Poor Outcomes in HIV Negative and HIV Positive Patients with Plasmablastic Lymphoma (PBL) in a Population Based Patient Study

Introduction:

Plasmablastic lymphoma (PBL) is a rare, aggressive lymphoma of differentiated B lymphocytes. PBL sits at the boundary between Diffuse Large B Cell Lymphoma (DLBCL) and plasmablastic myeloma, and is more frequent amongst HIV positive patients. With an increasing divergence between therapy for myeloma and lymphoma and a deficit of prospective studies, the nature of PBL and outcome on current therapy merits further assessment.

Methods:

We conducted a retrospective analysis of consecutive patients diagnosed between 1st January 2004 and 1st July 2015 with PBL at the HMDS, a regional NHS diagnostic centre serving a population in excess of 3 million and representative of the wider UK population. Using the Haematological Malignancy Research Network (HMRN), a population based epidemiological study overlapping with the population served by HMDS, patient demographics and clinical outcomes were investigated.

Results:

53 patients (18 female, 35 male) were diagnosed with PBL. The median age was 67 years (19-95 years). 56.8% were 65 years or older. Six patients were HIV positive and three patients were immunosuppressed following solid organ transplantation. 37/53 (69.8%) patients had extranodal PBL at presentation, including 16/53 (30.2%) with gastrointestinal tract disease. Patient demographics are summarised below.

Table 1: Plasmablastic lymphoma

	Number	Male (%)	Median age in years (range)	Extranodal disease (%)
Total	53	35(66.7)	67 (19-95)	37 (69.8)
HIV positive	6	4 (66.7)	47 (39-56)	4 (66.7)
HIV negative	44	38 (86)	70 (19-95)	33 (75)
Post solid organ transplant	3	3 (100)	60 (55-68)	1 (33.3)

B cell markers (CD19, CD20, PAX-5) were negative or weakly expressed in all cases. One third demonstrated weak CD45 positivity. The neoplastic cells retained CD79a and expressed plasma cell marker CD138 but were negative for CD56. IRF4/MUM1 and BLIMP1 were expressed in all tumours examined. A high KI67 index was common, with 60.4% of cases having a KI67 index of 80% or greater.

Bone marrow involvement was found in eight patients, raising the question of re-assignment to plasmablastic myeloma. Three additional patients had small monoclonal B cell populations detected by flow cytometry.

Where suitable material was available, FISH for MYC was performed. MYC/IgH translocation was detected in 23%. The MYC/IgH translocation occurred as a sole abnormality in one third of positive cases and as part of a complex karyotype in15.4%. Extra copies of MYC were found in 15.4% and MYC amplification occurred in a 7.7%. LMP-1 was performed in 17/53 cases, of which three (17.6%) were positive. EBER-ISH was positive in 60% cases tested. HHV8 was detected in a single HIV positive patient.

Eight patients were managed with palliative intent, with a median survival of 6 days (95% CI 1-30). Of the patients receiving at least one cycle of chemotherapy, the median was 474 days (95% CI 174-1580). Cyclophosphamide, doxorubicin, vincristine and prednisolone (CHOP) was used most frequently. The median survival in the overall group is 241 days (95% CI 141-613). The five year overall survival of the overall population was 21.4%.

Figure 1: Kaplan Meier survival curve for plasmablastic lymphoma patients

Conclusion:

Our unselected cohort represents a large population based series of a rare and challenging disease and describes the experience in a predominantly elderly, HIV negative UK population. The male predominance, propensity for extranodal sites and association with EBV infection mirrors published reports in HIV positive patients. Survival in this population remains dismal.

Further investigation is warranted to refine diagnostic criteria and explore molecular pathogenesis of PBL.  Innovative therapeutic modalities guided by insights into molecular pathogenesis and exploring options across both lymphoma and myeloma therapeutic strategies are required to improve patient outcome. Given the rarity of PBL, prospective randomised studies are unlikely to be achievable. Our cohort serves as a historical control for future phase II clinical trials.