Minimal change nephrotic syndrome in patients infected with human immunodeficiency virus: a retrospective study of 8 cases

Minimal change nephrotic syndrome (MCNS) is a glomerular disease characterized by massive selective proteinuria without glomerular lesions on light microscopy and with no immunoglobulin deposits on immunofluorescence study. MCNS has a higher incidence in children than in adults, but is, nevertheless, a frequent cause of nephrotic syndrome (NS) in adults, accounting for 10 to 25% of cases [1]. The pathogenesis of MCNS remains poorly understood, but there is compelling evidence, to suggest that it involves both immune system impairment and podocyte dysfunction [2]. In most cases, MCNS is considered to be an idiopathic glomerular disorder potentially triggered by immunological stimuli, such as viral infection, immunization or allergens [2]. MCNS may occur in association with chronic lymphoid neoplasms, such as Hodgkin disease [3], non-Hodgkin lymphoma [4] and thymoma [5], consistent with the hypothesis that it results from an immune system disorder. In the course of studies on molecular mechanisms underlying the pathophysiological processes involved in MCNS occurrence, we originally identified CMIP (c-maf inducing protein) as induced in T lymphocytes of patients with MCNS relapse but further studies have shown that CMIP is overproduced in podocytes of these patients, whereas it is downregulated in both tissues in MCNS remission [6, 7]. In addition, transgenic mice expressing selectively CMIP in podocytes develop heavy proteinuria without inflammatory lesions or immune complex deposits [6] and there is growing evidence that increase CMIP abundance could dramatically affect the function and survival of podocytes [2]. In rare cases, MCNS has also been reported in association with chronic viral infection [8, 9]. Human immunodeficiency virus (HIV) infection is a well-known cause of renal disease [10]. HIV-associated nephropathy (HIVAN) is the most common cause of glomerular injury in these patients [11]. A large number of studies have clearly demonstrated that HIVAN pathogenesis is triggered by the direct infection of glomerular epithelial cells and tubular epithelial cells by HIV [12]. In some studies, MCNS has been reported to account for 2–4% of all nephropathies associated with HIV [13, 14]. An extensive review of seven renal pathology studies including a total of 949 patients showed MCNS to be the main renal biopsy finding in 10 patients (1.05% of cases), suggesting that MCNS is a rare finding in HIV-infected patients [11]. It remains unclear whether there is a real pathophysiological link between MCNS and HIV infection, or just a fortuitous association.

In this study, we retrospectively identified patients with biopsy-proven MCNS and diagnosed with HIV infection, and we analyzed their clinical, histological, laboratory, and treatment data, to assess the significance of this rare association. We also performed in situ hybridization to determine whether HIV-RNA could be detected on renal biopsies from these patients and to assess possible CMIP expression, which has recently been shown to be associated with podocyte dysfunction [6].

HIV-related renal diseases include a large diversity of pathological entities, but the most relevant glomerular disorders observed in these patients are HIVAN and immune complex-mediated glomerular disease [11, 18]. The occurrence of MCNS in the setting of HIV infection has never been studied in detail and the main characteristics of this association remain to be determined. Over a 16-year period, we could find only eight patients with MCNS in a context of HIV infection, suggesting that this association may be fortuitous rather than indicative of a direct relationship. However, our study was not designed to accurately assess the prevalence of MCNS in HIV infected patients.

Genetic factors, including APOL1 variants, have recently emerged as key factors potentially increasing susceptibility to HIVAN in HIV-infected patients of African ancestry [19, 20]. In our study, APOL1 variants were not systematically checked, but the distribution of Caucasian and African patients in our cohort suggests that this was not a susceptibility factor for this association. Moreover a recent study suggested that two risk allele genotype is uncommon in patients with MCNS and mainly found in patient with FSGS steroid resistant nephrotic syndrome [21]. In a retrospective study, Lescure et al. described changes in the pattern of glomerular lesions in HIV patients [14]. They demonstrated that the incidence of HIVAN had strongly decreased over time, since the introduction of HAART, and that classical FSGS was the leading causes of glomerular disease in HIV-infected patients. Lescure et al. found that MCNS was the main pathological lesion in four patients (4.5% of cases). Strikingly, seven patients from our cohort were on HAART and HIV infection was considered to be under control at the onset of nephrotic syndrome. These data indicate that MCNS was probably not due to severe immunodeficiency in these patients. MCNS occurred exclusively in one patient newly diagnosed with HIV infection, with no obvious difference in clinical, biological and histological presentation compared to other patients in whom HIV infection was well controlled. In this specific case, we can not rule out that HIV infection in the context of immune susceptibility or particular genetic background might contribute to trigger MCNS relapse, as was recently reported for EBV infection [22]. In three patients, the diagnosis of MCNS led to change the antiretroviral drugs used, due to potential tenofovir-mediated renal toxicity in two cases and AKI in one case. Tenofovir administration has been identified as a potential cause of tubular injury, sometimes associated with significant but non-nephrotic proteinuria [23]. However, none of our patients on this drug had typical proximal tubular dysfunction or pathological lesions suggestive of toxic tubular necrosis. In a study of 95 adult patients with idiopathic MCNS, Waldman et al. found that AKI was present at initial presentation or during a relapse episode in 25.2% of cases, mostly due to acute tubular injury or interstitial inflammation [24]. Three patients from our cohort displayed AKI according to KDIGO criteria but, in these cases, analyses of renal biopsies provided no evidence of interstitial inflammatory lesions or tubular injury, suggesting a prerenal origin of AKI.

