Abstract
Background/Aim: Low human epidermal growth factor receptor 2 expression (HER2-low: 1+/2+ by immunohistochemistry without HER2 amplification) is emerging as defining a specific breast cancer (BC) subgroup owing to its distinct biological features. However, its prognostic role has not been confirmed in clinical practice. We conducted a systematic review and meta-analysis to determine the prognostic role of HER2-low status in patients with estrogen receptor-positive (ER+) early BC. Materials and Methods: We searched PubMed, EMBASE, and the Cochrane Library for prospective or retrospective studies that reported data on overall (OS) or disease-free (DFS) survival for HER2-low compared to HER2-negative BC. Data were pooled using hazard ratios (HR) with confidence intervals (CI) for OS/DFS of HER2-low vs. HER2-negative subgroups according to the random-effects model. OS was the primary outcome measure, and DFS and pathological complete response were the secondary endpoints. Results: An analysis was made of 25 studies collected, including 34,965 patients with HER2-low BC. A HER2-low status was associated with an HR for OS of 0.83 (95% CI=0.76-0.9, p<0.0.01). Similarly, a pooled HR of 0.89 (95% CI=0.840.94, p<0.0.01) showed that patients with HER2-low BC had an increased DFS. Pathological complete response was significantly lower in HER2-low BC in 13 studies (OR=0.72, 95% CI=0.58-0.91; p<0.01). Conclusion: Based on these data, HER2-low status should be identified as a potential prognostic factor in early stage ER+ BC. This should be taken into account when considering treatment in (neo)adjuvant settings, and it should be a potential stratification factor in future investigations.
Traditional classification and treatment of tumors with human epidermal receptor 2 (HER2/ERBB2) overexpression/amplification is based on hormone receptor expression status. Approximately two-thirds of HR-positive and one-third of triple-negative tumors present low HER2 expression by immunohistochemistry (IHC) [HER1+ or HER2+/in situ hybridization (ISH) not amplified]. Even though these tumors are considered HER2-negative, they still have a measurable quantity of HER2 protein on their cell membranes. HER2-low breast cancer (BCs) may be distinguished from other forms of BC based on their low level of HER2 expression. They also represent a newly defined clinical entity that can be categorized as luminal if positive for expression of hormone receptors or triple-negative (TNBC) if expression of hormone receptors is negative (1). There is variability within this subgroup with regard to hormone receptor expression, which accounts for approximately half of the entire BC population. Tumors in the HER2-low group have IHC scores of 1+ or 2+ and do not show evidence of gene amplification. In accordance with this criterion, only patients who qualify as HER2-negative have a score of 0 on IHC. Patients without HER2 amplification are not administered HER2-blockade treatment; however, they are potential candidates for antibody–drug conjugates (ADCs) other than standard chemotherapy (CT)/endocrine therapy. This is due to the typically lower number of HER2 receptors normally present in HER2-low BC cases compared to cases harboring HER2 gene amplification. Targeting the small number of HER2 receptors on the surface of cancer cells enables the delivery of ADCs to these cells (2, 3). As the payload of these medications can pass through cell membranes, it is possible that they cause the release of cytotoxic agents and kill neighboring cells that do not express the HER2 receptor through a process known as the bystander effect. Following treatment with the ADC trastuzumab-deruxtecan, a partial response was observed in 37% of patients with metastatic HER2-low BC, according to the findings of a phase Ib study by Modi et al. (4). A confirmatory phase III trial found that treatment with trastuzumab-deruxtecan led to significantly longer progression-free (PFS) and overall (OS) survival than treatment with chemotherapy of the physician’s choice (5).
Because HER2-low BC is still a relatively novel entity, there is a paucity of data regarding its biology, as well as the impact of HER2-low status on prognosis, particularly in early disease. Previous studies have shown that patients with HER2-low tumors have a lower chance of achieving pathological complete response (pCR) following treatment with neoadjuvant CT [analyzed in (6)]. However, this predictive influence cannot be relied upon as in a large dataset in which HER2-low was compared to HER2 IHC 0 BC, the two groups had similar pCR rates (7).
In this systematic review and meta-analysis, we evaluated the prognostic value of HER2-low status in BC (IHC 1+/2+ scores compared to IHC 0) in patients with hormone receptor-positive early BC.
Materials and Methods
We performed this systematic review and meta-analysis in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and the Cochrane Handbook for Systematic Reviews of Interventions (8).
