Over the past 20 years, the combination of breast-conservation surgery and radiation therapy has become the most common treatment for the majority of patients with invasive and in situ breast malignancies. Extensive data from both randomized trials[1-3] and large clinical databases have shown that this approach is equal to mastectomy in terms of survival and provides excellent quality of life and patient acceptance. Nevertheless 10% to 15% of patients treated with breast-conservation surgery and radiation will have a local recurrence within the ipsilateral breast (IBTR) within 10 years.[4-8] This paper will review the biology, clinical management, and outcome of patients with ipsilateral breast recurrence. Incidence Each year in the United States, there are about 200,000 new cases of invasive breast cancer. In addition, due to the widespread utilization of screening mammography, the incidence of ductal carcinoma in situ (DCIS) has increased dramatically over the past 20 years, and now there are 40,000 to 50,000 new cases of this entity diagnosed annually in the United States. If only half of these patients are treated with breast-conservation therapy, and the local recurrence rate is only 10%, that still amounts to almost 12,000 women in the United States who develop ipsilateral breast recurrence each year. That is greater than the number of annual new cases of Hodgkin's disease or testicular cancer. Despite these numbers, there have been few good prospective studies and no randomized trials addressing treatment options for these patients. Table 1 shows the ipsilateral breast recurrence rate for a variety of studies. In the classic National Surgical Adjuvant Breast and Bowel Project (NSABP) B-06 trial, the recurrence rate at 20 years was 39% for the group that underwent lumpectomy without irradiation and 14% for the group that received radiation.[1] In the major prospective randomized trial of breast conservation from Italy, the breast recurrence rate for the group who underwent quadrantectomy and radiation therapy was 8.8% at 20 years.[3] The rate of recurrence over time in the NSABP study is shown in Figure 1. In the group without radiation therapy, the annual hazard rate for IBTR was 8.5% for years 1 to 3 and then decreased in subsequent years, whereas in the group that received breast irradiation, the hazard rate remained about 1% annually throughout the follow- up period. Overall, the results of both prospective randomized trials and retrospective studies are remarkably consistent. The IBTR rate for patients who received partial mastectomy and irradiation is about 5% to 10% at 5 years, 10% to 15% at 10 years, and 15% to 20% at 15 years. Risk Factors for IBTR Several factors have been associated with an increased risk of IBTR. Most studies have found that young age, positive microscopic margins, gross multifocality, an extensive intraductal component (EIC), and lymphatic vessel invasion are associated with a higher risk of IBTR.[5,9-17] In the Yale series, the IBTR rate was 18% at 5 years and 30% at 10 years for women under 40, compared to 4% and 12%, respectively, for women over 40.[14] Similarly, a large European study[16] found the short-term local recurrence rate to be 10.6% for women under 40, compared to 3.7% for older women. The reason that young age is associated with local recurrence is not well understood. The association of young age with local relapse is not related to the primary tumor size, nodal status, receptor status, or other factors commonly associated with metastatic potential. EIC and Surgical Margin
Many studies have found gross multifocality (ie, the presence of more than one tumor on physical exam or mammogram) or microscopic EIC (ie, intraductal carcinoma occupying more than 25% of the tumor area and extending beyond the edge of the invasive tumor) to be risk factors for IBTR.[6,9,17] An EIC has been found to be a risk factor for local recurrence following breast-conservation surgery but not following mastectomy, and is not a risk factor for the development of distant disease.[9] The presence of EIC, however, is not a contraindication to breast conservation, as its importance is closely related to the extent of surgical margin. The definition of a satisfactory surgical margin has been somewhat controversial.[ 18] Almost all studies agree that if there is actually tumor at the inked margin, the IBTR rate is two to three times higher than if the margin is negative.[5,11,13,15,19,20] However the significance of a "close" margin varies in different series. It appears that in older patients with purely invasive tumors, just a negative margin is sufficient, whereas in younger women, or those with EIC, a wider negative margin results in a lower incidence of local recurrence. Convincing data suggest that for pure DCIS, the recurrence rate is inversely proportional to the extent of the margin up to 1 cm.[21] Treatment
Treatment also affects the IBTR rate. All studies agree that radiation therapy dramatically reduces the risk of recurrence in the breast.[1,5,6,22] However, it is now recognized that systemic adjuvant hormonal therapy or chemotherapy also results in a lower incidence of IBTR.[1,9,12,19,20] The increased use of systemic adjuvant therapy in recent years is resulting in a somewhat lower incidence of IBTR than was seen in the initial trials. Risk Model
