Prognostic Factors and Outcomes in Primary Effusion Lymphoma in HIV-Infected Patients: A Single Center Experience

Primary effusion lymphoma (PEL) is a human herpesvirus 8-associated non-Hodgkin lymphoma (NHL) that predominantly develops in serous body cavities leading to recurrent lymphomatous effusions. It is poorly understood and very uncommon, accounting for only 3% of all HIV-related NHL. While outcomes have improved in patients with HIV-associated NHL since the widespread use of HAART, the prognosis of PEL remains very poor and median survival is less than 6 months. This study reports our institution’s experience with PEL over the past decade.

Between 2004 and 2014, we treated 8 patients with HIV-associated PEL. All patients were male. The median age was 46.5 years old (range 35-63), three were Caucasian, four were Hispanic, and one was African-American. The primary site of involvement was pericardial in 3 patients, pleural in 3 patients, and extra-cavitary in 2 patients. All patients were Ann Arbor stage IV at the time of diagnosis. The mean LDH and CD4 count at time of diagnosis was 625 U/L and 161 cells/mm3, respectively.

Of the eight patients identified with PEL, four have achieved complete remission (CR). Prognostic factors associated with achieving CR include compliance with HAART therapy prior to PEL diagnosis as well as lower serum lactate dehydrogenase (LDH) levels (202±30 versus 1049±290, p=.03). Patients achieving CR also had a higher average CD4 count (228±93 versus 95±35, p=.22) at time of diagnosis. Median survival is nearly 6 years for those in CR whereas it was 32 days in patients that died (one from an acute stroke, one from septic shock, and two from progressive disease). All patients who achieved CR had previously diagnosed HIV whereas two of the four who died were previously undiagnosed with HIV. Interestingly, pericardial sites of involvement were associated with significantly better outcomes as compared to pleural sites in our case series, as well. There was no association between outcome and HIV viral load, duration of HIV infection prior to PEL diagnosis, or presence of AIDS-defining illnesses. The patients achieving CR were treated with HAART in addition to Hyper-CVAD (1a-3b) or 6 cycles of CHOP or EPOCH. In addition, two patients received bortezomib as initial therapy. Given its rarity, our knowledge of PEL relies heavily on case reports and small case series. Here we report our institution’s experience with PEL, identify LDH and CD4 as possible prognostic factors in PEL, and suggest bortezomib-based chemotherapy as an effective treatment option in HIV-associated PEL.