Pattern of Tumor Shrinkage during Neoadjuvant Chemotherapy Is Associated with Prognosis in Low-Grade Luminal Early Breast Cancer 1

Transcription

1 This copy is for personal use only. To order printed copies, contact Pattern of Tumor Shrinkage during Neoadjuvant Chemotherapy Is Associated with Prognosis in Low-Grade Luminal Early Breast Cancer 1 Original Research n Breast Imaging Ippei Fukada, MD Kazuhiro Araki, MD, PhD Kokoro Kobayashi, MD Tomoko Shibayama, MD Shunji Takahashi, MD, PhD Naoya Gomi, MD, PhD Yumi Kokubu, MD, PhD Katsunori Oikado, MD, PhD Rie Horii, MD, PhD Futoshi Akiyama, MD, PhD Takuji Iwase, MD Shinji Ohno, MD, PhD Kiyohiko Hatake, MD, PhD Naohiro Sata, MD, PhD Yoshinori Ito, MD, PhD Purpose: Materials and Methods: To evaluate the association between tumor shrinkage patterns shown with magnetic resonance (MR) imaging during neoadjuvant chemotherapy (NAC) and prognosis in patients with lowgrade luminal breast cancer. This retrospective study was approved by the institutional review board and informed consent was obtained from all subjects. The low-grade luminal breast cancer was defined as hormone receptor positive and human epidermal growth factor receptor 2 negative with nuclear grades 1 or 2. The patterns of tumor shrinkage as revealed at MR imaging were categorized into two types: concentric shrinkage (CS) and non-cs. Among 854 patients who had received NAC in a single institution from January 2000 to December 2009, 183 patients with low-grade luminal breast cancer were retrospectively evaluated for the development set. Another data set from 292 patients who had received NAC in the same institution between January 2010 and December 2012 was used for the validation set. Among these 292 patients, 121 patients with low-grade luminal breast cancer were retrospectively evaluated. 1 From the Department of Breast Medical Oncology, Breast Oncology Center (I.F., K.A., K.K., T.S., Y.I.), Department of Medical Oncology (S.T.), Department of Diagnostic Imaging (N.G., Y.K., K.O.), Department of Pathology (R.H., F.A.), Department of Breast Surgical Oncology, Breast Oncology Center (T.I.), Breast Oncology Center (S.O.), and Department of Hematology and Oncology (K.H.), the Cancer Institute Hospital of the Japanese Foundation for Cancer Research, Ariake, Koto-ku, Tokyo , Japan; and Department of Gastrointestinal Surgery, Jichi Medical University, Shimotsuke, Japan (N.S.). Received July 4, 2016; revision requested August 30; revision received February 21, 2017; accepted April 6; final version accepted May 12. Address correspondence to I.F. ( ippei. fukada@jfcr.or.jp). q RSNA, 2017 Results: In the development set, the median observation period was 67.9 months. Recurrence was observed in 31 patients, and 16 deaths were related to breast cancer. There were statistically significant differences in both the disease-free survival (DFS) and overall survival (OS) rates between patterns of tumor shrinkage (P,.001 and P,.001, respectively). Multivariate analysis demonstrated that the CS pattern had the only significant independent association with DFS (P =.001) and OS (P =.009) rate. In the validation set, the median follow-up period was 56.9 months. Recurrence was observed in 20 patients (16.5%) and eight (6.6%) deaths were related to breast cancer. DFS rate was significantly longer in patients with the CS pattern (72.8 months; 95% confidence interval [CI]: 69.9, 75.6 months) than in those with the non-cs pattern (56.0 months; 95% CI: 49.1, 62.9 months; P.001). The CS pattern was associated with an excellent prognosis (median OS, 80.6 months; 95% CI: 79.3, 81.8 months vs 65.0 months; 95% CI: 60.1, 69.8 months; P =.004). Multivariate analysis demonstrated that the CS pattern had the only significant independent association with DFS (P =.007) and OS (P =.037) rates. Conclusion: The CS pattern as revealed at MR imaging during NAC had the only significant independent association with prognosis in patients with low-grade luminal breast cancer. q RSNA, 2017 Radiology: Volume 286: Number 1 January 2018 n radiology.rsna.org 49

