Towards a better understanding of Kaposi sarcoma-associated immune reconstitution inflammatory syndrome

Abstract

After antiretroviral therapy (ART) initiation, some patients experience clinical deterioration despite increased CD4 T-lymphocyte counts and decreased plasma HIV viral loads [1]. This phenomenon has been described as immune reconstitution inflammatory syndrome (IRIS) and results from restored immunity to specific infectious or noninfectious antigens [2]. A low CD4 T-cell count at the start of ART has been consistently associated with a high risk of IRIS, particularly in those patients with CD4 T-cell counts below 50 cells/μl. The proportion of patients with IRIS reported in the literature ranges from 10% to more than 50% [3]. The wide variation in estimated rates of IRIS can be partially explained by real differences between populations; differences in the study design [4]; the prevalence of opportunistic disease [5]; the failure to distinguish between paradoxical and unmasking forms of IRIS; and last, but not least, the IRIS case definition. With the exception of IRIS in the context of tuberculosis (TB) [6] and cryptococcosis [7], there is no consensus case definition. Some of the most common IRIS-defining diagnoses are cytomegalovirus infection, disseminated Mycobacterium avium intracellulare, Pneumocystis pneumonia, varicella zoster, Kaposi sarcoma and non-Hodgkin lymphoma [8]. Most descriptions of Kaposi sarcoma-associated IRIS have been case reports or case series from literature review [9]. An incidence of 7% of Kaposi sarcoma-associated IRIS was originally reported by Bower et al.[10] in a prospective cohort of HIV-infected patients with Kaposi sarcoma who initiated ART in London; although in another prospective study [11] from Mozambique, the incidence of Kaposi sarcoma-associated IRIS was 31%. In a more recent cohort study including patients located in France and French overseas territories, it was found that the risk for developing Kaposi sarcoma was particularly high during the first 3 months on ART. CD4 cell counts on ART initiation were significantly lower in patients who developed Kaposi sarcoma. In this study, patients with a history of Kaposi sarcoma were not included [12]. At first sight, it is evident that the aforementioned studies can hardly be compared due to differences in study design, and as pointed out by Letang et al.[13] in a previous publication, by the lack of a uniform Kaposi sarcoma-associated IRIS case definition. In this issue, Letang et al.[14] published the largest and most comprehensive study exploring the incidence, possible risk factors and clinical outcomes of Kaposi sarcoma-associated IRIS. Distinctive features of this study were the inclusion of three prospective cohorts of ART-naive HIV-infected patients with Kaposi sarcoma constituting a large multinational cohort; examination of the contribution of Kaposi sarcoma-associated herpes virus (KSHV) viral load to Kaposi sarcoma-associated IRIS development and outcome; and the use of a standardized Kaposi sarcoma-associated IRIS case definition involving abrupt Kaposi sarcoma worsening after ART, time from ART (3 months), viral suppression, agreement of two primary investigators and exclusion of other causes of worsening. The incidence of Kaposi sarcoma-associated IRIS was higher in the African than in European cohorts (19.6 vs. 8.5%). The overall mortality rate was 3.3-fold higher in the African than in the London cohorts, and there were no Kaposi sarcoma-associated IRIS related deaths in the European cohort. However, even when the cause of death was known only for 11 of 19 patients and 54.5% was attributed to Kaposi sarcoma-associated IRIS, the authors considered that these findings might be mainly explained by the higher rate of Kaposi sarcoma-associated IRIS and the lack of adjunctive chemotherapy in the African cohorts. In this regard, it would be interesting to explore if other sub-diagnosed, concomitant opportunistic infections such as TB might also contribute to the elevated mortality of these patients. In fact, cutaneous colesional occurrence of AIDS-Kaposi sarcoma and microscopic TB has been reported [15]. Mycobacterium tuberculosis is able to induce pro-inflammatory responses [16], and it remains to be elucidated whether this would be a favourable microenvironment for Kaposi sarcoma progression. Another potential factor for mortality in the population studied by Letang et al. could be the common use of steroids for the treatment of an opportunistic infection, as the use of prednisone or dexamethasone has been associated with accelerated clinical progression of Kaposi sarcoma in HIV-infected patients [17]. A detectable baseline plasma level of KSHV DNA was identified by Letang et al. as an independent predictor of Kaposi sarcoma-associated IRIS development. However, it should be considered that the prevalence of KSHV infection is lower in London than in sub-Saharan Africa, where KSHV was endemic before the AIDS epidemic [18]. In South Africa, one in three adults are KSHV seropositive [19]. Therefore, the prevalence of KSHV infection should be a major determining factor for the number of Kaposi sarcoma cases. The risk factors for Kaposi sarcoma-associated IRIS, as well as the criteria for administration of chemotherapy in patients with high probability of Kaposi sarcoma-associated IRIS, have yet to be defined. As stated by the authors, the T1 Kaposi sarcoma stage might be a predictor of sarcoma-associated IRIS. This investigation has gone some way towards enhancing our understanding of Kaposi sarcoma-associated IRIS. The development of particular consensus case definitions for the different forms of HIV-associated IRIS will contribute to the rapid clinical diagnosis of these conditions and will also enable comparison of studies, pooling of data and meta-analyses. IRIS-associated morbidity and mortality is thus affected by the effectiveness of early HIV-diagnosis programmes; the prevalence of opportunistic infections presented as active diseases or as latent infections in the different geographical regions; the accessibility of facilities for the rapid diagnosis and treatment of such conditions before ART initiation; and the quality of the medical infrastructure. After all, we should never forget that IRIS is a direct consequence of late presentation for HIV care. Acknowledgements This work was supported by grants from the Mexican Government (Comisión de Equidad y Género de la Honorable Cámara de Diputados de la LXI Legislatura de México) and Fundación México Vivo (http://http://www.mexicovivo.org/). The funders had no role in preparation of the manuscript. Conflicts of interest The authors declare no conflicts of interest.