OUP user menu

A Comparison of Estrogen Receptor SP1 and 1D5 Monoclonal Antibodies in Routine Clinical Use Reveals Similar Staining Results

Jane E. Brock MBBS, PhD, Jason L. Hornick MD, PhD, Andrea L. Richardson MD, PhD, Deborah A. Dillon MD, Susan C. Lester MD, PhD
DOI: http://dx.doi.org/10.1309/AJCPSKFWOLPPMEU9 396-401 First published online: 1 September 2009

Abstract

Clinical therapies for breast cancer are guided by estrogen receptor (ER) status determined by immunohistochemical analysis. A previous retrospective study comparing the recently generated rabbit SP1 monoclonal antibody (MAb) with the conventionally used mouse 1D5 MAb reported that 8% of breast carcinomas were SP1+/1D5− (correlating with good outcomes), and 2% were SP1−/1D5+ (correlating with poorer outcomes). This study on mostly previously frozen tissue implied that 1D5 fails to identify some women who may benefit from endocrine therapy. The current prospective study compared SP1 and 1D5 immunostaining on routinely processed consecutive cases of breast carcinoma. ER was classified using the same positive threshold used in the prior study (<1% negative; ≥1% positive).

Of 508 carcinomas, 2 were SP1+/1D5–, and none were SP1–/1D5+. Although SP1 is our preferred antibody, with more intense nuclear staining, both MAbs give similar results in tissue from routine clinical samples with discrepant results in fewer than 0.5% of cases.

Key Words:
  • Estrogen receptor
  • SP1
  • 1D5
  • Monoclonal antibodies
  • Breast carcinoma

Immunohistochemical evaluation of estrogen receptor (ER) expression in breast carcinomas replaced ligand binding assays by the late 1990s. This followed the demonstration that it was at least equivalent at predicting response to adjuvant endocrine therapy and, in addition, allowed determination of ER status in small or sparsely cellular carcinomas.1 The early validations of immunohistochemical techniques using disease-free survival curves in patients receiving adjuvant endocrine therapy established an ER+ threshold at as few as 1% or fewer weakly staining cells or a score of 3 using the Allred score (a combination of total proportion of positively staining tumor cells on a 0–5 scale and an intensity score of 0–3).1,2 The 2 antibodies used in validation studies were mouse monoclonal antibodies (MMabs), MMab 6F11 and MMab 1D5.1,3,4 In the last few years, rabbit monoclonal antibodies (RabMabs) have become increasingly popular because they typically have higher affinities for the desired epitope, which translates into cleaner stains with more intense positive staining and lower background staining.5 The RabMab SP1 is one such antibody. It has an 8-fold higher affinity than that of MMab 1D5 for ER.6

A recent study comparing RabMab SP1 with MMab 1D5 in more than 4,000 invasive breast carcinomas found that 8% of cases were SP1+ and 1D5–.7 When these same cases (337 of 4,105) were compared with results of ligand binding assays, 92% were positive (>10.0 fmol/mg), and this correlated with disease-free survival curves and better outcomes. In turn, 1D5 identified 77 (1.9%) of 4,105 cases that were negative by SP1, and these cases were associated with worse outcomes. The implication of this study is that SP1 identifies more women who might benefit from endocrine therapy than does 1D5. However, this study used different techniques and analysis from those of routine clinical practice because all tissue tested had been previously frozen, and the majority of immunostains were performed and read on microarrays and not whole sections. The aim of this prospective study was to compare RabMab SP1 and MMab 1D5 in consecutive cases of breast carcinoma analyzed for routine clinical care during a 10-month period at Brigham and Women’s Hospital (BWH), Boston, MA, between June 2007 and March 2008. Additional information was collected that could possibly influence the results of the immunohistochemical study.

Materials and Methods

All BWH cases were received fresh and then formalin fixed and paraffin embedded. From the gross description of consult cases, it was noted whether these cases were received fresh and then formalin fixed or whether they were frozen before formalin fixation and paraffin embedding.

Four-micrometer-thick sections were immunostained according to manufacturer recommendations using the EnVision+ System-HRP (DAKO, Carpinteria, CA). In detail, slides were baked at 37°C overnight, then deparaffinized and rehydrated (100% xylene 4 times for 3 minutes each, 100% ethanol 4 times for 3 minutes each, and running water for 5 minutes). Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 10 minutes and washed under running water for 5 minutes. For heat-induced epitope retrieval, slides were placed in 10 mmol/L citrate buffer at a pH of 6.0 (Target Retrieval Solution, S1699, DAKO) and then pressure cooked (Biocare Medical, Concord, CA) at 122°C to between 14 and 17 psi with the cycle lasting, on average, 45 minutes and a cool-down period of approximately 20 minutes. Immunostains were performed on an automated instrument (DAKO Autostainer Plus, DAKO). A range of titers was tested for both antibodies, and titers were calibrated using internal positive control staining of normal breast epithelium.

