Clinicopathological study of glomerular diseases associated with sarcoidosis: a multicenter study

Sarcoidosis is a chronic multisystemic inflammatory disease of unknown origin, characterized by the presence of non-caseating, epitheliod granulomas in some tissues leading to organ dysfunction [1, 2]. Sarcoidosis typically affects young adults and its clinical course ranges from spontaneous resolution to chronic progressive disease. The severity and diversity of the clinical manifestations related to sarcoidosis depend on the extent of the infiltrating granulomatous lesions, which preferentially involve the lower respiratory tract. Granulomatous lesions can also affect extrapulmonary sites, such as the lymph nodes, heart, kidneys, central nervous system, liver, spleen, larynx and eyes [2, 3]. Pathophysiological mechanisms involved in the formation, maintenance or spontaneous resolution of sarcoidosis granuloma thought to be linked to genetic susceptibility and to unidentified environmental antigens which trigger an uncontrolled cell-mediated immune reaction involving macrophages and CD4 type 1 helper T (Th1) cells [2, 4, 5].

The prevalence of kidney impairment in sarcoidosis patients ranges from 10% to 20% of cases [6]. Renal injury is most commonly due to disorders in calcium homeostasis, with renal stone disease and nephrocalcinosis [6]. Granulomatous tubulointerstitial nephritis (GTIN) is a less common cause of renal lesions that occurs in approximately 20% of patients with sarcoidosis [7], and benefits from steroid therapy [8, 9]. By contrast, glomerular diseases have been rarely reported in patients with sarcoidosis [6]. Most publications reporting glomerular diseases in the context of sarcoidosis consisted exclusively in case reports. The nature of glomerular lesions seems to be diverse, with membranous nephropathy (MN) [10‐14]; minimal change nephrotic syndrome (MCNS) [15‐17]; focal segmental glomerulosclerosis (FSGS) [18‐20], imunoglobulin A nephropathy (IgAN) [21‐24] and proliferative glomerulonephritis [13, 25, 26].

In some sarcoidosis patients, the kidneys may be major sites of inflammation and granuloma formation, which leads to significant renal impairment [7]. In such cases, it is likely that infiltration of activated T cells and macrophages into the interstitium results in granuloma formation and interstitial nephritis. Alternatively, in rare cases of glomerular injury in the context of systemic sarcoidosis, the underlying molecular mechanisms of this association remain unclear. The aim of this retrospective was not to determine the incidence and/or the prevalence of glomerular involvement compared to interstitial nephritis related to sarcoidosis but exclusively to describe the spectrum of glomerular lesions observed in patients with sarcoidosis.

Our study demonstrated that a wide spectrum of glomerular lesions may be found in sarcoidosis patients. As these glomerular lesions do not share common pathophysiological mechanisms, the existence of a direct molecular link between sarcoidosis and glomerular pathology remains uncertain. Nevertheless, the close temporal relationship exhibited by nine patients, strongly suggests that glomerular disease and sarcoidosis may be linked in these patients.

