HER2 Immunohistochemistry in Invasive Micropapillary Breast Carcinoma: Complete Assessment of an Incomplete Pattern Open Access

After obtaining institutional review board approval, we performed a search of the surgical pathology database and identified all cases with a diagnosis of primary breast IMPC from January 2017 to September 2019 and available HER2 IHC staining performed at our institution. Information about the tumor, including histologic subtype, grade, and hormone receptor profile was extracted from the pathology reports. Clinical information including age at diagnosis, treatment, and follow-up was retrieved from the electronic medical records.

The hematoxylin-eosin (H&E) and HER2 IHC slides were reviewed by 2 authors. From morphologic review of the H&E slides, we assessed the percentage of MP component (≥90%, 60%–89%, 30%–59%, <30%), the distribution of tumor with MP morphology (focal, heterogeneous, or contiguous), the presence of mucin (≥90%, 60%–89%, 30%–59%, <30%), and the presence of apocrine cytology within the tumor. The pattern of distribution was defined as follows: focal—small, discrete area of MP carcinoma; heterogeneous—MP component consisting of multiple, scattered foci; and contiguous—diffuse and continuous areas of MP carcinoma or “pure,” with ≥90% of tumor showing MP pattern.¹ 

HER2 IHC (PATHWAY anti–HER2/neu [4B5], Ventana Medical Systems Inc, Tucson, Arizona) staining was performed on all cases of invasive carcinoma at the time of diagnosis. HER2 IHC was performed on the core biopsy specimen if available, and otherwise on the excisional material.¹⁷  All HER2 IHC equivocal cases were reflexed to HER2 FISH (HER2 IQFISH pharmDx, Dako, Carpinteria, California) in accordance with the ASCO/CAP guideline and the standard practices at our institution. Select HER2 IHC negative or positive cases were assessed by FISH either as per clinician request or for the purpose of this study. FISH was performed at the time of diagnosis for all HER2 IHC equivocal 2⁺ cases and performed on selected HER2 IHC 1⁺ cases at the discretion of the pathologist because of the specific MP morphology. For the purpose of this study, we performed FISH on selected HER2 IHC 3⁺ cases to confirm HER2 amplification.

The HER2 IHC–stained slides were reviewed to assess staining intensity (weak, moderate, or strong), distribution of the staining pattern (focal, heterogeneous, or contiguous), and the pattern of membranous staining (basolateral, complete, incomplete, or absent) (Figure 1, A through D). All HER2 IHC slides were reviewed and reevaluated. The 2018 ASCO/CAP HER2 guideline¹⁷  was applied, with the exception of the rules set forth for IMPC. Cases were excluded if IHC was equivocal and FISH results were not available. For the purpose of this study, IMPC with only basolateral membrane staining and strong intensity in more than 10% of tumor cells was classified as positive (3⁺) (Figure 2, A), and IMPC with only basolateral membrane staining with weak to moderate intensity in more than 10% of the tumor cells was classified as equivocal (2⁺) (Figure 2, B). Otherwise, as described in the guideline, cases with complete membrane staining demonstrating strong intensity in more than 10% of tumor cells were classified as HER2 IHC positive (3⁺), cases with weak to moderate complete membrane staining in more than 10% of tumor cells were classified as HER2 IHC equivocal (2⁺), cases with incomplete (not basolateral) membrane staining barely perceptible and in more than 10% of tumor cells were classified as HER2 IHC negative (1⁺), and cases with no staining or incomplete membrane staining barely perceptible and in less than 10% of tumor cells were classified as HER2 IHC negative (0).¹⁷  For cases with percentage of MP component below 90%, we documented the IHC staining pattern for each respective histologic subtype and noted if any significant differences were present. We also noted the predominant staining pattern (basolateral, complete, incomplete, or absent) for cases with more than 1 staining pattern. HER2 FISH results were classified according to the 2018 ASCO/CAP HER2 guideline¹⁷  and categorized into groups 1 to 5.

