Journal of the National Cancer Institute Advance Access originally published online on January 29, 2008
JNCI Journal of the National Cancer Institute 2008 100(3):218-221; doi:10.1093/jnci/djm270
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author 2008. Published by Oxford University Press.
Comparison of Estrogen Receptor Results From Pathology Reports With Results From Central Laboratory Testing
Affiliations of authors: Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA (LCC, JDM); Channing Laboratory, Department of Medicine, Brigham and Women’s Hospital, Boston, MA (HJB, RMT); Department of Epidemiology, Harvard School of Public Health, Boston, MA (HJB, RMT)
Correspondence to: Laura C. Collins, MD, Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Avenue, Boston, MA 02215 (e-mail: lcollins{at}bidmc.harvard.edu).
 |
ABSTRACT |
|---|
 Top Abstract Context and CaveatsFundingReferencesNotes |
|---|
We compared estrogen receptor (ER) assay results abstracted from pathology reports with ER results determined on the same specimens by a central laboratory with an immunohistochemical assay. Paraffin sections were cut from tissue microarrays containing 3093 breast cancer specimens from women enrolled in the Nurses’ Health Study, 1851 of which had both pathology reports and tissue available for central laboratory testing. All sections were immunostained for ER at the same time. The original assays were biochemical for 1512 (81.7%) of the 1851 specimens, immunohistochemical for 336 (18.2%), and immunofluorescent for three (0.2%). ER results from pathology reports and repeat central laboratory testing were in agreement for 87.3% of specimens (1615 of the 1851 specimens; kappa statistic = 0.64, P < .001). When the comparison was restricted to the specimens for which the ER assays were originally performed by immunohistochemistry, the agreement rate increased to 92.3% of specimens (310 of the 336 specimens; kappa statistic = 0.78, P < .001). Thus, ER assay results from pathology reports appear to be a reasonable alternative to central laboratory ER testing for large, population-based studies of patients with breast cancer.
Prior knowledge Breast cancer clinical trials use estrogen receptor (ER) results from pathology reports instead of assaying the specimens again in a central laboratory, but the reliability of these data is unclear. Study design Results of ER assays abstracted from pathology reports were directly compared with immunohistochemical ER assay results on the same specimens in a tissue microarray from a central laboratory. Contribution ER results from pathology reports and central laboratory testing were in agreement for 87.3% of specimens and for 92.3% of specimens when the results were restricted to specimens originally tested with an immunohistochemical test. Implications ER assay results from pathology reports appear to be a reasonable alternative to central laboratory ER testing for large clinical trials of breast cancer patients. Limitations Tissue blocks from all cancers could not be obtained, and so only 52% of cancers were represented in the tissue microarray for testing by the central laboratory.
|
Epidemiologic studies of patients with breast cancer commonly use information abstracted from pathology reports in lieu of central pathology review (1) because of the practical limitations of performing central pathology review on hundreds to thousands of specimens and the economic and logistical constraints inherent in obtaining breast cancer slides and paraffin blocks from decades past. The reliability of the information contained in breast cancer pathology reports used for this purpose is of concern because the original misclassification of pathologic features could have a substantial impact on study results. In particular, the potential impact on epidemiologic studies of using unverified results for estrogen receptor (ER) assays performed at many different institutions with a variety of methods over a long period of time has not been previously evaluated.
To address this issue, we compared the results of ER assays performed in a central laboratory with results of ER assays abstracted from the corresponding pathology reports. The central laboratory used tissue microarrays that were constructed from 3093 paraffin blocks containing tumor specimens from women who reported a breast cancer diagnosis after completing the baseline questionnaire in June 1976 through the 1996 questionnaire cycle (diagnosis dates of between June 1976 and June 1997) among women enrolled in the Nurses’ Health Study. After taking into account age and year of diagnosis, participants whose tumors were included in the tissue microarrays were very similar to those for whom we were unable to obtain tissue blocks (Table 1). Details of these cohorts have been described previously (2). Three cores (0.6 mm in diameter) were obtained from a representative paraffin block from each patient and embedded in recipient paraffin blocks by use of a manual tissue arrayer (Beecher Instruments, Sun Prairie, WI). None of the donor blocks were core needle biopsy specimens. After heat-induced epitope retrieval, 5-µm sections cut from the tissue microarray blocks were immunostained for ER. The immunohistochemical staining procedure consisted of sequential application of the primary antibody against ER (rabbit monoclonal antibody, clone SP1, Dako, Carpinteria, CA; dilution 1:400), the EnvisionPlus Detection System (Dako), and diaminobenzidine as the chromagen. Positive and negative controls consisted of a known ER-positive breast cancer specimen and the omission of the primary antibody, respectively. All slides were stained in a single staining run.
