Abstract
Background/Aim: The Oncotype DX Recurrence Score (ORS) predicts the likelihood of recurrence and the benefit of chemotherapy in patients with estrogen receptor (ER)-positive, human epidermal growth factor receptor 2 (HER2)-negative early-stage breast carcinoma (ESBC). Tumor budding (TB) is a poor prognostic factor in breast carcinoma. This study aimed to determine the clinicopathological significance of TB in predicting ORS in patients with ESBC. Patients and Methods: We included 359 patients with ER-positive, HER2-negative ESBC. The number of peritumoral TB was assessed, and the cases were categorized into TB-low (<10 buds) and TB-high (≥10 buds) groups. Results: Patients with TB-high ESBC (170/359; 47.4%) showed a significantly higher median ORS (15.0 vs. 13.0) than those with TB-low tumors (189/359; 52.6%). Multivariate analysis revealed that high TB level was an independent predictive factor for higher ORS in patients with ESBC. Conclusion: High TB in ESBC independently predicted higher ORS. TB may serve as a surrogate marker for predicting ORS in patients with ESBC.
Breast carcinoma is one of the most common malignancies and the second leading cause of carcinoma-related death in women worldwide (1-3). Despite considerable advances in both laboratory research and clinical practice, the global incidence and mortality rates of breast carcinoma continue to increase (4). Accumulating evidence suggests that breast carcinomas with different histological and molecular features exhibit distinct biological behaviors that lead to different management strategies and treatment responses (5).
Several multigene assays have been used to predict the risk of recurrence and the potential benefits of chemotherapy in patients with breast carcinoma (6, 7). The Oncotype DX (ODX; Exact Sciences, Marlborough, MA, USA) test is a 21-gene reverse transcriptase polymerase chain reaction assay that analyzes the activity of a group of genes that can affect how the tumor is likely to behave and respond to treatment. The ODX Recurrence Score (ORS; Exact Science) has been reported to predict the probability of distant recurrence and the benefit of chemotherapy in patients with estrogen receptor (ER)-positive, human epidermal growth factor receptor 2 (HER2)-negative early-stage breast carcinoma (ESBC) (8, 9). The Trial Assigning Individualized Options for Treatment (TAILORx) study found that patients with ESBC who were 50 years of age or older with ORS ≤25 and those who were younger than 50 years with ORS ≤15 were candidates for avoiding adjuvant chemotherapy (10). Based on the results of the TAILORx study, the National Comprehensive Cancer Network (NCCN) guidelines indicated that ODX can help guide treatment decisions for ER-positive, lymph node (LN)-negative breast carcinoma (11). The ORS has been shown to be associated with histological grade, progesterone receptor (PR) expression, and proliferative activity (12-14). Low ORS results in downstaging regardless of the tumor size, leading to modified management strategies for patients in this group because of the low chemotherapy benefit (9). Despite the efficacy of ODX, there are several limitations and drawbacks to its use, including high cost, limited availability in resource-poor settings, a several-week delay from the time of surgery to test results, and residual uncertainty in the management of patients with intermediate ORS (15).
Tumor budding (TB) is a well-recognized histological parameter of epithelial malignancies (16). TB is morphologically defined by the presence of isolated tumor cells or small clusters (16). Intratumoral TB refers to tumor buds in the central part of the tumor, whereas peritumoral TB represents those at the invasive tumor front (17). TB provides a histological basis for tumor invasion and metastasis and has been identified as a poor prognostic factor for various human malignancies (18, 19). Previous studies have reported that TB is associated with adverse clinicopathological characteristics, including larger tumor size, worse histological differentiation, lymphovascular invasion, LN metastasis, and resistance to chemotherapy or radiation therapy (20-23). There was also a significant correlation between the degree of TB and tumor recurrence in the liver, lungs, LNs, and peritoneum (24). Furthermore, TB is an unfavorable prognostic factor for most tumor types and is associated with worse overall and disease-free survival (18, 19, 25-28).
The clinicopathological and prognostic significance of TB has been reported in patients with breast carcinoma. However, the implementation of TB in clinical practice is currently limited by the lack of a standardized scoring system (29). This study aimed to evaluate the predictive value of TB in association with ORS and the clinicopathological characteristics of patients with ESBC. We hypothesized that high TB in ESBC would correlate with a higher ORS and that TB would serve as a significant predictor of ORS.
Patients and Methods
This study was reviewed and approved by the Institutional Review Board (approval number: 2023-09-013) of Kangbuk Samsung Hospital (Seoul, Republic of Korea). Between January 2016 and December 2022, 359 consecutive patients with primary breast carcinoma who underwent surgery for ER-positive, HER2-negative ESBC were identified from the institutional database. None of the patients received preoperative neoadjuvant chemotherapy. Patients with only precancerous disease (ductal carcinoma in situ) were excluded from the study. For patients who underwent the ODX test, the results were provided as a continuous ORS.
