Discussion
In this study, we extended our previous conclusions to patients who received NAC and BCS. Atypical hyperplasia at the margins also did not confer worse local ipsilateral breast recurrence-free survival, DMFS or OS in patients underwent BCS after NAC, and the observed IBTR rate was low (16/323, 5.0%).
Data from a Netherlands study showed higher rates of tumor-involved margins in patients treated with NAC and BCS versus primary BCS (23 vs. 10%)2. Consistent with this literature, we found a high re-excision rate 30.9% (100/323) in our study. The rate of tumor at margins requiring intraoperative re-excision was actually quite low (37 patients, 11.4%). Although 63 patients received intraoperative re-excision due to severe AH appeared at primary resection margins, the local recurrence rate of patients with severe AH was not higher than the patients without AH. In addition, Coopey et al . reported that patients with ADH and severe ADH were equally likely to develop invasive cancer and DCIS16. This indicates that if intraoperative frozen pathology can be accurately distinguish severe AH from DCIS, then there is no need to further resect severe AH at the margins. Our findings therefore lend further support for the oncology safety of residual severe AH at the margins to minimize unnecessary re-excisions.
AH of the breast refers to spectrum of abnormal epithelial proliferative lesions that does not qualify as carcinoma in situ. AH is found in approximately 10% of breast biopsies with benign lesions, and it is classified as a high-risk precancerous lesion due to its association with, and potential to progress to breast cancer18. Women with AH have an approximate four-fold increase in BC risk6,7. Why AH did not confer higher IBTR in the present study, the following reasons may address to this issue. First, the conclusion that AH contributes to higher risk of breast cancer was drawn from people with benign diseases but not patients who have already suffered from breast cancer. The tumor burden of breast cancer patients was significantly different from those with only benign disease. Holland et al.reported in their study, of the 282 patients with invasive cancers that underwent lumpectomy, 105 (37%) showed no other tumor foci in the residual breast, but other tumor foci were found in the remaining 177 breasts, of which 20% tumor foci were present within 2 cm off the reference tumor in the residual breast19. Therefore, it is clearly a negative margin that does not indicate that there is no residual cancer in the breast, however, the residual tumor can be controlled by radiotherapy and adjuvant treatments such as endocrine therapy and targeted therapy. The risk of IBTR of breast conserving therapy is about 0.5–2% per year with an increased risk during the first few years20,21, and the median time to IBTR is 36 months22. Additionally, results from NSBAPB-18 suggested that patients downstaged after NAC for BCS may have higher local recurrence rate23. However, the annual incidence of breast cancer caused by AH is not significantly higher than the risk of IBTR after BCS. Recent reports with long-term follow-up have demonstrated that absolute risk for developing breast cancer is in the range of 1-2% per year7,24. More recently, Menes and colleagues’ research which was large and contemporary, found that the 10 years cumulative risk for developing breast cancer in women with atypical ductal hyperplasia (ADH) is only 5.6%25. Therefore, the effect of AH on IBTR may be offset by the breast cancer itself. Thus, it is not surprising that AH involved at the margin does not contribute to a higher risk of IBTR. Second, endocrine therapy can reduce IBTR. Data from the NSABP P-1 trial showed that tamoxifen administered for 5 years decreased the risk of invasive and non-invasive breast cancer by approximately 50%26. In the present study over 70% of the patients received endocrine therapy. We postulate that this result is attributable in part to the use of endocrine therapy. Third, the follow-up period of our study may not be long enough to reveal the effectiveness of the difference between groups. Page et al. reported that the interval of ADH and ALH to breast cancer is 8.2 and 11.9 years27, but the median follow-up period of our study was approximately 48 months.
In addition to margin status, other important determinants of overall prognosis including the pCR rate, are being recognized. It has been reported that different chemotherapy responses and residual tumor burdens have different risks of cancer relapse, since patients who achieved a pCR will have favorable DFS and OS28,29. In the present study, patients who achieved a pCR had more excellent DMFS compared with those with residual lesions after NAC in our study. In patients undergoing NAC, the lack of an axillary pCR has been shown to be associated with less favorable overall survival30-32. Consistent with these published studies, we also found that positive lymph nodes post-NAC was associated with inferior OS.
Our study has some potential limitations due to its retrospective design. The chemotherapy regimens given to the patients varied among different subtypes, and this may affect the relationship between NAC response and DFS/OS. Furthermore, the number of patients enrolled in the study was limited. Future multi-centered studies and more patients enrolled with a longer follow-up are needed to achieve greater statistical power so as to address this issue more clearly.
In summary, our study did not find a higher risk of IBTR or distant metastasis or death events in patients with AH involved at the surgical margin who underwent BCS after NAC. In multivariate analysis, there was no association between AH status and IBTR, DMFS or OS. The lack of a pCR, and positive pathologic lymph nodes post-NAC were associated with inferior outcomes. Although further studies are needed to determine how AH status is related to IBTR in this population, the excellent long-term outcomes demonstrated in patients with AH at margins following NAC in this study suggest that AH involved at the margins may be acceptable in appropriately selected patients with breast cancer.