Introduction
The World Health Organization reports that breast cancer is the most common cancer among women worldwide.1 In 2022, it was estimated that there were 670,000 deaths from breast cancer globally,1 and in 2024, it is estimated that 3,300 Australian women will succumb to the disease.2 Fortunately, advances in prevention, early detection, and treatment have led to a decline in breast cancer deaths over the past three decades.3 Key determinants of breast cancer treatment protocols include the presence of estrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (HER2) on the primary tumor.4,5 These hormone receptors are essential for cancer cell growth and serve as primary targets in breast cancer treatment.5
Hormone receptor (HR) status is predictive of response to treatment, influencing treatment choices, and is prognostic, with the hormonal profile affecting the risk of metastasis, recurrence rates, and survival.6 Approximately 70% of patients diagnosed with breast cancer are hormone receptor-positive (i.e. ER and/or PR positive), which generally indicates a better prognosis.6 HER2 amplification occurs in approximately 20% of breast cancers and is associated with a poorer prognosis.5,6 The most aggressive form of breast cancer, with the poorest prognosis, is triple-negative (negative for ER, PR, and HER2), which occurs in about 15% of patients.5,6
Approximately 25–30% of patients will experience breast cancer recurrence during their lifetime.7 Recurrent breast cancer is defined as cancer that reappears after initial treatment, either at the primary site or as a metastasis. Historically, it was assumed that recurrent breast cancer would retain the same receptor profile as the primary tumor. However, a review of the literature has demonstrated discordance rates for ER (19.3%, 95% confidence interval (CI) 15.8–23.4), PR (30.9%, 95% CI 26.6–35.6), and HER2 (10.3%, 95% CI 7.8–13.6) between primary and recurrent cancers,8 indicating that such discordance may worsen survival outcomes.9
To identify receptor discordance, a re-biopsy of the recurrent cancer is needed to compare its hormone expression with that of the primary cancer. Some guidelines recommend re-biopsy for recurrent breast cancer;10,11 however, this is not consistent across all guidelines, and the decision ultimately falls on the treating physician.4,10,12–14 Biopsies may not be performed for various reasons, including the historical assumption that the receptor profiles remain the same, the risks and inconvenience of the procedure, and the potential impact on the patient’s quality of life.15,16 However, recent literature increasingly questions these assumptions.9,15,17
This systematic literature review (SLR) aimed to examine studies on receptor expression in primary and recurrent breast cancer published from 2013 to 2023 and synthesize the current understanding of receptor discordance, its impact on prognosis, and whether primary tumor heterogeneity plays a role in the discordance.
Results
Study characteristics
A total of 10 studies comparing receptor expression in primary and recurrent breast cancer were included in this SLR. The majority of these articles were published in the last three years (n = 7, since 2020),18–24 with the remaining three published in 2018,25 2017,26 and 2014.27 These studies represent a wide geographical distribution, including Germany,21,25 the USA,23,24 China,18,20,22 Australia,27 the Netherlands,24,26 and India (Table 1).19 Most studies were retrospective cohort studies (n = 8), with only two being prospectively planned (Table 1).21,22 The included studies had significant duration, ranging from four years to 50 years.19,27 Sample sizes varied significantly, ranging from 20 (for a study of six years’ duration) to 1,173 patients (for a study of nine years’ duration), indicating a broad spectrum of study scales (Table 1). While most studies included data from more than 100 patients (n = 6),18,20–24 some studies (n = 4) provided results from a smaller number of patients (20 to 55),19,25–27 limiting the generalizability of findings to the broader breast cancer patient population.
