Moderator's view: The use of calcineurin inhibitors in the treatment of lupus nephritis

Abstract

Lupus nephritis (LN) is one of the most severe manifestations of systemic lupus erythematosus (SLE), affecting ∼50% of patients, and both renal disease and treatment-related toxicity contribute to significant morbidity and mortality. Although our understanding of the aetiopathogenesis of LN is improving, treatment still remains a challenge, with the achievement of complete remission at 1 year in <50% of patients treated with current standard of care immunosuppressive therapy; this is associated with considerable short- and long-term side effects, some of which further contribute to non-adherence. Calcineurin inhibitors (CNIs) have been successfully used in organ transplantation and there is increasing evidence that cyclosporin A (CSA), and especially tacrolimus (TAC), are also effective in the treatment of LN. Randomised controlled trials showed similar efficacy for TAC when compared with mycophenolate mofetil (MMF) and multitarget therapy, including TAC and low-dose MMF, and resulted in significantly more complete remissions and overall responses compared with intravenous cyclophosphamide (CYC). Flares are observed in up to 45% of patients with LN, and an increase in relapse rate following induction with CNIs may be an issue. Most studies on this matter have been restricted to patients from Asia, and studies in more balanced cohorts are desirable. Moreover, there is a need to understand and determine the long-term effects of CNIs on renal function, proteinuria and comorbidities, with a special focus on cardiovascular risk. In this ‘Pros and Cons’ debate, the potential benefits and disadvantages of CNIs in the treatment of LN will be critically highlighted.

INTRODUCTION

The main goals of treatment in lupus nephritis (LN) are the preservation of renal function and prevention of flares by minimising treatment-related short-term and long-term side effects. Initial treatment requires intensive immunosuppressive therapy aimed at achieving complete, or at least partial, renal response followed by a less intensive maintenance therapy. Early renal response to immunosuppression is of prognostic value for a good long-term renal outcome. Side effects, particularly from corticosteroids, limit patient adherence, with subsequent impact on treatment efficacy. The current standard of care (SOC) is suboptimal and new therapeutic options, especially steroid-sparing regimens, are of great interest.

Observational studies addressing new biologic therapies have been promising, especially concerning the use of rituximab in refractory, relapsing or SOC-intolerant patients [1]. However, a recent randomised controlled trial (RCT) with rituximab, which aimed to improve clinical response in either proliferative (class III or IV) or membranous (class V) LN, failed to achieve the primary efficacy end point [2]. Moreover, ocrelizumab exhibited an imbalance of serious infections when used on top of SOC, leading to early termination of the BELONG trial [3], and atacicept (ATA LN trial) was associated with severe hypogammaglobulinaemia and subsequent serious infections when combined with MMF and corticosteroids [4].

A number of factors may have confounded the clinical trials of new lupus therapies. The optimisation of trial design and end point definition, as well as disease assessment, are ongoing matters of discussion in LN. Critical aspects were recently discussed in a review [5], including the following factors: (i) the definition of the key outcome measures, speculating that a more liberal proteinuria definition for complete remission may have resulted in a significant response rate in patients treated with abatacept [6]; (ii) the ‘on top’ design adding biologics on the SOC ‘background’ immunosuppressive therapy, particularly on concomitant glucocorticoids; (iii) the too optimistic assumption of expected differences in response rates between groups [these factors may have influenced the results of the LUNAR trial, where rituximab failed to meet its primary end point of 20% superiority in the proportion of patients achieving remission at 52 weeks when rituximab was given on top of SOC (MMF and corticosteroids) [2]]; (iv) short study durations and (v) the impact of ethnicity regarding drug response may have affected study results in the ALMS trial (a lower number of African Americans were recruited to the ALMS trial compared with the initial non-inferiority trial [7, 8]). The circumstances highlighted above leave us with the disappointing situation that, despite great effort, no new immunosuppressive drugs for the treatment of LN have been licensed so far.

Calcineurin inhibitors (CNIs), namely cyclosporin A (CSA) and tacrolimus (TAC), have emerged as treatment options in patients with LN; however, their use remains controversial, as will be discussed in the Polar Views in Nephrology (Pro and Con) papers by Dr Mok and Drs Fernandez Nieto and Jayne. We have been assigned to summarise their opposing views and would like to focus on the potential direct anti-proteinuric effects of CNIs, as well as on the efficacy of CSA and TAC for the preservation of renal excretory function, both in the initial and maintenance phases. In addition, we will comment on the impact of CNIs on cardiovascular risk factors and refer to remaining uncertainties and future directions.

ANTI-PROTEINURIC EFFECTS OF CALCINEURIN INHIBITORS

The first evidence for a direct effect of CSA on the prevention of proteinuria in murine LN was provided by Bergijk et al. [9], who demonstrated that early treatment with high-dose CSA (250 mg/kg/week) completely prevented the onset of proteinuria and glomerulosclerosis. Interestingly, the effect of CSA treatment was not immunologically mediated, but was instead associated with a stabilisation of the actin cytoskeleton by blocking the CNI-mediated dephosphorylation of synaptopodin [10]. In addition, in MRL/lpr mice, TAC treatment resulted in a reduction of proteinuria, improvement of renal function, preservation of podocyte numbers and the restoration of synaptopodin expression [11]. In a mouse model of nephrotic syndrome, calcineurin (CN) was increased in glomeruli during nephrosis. Suppression of CN along with amelioration of proteinuria could also be observed following TAC treatment [12]. Thus, direct effects on podocytes may account for an early reduction of proteinuria in patients treated with CNIs. These different effects may not be observed when using conventional immunosuppressive drugs.