Search strategy. References for this systematic review and meta-analysis were identified through searches of PubMed, Cochrane Library, SCOPUS, Web of Science, and EMBASE from their inception to December 1, 2022. The search terms included the following: (“erbB2 low”[All Fields] OR “HER2-low”[All Fields] OR “HER-2 low”[All Fields]) AND (“breast neoplasms”[MeSH Terms] OR (“breast”[All Fields] AND “neoplasms”[All Fields]) OR “breast neoplasms”[All Fields] OR (“breast”[All Fields] AND “cancer”[All Fields]) OR “breast cancer”[All Fields]). Relevant studies were manually selected based on the functions of the related articles. The citation lists of all retrieved articles were analyzed to identify other potentially relevant reports.
Study selection and data extraction. The following criteria for eligibility among studies were set before collecting the articles: i) Expression of HER2 defined as low according to the canonical definition: Tumors that are IHC 1+ or 2+ with ISH negativity; ii) survival information (OS or BC-specific survival or disease-free survival (DFS)] at specific follow-up was reported in the article as a hazard ratio (HR) with 95% confidence interval (95% CI) according to multivariate or univariate analysis or rates of pCR after neoadjuvant CT; iii) early BC with estrogen receptor (ER) expression (stages I-III); iv) articles published in the English language; and v) when several articles were published by the same authors or group, the newest or most informative single article was selected. The exclusion criteria were as follows: i) TNBC; ii) letters to editors/commentaries, reviews, and articles published in a book, or published in a non-English language; and iii) metastatic disease. When studies included some TNBC cases, they were included provided that they were the minority of presented cases, or if the analysis was split according to ER status.
Two Authors (FP and GT) independently conducted the search and identification, and one article was selected by consensus with a third Author (OG). The following information was independently extracted from each report by the two Authors: Author/year of publication, country, patient number, study type, systemic therapy exposure (rate), survival data (HRs), and pCR rates.
Statistical analysis. For the analysis of the survival results, the HRs were pooled to provide an aggregate value. In this analysis, all HRs with 95% CIs obtained from multivariate analysis or univariate analysis and available in the articles were combined to obtain prognostic information regarding HER2-low status compared to the HER2-negative counterpart. Pooled odds ratios (ORs) for pCR in the HER2-low and HER2-negative BC arms were also calculated. The primary endpoint was HR for OS, and the secondary endpoints were DFS [or relapse-free survival (RFS) if reported] and pCR. Sensitivity analysis for OS was performed according to the median follow-up (≥5 vs. <5 years), quality (high vs. low quality papers), initial tissue analysis (core biopsy vs. surgical specimen), and type of analysis (multivariate vs. univariate).
Data were entered into Review Manager software v5.4.1 [Review Manager (RevMan), Version 5.4. The Cochrane Collaboration, 2020]. Cochran’s test was used to assess the heterogeneity of the included studies. For heterogeneity tests, a value of p<0.05 was considered to indicate significance. A random-effects model was used to pool the estimates across studies using the Der Simonian–Laird method. By convention, an observed HR or Odds ratio of <1 implies better survival (or better pCR rate) for HER2-low BC. The Newcastle-Ottawa Scale (NOS) was used for quality assessment (9). This scale assesses the likelihood of bias in three domains: (i) Selection of study groups, (ii) comparability of groups, and (iii) ascertainment of exposure and outcome. Studies with scores ≥7, 4-6, and <4 were considered high, fair, and poor quality, respectively. We determined that the follow-up was adequate when the median follow-up period was >5 years.
Finally, we investigated the publication bias for OS meta-analysis with a visual inspection of funnel plots and with Begg–Mazumdar Kendall’s tau and Egger’s bias test (10, 11). Moreover, in the presence of publication bias for the primary analysis, we conducted a trim-and-fill-adjusted analysis (12) to remove the smallest studies from the positive side of the funnel plot and recalculated the effect size at each iteration until the funnel plot was symmetric about the new effect size. The risk of bias for nonrandomized studies (ROBINS) was also applied (13).
Results
In total, 839 potentially relevant citations were reviewed (Figure 1). Among these, 25 studies published between 2021 and 2022 that reported the prognostic value of HER2-low expression were analyzed. A total of 34,965 patients with HER2-low BC were included in this analysis. The major characteristics are presented in Table I (6, 7, 14-37).
Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram of included studies.
Characteristics of included studies.
