LAUSR

Introduction Primary adrenal lymphoma (PAL) and primary renal lymphoma (PRL) are rare extranodal lymphomas, predominantly of diffuse large B-cell lymphoma (DLBCL) subtype. Adrenal/renal involvement is one of six risk factors that define the Central Nervous System International Prognostic Index (CNS-IPI), the most commonly used tool in determining the risk of CNS relapse in DLBCL. However, guidelines regarding CNS prophylaxis in early stage PAL/PRL (Ann Arbor IE and IIE) and advanced stages without high-risk CNS-IPI, and optimal route of administration have not been clearly elucidated. Here we provide an analysis of over 400 cases of PAL/PRL, the largest collection to date, including data on disease biology, risk of CNS relapse, CNS prophylaxis, and outcomes. Methods We conducted a comprehensive literature review on adult PAL/PRL cases reported from Jan 1st 1998 to July 1st 2019 in PubMed. Additionally, we collected cases available at our institution and through international collaborators. We excluded cases with CNS involvement at onset, unknown staging, and histology other than DLBCL. Data were tabulated regarding the following variables: age, gender, stage, laterality, cell-of-origin (germinal center B-cell-like (GCB) vs. activated B-cell-like (ABC)), chemotherapy, CNS prophylaxis (and type), CNS relapse, and overall survival (OS). OS was calculated using the Kaplan-Meier method, and comparison between OS curves made using the log-rank test, with statistical significance set as p-value < 0.05. Results/ Discussion With over 700 PAL/PRL cases available for review, 405 met inclusion criteria (258 from PubMed, 147 through collaboration). Analyzing the fraction of reported cases, patients were predominantly advanced stage (68%; 274/405), male gender (77%; 167/218), older than 60 years (73%; 157/214), had bilateral involvement (70%; 130/186), and were of ABC cell-of-origin (69%; 101/146) (Fig. 1A). Only 45 patients had available CNS-IPI scores (13 low/intermediate and 32 high), which limits drawing strong conclusions. The low/intermediate-risk group did not have a CNS relapse event, with 46% (6/13) receiving CNS prophylaxis. The high-risk group had 22% (7/32) CNS relapse rate overall, with 21% (3/14) relapse among the CNS prophylaxis subset. For the CNS-IPI group overall, a strong correlation was noted between disease stage and CNS-IPI risk: 85% of the low-risk CNS-IPI group had early-stage disease (stages IE and IIE), and 84% of the high-risk IPI group had advanced stage. Therefore, we sought to examine the rate of CNS relapse and impact of CNS prophylaxis for early vs. advanced-stage PAL/PRL. CNS relapse was documented in 15% of all cases (60/405), but was notably higher in advanced stage (20%; 55/275) vs. early stage (4%; 5/131) disease (Fig. 1A). Intriguingly, CNS prophylaxis did not affect CNS relapse rate in the advanced-stage group (26% vs. 28% with prophylaxis), while in the early-stage group it was beneficial (17% vs. 0% with prophylaxis) (Fig. 1B). Nevertheless, CNS prophylaxis was associated with a prolonged OS regardless of the disease stage (Fig. 1C). We then analyzed whether type of CNS prophylaxis used - intrathecal methotrexate or cytarabine (IT-chemo) vs. high-dose systemic methotrexate (HD-MTX) with or without IT-chemo - affected OS and CNS relapse rate. The OS curve favored HD-MTX in a limited size of 26 cases (Fig. 1D). CNS relapse data was available in 35 cases. CNS relapse was not found with HD-MTX (0/10), while it was observed in 32% (8/25) of the IT-chemo group, all of which were advanced stage (8/20). This can explain, at least in part, the lack of difference in rate of CNS relapse between the prophylaxis and non-prophylaxis groups in the setting of advanced stage. Consequently, a strong trend toward a better OS (p=0.12) was observed when HD-MTX was used in the setting of advanced-stage disease. Conclusion PAL/PRL is a rare presentation of DLBCL, more likely to be associated with high-risk features such as ABC cell-of-origin and advanced-stage disease. CNS prophylaxis should be used regardless of disease stage, preferably with HD-MTX in the setting of advanced-stage disease. Given the close relationship between stage and CNS-IPI, our findings suggest that, akin to primary testicular DLBCL, PAL/PRL could be an independent and significant predictor for CNS relapse regardless of the CNS-IPI, implicating the need for CNS chemo-prophylaxis in all cases. Ferreri: Novartis: Membership on an entity's Board of Directors or advisory committees; Gilead: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees, Research Funding; Roche: Research Funding. Safah:Amgen: Honoraria, Speakers Bureau; Incyte: Speakers Bureau; Jazz: Speakers Bureau; Verastem: Honoraria, Speakers Bureau; Celgene: Speakers Bureau. Saba:Pharmacyclics: Consultancy, Speakers Bureau; Kyowa Kirin: Consultancy; AbbVie: Consultancy; Janssen: Consultancy, Speakers Bureau.