INITIAL THERAPY

Cyclosporin A experience in lupus nephritis

Recent experience comes from a multicentre RCT from the Czech Republic/Slovakia that recruited patients with newly diagnosed class III or class IV LN. After 9 months of follow-up, no difference in partial response was observed between patients treated either with CYC or with CSA (11 of 21 versus 8 of 19, respectively), while complete remission at 18 months was higher in the CSA arm. However, significantly more patients met the criteria for an improvement of renal function in the CYC arm [13]. During an open follow-up of 7.7 years, 68% of patients in the CSA arm and 47% in the CYC arm required blood pressure–lowering drugs. Two malignancies were observed in the CSA-treated patients and none in the control group [14]. In another study from the USA, 42 patients with pure membranous LN were randomised to either prednisolone monotherapy or prednisolone in combination with adjunctive regimens (CYC or CSA). After 12 months of follow-up, the probabilities of remission were 27% (prednisolone monotherapy), 60% (CYC) and 83% (CSA). However, patients assigned to CSA treatment developed significantly more relapses of nephrotic syndrome during long-term follow-up. Two patients in the CSA group developed new-onset diabetes, and dose adjustments were necessary in 9 of 12 patients due to an increase in serum creatinine or hypertension [15]. As outlined by Dr Mok, FK-506 (TAC) is a compound that is 10- to 100-fold more potent than CSA [16], and TAC-based immunosuppression became the SOC in renal transplantation [17].

Tacrolimus and multitarget therapy in lupus nephritis

In a small RCT (total number of patients = 40), an earlier response of proteinuria was observed in patients treated with steroids and TAC (initial trough level 6–8 ng/mL, adapted after induction and when kidney function was impaired) when compared with steroid- and CYC-treated subjects (750 mg/m²/month for six pulses), and there were also significantly more complete remissions at 12 months in the TAC regimen group [18]. A multicentre study from China compared the efficacy of TAC (trough level 5–10 ng/mL) to intravenous CYC (750 mg/m²/month, adjusted, for six pulses). Overall, 81 patients were included with a median proteinuria <0.5 g/day. Most patients showed a proliferative pattern (mainly class IV LN) on renal histopathology. A trend towards a better overall response (complete and partial remission) and a significantly marked decrease in proteinuria was observed after induction treatment [19] with TAC. Another Chinese study recruited 150 patients to receive either TAC or MMF. The rate of complete responses was comparable between both groups. However, creatinine clearance improved (+14 mL/min) with MMF, whereas it remained stable with TAC (–0.8 mL/min). Subgroup analysis of pure membranous LN indicated a trend towards a better response in the TAC group (100% remission versus 75% in the MMF group) [20]. This is in contrast to the study by Yap et al. [21], who investigated a small number of patients with pure membranous LN, observing significantly more complete responses in patients treated with MMF compared with those receiving TAC (57.1% versus 11.1%; P = 0.049).

Multitarget therapy (comprising TAC and low-dose MMF) compared with intravenous CYC demonstrated higher rates of complete remission after 6 and 9 months in a small, prospective study. Repeat biopsy after 6 months revealed no signs of CNI-induced nephrotoxicity, while activity indices significantly decreased [22]. However, biopsies after a short-term follow-up may not detect chronic damage attributable to CNIs. Nonetheless, this study paved the way for a larger RCT. In general, after 24 weeks, significantly more complete remissions and overall responses were achieved with multitarget therapy when compared with treatment with intravenous CYC. However, the results need to be interpreted with caution, since complete follow-up was available only in 65 of 362 included patients. In addition, there was an excess of serious adverse events observed in the multitarget therapy group (10 withdrawals in the multitarget therapy group compared with only three in the CYC group) [23]. The burden of adverse events caused by multitarget therapy was also highlighted by a small retrospective study from the USA performed by Lanata et al. [24], which used multitarget therapy for the treatment of refractory disease. In this series, remission was achieved in four of seven patients; however, diabetic ketoacidosis, pneumonia and severe muscle pain limited its further use.

Side effects of TAC, especially tremor, were also limiting factors in two large RCTs; of note, there was no excess related to new-onset diabetes mellitus (DM) in these studies [19, 22]. No information concerning the development of malignancies has been provided by either study [19, 22], but the observation period may have been too short to draw inferences over the long-term.