2 Neoadjuvant chemotherapy (NAC) (not including hormonal treatment) for breast cancer has been used mainly for locally advanced cancer or inflammatory cancer, aiming at downstaging. The National Surgical Adjuvant Breast and Bowel Project Protocol B-18 and B-27 trials showed similar overall survival (OS) and disease-free survival (DFS) rates between pre- and postoperative chemotherapy (1,2). Preoperative docetaxel administration increased the pathologic complete response (pcr) rate by 12% (2,3). In patients in whom pcr was obtained after NAC, both the OS and DFS rates were favorable (1). Recently, classifications based on biomarkers such as the expression status of hormone receptors or human epidermal growth factor receptor 2 (HER2) have been made, and treatment based on the subtypes has become mainstream. At the St Gallen International Breast Cancer Conference in 2013, breast tumors that were positive for either or both estrogen receptor (ER) and progesterone receptor (PR), were negative for HER2, and that showed a low Ki-67 labeling index were defined as luminal A type tumors (4). This type is hormone sensitive, and various subtype analyses have shown poor responses to chemotherapy in patients with operable luminal A Advances in Knowledge According to the categorization of tumor shrinkage pattern during neoadjuvant chemotherapy (NAC), median disease-free survival (DFS) rate was significantly longer in patients with the concentric shrinkage (CS) pattern (72.8 months) than in those with the non-cs pattern (56.0 months). Regarding overall survival (OS) rate, the CS pattern was associated with an excellent prognosis (median OS, 80.6 months vs 65.0 months). Multivariate analysis also demonstrated that the CS pattern had the only significant independent association with DFS and OS rates. breast cancer (5 7). The pcr rate by using NAC in ER-positive breast cancer is not high in comparison with hormone receptor negative breast cancer (8 10). As von Minckwitz et al reported, pcr was achieved in only 8.9% of patients with low-grade luminal breast cancer, defined as ER-positive and/or PR-positive, HER2-negative, and nuclear grade 1 or 2 (11). Therefore, it is crucial to identify prognostic factors other than pcr in low-grade hormone receptor positive breast cancer. Magnetic resonance (MR) imaging can be used to differentiate between therapy-induced nonvascularized fibrosis and residual vital tumors, so in daily practice, it is used to evaluate the response to NAC (12 15). The response to chemotherapy varies widely according to intrinsic subtypes. Therefore, to monitor the efficacy of NAC, it would be very useful to recognize the pattern of tumor shrinkage that may reflect the biologic characteristics of each intrinsic subtype. Despite our recently increased understanding of the effects of intrinsic subtypes on the response to and prognosis following NAC, it has been unclear whether the patterns of tumor shrinkage during NAC could associate with prognosis. The purpose of the present study was to evaluate the association between tumor shrinkage patterns during NAC and prognosis in patients with lowgrade luminal breast cancer. Materials and Methods Patients and Materials The indication criteria for NAC were a tumor diameter of greater than or Implications for Patient Care The CS pattern as revealed at MR imaging during NAC had the only significant independent association with prognosis in patients with low-grade luminal breast cancer. For patients with a non-cs pattern, a new strategy should be sought that would lead to CS and better clinical outcomes. equal to 3 cm or axillary lymph node metastasis as revealed by preoperative cytodiagnosis. Among 853 patients who had received NAC in our institution between January 2000 and December 2009, 523 patients underwent preoperative needle biopsy and could be reevaluated for the development set used in our study. Among these 523 patients, 183 who had low-grade luminal breast cancer and could be evaluated with MR imaging before and after chemotherapy were retrospectively examined. All patients participating in this study had operable early breast cancer; patients with distant metastases were excluded by means of bone scintigraphy, abdominal ultrasonography, computed tomography (CT), or positron emission tomography/ct. Comprehensive written informed consent for the use of specimen materials was obtained preoperatively from all patients participating as subjects in this study. The retrospective study was approved by the institutional review Content code: Radiology 2018; 286:49 57 Abbreviations: CI = confidence interval CS = concentric shrinkage DFS = disease-free survival ER = estrogen receptor HER2 = human epidermal growth factor receptor 2 NAC = neoadjuvant chemotherapy OS = overall survival pcr = pathologic complete response PR = progesterone receptor Author contributions: Guarantors of integrity of entire study, I.F., K.K., T.S., N.G., Y.I.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, I.F., K.A., T.S., S.T., N.G., S.O., N.S.; clinical studies, I.F., K.A., T.S., S.T., N.G., Y.K., K.O., R.H., T.I., S.O., K.H., N.S., Y.I.; experimental studies, I.F., K.A., T.S., S.O., N.S.; statistical analysis, I.F., K.A., T.S., S.O., N.S.; and manuscript editing, I.F., K.A., K.K., T.S., S.T., N.G., Y.K., K.O., F.A., S.O., K.H., N.S., Y.I. Conflicts of interest are listed at the end of this article. See also the editorial by Moy in this issue. 50 radiology.rsna.org n Radiology: Volume 286: Number 1 January 2018