We investigated the potential role of HIV infection in MCNS occurrence in this context, by performing ISH to determine whether the renal cells of these patients were infected. By contrast to HIVAN patients, in which large numbers of tubular and glomerular cells gave a positive signal, no HIV RNA was detected in the cells of two patients from our cohort with viral loads below the threshold of detection. Nevertheless, ISH detected a few tubular cells displaying positive staining, with a complete absence of staining in the glomerular area, in two other patients with MCNS in a context of HIV infection. In one case, HIV RNA was detected in the renal parenchyma, despite undetectable viremia. Winston et al. were the first to demonstrate, in a patient diagnosed with HIVAN that HIV-1 mRNA may persist in the renal epithelium even if viral RNA is undetectable in the plasma, following the initiation of HAART [25]. Consistent with this hypothesis, Canaud et al. found that kidney graft parenchyma may serve as a viral reservoir after kidney transplantation [17]. Another key finding of our study is the induction of CMIP expression in the podocytes of patients with MCNS in a context of HIV infection. CMIP upregulation has been described in renal biopsies from patients with idiopathic MCNS, FSGS and membranous nephropathy, whereas CMIP is not generally expressed in inflammatory and proliferative glomerular diseases, such as IgA nephropathy or active lupus nephritis [6, 26, 27]. In vivo and in vitro experiments have shown that CMIP plays a crucial role in podocyte dysfunction, by inactivating the nephrin and Akt signaling pathways responsible for altering podocyte signaling and cytoskeleton remodeling, an important event in the induction of proteinuria [6]. In addition, experimental data from several models of podocyte injury suggest that targeting CMIP with specific inhibitors taken up by podocytes may be a promising future therapeutic approach for some glomerular diseases associated with the overproduction of CMIP in podocytes [6, 26]. By contrast, in patients with HIV infection, pathophysiological processes involved in HIVAN seems to be closely related to viral integration into target cells. Thus, the expression of HIV proteins, such as Nef, in particular, in podocytes, leads to the activation of STAT3 signaling, which seems to be a key factor in HIVAN pathogenesis [28, 29]. Thus, the absence of HIV RNA in glomerular cells and the induction of CMIP in podocytes from patients with MCNS in a context of HIV infection suggest that the occurrence of MCNS in individuals with HIV infection may be coincidental and unrelated. As MCNS may primarily result from an immune system disorder [2], chronic HIV infection may triggers immune disturbances in some patients, promoting podocyte dysfunction.

In accordance with current recommendations for the treatment of initial episodes of idiopathic MCNS in adult patients [1], first-line treatment with steroids was initiated in six patients, leading to CR in five cases (the remaining patient being lost to follow-up). Two of the patients in our cohort presented spontaneous remission, consistent with the findings of Maas et al., who reported the occurrence of spontaneous progressive remission in 10% of patients not receiving specific therapy [30]. After a median of 20 months of follow-up, five of the seven patients for whom data were available presented relapses of MCNS (including three patients with steroid-dependent MCNS). Strikingly, rituximab was the treatment of choice most frequently prescribed for patients with steroid-dependent MCNS. It was also prescribed for one patient with frequent relapses. Rituximab has recently emerged as a promising treatment for steroid-dependent nephrotic syndrome [31‐33]. In a context of HIV infection, Rituximab is primary used for the treatment of patients with HIV related lymphoma and Multicentric Castleman’s disease [34]. However, Rituximab may increase the risk of infectious episodes in HIV patients. Our retrospective study was not designed to investigate the safety and tolerance of Rituximab in this population. In our cohort, non-fatal bacterial infections occurred in two patients on rituximab. Given that rituximab administration has been associated with an increase in the risk of hepatitis B reactivation, opportunistic infections and polyomavirus JC-related progressive multifocal leukoencephalopathy [35], this treatment should probably be used with caution in HIV-infected patients.

MCNS may be observed in HIV patients, but seems to be a rare glomerular disease in these patients. Our pathology studies suggested that HIV infection was not directly involved in podocyte dysfunction and that CMIP induction, through unknown mechanisms, might play a key role in podocyte dysfunction and the occurrence of proteinuria. First-line treatment for MCNS in patients with HIV infection should include steroids, but, given the high incidence of MCNS relapse and steroid dependence, the potential contribution of rituximab to treatment should be investigated further.