacy; preclinical and clinical data support this effect for vic-trastuzumab-duocarmazine (SYD985) and ARX788.^14,80 

SYD985 is a second-generation anti-HER2 ADC with a prodrug (seco-duocarmycin–hydroxybenzamide–azaindole [seco-DUBA]) as payload. When seco-DUBA is cleaved at the tumor site or intracellularly, it releases the active toxin DUBA. The linker drug is covalently bound to the monoclonal antibody trastuzumab.¹⁴  In the phase 3 TULIP study, including patients with previously treated metastatic HER2-positive BC, SYD985 improved PFS compared with treatment per physician’s choice.⁸¹  A different anti-HER2 ADC, ARX788, is engineered with a nonnatural amino acid that allows site-specific chemical conjugation of the linker and payload; this permits a homogeneous distribution of DAR 2.⁸²  ARX788 uses the potent anti-tubulin agent AS269 as a drug linker. This is a noncleavable linker, decreasing the off-target release of the payload. RC48-ADC is another second-generation anti-HER2 ADC, composed of the monoclonal antibody hertuzumab conjugated to monomethyl auristatin E (MMAE) through a cleavable linker.⁸³  Preclinical data suggest RC48-ADC is effective in HER2-low BC tumor environments.⁸⁴ 

More recently, the FDA approved margetuximab for metastatic HER2-positive BC. Margetuximab (MGAH22) has been engineered with increased affinity to both isoforms of CD16A with modifications that increase binding to both isoforms of CD16A Fc-γRIIIA stimulatory receptors on natural killer cells.⁸⁵  CD16A is critical for the antibody-dependent cell-mediated cytotoxicity properties of monoclonal antibodies.^85–88  Efficacy was evaluated in SOPHIA, a phase 3 trial including patients with metastatic HER2-positive BC, where the median PFS was 5.8 and 4.9 months for margetuximab/chemotherapy and trastuzumab/chemotherapy, respectively (hazard ratio, 0.76; 95% CI, 0.59–0.98; P = .03); however, the overall survival was not statistically different between treatment arms.^89,90  In the HER2-low BC setting, margetuximab has not been clinically effective.^91,92 

Bispecific antibodies are also under development. For instance, zenocutuzumab (MCLA-128) is a bispecific HER2/HER3 antibody that, by docking onto HER2, blocks NRG1 fusion protein binding, preventing HER2/HER3 heterodimerization.^93,94  Preclinical studies suggest that zenocutuzumab may be active against HER2-positive and HER2-low cells stimulated with heregulin.⁹⁵ 

The vaccines and tyrosine kinase inhibitors exhibit completely different mechanisms of action. The vaccine nelipepimut-S (NPS), an immunogenic peptide derived from the HER2 protein (HER2 369–377), has been primarily developed in the HER2-positive space.^96,97  In a phase 3 study, NPS combined with granulocyte-macrophage colony-stimulating factor (GM-CSF) versus placebo plus GM-CSF to prevent BC recurrence showed no difference in disease-free survival.¹⁵  However, a subgroup analysis of a phase 2b study investigating NPS + GM-CSF + trastuzumab + versus GM-CSF combined with trastuzumab (placebo) suggested that there may be a benefit for patients with TNBC who had HER2-low BC.⁹⁸ 

The tyrosine kinase inhibitors lapatinib, neratinib, and tucatinib are approved by the FDA for use in patients with HER2-positive BC. Pyrotinib combined with capecitabine has been approved in China to treat patients with metastatic HER2-positive BC.⁹⁹  The combination of pyrotinib with capecitabine has shown significantly longer median PFS versus lapatinib (12.5 versus 6.8 months; P < .001).⁹⁹  Pyrotinib, neratinib, and tucatinib have been evaluated to treat HER2 mutated solid tumors. In addition, pyrotinib is being tested in the HER2-low space.^3,100–102 

The growing interest in treating HER2-low BC is based on the increasing number of studies exploring treatments for this patient population. The evolution of HER2-low BC as an actionable biomarker of anti-HER2 therapies has raised questions that remain largely unanswered. For example, what is the lowest level of HER2 expression to allow second-generation anti-HER2 ADCs sufficient payload delivery to tissues? What are the advantages and disadvantages of different linkers? Are resistance mechanisms more related to the mechanism of action of the payload, to the linker, or to the antibody structure? Multiple completed or ongoing clinical trials evaluate several anti-HER2 therapies, specifically in the HER2-low BC setting (Table). These phase 1, 2, and 3 clinical trials include evaluating trastuzumab naked antibody, vaccines, ADCs, and a tyrosine kinase inhibitor, covering early-stage HER2-low BC and metastatic disease.

In light of the above, there is an increasingly critical need for clear communication and collaboration between pathologists and medical oncologists to improve treatment and expand patient research opportunities.