Through a search of our database 187 IMPC cases were identified and suitable for evaluation from 177 women and 4 men. The median age was 58 years, ranging from 30 to 95 years (Table 1). Procedure was core biopsy in 33% (61 of 187) of cases and surgical resection in 67% (126 of 187). Three patients had HER2 testing on both the biopsy and surgical specimens, 2 patients had multifocal carcinoma and testing was performed on each focus, and 1 patient had testing on both primary breast tumor and local recurrence. The tumor grade according to Nottingham system was grade 1 in 3% (6 of 187) of cases, grade 2 in 64% (119 of 187), and grade 3 in 33% (62 of 187). Receptor profile was ER⁺ PR⁺ HER2⁻ in 67% (125 of 187) of cases, ER⁺ PR⁻ HER2⁻ in 8% (15 of 187), ER⁺ PR⁺ HER2⁺ in 12.5% (24 of 187), ER⁺ PR⁻ HER2⁺ in 6.5% (13 of 187), ER⁻ PR⁺ HER2⁺ in 1% (2 of 187), ER⁻ PR⁻ HER2⁺ in 3% (5 of 187), and ER⁻ PR⁻ HER2⁻ in 2% (3 of 187) (Table 1).

A homogeneous (≥90%) MP component was found in 40% (75 of 187) of cases, 18% (33 of 187) had 60% to 89% MP component, 15% (28 of 187) had 30% to 59%, and 27% (51 of 187) had less than 30% (Table 2). In cases with an MP component below 30%, the pattern of distribution was predominantly focal. In cases with an MP component of 90% or more, the pattern of distribution was predominantly contiguous. The mixed histologic subtype in “non-pure” MP cases was invasive ductal carcinoma, not otherwise specified (IDC-NOS). Mucinous features were identified in 20% (38 of 187) of IMPC cases and apocrine cytology in 2% (4 of 187) of cases.

HER2 IHC staining pattern, distribution, and intensity were assessed for all cases (Table 3). The membranous staining pattern was predominantly basolateral in 49% (91 of 187) of the cases, complete in 14% (27 of 187), incomplete in 24% (44 of 187), and absent in 13% (25 of 187) (Figure 1, A through D). The staining intensity was absent in 13% (25 of 187) of the cases, weak in 42% (78 of 187), weak to moderate in 10% (19 of 187), moderate in 20% (37 of 187), moderate to strong in 1% (2 of 187), and strong in 14% (26 of 187). HER2 IHC was negative (0, 0 to 1⁺, or 1⁺) in 51% (95 of 187) of cases, equivocal (1⁺ to 2⁺, or 2⁺) in 35% (66 of 187), and positive (3⁺) in 14% (26 of 187) (Table 4).

HER2 FISH was performed in 41% (78 of 187) of cases, including 1 HER2 IHC negative (1⁺), 66 equivocal (1⁺ to 2⁺, 2⁺) and 11 positive (3⁺) cases (Table 4). FISH showed amplification in 26% (17 of 66) of HER2 IHC equivocal cases. Of those 17 IHC equivocal/FISH amplified cases, 76% (13 of 17) showed either entirely or a predominantly basolateral staining pattern (7 basolateral, 5 basolateral and complete, 1 basolateral and incomplete) and 24% (4 of 17) showed either entirely or a predominantly complete staining pattern (1 complete, 1 complete and basolateral, 2 complete and incomplete). Staining intensity was weak to moderate in 12% (2 of 17) of cases, moderate in 82% (14 of 17), and moderate to strong in 6% (1 of 17). The percentage of MP component in those 17 cases was homogenous (≥90%) in 53% (9 of 17) of cases, 60% to 89% in 29% (5 of 17) of cases, and less than 60% in 18% (3 of 17) of cases. Overall, 24 cases in this cohort showed basolateral staining pattern with weak to moderate intensity and 29% (7 of 24) were found to be HER2 amplified by FISH.

