A custom three-dimensional electron bolus technique for optimization of postmastectomy irradiation.

PURPOSE: Postmastectomy irradiation (PMI) is a technically complex treatment requiring consideration of the primary tumor location, possible risk of internal mammary node involvement, varying chest wall thicknesses secondary to surgical defects or body habitus, and risk of damaging normal underlying structures. In this report, we describe the application of a customized three-dimensional (3D) electron bolus technique for delivering PMI. METHODS AND MATERIALS: A customized electron bolus was designed using a 3D planning system. Computed tomography (CT) images of each patient were obtained in treatment position and the volume to be treated was identified. The distal surface of the wax bolus matched the skin surface, and the proximal surface was designed to conform to the 90% isodose surface to the distal surface of the planning target volume (PTV). Dose was calculated with a pencil-beam algorithm correcting for patient heterogeneity. The bolus was then fabricated from modeling wax using a computer-controlled milling device. To aid in quality assurance, CT images with the bolus in place were generated and the dose distribution was computed using these images. RESULTS: This technique optimized the dose distribution while minimizing irradiation of normal tissues. The use of a single anterior field eliminated field junction sites. Two patients who benefited from this option are described: one with altered chest wall geometry (congenital pectus excavatum), and one with recurrent disease in the medial chest wall and internal mammary chain (IMC) area. CONCLUSION: The use of custom 3D electron bolus for PMI is an effective method for optimizing dose delivery. The radiation dose distribution is highly conformal, dose heterogeneity is reduced compared to standard techniques in certain suboptimal settings, and excellent immediate outcome is obtained.

Factors predictive of outcome in patients with breast cancer refractory to neoadjuvant chemotherapy.

PURPOSE: The purpose of this study was to determine the clinical, pathological, and treatment factors that are predictive of local-regional recurrence and overall survival for patients with breast cancer that is refractory to neoadjuvant chemotherapy. PATIENTS AND METHODS: This study analyzed the data of the 177 breast cancer patients treated on our institutional protocols who had less than a partial response to neoadjuvant chemotherapy. The initial clinical stage of disease was II in 27%, III in 69%, and IV (supraclavicular lymph node involvement) in 4%. Surgery was performed in 94% of the patients, and 77% of these patients also received adjuvant chemotherapy. RESULTS: After a median follow-up of 5.2 years, 106 patients experienced disease recurrence, with 98 of these having distant metastases and 45 having local-regional recurrence. The 5- and 10-year overall survivals for the entire group were 56% and 33%, respectively. The factors that were independently associated with a statistically significant poorer overall survival in a Cox regression analysis were pathologically involved lymph nodes after surgery, estrogen receptor-negative disease, and progressive disease during neoadjuvant chemotherapy. The 5-year overall survival for patients with pathologically negative lymph nodes ranged from 84% (estrogen receptor-positive disease) to 75% (estrogen re-ceptor-negative disease), compared with rates for patients with pathologically positive lymph nodes of 66% (estrogen receptor-positive disease) and 40% (estrogen receptor-negative disease). The 5-year survival of patients with progressive disease was only 19%. The 5- and 10-year local-regional recurrence rates for the 177 patients were 27% and 34%, respectively. Significant factors on Cox analysis that predicted for local-regional recurrence were four or more pathologically involved lymph nodes and estrogen receptor-negative disease. For the 105 patients treated with surgery and postoperative radiation therapy, the 10-year local-regional recurrence rates for the subgroups with 0, 1, or 2 of these factors were 12%, 25%, and 44%, respectively. CONCLUSIONS: For patients with a poor response to neoadjuvant chemotherapy, conventional treatments achieve reasonable outcomes in those with lymph node-negative disease or estrogen receptor-positive disease. However, more active systemic and local therapies are needed for patients with estrogen receptor-negative disease and positive lymph nodes and for those with clinical evidence of progressive disease during neoadjuvant chemotherapy.