Table 1Main characteristics of studies included in the systematic literature review
| Author | Study design | Duration | No. of patients | Age (range) | % Receptor positive at diagnosis | Sites of metastatic biopsy | Setting |
|---|
| Hu et al., 202318 | Single centre Retrospective analysis | 6 years (2014–2019) | 130 | Median 55 (24,86) | ER: 63%; PR: 60%; HER2: 32% | Liver, lung, bone, other | China |
| Shanthala et al., 202319 | Prospectively planned retrospective single-centre cohort study | 4 years | 51 | Median 46 (24, 68) | ER: 92%; PR: 82%; HER2: 10% | Lung, liver, bone, ovaries, adrenal, cervix, brain, pleura | India |
| Lv et al., 202220 | Single centre Retrospective analysis | 9 years (2010–2018) | 1,173 | Median 46 (27–82) | ER: 57%; PR: 53%; HER2: 30% | Liver, lymph nodes, chest wall, lungs, breast, bone, stomach, colon, other | China |
| Kolberg-Liedtke et al., 202121 | Prospectively planned retrospective multi-centre cohort study | 30 year (1980–2010) | 592 | Unspecified | ER: 66%; PR: 60%; HER2: 16% | Visceral, bone, lymph node/soft tissues | Germany |
| Zhao et al., 202122 | Retrospective single institutional cohort | 12 years | 426 | Mean 45.9 | ER: 59%; PR: 50%; HER2: 21% | Bone, lung, liver, CNS, lymph nodes, soft tissue | China |
| Chen et al., 202023 | Retrospective single institution cohort | 21 years (1998–2019) | 390 | Unspecified | ER: 76% ER; PR: 55% PR; HER2: 22% | Bone, liver, lung, brain | USA |
| Hulsbergen et al., 202024 | Retrospective muti-institutional cohort | 17 years (2001–2018) | 219 | Mean 51.85 (SD 10.61) | ER: 53%; PR: 36%; HER2: 44% | Brain | USA, Netherlands |
| Thangarajah et al., 201825 | Retrospective cohort | 6 years (2013–2018) | 20 | Mean 56 (21,70) | HR: 50%; HER2: 25% | Supraclavicular | Germany |
| Szekely et al., 201726 | Autopsy study | 13 years (2001–2014) | 25 | Unspecified | ER: 64%; PR: 28%; HER2: 4% | Lung, bone, liver, adrenal, CNS, Gynae organs, other | Netherlands |
| Cummings et al., 201427 | Retrospective longitudinal single centre cohort study of autopsies | 50 years (1957–2007) | 55 | Median 52 | ER: 49%; PR: 58%; HER2: 23% | Lung, bone, liver, adrenal, CNS, Gynae organs, other | Australia |
The study participants were generally representative of women diagnosed with breast cancer, with ages ranging from 21 to 82 years. The mean/median age reported in seven studies was approximately 50 years,18–20,22,24,25,27 aligning with typical diagnosis ages for breast cancer in women.3 However, it is important to note that these women had experienced recurrences following initial treatment, underscoring age as an independent risk factor for breast cancer recurrence.28,29 The receptor expression of the primary breast cancer at diagnosis was provided in all included studies and was comparable to the reported literature.5,6 Hormone receptor expression varied from 28% PR-positive to 92% ER-positive.26,30 Overall, approximately 60% of the primary breast cancers were HR-positive at initial diagnosis, comparable to published reports of approximately 70% of patients having HR-positive cancer at diagnosis.6 HER2 expression averaged at just over 20%, with a range between 4% and 44%,24,26 aligning with expected rates of HER2 amplification in breast cancers at diagnosis.5,6
The most common sites for metastasis (Table 1) were bone,18,19,21,22,26,27 liver,18–20,22,23,26,27 lungs,18–20,22,23,26,27 and brain,19,22–24,26,27 consistent with literature reports on breast cancer metastasis.30,31
Methods used for determining receptor status
It has been postulated that receptor discordance may stem from a genuine biological manifestation of tumor heterogeneity or technical challenges, such as the inconsistent reproducibility of IHC, techniques.32–34 Therefore, the first consideration in evaluating the literature was the methods used to determine receptor expression. To ensure consistency and comparability, this study included literature that utilized IHC only (for ER and PR determination) or IHC with FISH in the case of HER2 (Table 2).18–27 The Royal College of Pathologists in Australia recommends including FISH for equivocal IHC findings for HER2.35
Table 2Methods used to determine receptor expression in the included studies
| Author | Cancer receptor determination method | Threshold for HR +ve | Threshold for HER2 | Method of determination |
|---|
| Hu et al., 202318 | Information of pathology records and patients’ record followed by paired slides review - statuses were determined using IHC +/− FISH | >1% IHC | IHC±FISH# | Independent analysis |
| Shanthala et al., 202319 | Pathology database of specimens and patients’ medical records followed by paired slides review using IHC +/− FISH according to ASCO/CAP 2013/2018 guidelines | >1% IHC | IHC±FISH# | Independent analysis |
