Ki-67 is a biological tumor marker that reflects tumor

Evaluation of Ki-67 Index in Core Needle Biopsies and Matched Breast Cancer Surgical Specimens Soomin Ahn, MD; Junghye Lee, MD; Min-Sun Cho, MD, PhD; Sanghui Park, MD, PhD; Sun Hee Sung, MD, PhD Context. The Ki-67 index is strongly prognostic and is used as a surrogate marker to distinguish luminal A from luminal B breast cancer types. Objective. To investigate differences in Ki-67 index between core needle biopsy samples and matched surgical samples in breast cancer. Design. We included patients with invasive breast cancer who did not receive neoadjuvant therapy. A total of 89 pairs of core needle biopsies and surgical specimens were collected, and the Ki-67 index was assessed in hot spot areas using an image analyzer. We applied a 14% Ki- 67 index to define low versus high groups. Results. The Ki-67 index was significantly higher in core needle biopsies than in surgical specimens (P,.001), with a median absolute difference of 3.5%. When we applied 14% as a cutoff, 16 of 89 cases (18%) showed discrepancy. Thirteen cases showed a high Ki-67 index in core needle biopsies but a low Ki-67 index in surgical samples. There were 10 cases (11.2%) that showed discordant luminal A/B types between core needle biopsy and the matched surgical specimen. The reasons for the discordance were poor staining of MIB1 accompanied by fixation issues and intratumoral heterogeneity of the Ki-67 index. Conclusions. A significant difference in the Ki-67 index between core biopsy and surgical specimens was observed. Our findings indicate that it may be better to perform the Ki-67 assay on the core needle biopsy and the surgical specimen than on only one sample. (Arch Pathol Lab Med. 2018;142:364 368; doi: 10.5858/ arpa.2017-0014-oa) Ki-67 is a biological tumor marker that reflects tumor proliferation. In breast cancer, the Ki-67 index is strongly prognostic 1 and supplies additional useful information. For the purpose of prognostication and treatment decision making, breast tumors are grouped into surrogate intrinsic subtypes, defined by routine histology and immunohistochemistry (IHC) data. 2 As gene expression tests are not universally available in daily practice, IHC-based classification including the Ki-67 index is important to differentiate luminal A and B types. According to the 2013 and 2015 St Gallen guidelines, decisions on systemic adjuvant therapies of luminal cases with unclear indications are based on intrinsic subtype, determined by estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), and Ki-67 assessment. 2,3 Luminal A disease is characterized by a positive hormone receptor (HR), absence of HER2 amplification, and low Ki- 67 (,14%), and requires only endocrine therapy. On the other hand, in cases of luminal B/HER2 (HR þ, HER2, Ki- 67 14%), luminal B/HER2 þ (HR þ, HER2 þ ), HER2-enriched (HR, HER2 þ ), and triple-negative disease (HR, HER2 ), Accepted for publication May 30, 2017. Published as an Early Online Release November 16, 2017. From the Department of Pathology, Seoul National University Bundang Hospital, Seongnam, Gyeonggi-do, Korea (Dr Ahn); and the Department of Pathology, Ewha Womans University Medical Center, Seoul, Korea (Drs Lee, Cho, Park, and Sung). The authors have no relevant financial interest in the products or companies described in this article. Reprints: Sun Hee Sung, MD, PhD, Department of Pathology, Ewha Womans University Medical Center, 1071, Anyangcheon-ro, Yangcheon-gu, Seoul 07895, Korea (email: sunhsung@ewha.ac.kr). chemotherapy is considered. 4 In the neoadjuvant chemotherapy setting, the Ki-67 index may serve as a predictive marker of response to chemotherapy and also serves as the primary or secondary endpoint. 5 7 Therefore, accurate estimation of the Ki-67 index is very important for proper management of patients with breast cancer. Multiple factors might affect Ki-67 staining and its measurement. These include type of biopsy, time to fixation, type of fixative, time in fixative, storage type, and measurement method. Regarding tissue type, the International Ki-67 in Breast Cancer working group 8 commented that both core biopsies and whole sections from excision are acceptable. Some studies 9,10 have suggested that whole sections may give higher Ki-67 scores than core biopsies. On the other hand, fixation is better controlled for core biopsies, allowing safer antigen preservation for IHC. 3 Fixation issues can frequently cause differences in the appearance of stained nuclei. 