Primary antibodies ER-1D5 (dilution 1:100; DAKO), ER-SP1 (dilution 1:200; Lab Vision, Fremont, CA), and progesterone receptor (PR; PgR636, dilution 1:200; DAKO) were incubated for 40 minutes at room temperature. A DAKO polymer secondary antibody system was used (Envision Poly K4011 for the SP1 RabMab and Envision Mono K4007 for 1D5 and PR MMabs) and incubated for 30 minutes in a humid chamber at room temperature. Sections were developed using 3,3′-diaminobenzidine (Sigma Chemical, St Louis, MO) as substrate and counterstained with Mayer hematoxylin. External positive controls were also run. The studies comparing ER antibodies were performed on the same day for a given case. Slides were scored for ER positivity by 2 pathologists (J.E.B., A.L.R., D.A.D., and S.C.L.) using the following 3 categories for ER and PR: less than 1%, negative; 1% to 10%, positive-low; and more than 10%, positive. Tumor cell nuclear positivity of 1% or more was also the score considered positive in the study by Cheang et al.7

Results

We tested 508 consecutive cases of breast carcinoma that required immunohistochemical staining for routine clinical care. The cases comprised 409 BWH cases and 99 consult cases. The specimens were from core needle biopsies, excisions, mastectomies, and biopsies for metastases (including lymph nodes, lung, skin, bone, liver, and gastrointestinal tract). Overall, 72.8% (370/508) were primary invasive carcinomas, 19.7% (100/508) were ductal carcinoma in situ (DCIS), and 7.5% (38/508) were metastases. Table 1 shows a summary of all ER immunoprofile results for invasive carcinomas, DCIS, and metastases using SP1 and 1D5.

View this table:
Table 1

No Discrepant Cases Identified Among DCIS Cases Tested

Of the 100 DCIS cases, there were 92 BWH cases (61 core biopsies and 31 excisions) and 8 consult cases (6 core biopsies and 2 excisions). Of 100 cases tested, 85 (85.0%) were ER+ by 1D5 and SP1, with no discrepant cases present. RabMab SP1 was consistently noted as having a greater intensity of staining compared with MMab 1D5, and, thus, was easier to read, but there was no noted difference in the total number of cells staining.

Image 1

Two discrepant cases (0.4%) showed negative staining with estrogen receptor (ER)-1D5 (<1% of total tumor cells staining) and positive-low staining with ER-SP1 (≥1%–10% of total cells staining). A, Invasive carcinoma (H&E, ×400). B, Invasive carcinoma (ER-1D5 immunohistochemical stain, ×400). C, Invasive carcinoma (ER-SP1 immunohistochemical stain, ×400); magnified inset, positive nuclear staining in a single cell (ER-SP1 immunohistochemical stain, ×600). D, Liver metastasis (H&E, ×400).

E, Liver metastasis (ER-1D5 immunohistochemical stain, ×400). F, Liver metastasis (ER-SP1 immunohistochemical stain, ×400); magnified inset, positive nuclear staining in a single cell (ER-SP1 immunohistochemical stain, ×600).

One Significant Discrepant Case Identified Among Invasive Carcinoma Cases Tested

The 370 cases of invasive carcinoma included 291 BWH cases (167 core biopsies and 124 excisions) and 79 consult cases (20 core biopsies and 59 excisions). Of 370 cases, 274 (74.1%) were positive for 1D5 and SP1 and 95 (25.7%) were negative for 1D5 and SP1. One invasive carcinoma (0.3%) was SP1+/1D5–. No cases were classified as SP1–/1D5+. The SP1+/1D5– case was a consult case of a poorly differentiated invasive carcinoma showing very rare weakly positive nuclei with 1D5 (<1%) and 5% to 8% strongly positive nuclei with SP1 Image 1A, Image 1B, Image 1C, and Table 2. This case was PR– and HER2/neu+ (3+). An additional 3 cases (0.8%) comprising 1 BWH core biopsy and 2 excision consult cases had different total percentage staining of tumor nuclei between 1D5 and SP1, but each antibody was still classified in the same category. In brief, 2 cases showed rare cells positive on both stains, with more tumor nuclei staining with SP1 than with 1D5 but not enough to reach a threshold of 1%. Cytoplasmic staining with 1D5 was also present for 1 case. The third case showed positive-low staining for 1D5 and SP1, with more tumor nuclei staining strongly with SP1 (8%–10%) vs 1D5 (2%–4%).

View this table:
Table 2

In 13 ER– cases (13/95), marked cytoplasmic staining was present on the 1D5 stain that was never seen with the SP1 stain. The cases with cytoplasmic ER-1D5 staining were histologically notable for the presence of a brisk lymphocytic infiltrate, and all were HER2/neu–. The significance or relevance of these observations is not known. No differences between a positive-low and a positive score were noted between 1D5 and SP1. RabMab SP1 was consistently noted as having a greater intensity of staining compared with MMab 1D5, and, thus, it was much easier to read positive-low results and negative results.

