Introduction:

Patients with chronic HBV infection (HBsAg+) have a higher risk of diffuse large B-cell lymphoma (DLBCL), more unfavorable features at presentation, and worse survival compared to HBV-negative patients (Ren et al. Blood. 2018). Data on the impact of past HBV infection on clinical features and outcomes of DLBCL are conflicting. HBV is a well-known risk factor for HCC but not for DLBCL. There are 3,000 new annual cases of HBV infection in Philadelphia and 25,000 persons live with HBV, mostly unaware, unvaccinated, and unscreened (https://www.hepb.org/). Largest at-risk groups are Asians and African Americans. Our goal was to characterize clinical features and estimate disparate impact of HBV infection (past [HBcAb+; HBsAg-] or chronic [HBsAg+; HBcAb+]) in patients with DLBCL, seen at Thomas Jefferson University Hospital (TJUH).

Methods:

We utilized EPIC SlicerDicer to retrospectively identify cases of DLBCL between June 10, 2010 and March 4, 2024 at TJUH. Eligible patients included pathologically confirmed DLBCL, age greater than or equal to 18, and positive HBV serology (HBsAg+, HBcAb+, or both). Patients with a history of human immunodeficiency virus (HIV) or hepatitis C virus (HCV) infections were excluded. Germinal center (GC) and non-GC B-cell type was determined by the Hans algorithm. The primary endpoint, actuarial progression-free survival (PFS), was defined as the interval between diagnosis and relapse, progression, or death from any cause, whichever came first, and was calculated by the Kaplan-Meier method. All analyses were conducted with SAS version 9.4.

Results:

The initial Slicer Dicer search, using ICD-10 codes for eligibility criteria, identified 71 patients. After manual review, 22 patients were excluded due to lack of documented HBV serology data or pathologic confirmation of DLBCL. Forty-nine patients were included in the final analysis. A plurality of patients self-identified as Asian (43%; 21/49); 33% (16) as white; and 18% (9) as Black or African American. Median age at diagnosis was 64.7 years. Of the whole cohort (49), all but 1 had documented HBcAb positivity (the one HBcAb-unknown patient was HBsAg+). Twelve patients (24%) were HBsAg+ and 37 (76%) were HBsAg-. Two-thirds of patients had advanced-stage (Ann Arbor III/IV) at the time of diagnosis (67%; 33/49); 41% (20/49) had extranodal disease. Two-thirds (10/15) of the National Comprehensive Cancer Network (NCCN) IPI-evaluable patients were high-intermediate or high risk. Of 37 patients evaluable for cell of origin (COO), 57% (21/37) were GC type and 43% (16/37) were non-GC type.Of 38 patients with FISH cytogenetics data for MYC and BCL2/BCL6 rearrangements, 5 (13%) were double-hit (C-MYC plus either BCL-2 or BCL-6). Two of these 5 patients were HbsAg+. Median pre-treatment albumin was 3.70 g/dL. Forty-seven of 49 patients received front-line DLBCL chemoimmunotherapy with curative intent. With an actuarial median follow-up of 3.3 years (range 0.3 - 14 years), the actuarial PFS estimate at 5 years was 48.2% (95% CI: 31.5% - 64.9%). We did not observe statistically significant associations between PFS and HBsAg status, C-MYC/BCL-2/BCL-6 status, NCCN-IPI score, or other clinical parameters.

Conclusion:

We report a cohort of Philadelphians with past or chronic HBV infection and DLBCL from TJUH. A high percentage had advanced stage (67%), extranodal disease (41%), high or high-intermediate NCCN-IPI (67%), and double hit cytogenetics (13% of subset, consistent with high GC type prevalence [57%]). Estimated PFS at 5 years was <50%, appearing lower than reported in the literature in HBV-unselected patients (Lenz et al. N Engl J Med. 2008). Two-thirds of the patients were Asian and African American, highlighting minority overrepresentation in this poor risk subset of DLBCL and identifying an important disparity. These findings support data from HBV-endemic world regions showing that DLBCL patients with past or current HBV infections have unique clinical features and inferior survival outcomes (Cheng et al, Oncologist. 2020; Ren et al. Blood. 2018). Due to selection bias, this dataset needs to be taken with caution, but it is hypothesis-generating and can inform public healh strategies to address health disparities in DLBCL in Philadelphia. Further studies should assess if MYC and BCL2/BCL6 rearrangements are more prevalent in this population and confirm inferior survival outcomes.

Disclosures

Porcu:Kiowa Kirin: Honoraria, Research Funding; Innate Pharma: Membership on an entity's Board of Directors or advisory committees, Research Funding; SOBI: Consultancy, Honoraria, Speakers Bureau; Viracta: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; TEVA: Consultancy, Research Funding; ONO: Consultancy, Research Funding; Daiichi: Consultancy, Honoraria.

This content is only available as a PDF.
Sign in via your Institution