BACKGROUND: Kaposi sarcoma herpesvirus (KSHV, also called human herpesvirus-8) is the causative agent of 3 disorders: primary effusion lymphoma (PEL), Kaposi sarcoma (KS), and a plasmablastic form of multicentric Castleman disease (KSHV-MCD). It also causes KSHV inflammatory cytokine syndrome (KICS), which is characterized by inflammatory symptoms and an elevated KSHV viral load. Multiple KSHV-associated diseases, which usually develop in HIV-infected patients, can present together in the same patient. Effusions can occur in each of these diseases, thereby presenting a diagnostic challenge. Identifying PEL is especially crucial as prompt treatment with multi-agent chemotherapy can be curative. We analyzed effusions from patients with KICS, PEL, and KSHV-MCD to identify distinct immunologic characteristics and virologic profiles that may aid diagnosis, inform treatment and elucidate pathogenesis.

PATIENTS AND METHODS: We identified 22 HIV-infected patients with effusions [pleural effusions (20), ascites (1) and pericardial effusion (1)] with diagnoses of PEL (9 patients), KICS (8 patients), or KSHV-MCD (5 patients). All patients had a concurrent diagnosis of KS. We obtained clinical and immunologic characteristics from effusions and paired serum samples at the same timepoint for each patient. Serum and effusion cytokine levels of interleukin (IL)-1ß, IL-2, IL-4, IL-6, IL-8, IL-10, IL-13, and IL-12p70; interferon gamma (IFN-g); tumor necrosis factor alpha (TNF-a); vascular endothelial growth factor (VEGF); and inducible protein 10 (IP-10) were evaluated using a commercial multiplex assay. Peripheral blood mononuclear cell (PBMC) and effusion- associated KSHV and Epstein Barr virus (EBV) viral DNA (KSHV-VL, EBV-VL) in PBMC and cells were quantified using PCR with primers to KSHV K6 and EBV pol. Effusion and serum immunologic and virologic characteristics were compared within each disease and separately between diseases using the Wilcoxon sign rank test and Wilcoxon rank sum test, respectively. In these exploratory analyses with few patients, no correction was made for multiple comparisons.

RESULTS: In patients with PEL, the median (med) age was 42 years with med CD4+ count of 54 T-cells/μL and HIV viral load (VL) of 325 copies/mL. In those with KICS, the med age was 32 years, with a med CD4+ count of 119 T-cells/μL and HIV VL of 48 copies/mL. In patients with KSHV-MCD, the med age was 31 years, med CD4+ count of 118 T-cells/μL and HIV VL was 75 copies/mL.

IL-13 was substantially higher in PEL effusions as compared to serum levels (med 16.9 vs. <0.12 ng/ml; p=0.007). In addition, patients with PEL had significantly higher levels of 6 other cytokines (IL-12p70, IL-1ß, IL-2, IL-4, IL-6 and IP-10) in effusions as compared with serum (Table 1, p<0.05). In both KSHV-MCD and KICS, IL-12p70, IL-2 and IL-4 levels were higher in the effusion as compared with serum (p<0.05).

In analyses comparing serum and cytokine differences among diseases, effusions from patients with PEL had detectable levels of IL-13 (med 16.9 ng/ml; interquartile range 9.7-26.9 ng/ml) compared to patients with KSHV-MCD (med <0.114 ng/ml; p=0.0037) or KICS (med <0.114 ng/ml; p=0.0003). PEL effusions had higher IL-1ß levels as compared with KICS effusions (p=0.0028). Serum IL-10 levels were also higher in PEL as compared with KICS (med 51.6 vs. 2.5ng/mL; p=0.0015). KSHV VL levels were significantly higher in PEL effusions as compared to KICS effusions (med 31,428,571 vs. 569 copies/mL; p=0.0005) and KSHV-MCD (med 231,884 copies/mL; p=0.02).

CONCLUSIONS: In HIV-infected patients, effusions can indicate a diagnosis of PEL, KSHV-MCD and/or KICS. Quantifying KSHV VL in the effusion may be useful in the diagnosis of PEL. Effusions in PEL had a distinct profile compared to the circulation or other KSHV-associated diseases, particularly with regard to elevated IL-13, which may aid in diagnosis. In contrast, the inflammatory and virologic findings in KSHV-MCD and KICS effusions roughly paralleled the levels seen in the circulation. This study suggests that KSHV upregulation of IL-13 is a unique feature of PEL, which may contribute to PEL pathogenesis through STAT6 activation and be a potential future therapeutic target.


Uldrick:Celgene: Patents & Royalties: 10,001,483 B2; Celgene: Research Funding; Merck: Research Funding. Yarchoan:Celgene Corp.: Research Funding; NIH: Patents & Royalties: Patents on IL-12 for KS and cereblon-binding drugs for KSHV diseases. Spouse has patent on KSHV IL-6. Patents assigned to DHHS/NIH..

Author notes


Asterisk with author names denotes non-ASH members.