Discussion: With an extensive median follow-up of 20 years, this study reports an 8% rate of IBTR after treatment with breast-conserving surgery and radiotherapy. After excluding patients who had an inoperable recurrence or distant metastases at IBTR, the remaining 146 patients were definitively managed with either SM or SBCS. Ipsilateral breast tumor recurrence after breast-conserving surgery plus radiotherapy is associated with an increased risk of distant metastases and death compared with breast cancer patients who do not have a local recurrence. The relative risk of breast-cancer–associated death has been reported to increase by a factor of 3.4–4.6 after an in-breast relapse. The increase in mortality rate is likely related to the biology of disease, rather than a cause of distant metastases, although this remains a controversial viewpoint. Many patients do have favorable outcomes after an invasive local relapse. In our study the 10-year survival after IBTR was 64.4%, with a cause-specific survival of 71.2%. A large series from the University of Pennsylvania of local recurrence after breast conservation therapy found similarly encouraging long-term results with aggressive surgical salvage. The major cause of death in both series was distant metastatic spread from breast cancer.

Mastectomy is considered standard salvage therapy for a local relapse in patients previously treated with breast conservation therapy. However, a second local relapse in the form of a chest wall recurrence occurs in 3%–32% of patients treated with SM. Our series reported an 8% rate of chest wall recurrence after SM. It is important to note SM does not necessarily eradicate the risk of subsequent local relapse.

Mastectomy does, however, provide more complete pathologic information to tailor further therapy and assess the risk of disease progression. Review of mastectomy specimens from patients previously treated with breast-conserving surgery as part of protocol NSABP B-06 yielded a 14% incidence of multicentricity. Such patients in this pathologic study would not be suitable for SBCS. However, clinical findings on physical examination and modern imaging studies are more likely to identify patients who require mastectomy.

There are limited published series on the outcomes of patients treated with breast conservation surgery at the time of a local recurrence. In general, there is relatively short follow-up and a small number of patients studied. An early series by Kurtz  examined the feasibility of wide local excision for IBTR after previous breast conservation therapy. Sixteen (32%) of the 50 patients had second local relapses with a median follow-up of 51 months. Five-year local control rates were higher in patients with negative margins and a longer time interval to recurrence (>5 years). The Dutch Study Group on Local Recurrence after Breast Conservation reported a local recurrence rate of 38% with a median follow-up of 52 months after SBCS. There was no significant difference in the rate of second local relapse associated with SM or SBCS.

The largest series evaluating SBCS vs. SM for in-breast local relapse was reported by Salvadori  from Milan, Italy. This retrospective study evaluated 134 patients who underwent total mastectomy and 57 patients who received salvage local excision. With a median follow-up of 73 months after local relapse, the 5-year survival was 70% after SM and 85% after SBCS, and there was no difference in disease-free survival. Second local relapse was more common after salvage local excision (19% vs. 4%). This rate was lower than previous series and may be related to the small tumor size in the SBCS cohort (60% of the tumors were <1 cm). A recent series from Japan reported excellent preliminary outcomes after repeat lumpectomy for patients with a small recurrence, low histologic grade, and no lymphovascular invasion. Based on these studies and our reported outcomes, survival does not appear to be compromised after SBCS. Second local recurrence occurred in roughly one third of patients, although our series reported a lesser risk with lengthy follow-up after IBTR. This discrepancy may be related to the smaller median tumor size (<1 cm) in the SBCS cohort as well as the prolonged median time to local relapse in our series of 4.9 years.

Prognostic factors at IBTR have mainly focused on identifying patients at risk for metastases or mortality. Patients with an invasive breast recurrence have less favorable outcomes than patients with ductal carcinoma in situ histology at relapse (9). The interval from diagnosis to recurrence has been found to be a significant predictor of survival in several series. Clinical size of the recurrence as well as skin involvement has also been shown to be important predictors of outcome (1, 27). A large series from Voogd et al. (11) found the size of the recurrence, skin involvement, vascular invasion, and lymph node status of the primary to predict for distant metastases after local relapse. Our series utilized known pathologic factors from mastectomy specimens to assess for a localized relapse that could potentially be treated with SBCS. Two thirds of the patients had an IBTR with size <3 cm, ?3 positive lymph nodes, no lymphovascular invasion or skin extension, and concordant preoperative mammogram and physical examination. Early detection with mammography correlated with a localized recurrence, and conversely, multicentric disease at IBTR was associated with detection by physical examination. Favorable biology also predicted for a localized relapse, including ER positive at diagnosis and diploid at recurrence. Lymph node metastases at IBTR were predictive of multicentric disease. An important finding in our series is that half of the women with BRCA1/2 germline mutations were found to have multicentric disease at the time of mastectomy. Therefore, patients with genetic breast cancer may be less appropriate for salvage with breast conservation.

The addition of systemic therapy or focused radiotherapy to SBCS may further improve results. Systemic therapy at the time of IBTR may improve the ability to perform salvage conservation surgery. Tamoxifen after local relapse has been shown to improve disease-free survival compared with observation, and ER-positive tumors had a prolonged median disease-free survival by more than 4.5 years in a Phase III Swiss study. Reirradiation of the tumor bed after SBCS may further reduce the incidence of second local recurrence. Select series have delivered repeat radiotherapy using interstitial brachytherapy, although there is only one prospective study on reirradiation for IBTR. In this study from Vienna University, there were no local recurrences after partial breast irradiation with interstitial pulse dose rate brachytherapy to a dose of 40–50 Gy, and cosmetic outcomes were acceptable. We are currently in the process of developing a Phase I/II prospective multi-institutional trial to evaluate SBCS with three-dimensional conformal external beam partial breast reirradiation to the tumor bed. This study will prospectively evaluate the toxicity and local control rates of reirradiation to a partial breast field after repeat lumpectomy in a selected cohort of locally recurrent conservatively managed breast cancer patients.
Conclusion

This article presents information on long-term outcomes for a large series of women who underwent definitive surgical salvage at IBTR with either mastectomy or repeat breast-conserving surgery. There is potential for long-term survival after IBTR, highlighting the importance of salvage therapy. While mastectomy remains the standard of care for in-breast relapse after breast-conserving surgery and radiotherapy, SBCS appears feasible for select patients who have favorable tumor biology and early detection of their local relapse. Women with BRCA1/2 mutations may be less appropriate for SBCS owing to their propensity for multicentric disease in our series. Patients who opt for SBCS remain at continued risk for a second in-breast relapse, and strategies to minimize this risk warrant further investigation.