Share this article on:

Immune reconstitution inflammatory syndrome associated with Kaposi sarcoma: higher incidence and mortality in Africa than in the UK

Letang, Emilioa,b; Lewis, James J.c; Bower, Markd; Mosam, Anisae; Borok, Margarethf; Campbell, Thomas B.g; Naniche, Denisea,b; Newsom-Davis, Tomd; Shaik, Fahmidae; Fiorillo, Suzanneg; Miro, Jose M.h; Schellenberg, Davidi; Easterbrook, Philippa J.e

doi: 10.1097/QAD.0b013e328360a5a1
Clinical Science

Objectives: To assess the incidence, predictors, and outcomes of Kaposi sarcoma-associated paradoxical immune reconstitution inflammatory syndrome (KS-IRIS) in antiretroviral therapy (ART)-naive HIV-infected patients with Kaposi sarcoma initiating ART in both well resourced and limited-resourced settings.

Design: Pooled analysis of three prospective cohorts of ART-naive HIV-infected patients with Kaposi sarcoma from sub-Saharan Africa (SSA) and one from the UK.

Methods: KS-IRIS case definition was standardized across sites. Cox regression and Kaplan–Meier survival analysis were used to identify the incidence and predictors of KS-IRIS and Kaposi sarcoma-associated mortality.

Results: Fifty-eight of 417 (13.9%) eligible individuals experienced KS-IRIS with an incidence 2.5 times higher in the African vs. European cohorts (P = 0.001). ART alone as initial Kaposi sarcoma treatment (hazard ratio 2.97, 95% confidence interval (CI) 1.02–8.69); T1 Kaposi sarcoma stage (hazard ratio 2.96, 95% CI 1.26–6.94); and plasma HIV-1 RNA more than 5 log10 copies/ml (hazard ratio 2.14, 95% CI 1.25–3.67) independently predicted KS-IRIS at baseline. Detectable plasma Kaposi sarcoma-associated herpes virus (KSHV) DNA additionally predicted KS-IRIS among the 259 patients with KSHV DNA assessed (hazard ratio 2.98, 95% CI 1.23–7.19). Nineteen KS-IRIS patients died, all in SSA. Kaposi sarcoma mortality was 3.3-fold higher in Africa, and was predicted by KS-IRIS (hazard ratio 19.24, CI 7.62–48.58), lack of chemotherapy (hazard ratio 2.35, 95% CI 1.09–5.05), pre-ART CD4 cell count less than 200 cells/μl (hazard ratio 2.04, 95% CI 0.99–4.2), and detectable baseline KSHV DNA (hazard ratio 2.12, 95% CI 0.94–4.77).

Conclusion: KS-IRIS incidence and mortality are higher in SSA than in the UK. This is largely explained by the more advanced Kaposi sarcoma disease and lower chemotherapy availability. KS-IRIS is a major contributor to Kaposi sarcoma-associated mortality in Africa. Our results support the need to increase awareness on KS-IRIS, encourage earlier presentation, referral and diagnosis of Kaposi sarcoma, and advocate on access to systemic chemotherapy in Africa.

Supplemental Digital Content is available in the text

aBarcelona Centre for International Health Research (CRESIB, Hospital Clínic-Universitat de Barcelona), Barcelona, Spain

bCentro de Investigação em Saúde de Manhiça (CISM), Manhiça, Maputo Province, Mozambique

cDepartment of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine (LSH&TM)

dDepartment of Oncology, Chelsea and Westminster Hospital, London, UK

eDepartment of Infectious Diseases, Nelson R Mandela School of Medicine, University of KwaZulu Natal, Durban, South Africa

fDepartment of Medicine, University of Zimbabwe College of Health Sciences, Harare, Zimbabwe

gDivision of Infectious Diseases, Department of Medicine, University of Colorado Denver, Aurora, USA

hInfectious Diseases Service, Hospital Clínic/IDIBAPS, Universitat de Barcelona, Barcelona, Spain

iDepartment of Disease Control, London School of Hygiene and Tropical Medicine (LSH&TM), London, UK.

Correspondence to Emilio Letang, Barcelona Centre for International Health Research (CRESIB, Hospital Clínic-Universitat de Barcelona), Rosselló 132, 4th floor, Barcelona 08036, Spain. Tel: +34 932 275 400x4123; e-mail:

Received 31 October, 2012

Revised 17 February, 2013

Accepted 26 February, 2013

Preliminary results of this study were presented at the 18th Conference on Retroviruses and Opportunistic Infections, CROI 2011, Boston, USA, poster #857.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (

© 2013 Lippincott Williams & Wilkins, Inc.