| Lv et al., 202220 | Pathology reports of patients - methodology unspecified | Unspecified | Unspecified | Patient pathology report |
| Kolberg-Liedtke et al., 202121 | Information of pathology records and patients’ record - statuses were determined using IHC | Unspecified | Unspecified | Patient pathology report |
| Zhao et al., 202122 | Information of pathology records and patients’ record followed by paired slides review - statuses were determined using IHC +/− FISH | >10% IHC | IHC±FISH# | Independent analysis |
| Chen et al., 202023 | Pathology database of specimens followed by paired slides review using IHC +/− FISH according to ASCO/CAP 2013/2018 guidelines | >1% IHC | IHC±FISH# | Independent analysis |
| Hulsbergen et al., 202024 | Information of pathology records and patients’ record - statuses determine using IHC +/− FISH | >10% IHC | IHC±FISH# | Patient pathology report |
| Thangarajah et al., 201825 | Information of pathology records and patients’ record - statuses were determined using IHC +/− FISH | >10% IHC | IHC±FISH# | Patient pathology report |
| Szekely et al., 201726 | Pathology database of specimens and patients’ medical records followed by available specimen slides and autopsy extraction of tissue samples using IHC +/− FISH according to ASCO/CAP 2013 guidelines | >1% IHC | IHC±FISH# | Independent analysis |
| Cummings et al., 201427 | Review of pathology and autopsy database of specimens using IHC +/− FISH | Any level of IHC | IHC±FISH# | Independent analysis |
IHC is widely used by pathologists to detect the presence of specific antigens or receptors in tissue samples, aiding in cancer identification and differentiation. In terms of breast tissue, there is a large range of markers that could be used to identify cancer, to differentiate benign lesions from carcinoma, or to differentiate between carcinoma in situ and invasive carcinoma.36 For this review, HR expression discordance was chosen as HR status is predictive of treatment and is a good prognostic indicator, influencing metastasis risk, recurrence rates, and survival.6 In contrast to IHC, FISH examines DNA expression related to a specific receptor, such as HER2 in breast cancer. In cases where the IHC result for HER2 is equivocal, FISH can identify patients with HER2-positive tumors who could benefit from HER2-targeted therapy.37 Identification of HER2 amplification is crucial as it is associated with a poor prognosis and a more aggressive form of breast cancer.6,37 IHC with FISH was used for HER2 status determination in 8 of the included studies.18,19,22–27 Two studies did not indicate the HER2 threshold used.20,21
Another important consideration was the threshold used to confirm receptor expression, commonly defined as the percentage of positive cells. The use of a threshold for scoring IHC results by counting the percentage of positive cells is common and was the first scoring system used by pathologists.38,39 A threshold of at least 1% positive cells indicates eligibility for hormonal therapy.36 The 2021 St. Gallen International Consensus Guidelines categorize breast cancer based on the percentage of positive cells: responsive (10%), response uncertain (1–9%), and nonresponsive (0%).12 Four studies used a 1% threshold for positive ER or PR receptor expression,18,19,23,26 while three studies used a 10% threshold.22,24,25 The remaining three studies did not report the threshold used.20,21,27
The final distinction was whether the researchers determined their own receptor expression results or used pathology department reports. Most of the included studies (n = 6) examined pathology database slides or biopsy samples independently,18,19,22,23,26,27 minimizing inter-observer variability and strengthening the findings. Four studies based their analysis on pathology reports produced at the time of clinical diagnosis.20,21,24,25
Results relating to receptor discordance
The primary purpose of this SLR was to examine ER, PR, and HER2 receptor expression in primary and recurrent breast cancer tissue. Table 3 provides an overview of receptor concordance, discordance, gain, and loss across the 10 included studies published between January 2013 and December 2023.18–27 Receptor concordance refers to unchanged receptor status between primary and recurrent breast cancer. Receptor discordance indicates a change in receptor status between primary and recurrent breast cancer. Receptor gain occurs when a receptor that was not present in the primary tumor is detected in the recurrent tumor, while receptor loss refers to when a receptor that was present in the primary tumor is no longer detected in the recurrent tumor.
