Lancet. 2021 May 29; 397(10289): 2070-2080.
doi: 10.1016/S0140-6736(21)00578-X
Efficacy and safety of voclosporin versus placebo for lupus nephritis (AURORA 1): a double-blind, randomised, multicentre, placebo-controlled, phase 3 trial
Brad H Rovin, Y K Onno Teng, Ellen M Ginzler, Cristina Arriens, Dawn J Caster, Juanita Romero-Diaz, Keisha Gibson, Joshua Kaplan, Laura Lisk, Sandra Navarra, Samir V Parikh, Simrat Randhawa, Neil Solomons, Robert B Huizinga
PMID: 33971155
Introduction
Lupus nephritis (LN) is a common and serious complication of systemic lupus erythematosus (SLE). The clinical presentation of LN is highly variable, ranging from asymptomatic hematuria and proteinuria, to full-blown nephrotic syndrome, to rapidly progressive glomerulonephritis with accelerated decline in kidney function. Approximately 10-30% of patients with SLE+LN progress to end stage kidney disease (ESKD) within 10 years of diagnosis (Mahajan et al., Lupus 2020; Almaani, et al., CJASN 2016). Proteinuria, specifically reduction in proteinuria in the first 6-12 months after treatment initiation, has come to be an important prognostic marker. This reduction in proteinuria is the best predictor of reduced risks of disease flare, ESKD, and death (Dall’Era, et al., Arthritis Rheumatol 2015; Tamirou et al; Lupus Sci Med 2015). As such, it has come to be a standard for trials evaluating therapeutic interventions for LN. Unfortunately, up to 60% of patients are unable to achieve the current guideline target (proteinuria ratio <0.5-0.7 mg/mg within first year of therapy) with current agents, many of which are used off-label. Most standard therapeutic regimens for LN have this disappointing response rate and patients left at the mercy of considerable toxicities (especially in those regimens containing cyclophosphamide and high-dose steroids).
The treatment of lupus nephritis has evolved considerably since the mid-1980s (summarized in a 2019 RFN Blog). Although calcineurin inhibitors (CNI) were initially tried early, they truly came onto the LN scene around the 2010s (Figure 1). Two open-label randomized controlled trials (Bao, et al. JASN 2008 n=40; Liu, et al. Annals Int Med 2015 n=368), compared the CNI tacrolimus + mycophenolate mofetil (MMF) with cyclophosphamide and showed significantly increased response rate for CNI + MMF, but no long-term difference in outcomes.
Landmark trials in treatment of lupus nephritis from Landmark Nephrology
The BLISS-LN trial published in 2020 was summarized here on 11/24/20 (and also on Freely Filtered 2/14/2021 Ep 29).
Voclosporin is a novel CNI that was developed for treatment of LN. It is structurally similar to cyclosporin, but contains a modification that increases calcineurin binding. As with other CNI agents, Voclosporin decreases lymphocyte proliferation, blocks T-cell mediated immune responses, and increases podocyte integrity in the kidney (Kuglstatter etal., Acta Crystallogr D Biol Crystallogr 2011; Faul et al., Nat Med 2008). This particular agent may have some advantages over other available CNIs. One, its pharmacokinetic profile and oral bioavailability are more consistent, which eliminates the need for drug level monitoring (Mayo et al., J Clin Pharmacol 2013; Mayo et al., J Clin Pharmacol 2014). Unlike cyclosporin, voclosporin does not appear to have drug interaction with MMF (van Gelder et al., Nephrol Dial Transplant 2021). Further, voclosporin may have a more favorable effect of lipid and glucose profiles than other CNIs. In a phase 2 study of voclosporin vs tacrolimus in kidney transplant recipients, voclosporin showed similar efficacy in reducing acute rejection and decreased incidence of new-onset diabetes after transplantation (NODAT 1.6% vs 16.4%, with voclosporin vs tacrolimus, p=0.031) (Busque et al., Am J Transplant 2011). In AURA-LV, the incidence of diabetes during the 12-month study period was the same (1.1%) in the low-dose voclosporin and placebo groups (Rovin et al., KI 2019). The incidence of NODAT in transplant patients on 1 year of tacrolimus therapy was 25% (Torres et al., Kidney Int Rep 2018). While these are two different populations, there may be a physiologic mechanism to explain the lower incidence. Cell culture studies have (Kolic et al., Endocrinol 2020) demonstrated that tacrolimus, but not voclosporin, inhibits the response of glucose-stimulated insulin secretion from human islets (Kolic et al., Endocrinol 2020).