In all publications, a retrospective analysis of the prognostic effect of HER2-low status was presented, with IHC tests performed on core biopsies (n=7) or surgical specimens (n=14). According to race, 14 publications included Asians, seven were on European patients, and the remaining were international or American cohorts. All were early or locally advanced ER+ BC; in nine studies, stage was not reported or TNBC patients were included in the stage classification. The quality of the studies as expressed by the NOS scores ranged from 5 to 9, with 32% including high-quality studies (NOS scores from 7 to 9). Only in one series did patients only undergo surgery and not receive systemic therapies (neither in adjuvant nor in neoadjuvant settings); in five articles, adjuvant therapies were not reported.
Meta-analysis of OS. Eighteen trials had OS data available. Although the heterogeneity test was not significant (I2=20%, p=0.2), a random-effects model was applied because of the retrospective nature of the studies. The pooled HR of 0.83 (95% CI=0.76-0.9; p=0.01) indicated that patients with HER2-low BC had significantly reduced risk of death (Figure 2).
HER2-low status and overall survival in breast cancer. CI: Confidence interval; HR: hazard ratio.
Meta-analysis of DFS/RFS and pCR rate. Twenty-one studies had DFS/RFS data; the heterogeneity test did not show any heterogeneity (I2=0%, p=0.68) (Figure 3). A random-effects model was used. A pooled HR of 0.89 (95% CI=0.84-0.94, p<0.01) showed that HER2-low BC was associated with a significantly increased DFS/RFS. The rates of pCR were significantly lower in HER2-low BC in 13 studies (odds ratio=0.72, 95% CI=0.58-0.91; p<0.001; data not shown).
HER2-low status and disease-free survival in breast cancer. CI: Confidence interval; HR: hazard ratio.
Sensitivity analysis. OS was better in studies where multivariate analysis was conducted (HR=0.85, 95% CI=0.77-0.93; p<0.001), in high quality studies (HR=0.9, 95% CI=0.8-1; p=0.06), in studies with longer follow-up (HR=0.88, 95% CI=0.82-0.95; p<0.01) and in neoadjuvant studies (HR=0.7, 95% CI=0.55-0.9; p<0.01).
Publication bias. A funnel plot (Figure 4, Begg’s test p=0.08) and Egger’s test (p=0.01) indicated the presence of potential publication bias. Trim-and-fill analysis also did not change the pooled estimates of the meta-analysis, and under the random-effect model, the pooled estimate for OS was 0.87 (95% CI=0.79-0.96).
Funnel plot for overall survival analysis.
Discussion
Very exciting results from recent clinical trials of novel HER2-targeting ADCs have focused on a newly defined subset of HER2-low BCs. However, the clinical and molecular features of HER2-low BC remain to be elucidated. This is an extremely important issue because the proportion of patients with HER2-low BC is substantial.
Several earlier studies have examined the role of low HER2 expression, namely HER2 IHC 2+ level (with negative ISH), as a potential independent prognostic factor compared to the HER2 IHC 0/1+, although the results have been inconsistent (38-41).
In contrast, our meta-analysis suggests that patients with early luminal BC expressing low levels of HER2 may have a longer OS than their HER2 IHC 0 counterparts. The risk of death from any cause was reduced by approximately 20% in this cohort. Similarly, in all 21 studies with available data, DFS and RFS were considerably enhanced, but to a lesser extent.
These findings contradict recent studies suggesting that there is no significant difference in OS between the HER2-low and HER2 IHC-0 groups. Won et al. studied 30,491 I-III stage BC patients from the Korean Breast Cancer Registry and found no significant difference in OS between the two groups. Despite hormone receptor status, HER2-low BC was associated with considerably improved BC-specific survival compared to HER2 IHC 0 BC. Intriguingly, in multivariate analysis, the impact of low HER2 expression on BC-specific survival was significant only in the subgroup with TNBC (16).
Additionally, findings from a pooled analysis of individual patient data have recently been reported (6). This study explored a cohort of 2,310 patients with HER2-non-amplified primary BC who were treated with neoadjuvant CT in four prospective clinical trials. The authors demonstrated that patients with HER2-low BC had a significantly longer DFS and OS than those with HER2 IHC 0 BC. These changes were detected only in hormone receptor-negative tumors when subgroup analysis was performed. In addition, the percentage of patients who achieved pCR was noticeably lower for those with HER2-low tumors than for HER2 IHC 0 tumors. This second finding, which is noteworthy, is consistent with the results of our study, which found that patients with HER2-low BC had significantly lower odds of complete pathological downstaging (odds ratio=0.72, 95% CI=0.58-0.91; p=0.01).