The 2018 ASCO/CAP guideline for HER2 testing in BC defined algorithms for evaluating HER2 protein expression by IHC and HER2 gene amplification by ISH and outlined the interpretation criteria.²  Quality assurance measures to ensure HER2 testing and reporting accuracy have mainly focused on detecting HER2-positive BC to assess eligibility for trastuzumab HER2-targeted therapy. In BC with HER2 expression below the threshold for HER2-positive disease, there is a continuous distribution of HER2 expression. According to the current guideline, HER2 IHC 0 and 1+ are diagnosed as HER2 negative and require no further testing if there is no apparent histopathologic discordance. Although the ASCO/CAP guideline clearly defines IHC 0 and 1+, there has been limited clinical significance in distinguishing between these categories because the first generation of HER2-targeted therapies offered no clinical benefits in patients with HER2-negative disease.⁴⁷ 

Moreover, the 2 groups, HER2 IHC 0 and 1+, are often combined as HER2 IHC negative in interpreting and reporting.^2,103  Still, recent clinical studies demonstrated the clinical efficacy of anti-HER2 ADCs in patients with low levels of HER2 expression.^3,14,79  Accordingly, the new therapeutic category HER2-low BC has been introduced.¹⁰²  Therefore, regardless of the development of other assays, the distinction between HER2 IHC 0 and IHC 1+ following the current ASCO/CAP guideline has become critical. There is an urgent need for awareness, clinical implementation, and quality assurance for the optimal identification of HER2-low BC.

The growing need for molecular analysis in clinical oncology demands that companion diagnostic assays, such as HER2, be accurate and reproducible. Variability of the molecular integrity of BC tissues could result in erroneous biomarker results, with a potential negative impact on patient care. Biomarker testing must be judiciously controlled, with appropriate quality assurance measures. Accordingly, all laboratories performing BC biomarker testing must comply with quality control and quality assurance recommendations outlined in published national guidelines, including the standardization of preanalytic variables.² 

Preanalytic factors can impact the quality of ER, progesterone receptor (PR), and HER2 testing¹⁰⁴  and can adversely affect the integrity and molecular repertoire of tumor tissue.¹⁰⁵  Preanalytic factors include different processes beginning with the warm ischemic time (time from ligation of tissue blood supply to the removal from the patient); cold ischemic time (time from the removal from the patient until the tissue is placed in an appropriate fixative and stabilized for a proper length of time); and ending with tissue processing and embedding, in preparation for morphologic and molecular analysis.¹⁰⁶  The ligation of the blood supply during surgery will lead to tissue hypoxia, ischemia, and the progressive degradation of macromolecules of potential clinical interest.¹⁰⁷ 

In a study by Yildiz-Aktas et al,¹⁰⁸  breast resection specimens were subjected to varying cold ischemic times within the refrigerator and at room temperature.¹⁰⁸  These samples were processed and stained for ER, PR, and HER2. The results were compared with the prior needle core biopsies from the same patient. IHC staining for HRs and HER2 was significantly reduced after 4 hours for refrigerated samples and 2 hours for nonrefrigerated samples. Results such as this raise a significant concern that prolonged cold ischemic time for tumor tissue samples could potentially result in patients having an incorrect receptor-negative classification.²  This may be particularly important when detecting biomarkers at lower expression levels, such as HER2-low BC, further emphasizing the need to standardize preanalytic variables for BC specimens. In addition to having high-quality specimens for HER2 analysis, careful attention must be paid to the interpretation of assay results, requiring appropriate training and the application of the criterion set forth by the ASCO/CAP recommendation.^2,109 Figure 3 shows the criteria proposed by the ASCO/CAP panel to interpret HER2 IHC assay results.^2,110  The promising efficacy of novel HER2-targeted therapy in advanced/metastatic HER2-low BC has raised the possibility of changing the clinical interpretation of HER2 status in BC to include an HER2-low BC category. The accurate interpretation of HER2-low BC will require a careful evaluation of HER2 IHC assays at lower receptor expression levels.

