Skip to content


Infectious Agents and Cancer

Open Access

KSHV activation, human and viral IL-6 production, and other cytokine dysregulation: Association with the symptomatology of KSHV-associated multicentric Castleman's disease

  • Mark N Polizzotto1Email author,
  • Thomas S Uldrick1,
  • Victoria Wang1,
  • Karen Aleman1,
  • Kathleen M Wyvill1,
  • Vickie Marshall3,
  • Stefania Pittaluga2,
  • Dierdre O'Mahony1,
  • Denise Whitby3,
  • Giovanna Tosato4,
  • Seth M Steinberg5,
  • Richard F Little1 and
  • Robert Yarchoan1
Infectious Agents and Cancer20105(Suppl 1):A47

Published: 11 October 2010


Viral LoadPeripheral Blood Mononuclear CellStrong PredictorInterferon GammaMultiple Linear Regression Model


KSHV-associated multicentric Castleman’s disease (MCD) is a frequently fatal lymphoproliferative disorder characterized by inflammatory flares of fever, cytopenias, hypoalbuminemia, hyponatremia, and splenomegaly. Most cases occur in HIV-infected patients. KHSV viral interleukin-6 (vIL-6), human IL-6 (hIL-6), and possibly other proinflammatory cellular cytokines are believed to contribute to the pathophysiology of MCD flares.


We identified MCD patients with clinical flares. KSHV viral load (VL) in peripheral blood mononuclear cells, vIL-6, and the cellular cytokines IL-6, IL-1b, IL-8, IL-10, IL-12p70, interferon gamma, and tumor necrosis factor alpha were measured during flares and remissions to identify parameters best characterizing flares. The assay for vIL-6 was modified from Aoki Y. et al., Blood, 97, 2526, 2001; the cutoff of detection was 1560 pg/ml. Factors statistically associated with flares (p<0.01) were explored in relationship to common disease manifestations with multiple linear regression models.


20 patients (18 male, 2 female) were studied during 33 flares (range 1-3 per patient) and, in 18 patients, remission with therapy. Median (range) values of key parameters during flares included hemoglobin 9.9 mg/dL (6.8-14.4); platelet count 97 K/μL (6-377); sodium 133 mEq/L (127-143); albumin 2.7 mg/dL (1.2-3.9); spleen size 14.5 cm (9-28); temperature 38°C (36.1-40.5); CD4 count 240 cells/μL (24-1319); HIV VL <50 copies/mL (<50-64100). Flares were associated with elevated KSHV VL (median 23448 copies/mL; range 0-3913043; p<0.0001 compared with remission), vIL-6 (2575 pg/mL; <1560-20497; p=0.0039), hIL-6 (24.2 pg/mL; 1.4-171.5; p=0.0034), hIL-10 (783.9 pg/mL; 2.8-26021; p=0.0027), and hIL-1b (1.3 pg/mL; 0-11.3; p=0.0074). In two of the 33 flares vIL-6 was elevated but hIL-6 was not; in 14 hIL-6 was elevated but vIL6 was undetectable; and in 15 both were elevated. Neither was initially elevated in 2 flares, but hIL-6 later became elevated in both. Disease manifestations did not differ among flares with differing vIL-6/hIL-6 profiles. In multiple regression analysis, elevated KSHV VL was the strongest predictor of level of hemoglobin (p<0.0001), sodium (p<0.0001), albumin (p<0.0001), and spleen size (p=0.0002); hIL-6 the strongest predictor of thrombocytes (p=0.0011), and KSHV and hIL-6 together the strongest predictors of body temperature (p<0.0001). For hemoglobin, but not other parameters, vIL-6 and hIL-6 in combination were stronger predictors than either independently (p=0.0002), though less strong than KSHV VL alone.


KSHV activity, vIL-6 production, and associated human hIL-6 dysregulation are key determinants of the clinical manifestations of MCD. vIL-6 and hIL-6 each appear sufficient to induce flares without the other. hIL-10 and hIL-1ß are also elevated in MCD flares, but their contribution to symptomatology remains to be determined.



This article has been published as part of Infectious Agents and Cancer Volume 5 Supplement 1, 2010: Proceedings of the 12th International Conference on Malignancies in AIDS and Other Acquired Immunodeficiencies (ICMAOI).The full contents of the supplement are available online at

Authors’ Affiliations

HIV and AIDS Malignancy Branch, Center for Cancer Research, National Cancer Institute, Bethesda, USA
Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, Bethesda, USA
Viral Oncology Section, AIDS and Cancer Virus Program, National Cancer Institute, Bethesda, USA
Laboratory of Cellular Oncology, Center for Cancer Research, National Cancer Institute, Bethesda, USA
Biostatistic and Data Management Section, Center for Cancer Research, National Cancer Institute, Bethesda, USA


© Polizzotto et al; licensee BioMed Central Ltd. 2010

This article is published under license to BioMed Central Ltd.