One Significant Discrepant Case Identified Among Cases of Breast Carcinoma Metastases Tested

We tested 38 metastases, including 26 BWH cases and 12 consult cases. Of the 38 cases, 24 (63%) were ER+ by 1D5 and SP1 and 13 (34%) were ER– by 1D5 and SP1. One case (3%) was classified as SP1+/1D5– Image 1D, Image 1E, and Image 1F. No case was SP1–/1D5+. The SP1+/1D5– case was a liver metastasis in a patient with a 3-year history of a poorly differentiated invasive ductal carcinoma of the breast that had previously been reported in a primary excision and a chest wall recurrence as ER– (0%), PR– (0%), and HER2/neu– (0) at 1 outside hospital using the ER 6F11 antibody. In addition, the chest wall recurrence was stained at a second outside hospital where it was reported as ER+ (20%), PR+ (60%), and HER2/neu–. Neither set of immunostains was available for review at BWH, and the antibody used by the second institute is unknown. The liver metastasis was morphologically identical to the primary tumor on review at BWH, and on staining of the metastasis at BWH, it was found to be SP1 positive-low with 5% of nuclei staining strongly and rare cells (<1% of nuclei) staining strongly with 1D5 (Images 1D through 1F and Table 2). In addition, PR was positive-low (5%) and HER2/neu negative (0).

An additional 2 cases showed different total percentages of tumor cells staining but were still classified in the same group for each antibody. One of these had no positive nuclei with 1D5 and rare cells positive (<1%) with SP1. The other case was classified as positive-low with 1D5 and SP1 but was notable for having 2% to 3% of nuclei positive with 1D5 and 8% to 10% positive with SP1. Cytoplasmic staining for 1D5 was seen in 4 of 38 cases, including 1 of the 3 cases described. SP1 was consistently noted as having a greater intensity of staining compared with 1D5, and, thus, it was much easier to read positive-low results and negative results.

Discussion

In this prospective comparison of SP1 and 1D5, 2 breast carcinomas of 508 cases (<0.5%) showed significantly discrepant results as they were scored as “positive-low” with RabMab SP1 and “negative” with MMab 1D5 using a positive threshold of 1% or more total tumor nuclei staining.

Our results differ from those in the recent study by Cheang et al,7 who reported 8% of cases were SP1+/1D5– using the same threshold of 1% or more total tumor nuclei staining for positivity and an identical immunohistochemical technique including conventional antigen retrieval (written communication, A.M. Gown, MD, PhenoPath Laboratories, Seattle, WA, February 27, 2008).

Why might these 2 ER antibodies give different results? First, they recognize different epitopes. The SP1 antibody recognizes the C-terminus of human ERα protein and, by contrast, 1D5 recognizes the N-terminus. Many mutant forms of ERα messenger RNA are expressed in ER+ and ER– tumors in vivo, including exon deletions and duplications in all the functional domains.8 However, the in vivo clinical significance of these mutant forms of ER and their effect on hormone sensitivity have yet to be demonstrated. Thus, loss of the epitope in mutant ERα proteins, particularly in truncated forms of the ER protein lacking the C-terminus, is possible. But there is no evidence that C-terminus–directed ER antibodies are inferior to N-terminus–directed antibodies when comparing them with enzyme immunoassay results or ligand binding assays in tumors.7,9

A second possible reason for a different result is that the epitopes could be altered differentially by postexcision tissue processing: 10% phosphate-buffered formalin is a widely used standard fixative, but many fixatives have different or additional components, including metal salts and strong acids. For example, Zenker, Bouin, and decalcifying agents are known to impair preservation of epitopes for immunohistochemical analysis.10 All of the tissues we tested were received fresh and rapidly fixed in 10% buffered formalin. Similarly, cryopreservation also alters epitopes, and, although it is typically regarded as less epitope-destructive than many fixatives, proteins do degrade over time in cryopreserved tissues. Of note, almost all of the cases tested by Cheang et al7 had been previously frozen, and it is possible that the SP1 epitope is more robust than the 1D5 epitope in these preservation conditions. It is clear from our study that in routine clinical samples that have not been frozen, these 2 monoclonal antibodies give very similar results.

A third potential reason for discrepant results may be the impact of different antibody sensitivities when protein levels are very low. SP1 has an 8-fold higher affinity for its epitope than does 1D5, and it may simply detect more positive cells than does 1D5 when protein levels are extremely low. The evidence to date suggests that detection of very low ER protein levels in tumors should not be considered a false-positive result since the study by Cheang et al7 showed that SP1+/1D5– cases correlate with better outcomes than SP1–/1D5+ cases.

RabMab SP1 is now the preferred antibody at our institution for 2 reasons. First, it is easier to read positive-low results and negative results owing to its more intense nuclear staining and lower background staining. Second, the cytoplasmic staining seen with MMab 1D5 in ER– cases is never seen with RabMab SP1. Use of RabMab SP1 reduces the unlikely but real chance of error in reporting a case as ER+ that is, in fact, ER–. (We have observed this error in our consult cases.) Although SP1 is our preferred antibody, both monoclonal antibodies give similar results in tissue from routine clinical samples with discrepant results in fewer than 0.5% of cases.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
View Abstract