A prospective analysis of two studies that used the 5-mm interval slices and 5-mm margin-free method for ipsilateral breast tumor recurrence after breast-conserving surgery without radiotherapy

Breast-conserving surgery with radiotherapy is one of the standard treatments for early breast cancer. There were no differences in disease-free and overall survival between breast-conserving treatment with radiotherapy and mastectomy in randomized trials and meta-analysis [1‐8]. Radiotherapy was demonstrated to markedly reduce ipsilateral breast tumor recurrence (IBTR) in many randomized trials and meta-analysis [7, 9‐12]. However, radiotherapy is time-consuming and costly, and can cause severe treatment-related adverse events [13].

Randomized trials have been conducted to identify the population not requiring radiotherapy after breast-conserving surgery [14]. Accordingly, it has been regarded as an option to treat elderly patients with small hormone receptor-positive breast cancer by breast-conserving surgery and hormone therapy without radiotherapy. However, the 10-year IBTR rate in patients treated without radiotherapy was around 10% in the randomized trials. As such, a good method to reduce the IBTR rate without radiotherapy is needed. We therefore conducted two sequential prospective studies to examine the feasibility of breast-conserving surgery without radiotherapy using our 5-mm interval slice and 5-mm margin-free method since 2002.

Primary female breast cancer patients who fulfilled the following eligibility criteria were prospectively enrolled in two sequential studies named WORTH 1 and 2: (1) a tumor of 3 cm or less by palpation, (2) pathologically node negative by axillary dissection or sentinel node biopsy and M0, (3) no treatment before surgery, (4) 50 years of age or older at the time of surgery and postmenopausal, (5) no tumor cells within 5 mm from the margins histologically, (6) no lymphatic invasion around the primary cancer, (7) estrogen receptor-positive judged by each institution, and (8) within 8 weeks after definitive surgery. The surgical mammary gland materials must have been sliced in 5-mm intervals and all the slices must have been examined microscopically. The status of hormone receptors and HER2 was judged at each institution. We regarded HER2 in immunohistochemistory (IHC) 3 + , or 2 + and FISH amplification as positive. But we included HER2 IHC 2 + without the information of FISH in negative in Table 1. Postoperative radiotherapy was prohibited, but adjuvant chemotherapy was allowed. The patients were administered tamoxifen or anastrozole in WORTH 1 and anastrozole in WORTH 2 as postoperative adjuvant endocrine therapy for 5 years. The exclusion criteria were: (1) suspected multifocality of the tumor, (2) history of breast cancer or bilateral breast cancer, or (3) psychological disease.

IBTR-free survival and distant relapse-free survival (DRFS) were recorded as the interval from initial definitive surgery until IBTR or distant relapse, respectively. Only IBTR that occurred as the first event was regarded as events of IBTR-free survival. IBTR that developed at the same time as other relapses or within 2 months of the time of other relapses was regarded as an event of IBTR-free survival. Patients who did not develop IBTR or distant relapse were censored at the time of last follow-up or death as the IBTR-free rate.

The primary end point was IBTR-free survival, and secondary end points were DRFS and overall survival.

As safety monitoring, interim monitoring by Bayesian approach was carried out. As stopping criteria based on the posterior distribution, WORTH 1 was planned to stop if the posterior probability exceeding 1% in the annual IBTR rate exceeded 95%. WORTH 2 adopted a criterion of 0.5% in the annual IBRT rate. The determination of the sample size for WORTH 1 can be shown with an accuracy such that the upper limit of the 95% confidence interval of the estimated IBRT rate is about 10% when the 5-year IBRT rate is assumed to be 5%. In WORTH 2, assuming that the 5-year IBRT rate is 2.5%, the accuracy that the upper limit is within about 5% is obtained. The sample size for WORTH 1 was set 120 and that for WORTH 2 was 200.

The protocols were approved by the ethics committee at each participating institution.

Written informed consent was received from all participants.

Registration of WORTH 1 was not required in 2002 when it was started. WORTH 2 was registered in UMIN-CTR, numbered UMIN000000534, on December 10, 2006.

The number of patients who participated in the two trials was 123 and 198 in WORTH 1 and WORTH 2, respectively. The characteristics of patients, and their tumors and treatments are shown in Table 1. Almost all patients were HER2 negative and others in histological types in Table 1 had invasive cancers other than invasive ductal carcinoma.

Breast-conserving surgery with radiotherapy to the breast is one of the standard local treatments for early breast cancer because randomized trials and meta-analyses demonstrated it to have equivalent effects to mastectomy in terms of disease-free and overall survival [1‐8]. Radiotherapy is regarded as an essential component of breast-conserving treatment because it markedly reduced local recurrence in the randomized trials [7, 9‐11], and meta-analysis [12] demonstrated that radiotherapy significantly reduced not only locoregional recurrence, but also distant recurrence and breast cancer deaths. However, pathologically negative node, old age, and positive hormonal receptor status are known to reduce local recurrence rates. Therefore, randomized trials comparing radiotherapy with no radiotherapy after breast-conserving surgery have been conducted for elderly patients with node-negative and estrogen receptor-positive breast cancer treated using adjuvant hormonal agents [14]. Radiotherapy reduced the risk of IBTR, but it did not impact distant recurrence or overall survival of early breast cancer treated by breast-conserving surgery and tamoxifen in elderly women (70 years or older) in CALGB 9343. The IBTR rate was 9% at 10 years in the no radiotherapy group. McCormick et al. reported that 88% of women aged 70 years and older with stage I, estrogen receptor-positive breast cancer received radiotherapy after breast-conserving surgery, even after the National Comprehensive Cancer Network (NCCN) guidelines for elderly breast cancer patients stated the omission of radiotherapy after breast-conserving surgery in women meeting the CALGB 9343 entry criteria who were prescribed hormonal agents in 2009 [15]. Two other groups also reported low IBTR rates in 65-year-old or older breast cancer patients [16, 17]. One study is a randomized trial of PRIME II. This trial randomly assigned women with hormone receptor-positive and node-negative breast cancer of 3 cm or less and negative excision margins (1 mm or more) into either whole-breast radiotherapy or no radiotherapy. At median follow-up of 5 years, actuarial ipsilateral breast cancer recurrence was 1.3% in women allocated whole-breast radiotherapy and 4.1% in those assigned no radiotherapy. The other study demonstrated a very low rate of IBTR after breast-conserving surgery without radiotherapy [17]. This was a single arm study. They reported an IBTR rate of 1.2% at 5 years in 601 participants with hormonal agents. They used the criteria of T1 tumor, node-negative, ER-positive, and Elston–Ellis histological grade 1 or 2. However, they used sector resection of the tumor, the area of which was larger than that of local excision that we used.