8 More rapidly fixed core biopsies consistently showed well-circumscribed, uniformly staining nuclei, whereas nuclei in whole sections often showed areas of highly variable staining. This variability did not disrupt scoring by manual assessment. However, it can be difficult to deal with in digital image analysis procedures. 8 For ER, PR, and HER2, IHC is recommended in preoperative biopsy specimens. In cases of negativity of ER, PR, and HER2 in the biopsy specimen, it is advisable to retest for them in the surgical specimen because of putative tumor heterogeneity. 3 However, there is no established guideline for Ki-67 measurement. In this study, we measured the Ki-67 index with a digital image analyzer in core needle biopsies and matched surgical specimens of breast cancer. Further- 364 Arch Pathol Lab Med Vol 142, March 2018 Ki-67 Index in Breast Cancer Ahn et al

Table 1. Patient Characteristics (N ¼ 89) Ki-67 Index Low (,14%), No. (%) High (14%), No. (%) Total No. Cases P Breast surgery types Partial mastectomy 20 (76.9) 48 (76.2) 68.99 Total mastectomy 6 (23.1) 15 (23.8) 21 pt stage 1 21 (80.8) 27 (42.9) 48.001 a 2 3 (11.5) 33 (52.4) 36 3 2 (7.7) 3 (4.8) 5 pn stage 0 18 (69.2) 40 (63.5) 58.85 a 0þ 2 (7.7) 5 (7.9) 7 1 5 (19.2) 10 (15.9) 15 2 0 (0) 4 (6.3) 4 3 1 (3.9) 4 (6.3) 5 Pathology type Invasive ductal carcinoma 24 (92.3) 62 (98.4) 86.20 Others 2 (7.7) 1 (1.6) 3 Estrogen receptor Positive 24 (92.3) 36 (57.1) 60.001 Negative 2 (7.7) 27 (42.9) 29 Progesterone receptor Positive 23 (88.6) 32 (50.8) 55.001 Negative 3 (11.5) 31 (49.2) 34 HER2 Positive 0 (0) 14 (22.2) 14.008 Negative 26 (100) 49 (77.8) 75 a Analyzed by Fisher exact test. more, we evaluated the trend of Ki-67 index between 2 sample types and also analyzed discordant cases. MATERIALS AND METHODS Patient Population From March to July 2016, 147 patients underwent partial or total mastectomy at Ewha Womans Hospital (Seoul, Korea) for primary breast cancer. Preoperative core needle biopsy was performed for all patients using at least two 14-gauge cores with ultrasound guidance. Patients who were diagnosed with invasive tumor both in core needle biopsy and in surgical specimen were included. Patients who received neoadjuvant chemotherapy or any intervening therapy were excluded. After exclusion, a total of 89 pairs of core needle biopsies and surgical specimens were included in this study. These core needle biopsy samples included those of both inhouse patients (n ¼ 48) and outside-hospital patients (n ¼ 41). Inhouse core biopsy specimens were immediately immersed in 10% formalin. For the surgical excision specimens, the time to fixation was 0.2 to 3 hours, and the fixation time was 12 to 24 hours. This study was approved by the institutional review board of our institution, and the need for informed consent was waived. Immunohistochemistry Immunohistochemistry was performed on 5-lm slidesofforma- lin-fixed and paraffin-embedded archival tumor tissue. Antigen retrieval was performed in a micro-oven in citrate buffer ph 6 for 20 minutes. The Ki-67 antibody (clone MIB1, Dako, Glostrup, Denmark) was diluted 1:500, incubated for 25 minutes in a TechMate 500 plus (Dako), and visualized with diaminobenzidine. Immunohistochemistry for ER, PR, HER2, and Ki-67 was performed on core needle biopsy samples. Immunohistochemistry for ER, PR, and HER2 was selectively repeated on the surgical resection specimens; Ki-67 IHC was repeated on the surgical resection specimen for all cases. For HER2 2 þ cases, fluorescence in situ hybridization was performed using a PathVysion HER2 DNA Probe Kit (Vysis, Downers Grove, Illinois). The Allred score was used for ER and PR scoring, and scores from 3 to 8 were regarded as positive. 