Risk factors for breast cancer mortality after ductal carcinoma in situ diagnosis differ from those for invasive recurrence.

BACKGROUND: Breast cancer mortality after ductal carcinoma in situ is rare, making it difficult to predict which patients are at risk and to identify whether risk factors for this outcome are the same as those for invasive recurrence. We aimed to identify whether risk factors for invasive recurrences are similar to those for breast cancer death after a diagnosis of pure ductal carcinoma in situ. METHODS: The Surveillance, Epidemiology, and End Results Program was queried for female patients diagnosed with pure ductal carcinoma in situ. Cumulative incidence was estimated by treatment group using competing risks. Competing risks regression was then performed for the development of in-breast invasive recurrence with competing risks of breast and non-breast cancer death. Competing risks regression was then again performed for development of breast cancer mortality with the competing risk of non-breast cancer death. RESULTS: A total of 29,515 patients were identified. Of them, 164 patients suffered breast cancer mortality without an intervening invasive recurrence, and 44 suffered breast cancer mortality after an invasive in-breast recurrence. On competing risks analysis for invasive in-breast recurrence, significant factors included lesion size >5 cm (hazard ratio = 1.59, 95% confidence interval 1.24-2.04, P < .001), diffuse disease (hazard ratio = 0.0005, 95% confidence interval 0.0003-0.0007, P < .001), other race (hazard ratio = 1.29, 95% confidence interval 1.10-1.52, P = .002), Black race (hazard ratio = 1.21, 95% confidence interval 1.01-1.46, P = .04), age at diagnosis (hazard ratio = 0.99, confidence interval 0.98-1.00, P = .02), low-grade disease (hazard ratio = 0.79, 95% confidence interval 0.64-0.96, P = .02), lumpectomy with radiation (hazard ratio = 0.67, 95% confidence interval 0.58-0.77, P < .001), and mastectomy (hazard ratio = 0.36, 95% confidence interval 0.30-0.44, P < .001). Significant factors for breast cancer mortality included age at diagnosis (hazard ratio = 1.04, 95% confidence interval 1.03-1.05, P < .001), Black race (hazard ratio = 2.88, 95% confidence interval 2.08-3.99, P < .001), diffuse disease (hazard ratio = 6.02, 95% confidence interval 1.39-26.07, P = .02), lumpectomy with radiation (hazard ratio = 0.51, 95% confidence interval 0.36-0.72, P < .001), and mastectomy (hazard ratio = 0.60, 95% confidence interval 0.50-0.92, P = .02). CONCLUSION: Our results suggested that risk factors for in-breast invasive recurrence after a diagnosis of pure ductal carcinoma in situ differ from risk factors for breast cancer mortality and development of metastatic recurrence. In-breast invasive recurrence is not the only consideration for breast cancer specific mortality in ductal carcinoma in situ patients.