Although the Japanese multicenter phase II trial in localized primary gastric diffuse large B-cell lymphoma (PG-DLBCL), which evaluated three cycles of CHOP followed by radiotherapy (RT), showed good prognosis (Cancer Sci 2005; 96: 349), reports about outcomes and prognostic factors of localized PG-DLBCL patients in the rituximab era are limited. Recently, it has been reported that the concurrent expression of MYC and BCL2 predicts unfavorable outcome in DLBCL patients treated with R-CHOP (J Clin Oncol 2012; 30: 3460). However, the impact of the concurrent expression of MYC and BCL2 on outcomes of localized PG-DLBCL patients has never been reported.

Patients and Methods

We retrospectively analyzed 52 consecutive patients diagnosed as having localized (stage I or II according to the Lugano Staging System for Gastrointestinal Lymphomas) PG-DLBCL who were initially treated at our institution between 2003 and 2013. Positivity of MYC in immunohistochemistry was defined as labeling of tumor cells of more than 40% and positivity of BCL2 was defined as more than 70%. The lymphoma cells were assigned a GCB or non-GCB phenotype using the Hans algorithm for determining the cell-of-origin (COO) subtyping.


Twenty-four (46%) patients were male and 28 (54%) female, with a median age of 62 years (range: 29-85). Thirty (58%) patients presented with stage I disease, 15 (29%) with stage II1, two (4%) with stage II2 and five (9%) patients with stage IIE. Most patients (47 patients; 90%) had low or low-intermediate risk based on the International Prognostic Index. Fifty (96%) patients received R-CHOP with or without RT, and one each received CHOP plus RT, and total gastrectomy followed by rituximab. The median number of CHOP cycles was three (range: 2-8). The majority (43 patients; 83%) of patients were treated with R-CHOP followed by RT. COO subtype could be determined in 48 of the 52 patients (63% GCB and 37% non-GCB). Both MYC and BCL2 expression could be assessed in 47 of the 52 patients, and the concurrent expression of MYC and BCL2 was confirmed in seven (15%) patients. In this analysis, no patients showed positivity for EBER-1 in situ hybridization, which was reported as an adverse prognostic factor of localized PG-DLBCL in the pre-rituximab era. Median follow-up duration was 76 months (range: 4-127 months). Fifty (96%) patients achieved complete responses, and the remaining two without concurrent expression of MYC and BCL2 had primary refractory disease. The estimated 5-year overall and progression-free survival rates of all 52 patients were 90% (95% CI, 75-96%) and 89% (95% CI, 75-95%), respectively (Fig. 1). The estimated 5-year overall survival rates of GCB phenotype and non-GCB phenotype cases were 86% (95% CI, 63-96%) and 93% (95% CI, 59-99%), respectively, with no statistically significant difference (p=0.96). The estimated 5-year overall survival rates of the patients with and without concurrent expression of MYC and BCL2 were 100% and 88% (95% CI, 72-96%), respectively (Fig. 2). There was no significant difference between the two cohorts (p=0.74).


The results of our analysis showed good prognosis, and revealed that COO subtype and concurrent expression of MYC and BCL2 did not influence the outcome in patients with localized PG-DLBCL treated with rituximab-containing chemotherapy with or without RT. Further investigations, especially a prospective cohort study, are needed to confirm our results.


Maruyama:Eisai Co., Ltd: Honoraria. Kobayashi:Otsuka Pharmaceutical Co., Ltd.: Research Funding; ARIAD Pharmaceuticals, Inc.: Research Funding; Boehringer Ingelheim GmbH: Research Funding. Tobinai:Zenyaku Kogyo: Research Funding; Chugai Pharmaceutical: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.