Freedman et al[12] recently used an interesting recursive partitioning model to identify patients at high and low risk for IBTR. This methodology uses decision trees to analyze subgroups of patients at significantly different risk. The first split implemented was age, with women over 55 having different risk factors than women under 55. For women over 55, the only other significant risk factor was tamoxifen(Drug information on tamoxifen) use. Women over 55 on tamoxifen had a 10-year IBTR rate of 2%, compared to 5% for those not on tamoxifen. For women under 55, the presence of EIC, margin positivity, tamoxifen use, and age (< 35 vs 36 to 55) all contributed to the definition of risk. For example, a woman under 35 with an EIC-negative tumor had a 10-year IBTR rate of 3% if the margin was negative, compared to 34% if the margin was positive. A woman aged 36 to 55 had a 10-year IBTR rate of 5% with tamoxifen, compared to 20% without tamoxifen. True Recurrence vs New Primary It recently became clear that a significant portion of patients who experience IBTR following conservative surgery and radiation therapy actually have new primary tumors as opposed to true local recurrences.[9,23,24] At Yale, Smith et al[24] defined the second breast tumor as a new primary if it was distinctly different from the original tumor with respect to histologic subtype, if it was in a different location in the breast, or if flow cytometry changed from aneuploid to diploid. The time interval between the original primary and the second tumor was considerably greater for new primaries compared to true recurrences (7.3 vs 3.7 years, P < .0001). Furthermore, 10-year overall survival (75% vs 55%, P < .0001) and distant disease-free survival (85% vs 41%, P < .0001) were much better for patients with new primaries compared to those with true recurrences. Huang et al[23] at M. D. Anderson also found that new primaries tended to occur later than true recurrences, and that overall survival and distant disease-free survival were much better for new primaries than true recurrences. In addition, they found that patients with new primaries were significantly less likely to develop a second recurrence following salvage therapy (2% vs 18%, P = .008). Interestingly, however, patients with new primaries were more likely to develop a new contralateral breast carcinoma (29% vs 8%, P = .004). Therefore, the diagnosis of a new primary as opposed to a true recurrence clearly implies a different natural history and prognosis, and has different implications for therapeutic management. Unfortunately, most series addressing IBTR do not adequately distinguish between the two. Differentiating between a new primary and a true recurrence is particularly valuable in understanding the biology of breast cancer in young women with BRCA1/2 mutations. Although early results have indicated acceptable results with breast conservation in this population, more prolonged follow-up has revealed a continued increased incidence of IBTR due to new primaries, as well as an increased incidence of new contralateral breast cancers.[25] Further study of this group using chemoprophylaxis or bilateral mastectomy is warranted. Treatment Options for IBTR Mastectomy
Most patients with IBTR have undergone salvage therapy with mastectomy, and this has been considered the standard treatment.[26-30] Mastectomy generally results in long-term local control rates of 85% to 95% and 10-year overall survival of 60% to 70%. In an early study of mastectomy for IBTR, Fowble et al[28] achieved local control in 95% of patients; at 5 years, the overall survival rate was 84% and distant disease-free survival was 59%. The authors could not identify any factors that would predict for no disease in the mastectomy specimen, so they recommended mastectomy as the preferred treatment. In another early study, Kurtz et al[31] noted locoregional control in 88% at 5 years following salvage mastectomy, and in only 64% after further breast-conserving salvage procedures, although the type of salvage operation did not affect survival. In more recent reports, Dalberg et al[32,33] studied the incidence of uncontrolled local disease, defined as the appearance of cancer in the remaining breast or chest wall that could not be eradicated within 3 months by further treatment. The cumulative incidence of uncontrolled local disease at 5 years following salvage mastectomy was 12%, compared to 33% after salvage reexcision. Development of uncontrolled local disease was a very grave prognostic sign: Indeed, 78% of such patients died of disseminated breast cancer, compared to only 10% of patients with IBTR who were salvaged without further local recurrence. Certainly this is one of the strongest arguments for using mastectomy to salvage patients with IBTR. Reconstruction
If mastectomy is necessary, it would certainly be helpful to be able to offer patients breast reconstruction. Unfortunately, there have been only a few studies of breast reconstruction following IBTR. Forman et al[34] reported on 10 breast reconstructions using tissue expanders and implants after salvage mastectomy for IBTR. Unfortunately one implant extruded, another became infected, two did not expand because the irradiated tissue lacked compliance, and two more developed severe capsular contractures. Overall, 60% of the patients had a complication or an unfavorable result. Most plastic surgeons probably prefer autogenous tissue reconstruction in patients who have received breast irradiation. However even pedicled transverse rectus abdominis myocutaneous (TRAM) flaps have been reported to have complication rates of up to 33% in irradiated tissue beds. Moran et al[35] reported on 14 patients who underwent free TRAM flaps with anastomosis to the thorocodorsal vessels. Their complication rate was only 14%, and the aesthetic result was rated as excellent. Repeat Lumpectomy With or Without Additional Radiation