Table 3Rates of reported receptor concordance, discordance, or change in the included studies
| Study | Estrogen receptor
| Progesterone receptors
| HER2 receptors
|
|---|
| Concordance | Discordance | ER gain | ER loss | Concordance | Discordance | PR gain | PR loss | Concordance | Discordance | HER2 gain | HER2 loss |
|---|
| Hu et al., 202318 | 44.44% | 55.56% | 29.17% | 26.39% | 47.23% | 52.77% | 33.33% | 19.44% | 74.62% | 25.38% | 8.96% | 16.42% |
| Shanthala et al., 202319 | 52.94% | 47.06% | 23.53% | 23.53% | 31.37% | 68.63% | 23.53% | 45.10% | 94% | 6.00% | 1.96% | 3.92% |
| Lv et al., 202220 | 86.90% | 13.10% | 6.90% | 10.60% | 68.70% | 31.30% | 9.10% | 22.10% | 86.20% | 13.80% | 10.10% | 3.80% |
| Kolberg-Liedtke et al., 202121 | 81.3%# | – | 5.5%# | 13.2%# | see ER# | see ER# | see ER# | see ER# | 78.40% | 21.60% | 14.90% | 6.70% |
| Zhao et al., 202122 | 78.90% | 21.10% | 6.30% | 14.80% | 66.80% | 33.20% | 5.50% | 27.70% | 88.40% | 11.60% | 6.10% | 5.50% |
| Chen et al., 202023 | 81.66% | 18.34% | 2.58% | 15.76% | 59.71% | 40.29% | 9.28% | 31.01% | 86.29% | 13.71% | 5.92% | 7.79% |
| Hulsbergen et al., 202024 | 83.34% | 16.66% | 1.85% | 14.81% | 74.76% | 25.24% | 2.86% | 22.38% | 89.55% | 10.45% | 7.96% | 2.49% |
| Thangarajah et al., 201825 | 80.00% | 20.00% | – | – | 63.20% | 36.80% | – | – | 70.0% | 30.0% | – | – |
| Szekely et al., 201726 | – | | – | 62.50% | – | – | – | 90.91% | – | – | – | 50.0% |
| Cummings et al., 201427 | 56.40% | 43.60% | – | – | 67.30% | 32.70% | – | – | 96.40% | 3.60% | – | – |
| Average* | 80.69% | 19.31% | 7.19% | 14.22% | 65.51% | 34.49% | 9.54% | 25.61% | 84.96% | 15.04% | 9.59% | 5.97% |
ER concordance was reported in eight out of the 10 studies (Table 3).18–20,22–25,27 In one study,21 ER and PR were reported together as hormone receptors. On average, across the eight studies, ER concordance was reported to be 80.69% (discordance 19.31%). This is consistent with previously reported ER discordance rates.8 ER concordance ranged from 44.44% to 86.90%.18,20 Hu et al.18 reported a concordance of 44.44%, meaning that 44.44% of patients (or n = 58) of the 130 included patients had similar levels of ER expression in their primary and recurrent breast cancer. ER gain was reported in six out of 10 studies,18–20,22–24 with an average ER gain of 7.19%, ranging from 1.85% to 29.17%.18,24 ER loss was reported in seven out of 10 studies,18–20,22,24,26 with an average ER loss of 14.2%, ranging from 10.6% to 62.5%.20,26 These findings indicate dynamic changes in ER status over time in primary and recurrent breast cancer.
Similarly, PR concordance was reported in 8 out of the 10 studies (Table 3).19,20,22–27 On average, PR concordance was 65.51%, with discordance of 34.39%, aligning with previously reported rates.8 PR concordance ranged from 31.37% to 74.76%.18,24 This was again similar to the PR discordance of 30.9% reported previously.8 PR gain was reported in six out of 10 studies,18–20,22–24 with an average PR gain of 9.54%, ranging from 2.86% to 33.33%.18,24 PR loss was reported in seven out of 10 studies,18–20,22–24,26 with an average PR loss of 25.61%, ranging from 22.1% to 90.91%.20,26 These findings also indicate dynamic changes in PR status over time in primary and recurrent breast cancer.
