Evaluation of estrogen receptor (ER), progesterone receptor (PR), and HER2/neu (HER2) biomarkers is standard of care for all cases of newly diagnosed invasive, recurrent, and metastatic breast cancer. Repeat analysis is also performed in select cases per College of American Pathologists/American Society of Clinical Oncology guidelines and other clinical indications. However, in specific scenarios, preanalytic and analytic variables may pose distinct challenges to testing.
To provide a review of select challenges in the testing of commonly performed breast cancer biomarkers ER, PR, and HER2 and outline best practices for overcoming these challenges.
Review of College of American Pathologists/American Society of Clinical Oncology recommendations, current literature, and personal experience of the author.
Attention must be given to specimen handling to ensure accurate ER, PR, and HER2 biomarker assessment and appropriate management of breast cancer patients.
Estrogen receptor (ER), progesterone receptor (PR), and HER2/neu (HER2) status have proven prognostic and therapeutic significance in patients with breast cancer.1 In general, good-prognosis, well-differentiated invasive carcinomas of luminal A type have the highest levels of ER/PR expression, often with strong, diffuse tumor cell positivity. Conversely, HER2-positive tumors are more aggressive and more frequently higher grade with high proliferative rate. However, the introduction of HER2-targeted therapy has resulted in improved outcomes, with 22% to 63% of patients achieving pathologic complete response after neoadjuvant chemotherapy (NAC) with HER2-directed therapies.2–8 Additionally, ER/PR and HER2 biomarkers may change with disease progression and in response to therapy.9–11 Thus, College of American Pathologists (CAP)/American Society of Clinical Oncology (ASCO) guidelines recommend ER, PR, and HER2 analysis for all new diagnoses of primary, recurrent, and metastatic breast cancer.12,13
In current practice, ER and PR are typically assessed by immunohistochemical (IHC) staining, evaluating protein expression. Just 1% tumor staining is often deemed positive because early studies such as that by Harvey et al14 showed improved disease-free survival, even at low levels, but with better disease-free survival correlating with higher expression. Of note, this study14 focused on patients who received adjuvant endocrine therapy alone. The 1% threshold for positivity has been challenged, though, as in the study by Fujii et al,15 which found that patients with ER/PR low-positive (1%–9%) (and HER2-negative) tumors who underwent NAC derived little benefit from adjuvant endocrine therapy, behaving more like those with triple (ER, PR, and HER2)–negative tumors. Example cases with variable ER IHC staining are shown in Figure 1, A through D.
Variation in estrogen receptor staining in invasive breast cancers, ranging from negative (0%) (A) to focal, low positive (1% to 9%) (B), patchy positive (≥10%) (C), and diffuse positive (100%) (D) (immunohistochemistry, original magnification ×200).
Variation in estrogen receptor staining in invasive breast cancers, ranging from negative (0%) (A) to focal, low positive (1% to 9%) (B), patchy positive (≥10%) (C), and diffuse positive (100%) (D) (immunohistochemistry, original magnification ×200).
HER2 is assessed by IHC to evaluate protein expression and/or by in situ hybridization (ISH) methods that evaluate gene amplification. Current CAP/ASCO guidelines do not recommend which test to run primarily but recommend reflexing to the other method when the first provides an equivocal result.13 However, because of faster turnaround time and lower cost, most laboratories start with IHC and reflex to ISH. HER2 reporting has changed in different iterations of CAP/ASCO guidelines. Immunohistochemical staining is graded in a semiquantitative manner, with no (0) to incomplete, faint (1+) membrane staining being negative; incomplete and/or weak/moderate (2+) membrane staining being equivocal; and complete, intense (3+) membrane staining being positive. Current thresholds have established that more than 10% of tumor cells must have staining for interpretation in a category (eg, just 5% of tumor cells with complete, intense membrane staining would constitute a 2+ equivocal and not a positive result).13 In previous CAP/ASCO guidelines this threshold was 30% of tumor cells.16 Example cases with variable HER2 IHC staining are shown in Figure 2, A through D.
Variation in HER2/neu immunohistochemical staining in invasive breast cancers, ranging from 0 (A) to 1+ (B) (negative) to 2+ (equivocal) (C) and 3+ (positive) (D) for overexpression (immunohistochemistry, original magnification ×200).
Variation in HER2/neu immunohistochemical staining in invasive breast cancers, ranging from 0 (A) to 1+ (B) (negative) to 2+ (equivocal) (C) and 3+ (positive) (D) for overexpression (immunohistochemistry, original magnification ×200).