|
Two of the study pathologists (L. C. Collins and J. D. Marotti) reviewed the immunostained sections under a microscope and estimated the percentage of tumor cells showing nuclear immunoreactivity for ER in every tissue core. Among the 1851 specimens, 1566 (84.6%) had three evaluable cores, 197 (10.6%) had two evaluable cores, and 88 (4.8%) had only one evaluable core. Cores that were not scored had technically suboptimal staining or lacked tumor cells. A specimen was scored as ER positive when more than 10% of the tumor cell nuclei in any of the three cores from that specimen had staining, as low positive when 1%–10% of nuclei in any of the three cores had staining, and as ER negative when no nuclear ER staining was observed in any core. Given that a consensus development panel of the US National Institutes of Health recommended that any ER staining in breast cancers should be sufficient to consider a tumor to be ER positive and the patient to be a suitable candidate for endocrine therapy (3), for this analysis the categories of positive and low positive were combined into a single "ER-positive" category. Results of the original ER assays were abstracted from pathology reports. Raw concordance data and kappa statistics (4) were used to assess the level of agreement between the ER assay results as determined from the tissue microarray sections and the ER status provided in the pathology reports. The kappa statistic was chosen for this analysis because neither original testing results nor central immunohistochemical testing on the tissue microarray slides can be considered to be the gold standard for determination of sensitivity and specificity; each technique has its limitations.
Of the women with breast cancer, ER results from both the pathology reports and the tissue microarrays were available for 1851; these 1851 patients constituted the population for this analysis. The original ER assays had been performed over a 20-year period in laboratories from 37 states in the United States and in various laboratories in Canada and Saudi Arabia. By report, the ER status was originally obtained with a biochemical assay for 1512 (81.7%) of the 1851 patients, an immunohistochemical assay in 336 (18.2%), and an immunofluorescent assay in three (0.2%).
The ER status as abstracted from the pathology reports and the ER status as determined on the tissue microarrays were in agreement for 1615 (87.3%) of the 1851 specimens (Table 2). The kappa statistic for this comparison (0.64; P < .001) indicated substantial agreement (4). In 99 (5.4%) of the 1851 specimens, the reported ER was positive but the tissue microarray ER was negative; in 137 (7.4%), the reported ER was negative and the tissue microarray ER positive. The original assay results and tissue microarray assay results were in agreement for 1302 (86.1%) of the 1512 specimens for which the original assay was biochemical and for 310 (92.3%) of the 336 specimens for which the original assay was immunohistochemical (Table 2). The kappa statistic for the comparison between original immunohistochemical results and tissue microarray results (0.78; P < .001) indicated almost perfect agreement (4). In this subset of specimens for which only immunohistochemical results were available and a tumor was scored as positive when there was any nuclear staining in tumor cell nuclei (Table 2), there were nine (2.7%) specimens for which the reported ER was positive but the tissue microarray ER was negative and 17 (5.1%) for which the reported ER was negative and the tissue microarray ER was positive. The level of concordance between the original assay results and tissue microarray results was lower when the analysis was restricted to specimens for which a tumor was scored as positive when more than 10% of the tumor cell nuclei showed ER staining (Table 2).
|
Information from all evaluable cores was used to determine ER status for each patient. However, we found, by use of receiver operator curve analyses, that the intrinsic information contained in ER measurements from the tissue microarray blocks was strongly associated with ER status even when only one core was interpretable. Little gain was realized from adding two or three core measurements (data not shown).
A limitation of the current study is that we were unable to obtain tissue blocks from all cancers. Our success was highly related to the time between diagnosis and the initiation of our tissue block collection because many hospitals destroy tissue blocks after 5–10 years. After taking into account the effect of age and year of diagnosis, the women for whom we were able to obtain tissue blocks were very similar to those for whom we were unable to obtain tissue blocks. Moreover, the frequency of ER status positivity was very similar to that of other populations, indicating that samples included in this study were representative of the underlying population. Additionally, because lesions had to be large enough to obtain multiple cores, it is possible that we may have biased our sample toward larger lesions. Again, the frequency of finding ER-positive tumors in our study indicates that the women included in our study were representative of patients with breast cancer in the US population.
The practice of using information abstracted from pathology reports for large population-based studies has many logistical and economic advantages over central pathology review. However, these advantages need to be balanced against the potential for misclassification of diagnoses and pathology details when using information abstracted from pathology reports, particularly those collected from many institutions over a long period of time during which diagnostic criteria, reporting practices, and test methodologies have changed or evolved. In fact, previous studies have indicated that there is considerable variation in the content and the quality of the information available in breast cancer pathology reports (5,6). Such clinically important elements as margin status, lymphovascular space invasion, and the presence or absence of ductal carcinoma in situ have not always been well documented. Furthermore, interlaboratory and intralaboratory variation in ER assay methodology and reporting of results have varied considerably since the ER test was first introduced into clinical practice in the 1970s (7); this variation raises concerns about the validity of using ER results from pathology reports in epidemiologic studies of patients with breast cancer.
