香京julia种子在线播放

Endocrine therapy is inevitable for patients with hormonal status–positive breast cancer to prevent local recurrence and distant metastasis (1–3). Generally, patients with estrogen receptor (ER)- or progesterone receptor (PR)-positive breast cancer are treated with adjuvant endocrine therapy for five years after surgical treatment (4, 5). It is important to evaluate the recurrence risk and determine whether to maintain or stop endocrine therapy after five years based on side effects, such as postmenopausal symptoms, and patient quality of life (6–8). Therefore, it is necessary to decide whether to stop or continue endocrine therapy after weighing the side effects of therapy and the risk for recurrence or metastasis of breast cancer.

Recently, Dowsett and colleagues introduced a tool called the Clinical Treatment Score post–five years (CTS5) as a scoring system to help decide whether to stop or continue treatment after five years of endocrine therapy using several clinicopathologic parameters including tumor size, nodal status, and histopathologic grade (9, 10). This scoring system was developed using data from the Arimidex, Tamoxifen, Alone or in Combination (ATAC) trial, which included postmenopausal women with ER-positive or ER-unknown early breast cancer (11, 12). The ATAC trial categorized patients into three risk groups (low, intermediate, and high) for estimating the prognostic performance for late distant metastasis.

However, the CTS5 scoring system might not be as effective in Asian countries because there are many more young breast cancer patients than in Western society (13). In prior research, CTS5 was applied to postmenopausal women in the ATAC and BIG 1-98 study cohorts at diagnosis, and the algorithm was not applied to premenopausal patients (11, 14, 15). CTS5 provides a convenient way to predict distant recurrence (DR) but has limitations in extending its use to all ER- or PR-positive breast cancer patients.

In the present study, we aimed to validate the CTS5 score and develop a modified scoring system to predict distant metastasis not only in postmenopausal women, but also in premenopausal women. We used data from a single institution as the validation set and analyzed participants after subdividing them into pre- and postmenopausal groups to differentiate existing CTS5 scores and identify the prognostic value of CTS5 according to menopausal status.

The demographics and baseline characteristics of this study are described in Table 1 . Among 2,605 patients included in the present study, 1,749 (67.14%) were premenopausal, and 856 (32.86%) were postmenopausal. The median follow-up period was 94.69 months [59.97 – 233.85]. The median age of the premenopausal women was 44.00 years, and that of the postmenopausal women was 56.50 years; overall, the average median age of the study population was 46.00 years. Nodal status, tumor grade, and tumor size were not significantly different between the two groups. A total of 1,902 patients (73.04%) received adjuvant chemotherapy, and more premenopausal women than postmenopausal women received chemotherapy (77.02% vs. 64.91%; p < 0.0001). There was no difference between pre- and postmenopausal patients in terms of tumor size, tumor grade, or nodal status. Tamoxifen was administered in 1,481 patients (56.85%) total, 1,325 of whom were premenopausal. Goserelin treatment as a subcutaneous injection of a depot formulation was observed in 159 patients, all of whom were premenopausal. As for aromatase inhibitors, 336 postmenopausal patients (39.25%) received anastrozole, and 278 (32.48%) received letrozole. During the first five years of treatment, 126 patients were switched to aromatase inhibitors, and all 126 patients were classified as premenopausal. Moreover, 110 (4.22%) cases of late DR were recorded, with an annual hazard rate of 1.41% (95% CI: 1.16%–1.70%). There was no significant difference between the two groups in terms of DR cases or annual rate of DR after five years of adjuvant endocrine therapy. The rate for late DR of HER2-positive patients was significantly lower than that of HER2-negative patients (1.64% vs. 4.49%; p = 0.0351) ( Supplementary Table 1 ).

Tumor size, tumor grade, and nodal status were arranged in a separate table according to CTS5 risk category status in pre- and postmenopausal women ( Table 2 ). Overall, 86.79% (n = 2,261 patients) were categorized as low risk, 5.95% (n = 155 patients) were categorized as intermediate risk, and 7.26% (n = 189 patients) were categorized as high risk for late DR. Notably, more than 90% of patients in the moderate to poor tumor grade group were categorized high risk, as were all patients with more than two positive nodes. Combined ATAC and BIG 1-98 cohort data and data from participants of this study are compared in Table 3 . Compared with the ATAC and BIG 1-98 cohorts, there was no difference in tumor size or nodal status. The rate of chemotherapy in our study was more than 70% higher than that of the ATAC and BIG 1-98 cohorts, representing a significant difference, and low risk was identified in more than 80% of patients with the CTS5 score at the time of validation. The HR of five- to 10-year DR risk among premenopausal women was 0.817 (95% CI: 0.547–1.221; p = 0.3236), which was lower than that among postmenopausal women ( Figure 2 ).