3 board of the Cancer Institute Hospital of the Japanese Foundation for Cancer Research (approval number, ). Biopsy specimens were embedded in paraffin and cut into thin slices (4 mm), and immunohistochemical status was evaluated in preoperative specimens. They were stained with hematoxylin-eosin and were immunohistochemically examined for ER, PR, and HER2. By using hematoxylin-eosin stained slices, the histologic subtype of cancer and nuclear grade of cancer cells were evaluated according to the general rules for clinical and pathologic assessment of breast cancer, as edited by the Japanese Breast Cancer Society (16). Histologic subtypes were divided into three groups based on the size and structure of invasive components: papillotubular carcinoma, solidtubular carcinoma, and scirrhous carcinoma. Nuclear grade was evaluated via a combination of nuclear atypia and mitotic counts and was divided into three groups in order of increasing atypia: 1, 2, and 3. Immunohistochemical assessment of ER and PR expression was performed by using antibodies for ER (clone 1D5 [Dako Japan, Tokyo, Japan]) and for PR (clone PgR636 [Dako Japan]). Positive reactions for ER and PR were defined as nuclear staining in 10% or more of cancer cells, and negative reactions were defined as nuclear staining in less than 10%. Scoring of ER and PR status was also evaluated by the Allred method, which takes into consideration the proportion of positive cells (scored on a scale of 0 5) and staining intensity (scored on a scale of 0 3). Immunohistochemical detection of HER2 protein was performed by using the HercepTest (Dako Japan). Expression of HER2 protein was classified into four groups: 0, 1+, 2+, and 3+. In those tumors that were classified as 2+, HER2 genetic testing by fluorescence in situ hybridization was performed by using a PathVysion HER-2 DNA Probe Kit (Abbott Molecular, Des Plaines, Ill). Both protein and genetic status were estimated based on the guidelines for HER2 testing in breast cancer, as edited by the American Society of Clinical Oncology and the College of American Pathologists (17). HER2 positivity was defined as HER2 protein 3+ or HER2 gene amplification. Standard chemotherapy including anthracycline and taxane was administered for NAC. The regimens of anthracycline-containing chemotherapy used in this study were fluorouracil 500 mg/m 2, doxorubicin 50 mg/m 2, and cyclophosphamide 500 mg/m 2, as well as fluorouracil 500 mg/m 2, epirubicin 100 mg/m 2, and cyclophosphamide 500 mg/ m 2. The taxane regimens were a weekly dose of paclitaxel 80 mg/m 2, a triweekly dose of paclitaxel 175 mg/m 2, and a triweekly dose of docetaxel 75 mg/m 2. Definition of Low-Grade Luminal Breast Cancer By using the von Minckwitz criteria, low-grade luminal breast cancer was defined as ER-positive and/or PR-positive, HER2-negative, and nuclear grade 1 or 2 in our study (12). MR Imaging Technique Breast MR imaging examinations were performed by using a 1.5-T MR imaging unit (Signa HD; GE Healthcare Japan, Tokyo, Japan). A commercially available, dedicated four-chael breast array coil was used with the 1.5-T MR imaging unit. The MR imaging examination was performed with the patient in the prone position. Our imaging protocol included a localizing sequence followed by unilateral fast spin-echo T2-weighted imaging (repetition time msec/echo time msec, 4800/85; echo train length, 16; matrix, ) with fat suppression by a chemical shift selective imaging sequence. Other parameters were as follows: field of view, 260 mm; section thickness, 3 mm; and intersection gap, 0 mm. This examination was followed by combined dynamic contrast material enhanced unilateral coronal breast imaging. An enhanced T1-weighted examination using a threedimensional gradient-echo fat-suppressed sequence by spectral inversion recovery was performed before and after contrast material injection. The image parameters were as follows: 3.6/1.0; flip angle, 15 ; field of view, cm; matrix, ; section