Improving Outcomes in Plasmablastic Lymphoma: A Multicenter Experience from an American Cohort

Introduction:

Plasmablastic lymphoma (PBL) is a rare subtype of large B-cell lymphoma first described in 1997 in a small cohort of persons living with HIV (HIV-PBL). Since then, it has been observed post-transplant (PTLD-PBL) and even in otherwise immunocompetent patients (IC-PBL). With no current well-defined standard of care, prognosis of PBL has historically been poor (median overall survival (OS) 7-15 months), although recent studies suggest modest improvements in survival. We aimed to better characterize disease characteristics, treatment patterns, and factors associated with outcomes in a large contemporary US cohort.

Methods:

Patients (pts) ≥ 18 years old with PBL diagnosed between 1/2005 and 12/2022 from 20 academic centers in the United States were identified. Data on pt and disease characteristics and treatment were collected from medical records. OS, progression-free survival (PFS), lymphoma-specific survival (LSS), treatment-related mortality (TRM), and non-relapse mortality (NRM) were calculated using the Kaplan-Meier method. Cox regression and Fine-Gray competing risk regression were used to assess the association between clinical factors and survival outcomes.

Results:

331 pts with PBL were identified. The median age at diagnosis was 53 years (range 20-91y). 78% were male (n = 257). 46% had HIV-PBL (n = 151), 35% had IC-PBL (n = 117), 6% had PTLD-PBL (n = 20), and 13% had PBL with other immunosuppressed states (n = 43). 66% of pts were White (n = 219), 18% were Black/African American (n = 58), and 31% were Hispanic (n = 104). 74% of Black/African American and 52% of Hispanic pts were diagnosed with HIV-PBL, compared with only 37% of White pts. The median age at diagnosis was greater in the IC-PBL group (68y) than in the HIV-PBL and PTLD-PBL groups (46y and 55y, respectively; p < 0.001).

Most pts presented with Stage IV disease (61%, n = 220). Bone marrow (BM) involvement was present in 22% (n = 73). 92% of pts had extranodal sites involved at presentation (n = 302), most commonly the GI tract, BM, and oral cavity/jaw. Most tumors were CD138+, CD20-, MYC+, and EBV+ by either EBER or LMP1. EBV positivity was observed in 74% and 81% of PTLD-PBL and HIV-PBL cases, respectively, versus only 50% of IC-PBL cases (p < 0.001).

Of the 298 pts who received chemotherapy, the most common first line regimen used was DA-EPOCH (70%), followed by CHOP/CHOP-like regimens (14%), hyper-CVAD (4%), and CODOX-M/IVAC (4%). As part of their initial treatment, 35% of pts received a proteasome inhibitor (n = 104), 20% received rituximab (n = 58), and 3% received daratumumab (n = 10). 35% received CNS prophylaxis with either IT MTX/AraC or HD MTX (n = 103), 19% received radiation in the first line setting (n = 56), and 8% underwent a consolidative autologous HCT in first remission (n = 24).

Median FU for the entire cohort was 1.8y (range 0-17y). Median OS was 4.5y, with a median PFS of 1.4y. Median OS was 6.1y for HIV-PBL, 4.1y for IC-PBL, and 1.1y for PTLD-PBL. Factors associated with poor OS and PFS on multivariable analysis (MVA) included advanced age, stage III+, BM involvement, and EBV negativity. Female sex was associated with worse OS on MVA (HR 1.70, p = 0.027) and worse PFS on univariate analysis (HR = 1.53, p = 0.009). Use of CNS prophylaxis was associated with improved OS (HR 0.59, p = 0.031) and PFS on MVA (HR 0.68, p = 0.035). We found no significant difference in OS, PFS, and LSS for pts treated with DA-EPOCH versus pts treated with CHOP. On MVA, DA-EPOCH had a higher TRM compared to CHOP (p < 0.001). The addition of a proteasome inhibitor to 1L chemotherapy did not have a significant impact on OS, PFS, and LSS. While use of radiation and AHCT consolidation were associated with better PFS, OS was unaffected.

Conclusion:

In this largest US cohort of PBL patients to date, we observed improved outcomes compared to previous studies with 1- and 2y OS rates of 69% and 58%, respectively. Factors associated with worse prognosis were advanced age and stage, BM involvement, and EBV negativity, while HIV status was not associated with survival, consistent with other studies. Also, the absence of an OS benefit with DA-EPOCH and HCT with added toxicity is consistent with findings from other cohorts. We report several novel findings, including worse survival in female patients. More detailed analyses of the impact of immune status, CNS prophylaxis, and second line therapies on outcomes are pending.