Two board-certified pathologists specializing in breast oncology thoroughly examined all the available hematoxylin and eosin (H&E)-stained and immunostained slides to confirm the pathological diagnosis. They also reviewed the electronic medical records and pathology reports to collect the following clinicopathological information: patients’ age at the time of initial diagnosis, greatest dimension of the tumor, pathological tumor stage (pT), LN metastasis, pathological nodal stage (pN), histological type, nuclear grade, mitotic count (per 10 high-power fields), hormone receptor expression status, and Ki-67 labeling index. The expression statuses of hormone receptors and Ki-67 were determined by immunostaining. ER and PR immunoreactivities were analyzed using the Allred scoring system (30). For both proteins, Allred scores of 0-2, 3-4, and 5-8 were interpreted as negative, weakly positive, and strongly positive, respectively. To estimate the Ki-67 labeling index, three 20× objective fields were digitally captured using the GenASis Capture and Analysis Platform (Applied Spectral Imaging, Carlsbad, CA, USA), and the index was measured using a GenASis HiPath system (Applied Spectral Imaging).
Peritumoral TB was assessed in H&E-stained slides obtained from surgically resected tumor specimens. In accordance with the recommendations of the International Tumor Budding Consensus Conference (ITBCC) (31), TB was defined as the presence of a single tumor cell or small cellular clusters comprising four or fewer tumor cells. TB was classified using a two-tiered system (16). High TB was defined as the presence of ≥10 buds per 20× objective field and low TB was defined as <10 buds. Figure 1 shows the representative photomicrographs demonstrating high (Figure 1A and B) and low (Figure 1C and D) TB in ESBC.
Representative photomicrographs showing the degree of tumor budding (TB) in early-stage breast carcinoma. (A) High TB (20 blue circles). (B) High TB (17 blue circles). (C) Low TB (three green circles). (D) Low TB (three green circles). Staining method: hematoxylin and eosin staining. Original magnification: A-D, 100×.
Using formalin-fixed, paraffin-embedded tumor tissues, ODX quantified the expression of 16 breast carcinoma-related genes and five reference genes, including proliferation-related, metastasis-related, HER2-related, sex hormone-related, and internal control genes (32). The numerical ORS was calculated based on the ODX formula (33) and ranged from 0 to 100. The highest weight was assigned to proliferation-, HER2−, and sex hormone-related genes (34). According to the recommendation by the NCCN for the use of ODX for ER-positive, HER2-negative ESBC (11, 15), the tumors were classified into two groups: ORS-lower (<16) and ORS-higher (≥16) ESBCs.
Statistical analyses were conducted using the R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). We used an independent two-sample t-test, Mann–Whitney U-test, Pearson chi-square test, or Fisher’s exact test to investigate the association between the degree of peritumoral TB and the clinicopathological characteristics of patients with ESBC. Univariate and multivariate logistic regression analyses were performed to identify clinicopathological parameters that significantly predicted a higher ORS. Statistical significance was set at p<0.05.
Results
Table I summarizes the baseline clinicopathological characteristics of the 359 patients with ER-positive and HER2-negative ESBC. The median patient age was 50 years [interquartile range (IQR)=45-60 years]. One hundred and eighty-two patients (50.7%) were aged 50 years or younger, and 177 patients (49.3%) were older than 50 years. The median tumor size was 1.5 cm (IQR=1.0-1.9 cm), and most patients (296/359; 74.9%) had pT1 tumors. Most patients (331/359, 92.2%) did not show LN metastasis (pN0), whereas 28 patients (7.8%) were diagnosed with pN1 tumors. Most tumors (297/359, 82.7%) were invasive ductal carcinomas (IDC), and 42 (11.7%) were invasive lobular carcinomas (ILC). More than 95% of the tumors exhibited either nuclear grade 1 (163/359; 45.4%) or 2 (183/359; 51.0%). Regarding hormone receptor expression status, almost all tumors (356/359; 99.2%) displayed ER Allred score ≥5. Similarly, 82.2% (295/359) of the tumors showed PR Allred score ≥5. The median mitotic count and Ki-67 labeling index were 2.0 (IQR=1.0-5.0) and 7.4% (IQR=4.4-12.9%), respectively. The median number of TB per 20× objective field was 8.0 (IQR, 3.0-23.5). More than half of the patients (210/359; 58.5%) were classified into the ORS-lower group (ORS <16), whereas 41.5% (149/359) of the patients with ORS ≥16 were classified into the ORS-higher group.