thickness, 3.0 mm; intersection gap, 0 mm; and acquisition time, 60 seconds. The dynamic study in the coronal plane was performed before and 60 seconds, 120 seconds, 180 seconds, and 240 seconds after initiating an intravenous injection of 0.2 mmol per kilogram of gadodiamide hydrate (Omniscan; Daiichi-Sankyo, Tokyo, Japan) at a rate of 3 ml/ sec, followed by a 20-mL saline flush at the rate of 3 ml/sec. In addition, a contrast-enhanced bilateral axial breast imaging study with volume imaging for breast assessment three-dimensional gradient-echo imaging with fat suppression by spectral inversion recovery was performed 300 seconds after the initiation of an intravenous injection. The image parameters were as follows: 4.9/2.3; flip angle, 10 ; field of view, cm; matrix, ; section thickness, 2.0 mm; intersection gap, 0 mm; and acquisition time, 234 seconds. The breast MR imaging examination was performed before NAC, after NAC, and at the time of changing the chemotherapy regimen. MR Imaging Interpretation All MR images from the development set were interpreted by a single breast radiologist (N.G.) with 25 years of experience. Tumor extent, morphology, and relative enhancement were assessed during initial and late enhancement at baseline MR imaging and at MR imaging during NAC. The extent of each tumor was assessed by its largest diameter in three reformatted planes (sagittal, axial, and coronal) at initial and late enhancements. We initially divided the pattern of shrinkage by NAC into six groups: concentric shrinkage (CS), reduction to small foci, diffuse decrease, decrease of intensity only, no change, and enlargement. Then, we simply divided them into the CS pattern and non-cs pattern. The decision to use CS versus non-cs was made after the initial MR imaging interpretation (but before the relationship to prognosis was investigated) based upon the need for a large enough sample size. Radiology: Volume 286: Number 1 January 2018 n radiology.rsna.org 51

4 Figure 1 Figure 1: Images show patterns of tumor shrinkage during neoadjuvant chemotherapy. Breast MR imaging examination was performed before and after NAC. Dynamic study in the coronal plane was performed 60 seconds and 240 seconds after initiating an intravenous injection. (a) Simple concentric shrinkage pattern, (b) concentric shrinkage to small foci, (c) diffuse decrease, and (d) decrease of intensity only. The CS pattern consisted of two types, the simple CS pattern (Fig 1a) and CS to small foci (Fig 1b). The simple CS pattern was defined as follows: first, mass enhancement without nonmass enhancement and any daughter lesion around the main tumor before NAC; second, the simple CS pattern, in which the largest diameter was reduced and during NAC did not yield any nonmass-enhancement lesion or small mass enhancement around the 52 radiology.rsna.org n Radiology: Volume 286: Number 1 January 2018

5 main tumor. The CS to small foci was defined as follows: first, main mass enhancement with or without nonmass enhancement and daughter lesions around the main tumor before NAC; second, a main mass enhancement lesion showing a CS pattern, but residual foci allowed. The focus was defined as a dot of enhancement so small (,5 mm) that it could not be otherwise characterized (18). The other shrinkage patterns, which could not be divided into the above two concentric categories, were defined as non-cs patterns, including diffuse decrease (Fig 1c), decrease of intensity only (Fig 1d), no change, and enlargement. In regard to patterns before NAC, we divided the preoperative MR imaging patterns into two groups: mass enhancement only or with nonmass enhancement. Definition of Histologic Therapeutic Effects pcr was defined according to the criteria established by the University of Texas MD Anderson Cancer Center s trial pcr criteria: ypt0/is ypn0 (no invasive residual in breast or nodes; noninvasive breast residuals allowed) (19). The yptinv ypn0 (invasive residual in breast and no invasive or noninvasive residual in breast or nodes) and yptis/ inv ypn+ (invasive or noninvasive residuals in the breast, infiltrated lymph nodes) were also used. Patients for Validation of the Tumor Shrinkage Pattern For validation of the tumor shrinkage patterns, another data set from 292 patients who had received NAC in our institution between January 2010 and December 2012 was used. Among these 292 patients, 121 who had low-grade luminal breast cancer and could be evaluated with MR imaging before and after chemotherapy were examined. In