Lastly, consideration must be given to potential analytic differences in the performance of HER2 IHC assays. Currently, several FDA-approved IHC assays are cleared for clinical HER2 assessment. These use different antibodies, detection, and retrieval systems and have different performance characteristics. Although comparative studies of agreement and analytic validation of these assays have been well studied in identifying HER2-positive disease, less is known about their comparative performance in detecting tumors with lower levels of HER2 expression. For example, recent studies comparing 2 widely used HER2 IHC assays, the HercepTest (A0485, Dako/Agilent) versus the Ventana PATHWAY Anti-HER2/neu (4B5, Roche), suggest comparative differences in assay sensitivity in tumors with low levels of HER2 ISH amplification¹¹¹  as well as antibody-related differences in interobserver and interantibody reproducibility in HER2-low BC.¹¹²  Scott et al¹¹³  studied 500 BC samples using the 4B5 assay and the Hercep test. Results showed that the 4B5 assay identified 28.0% of HER2 IHC 1+/2+ samples compared with 11.6% using the Hercep test.¹¹³  A recent study by Rüschoff et al¹¹⁴  compared the performance of the Hercep test mAb pharm Dx (GE001, DAKO Omnis) versus the 4B5 assay in 119 BC samples covering the entire range of HER2 IHC expression. Although there was a high concordance (98.2%) for the distinction of HER2 negative (IHC 0, 1+, 2+ and fluorescence in situ hybridization [FISH] negative) and HER2 positive (HER2 IHC 3+, 2+ and FISH positive), there was only a complete agreement for individual scores in 69.7% of cases. The Hercep test was significantly more sensitive for identifying HER2-low–expressing samples (35% versus 19%; P < .01).¹¹⁴ 

The ASCO/CAP 2018 guideline has classified invasive BC into 2 categories: HER2 positive (either using 3+ IHC or ISH amplification) and the remainder of BC into HER2 negative.²  These efforts were predicated on the importance of clinical linkage between existing pharmacologic therapies (eg, trastuzumab and pertuzumab) and identifying patients for whom these therapies would provide clinical benefit. The ASCO/CAP 2018 guideline did not require reporting and resulting from enumerating the heterogeneity of HER2 expression that exists at the 1+, 2+, and 3+ levels, mainly because no clinically actionable therapy required capturing heterogeneity in great detail, despite the knowledge that tumors with less than 100% HER2 overexpression/gene amplification do not respond as favorably to anti-HER2 treatment.^61,115  The ITH of HER2-low BC is depicted in Figures 4 through 6.

In contrast to the use of CAP synoptic tumor reporting templates, HER2 IHC/ISH reporting is currently done using free text.^116,117  Given the current ASCO scoring guideline, this makes sense because there is no need to capture fine detail to enumerate HER2 heterogeneity. However, there is value in moving toward a synoptic reporting template for capturing the percentage of tumor cells with membrane staining at each intensity level (from 0 to 3+) to capture HER2-low BC and ITH. Synoptic reports provide the opportunity to capture tumor HER2 membrane staining heterogeneity and offer scoring reminders for clinicians to recall clinically actionable cutoffs. Moreover, they capture all the information in discrete database fields, which is valuable for supporting downstream clinical decision-making.^116,118  Modern electronic health record (EHR) systems provide for best practice alerts and clinical decision guidelines if discrete data are available to the EHR. Discrete data stored as data elements within a synoptic report can enable reminders of best practices, link specific therapeutic options, and monitor adherence to the National Comprehensive Cancer Network practice guideline as therapies for HER2-low BC become available. EHR systems from vendors like Epic and Cerner Corporation create synoptic reports linked to downstream clinical decision support as part of their HER2 and oncology management systems.

Manual quantification of HER2 IHC can suffer from interobserver and intraobserver variability even when the ASCO/CAP HER2 interpretation guidelines are followed strictly, particularly for the identification of HER2-low BC.^119,120  The recent report by Fernandez et al¹²⁰  showed a concordance of only 26% among 18 pathologists for the distinction of 0 versus 1+ HER2 IHC groups of immunostained BC biopsies.¹²⁰  Although quality improvement requires a holistic approach, once the preanalytic factors are addressed, there is great potential and opportunity for using digital pathology and artificial intelligence to address the quality of HER2 interpretation and scoring. Because of digital pathology and artificial intelligence advancement, computer-aided analysis of HER2 using whole slide images is gaining wider adoption to facilitate accuracy and reproducibility of HER2 IHC scoring/interpretation in clinical practice.^{119,121–123}  Many commercially available HER2 quantitative image analysis (QIA) algorithms have been approved by the FDA or are developed as laboratory-developed tests. Recently, there has been a growing interest in developing neural networks and deep-learning–based HER2 QIA.^124–127  There is a higher reproducibility in manually quantitated biomarkers at extreme expression ranges (HER2 IHC 0 or 3+) than in tumors with HER2 IHC 2 or 1+ to 2+ expression.¹²⁰  For the HER2 IHC 2+ tumors, HER2 ISH testing can be an arbitrator to distinguish the true HER2-positive tumors.²  Unfortunately, no accepted or clinically available analytic method exists beyond IHC to segregate HER2 IHC-negative from HER2-low BC. The accurate and reproducible nature of the computer-aided QIA provides a great potential solution to this practical issue.¹²¹  In addition, some HER2-positive tumors do not respond to the conventional HER2 regimens because of ITH; QIA can recognize these variable patterns within the tumor and support HER2 scoring to identify patients appropriately for HER2-targeted therapy.