11 Hormone receptor positivity was defined by any ER or PR positivity. Image Analysis Ki-67 stained slides were captured digitally at a hot spot at 3200 magnification. One hot spot area was selected for each case, and the measured area was 0.27 mm 2. The Ki-67 labeling index was measured using digital image analysis software (Tissue Studio 64 Dual, version 3.5, Definiens, Munich, Germany). Image analysis was performed by one experienced pathologist (S.A.). The Definiens software is based on Definiens Cognition Network Technology and is object oriented, multi-scale, context driven, and knowledge based. 12 During the process of area selection, normal breast tissue and nontumor cells were manually excluded by the pathologist. 13 Despite area selection, some contamination with nontumor cells, including stromal and lymphoid cells, was inevitable. 13 In order to distinguish non carcinoma cell elements from carcinoma cell nuclei on the digitized image, nuclei with small areas (,32 lm 2 gross area, which was decided by a mean nuclear area of 50 infiltrating lymphocytes, 31.7 lm 2, and by mean nuclear area of 50 normal ductal cells, 31.1 lm 2 ) and spindle features (greater than 0.5 oval rate) were considered as lymphocytes, normal ductal cells, and stromal cell nuclei, respectively, and were excluded. 13 Statistical Analysis Statistical analysis was performed using SPSS version 18.0. Contingency tables and v 2 tests were used to correlate Ki-67 results with clinicopathologic variables including pt and N stages and ER and HER2 statuses. The Ki-67 indexes of core needle biopsies and surgical specimens were compared by Student t test. RESULTS Patient Demographics and Clinicopathologic Features Clinicopathologic characteristics of the 89 patients are summarized in Table 1. Twenty-one patients underwent radical mastectomy, and 68 patients underwent partial Arch Pathol Lab Med Vol 142, March 2018 Ki-67 Index in Breast Cancer Ahn et al 365

mastectomy with lymph node dissection or sentinel node biopsy. The median tumor size was 2.1 cm (range, 0.6 10 cm). Of 89 total patients, 24 (27%) had lymph node metastasis. The histologic type was invasive ductal carcinoma for 86 of 89 patients. The other 3 patients were diagnosed with metaplastic carcinoma, papillary carcinoma, and mucinous carcinoma, respectively. Estrogen receptor and PR were expressed in 60 (67.4%) and 55 cases (61.8%), respectively. The HER2 IHC was negative in 57 cases (64%), equivocal in 23 cases (25.8%), and positive in 9 cases (10.1%). Of 23 HER2 IHC equivocal cases, 5 showed HER2 amplification by fluorescence in situ hybridization. In IHC classification of the preoperative biopsy, the numbers of luminal A, luminal B/HER2 negative, luminal B/HER2 positive, HER2-positive, and triple-negative type cancers were 25 (28.1%), 28 (31.5%), 7 (7.9%), 7 (7.9%), and 22 (24.7%), respectively. Ki-67 index was associated with pt stage (P ¼.001), ER status (P ¼.001), PR status (P ¼.001), and HER2 status (P ¼.008) (Table 1). Ki-67 Index and IHC Subtype in Core Needle Biopsy and Matched Whole Resection Specimens The Ki-67 index was significantly higher in the core needle biopsy compared with the surgical specimen (P,.001), with a median absolute difference of 3.5% (Figure 1). The Ki-67 index in the core needle biopsy ranged from 0.30 to 78 (median, 21.0; mean, 25.58), whereas it ranged from 0.20 to 67 (median, 17; mean, 22.1) in the surgical specimen. When we applied 14% as a cutoff, 16 of 89 cases (18%) showed discrepancy (Table 2). Thirteen cases showed a high (14%) Ki-67 index in the core needle biopsy but a low (,14%) Ki-67 index in surgical samples. Twelve of these 13 cases were positive for ER, 3 of which were HER2 positive. In terms of IHC subtype, 9 discordant cases were classified as luminal B/HER2 in the core needle biopsy and as luminal A in the surgical specimen. On the other hand, 3 cases showed a low Ki-67 index in the core needle biopsy but a high Ki-67 index in surgical specimens. Of these cases, one case was triplenegative type, and another case was luminal B/HER2 þ type. The other case was classified as luminal A type in thecoreneedlebiopsyandasluminalb/her2 type in the surgical