addition to the previous blinded reviewer (N.G), two blinded independent reviewers (Y.K., K.O.; breast and chest radiologists with 20 years of experience, respectively) evaluated the tumor shrinkage patterns. If all three radiologists agreed that concentric pattern was present, then that term was used for the categorization of tumor response. However, if any radiologist felt that the pattern was not concentric, then the response was categorized as nonconcentric. Statistical Analysis Software (SPSS, version 17.0; IBM, Armonk, NY) was used for statistical analysis. At univariate analysis, the DFS and OS rates were analyzed by using the Kaplan-Meier method with the logrank test. At multivariate analysis, Cox regression analysis was used. A P value of,.05 was considered to indicate statistical significance. Results Development Set: Patient Characteristics and Shrinkage Pattern Distribution Patient characteristics as well as the distribution of tumor shrinkage patterns are shown in Table 1. The median observation period was 67.9 months from surgery, and recurrence was observed in 31 patients (16.9%). Sixteen (8.7%) deaths were related to breast cancer. The 5-year DFS and OS rates for all patients were 84.7% and 94.3%, respectively. The CS pattern was observed in 104 (56.8%) of the 183 patients, and the non-cs pattern was found in 79 patients (43.2%). In regard to pattern before NAC, mass enhancement only was observed in 110 (60.1%), with nonmass enhancement observed in 73 (39.9%). As to the histologic therapeutic effects of NAC, only three patients (1.6%) showed pcr while 180 patients (98.4%) showed non-pcr. Although there was no important difference in age, hormone status, nuclear grade, pathologic type, or NAC regimen among the shrinkage patterns, significant differences between shrinkage patterns and tumor size (P =.002), number of metastatic lymph nodes (P =.002), pattern before NAC (, 0.001), and histologic therapeutic effect (P =.017) were observed. Development Set: Prognosis according to the Tumor Shrinkage Pattern as Shown by MR Imaging We observed six recurrences (3.3%) in the CS pattern group and 25 recurrences (13.7%) in the non-cs pattern group. Two deaths (1.1%) occurred in patients with the CS pattern and 14 deaths (7.7%) occurred in patients with the non-cs pattern. The Kaplan-Meier plots for DFS and OS rates are shown in Figure 2, A and B. According to our categorization, median DFS rate was significantly longer in patients with the CS pattern (115.9 months; 95% confidence interval [CI]: 110.9, months) than in those with the non-cs pattern (78.3 months; 95% CI: 70.2, 86.3 months; P.001). Regarding OS rate, the CS pattern was associated with excellent prognosis (median OS, months; 95% CI: 117.8, months vs 91.5 months; 95% CI: 85.8, 97.2 months; P.001). Development Set: Univariate and Multivariate Cox Regression Analyses for Association with DFS and OS Rates The univariate and multivariate analyses are shown in Table 2. At univariate analysis, more than four metastases of lymph nodes at the time of surgery (P,.001 and P =.012), larger tumor size before NAC (P =.003 and P =.002), and the CS pattern (P,.001 and P,.001) had the significant association with DFS and OS rates. Multivariate analysis demonstrated that more than four metastases of lymph nodes (P =.021), the CS pattern (P =.001), and pathologic subtype (P =.042) had the significant association with DFS rate. Moreover, the CS pattern had the only significant independent association with OS rate (P =.009). Patterns before NAC were associated only with DFS rate at univariate analysis; no significant difference at multivariate analysis was observed. Validation Set: Prognosis, Univariate and Multivariate Cox Regression Analyses for Association with DFS and OS Rates This index was applied to each of the 121 patients with low-grade luminal breast cancer for validation. At the time of analysis, the median follow-up was 56.9 months (range, months); recurrence was observed in 20 patients (16.5%). Eight (6.6%) deaths were related to breast cancer. Radiology: Volume 286: Number 1 January 2018 n radiology.rsna.org 53