Recently, a meta-analysis evaluated the evidence of TAC as a potential agent in the treatment of LN, either when compared with CYC or MMF or as part of a multitarget therapy. The analysis showed better efficacy of TAC compared with CYC for the induction of complete renal remission, while there was no difference in comparison with MMF. The analysis also showed superiority for multitarget therapy compared with CYC. Notably, two of the studies included have only been reported as abstracts; the results also need to be interpreted with caution in terms of sample size and the heterogeneity of the studies included, and because of the lack of long-term data [25]. TAC is a clear therapeutic option in pregnancy, since it is assumed to be safe. A single-centre experience showed prevention of relapses in three pregnancies, and TAC was successfully used for the treatment of nephrotic flares while on azathioprine (AZA) as an add-on or as a substitute. All responded, having live births with weights appropriate for gestational age [26].

Of particular importance, to date no data from RCTs are available supporting the use of CNIs in LN in Caucasians and African Americans.

MAINTENANCE THERAPY

Several publications have assessed the role of CNIs as maintenance therapy in LN. Moroni et al. [27] reported a trial of 75 patients who received either CSA or AZA after induction with steroids and CYC. The decline in proteinuria and the number of flares were similar in both groups, while eGFR changes after 2 years were in favour of AZA (12 mL/min/1.73 m² from baseline). However, no difference was found in the chronicity index in repeat kidney biopsies. In a Chinese cohort of 70 patients, AZA or TAC in conjunction with low doses of prednisolone (10 mg/day) had similar low flare rates (two flares with AZA, none with TAC) at 6 months [28], although the trial was too short to draw any definite conclusions on efficacy regarding relapses. A long-term issue will be the rate of relapse on cessation of CNIs.

Another question addresses the choice of the agent used for maintenance therapy in subjects initially treated with CNIs. In the study by Mok et al. [20], both groups (MMF and TAC) were switched to AZA maintenance therapy, and there was a trend towards an excess of renal relapses in the group of patients receiving TAC as induction therapy (proteinuric and nephritic flares: 35% and 27% versus 24% and 18%, respectively), although this difference did not reach the level of statistical significance.

IMPACT OF CALCINEURIN INHIBITORS ON CARDIOVASCULAR PARAMETERS

A possible relationship between autoimmune inflammation and atherosclerosis has been discussed in various autoimmune diseases. Patients with systemic lupus erythematosus (SLE) have an increased risk of cardiovascular morbidity very early in their disease course. In a recent multicentre analysis, patient age, hypertension, hypercholesterolaemia and smoking were independent risk factors for myocardial infarction in patients with LN [29]. Data from long-standing experience of renal transplantation indicate that the use of CNIs is associated with aggravation of arterial hypertension, post-transplantation DM or worsening of DM and dyslipidaemia [30]. Long-term data on cardiovascular morbidity in patients with LN treated with CNIs are still needed. Cardiovascular risk factors have been investigated in various studies. One study assessed the LDL:HDL ratio after 6 months of induction treatment. The overall decline in the ratio was 0.73 in the MMF- and 0.46 in the TAC-treated patients, respectively [20]. A clear trend towards higher rates of new-onset hypertension and hyperglycaemia has been reported in a recent meta-analysis of different RCTs under CNI therapy [25]. In contrast to the assumption that CNIs worsen atherosclerosis, CSA treatment was an independent predictor of lower carotid intima media thickness, despite the association of CNI use and hypertension observed in a single-centre study [31].

REMAINING UNCERTAINTIES AND FUTURE AIMS/DIRECTIONS

CNIs have been acknowledged as treatment alternatives, especially in patients with pure membranous LN. However, the evidence of benefit, as assessed in the current Polar View, has also been reported for proliferative LN forms. Nevertheless, uncertainties remain and well-designed trials are required to define the role of CNIs in LN. Thus, the suitability of using CNIs in patients who already have severely impaired renal function remains unclear and there is a lack of data on long-term stability following initial response, which may also be partly attributable to haemodynamic changes [32]. Additionally, the question of whether CNIs have a direct effect on podocytic structures independent from immunosuppression needs to be clarified. Acute nephrotoxicity was observed in several studies [15, 20] and was reversible following dose adjustment, but more information on the long-term nephrotoxicity of CNIs in LN is still required. In general, no recommendations can be derived regarding the duration of therapy or the tapering of agents in the multitarget therapy setting. From other studies we have learned that, for the evaluation of immunosuppressive drugs, an efficacy end point of <12 months might be inadequate in LN and longer follow-ups are desirable. Despite the efficacy signals of CNIs reported in Asian studies, less convincing data are available from non-Asian populations. Thus, for the time being, the SOC remains MMF- or CYC-based induction therapy and CNIs are recommended as an alternative in the treatment of class V LN. Exciting data are expected from an ongoing international, multicentre trial (AURA-LV) that is investigating the value of voclosporin, a novel CNI, in patients with LN classes III–V (NCT02141672, www.clinicaltrials.gov).

CONFLICT OF INTEREST STATEMENT

None declared. (See related article by Fernandez Nieto and Jayne. Con: The use of calcineurin inhibitors in the treatment of lupus nephritis. Nephrol Dial Transplant 2016; 31: 1567–1571; Mok. Pro: The use of calcineurin inhibitors in the treatment of lupus nephritis. Nephrol Dial Transplant 2016; 31: 1561–1566)

REFERENCES