specimen. There were a total of 10 cases (11.2%) that showed discordant intrinsic subtypes between the core needle biopsy and its matched surgical specimen. Analysis of Discordant Cases For 16 discordant cases, the images used for digital image analysis and Ki-67 slides were reviewed again by 2 pathologists (S.A. and S.S.). Of 13 cases showing a higher Ki-67 index in the core needle biopsy (Figure 2, A), 7 revealed a poor staining pattern in surgical specimens (Figure 2, B). Six cases showed a good quality of Ki-67 staining in both the core needle biopsy and the surgical specimen. For these cases, a heterogeneous distribution of Ki-67 positive tumor cells was noted (Figure 3). There were 3 cases showing a higher Ki-67 index in the surgical specimen. Two cases showed a heterogeneous staining pattern. One patient underwent core needle biopsy (Figure 4, A) at a primary health care center, and the fixation quality was poor in the core needle biopsy specimen (Figure 4, B). DISCUSSION Accurate Ki-67 measurement is important for proper treatment and management of breast cancer patients in clinical practice. In particular, the Ki-67 index is used as a surrogate marker of luminal A and luminal B types in ERpositive and HER2-negative breast cancers, and 14% is currently used as a cutoff. 2 We analyzed the concordance of Ki-67 index in core needle biopsies and matched surgical specimens of breast cancer. All Ki-67 analysis was performed by digital image analyzer in a hot spot area. In our study, the core biopsy revealed a higher Ki-67 index compared with the surgical specimen. Previous studies investigating concordance between core needle biopsy and surgical specimens showed various results. 9,10,14 18 Although some studies showed a higher Ki-67 index in core biopsy specimens, 14,15 other studies showed a higher Ki-67 index in surgical specimens. 9,10 Most studies reported a moderate rate of concordance between the 2 tissue types. However, the discordance rates of Ki-67 index were higher than those of ER and HER2 stains. In our study investigating 89 pairs of core biopsy and surgical specimens, there were 13 cases with a high (14%) Ki-67 in the core biopsy specimen and a low (,14%) Ki-67 in the surgical specimen. Of note, most cases were ER positive, and there was some discordance of luminal A/B type classification between the core needle biopsy and the surgical specimen. The fixation issue in surgical specimens accounted for many discordant cases. The surgical specimens often showed incomplete fixation in the center with good fixation at the periphery, which led to better staining at the periphery than at the center. In poorly fixed surgical specimens, the MIB1 stained many fewer tumor cells than it did in core needle biopsy samples. Unlike core needle biopsy samples, surgical specimens are inevitably exposed to varying periods of ischemia during tumor removal. 14 This hypoxic damage could result in apoptosis of cancer cells in surgical samples, causing a lower proliferation index. 14 In addition, the fatty characteristic of breast tissue often leads to inadequate fixation in mastectomy specimens. Poor fixation causes not only a decreased number of positive tumor cells, but also poor staining of MIB1. In some mastectomy specimens, the MIB1 was stained very faintly, and the digital image analyzer failed to accurately interpret faint staining. In cases showing poor quality of staining of MIB1, staining of ER, PR, and HER2 showed relatively good quality in the same specimen. It is known that MIB1 is a sensitive marker that is largely influenced by tissue quality. 19 Although the threshold of detection was modified several times in the digital analyzer, accurate detection was not successfully achieved. In this case, manual analysis could be more accurate. A recent study reported that digital image analysis outperforms manual biomarker assessment in breast cancer. 20 However, caution is required in interpretation of weak or unsatisfactory staining in digital image analysis. On the other hand, we also experienced one case that showed a higher Ki-67 index in a surgical specimen due to poor staining quality in the core needle biopsy from an outside hospital. This result also emphasizes the importance of fixation in breast samples. Another major cause of discordance is intratumoral heterogeneity of the Ki-67 index in breast cancer. We measured the Ki-67 index in a hot spot area. In breast cancer, hot spot areas are usually located in the periphery of the tumor. Core needle biopsy contains considerably fewer 366 Arch Pathol Lab Med Vol 142, March 2018 Ki-67 Index in Breast Cancer Ahn et al