Although HER2 ISH can be done by single-probe assay, currently most laboratories use a dual-probe method, allowing quantification of HER2 and control (typically CEP17) signals, with HER2:CEP17 ratio of 2 or greater and 4 or more HER2 copies/cell (group 1) constituting an amplified or positive result and HER2:CEP17 ratio less than 2 and average HER2 signals/cell less than 4 (group 5) constituting a nonamplified or negative result. Other HER2 ISH groups 2 (HER2:CEP17 ratio ≥2 and average HER2 signals/cell <4), 3 (HER2:CEP17 ratio <2 and average HER2 signals/cell ≥6), and 4 (HER2:CEP17 ratio <2 and average HER2 signals/cell >4 but <6) are now interpreted in combination with IHC to give a definitive positive or negative classification. Of note, in the typical algorithm of reflexing only IHC (2+)–equivocal cases, this would result in group 3 having positive and group 2 and 4 having negative overall HER2 status.17 Example amplified/positive (group 1) and nonamplified/negative (group 5) cases of HER2 fluorescence ISH are shown in Figure 3, A and B.
Example cases of HER2/neu fluorescence in situ hybridization with nonamplified (group 5) (A) and amplified (group 1) (B) results (original magnification ×600).
Example cases of HER2/neu fluorescence in situ hybridization with nonamplified (group 5) (A) and amplified (group 1) (B) results (original magnification ×600).
Current ER, PR, and HER2 CAP/ASCO guidelines have well-outlined, comprehensive standards for evaluation that allow laboratories to develop standard protocols that frequently result in timely, accurate reporting of results. However, there are relatively common scenarios in which preanalytic and analytic variables pose challenges to accurate evaluation.
PREANALYTIC VARIABLES
Warm Ischemia Time
Warm ischemia is the state in which a tumor has experienced loss of blood supply but remains in the warm environment of the body and undergoes enzymatic degradation. Prolonged warm ischemia time is more often seen with difficult, protracted surgical procedures or can be seen in portions of tumors that undergo rapid enlargement, where blood supply has been outgrown. Because warm ischemia is difficult to control or measure, effects of tumor prolonged warm ischemia time are less understood.
Cold Ischemia Time
Cold ischemia time (CIT) is the time from tumor removal from the patient until exposure to formalin fixation. Delay to formalin fixation, or prolonged CIT, has shown to cause decline in ER, PR, and HER2 in numerous studies, with increasing possibility for status change, or false-negative results, with increasing delay to formalin fixation.18–22 For ER/PR, false negatives secondary to prolonged CIT appear most likely to occur in tumors with low levels of expression (Figure 4, A and B).23
Example case with conversion of estrogen receptor staining from low positive (A) on biopsy to negative (B) on resection with cold ischemia time of more than 8 hours (immunohistochemistry, original magnification ×200).
Example case with conversion of estrogen receptor staining from low positive (A) on biopsy to negative (B) on resection with cold ischemia time of more than 8 hours (immunohistochemistry, original magnification ×200).
The CAP/ASCO optimal handling requirements state that CIT should be recorded and recommend CIT of 1 hour or less.12 However, some authors have suggested that CIT of up to 4 hours may be acceptable, especially if the specimen is chilled, as biomarker changes are minimal if CIT is within this timeframe.21,22,24,25
For core biopsy specimens, CIT is usually within minutes, but optimally should be recorded on the specimen requisition for quality assurance. For resection specimens, recording of accurate CIT requires knowledge of 2 specific time points: excision time, or time of removal from the patient, and time in formalin, or time that the tumor is exposed to formalin. Of note, time in formalin is not accurate when a large specimen is simply placed in a container of formalin, as tumor within resection specimens is typically surrounded by a layer of normal breast tissue that is only very slowly penetrated by formalin. Protocols for rapid specimen delivery, special triaging or grossing, and/or injection of the tumor with formalin can minimize CIT.26–28 Despite these protocols, optimal CIT is often more difficult to achieve for resection specimens because of delays in specimen delivery from the operating room and handling in the laboratory.