Relationship of sentinel and axillary level I-II lymph nodes to tangential fields used in breast irradiation.

PURPOSE: To evaluate the volume of nodal irradiation associated with breast-conserving therapy, we defined the anatomic relationship of sentinel lymph nodes and axillary level I and II lymph nodes in patients receiving tangential breast irradiation. METHODS AND MATERIALS: A retrospective analysis of 65 simulation fields in women with breast cancer treated with sentinel lymph node surgery and 39 women in whom radiopaque clips demarcated the extent of axillary lymph node dissection was performed. We measured the relationship of the surgical clips to the anatomic landmarks and calculated the percentage of prescribed dose delivered to the sentinel lymph node region. RESULTS: A cranial field edge 2.0 cm below the humeral head the sentinel lymph node region was included or at the field edge in 95% of the cases and the entire extent of axillary I and II dissection in 43% of the axillary dissection cases. In the remaining 57%, this field border encompassed an average of 80% of cranial/caudal extent of axillary level I and II dissection. In 98.5% of the cases, all sentinel lymph nodes were anterior to the deep field edge and 71% were anterior to the chest wall-interface, whereas 61% of the axillary dissection cohort had extension deep to the chest wall-lung interface. If the deep field edge had been set 2 cm below the chest wall-lung interface, the entire axillary dissection would have been included in 82% of the cases, and the entire sentinel lymph node would have been covered with a 0.5-cm margin. The median dose to the sentinel lymph node region was 98% of the prescribed dose. CONCLUSIONS: By extending the cranial border to 2 cm below the humeral head and 2 cm deep to the chest wall-lung interface, the radiotherapy fields used to treat the breast can include the sentinel lymph node region and most of axillary levels I and II.

The influence of pathologic tumor characteristics on locoregional recurrence rates following mastectomy.

PURPOSE: The objective of this study was to evaluate the influence of pathologic factors other than tumor size and number of involved axillary nodes on the risk of locoregional recurrence (LRR) following mastectomy. PATIENTS AND METHODS: We reviewed the medical records of 1031 patients treated with mastectomy and doxorubicin-based chemotherapy without radiation on 5 prospective clinical trials. Median follow-up was 116 months (range, 6-262 months). RESULTS: Patients with gross multicentric disease were at increased risk of LRR (37% at 10 years). However, patients with multifocal disease and those with microscopic multicentric disease did not experience higher rates of LRR than those with single lesions (17% at 10 years). Patients with lymph-vascular space invasion (LVSI) or involvement of the skin or nipple also experienced high rates of LRR (25%, 32%, and 50%, respectively). The presence of close (<5 mm) or positive margins was associated with an increased risk of LRR (45%). The increased risk of LRR observed for patients with pectoral fascial invasion (33%) was not reduced when negative deep margins were obtained. On multivariate analysis, the presence of 4 or more involved axillary nodes, tumor size of greater than 5 cm, close or positive surgical margins, and gross multicentric disease were found to be independent predictors of LRR (all, p < 0.01). In a separate analysis including only patients with 1-3 involved axillary nodes, microscopic invasion of the skin or nipple, pectoral fascial invasion, and the presence of close or positive margins were significant predictors of LRR. CONCLUSION: In addition to the extent of primary and nodal disease, other factors that predict for high rates of LRR include the presence of LVSI, involvement of the skin, nipple or pectoral fascia, close or positive margins, or gross multicentric disease. These factors predict for high LRR rates regardless of the number of involved axillary nodes.

Reduction mammoplasty improves breast conservation therapy in patients with macromastia.