6 Table 1 Patient Characteristics and Shrinkage Pattern Distribution Parameter Age (y) CS Pattern (n = 104) Development Set The 5-year DFS and OS rates for all patients were 82.4% and 93.2%, respectively. Non-CS Pattern (n = 79) P Value CS Pattern (n = 74) Validation Set Non-CS Pattern (n = 47) P Value Median Range Tumor size T T T T No. of metastatic lymph nodes Hormone status (Allred score) ER7, PR ER7,.PR ER7,.PR ER7, PR Nuclear grade Pathologic subtype Papillotubular Solid-tubular Scirrhous NAC Anthracycline Anthracycline followed by taxane Taxane Postoperative chemotherapy Anthracycline Taxane Pattern before NAC With nonmass enhancement 25 48, ,.001 Mass only Histologic therapeutic effects ypt0/is ypn yptinv ypn yptis/inv ypn Prognosis Recurrence Death The CS pattern was observed in 74 (61.2%) of the 121 patients, and the non-cs pattern was found in 47 patients (38.8%) (Table 1). In 74 sets of images, all three radiologists agreed that the CS pattern was present. In 44 sets of images, all three radiologists agreed on the non-cs pattern. In three of 121 sets of images (2.5%) there was disagreement on the pattern, and these were counted as non-cs patterns. The Kaplan-Meier plots for DFS and OS rates are shown in Figure 3, A and B. According to our categorization, median DFS rate was significantly longer in patients with the CS pattern (72.8 months; 95% CI: 69.9, 75.6 months) than in those with the non-cs pattern (56.0 months; 95% CI: 49.1, 62.9 months; P.001). Regarding OS rate, the CS pattern was associated with an excellent prognosis (median OS, 80.6 months; 95% CI: 79.3, 81.8 months vs 65.0 months; 95% CI: 60.1, 69.8 months; P =.004). Multivariate analysis demonstrated that the CS pattern had the only significant independent association with DFS (P =.007) and OS (P =.037) rates (Table 2). Discussion Our results showed the CS pattern as revealed at MR imaging during NAC had the only significant independent association with prognosis in patients with low-grade luminal breast cancer. Loo et al reported changes in tumor diameter found at MR imaging during NAC (20). In their study, massive tumor regression after chemotherapy was observed only in triple-negative and HER2-positive tumors. The presence of residual tumors in surgical specimens after NAC in ER-positive/ HER2-negative tumors was not correlated with the change in largest diameter of late enhancement during NAC. They commented that this striking observation might be explained by the heterogeneous appearance and the low rate of pcr in the ER-positive/HER2- negative group. We showed that these shrinkage patterns by NAC might reflect biologic characteristics. Therefore, residual cells in the non-cs pattern in patients who did not show CS may harbor cancer cells that are more resistant to 54 radiology.rsna.org n Radiology: Volume 286: Number 1 January 2018

7 Figure 2 Figure 2: Kaplan-Meier plots show, A, DFS rate and, B, OS rate in the development set. There were statistically significant differences in both the median DFS and OS rates between the two patterns of tumor shrinkage (P,.001 and P,.001, respectively). Table 2 Univariate and Multivariate Cox Regression Analyses for Association with DFS and OS Factor Univariate Analysis Development Set Multivariate Analysis Univariate Analysis Validation Set Multivariate Analysis P Value DFS OS P Value DFS OS DFS OS Relative Risk P Value Relative Risk P Value DFS OS Relative Risk P Value Relative Risk P Value Age (y) 50, Tumor size (before NAC) T1+T2, T3+T Lymph node metastases.n4, n4, Hormone status (Allred score) ER7 and PR7, others Nuclear grade 1, Pathologic subtype Papillotubular, solid-tubular carcinoma, scirrhous carcinoma Shrinkage patterns CS, non-cs,.001, , Pattern before NAC With nonmass enhancement vs mass only Histologic therapeutic effects ypt0/is ypn0, yptinv ypn0, yptis/inv ypn preoperative chemotherapy. One possible cause for the non-cs pattern being a significant association with prognosis might be the characteristic growth pattern of luminal breast cancer. This type of breast cancer tends to grow Radiology: Volume 286: Number 1 January 2018 n radiology.rsna.org 55

8 Figure 3 Figure 3: Kaplan-Meier plots show, A, DFS rate and, B, OS rate in the validation set. There were statistically significant differences in both the median DFS and OS rates between the two patterns of tumor shrinkage (P,.001 and P =.004, respectively). slowly, and it shows a low apoptosis rate and low genetic instability. Foulkes et al proposed a hypothesis that tumors with low growth ability cause heterogeneity within the tumor and that many tumor cells acquire the ability for distant metastasis (21). The limitation of the present study was that no information about proliferative marker indexes such as Ki-67 exists. The patients with luminal breast cancer whom we examined harbored both subtypes with low-proliferative luminal A and high-proliferative