Compared to ER and PR, HER2 concordance shows the least variation between primary and recurrent breast cancer (Table 3). A previous meta-analysis reported a discordance rate of 10.3% for HER2 between primary and recurrent breast cancer.8 HER2 concordance was recorded in nine out of the 10 included studies.18–25,27 On average, HER2 concordance was 84.97% (discordance 15.04), although HER2 concordance ranged from 70% to 96.4%.25,27 HER2 gain was reported in seven out of 10 studies,18–24 with an average HER2 gain of 9.59%, ranging from 1.96% to 14.9%.19,21 HER2 loss was reported in eight out of 10 studies,18–24 with an average HER2 loss of 5.97%, ranging from 2.49% to 50.0%.24,26 These findings indicate some dynamic changes in HER2 status over time. Due to the lack of data, no correlations between HR loss and HER2 receptor gain could be determined, but this remains an important area for future research.
Impact of hormone receptor loss on prognosis or metastasis
A measure of prognosis was reported in seven of the 10 included studies, although it was measured in various ways (Table 4).18,19,21-24,26,27 Three studies utilized Disease-Free Survival as a measure of prognosis,18,22,23 which is the time between initial diagnosis and recurrence. Three studies utilized Post-Recurrence Survival, or survival time from recurrence to death,21,22,24 and three studies utilized Overall Survival, which accounts for the time from initial diagnosis to death.22,23,27
Table 4Impact of receptor discordance on prognosis and metastasis
| Author | Disease free survival | Overall survival | Post recurrence survival | Other | HR influences on prognosis | HER2 influence on prognosis | Reported discordance in metastasis |
|---|
| Hu et al., 202318 | X | | | Ki67 marker | +ve receptor status had better prognosis than those with -ve receptor status | Trend (not significant) for HER2 expression or HER2 gain → prolonged survival | |
| Shanthala et al., 202319 | | | | Correlation with staging of cancer | ER loss more frequently associated with worse prognosis | No comment | Yes |
| Kolberg-Liedtke et al., 202121 | | | X | | +ve receptor status had better prognosis than those with -ve receptor status | HER2 loss was associated with poorer post recurrence survival compared to concordant HER2 | |
| Zhao et al., 202122 | X | X | X | | +ve receptor status had better prognosis than those with -ve receptor status | Prognosis related to treatment not HER2 expression | Yes |
| Chen et al., 202023 | X | X | | | +ve receptor status → better prognosis than those with -ve receptor status | Trend (not significant) for HER2 gain → prolonged survival | Yes |
| Hulsbergen et al., 202024 | | | X | | ER loss was identified with worse prognosis | No comment | |
| Szekely et al., 201726 | | | | | | | Yes |
| Cummings et al., 201427 | | X | | | +ve receptor status had better prognosis than those with -ve receptor status | No comment | |
All seven included studies that reported on prognosis observed a poorer prognosis associated with receptor loss.18,19,21–24,27 Two studies also noted that patients with receptor gain had a better prognosis with treatment.22,24 The impact of receptor loss in recurrent breast cancer on prognosis is important, as the included studies showed a higher degree of HR (ER and PR) loss compared to receptor gain (see the average scores at the bottom of Table 3; ER 14% loss, 7% gain; PR 26% loss, 9% gain).20–23 Two studies suggested that adjuvant endocrine treatment for primary cancer was likely associated with the loss of PR and ER in recurrent breast cancer.20,22 Similar to antimicrobial resistance, receptor discordance may be due to tumor heterogeneity in primary cancer, where some cells are successfully treated while others are resistant and seed recurrences.