Of 26 cases with HER2 3⁺ results, 46% (12 of 26) showed only a basolateral staining pattern, 23% (6 of 26) showed only a complete staining pattern, and 31% (8 of 26) showed mixed complete and basolateral staining patterns (5 predominantly basolateral and 3 predominantly complete), with all showing a strong intensity. Therefore, 65% (17 of 26) showed some component of a basolateral staining pattern and 35% (9 of 26) some component of complete staining. A basolateral staining pattern and strong intensity were considered HER2 IHC positive (3⁺), even in the absence of complete membranous staining, after consensus agreement amongst 3 authors (Figure 2, A). Of the 12 cases with basolateral staining pattern and strong intensity defined as HER2 positive (3⁺), confirmatory HER2 FISH was performed on 1 case at the time of diagnosis per clinical request and showed amplification. For the purpose of the study, HER2 FISH was performed on 9 of the remaining 11 cases in which material was available and amplification was confirmed. HER2 FISH analysis was performed only on 1 IHC negative (1⁺) case in our cohort, which was not amplified. Representative examples of a HER2 IHC equivocal/FISH amplified case and a HER2 IHC positive/FISH amplified case, each with basolateral staining pattern, and weak to moderate or strong intensity, respectively, are displayed in Figure 3, A through F.

Upon reevaluation of the HER2 IHC slides in this cohort, we changed the HER2 IHC result for 16 cases. Five of 16 cases were originally classified as HER2 IHC 1⁺ and we changed them to an IHC score of equivocal (2⁺). FISH was performed on all of those 5 cases at the original time of diagnosis specifically owing to the presence of MP morphology. The FISH results on these 5 cases were as follows: group 1, n = 1; group 3, n = 1; group 5, n = 2; test failure, n = 1. Eleven of 16 cases were originally classified as HER2 IHC 0 and we changed them to an IHC score of 1⁺ (no FISH was performed on any of these 11 cases).

The 2018 updated ASCO/CAP HER2 guideline was available online May 30, 2018.¹⁷  Of the 5 cases we reclassified as HER2 IHC equivocal (2⁺) for the purpose of this study, 3 cases (1 group 1, 1 group 5, and 1 test failure) were originally reported before the updated guideline and 2 (1 group 3 and 1 group 5) after the updated guideline.

HER2 FISH was performed on 78 cases, with amplification in 36% (28 of 78) (Table 4). Of those cases, 93% (26 of 28) were classified as group 1 (15 IHC equivocal, 11 IHC positive) and 7% (2 of 28) as group 3 (2 IHC equivocal). Of the cases that were not HER2 amplified, 66% (33 of 50) were classified as group 5 (32 IHC equivocal, 1 IHC negative), 30% (15 of 50) were classified as group 4 (15 IHC equivocal), and 4% (2 of 50) had a test failure (2 IHC equivocal).

Of the 41 HER2 IHC equivocal 2⁺ cases showing basolateral membranous staining pattern with weak to moderate, and moderate intensity (Table 3), 27% (11 of 41) were classified as group 1; 2% (1 of 41) as group 3; 22% (9 of 41) as group 4; and 49% (20 of 41) as group 5 (Table 5). Of the 15 HER2 IHC equivocal 2⁺ cases showing complete membranous staining pattern with weak to moderate, and moderate intensity (Table 3), 20% (3 of 15) were classified as group 1; 6.5% (1 of 15) as group 3; 20% (3 of 15) as group 4; 47% (7 of 15) as group 5; and 6.5% (1 of 15) as test failure (Table 5).

At a median follow-up time of 30 months (range, 3–160) for the 43 patients with HER2-positive disease, all were alive. Seven patients experienced a recurrence (6 with local recurrence in breast or chest wall, 1 with distant metastasis in the brain), with a median time to recurrence of 50 months (range, 17–133). Of these 7 patients, 6 were ER⁺ and 1 was ER⁻. All patients with recurrence received HER2-targeted therapy. Of note, the 11 patients from the 12 cases with basolateral and strong intensity staining had a median follow-up time of 35 months (range, 3–89) and all were alive. One patient experienced a local recurrence in the chest wall at 51 months.

