Socio-economic deprivation independently predicts early recurrence in estrogen receptor positive breast cancer.

CTRC-AACR San Antonio Breast Cancer Symposium: 2008 Abstracts Abstract #6081 Introduction: Socio-economic deprivation is associated with reduced survival in breast cancer, but little is known about how it affects the pattern of disease recurrence. We investigated the relationships between deprivation, other risk factors and recurrence in post-menopausal women with operable ER+ breast cancer. Methods: Patients who underwent potentially curative surgery for ER+ breast cancer between 1995 and 2005 at four UK centres were studied. Those patients whose area of residence was ranked in the highest quintile group according to the appropriate national Index of Multiple Deprivation were classed as deprived. The cumulative risk of recurrence was calculated using the Kaplan-Meier method, and hazard ratios were calculated using conventional and time-dependent Cox regression analysis stratified by centre. Results: 4110 patients were available for analysis, of whom 821 (20.0%) were classed as deprived. The median age at diagnosis was 62 years in both deprived and non-deprived groups. Tumour size and grade did not vary with deprivation status, but nodal involvement (38.0% vs 34.3%, p=0.05, Fisher's exact test) and lymphovascular invasion (28.1% vs 23.6%, p=0.01) were more prevalent in deprived patients. The latter were also more likely to undergo mastectomy (53.3% vs 47.2%, p=0.002) and less likely to receive adjuvant radiotherapy (50.4% vs 58.5%, P<0.001) or chemotherapy (19.1% vs 22.8%, p=0.03). Median follow up was 5 years from diagnosis. Raw cumulative recurrence at 2.5 years was 8.4% (95% confidence interval 6.4-10.4) in deprived and 6.1% (5.3-6.9) in non-deprived patients. At 5 years the corresponding rates were 15.8% (13.1-18.5) and 13.3% (11.9-14.7). Only distant recurrence contributed to this difference, the local and contralateral recurrence rates being almost identical in deprived and non-deprived groups. Time-dependent Cox regression analysis showed that deprivation was a significantly stronger predictor of recurrence within 2.5 years of diagnosis than at later intervals, with or without adjustment for other prognostic variables (p<0.05). For recurrence within 2.5 years, the unadjusted hazard ratio for deprivation was 1.64 (95% CI 1.22-2.22, p=0.001). After adjustment for age, mode of presentation, pathological variables and treatment, the hazard ratio was 1.51 (1.07-2.14, p=0.02). Conclusions: Post-menopausal women with ER+ breast cancer who are resident in areas of high deprivation are at increased risk of early recurrence, independently of known differences in tumour pathology and treatment. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 6081.