8 Locoregional Treatment Outcomes for Patients with Ipsilateral Supraclavicular Metastases at Diagnosis of Breast Cancer E. H. Huang, E. A. Strom, V. Valero, G. H. Perkins, N. R. Schechter, K. K. Hunt, F. Meric-Bernstein, M. D. McNeese, G. N. Hortobagyi, T. A. Buchholz The University of Texas M. D. Anderson Cancer Center, Houston, TX Purpose/Objective: The American Joint Committee on Cancer recently changed the staging of breast cancer so that ipsilateral supraclavicular (SCV) nodal metastases are now considered locoregional disease. However, there are very few published data regarding the efficacy of locoregional treatment for this subset of patients. The purpose of this study was to investigate the efficacy of trimodality treatment for breast cancer patients who presented with SCV disease without systemic metastases. Materials/Methods: We retrospectively reviewed the hospital records of 71 patients with ipsilateral SCV nodal involvement at presentation that were treated at our institution between 1974 and 2000 for a noninflammatory, nonmetastatic breast cancer. All of the patients were treated with curative intent using doxorubicin-based neoadjuvant chemotherapy, mastectomy or breast-conserving surgery (BCT, n = 11), and radiation to the breast or chest wall and the regional lymphatics. Forty-five patients had ultrasound evaluations before and after chemotherapy. Fifty-five patients had palpable SCV lymphadenopathy at presentation, and 16 had SCV lymphadenopathy identified by ultrasound without palpable disease. Total locoregional recurrence (LRR), disease-free survival (DFS), and overall survival (OS) were calculated according to the Kaplan-Meier method and differences between groups were compared using the log-rank test. Results: The median follow-up of the surviving patients was 5.1 years. The 5-year rates of LRR, DFS, and OS were 23%, 31%, and 47%, respectively. Of the 15 patients who had a LRR, 7 failed in the SCV nodes (5 had isolated SCV recurrences) and 10 had chest wall failures (6 had isolated chest wall recurrences). All 7 SCV failures recurred within the radiation field. Patients with palpable SCV lymphadenopathy at presentation had a higher rate of LRR than those with SCV disease detected only by ultrasound (31% vs 0%, p = 0.017). The extent of residual SCV disease after neoadjuvant chemotherapy was also associated with worse outcomes. The patients who did not achieve a complete response (CR) of the SCV disease by clinical exam had a higher rate of LRR (36% vs 14%, p = 0.026), as did those who did not achieve a CR by ultrasound (34% vs 4%, p = 0.007). Of the patients who had a CR of the SCV disease by clinical exam, those with a persistently abnormal ultrasound after chemotherapy had a higher rate of LRR, but this difference was not statistically significant (18% vs 4%, p = 0.19); however, the DFS was significantly different between the two groups (0% for clinical CR with abnormal ultrasound vs 55% for clinical CR with normal ultrasound, p = 0.03). Two other factors that correlated with LRR were having clinical T4 disease (35% vs 8%, p = 0.009) or estrogen receptor-negative disease (32% vs 15%, p = 0.056). The use of BCT was not associated with a higher LRR rate (18% for BCT vs 24% for mastectomy, p = 0.80). In addition, higher radiation doses delivered to the SCV field were not associated with a lower rate of LRR, although the higher radiation doses were associated with having gross residual SCV disease at the time of radiation (median dose: 60 Gy for those with gross disease vs 56 Gy for those who had a CR). Conclusions: Radiation achieved excellent local control after surgery for patients with ipsilateral SCV metastases who achieved a CR of the SCV disease after neoadjuvant chemotherapy. However, LRR remained a persistent problem despite trimodality therapy for those with residual SCV lymphadenopathy after chemotherapy. For patients who achieved a CR of the SCV disease by clinical exam, ultrasound evaluation of the SCV fossa may help discriminate the risk of disease recurrence. SCV involvement should not be considered a contraindication for BCT. 9 Are There Patient Subsets at High Locoregional Failure Risks After Mastectomy for pT1/2 pN0 Breast Cancer that May Warrant Radiotherapy P. T. Truong, 1,2 M. Lesperance, 3 A. Culhaci, 3 C. Speers, 1 I. A. Olivotto 1,2 1 Radiation Therapy Program and Breast Cancer Outcomes Unit, BC Cancer Agency, Victoria, BC, Canada, 2 University of British Columbia, Victoria, BC, Canada, 3 Dept of Mathematics and Statistics, University of Victoria, Victoria, BC, Canada Purpose/Objective: Postmastectomy radiotherapy (PMRT) improves locoregional control and survival in patients with tumors 5cm or node-positive disease but little information is available to guide PMRT decisions in women with T1/T2 N0 breast cancer. This study aims to identify subsets with T1/T2, N0 breast cancer at sufficiently high risks of locoregional failure (LRF) who may benefit with PMRT. Materials/Methods: Data were analyzed for 2024 women referred from 1989 –1998 with pT1/T2, pN0, M0 breast cancer treated with mastectomy without adjuvant radiotherapy. Patient, tumor, and treatment factors analyzed were: age at diagnosis, tumor size, location, histology, grade, lymphovascular invasion (LVI), # nodes removed, estrogen receptor (ER) status, margin status, and systemic therapy use. Recursive partitioning analysis, a method of building decision trees using independent prognostic factors, was performed to classify subsets at different LRF risks based on combinations of prognostic factors. Results: Median follow-up was 7.5 years. Local, regional, and combined locoregional failure rates were 4.5%, 2.7%, and 6% respectively. On multivariate analysis, independent factors predicting LRF were histologic grade (p = .001), LVI (p = .003), and T stage (p = .001). On recursive partitioning analysis, the first split in the partition tree was histologic grade. Among 1370 patients without high grade histology, LRF was 4.2%. Among 654 patients with grade III disease (LRF 10.2%), the concomitant presence of LVI (N = 168) was associated with LRF rates of 17%. In 486 patients with grade III disease without LVI, T2 tumors (N = 211) conferred LRF rates of 12%, which was increased to 20.5% among patients who did not receive systemic therapy. Conclusions: Women with pT1/T2 N0 breast cancer experienced high LRF rates 15% in the presence of grade III disease with LVI or T2 tumors not receiving systemic therapy. These high-risk subsets of patients with node-negative breast cancer warrant consideration of PMRT. S134 I. J. Radiation Oncology ● Biology ● Physics Volume 60, Number 1, Supplement, 2004