Background: Primary breast diffuse large B-cell lymphoma (DLBCL) has poor outcomes with frequent extranodal failures, particularly in the central nervous system (CNS). To prevent CNS recurrence, we designed this phase II trial that addressed feasibility and activity of conventional immunochemotherapy and CNS prophylaxis. Methods: This prospective, multicenter, single-arm phase II study was conducted to evaluate efficacy and safety of 6 cycles of conventional rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone every 21 days (R-CHOP) with the addition of 4 doses of intrathecal methotrexate (IT MTX; 12mg) during the first 4 cycles of R-CHOP in patients with primary breast DLBCL. Primary breast lymphoma was defined as lymphoma involving one or both breasts as a sole extranodal site regardless of specific nodal involvement status. The primary end-point was 2-year progression-free survival (PFS). Secondary end-points included cumulative incidence of CNS recurrence, overall survival (OS), and safety. All patients provided written informed consents and the study was registered at www.clinicaltrials.gov as #NCT01448096. Results: Thirty-three patients with primary breast DLBCL were enrolled between Jan 2012 and Jul 2017 in the Consortium for Improving Survival of Lymphoma (CISL) member institutions. The median age was 50 years at diagnosis (range, 29-75) and all were female. Right breast involvement was more common than left (18 [55%] vs 14 [42%]) and bilateral breast involvement was found in one patient (3%). Nodal involvement was present in 16 patients (49%), primarily in regional nodes (14 patients). Thus, the Ann Arbor stage was IE in 17 (52%), IIE in 13 (39%), IIIE in 2 (6.1%), and IV in 1 (3.0%). ECOG performance status was ≥2 in 1 patient (3%) and serum LDH level was elevated in 9 (27%). Therefore, the IPI and the CNS-IPI risk were mainly low (28 patients, 85%; respectively). No patients had CNS involvement at diagnosis. 32 (97%) of the 33 patients completed R-CHOP as planned, and the remaining patient withdraw a consent after four cycles of R-CHOP because of poor tolerance. CNS prophylaxis using IT MTX was completed as planned in 31 patients (94%), but it was discontinued in 2 patients because of patient's refusal. These 2 patients received two and three IT MTX doses, respectively. 32 patients (97%) were evaluable for treatment response and all these patients achieved a complete response. At the cutoff date of this analysis (10 Jul 2019), all patients who entered a follow-up phase had at least 24.0 months of follow-up. With a median follow-up duration of 46.1 months (IQR 31.1-66.8), 6 patients had experienced treatment failure and 3 of these died. The 2-year PFS and OS were 81.3% (95% CI, 67.7-94.8) and 93.5% (95% CI, 84.9-100.0), respectively (fig 1A and B). Of the 6 patients with treatment failure, diseases involved CNS with or without lymph nodes in 4 patients and breasts in 2 patients (1 ipsilateral and 1 contralateral breast recurrence). 3 of the four patients with CNS recurrence had isolated CNS recurrences (2 brain parenchymal and 1 meningeal disease) and one had a concurrent meningeal and lymph nodal recurrence. All 4 patients with CNS recurrence had received prophylactic IT MTX as planned by protocol. The 2-year cumulative incidence of CNS recurrence, taking into account the competing risk of death, was 12.5% (95% CI, 0.3-23.2, fig 1C). Although the number of patients with intermediate CNS-IPI risk was small (5 patients, 15%), the cumulative incidence of CNS recurrence did not differ significantly according to the CNS-IPI risk group. All CNS recurrences occurred within the first 2 years after enrolment. Toxicities were generally manageable during the R-CHOP and IT MTX treatment. No deaths as a result of toxicity occurred during treatment. Conclusion: Our study shows that conventional R-CHOP with prophylactic IT MTX is feasible in patients with primary breast DLBCL. However, given a substantially high rate of CNS recurrence, further studies to properly define the best strategy for CNS prophylaxis should be needed in patients with primary breast DLBCL. Figure 1 Disclosures Yoon: F. Hoffmann-La Roche Ltd: Research Funding. Kim:Celltrion: Research Funding; Novartis: Research Funding; Donga: Research Funding; Kyowa-Kirin: Research Funding; Novartis: Research Funding; J + J: Research Funding; F. Hoffmann-La Roche Ltd: Research Funding.