Pfeiffer et al. found that hormone receptor expression was associated with a higher likelihood of HER2-low status in the largest series ever published, including more than 700,000 HER2-low BC cases from the US National Cancer Database (42). Histologically, in non-ductal tumors (e.g., invasive lobular cancer), a higher proportion of HER2-low subtype was observed compared to ductal carcinomas. Additionally, 23.6% of HER2-negative patients experienced pCR, compared to 16.3% of HER2-low patients. The outcome was almost similar, with minimal association with better OS, particularly in stage III to IV hormone receptor–positive cancer.
HER2-low classification was also debated in the recent American Society of Clinical Oncology and College of American Pathologist updated review of HER2 testing in BC (43). The authors concluded that a new subclassification (HER2-low vs. HER2 IHC 0 vs. HER2-positive) may not be useful at this time. The reason is that to determine if there can be differential clinical benefits based on HER2 expression levels, additional clinical trials would be needed to confirm that IHC 0 samples include targetable levels of the HER2 protein. Additionally, the authors state that there is a risk that IHC test subclassification may yield false-negative or false-positive results, which might affect treatment decisions. Accordingly, the HER2 guidelines remain unchanged without adding any new subclassifications, as no new data support changes to the current scoring and reporting recommendations.
In three previous meta-analyses involving less updated literature searches, similar conclusions were reached. Yang et al. (in 9 and 8 studies available for OS and DFS in hormone receptor-positive early BC) found that patients with HER2-low BC had better survival outcomes than those with HER2 IHC 0, hormone receptor-positive BC (OS: HR=0.87, 95% CI=0.81-0.93; DFS/RFS: HR=0.91, 95% CI=0.85-0.96) (44). Ergut et al. found that there were 9 and 11 studies available for OS and DFS, respectively, in patients with hormone receptor-positive BC (45). There was a significant improvement for the HER2-low group in DFS (HR=0.88, 95% CI=0.83-0.94; p=0.001), and OS (HR=0.87, 95% CI=0.78-0.96; p=0.001). Furthermore, the HER2-low group less frequently had unfavorable risk factors. Lastly, Tang et al. analyzed 15 and 9 studies with OS and DFS data and found similar results, including lower pCR rates (HR=0.98, 95% CI=0.96-0.99; p=0.001; and HR=0.96, 95% CI=0.94-0.99; p=0.003) (46).
These cumulative findings suggest that the biology and clinical impact of low HER2 expression can differ according to hormone receptor status, and support the need for further investigation to better elucidate the biology of this clinicopathological entity.
To date, there is no compelling evidence to explain why patients with early-stage BC and HER2-low tumors have better survival outcomes than those with HER2 IHC 0. Various hypotheses have been developed, some of which include the different intrinsic subtypes of HER2-low and HER2 0 BC, factors commonly associated with reduced aggressiveness, and the existence of a more complex, unidentified biology associated with HER2-low BC. According to Denkert et al.’s study of translational observations (6), a significantly higher proportion of HER2-low malignancies was observed in the hormone receptor-positive BC subgroup than HER2 IHC 0 tumors. In addition, these hormone receptor-positive cancers are typically less aggressive than HER2 IHC 0 BCs. Compared to HER2 IHC 0 tumors, low HER2 positivity was highly correlated with a lower prevalence of grade 3 tumors, lower Ki-67 status, and fewer TP53 mutations. These factors are commonly associated with reduced tumor aggressiveness, and a higher ER expression status may explain the reduced pCR rate for neoadjuvant therapy in the HER2-low population. In some of the included series, where clinicopathological status was compared, we observed more favorable prognostic variables (less Ki-67%, more ER positive cases, and more grade 2 disease) in the HER2-low population compared to their HER2 IHC 0 counterparts.
One may ask whether histology matters in HER2-low BC (most commonly associated with ductal carcinomas). In a series of 666 invasive lobular BCs, most clinicopathological features did not differ between HER2-low and HER2-negative cases (47). However, when adjusting for tumor size, number of positive nodes, hormone receptor status, and local therapy received, patients with HER2-low status and lobular BC had worse DFS than those with HER2-negative tumors (HR=2.0, 95% CI=1.0-4.1; p=0.05).