At a median follow-up of 29 months (range, 1–402) for the 138 patients with HER2-negative disease, 1 patient had died of disease. Fourteen patients experienced a recurrence (11 with local recurrence in breast, chest wall, or axilla; 3 with distant metastases in lymph nodes or lung) with a median time to recurrence of 33 months (range, 2–197). Of these 14 patients, 12 were ER⁺ and 2 were ER⁻.

IMPC is a rare and distinct histologic subtype of IBC considered to have an aggressive behavior with frequent lymphovascular invasion and axillary lymph node involvement when compared to IDC-NOS.^1,6,20–22  Thus, recognition of this pattern is of clinical importance. However, whether the overall prognosis of IMPC is worse than IDC-NOS remains controversial. Recent meta-analysis regrouping 1888 patients with IMPC from 14 studies showed that even though patients have a higher rate of local recurrence than with IDC-NOS, the overall survival, disease-specific survival, or distant metastasis–free survival were not significantly different.²³  Surprisingly, some studies reported a better outcome for IMPC.^24,25  An explanation for this may be that IMPC is often hormone receptor positive, a better prognosis when comparing to triple-negative breast cancer. Many of these studies did not compare the impact of HER2 status because they were conducted before HER2 testing was routinely available in clinical practice. Recently, Lewis et al²⁶  compared the outcome of IMPC with known hormone receptor and HER2 status and found that similar to IDC-NOS, triple-negative IMPCs have a worse outcome. They did not identify any outcome difference between HER2-positive status and hormone receptor–positive, HER2-negative IMPC,²⁶  and this is possibly due to successful treatment with HER2-targeted therapy.

HER2 is a predictive and prognostic biomarker in IBC, and accurate assessment of the HER2 status is needed to guide treatment.¹⁷  HER2-targeted therapies have been approved for clinical use and include HER2-directed monoclonal antibodies (ie, trastuzumab and pertuzumab) and small-molecule HER2 kinase inhibitors (ie, lapatinib). To assess the eligibility of patients for these treatments, the 2018 ASCO/CAP HER2 guideline for breast carcinoma¹⁷  describes HER2 IHC positive (3⁺) as circumferential (complete) and strong staining of the cell membrane in more than 10% of tumor cells. Incomplete or basolateral and weak staining in more than 10% of tumor cells is considered HER2 IHC negative (1⁺) and no staining or membrane staining that is incomplete or barely perceptible in less than or equal to 10% of tumor cells is HER2 IHC negative (0). In contrast, weak to moderate complete membranous staining in more than 10% of tumor cells is classified as HER2 IHC equivocal (2⁺), and reflex testing by in situ hybridization (ISH) is required to determine the HER2 amplification status.¹⁷  The distinct morphology of IMPC prompted modifications of the guideline for interpretation of HER2 IHC staining in 2013, as IMPC was considered to have an unusual staining pattern showing moderate to intense but incomplete (basolateral) staining, which can be found to be HER2 amplified and should be considered HER2 IHC equivocal (2⁺).¹⁵ 