Although mastectomy has been considered the standard therapy for IBTR, there is increasing interest in defining a patient population for whom breast preservation may be appropriate. In a large study of 197 patients with operable IBTR, Salvadori et al[36] treated 70% with mastectomy and 30% with further local resection. A second IBTR was more common at 5 years in the reexcision group (19% vs 4%), but there was no difference in disease-free survival. Because the type of surgery did not affect survival, the authors concluded that breast conservation can be considered in selected patients with IBTR. Komoike et al[37] treated 30 IBTR patients with repeat lumpectomy. Nine of these patients developed a second local relapse, but those who did not tended to have small tumors (< 1 cm) with a low histologic grade and no lymphatic invasion. In contrast, those who failed this therapy tended to be young (age < 35), to have positive family histories, and to not receive adjuvant systemic treatment. The literature contains some experience with a second course of radiation therapy following local resection of an IBTR. Deutsch et al[38] treated 39 women who had had repeat lumpectomy for IBTR with 5,000 cGy to the operative area using electrons. The treatment was well tolerated, and the only late sequelae were skin pigmentation changes. Nine (23%) of these patients developed a second IBTR. The overall and disease-free survival of the entire group at 5 years was 78% and 68%, respectively. Resch et al[39] treated 17 IBTR patients with pulse dose-rate brachytherapy following repeat lumpectomy. Twelve did not develop further breast recurrence, and side effects were limited to moderate (grade 1/2) fibrosis. Both authors concluded that repeat radiation therapy may be an acceptable alternative to mastectomy for selected patients. Systemic Adjuvant Therapy
IBTR is a strong independent predictor for the subsequent development of distant disease. For that reason, several authors have recommended adjuvant systemic therapy at the time of IBTR,[40,41] and many oncologists use this strategy based on extrapolation of the data for adjuvant treatment of primary breast cancer. Unfortunately there are no published prospective trials addressing the issue.[42] There is one prospective randomized trial of hormone therapy in 167 patients with chest wall recurrence following mastectomy.[ 43] Tamoxifen was found to significantly increase 5-year diseasefree survival from 36% to 59% compared to placebo, but there was no improvement in overall survival. One recent retrospective study of IBTR following lumpectomy found a benefit for ovarian suppression and chemotherapy at the time of IBTR in premenopausal patients but not in postmenopausal patients.[44] Clearly more studies are needed. The International Breast CancerStudy Group recently initiated a randomized trial of adjuvant chemotherapy following IBTR, and the NSABP plans to collaborate and open the trial in the United States.[45] This would seem to address an important clinical question, and hopefully answers will be forthcoming. Outcome of Treatment of IBTR Table 2 shows the overall survival and disease-free survival following treatment of IBTR in several series. The overall survival averages 74% at 5 years and 65% at 10 years; and the disease-free survival averages 60% at 5 years and 57% at 10 years. When these data first became available more than a decade ago, it was impressive that the prognosis for patients following IBTR was so favorable, particularly when compared to that of patients who developed chest wall recurrence following mastectomy. The overall 5-year survival for patients with chest wall recurrence averaged 35%, with disease-free survival of 30%. Clearly, salvage following IBTR produced survival rates that were about twice as high.[5] Therefore, it became accepted that patients need to be followed closely following lumpectomy, so that if IBTR developed, it could be detected early and the patient would still have a good chance of cure with salvage therapy. When considering survival for patients with IBTR, it is important to note whether that parameter is measured from the date of the IBTR, or from the date of diagnosis of the original tumor. Although, as mentioned above, the survival measured from the date of IBTR is relatively good, if we examine survival from the date of original diagnosis in patients who do or do not develop IBTR, it is clear that IBTR is a strong independent risk factor for the development of distant metastases. Most studies have found that the relative risk of distant metastases with IBTR ranges from three- to fivefold.[ 19,20,40,46-50] For example, in the NSABP B-06 study,[40] the risk of distant disease was 3.41 times greater in patients who developed an IBTR, even when adjusted in multivariate analysis for all other prognostic variables, such as patient age, nodal status, nuclear grade, tumor type, and maximum pathologic tumor size. IBTR is actually one of the strongest prognostic variables available to predict distant metastases or death from breast cancer. The interpretation of this association, however, has varied widely.