In our study, the original ER assays had been performed in numerous laboratories across the United States, Canada, and Saudi Arabia by use of a wide range of assays over a 20-year period. Nonetheless, we found a very high level of concordance between ER results abstracted from pathology reports and the results obtained by repeating all ER analyses at the same time in a central laboratory that used one immunostaining technique on tissue microarray sections. Our findings, therefore, indicate that using ER assay results from pathology reports is a reasonable alternative to using central laboratory testing to obtain ER results in large, population-based studies of women with breast cancer.
|
Funding |
|---|
|
TopAbstractContext and CaveatsFundingReferencesNotes |
|---|
GlaxoSmithKline (WE234 EPI40307); National Cancer Institute, National Institutes of Health, Department of Health and Human Services (Public Health Service Grants CA087969 and SPORE in Breast Cancer CA089393).
|
NOTES |
|---|
|
TopAbstractContext and CaveatsFundingReferencesNotes |
|---|
The authors thank Drs Graham A. Colditz and Stuart J. Schnitt for their advice and support in the preparation of the manuscript.
The authors had full responsibility in the design of the study, the collection of the data, the analysis and interpretation of the data, the decision to submit the manuscript for publication, and the writing of the manuscript.
|
REFERENCES |
|---|
|
TopAbstractContext and CaveatsFundingReferencesNotes |
|---|
1. Colditz GA, Rosner BA, Chen WY, Holmes MD, Hankinson SE. Risk factors for breast cancer according to estrogen and progesterone receptor status. J Natl Cancer Inst. (2004) 96:218–228.
2. Colditz GA, Hankinson SE. The Nurses’ Health Study: lifestyle and health among women. Nat Rev Cancer (2005) 5:388–396.[CrossRef][Web of Science][Medline]
3. Eifel P, Axelson JA, Costa J, et al. National Institutes of Health Consensus Development Conference Statement: adjuvant therapy for breast cancer, November 1–3, 2000. J Natl Cancer Inst. (2001) 93:979–989.
4. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics (1977) 33:159–174.[CrossRef][Web of Science][Medline]
5. Apple SK. Variability in gross and microscopic pathology reporting in excisional biopsies of breast cancer tissue. Breast J. (2006) 12:145–149.[CrossRef][Web of Science][Medline]
6. Wilkinson NW, Shahryarinejad A, Winston JS, Watroba N, Edge SB. Concordance with breast cancer pathology reporting practice guidelines. J Am Coll Surg. (2003) 196:38–43.[CrossRef][Web of Science][Medline]
7. Schnitt SJ. Estrogen receptor testing of breast cancer in current clinical practice: what's the question? J Clin Oncol. (2006) 24:1797–1799.
Manuscript received August 3, 2007; revised November 13, 2007; accepted November 16, 2007.
Related Article in JNCI
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2008 100: 157.
This article has been cited by other articles:
|
|
 |
|
  J. M. Bueno-de-Mesquita, D. S. A. Nuyten, J. Wesseling, H. van Tinteren, S. C. Linn, and M. J. van de Vijver The impact of inter-observer variation in pathological assessment of node-negative breast cancer on clinical risk assessment and patient selection for adjuvant systemic treatment Ann. Onc., July 21, 2009; (2009) mdp273v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Milne, J. Benitez, H. Nevanlinna, T. Heikkinen, K. Aittomaki, C. Blomqvist, J. I. Arias, M. P. Zamora, B. Burwinkel, C. R. Bartram, et al. Risk of Estrogen Receptor-Positive and -Negative Breast Cancer and Single-Nucleotide Polymorphism 2q35-rs13387042 J Natl Cancer Inst, July 15, 2009; 101(14): 1012 - 1018. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Merglen, H. M. Verkooijen, G. Fioretta, I. Neyroud-Caspar, V. Vinh-Hung, G. Vlastos, P. O. Chappuis, M. Castiglione, E. Rapiti, and C. Bouchardy Hormonal therapy for oestrogen receptor-negative breast cancer is associated with higher disease-specific mortality Ann. Onc., May 1, 2009; 20(5): 857 - 861. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Wilson, L. A. Mucci, E. Cho, D. J. Hunter, W. Y. Chen, and W. C. Willett Dietary Acrylamide Intake and Risk of Premenopausal Breast Cancer Am. J. Epidemiol., April 15, 2009; 169(8): 954 - 961. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||