Multicentric Castleman disease: viral and cellular targets for intervention

Kaposi sarcoma (KS) is now the most common malignancy in much of sub–Saharan Africa, due to widespread infection with KS-associated herpesvirus (KSHV, also known as human herpesvirus 8) and human immunodeficiency virus (HIV). Two rarer but nonetheless serious neoplasms, primary effusion lymphoma (PEL) and plasmablastic multicentric Castleman disease (MCD), also are caused by KSHV and markedly augmented by HIV.

Patients with MCD have lymphadenopathy typically with hepatosplenomegaly. Many are severely ill with chronic or intermittent malaise, fever, weight loss, night sweats, respiratory symptoms, peripheral edema, and skin rashes. Anemia, thrombocytopenia, elevations of γ-globulin and acute-phase reactants, hypoalbuminemia, and proteinuria occur in the majority of patients. About one quarter develop central nervous system disease. Especially with HIV coinfection, a substantial fraction progresses to PEL, plasmablastic, or anaplastic large-cell lymphomas.^1-3 

MCD and PEL arise from B cells, the primary reservoir for KSHV. In MCD-affected lymph nodes, KSHV-latent nuclear antigen can be detected consistently in immunoglobulin M (IgM)–expressing, λ-light-chain–restricted B-cell plasmablasts in the mantle zone.²  Interleukin-6 (IL-6) is a potent growth factor for plasmablasts and both human (huIL-6) and KSHV (vIL-6) homologues may contribute to its pathogenesis.³ 

Vinca alkaloids and other cytotoxic agents, with or without high-dose corticosteroids, produce frequent but short-lived responses. Similarly, treatment with an anti–huIL-6 monoclonal antibody temporarily relieved the signs and symptoms of MCD but not its underlying pathophysiology.⁴  Cidofovir has potent in vitro activity against KSHV replication, but no MCD responses have been reported.

MCD exacerbations are correlated with huIL-6, IL-10, and C-reactive protein serum levels and with the level (viral load) of KSHV DNA in peripheral blood cells. One patient obtained complete relief of severe MCD symptoms for at least 14 months following one dose of rituximab, an anti-CD20 monoclonal antibody.⁵  Rituximab targets B cells, a KSHV reservoir and probably a major source of the IL-6 and other soluble mediators of MCD. In the current issue, Marcelin and colleagues (page 2786) report on treating 5 more MCD patients with rituximab, 2 of whom died quickly with MCD complications and other conditions. The other 3 patients experienced complete clinical remissions of MCD that lasted 3, 6, and 12 months, respectively, after receiving 4 doses of rituximab. Peripheral blood B-cell and C-reactive protein levels fell abruptly in responding patients, whereas KSHV viral load declined inconsistently, illustrating the presence of non-CD20 viral reservoirs.

Some patients developed MCD despite complete suppression of HIV with highly active antiretroviral therapy (HAART) and CD4 counts above 200 cells/μL. Rituximab apparently had no drug interactions with HAART, nor did it have any effect on HIV viral load or CD4 count. However, because KS worsened in 2 patients, the risk of de novo or worsening KS in other rituximabtreated patients should be evaluated.

Although it is apparently effective and relatively safe, rituximab does not eliminate KSHV infection and almost certainly does not cure MCD. Rather, as with MCD responses following splenectomy or excision of a lymph node mass,⁴  rituximab reduces the burden of KSHV-infected cells and B-cell growth factors, particularly IL-10. More experience in treating MCD with rituximab is needed before it can be considered the standard of care.