Baseline clinicopathological characteristics of 359 patients with early-stage breast carcinoma.
We examined whether peritumoral TB status was significantly associated with the clinicopathological characteristics of patients with ESBC. According to the cutoff value (10 buds), the cases were classified into TB-low (189/359; 52.6%) and TB-high groups (170/359; 47.4%). The median number of TB of the latter group (24.0; IQR=15.0-37.0) was significantly higher than that of the former group (3.0; IQR=0.0-5.0). As shown in Table II, patients with TB-high ESBC were more likely to develop LN metastasis than those with TB-low ESBC (p=0.014). Most ILCs (35/42; 83.3%) were TB-high (p<0.001). Patients with TB-high ESBC had significantly higher ORS scores than those with TB-low ESBC (p=0.019). Approximately half of the patients with TB-high ESBC (82/170; 48.2%) had a higher ORS, whereas approximately two-thirds of the patients with TB-low tumors (122/189; 64.6%) had a lower ORS. The median ORS of patients with TB-high ESBC (15.0; IQR=10.0-20.0) was significantly higher than that of patients with TB-low ESBC (13.0; IQR=9.0-18.0). The differences in median age (p=0.222), pT (p=0.151), nuclear grade (p=0.960), median ER Allred score (p=0.217), median PR Allred score (p=0.087), median Ki-67 labeling index (p=0.443), and median mitotic count (p=0.270) between TB-low and high groups were not significant.
Relationship between tumor budding (TB) and clinicopathological characteristics of patients with early-stage breast carcinoma.
Table III summarizes the results of the logistic regression analyses, which were performed to identify independent predictive parameters for higher ORS in patients with ESBC. In univariate logistic regression analysis, nuclear grade 3 [p=0.004; odds ratio (OR)=2.24; 95% confidence interval (CI)=1.69-12.88], higher PR Allred score (p<0.001; OR=5.76; 95%CI=3.12-10.66), higher Ki-67 labeling index (p=0.013; OR=1.99; 95%CI=1.16-3.42), and high TB (p=0.014; OR=1.70; 95%CI=1.11-2.59) were found to be significantly associated with higher ORS. Patients’ age (p=0.107) and tumor size (p=0.738) were not significantly correlated with the ORS status. The nuclear grade 3 [p=0.005; hazard ratio (HR)=10.12; 95%CI=2.05-50.02], higher PR Allred score (p<0.001; HR=6.16; 95%CI=3.25-11.66), and high TB (p=0.030; HR=1.67; 95%CI=1.05-2.64) were found to maintain statistical significance in multivariate logistic regression analysis, indicating that these are independent predictive factors for higher ORS in patients with ESBC. In contrast, the Ki-67 labeling index did not independently correlate with ORS (p=0.063; HR=1.80; 95% CI=0.97-3.33).
Factors predicting Oncotype DX Recurrence Score in patients with early-stage breast carcinoma.
Discussion
Over the years, multigene assays have been developed to guide treatment based on the prediction of recurrence risk. ODX is the most widely used assay and has been incorporated into major international consensus guidelines for breast carcinoma treatment (35, 36). Patients with ER-positive ESBC may benefit from the implementation of ODX, which helps to determine the need for chemotherapy. This would ensure that high-risk patients are prescribed systemic treatment and other patients are protected from the unnecessary use of chemotherapy that might not add any benefit to their survival (37). ORS correlates with a combined scoring system using immunoreactivities for ER, PR, Ki-67, and HER2 (38, 39). Allison et al. (40) found a strong correlation between low ORS and histological grade 1, strong PR expression, and low Ki-67 labeling index (≤10%), and between high ORS and histological grade 3, low-to-absent PR expression, and high Ki-67 labeling index (>10%). Given that patients with breast carcinoma with a higher Ki-67 labeling index achieved a higher pathological complete response rate, the Ki-67 labeling index is considered a strong predictive factor for the effectiveness of neoadjuvant and adjuvant chemotherapy (39). Consistent with previous findings, this study also observed a significantly higher ORS in patients with a high Ki-67 labeling index and nuclear grade 3.
In contrast, although scoring systems that can closely estimate ORS have been developed using a combination of conventional prognostic parameters (nuclear grade, mitotic count, hormone receptor expression status, and HER2 expression status), high concordance rates were only observed in low- or high-risk groups, but not in intermediate-risk cases (32). Previous studies reported conflicting results regarding the association between ORS and conventional clinicopathological features. Wolf et al. (41) found no significant correlations between well-known clinicopathological variables and ORS in 109 patients with breast carcinoma. They also reported that neither standard prognostic parameters nor commonly used assessment tools reliably predicted ORS, indicating the need for additional studies on the predictive role of ORS. Aaron et al. (42) also found that the Ki-67 labeling index and conventional prognostic markers were not significantly correlated with ORS and concluded that the Ki-67 labeling index is not a substitute for ODX in settings where conventional prognostic markers do not show a clear indication for or against adjuvant chemotherapy. Similarly, in recent studies examining the correlation between circulating and disseminating tumor cells, a higher ORS was not significantly associated with the presence of circulating tumor cells in the blood or the dissemination of tumor cells in the bone marrow (43, 44).