Figure 1. Ki-67 index in core needle biopsies and matched surgical specimens. Figure 2. A, Ki-67 index in core needle biopsy: 18%. B, Ki-67 index in surgical specimen: 9%. The poor fixation of surgical specimen caused weak MIB1 staining. Nuclei in whole sections often showed areas of highly variable staining (MIB1 stain, original magnification 3200). Figure 3. This tumor reveals a heterogeneous pattern of Ki-67 index. Tumor cells on the left side show a higher Ki-67 index compared with those on the right (MIB1 stain, original magnification 340). Figure 4. A, Ki-67 index in core needle biopsy: 13%. B, Ki-67 index in surgical specimen: 53%. The poor fixation of the biopsy specimen caused weak MIB1 staining (MIB1 stain, original magnification 3200). Arch Pathol Lab Med Vol 142, March 2018 Ki-67 Index in Breast Cancer Ahn et al 367

Case No. Ki-67 in Core Needle Biopsy Ki-67 in Surgical Specimen Table 2. Estrogen Receptor Analysis of Discordant Cases Progesterone Receptor HER2 Cause of Discordance 1 32 5 Positive Positive Negative Poor fixation in surgical specimen 2 33 3 Negative Negative Negative Poor fixation in surgical specimen 3 24 0.5 Positive Positive Negative Poor fixation in surgical specimen 4 24 9 Positive Positive Positive Poor fixation in surgical specimen 5 20 11 Positive Positive Negative Heterogeneity of proliferation index 6 20 3 Positive Positive Negative Heterogeneity of proliferation index 7 20 12 Positive Positive Negative Poor fixation in surgical specimen 8 21 11 Positive Positive Negative Heterogeneity of proliferation index 9 18 9 Positive Positive Negative Poor fixation in surgical specimen 10 17 6 Positive Positive Negative Poor fixation in surgical specimen 11 15 5 Positive Negative Positive Heterogeneity of proliferation index 12 18 11 Positive Positive Positive Heterogeneity of proliferation index 13 17 6 Positive Positive Negative Heterogeneity of proliferation index 14 10 21 Positive Positive Positive Heterogeneity of proliferation index 15 13 53 Negative Negative Negative Poor fixation in core biopsy specimen 16 8 17 Positive Positive Negative Heterogeneity of proliferation index tumor cells than surgical specimens and is acquired mostly from the central portion of the tumor. An agreement on Ki- 67 measurement methodology has not been reached. There is no consensus on which region to score or the superiority of a digital image analyzer versus manual analysis. Although some institutes adopt the average measurement, many institutes adopt the hot spot area measurement for the Ki-67 index. In our institute, we measured Ki-67 in a hot spot area using a digital image analyzer. It is known that Ki-67 index measured in a hot spot area shows a higher level of interobserver and intraobserver variability. 21 This study has several limitations. We did not perform gene expression tests in breast cancer. Therefore, the correlation of IHC-based subtype and gene expression based subtype was not determined. In addition, the prognostic significance of Ki-67 was not investigated because of the short follow-up period. A further large-scale study with long-term follow-up is warranted to investigate this issue and suggest an established guideline for Ki-67 measurement. In conclusion, we report a substantial discordance in the Ki-67 index between core needle biopsies and matched surgical specimens. We recommend that Ki-67 index measurement be measured in core needle biopsies and surgical specimens for accurate subtyping of breast cancer. Good fixation quality should be assured for accurate Ki-67 index measurement. References 1. Yerushalmi R, Woods R, Ravdin PM, Hayes MM, Gelmon KA. Ki67 in breast cancer: prognostic and predictive potential. Lancet Oncol. 