Article Text

Download PDFPDF

The effect of delay in fixation on HER2 expression in invasive carcinoma of the breast assessed with immunohistochemistry and in situ hybridisation
  1. Andrew H S Lee1,
  2. Heather P Key1,
  3. Jane A Bell2,
  4. Patrick Kumah1,
  5. Zsolt Hodi1,
  6. Ian O Ellis1
  1. 1Department of Histopathology, Nottingham University Hospitals, City Hospital Campus, Nottingham, UK
  2. 2Source Bioscience plc, Nottingham Business Park, Nottingham, UK
  1. Correspondence to Dr A H S Lee, Department of Histopathology, Nottingham University Hospitals, City Hospital Campus, Hucknall Road, Nottingham NG5 1PB, UK; andrew.lee{at}nuh.nhs.uk

Abstract

Aims Accurate assessment of HER2 status is essential for selection of patients for HER2-targeted treatment such as trastuzumab. This study investigated the hypothesis that delayed fixation impairs HER2 assessment.

Methods 9 carcinomas were received fresh, and samples were fixed immediately or put in fixative at time intervals up to 24 h. All carcinomas were scored as 3+ with immunohistochemistry in properly fixed tissue.

Results 2 of 9 carcinomas (95% CIs 6% to 56%) showed reduced immunohistochemical staining with delays in fixation of 1 and 8 h. One carcinoma showed low-level amplification with fluorescence in situ hybridisation (FISH) when properly fixed and was not amplified after delayed fixation. The other carcinomas were amplified at all time points.

Conclusions Delayed fixation impaired HER2 protein expression assessed using immunohistochemistry in 22% of 3+ carcinomas. HER2 amplification assessed using FISH may be less affected.

  • Breast
  • Immunohistochemistry
  • Fixation
  • In Situ Hybridisation
View Full Text

Statistics from Altmetric.com

Introduction

Accurate assessment of HER2 status of invasive carcinoma of the breast is vital for selection of patients for HER2-targeted treatment such as trastuzumab. The UK guidelines recommend using a combination of immunohistochemistry and in situ hybridisation to assess HER2 status, with in situ hybridisation performed if the immunohistochemistry result is equivocal.1 Delay in fixation of the specimen on which HER2 is to be assessed could potentially compromise this assessment.

Fixation is essential to preserve the tissue and ensure good quality cellular morphology necessary for diagnostic evaluation. Delay in fixation affects assessment of mitoses resulting in potential undergrading of mammary carcinomas.2 Delay in fixation as short as 2 or 3 h can result in reduction in immunohistochemical staining for oestrogen receptor.3–5 By contrast, such short delays in fixation appear to have less effect on HER2 immunohistochemisty.5 ,6 There are limited data on the effect of longer delays in fixation on HER2 expression, as occurs, for example, if a tumour in a surgical specimen is not incised.5 ,7 ,8

The aim of this study was to test the hypothesis that delay in fixation of more than 2 h impairs detection of HER2 protein overexpression and DNA amplification.

Materials and methods

All surgical specimens of carcinomas of the breast are routinely received fresh in our laboratory. The time from theatre to laboratory for most samples is 10–20 min and for almost all is less than 30 min. For this study, small samples about 8×3×3 mm were taken. For each tumour, one or two fragments were fixed immediately and the remainder were put in fixative at various time intervals up to 24 h. Usually two fragments of carcinoma were examined for each time point. All samples were fixed for a minimum of 10 h. All carcinomas were scored as 3+ in immediately fixed tissue. Five were fixed in 10% phosphate buffered formalin and four in 10% neutral buffered formalin.

Immunohistochemistry for HER2 was performed with three different antibodies. Staining with the HercepTest kit (Dako UK Ltd, Ely, UK) was performed with the Dako autostainer as previously described.9 Staining using the Leica Oracle HER2 staining kit (TA9145) was performed on Bond Max Automated staining machines (Leica Microsystems, Milton Keynes, UK). Antigen retrieval was performed using Epitope Retrieval solution 1 (a ready-to-use citrate-based pH6.0 solution). Staining was also performed using the Ventana Pathway Her2/neu (4B5) rabbit monoclonal primary antibody (790-2991) with Ventana UltraView Universal DAB Detection kit (760-500) on a Ventana BenchMark XT platform (Ventana Medical Systems Inc., Tucson, Arizona, USA).

All immunohistochemical sections were independently assessed by two of three experienced observers using the standard system described in the UK guidelines.1 In addition, the percentage of invasive carcinoma cells with complete membrane staining was estimated and the average of the two scores was calculated. Any discrepancies were resolved by discussion between all three observers.

All tumours had fluorescence in situ hybridisation (FISH) performed at the Source BioScience laboratory using the PathVysion HER2 DNA FISH assay (Abbott Laboratories) as described previously.9 Samples with a HER2/chromosome 17 centromere ratio of at least 2.0 were regarded as amplified.