Our meta-analysis had several limitations. Firstly, most studies had a relatively short follow-up period, meaning that late events may not have been recorded. Despite this, even studies with longer follow-up duration showed a significantly improved OS for the HER2-low subgroups. Secondly, there is a general tendency for disagreement among pathologists regarding HER2 testing; hence, a possible bias may have affected our findings. Thirdly, natural ER and HER2 receptor expression fluctuations before and after neoadjuvant CT may have selected BCs with a better prognosis. Fourthly, we included heterogeneous and retrospective studies with obvious intrinsic biases. Specifically, the evaluation of the patient’s hormone receptor, and HER2 statuses was based solely on local pathology reports, and no central validation of the patient’s pathological condition was performed. Fifthly, we were only able to compare HER2 IHC 1+ and 2+ with HER2 IHC 0; however, the difference, if any, between the two scores was not well defined. According to a series of studies conducted by Cherifi et al. (48), HER2 IHC2+/ISH non-amplification is an independent predictive factor for higher DFS and OS in patients treated with neoadjuvant CT compared to those with HER2 IHC 1+. This demonstrates that this score is a continuum rather than a binary variable. Finally, our research describes a prognostic effect; however, a better strategy to treat early-stage BCs with HER2-low disease has not yet been established, and general information, such as stage, age, Ki-67 status, and grade, remain the preferred decision-making criteria. However, we present the first aggregate analysis of a large HER2-low series and report a small but statistically significant benefit in terms of OS and DFS in this BC subgroup. Although our results showed a potential association between HER2-low status and favorable outcomes, further research is needed to determine whether this truly has any clinical significance.
In the phase III study DESTINY 04 (5), the researchers evaluated trastuzumab-deruxtecan, an HER2-directed immunoconjugate antibody, in a patient population which had received one or two lines of chemotherapy prior to inclusion. Trastuzumab-deruxtecan treatment significantly improved OS relative to chemotherapy in the ER+/HER2-low population, with a median of 23.9 versus 17.5 months (reduction in the risk of death: 36%). Another similar agent is sacituzumab govitecan, a first-in-class TROP-2-directed ADC, with a TROP-2 antibody conjugated to SN-38, a topoisomerase inhibitor, via a hydrolysable linker. In an unplanned subgroup analysis of the TROPICS-02 STUDY (49), this drug appeared to have similar efficacy in both HER2 IHC 0 and HER2-low subgroups, with median PFS of 6.4 versus 4.2 months (HR=0.58, 95% CI=0.42-0.79; p<0.001) compared to physician choice in the standard arm. These trials explained that, at least in metastatic disease, HER2-targeted therapy still plays a role compared with standard CT in patients with HER2-low BC.
Anti-HER2 therapies are currently being studied to determine whether they can improve the response rate when combined with endocrine agents or neoadjuvant CT (50, 51). Molecular profiling, broader genomic analyses and functional studies are required to clarify the uncertain aspects of this BC subtype. In addition, the observations observed in luminal BC are not consistent with those for triple-negative cancers. In a study by Jacot et al. (52), HER2 1+/2+ scores did not predict a different OS or pCR in patients with TNBC. Denkert et al. came to the opposite conclusion, whereas Tarantino et al. discovered that HER2-low status had no impact on outcome (6, 7).
In a consensus regarding the state-of-the-art for HER2-low BC by Tarantino and colleagues, a dedicated panel discussed the molecular and clinical features of the HER2-low entity (53). As noted by the authors, HER2-low cancer should not be viewed as a distinct molecular entity but rather as a heterogeneous group of tumors with similar biology and prognosis to HER2 IHC 0 cancer. Adjuvant settings were not discussed, and they recommended that clinical trials with HER2-low BC should be powered appropriately for hormone receptor-positive and hormone receptor-negative subgroups within dedicated clinical trials, stratified populations, or studies comparing subgroups based on hormone receptor expression. Pathologists should always report HER2 IHC scores (0, 1+, 2+, or 3+) for HER2 testing results according to the current American Society of Clinical Oncology and College of American Pathologist algorithm. Depending on whether HER2-low expression is reported, the clinician can decide whether targeted agents are appropriate or if other trials are available. Pathology reports should not mention “HER2-low”, while in clinical practice it remains a valid interpretation of HER2 status.
In conclusion, according to the findings of this study, low HER2-low status may be used to determine the prognosis of early-stage luminal BC. However, ad hoc studies are required to evaluate the response to therapy and the prognosis of early HER2-low BC.
Footnotes
Authors’ Contributions
All Authors contributed equally to the article writing.
Conflicts of Interest
None to declare.
- Received July 16, 2023.
- Revision received August 12, 2023.
- Accepted September 1, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.