BACKGROUND: Macromastia has been considered a contraindication to breast conservation therapy because of difficulties with radiation therapy. This study evaluates the feasibility of bilateral reduction mammoplasty as a component of breast conservation therapy for breast cancer patients with pendulous breasts. METHODS: Of 153 patients undergoing reduction mammoplasty at the University of Texas M. D. Anderson Cancer Center, 28 were identified as breast cancer patients with macromastia receiving breast conservation therapy. Median follow-up was 23.8 months. RESULTS: Median patient age was 55 years. Nearly all patients were described as obese. Median weight of the reduction mammoplasty specimen on the cancerous side was 766 g. One patient (4%) required completion mastectomy for inadequate margin control. Major postoperative complications occurred in 2 patients (7%). There were no major postradiation complications. Patient survey revealed a satisfaction rate of 86%. CONCLUSION: Bilateral reduction mammoplasty is a reasonable and safe option for breast cancer patients with macromastia who desire breast conservation therapy.

Recursive partitioning analysis of locoregional recurrence patterns following mastectomy: implications for adjuvant irradiation.

PURPOSE: Postmastectomy irradiation improves overall survival for breast cancer patients at high risk for locoregional recurrence (LRR). The objective of this study was to use recursive partitioning analysis (RPA) to define patient subgroups at high risk for LRR following mastectomy. PATIENTS AND METHODS: A cohort of 1031 patients treated on prospective trials with mastectomy and doxorubicin-based chemotherapy without irradiation was analyzed. The variables considered in the RPA were tumor size, number of involved nodes, number of nodes examined, and percentage of nodes involved (nodes involved/nodes examined). The endpoint was LRR +/- distant metastasis. Only patients with complete data were analyzed (n = 913). Median follow-up was 8 years (range, 0.7-22 years). RESULTS: Involvement of 20% or more of the lymph nodes examined was the most significant variable predicting LRR. Three risk categories were defined. Patients with 20% or more involved nodes and tumors of 3.5 cm or more were at greatest risk for LRR (41% at 8 years). An intermediate-risk group included patients with 20% or more involved nodes and tumors of less than 3.5 cm as well as those with less than 20% involved nodes and tumor size of 5 cm or greater (18% at 8 years). Patients with less than 20% involved nodes and tumor size of less than 5 cm were at lowest risk for LRR (10% at 8 years). CONCLUSION: Tumor size and extent of nodal involvement play interrelated roles in predicting LRR following mastectomy and systemic therapy. Patients with 20% or greater involved nodes and those with less than 20% nodes and tumors of 5.0 cm or greater are at significant risk of LRR and should be considered for postoperative irradiation.

Impact of systemic treatment on local control for patients with lymph node-negative breast cancer treated with breast-conservation therapy.

PURPOSE: To determine the impact of tamoxifen and chemotherapy on local control for breast cancer patients treated with breast-conservation therapy. PATIENTS AND METHODS: The data from 484 breast cancer patients who were treated with breast-conserving surgery and radiation were analyzed. Only patients with lymph node-negative disease were studied to provide comparative groups with a similar stage of disease and a similar competing risk for distant metastases. Actuarial local control rates of the 277 patients treated with systemic therapy (128, chemotherapy with or without tamoxifen; 149, tamoxifen alone) were compared with the rates for the 207 patients who received no systemic treatment. Only 10% of the patients had positive (2%), close (3%), or unknown margin status (5%). RESULTS: Patients treated with systemic therapy had improved 5-year (97.5% v 89.8%) and 8-year (95.6% v 85.2%) local control rates compared with those that did not receive systemic treatment (P =.004, log-rank test). There was no statistical difference in local control between patients treated with chemotherapy and patients treated with tamoxifen alone (P =.219). Systemic treatment, margin status, young patient age, estrogen and progesterone receptor status, and primary tumor size were analyzed in a Cox regression analysis. The use of systemic treatment was the most powerful predictor of local control: patients who did not receive systemic treatment had a relative risk of local recurrence of 3.3 (95% confidence interval, 1.5 to 7.5; P =.004). CONCLUSION: In this retrospective analysis, systemic therapy appears to contribute to long-term local control in patients with lymph node-negative breast cancer treated with breast-conservation therapy.