The literature suggests that ER and PR expression is not only discordant during breast cancer recurrence (following diagnosis and subsequent treatment) but also unstable during the metastatic process.40–43 Shanthala and associates observed that PR loss may indicate a shift toward a more aggressive phenotype.19 Zhao et al.22 reported higher ER discordance in distant metastasis compared to local metastasis. Receptor discordance has also been noted in axillary lymph node metastases, identified during primary tumor biopsies and when assessing patients with multiple primary breast tumors.40,44–46 Chen et al.19 also observed discordance within the same organ and reported discordance between two different metastatic sites. Szekely found greater discordance of primary and recurrent tumors than between two separate recurrent tumor sites.26 The hypothesis is that HR receptor loss leads to a shift toward a more aggressive phenotype, which is more likely to metastasize, is associated with tumor recurrence, and is less responsive to treatment, leading to poorer prognosis (Fig. 2).32
Discussion
This systematic review of the literature provides evidence for receptor discordance between primary and recurrent breast cancer. This review adds to the existing evidence refuting the assumption that primary and recurrent breast cancer cells have uniform receptor profiles.16,47–49 The current study identified discordance rates comparable to those in a previous meta-analysis,8 for ER (19.3% vs 19.3%) and PR (34.9% vs 30.9%), but higher HER2 discordance (15.04% vs 10.3%). The precise cause of receptor discrepancy between primary and recurrent breast cancer remains unclear. Possible mechanisms include tumor heterogeneity, bio-evolution of the tumor, drug resistance, and differences in sampling/assay techniques.11,50–54 Very little has been explored in the literature to date on tumor heterogeneity as the mechanism of receptor discordance. Tumor heterogeneity, the concept that not all cells in a tumor are identical, may be the cause of the observed variability in the expression of ER, PR, and HER2.16,55,56
At its crudest level, the treatment of breast cancer assumes uniformity within the tumor. However, tumor heterogeneity has been observed both between different tumor lesions in the same patient (inter-tumoral heterogeneity) and within a single lesion (intra-tumoral heterogeneity).18,20,27,56 Tumor heterogeneity could lead to reduced treatment response, potentially affecting patient management and prognosis.57 While it adds complexity to treatment decisions, recognizing tumor heterogeneity is essential for personalizing treatment strategies. Two studies in the current review noted that primary cancer adjuvant endocrine treatment was associated with a loss of PR and ER in recurrent breast cancer.20,22 This may result from inherent heterogeneity within the primary tumor itself. This heterogeneity could manifest in two ways: adaptive evolution, where tumor cells develop resistance to treatments and evolve over time, resulting in receptor status changes; and selective replication, where a subpopulation of cells lacking certain receptors, such as ER and PR, survive treatment and continue to replicate, contributing to overall heterogeneity and receptor discordance in recurrent tumors.
Understanding tumor heterogeneity is crucial for comprehending disease progression and treatment outcomes, and it necessitates further research. Biopsies of recurrent cancer can offer insights into disease behavior, treatment resistance, and tumor cell evolution, potentially leading to improved targeting of treatments and better patient outcomes. However, whether metastatic tumors should be biopsied remains debated, despite numerous publications and some updated guidelines recommending it.4,10–14 All of the studies included in this SLR recommended re-biopsy of metastatic lesions if possible.18–27
The available research on the clinical significance of primary and recurrent tumor profiles is variable, with limited prospective data to guide clinical practice. Understanding the impact of ER, PR, and HER2 conversion on treatment schedules and breast cancer patient survival remains incomplete. The available data are limited, as is the optimal time for retesting tumor biology. In this context, clinical judgment remains crucial for guiding a reassessment of tissue biology. The decision to re-biopsy is multifactorial, involving considerations of patient impact (e.g., safe biopsy locations, acceptance of repeat biopsies, pain/discomfort, and treatment planning) versus the benefits of confirming receptor expression and making appropriate treatment choices. In the future, non-invasive diagnostics (liquid biopsies) could identify tumor cell markers, enhance clinical decision-making, and increase confidence in treatment choices for tumor recurrences, as well as being valuable aids for future research in this field.58
Agreement on study design or techniques is vital for ensuring the reliability and validity of results, allowing the field to progress. Collection of consistent demographic data, including treatments undertaken, age at diagnosis of primary and recurrent breast cancer, and patient menopausal status, would facilitate comparisons between studies. The retrospective nature of many of the included studies likely limited data availability, preventing comparisons between receptor discordance and treatments undertaken. Unravelling the complexity of tumor heterogeneity will likely involve a mix of methodologies, such as longitudinal studies, randomized controlled trials, and molecular profiling techniques. More retrospective studies could help identify patients most likely to benefit from certain treatments or at higher risk of recurrence. Focusing on these considerations would enhance our understanding of breast cancer, improve clinical practices, and revolutionize breast cancer research and clinical trial design, leading to more effective, personalized treatment strategies.