Therefore, we aimed to verify whether TB could be a novel biomarker for predicting aggressive clinical behavior in patients with ESBC. The prognostic or predictive role of TB in breast carcinoma has not been thoroughly investigated (45), and there are no standardized criteria for quantifying the number of TB in breast carcinoma. Methodological variations in TB grading make it less straightforward to compare the results of different studies. The development of international, evidence-based, and standardized scoring systems is essential to clarify the prognostic role of TB in breast carcinoma and conduct multi-institutional clinical trials. Liang et al. (28) conducted receiver operating characteristic analysis to determine whether seven TBs per 0.950 mm2 were indicative of high TB, whereas Salhia et al. (46) used the mean number of TB across 10 high-power fields. Liang et al. (28) demonstrated that TB was associated with adverse clinicopathological features of IDC, including larger tumor size, lymphovascular invasion, LN metastasis, high tumor-stroma ratio, and low inflammatory infiltrate. They also reported that higher TB was associated with lower overall and cancer-specific survival rates in patients with IDC. Salhia et al. (46) investigated the value of TB as a predictive marker of LN metastasis by comparing 148 resected specimens and 99 matched preoperative biopsy specimens with IDC. They found that high TB observed in resection and biopsy specimens was associated with LN metastases and lymphovascular invasion, highlighting the possibility that this feature may help in therapeutic decision-making. In this study, we used a cutoff value of 10 buds per 20× objective field, in accordance with the 2016 ITBCC recommendations (31). Our observation that the number of TB was higher in patients with ESBC with LN metastasis is consistent with previous data. More importantly, we found that patients with TB-high ESBC had significantly higher ORS than those with TB-low ESBC. Moreover, in multivariate analysis, high TB was found to be statistically significant, indicating that it is an independent predictive factor for higher ORS in patients with ESBC. To our knowledge, this is the first study to demonstrate the significant predictive value of TB in association with ORS in patients with ESBC.
A meta-analysis by Lloyd et al. (25) showed that ER− and HER2-positive breast carcinomas had higher levels of TB than triple-negative tumors, suggesting that TB may be driven by hormones and that the numbers of TB may vary among different breast carcinoma types. Although we did not find any significant correlation between the number of TB and hormone receptor status in this study, further investigations are necessary to clarify whether there is a significant difference in the number of TB according to the expression status of ER, PR, and HER2 as well as the histological types. Interestingly, our observation that ILC showed significantly higher numbers of TB than IDC was consistent with the theoretical mechanism of TB in association with the down-regulation of E-cadherin (45). However, we found that in a subset of ILC, it was difficult to distinguish isolated tumor cells and small cellular clusters from TB.
This study had some limitations. First, we enrolled patients with ESBC who underwent surgery at a single institution. As this study did not include patients who were diagnosed with HER2-positive or triple-negative breast carcinoma, the cohort size was relatively small. Second, we did not analyze the statistical differences in survival according to ORS. Further investigations using more detailed prognostic information obtained from larger ESBC cohorts are necessary. Third, we did not use morphometric or computer-assisted analyses for the quantification of TB. Comparing the manual counting and computerized analysis of TB may better address its clinicopathological significance.
In conclusion, we showed that a higher number of peritumoral TB correlated with a higher probability of LN metastasis and higher ORS as a novel finding in ESBC. To the best of our knowledge, this study is the first to show the significant value of TB in predicting a higher ORS in patients with ESBC. Our observations raise the possibility that, in a setting where ODX is not available, TB may play a significant role in aiding decision-making regarding adjuvant chemotherapy for breast carcinoma.
Acknowledgements
This work was supported by the Samsung Medical Center Grant (SMO1230291), the National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIT) (2023R1A2C2006223), and the Medical Research Funds from Kangbuk Samsung Hospital.
Footnotes
Authors’ Contributions
All Authors made substantial contributions to the conceptualization and design of this study; the acquisition, interpretation, and validation of the data; drafting and critical revision of the manuscript; and final approval of the version to be published.
Conflicts of Interest
The Authors declare that they have no conflicts of interest or financial ties related to this study.
- Received October 5, 2023.
- Revision received November 1, 2023.
- Accepted November 2, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