Coronary artery dosimetry in intact left breast irradiation.

PURPOSE: The purpose of this article is to report dose-volume histograms of coronary vessels from irradiation of the intact left breast. PATIENTS AND METHODS: Fifteen women with cancer of the left breast underwent computed tomographic treatment planning for radiation treatments of an intact left breast. Images through the heart were reconstructed at 1-mm increments to permit contouring of the coronary vessels. Five treatment plans were created for each patient; one plan from the simulated treatment fields and four additional plans that were generated from virtual treatment fields created by shifting the isocenter 5 mm and 10 mm both superficially and deep. The radiation dose was calculated using a three-dimensional treatment planning system that incorporated heterogeneity correction factors. RESULTS: With no adjustment to the perpendicular lung distance, a mean volume of 12% of the left anterior descending coronary artery received 20 Gy, 6% received 30 Gy, and 3% received 40 Gy. The dose to the left anterior descending coronary artery varied significantly with changes in the perpendicular lung distance. From the mean perpendicular lung distance of 1.87 for the simulated fields, a 5-mm increase in the perpendicular lung distance resulted in an increase of 20%, 15%, and 12% in the percentage of the left anterior descending coronary artery treated to 20 Gy, 30 Gy, and 40 Gy, respectively. With a 10-mm increase, the respective volumes were increased to 49%, 41%, and 34%, respectively. A 5-mm reduction of lung distance in the original plan resulted in a decrease of 10%, 5%, and 3% in the percentage treated to 20 Gy, 30 Gy, and 40 Gy, respectively. The dose to the left main coronary artery, the right main coronary artery, and the left circumflex coronary artery was limited to scatter and was less than 7 Gy. Changes in the perpendicular lung distance did not significantly affect the dose administered to these vessels. DISCUSSION: The left anterior descending coronary artery is anatomically located at the edge of the cardiac silhouette on traditional treatment films. Small changes in the perpendicular lung distance can significantly change the dose delivered to this vessel. A fundamental change in the shape of the dose-volume histogram occurs at a perpendicular lung distance of 2.3 cm, whereas the dose is very low when the perpendicular lung distance is less than 1.3 cm. These points may serve as clinically important values in the treatment planning for cancer of an intact breast.

Local-regional control in breast cancer patients with a possible genetic predisposition.

PURPOSE: Local control rates for breast cancer in genetically predisposed women are poorly defined. Because such a small percentage of breast cancer patients have proven germline mutations, surrogates, such as a family history for breast cancer, have been used to examine this issue. The purpose of this study was to evaluate local-regional control following breast conservation therapy (BCT) in patients with bilateral breast cancer and a breast cancer family history. METHODS AND MATERIALS: We retrospectively reviewed records of all 58 patients with bilateral breast cancer and a breast cancer family history treated in our institution between 1959 and 1998. The primary surgical treatment was a breast-conserving procedure in 55 of the 116 breast cancer cases and a mastectomy in 61. The median follow-up was 68 months for the BCT patients and 57 months for the mastectomy-treated patients. RESULTS: Eight local-regional recurrences occurred in the 55 cases treated with BCT, resulting in 5- and 10-year actuarial local-regional control rates of 86% and 76%, respectively. In the nine cases that did not receive radiation as a component of their BCT, four developed local-regional recurrences (5- and 10-year local-regional control rates of BCT without radiation: 49% and 49%). The 5- and 10-year actuarial local-regional control rates for the 46 cases treated with BCT and radiation were 94% and 83%, respectively. In these cases, there were two late local recurrences, developing at 8 years and 9 years, respectively. A log rank comparison of radiation versus no radiation actuarial data was significant at p = 0.009. In the cases treated with BCT, a multivariate analysis of radiation use, patient age, degree of family history, margin status, and stage revealed that only the use of radiation was associated with improved local control (Cox regression analysis p = 0.021). The 10-year actuarial rates of local-regional control following mastectomy with and without radiation were 91% and 89%, respectively. CONCLUSIONS: Patients with a possible genetic predisposition to breast cancer had low 5-year rates of local recurrence when treated with breast conserving surgery and radiation, but the local failure rate exceeded 50% when radiation was omitted. Our data are consistent with the hypothesis that patients with an underlying genetic predisposition develop cancers with radiosensitive phenotypes.