The time-dependent AUC at 10 years is presented with 95% CI value ( Figure 3 ). The AUC for all patients was 64.71 (95% CI: 57.75–71.67). Among postmenopausal women, the AUC exceeded the total population AUC at 72.71 (95% CI: 57.75–71.67), whereas that in premenopausal women was 61.75 (95% CI: 52.97–70.53).

Histograms for CTS5 score are shown according to menopausal status with validated prognostic values of CTS5 for risk of DR between five and 10 years ( Figure 4 ). Importantly, the premenopausal group included a greater proportion of patients at low risk than did the postmenopausal group. That is, premenopausal women unexpectedly had lower CTS5 scores than postmenopausal women, and this result correlated with the many patients at low risk.

There is a need for prognostic tools that can predict late recurrence rate after five years of endocrine therapy; CTS5 is a useful tool for satisfying this need and supporting clinicians in decision-making regarding extension of endocrine therapy (9, 17). This study is significant in that CTS5 was validated in premenopausal women, who account for the majority of breast cancer patients in Korea. After validation, we found that premenopausal women occupied a large portion of the low-risk recurrence group—in other words, the risk for late DR was underestimated by CTS5 in premenopausal women. Therefore, development of a predictive late DR model for premenopausal women is necessary.

Notably, compared with our study population, the ATAC cohort, used as the training set in CTS5, and the BIG 1-98 cohort, used as the validation set, both included postmenopausal women. Nodal status, tumor grade, and tumor size were similar between the population in our study and the ATAC plus BIG 1-98 cohort. In our study, there was no significant difference in tumor size, tumor grade, or nodal status between pre- and postmenopausal women. There was also no difference in DR or annual DR between the two groups. However, when risk was divided according to the cutoff value of CTS5, patients at low risk were more than twice as numerous relative to the cohorts of the ATAC and BIG 1-98 trials. In addition, the premenopausal group contained a greater number of patients at intermediate and high risk than did the postmenopausal group. Interestingly, the rate of receiving chemotherapy was close to 80% in our study, while about 20% of the ATAC plus BIG 1-98 cohort received adjuvant chemotherapy. For this reason, it is thought that the frequency of adjuvant chemotherapy was high among premenopausal women. The AUC had a greater predictive rate for postmenopausal women than for total patients but was less predictive for premenopausal women than for the entire cohort. In other words, as CTS5 was created based on data of postmenopausal women, there is a high probability that it cannot efficiently be applied to premenopausal women in many Asian countries (18). Premenopausal women are likely to be classified in an underestimated risk group when CTS5 is used, and it is necessary to introduce a new scoring system to address this.

Recently, many studies have analyzed the prognostic value of CTS5. Villasco et al. (19) found that CTS5 has prognostic value in predicting late DR in both pre- and postmenopausal women by testing its clinical validity in a retrospective cohort, while Lee et al. (20) similarly concluded that CTS5 is a good prognostic tool for evaluating the risk of late distant recurrence in both pre- and postmenopausal women using the Ki-67 labeling index and confirmed its prognostic performance in premenopausal women. Although Lee et al. reported that their risk groups presented differences in tumor grade relative to the ATAC and BIG 1-90 cohorts, there was no significant difference in tumor grade between our group and the ATAC plus BIG 1-90 cohort. We think that the reason for the different risk groups might be related to whether or not adjuvant chemotherapy was available.

HER2 gene amplification is known to have an impact on breast cancer, and the intracellular signaling pathway of estrogen receptors and HER2 has a complex connection (21). Wang et al. (22) recently suggested that HER2 status has an effect on late DR in hormone receptor–positive breast cancer. In HER2-positive patients, such as those with luminal type B disease, a less effective prognostic value was found in their study. Cases of HER2-positive breast cancer are considered high-risk for DR but, due to the development of anti-HER2–targeted therapy, it is not effective for the HER2-positive model, and a further prediction model is required.

