Abstract P6-06-01: Improved outcomes with occult versus known primary breast cancer: A SEER analysis

Abstract

Abstract Purpose: Controversy exists regarding the optimal management of the breast in the setting of lymph node presentation of an occult primary breast cancer (OPBC). Treatment includes radiation, a mastectomy or no treatment. Recent data confirms the need for a mastectomy or radiation for optimal outcomes with no clear difference between the two. In an effort to further elucidate this finding, we report outcomes of OPBC and known primary breast cancer (KPBC) who require local surgery using a large, population-based database. Materials and Methods: Using the Surveillance, Epidemiology and end Results (SEER) registry, data was obtained for all adult females diagnosed with T0-1 N1-3 M0 breast cancer who underwent definitive local therapy between 1990 and 2009. Patient characteristics that were analyzed included age at diagnosis, race, number of examined lymph nodes, number of positive lymph nodes, estrogen- and progesterone- receptor (ER, PR) positivity, and tumor grade. Propensity score modeling and Cox regression analysis were performed to compare overall survival (OS) and cause-specific survival (CSS) of patients with T0 and T1 breast cancer. Multivariate analysis (MVA) was used to determine independent predictors for OS and CSS. Given the large dataset, the propensity score model analyzed a randomly selected sample of 8000 subjects from the KPBC group, which was matched to the OPBC group. Results: 87,566 patients were identified of which 384 had OPBC and 87,182 had KPBC. The majority of patients were white (73.3%) with a median age of 58 and 56 years in the OPBC and KPBC groups, respectively. At baseline, the OPBC population had a greater median number of lymph nodes examined (18 vs. 16; p<.0001) and involved (4 vs. 2; p<.0001) lymph nodes, less ER+ (46.9% vs. 68.4%; p<.0001) and PR+ (33.3% vs. 57.7%; p<.0001) compared to the KPBC group. Regarding grade, the OPBC group had fewer moderately to well differentiated tumors compared to the KPBC group (7.8% vs. 47.5%; p<.0001). With a median follow-up of 74 months T0 OPBC had superior 5-/10-year OS (84%/68% vs. 81%/63%, p = 0.07) and 5-/10-year CSS (88%/80% vs. 88%/75%, p = 0.01) in comparison to KPBC. On MVA, the hazard of death is higher in the KPBC group for both OS [HR = 1.63 (95% CI 1.16-2.30), p = .005] and CSS [HR = 2.25 (95% CI 1.40-3.62), p = .0008]. Older age and a greater number of positive lymph nodes were associated with worse OS: an increase of one year in age at diagnosis was associated with a 4% increase in the hazard of death [HR = 1.04 (95% CI 1.03-1.05), p<.0001]; an increase of a single positive lymph node was associated with an 8% increase in the hazard rate [HR = 1.08 (95% CI 1.06-1.10), p<.0001]. For CSS, a greater number of positive lymph nodes were significantly associated with worse CSS [HR = 1.11 (95% CI 1.08-1.14), p<.0001]. Conclusions: Using the SEER registry, results from this large dataset demonstrate a significant improvement in OS and CSS in OPBC compared to KPBC, though only the CSS reached true statistical significance. These results help to clarify the controversial results of previously reported data for superior outcomes in OPBC. Future research is needed to identify additional predictors of outcomes. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P6-06-01.