Sarcoidosis is characterized by an infiltration of Th1 cells and macrophages into sites of inflammation and by a dominant expression of Th1 cytokines with low levels of Th2 cytokines [38, 39]. Recent studies showed that sarcoidosis is associated with an expansion of a regulatory T cell subset with antiproliferative activity [40]. Most glomerular diseases found in patients with sarcoidosis are thought to be caused by functional disorders of the immune system. Compelling evidence suggests that primary MCNS and FSGS with relapse also result from perturbations of immune system [41]. Thus, we can postulate that cytokines produced locally by granulomas may contribute to alterations in the glomerular filtration barrier. Indeed, increased production of TNF-alpha by granuloma macrophages and activated T cells has been reported [42]. Experimental observations suggest that TNF-alpha may increase the permeability of the glomerular filtration barrier [43]. In accord with previous observations, we found that MN was the most frequent glomerular disease associated with sarcoidosis [6, 10] and often preceded the sarcoidosis diagnosis. Sex ratio, mean age, and proportion of patients with nephrotic syndrome in patients with MN associated with sarcoidosis seem to be similar to those in previous series of patients with idiopathic membranous nephropathy [44‐47]. The molecular link between these two conditions remains unclear since no target antigen or specific antibodies have yet been identified. Circulating immune complexes and hypergammaglobulinemia are a common finding in sarcoidosis patients [27]. Recently, Knehtl et al. described a patient with sarcoidosis and MN associated with anti-phospholipase A2 receptor antibodies (anti-PLA2R), suggesting that primary and secondary MN may share certain pathophysiological processes [14]. Three patients displayed IgAN simultaneously with active sarcoidosis. This close temporal relationship strongly suggests that immune dysregulation in sarcoidosis may lead to IgA deposition in the glomerular mesangium. Indeed, cytokine production may also play a critical role in the production and glycosylation of systemic IgA leading to IgAN [48]. The coexistence of SLE and sarcoidosis seems to be a very rare finding [49] even if a recent study suggests an increase of prevalence of SLE among sarcoidosis patients [50]. Weinberg et al. showed that the presence of anti-DNA antibodies in the serum of sarcoidosis patients did not predict the subsequent development of SLE [51]. We described here two interesting cases of lupus nephritis occurring in steroid-resistant sarcoidosis. In these two patients, we cannot exclude the possibility of a fortuitous association without known pathophysiologic relationship between both conditions. Nevertheless, these two cases emphasize the need to systematically screen sarcoidosis patients with renal impairment for SLE antibodies.

Mahevas et al. have previously reported that GTIN lesions occurred simultaneously with the first clinical manifestations of sarcoidosis in 81% of these cases [8]. In our study, only nine patients (35%) displayed biopsy-proven glomerular disease occurring simultaneously with sarcoidosis. Our study demonstrated that all types of glomerulopathies, except lupus nephritis, may be identified in patients with a first relapse of sarcoidosis. Among 11 patients with sarcoidosis that preceded the diagnosis of glomerular disease, none displayed sarcoidosis relapse at the time of renal biopsy. Nevertheless, three patients had untreated persistent sarcoidosis and two others presented steroid-resistant or steroid-dependent sarcoidosis. These five additional cases suggest that persistent granuloma activity may secondarily promote the development of glomerular disease.

Regardless of the types of glomerular lesions found in kidney biopsies, one major issue from our study was to analyze the response to steroid therapy in patients with glomerular diseases occurring simultaneously with sarcoidosis. A recent survey conducted in 47 patients showed that two criteria (the initial fibrotic score at renal biopsy and the response after one month of steroid therapy) were significantly associated with the renal prognosis of patients with GTIN related to sarcoidosis [8]. In our study, exclusive steroid therapy was administered to eight of the nine patients with simultaneous glomerular disease and sarcoidosis. Complete remission of both diseases occurred in a single case (pt7) whereas partial remission of the glomerulopathy and complete disappearance of sarcoidosis related symptoms occurred in three patients (pt5, pt8 and pt9). Despite steroid treatment, one patient (pt 4) displayed both persistent sarcoidosis and significant proteinuria. Finally, we found that sarcoidosis was frequently resistant to or dependent on steroid treatment (4/8 patients). However, the failure of steroid therapy to control sarcoidosis symptoms did not preclude the remission of glomerular disease in three patients (pt1, pt 3 and pt6). These observations are different from those of existing case reports, in which glomerular disease activity appeared to be influenced by the response of the sarcoidosis to steroid therapy [17, 20, 24, 52]. Altogether, these observations show that the steroid responsiveness of sarcoidosis does not correlate with that of glomerular disease and suggest that both diseases require possibly a more aggressive therapy. Indeed, with a mean follow-up of 101 months, six of the 26 patients (23%) had chronic renal impairment requiring intermittent hemodialysis and/or renal transplantation, although their mean GFRs at the time of renal biopsy were not significantly impaired. In contrast, among patients with isolated GTIN who were followed for a median of 84 months, Rajakariar et al. demonstrated that exclusive steroid treatment may significantly improve GFR [9]. Similar results were found by Mahevas et al., suggesting that glomerular lesions are less steroid-sensitive than interstitial injury [8].

Our study showed that the spectrum of glomerular lesions observed in the context of sarcoidosis is heterogeneous and that only 35% of the patients displayed both diseases simultaneously. For these cases, our study suggests that exclusive steroid therapy is less efficient that previously reported. Therefore, optimal therapeutic management can only be determined after further investigations.