Fixatives
ER, PR, and HER2 studies are generally validated for use only with formalin-fixed, paraffin-embedded tissue that has been fixed in 10% neutral buffered formalin between 6 and 72 hours; when there is any deviation from this, it must at least be documented.29 Prolonged fixation has been shown to cause loss of biomarker expression, with the potential for false-negative results. Inconsistencies in ER, PR, and HER2 expression have also been seen with formalin fixation of less than 6 hours.30
In recent years there has also been significant work in validating and optimizing methods that allow for accurate evaluation of breast biomarkers using alcohol-based and dried smears used in cytology.31–33 Additionally, decalcification agents, which are notable for their deleterious effects on fluorescence ISH studies, fortunately only appear to have modest negative effects on ER, PR, and HER2 IHC.34 However, when a patient has a history of breast cancer and/or breast cancer metastasis is high on the differential diagnosis, cytology material can be preferentially reserved for cell block preparation. Likewise, if a bone metastasis is favored, there is often tissue that, because of replacement by carcinoma, can and should be preferentially processed without decalcification to avoid potential false-negative or indeterminate (failed test) results.
ANALYTIC VARIABLES
General
Laboratories performing ER, PR, and HER2 studies should meet accreditation guidelines. Proper optimization of antibodies and validation should be done prior to performing patient testing. Quality should be assessed via review of internal and external controls. External proficiency testing should also be performed to ensure accurate testing and interpretation. Additionally, CAP proficiency standards outline recording of the percentage of ER- and HER2-positive results, as a comparison with published benchmarks (eg, having no more than 30% of tumors being ER negative). Notably, if the laboratory falls outside of established benchmarks, further investigation may be done to assess if the deviation is appropriate or inappropriate. For example, a greater frequency of high-grade, ER-negative cancers may be appropriate at an institution with a younger, minority-enriched population.
Controls
When evaluating ER/PR IHC, internal control elements are often present in primary breast specimens. Internal as well as external controls should be evaluated alongside all patient tumors to help ensure appropriate results, especially when negative. If there is control failure, the assay should be repeated. In patients with specific well-differentiated invasive carcinomas (Figure 5, A through D), the negative ER/PR result (or the diagnosis/grade) should always be questioned and repeat ER/PR and/or additional studies are warranted.
Examples of well-differentiated invasive tubular (A), ductal (B), mucinous (C), and lobular (D) breast cancer subtypes, which are nearly always estrogen receptor positive (hematoxylin and eosin, original magnification ×200).
Examples of well-differentiated invasive tubular (A), ductal (B), mucinous (C), and lobular (D) breast cancer subtypes, which are nearly always estrogen receptor positive (hematoxylin and eosin, original magnification ×200).
For HER2 studies, there is often no internal positive control, and external controls are more relied upon. Of note, high-grade ductal carcinoma in situ is frequently HER2 positive and thus may serve as a positive control. However, care must be taken to evaluate HER2 in invasive carcinoma, not ductal carcinoma in situ, which it is frequently intermixed with, as HER2 status has prognostic and therapeutic significance only in the setting of invasive carcinoma.
SPECIFIC SCENARIOS
Testing Multiple Tumors
In the most recent (2010) CAP/ASCO ER/PR guidelines, recommendations for patients with multiple synchronous tumors state that “testing should be performed on at least one of the tumors, preferably the largest.”12 However, testing is warranted on smaller tumors with different histology and/or grade, especially if it is likely to result in a meaningfully different biomarker profile (for which there should be consideration of features and primary tumor results).35
Retesting Resection Specimens
ER, PR, and HER2 testing can usually be done reliably on biopsy material, providing valuable information to guide clinical therapy, especially in the decision to treat with NAC or to proceed to surgery first. However, retesting of tumor in resection material is relatively frequently performed. Specific CAP/ASCO indications for repeat testing include biomarker/histology discordance, small focus, indeterminate or equivocal result, and lack of internal control (ER/PR) on biopsy, among others.12,13
As noted previously, retesting of resection tumors can be problematic, as false negatives can occur because of prolonged CIT, especially at low/borderline biomarker levels, when specimen delivery and triage protocols are not in place.
Biomarker conversion of ER/PR and/or HER2 after NAC in patients without pathologic complete response has been shown to have prognostic significance, with worse overall survival for patients converting to triple-negative status.36 However, the decision to retest in the setting of residual tumor after NAC should be based on tumor characteristics (pre-NAC and post-NAC) and multidisciplinary discussion.
DISCUSSION
Accurate ER, PR, and HER2 evaluation is imperative for appropriate patient diagnosis and management. Detailed requirements and guidelines for testing help ensure accurate evaluation. However, certain variables, such as alternate fixatives/processing techniques, prolonged CIT, and shortened or prolonged formalin fixation can result in false-negative or indeterminate results. Fortunately, understanding of these variables and the scenarios in which they are likely to occur, as well as coordination between clinical and laboratory providers, can help circumvent biomarker misclassification and ensure accurate patient results.