Our first study, WORTH 1, was started to identify a population with breast cancer who had a low risk of IBTR after breast-conserving surgery without radiotherapy shortly after the first reports of the study of Fyles et al. [18] and the CALGB 9343 study [19] in 2001. We included the “5-mm-thick slice and 5-mm free-margin” method to reduce the IBTR rate, and “no lymphatic invasion around the tumor”, which was a possible factor to avoid inflammatory-type IBTR, in the eligibility criteria. After WORTH 1 was started, the IBTR rate was low. Therefore, we started the second study, WORTH 2, with the same eligibility criteria as WORTH 1 to ensure low IBTR rates.

The 5-year IBTR-free rates were 94.8%, 98.4%, and 97.0% in WORTH 1, WORTH 2, and in combination, respectively. The rate was lower in WORTH 1 than in WORTH 2 even though we used the same eligibility criteria. One reason for the difference is the hormonal agents used. The most frequently used agent was tamoxifen in WORTH 1, but it was anastrozole in WORTH 2. Aromatase inhibitors, including anastrozole, are more effective at reducing hormone receptor-positive breast cancer recurrence [20]. The second possible reason was the difference in PR-negative rates. Thirty-seven of 123 patients were PR negative in WORTH 1, whereas 27 of 198 were PR negative in WORTH 2 (30.1% vs. 13.6%, P = 0.0003 by chi-square test). The last possible factor was MRI use before the surgery. Recently, we often use MRI to make sure that multifocality or multicentricity is not present before the surgery in Japan, but we did not collect the data of MRI use.

The “5-mm interval slice and 5-mm margin-free” method were used in both WORTH 1 and 2. This method had been used to judge pathologically negative margins in Japan. This method is labor- and time-consuming, but effective in reducing the IBTR rate because the combined 5-year IBTR-free rate was 97.0%, which was high.

Two factors, age at the time of surgery and PR status, were identified as independent prognostic factors for the IBTR rate in this combined analysis. The IBTR-free rate of patients aged 65 years or older at the time of surgery with PR-positive tumors was 98.4% at both 5 and 10 years. Wickberg et al. reported that luminal A tumors have a lower IBTR rate than luminal B tumors after breast-conserving surgery with or without radiotherapy [21]. Based on this favorable rate, we concluded that patients aged 65 years or older with both estrogen and progesterone receptor-positive breast cancer of 3 cm or smaller by palpation, who underwent breast-conserving surgery, had histologically negative margins judged by the “5-mm interval slice and 5-mm margin-free” method and no lymphatic invasion around the tumor do not need radiotherapy.

The strong points of our studies were the relatively large sample size and prospective design, but there were some limitations. First, our studies were not randomized. Therefore, we cannot measure the effects of radiotherapy. However, this was not the purpose of our studies. Second, it is unclear whether our “5-mm interval slice and 5-mm margin-free” method is effective for reducing the IBTR rate. ASCO recommends the criteria of no tumor cells on inked margins for histologically negative margins [22]. However, ASCO guidelines only consider patients who will receive radiotherapy. Our “5-mm interval slice and 5-mm margin-free” method may be useful to select patients who do not need radiotherapy. The previously mentioned study reported a low IBTR rate of 1.2% at 5 years [17]. They enrolled patients with PR-positive breast cancer, comprising 89.1% of the cohort and used sector resection of the breast. The area of sector resection is larger than that of local excision which we used. So, we believe that our “5-mm interval slice and 5-mm margin-free” method was effective to reduce IBTR in local excision. Third, the IBTR rate of WORTH2 was lower than that of WORTH1. The reasons for this were thought to be the higher rates of PR positivity and use of aromatase inhibitors in WORTH2, but there might have been a higher rate of the use of MRI before surgery in WORTH2. MRI might have detected the multiple cancers in the breast. We did not collect the data on MRI use. Therefore, this might have been one of the reasons. Lastly, the percentages of ER and/or PR-positive tumor cells in the tumors were not obtainedd. However, patients with higher percentages of these receptors may have been enrolled in the studies.

Our “5-mm-thick slice and 5-mm margin-free” method may be effective for selecting patients who can be treated by breast-conserving surgery and hormone therapy without radiotherapy. Patients who fulfill our eligibility criteria for WORTH 1 and 2 and HER2 negative, and aged 65 years or older at the time of surgery with PR-positive tumors, may not need radiotherapy if they receive hormonal agents.