Recently, Noordhoek et al. (23) published a validity and accuracy study of the CTS5 for predicting the rate of late DR in the TEAM and IDEAL trials, arguing the CTS5 overestimates the risk of late DR in high-risk subgroups and suggesting that CTS5 should be used cautiously for interpreting the DR rate among patients at high risk despite its ease of use. Based on these findings, CTS5 must be applied carefully, and unique validation is justified in Korea, where there are many young breast cancer patients.

Various multigene assays have been developed to complement predictions using existing clinicopathologic parameters because of the lack of predictive values for late DR. Among them, GenesWell™ BCT has been validated as a tool to predict late DR (24). In some patients who underwent CTS5 validation, GenesWell™ BCT also was applied, and the CTS5-derived risk group correlated with the gene assay risk group ( Supplementary Figure 1 ). However, the use of multigene assays is controversial in predicting late DR because some assays have not been validated, and the importance of clinical risk has been emphasized, especially in multigene panels of patients under 50 years of age (25). In conclusion, balanced application of a clinicopathologic prediction model and a multigene assay should be conducted to predict late DR, and further study is needed for proper selection of patients to receive extended adjuvant endocrine therapy.

The main limitation of this study is that it was a retrospective study conducted in a single center. It is necessary to conduct multicenter studies to further explore the limitations of CTS5, and a modified version of CTS5 should be developed and validated in large sets through multicenter research in South Korea. In addition, it is necessary to accurately stratify risk groups by creating new cutoff values suitable for Korean patients. Accordingly, we plan to create a web-based search tool suitable for use in Korea.

When the CTS5 calculator was developed, HER2-positive patients were included despite the small population. Because the diagnostic technique has not been applied, such as silver-enhanced in situ hybridization in our data prior to 2003, it is not clear to describe the HER2 status. Due to development of HER2 gene amplification diagnosis, HER2 status can be accurately described, allowing not only chemotherapy, but targeted therapy to be administered. HER2-positive breast cancer is different from hormone receptor-positive breast cancer in terms of molecular biology, and further study is needed to predict late DR excluding the HER2-positive subtype. In addition, this study did not confirm the difference in CTS5 risk among premenopausal women according to use of GnRH agonists because of the small population (156 patients) of premenopausal patients received GnRH agonists. GnRH agonists have been used to suppress ovarian function in young patients with luminal-type breast cancer and premenopausal patients with high risk for DR due to poor prognosis (26, 27). In the SOFT-TEXT and ASTRRA trials, it was found that the prognosis of young breast cancer patients could be improved depending upon the use of GnRH agonists (28–31). In the future, additional CTS5 evaluation is required according to the use of GnRH agonists in premenopausal women, and it is necessary to evaluate the late prediction rate according to the combination of CTS5 and multigene assay by GnRH agonist use.

CTS5 can be prognostic, but risk evaluation is dependent upon traditional clinicopathologic parameters such as tumor size, tumor grade, and nodal status; it cannot provide customized guidance for each individual. Therefore, its combination with a multigene assay is important for deciding whether to extend endocrine therapy (17, 32). Furthermore, in addition to a clinical calculator, such as CTS5, and a multigene assay, further combination with an immunohistochemistry assay and radiographic imaging (computed tomography, positron-emission tomography, and magnetic resonance imaging) or tumor marker assessments is important for predicting patient prognosis (25, 33).

In conclusion, although CTS5 was created to support decision-making by clinicians about extending adjuvant endocrine therapy in postmenopausal women with positive hormonal receptors, there are limitations in predicting late DR in premenopausal women. For populations of premenopausal women with greater rates of breast cancer, a modified scoring system for late DR prediction is needed so as not to underestimate the recurrence risk.

The original contributions presented in the study are included in the article/Supplementary Materials, further inquiries can be directed to the corresponding author.

The present study was approved by the Review Committees (no. 2020-09-143). The patients/participants provided their written informed consent to participate in this study.

J-HL performed literature search, the data analysis, and drafted the manuscript. SL and BC performed revised the manuscript. JY performed literature investigation and reviewing. JEL SK, and SN performed supervision and reviewing. JR designed the concept of article. All authors contributed to the article and approved the submitted version.

This study was supported by the Korea Breast Cancer Foundation (no. PHO020413) and the biostatistics team of the Statistics and Data Center, Research Institute for Future Medicine, Samsung Medical Center.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.