Author:
O’Keefe Thomas J,Harismendy Olivier,Wallace Anne M
Abstract
AbstractPURPOSEThe current trend in ductal carcinoma in situ (DCIS) research is towards treatment de-escalation. Problematically, prognostic indicators for patients at high risk of breast cancer mortality (BCM), rather than recurrence, have not been identified. We aim to identify prognostic factors for the development of metastatic disease and mortality.EXPERIMENTAL DESIGNPatients diagnosed with DCIS in a local cancer registry as well as in the National Cancer Database (NCDB) and the Surveillance, Epidemiology and End Results (SEER) program were assessed for factors prognostic of metastatic disease, overall, and breast-cancer specific survival. Cox and competing risks regressions were developed.RESULTSAmong 5 patients who developed distant metastatic disease in the cancer registry, 3 had identifiable growth distribution; all 3 were diffuse type. None had in-breast invasive or DCIS recurrences before metastasis. In NCDB and SEER, cumulative incidence of any cause mortality (ACM) and BCM at 10 years was 12%/5.0% for diffuse lesions; 8%/3.6% for patients with microinvasive disease, 7.4%/2.3% for lesions >5 cm, 5.6%/1.4% for lesions 2-5 cm and 5.5%/1.5% for lesions <2 cm. Multivariate hazard ratios for ACM in NCDB and BCM in SEER were 2.0 and 5.3 (p=0.03 and 0.02, respectively). Among patients with diffuse lesions, cumulative incidence ACM at 10 years was 15.0% among those undergoing unilateral mastectomy vs. 2.5% among those undergoing bilateral mastectomy (p=0.11).CONCLUSIONDiffuse DCIS represents an uncommon but deadly subtype for whom treatment escalation, rather than de-escalation, is likely necessary. Further studies elucidating the mechanism of metastasis and best treatment course are needed.
Publisher
Cold Spring Harbor Laboratory