Primary treatment of cystosarcoma phyllodes of the breast.

BACKGROUND: Cystosarcoma phyllodes is a rare sarcoma of the breast. Although surgical removal is the mainstay of treatment, the extent of surgery required (excision vs. mastectomy) and the need for additional local therapy, such as radiotherapy, are unclear. The current study evaluated the rate of local and distant failure, as well as potential prognostic factors, to better define appropriate treatment strategies. METHODS: One hundred one patients treated primarily for cystosarcoma phyllodes of the breast were evaluated. These tumors were classified histologically into benign (58%), indeterminate (12%), and malignant (30%) based on well defined criteria. Stromal overgrowth (29%) was considered separately. Surgery was comprised of local excision with breast conservation (47%) or mastectomy (53%). Microscopic surgical margins were negative in 99% of cases. Six patients received adjuvant radiotherapy. RESULTS: Overall survival for the 101 patients was 88%, 79%, and 62% at 5, 10, and 15 years, respectively. For patients with nonmalignant (benign or indeterminate) and malignant cystosarcoma phyllodes, the overall survival was 91% and 82%, respectively, at 5 years, and 79% and 42%, respectively, at 10 years. Similar rates were observed based on the presence or absence of stromal overgrowth. Local recurrence occurred in 4 patients, with an actuarial 10-year rate of 8%. Eight patients developed distant metastases, with an actuarial 10-year rate of 13%. Multivariate analysis using Cox proportional hazards regression revealed stromal overgrowth to be the only independent predictor of distant failure. CONCLUSIONS: Local failure in this group of largely margin negative patients with cystosarcoma phyllodes of the breast was low, showing that breast-conserving surgery with appropriate margins is the preferred primary therapy. The current study data do not support the use of adjuvant radiotherapy for patients with adequately resected disease. Patients with stromal overgrowth, particularly when the tumor size was > 5 cm, were found to have a high rate of distant failure; such patients merit consideration of a trial that examines the efficacy of systemic therapy.

Ductal carcinoma-in-situ: long-term results of breast-conserving therapy.

BACKGROUND: The role of breast-conserving therapy (BCT) in the management of ductal carcinoma-in-situ (DCIS) is controversial because of reported high recurrence rates. We reviewed our experience to determine whether the rate and pattern of locoregional recurrence after BCT were similar in patients with DCIS and patients with early-stage (T1) invasive breast tumors and whether local recurrence affected survival. METHODS: Between 1973 and 1994, 87 patients with DCIS alone, 22 patients with DCIS with microinvasion (DCIS-M), and 646 patients with invasive breast cancer 2 cm or smaller in diameter were treated with BCT (wide local excision with radiotherapy) at The University of Texas M. D. Anderson Cancer Center. Survival was calculated by the Kaplan-Meier method. The median follow-up times were 11 years for patients with DCIS alone, 12 years for patients with DCIS-M, and 8 years for patients with invasive breast cancer. RESULTS: Eleven (13%) of 87 patients with DCIS and 5 (23%) of 22 patients with DCIS-M had developed locoregional recurrences at follow-up. Two patients with DCIS with locoregional recurrence died of breast cancer. Of the 646 patients with invasive breast cancer, 56 (9%) had a locoregional recurrence, and 16 (2%) died of breast cancer. The median time to locoregional recurrence was significantly longer in patients with DCIS or DCIS-M (9-10 years) than patients with invasive tumors (5 years). CONCLUSIONS: DCIS is a favorable disease with an excellent long-term survival. The locoregional recurrence rate in patients with DCIS treated with BCT is similar to that in patients with early-stage invasive breast cancer treated with BCT, but time to locoregional recurrence is significantly longer in patients with DCIS. In patients with DCIS treated with BCT, intense surveillance for locoregional recurrence needs to be maintained for the patient's lifetime.

