OBJECTIVETo analyze the clinical features, therapeutic methods and prognosis of primary breast lymphoma (PBL).

METHODSTwenty-one PBL patients treated in Ruijin Hospital from January 2003 to December 2013 were included in this study, with 17 diffuse large cell lymphoma (DLBCL), 1 mucosa-associated lymphoid tumor (MALT), 1 follicular lymphoma (FL), 1 Burkitt lymphoma and 1 subcutaneous peniculitis T-cell lymphoma according to the WHO 2008 classification. Of 21 patients, only one patient with MALT has bulged tumor mass (>7 cm), other patients had tumor mass <5 cm. Six patients had core needle biopsy of tumor, 2 modified radical operation, and others tumor excision for diagnosis. All the patients received chemotherapy. The impacts of surgery, rituximab and prophylaxis with lumbar puncture on the outcomes of patients were analyzed. Survival was estimated using Kaplan-Meier method and compared by log-rank test. All the results were analyzed by SPSS 10.0.

RESULTSAmong 21 PBL patients, 19 achieved complete remission (CR), 1 partial remission (PR) and 1 disease progression (PD). Followed-up till July 2014, with median follow-up of 14 months (6-75 months), only one patient died, with 3-year survival of 92.3%. Compared with chemotherapy alone, the progression-free survival (PFS) of combination therapy (surgery plus chemotherapy) was significant longer (P=0.015), but without statistic differences of CR rate and overall survival (OS) between two groups. Among the 20 patients with CD20-positive tumor cells, 17 received Rituximab. PFS and OS, as well as CR rate of PBL had no difference between the treatment with and without Rituximab. The incidence of central nervous system (CNS) infiltration had no difference between patients with and without CNS prophylaxis through lumber-puncture and intrathecal injection.

CONCLUSIONWith the common subtype of DLBCL, PBL patients had good outcome. Surgery, as a method to obtain tumor samples for diagnosis, could not prolong OS of patients. Therefore, radical operation shouldn't be recommended. PBL was reported to have high risk of CNS events, but prophylaxis with lumber puncture and intrathecal injection couldn't decrease the incidence of CNS infiltration.