luminal B. However, Ki-67 methodology has not been standardized (22) and it is premature to use Ki-67 to distinguish between luminal A and B breast cancer. It has been reported that breast cancers with high expression of Ki-67 also tend to have a high nuclear grade (23 25). Moreover, the St Gallen International Expert Consensus on Primary Therapy for Early Breast Cancer 2015 also showed that immunohistochemistry measurement of proliferative activity using the Ki-67 assay has proved controversial (26). Defining a single useful cut-off point has proved elusive both because Ki-67 displays a continuous distribution (27) and because of analytic and preanalytic barriers to standardized assessment (28). In a previous report, von Minckwitz et References 1. Wolmark N, Wang J, Mamounas E, Bryant J, Fisher B. Preoperative chemotherapy in patients with operable breast cancer: nineal reported that pcr is not a suitable surrogate end point for patients with luminal breast cancer. Among the few reports on the prognosis of luminal A breast cancer, Borges et al revealed that 3-year DFS and OS rates in patients with luminal A breast cancer, which was defined as positive for ER and PR with a low Ki-67 labeling index (cut-off of 14%), were 86.03% and 93.38%, respectively (88.89% and 88.89% in stage III B, respectively) (29). In regard to the prognosis of patients in our study, the 3-year DFS and OS rates were 90.6% and 97.2%, respectively, which were better than shown in their report. In our study, the CS pattern showed significant association with prognosis, suggesting high heterogeneity inside the tumors and involvement of scattered remaining cells in resistance to therapy. Thus, the effects of the current standard therapy are limited in pursuing higher pcr rates in highly heterogeneous luminal breast cancer. Other methods for quantification of tumor response to NAC have been recently evaluated. Park et al found that patients with locally advanced breast cancer and a low apparent diffusion coefficient before treatment were more likely to respond to NAC (30). Wu et al reported the early changes in MR imaging vascular maps and apparent diffusion coefficient were associated with the final treatment response (31). If the effects of standard therapy are unsatisfactory for patients with a non-cs pattern, then a new therapeutic strategy and MR imaging methodology should be sought that would lead to CS and better clinical outcomes. Disclosures of Conflicts of Interest: I.F. disclosed no relevant relationships. K.A. disclosed no relevant relationships. K.K. disclosed no relevant relationships. T.S. disclosed no relevant relationships. S. T. disclosed no relevant relationships. N.G. disclosed no relevant relationships. Y.K. disclosed no relevant relationships. K.O. disclosed no relevant relationships. R.H. disclosed no relevant relationships. F.A. disclosed no relevant relationships. T.I. disclosed no relevant relationships. S.O. disclosed no relevant relationships. K.H. disclosed no relevant relationships. N.S. disclosed no relevant relationships. Y.I. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author reports grants from Chugai, Novartis, Parexel, Esai, Taiho, EPS, Dai-Ichi-Sankyo, Sanofi, MSD, and Austra Zeneca. Other relationships: disclosed no relevant relationships. 56 radiology.rsna.org n Radiology: Volume 286: Number 1 January 2018

9 year results from National Surgical Adjuvant Breast and Bowel Project B-18. J Natl Cancer Inst Monogr 2001;(30): Bear HD, Anderson S, Brown A, et al. The effect on tumor response of adding sequential preoperative docetaxel to preoperative doxorubicin and cyclophosphamide: preliminary results from National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol 2003;21(22): Symmans WF, Peintinger F, Hatzis C, et al. Measurement of residual breast cancer burden to predict survival after neoadjuvant chemotherapy. J Clin Oncol 2007;25(28): Goldhirsch A, Winer EP, Coates AS, et al. Personalizing the treatment of women with early breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer A Oncol 2013;24(9): Pritchard KI, Shepherd LE, O Malley FP, et al. HER2 and responsiveness of breast cancer to adjuvant chemotherapy. N Engl J Med 2006;354(20): Early Breast Cancer Trialists Collaborative Group (EBCTCG). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 2005;365(9472): Swain SM, Jeong JH, Geyer CE Jr, et al. Longer therapy, iatrogenic amenorrhea, and survival in early breast cancer. N Engl J Med 2010;362(22): Guarneri V, Broglio K, Kau SW, et al. Prognostic value of pathologic complete response after primary chemotherapy in relation to hormone receptor status and other factors. J Clin Oncol 2006;24(7): Toi M, Nakamura S, Kuroi K, et al. Phase II study of preoperative sequential FEC and docetaxel predicts of pathological response and disease free survival. Breast Cancer Res Treat 2008;110(3): Alba E, Calvo L, Albanell J, et al. Chemotherapy (CT) and hormonotherapy (HT) as neoadjuvant treatment in luminal breast cancer patients: results from the GEICAM/ , a multicenter, randomized, phase-ii study. A Oncol 2012;23(12): von Minckwitz G, Untch M, Blohmer JU, et al. Definition and impact of pathologic complete response on prognosis after neoadjuvant chemotherapy in various intrinsic breast cancer subtypes. J Clin Oncol 2012;30(15): Loo CE, Teertstra HJ, Rodenhuis S, et al. Dynamic contrast-enhanced MRI for prediction of breast cancer response to neoadjuvant chemotherapy: initial results. AJR Am J Roentgenol 2008;191(5): Manton DJ, Chaturvedi A, Hubbard A, et al. Neoadjuvant chemotherapy in breast cancer: early response prediction with quantitative MR imaging and spectroscopy. Br J Cancer 2006;94(3): Padhani AR, Hayes C, Assersohn L, et al. Prediction of clinicopathologic response of breast cancer to primary chemotherapy at contrast-enhanced MR imaging: initial clinical results. Radiology 2006;239(2): Pickles MD, Lowry M, Manton DJ, Gibbs P, Turnbull LW. Role of dynamic contrast enhanced MRI in monitoring early response of locally advanced breast cancer to neoadjuvant chemotherapy. Breast Cancer Res Treat 2005;91(1): Sakamoto G, Inaji H, Akiyama F, et al. General rules for clinical and pathological recording of breast cancer Breast Cancer 2005;12(Suppl):S1 S Wolff AC, Hammond ME, Hicks DG, et al. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update. Arch Pathol Lab Med 2014;138(2): D Orsi CJ, Sickles EA, Mendelson EB, Morris EA. ACR BI-RADS Atlas, Breast Imaging Reporting and Data System. 5th ed. Reston, Va: American College of Radiology, Green MC, Buzdar AU, Smith T, et al. Weekly paclitaxel improves pathologic complete remission in operable breast cancer when compared with paclitaxel once every 3 weeks. J Clin Oncol 2005;23(25): Loo CE, Straver ME, Rodenhuis S, et al. Magnetic resonance imaging response monitoring of breast cancer during neoadjuvant chemotherapy: relevance of breast cancer subtype. J Clin Oncol 2011;29(6): Foulkes WD, Reis-Filho JS, Narod SA. Tumor size and survival in breast cancer a reappraisal. Nat Rev Clin Oncol 2010;7(6): Luporsi E, André F, Spyratos F, et al. Ki- 67: level of evidence and methodological considerations for its role in the clinical management of breast cancer: analytical and critical review. Breast Cancer Res Treat 2012;132(3): Barbareschi M, Girlando S, Mauri FM, et al. Quantitative growth fraction evaluation with MIB1 and Ki67 antibodies in breast carcinomas. Am J Clin Pathol 1994;102(2): Keedy JC, el-badawy N, DeRose PB, Cohen C. Comparison of cell proliferation in breast carcinoma using image analysis (Ki-67) and flow cytometric systems. Anal Quant Cytol Histol 1992;14(4): Gasparini G, Pozza F, Meli S, Reitano M, Santini G, Bevilacqua P. Breast cancer cell kinetics: immunocytochemical determination of growth fractions by monoclonal antibody Ki-67 and correlation with flow cytometric S-phase and with some features of tumor aggressiveness. Anticancer Res 1991;11(6): Coates AS, Winer EP, Goldhirsch A, et al. Tailoring therapies improving the management of early breast cancer: St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer A Oncol 2015;26(8): Denkert C, von Minckwitz G. Reply to Ki67 in breast cancer: a useful prognostic marker! A Oncol 2014;25(2): Polley MY, Leung SC, McShane LM, et al. An international Ki67 reproducibility study. J Natl Cancer Inst 2013;105(24): Borges GS, Colchon PH, Staak MC Jr, et al. Analysis of disease-free survival and overall survival in patients with luminal a breast cancer stratified according to TNM. Klin Onkol 2014;27(3): Park SH, Moon WK, Cho N, et al. Diffusion-weighted MR imaging: pretreatment prediction of response to neoadjuvant chemotherapy in patients with breast cancer. Radiology 2010;257(1): Wu LA, Chang RF, Huang CS, et al. Evaluation of the treatment response to neoadjuvant chemotherapy in locally advanced breast cancer using combined magnetic resonance vascular maps and apparent diffusion coefficient. J Magn Reson Imaging 2015;42(5): Radiology: Volume 286: Number 1 January 2018 n radiology.rsna.org 57