Locoregional recurrence patterns after mastectomy and doxorubicin-based chemotherapy: implications for postoperative irradiation.

PURPOSE: The objective of this study was to determine locoregional recurrence (LRR) patterns after mastectomy and doxorubicin-based chemotherapy to define subgroups of patients who might benefit from adjuvant irradiation. PATIENTS AND METHODS: A total of 1,031 patients were treated with mastectomy and doxorubicin-based chemotherapy without irradiation on five prospective trials. Median follow-up time was 116 months. Rates of isolated and total LRR (+/- distant metastasis) were calculated by Kaplan-Meier analysis. RESULTS: The 10-year actuarial rates of isolated LRR were 4%, 10%, 21%, and 22% for patients with zero, one to three, four to nine, or >/= 10 involved nodes, respectively (P <.0001). Chest wall (68%) and supraclavicular nodes (41%) were the most common sites of LRR. T stage (P <.001), tumor size (P <.001), and >/= 2-mm extranodal extension (P <.001) were also predictive of LRR. Separate analysis was performed for patients with T1 or T2 primary disease and one to three involved nodes (n = 404). Those with fewer than 10 nodes examined were at increased risk of LRR compared with those with >/= 10 nodes examined (24% v 11%; P =.02). Patients with tumor size greater than 4.0 cm or extranodal extension >/= 2 mm experienced rates of isolated LRR in excess of 20%. Each of these factors continued to significantly predict for LRR in multivariate analysis by Cox logistic regression. CONCLUSION: Patients with tumors >/= 4 cm or at least four involved nodes experience LRR rates in excess of 20% and should be offered adjuvant irradiation. Additionally, patients with one to three involved nodes and large tumors, extranodal extension >/= 2 mm, or inadequate axillary dissections experience high rates of LRR and may benefit from postmastectomy irradiation.

Locoregional irradiation for inflammatory breast cancer: effectiveness of dose escalation in decreasing recurrence.