The sensitivity of IHC in detecting very low clinically relevant HER2 has not been studied so far. Negative HER2 IHC expression does not mean the complete absence of HER2 protein expression, a limitation of the IHC methodology.¹²⁸  Potential methods of evaluating HER2 status that might be clinically relevant include RNA-based methods, such as reverse transcription–quantitative real-time polymerase chain reaction (RT-qPCR) used in the oncotype DX assay.^129–131  The comparison between RT-qPCR and the FDA-approved methods has shown conflicting results in detecting HER2 3+ and/or HER2 amplified tumors.¹³²  This is an area of further research to evaluate if RNA-based methods are more sensitive in detecting HER2-low BC.

The currently used tests for evaluating HER2 status, including IHC staining and ISH, do not directly quantify HER2 protein levels. Although these tests are helpful surrogates for determining HER2 protein overexpression, they are not perfect for selecting patients for anti-HER2 therapy. In recent years, the promising potential of targeted mass spectrometry (MS) for quantitative evaluation of HER2 protein levels using formalin-fixed, paraffin-embedded (FFPE) tissues has been reported.

Hembrough et al¹³³  reported the development of a multiplexed MS-based selected reaction monitoring (SRM) assay to measure attomole (10⁻¹⁸ mol) levels of HER2 in 29 FFPE tumor tissues.¹³³  In this study, SRM analysis of HER2 correlated with IHC scoring; HER2 immunostaining of 2+ or lower showed up to 344 amol HER2 per microgram, and those with 3+ staining showed a range of HER2 protein levels from 175 to 16 900 amol of HER2 per microgram tumor protein. This study demonstrated the feasibility of quantitative multiplexed analysis of proteomic markers in FFPE tissue, with a promising potential for selecting patients for anti-HER2 therapy. Steiner et al¹³⁴  reported the utility of SRM for evaluating HER2 peptide levels. They developed and validated a multiplexed SRM assay using 6 HER2 peptides in 40 FFPE tumor tissues with varying HER2 status determined by IHC and FISH.¹³⁴  The amounts of the 6 peptides were highly and significantly correlated with each other, indicating that peptide levels can serve as surrogates of protein levels in FFPE tissues. SRM peptide measurements predicted 90% of cases based on HER2 amplification, with a high agreement with IHC and FISH and good analytic performance concerning linearity, precision, and lower limit of quantitation of the peptides.

The most extensive clinical study using SRM for quantitation of HER2 protein levels, by Nuciforo et al,¹³⁵  tested the potential of this proteomic assay to quantitate HER2 peptide levels in 277 FFPE tumor tissues, including 142 HER2-positive specimens based on the results of IHC and ISH. They established an HER2 peptide threshold of 740 amol/μg to predict HER2 status by IHC/ISH and determined survival benefits after anti-HER2 therapy. This threshold showed an agreement of 94% with IHC and ISH. HER2 levels greater than 2200 amol/μg were significantly associated with patient outcomes. This study provided evidence that quantitative measurement of HER2 peptide levels by SRM was useful in predicting patient outcomes after anti-HER2 therapy.

Proteomic assays such as SRM MS have the potential to overcome the limitations of IHC by allowing the quantitation of HER2 protein levels with high sensitivity, specificity, and reproducibility. In addition, they can also enable multiplexing for the estimation of several biomarkers simultaneously. The clinical utility of SRM for determining absolute HER2 levels in the dynamic range and identifying HER2-low BC for selecting patients for anti-HER2 therapy needs to be established in prospective clinical trials.

The availability of anti-HER2 ADC treatment opens exciting new possibilities for treating HER2-low BC because of the retention of cytotoxic agents’ antitumor properties and an improved tumor-specific HER2-targeting homing effect.¹³⁶  These anti-HER2 agents allow the treatment of patients who were traditionally ineligible for treatment. A collaborative multidisciplinary approach between pathologists and oncologists is imperative to understand these patients’ clinical characteristics and prognosis, the patient journey, and the management of BC.^2,137 Figure 7 shows a proposed algorithm of HER2 status assignment by pathologists impacting treatment decisions made by medical oncologists.

The introduction of the category of HER2-low BC with the possibility of targeted therapy necessitates a more accurate assessment of HER2 protein expression at the lower end of the spectrum by practicing pathologists. Moreover, the commonly used reflex strategy based on IHC results or applying first-line ISH may deny potentially beneficial targeted therapy. Thus, routinely performing IHC and ISH assays in all BC cases will become necessary because some BCs show discordance between IHC and ISH.¹³⁸  These advancements may provoke further updates in the ASCO/CAP HER2 guideline to create an HER2-low BC category.

Medical writing and editorial assistance were provided by Elbalejandra Baquero, MSc, of Creative Lynx Ltd, and funded by Daiichi Sankyo Inc and AstraZeneca.