In our cohort, HER2 IHC equivocal (1⁺ to 2⁺ or 2⁺) cases were identified in 35% of IMPC cases and FISH was performed for all of them. FISH showed amplification in 26% of HER2 IHC equivocal cases. Of those cases, 76% showed either entirely or a predominantly basolateral staining pattern and staining intensity was weak to moderate in 12% of cases, moderate in 82%, and moderate to strong in 6%. Vingiani et al⁶  first reported this discordance between HER2 IHC equivocal cases of IMPC and HER2 FISH. The authors tested 22 IMPC cases considered to be HER2 IHC 1⁺ according to the 2007 ASCO/CAP HER2 guideline²⁷  and found that 1 case was HER2 amplified by FISH.⁶  This positive case had faint to moderate basolateral HER2 IHC staining. In another study conducted by Yang et al,¹⁹  a total of 52 IMPCs were classified as HER2 IHC negative (0) in 15 cases, negative (1⁺) in 25 cases, and equivocal (2⁺) in 12 cases, according to the 2007 ASCO/CAP HER2 guideline.²⁷  All were tested by HER2 FISH. One HER2 IHC negative (1⁺) case showed HER2 amplification. This case had basolateral staining with moderate intensity.¹⁹  Stewart et al¹⁸  also tested a total of 45 IMPCs by FISH and 21 cases were HER2 amplified. According to the 2013 ASCO/CAP HER2 guideline¹⁶  and irrespective of the IMPC morphology, the 21 HER2-amplified cases were scored as HER2 IHC negative (1⁺) in 9 cases, HER2 equivocal (2⁺) in 11 cases, and HER2 positive (3⁺) in 1 case.¹⁸  The equivocal (2⁺) cases showed a HER2 IHC staining pattern that was nonuniform, with deposit along the cell membrane at the apical surface, lateral surface, or at the periphery of the morula with weak to moderate intensity. Staining pattern appeared to be more incomplete and weak in the 9 HER2 IHC negative (1⁺) cases and it is questionable if those patients would have benefited from HER2-targeted therapy.^18,28,29  More recently, Zhou et al³⁰  tested 147 IMPC tumors and identified 117 as having a basolateral pattern with moderate intensity, classified as HER IHC equivocal (2⁺); of those cases, 32% were HER2 amplified by FISH. In our cohort, 24 cases showed basolateral staining pattern with weak to moderate intensity and 29% (7 of 24) were found to be HER2 amplified by FISH. The proportion of HER2 IHC equivocal (2⁺) cases amplified by HER2 FISH in our cohort is similar to that reported overall in breast cancer.³¹  Taken together, our results in addition to prior studies provide evidence for refining the definition of HER2 IHC equivocal staining in IMPC to basolateral membranous staining with weak to moderate intensity in more than 10% of tumor cells.

In our cohort, HER2 IHC was positive (3⁺) in 14% of cases (26 of 187), of which 46% (12) showed basolateral staining pattern, 23% (6) complete membranous staining pattern, and 31% (8) mixed complete and basolateral (5 predominantly basolateral and 3 predominantly complete), all showing a strong intensity. Of the mixed cases, 5 showed predominantly basolateral staining and overall 65% (17 of 26) of cases demonstrated predominantly basolateral staining. Of the 12 cases showing only basolateral pattern staining and strong intensity, 83% (1 case at time of diagnosis and 9 cases for the purpose of this study [10 of 12]) were tested by HER2 FISH and all were amplified. In the current 2018 ASCO/CAP HER2 guideline,¹⁷  all those cases are considered HER2 IHC equivocal (2⁺) and ISH should be performed to confirm HER2 gene amplification. We observed that interpretation of HER2 IHC with a basolateral membranous staining pattern with strong intensity can be very subjective. In some cases, pathologists may interpret the membranous staining pattern inside of the morula-like clusters as complete, which can explain why HER2 FISH was performed on only 1 case at time of diagnosis. In 2004, Kuroda et al¹⁰  aimed to characterize HER2 IHC in IMPC. They considered HER2 positive (3⁺) cases with moderate or strong intensity. In figure 3 of their article, they showed an example of a case considered HER2 IHC positive (3⁺) with basolateral staining and strong intensity. They reported HER2 IHC positivity in 25.9% of cases but no HER2 FISH analysis confirmation was performed at that time. Yang et al¹⁹  reported 12 cases with basolateral and strong intensity and all were HER2 amplified by FISH. Strong IHC staining of HER2 was present at the cell-cell membrane or basolateral membrane of the MP structure, but not in the cytoplasmic membrane facing the stroma. The authors suggested that interpretation of HER2 IHC staining should be based on membrane staining intensity, but not the completeness of the membrane staining in IMPC. Zhou et al³⁰  showed that all 30 IMPC tumors presenting with HER2 IHC basolateral staining and strong intensity were HER2 amplified by FISH. They proposed that IMPC with intense and basolateral staining pattern be classified as HER2 IHC positive (3⁺), even if the staining is incomplete, avoiding unnecessary reflex ISH testing. Our results are concordant with their findings, as all cases with basolateral staining and strong intensity in our cohort, and with sufficient material for FISH (10 of 12), showed amplification. Our suggested interpretation can affect the clinical judgment and practice at some point. Two cases were called HER2 IHC 1⁺ at time of diagnosis and were reclassified as HER2 equivocal 2⁺ following our new suggestion for the purpose of this study. FISH was performed on those 2 cases at time of diagnosis because of the specific MP morphology and both were found to be amplified by FISH. Based on our modified IHC interpretation guideline, clinical management would have changed for 1 of 16 reclassified cases, based on the IHC score change from 1⁺ to 2⁺—the 1 case categorized as group 3 would be treated as amplified with an IHC score of 2⁺; however, it would be categorized as nonamplified with a concurrent IHC score of 1⁺. The group 1 case would have been treated as HER2 amplified regardless of the IHC score.