PURPOSE: To evaluate the effect of radiation dose escalation on locoregional control, overall survival, and long-term complication in patients with inflammatory breast cancer. PATIENTS AND METHODS: From September 1977 to December 1993, 115 patients with nonmetastatic inflammatory breast cancer were treated with curative intent at The University of Texas M. D. Anderson Cancer Center. The usual sequence of multimodal treatment consisted of induction FAC or FACVP chemotherapy, mastectomy (if the tumor was operable), further chemotherapy, and radiation therapy to the chest wall and draining lymphatics. Sixty-one patients treated from September 1977 to September 1985 received a maximal radiation dose of 60 Gy to the chest wall and 45-50 Gy to the regional lymph nodes, 22 treated once a day at 2 Gy per fraction, and 35 were treated b.i.d. (32 after mastectomy and all chemotherapy was completed, and 2 immediately after mastectomy; one patient had distant metastases discovered during b.i.d. irradiation, and treatment was stopped). Four additional patients received preoperative radiation with standard fractionation. Based on the analysis of the failure patterns of the patients, the dose was increased for the b.i.d. patients in the new series, with 51 Gy delivered to the chest wall and regional nodes, followed by a 15-Gy boost to the chest wall with electrons. From January 1986 to December 1993, 39 patients were treated b.i.d. to this higher dose after mastectomy and all the chemotherapy was completed; and 8 additional patients received preoperative irradiation with b.i.d. fractionation to 51 Gy. During this period, another 7 patients were treated using standard daily doses of 2 Gy per fraction to a total of 60 Gy, either because they had a complete response or minimal residual disease at mastectomy or because their work schedule did not permit the b.i.d. regimen. Comparison was made between the groups for locoregional control, disease-free and overall survival, and complication rates. RESULTS: The median follow-up time was 5.7 years (range, 1.8-17.6 years). For the entire patient group, the 5- and 10-year local control rates were 73.2% and 67.1%, respectively. The 5- and 10-year disease-free survival rates were 32.0% and 28.8%, respectively, and the overall survival rates for the entire group were 40.5% and 31.3%, respectively. To evaluate the effectiveness of dose escalation, a specific comparison of patients who received b.i.d. radiation after mastectomy and completion of adjuvant chemotherapy was performed. There were 32 patients treated b.i.d. to 60 Gy in the old series versus 39 patients treated b.i.d. to 66 Gy in the new series. There was an significant improvement in the rate of locoregional control for the b.i.d. patients for the old vs. new series, from 57.8% to 84.3% and from 57.8% to 77.0% (p = 0.028) at 5 and 10 years, respectively. Chemotherapy regimens did not change significantly during this time period.Long-term complications of radiation, such as arm edema more than 3 cm (7 patients), rib fracture (10 patients), severe chest wall fibrosis (4 patients), and symptomatic pneumonitis (5 patients), were comparable in the two groups, indicating that the dose escalation did not result in increased morbidity. Significant differences in the rates of locoregional control (p = 0.03) and overall survival (p = 0.03), and a trend of better disease-free survival (p = 0.06) were also observed that favored the recently treated patients receiving the higher doses of irradiation. CONCLUSION: Twice-daily postmastectomy radiation to a total of 66 Gy for patients with inflammatory breast cancer resulted in improved locoregional control, disease free survival, and overall survival, and was well tolerated.

Prognostic implications of pathological lymph node status after preoperative chemotherapy for operable T3N0M0 breast cancer.

BACKGROUND: Although preoperative chemotherapy has become the standard of care for inoperable locally advanced breast cancer, its role for downstaging resectable primary tumors is still evolving. The purpose of this study was to determine whether the prognostic information from an axillary node dissection in patients with clinical T3N0 breast cancer was altered by preoperative chemotherapy compared with surgery de novo. METHODS: Between 1976 and 1994, 91 patients with clinically node-negative operable T3 breast cancer received doxorubicin-based combination chemotherapy on protocol at one institution. Fifty-three patients received both preoperative and postoperative chemotherapy (PreopCT), and 38 received postoperative chemotherapy only (PostopCT). All patients underwent axillary lymph node dissection as part of their definitive surgical treatment. There were no differences between the PreopCT and PostopCT groups in median age (51 vs. 49 years), median tumor size at presentation (6 cm vs. 6 cm), tumor grade, or estrogen receptor status (estrogen receptor negative 38% vs. 32%). The median follow-up time was 7 years. RESULTS: Patients in the PreopCT group had fewer histologically positive lymph nodes (median, 0 vs. 3, P < .01), and a lower incidence of extranodal extension (19% vs. 42%, P = .02). By univariate analysis, the number of pathologically positive lymph nodes (P < .01) and extranodal extension (P < .01) were predictors of disease-specific survival in PreopCT patients. Multivariate analysis showed that extranodal extension was the only independent prognostic factor in PreopCT patients (P < .01). Overall, PreopCT and PostopCT patients had similar 5-year disease-free survival rates (66% vs. 57%); however, PreopCT patients had worse disease-free (P = .01) and disease-specific survival (P = .04) when survival was compared after adjustment for the number of positive lymph nodes. Furthermore, PreopCT patients with 4-9 positive lymph nodes had a lower 5-year disease-free survival rate than PostopCT patients with 4-9 positive nodes (17 vs. 48%, P = .04). CONCLUSIONS: Axillary lymph node status remains prognostic after chemotherapy. Pathologically positive lymph nodes after preoperative chemotherapy are associated with a worse prognosis than the same nodal status before chemotherapy.