Diffuse Large B-Cell Lymphoma with Testicular Involvement: Outcome and Risk of CNS Relapse in the Rituximab Era

Abstract 780FN2 Background: Testicular involvement of diffuse large B-cell lymphoma (Te-DLBCL) is associated with a poor outcome and a high risk of central nervous system (CNS) relapse. Rituximab is now routinely used into the treatment of both limited (LIM) and advanced (ADV) stage disease but it is unclear if this can reduce the rate of CNS relapse given it's limited penetration of the blood brain barrier and frequent parenchymal relapses in Te-DLBCL. A recent phase II study by the IELSG using R-CHOP, contralateral testicular radiation (RT) and IT prophylaxis in LIM stage patients demonstrated improved survival rates and a lower rate of CNS relapse compared with historical series. Herein, we evaluated the impact of R-CHOP on the natural history in all patients with Te-DLBCL Methods: The Centre for Lymphoid Cancer Database was used to identify patients with Te-DLBCL who were treated with curative intent. LIM stage was defined as stage I/II without bulk (< 10 cm) or B symptoms and also included patients with discordant involvement of the bone marrow with a low grade lymphoma whose stage was otherwise limited stage, and ADV stage included all others. Results: 109 patients with Te-DLBCL treated between 1985–2011 were identified including 10 patients (11%) with discordant non-follicular low grade lymphoma in the bone marrow (9) or lymph node (1). Twenty-one patients were excluded: chemotherapy refusal (7), palliative (n=9), HIV + (2), peritesticular and not testicular involvement (2), clinical information N/A (1). Of the remaining 88 patients, 40 received CHOP-like chemotherapy and 48 received R-CHOP. The median age at diagnosis was 68 y (range 26–83 y) and half the patients had LIM stage disease. Most patients received either prophylactic contralateral testicular RT (59) or had bilateral orchiectomy (11) with few exceptions: refusal (3), PD or death on therapy (12), unknown (3). Five patients with CNS disease at presentation (R-CHOP n=4; CHOP n=1) also received HD methotrexate (HDMtx) and IT CHT. Only nine patients (7 ADV) received CNS prophylaxis (R-CHOP n=4; CHOP n=5) either IT (8) or IT + HDMtx (1) R-CHOP treated patients were more likely to have > 1 extranodal (EN) sites involved (p=0.030), and there was a trend to a greater proportion of patients with ADV disease (p=.087). With a median follow up of 60 mos, the 5 year TTP and OS for the CHOP and R-CHOP treated patients was similar (TTP 52% vs 67.5%, p=.181; OS 52.5% and 57%, p=0.262). However, an improved TTP (p=.025) and a trend to improved OS (p=.085) was observed in ADV stage patients but not LIM stage patients (TTP p=.617, OS p=.407) treated with R-CHOP. In univariate analysis, stage, EN sites > 1, IPI (0-2 v 3–5) and urinary tract (UT) involvement (kidney, adrenal or ureter) were prognostic for both TTP and OS. In multivariate analysis treatment with rituximab, > 1 EN sites and UT disease were prognostic for both TTP (rituximab p=.006, > 1 EN sites, p=.014, and a trend for UT disease p=.067) and OS (rituximab, p=.009, > 1 EN sites p=.025, UT disease p=.016) Excluding patients with CNS disease at diagnosis, there was no difference in the time to CNS relapse (TTCNS) in R-CHOP (5 y 27%) compared to CHOP treated patients (5 y 23%) (p=.789) in both LIM (5 y 16.5% vs 23.5%, p=.901) and ADV disease (5 y 27% vs 23%, p=0.386). In univariate analysis, the IPI, EN sites, PS and UT disease were associated with an increased risk of CNS relapse. In multivariate analysis, only UT disease (HR 9.18, p<.0001) was predictive of CNS relapse. CNS relapse in LIM patients (12) was typically late (≤ 2 years n=9), commonly parenchymal, (n=9 vs leptomeningeal (LM) n=2, both n=1) and usually in the absence of systemic disease (9). In contrast, CNS relapse in ADV patients (12) was usually early (< ∼ 1 year 10/12), preferentially involved the LM either alone (8) or in concert (3) with parenchymal lesions but also in the absence of systemic disease (10/12). Conclusion: Although survival has improved in Te-DLBCL patients since the introduction of rituximab, the benefit is likely primarily through systemic disease control as there remains an inherently high risk of CNS relapse. Patients with ADV stage disease and particularly with UT tract involvement, have frequent early leptomeningeal relapse and should be considered for up-front CNS treatment. Given the propensity for late parenchymal relapses in LIM stage patients, strategies to test agents with CNS penetration need further investigation. Disclosures: Villa: Roche: Research Funding. Shenkier:Roche: Research Funding. Sehn:Roche: Honoraria, Research Funding. Klasa:Roche: Research Funding. Gascoyne:Roche: Research Funding. Connors:Roche: Research Funding. Savage:Roche: Honoraria, Research Funding.