HER2 amplification is also capable of driving malignant transformation and tumor growth in gastric carcinoma. The guideline for HER2 testing and interpretation has been validated for gastric cancer as well.^32,33  Basolateral HER2 IHC staining is more common in gastric carcinoma than breast carcinoma owing to the higher frequency of glandular formation in gastric tissue.³³  Consequently, a HER2 IHC positive (3⁺) result is considered when moderate to strong complete or basolateral membranous reactivity is observed.³²  Thus, the distinct morphology of IMPC with basolateral staining with strong intensity is similar to the staining pattern considered positive in gastric carcinoma.

Ultrastructural studies of IMPC show the external cellular surface covered by microvilli, which are directed toward the surrounding stroma.^1,12  Immunohistochemical studies of IMPC reported that MUC1/EMA shows positivity only at the apical surface membrane, facing the stromal space, which is characteristic of the inside-out growth pattern.^8,14  E-cadherin shows a basolateral pattern with absence of staining at the apical pole facing the stroma.^14,34  This pattern of staining is similar to that seen by HER2 IHC, reflecting how the distinct morphology of IMPC has an impact on IHC staining.

There are a few limitations to this study. This is a retrospective analysis and HER2 FISH analysis was not available for all cases in the cohort. In addition, the short follow-up period is a limitation as the cohort is fairly recent and more time would be necessary to determine the prognostic significance of the aforementioned suggested IHC modifications to the guideline.

In conclusion, our results show that staining pattern and intensity are key features in the assessment of HER2 IHC in IMPC. The most frequent staining pattern in this large cohort was basolateral, seen in 49% (91 of 187) of cases, including 65% (17 of 26) of HER2 IHC positive (3⁺) cases and 76% (13 of 17) of HER2 IHC equivocal/FISH amplified cases. All HER2 IHC positive (3⁺) cases showed strong intensity, including cases with basolateral and complete staining pattern. If a basolateral pattern and weak to moderate staining is observed in IMPC, HER2 IHC should be reported as equivocal (2⁺) and alternative testing should be performed to confirm the HER2 status, as 29% (7 of 24) of equivocal cases with this staining pattern in our cohort were HER2 amplified by FISH. Overall, the distinct morphology of IMPC has an impact on HER2 IHC staining. Updates to the guideline for the equivocal and positive HER2 IHC categories for IMPC are needed to ensure that patients with this tumor type receive accurate HER2 classification and targeted therapy when indicated.