2010;11(2): 174 183. 2. Ignatiadis M, Buyse M, Sotiriou C. St Gallen International Expert Consensus on the primary therapy of early breast cancer: an invaluable tool for physicians and scientists. Ann Oncol. 2015;26(8):1519 1520. 3. Senkus E, Kyriakides S, Ohno S, et al. Primary breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2015;26(suppl 5):v8 v30. 4. Goldhirsch A, Wood WC, Coates AS, et al. Strategies for subtypes dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011. Ann Oncol. 2011;22(8):1736 1747. 5. Dowsett M, Smith IE, Ebbs SR, et al. Short-term changes in Ki-67 during neoadjuvant treatment of primary breast cancer with anastrozole or tamoxifen alone or combined correlate with recurrence-free survival. Clin Cancer Res. 2005;11(2, pt 2):951s 958s. 6. Penault-Llorca F, Andre F, Sagan C, et al. Ki67 expression and docetaxel efficacy in patients with estrogen receptor-positive breast cancer. J Clin Oncol. 2009;27(17):2809 2815. 7. Hugh J, Hanson J, Cheang MC, et al. Breast cancer subtypes and response to docetaxel in node-positive breast cancer: use of an immunohistochemical definition in the BCIRG 001 trial. J Clin Oncol. 2009;27(8):1168 1176. 8. Dowsett M, Nielsen TO, A Hern R, et al. Assessment of Ki67 in breast cancer: recommendations from the International Ki67 in Breast Cancer working group. J Natl Cancer Inst. 2011;103(22):1656 1664. 9. Knutsvik G, Stefansson IM, Aziz S, et al. Evaluation of Ki67 expression across distinct categories of breast cancer specimens: a population-based study of matched surgical specimens, core needle biopsies and tissue microarrays. PLoS One. 2014;9(11):e112121. 10. Chen X, Sun L, Mao Y, et al. Preoperative core needle biopsy is accurate in determining molecular subtypes in invasive breast cancer. BMC Cancer. 2013;13: 390. 11. Collins LC, Botero ML, Schnitt SJ. Bimodal frequency distribution of estrogen receptor immunohistochemical staining results in breast cancer: an analysis of 825 cases. Am J Clin Pathol. 2005;123(1):16 20. 12. Baatz M, Zimmermann J, Blackmore CG. Automated analysis and detailed quantification of biomedical images using Definiens Cognition Network Technology. Comb Chem High Throughput Screen. 2009;12(9):908 916. 13. Arihiro K, Oda M, Ohara M, et al. Comparison of visual assessment and image analysis in the evaluation of Ki-67 expression and their prognostic significance in immunohistochemically defined luminal breast carcinoma. Jpn J Clin Oncol. 2016;46(12):1081 1087. 14. Romero Q, Bendahl PO, Klintman M, et al. Ki67 proliferation in core biopsies versus surgical samples a model for neo-adjuvant breast cancer studies. BMC Cancer. 2011;11:341. 15. Iqbal S, Anderson TJ, Marson LP, Prescott RJ, Dixon JM, Miller WR. MIB-1 assessments in breast cancers. Breast. 2002;11(3):252 256. 16. Kwok TC, Rakha EA, Lee AH, et al. Histological grading of breast cancer on needle core biopsy: the role of immunohistochemical assessment of proliferation. Histopathology. 2010;57(2):212 219. 17. Ricci MD, Calvano Filho CM, Oliveira Filho HR, Filassi JR, Pinotti JA, Baracat EC. Analysis of the concordance rates between core needle biopsy and surgical excision in patients with breast cancer. Rev Assoc Med Bras (1992). 2012;58(5):532 536. 18. Kim HS, Park S, Koo JS, et al. Risk factors associated with discordant Ki-67 levels between preoperative biopsy and postoperative surgical specimens in breast cancers. PLoS One. 2016;11(3):e0151054. 19. Arima N, Nishimura R, Osako T, et al. The importance of tissue handling of surgically removed breast cancer for an accurate assessment of the Ki-67 index. J Clin Pathol. 2016;69(3):255 259. 20. Stalhammar G, Fuentes Martinez N, Lippert M, et al. Digital image analysis outperforms manual biomarker assessment in breast cancer. Mod Pathol. 2016; 29(4):318 329. 21. Chung YR, Jang MH, Park SY, Gong G, Jung WH; Korean Breast Pathology Ki-67 Study Group. Interobserver variability of Ki-67 measurement in breast cancer. J Pathol Transl Med. 2016;50(2):129 137. 368 Arch Pathol Lab Med Vol 142, March 2018 Ki-67 Index in Breast Cancer Ahn et al