The study was discussed with the chair of one of the Nottingham research ethics committees; the written opinion was that ethical approval was not necessary.

Results

The details of immunohistochemical staining at different times of delayed fixation are shown in table 1. Two carcinomas (95% CIs 6% to 56%) showed a clear reduction in staining intensity after delays in fixation of 1 h and 8 h (figure 1). Similar results were seen with three different antibody systems. The one carcinoma with discrepant results with different antibodies (tumour 9) had strong staining near the 30% threshold.

Table 1

Delay in fixation and the score and percentage of complete membrane staining with immunohistochemistry using three different antibody systems

Figure 1

Carcinoma that was (A) 3+ when properly fixed and (B) 0+ after 8 h delay in fixation. Carcinoma that (C) 3+ when properly fixed and (D) 3+ after 8 h delay in fixation, although staining was more patchy.

The details of FISH results at different times of delayed fixation are shown in table 2. One carcinoma showed low-level amplification when immediately fixed and was not amplified after delayed fixation. The other carcinomas showed amplification at all time points.

Table 2

Delay in fixation and the ratio of HER2 to chromosome 17 centromere signals with fluorescence in situ hybridisation

Discussion

Two of nine carcinomas in this study showed a clear reduction in staining for HER2 protein using immunohistochemistry after delayed fixation. This effect was seen with three different antibody systems. The two tumours with reduced staining were fixed in phosphate buffered formalin. However, we have seen the same effect in a carcinoma fixed in neutral buffered formalin. The carcinoma was 3+ in the core biopsy. The corresponding surgical specimen was not incised on the day of surgery and as a result fixation was poor and HER2 was negative. In our hospital, all surgical specimens for breast cancer are received fresh with most received within 20 min. This degree of delay does not appear to affect HER2 status. In a recent study, we found concordant HER2 status in 98% of core biopsies (which are well-fixed) and corresponding surgical specimens.10 Other studies also found that short delays in fixation have little or no effect on HER2 immunohistochemistry.5 ,6

There are limited data on the effect of longer delays in fixation on HER2 expression in breast cancer. Three studies each found no effect in a single 3+ carcinoma.5 ,7 ,8 One study found a reduction of staining was seen in some of the 2+ tumours after 4 h or more delayed fixation, but no details of the proportion affected is given.5

In the present study, delayed fixation appeared to have less effect on gene copy number assessed using FISH with only one carcinoma showing possible deterioration in the present study. A previous study suggested that FISH evaluation of gene copy number is compromised after 2 h delay in fixation, but no details of the effect on the scores were given.7

One approach to avoiding the problems of delayed fixation in surgical specimens is to assess HER2 on core biopsies. Two large studies have shown 98–99% agreement between HER2 assessed in core biopsies and surgical specimens using a combination of immunohistochemistry and FISH.10 ,11 However, in some cases, assessment can only be performed on the surgical specimen, for example, if there is only ductal carcinoma in situ in the core biopsy and invasive carcinoma is only present in the surgical specimen.

The results of this study reinforce the evidence of the importance of adequate fixation. Formalin penetrates slowly into tissues (about 1 mm/h),12 so to ensure good fixation of surgical specimens it is essential to incise the specimen as soon as possible. We receive all breast cancer resections fresh in the laboratory.

The present study is small and ideally a larger number of carcinomas with different fixatives, including 2+ carcinomas with good fixation, should be assessed. Also, the effect of more prolonged delays in fixation should be investigated.

In conclusion, this study shows that delayed fixation can adversely affect HER2 immunohistochemistry. Good fixation is also important for assessing morphology, histological grade, vascular invasion, and other immunohistochemistry such as for oestrogen receptor. The solution is for surgical specimens to be taken quickly to the pathology department for prompt incision of the tumour and immersion in fixative. Accurate assessment of HER2 is essential for selecting patients for HER2-targeted treatment, such as trastuzumab. An important part of the methodology of such assessment is adequate fixation.

Take home messages

  • Delayed fixation can reduce immunohistochemical staining for HER2 potentially compromising treatment decisions.

  • This reduced immunohistochemical staining is seen with different antibody systems.

  • Delayed fixation appears to have less effect on HER2 in situ hybridisation.

  • These results emphasise the importance of incision of mammary carcinomas in surgical specimens.

Acknowledgments

We thank Roche Products Ltd for a grant. The company had no influence on the design, conduct, analysis or writing up of this study.

References

View Abstract

Footnotes

  • Contributors All authors contributed to the design of the study, interpretation of the results, and drafting and approval of the paper. AHSL and PK collected the samples. HPK performed the immunohistochemistry. AHSL, ZH and IOE interpreted the immunohistochemistry. JAB performed the FISH, and AHSL is the guarantor.

  • Funding Roche Products Ltd.

  • Competing interests AHSL and IOE have received fees for lecturing from Roche. IOE has undertaken consultancy and advisory board work for Roche. IOE and ZH both work part time for Source Bioscience.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All the data are included in the paper.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.