AURA-LV, mentioned above, was a phase 2 study of voclosporin for treatment of lupus nephritis was published in 2019 (Figure 2). On a background of steroids + mycophenolate, two different voclosporin doses (23.7 mg and 39.5 mg) were compared to placebo to assess efficacy and adverse events. There was an increase in complete response at 6 months for both doses, but only reaching statistical significance for the 23.7 mg twice daily dose. Serious adverse events occurred more often in both voclosporin groups, and more deaths occured in the low-dose group compared to placebo and high-dose voclosporin groups (Rovin et al., KI 2019).
The AURA-LV Visual Abstract (Rovin et al., KI 2019)
This study at hand, AURORA 1, is a phase 3 trial to follow-up on AURA-LV. The goal was to verify the efficacy and determine the safety of 23.7 mg twice daily voclosporin (in addition to MMF and rapidly tapered steroids) in the treatment of lupus nephritis. Additionally, it aims to expand on the promising results observed with CNI (tacrolimus) + MMF + prednisone observed in a predominantly Asian population (Bao, et al. JASN 2008; Liu, et al. Annals Int Med 2015).
The Study
Design
This study was a global multicenter, double-blind, randomized, placebo-controlled phase 3 trial. Sites were spread across 142 centers across 27 countries in North and South America, Europe, Africa, and Asia.
Intervention
All patients received IV methylprednisolone (0.5 g/day for those >45 kg, 0.25 g/day for those <45 kg) on days 1 + 2
Followed by a rapid oral prednisone taper (20-25 mg/day gradually decreased to 2.5 mg/day by week 16)
All patients received MMF, goal dose 1 g twice daily
Those MMF naive were started at 1 g daily on days 1-7, then increased to 1g twice daily on day 8
Those already on MMF, continued their dose without interruption or adjustment
Patients in study group received 23.7 mg voclosporin twice daily for 52 weeks
Patients in control group received placebo (identical in appearance, taste, smell) twice daily for 52 weeks
Study population
Diagnosis of systemic lupus erythematosus (by ACR criteria) + Kidney biopsy showing lupus nephritis (class III, IV, V; alone or in combo) within 2 years of screening
Evidence of active disease
UPCR ≥ 1.5 mg/mg (or ≥ 2 mg/mg if isolated class V)
If biopsy more than 6 months out from screening, needed doubling or more increase in UPCR
Excluded if eGFR < 45 mL/min
Full Inclusion/Exclusion criteria are listed in the study’s Appendix
Measures
All study visits
Basic chemistries including serum creatinine to calculate eGFR
Urinary protein:creatinine ratio (UPCR), first morning void
Blood pressure
At screening + weeks 24 and 52
24-hour urine for protein quantification
anti-double-stranded DNA
Outcomes
The primary endpoint was complete response at week 52 as defined by (all of the following required):
UPCR ≤ 0.5 mg/mg
eGFR ≥ 60 mL/min (or no decrease of more that 20% from baseline)
No need for rescue medication for SLE/LN
No more than 10 mg prednisone equivalent per day weeks 44-52
Subgroups
Age - ≤ 30 vs > 30 years
Sex - male vs female
Race - White vs Asian vs Black vs Other
Region - Asia-Pacific, Europe & South Africa, Latin America, North America
LN Class - isolated class V vs other
Use of MMF at screening - yes vs no
Max daily MMF dose during study period - ≤ 2 vs > 2 g
The hierarchical secondary endpoints
Time to UPCR ≤ 0.5 mg/mg
Partial response at week 24 or 52, defined by ≥ 50% reduction in UPCR
Time to partial response
Complete response at week 24 (defined as above with steroid dosing assessed weeks 16-24)
Additional secondary endpoints
See supplement
Statistics
Planned sample size of 162 patients per group to have 80% power to detect a difference between complete response rate of 20% in the placebo group and 34% in the voclosporin group. This odds ratio 2:1 was determined by two-group continuity corrected 𝛘2 test.
The primary and secondary endpoints analyses were performed with intention-to-treat and done using logistic regression modeling.
Subgroup analyses (pre-specified covariates) of the primary endpoint was completed with logistic regression modelling.
Funding
Study was sponsored by Aurinia Pharmaceuticals (developer/manufacturer of LupkynisTM 7.9 mg voclosporin capsules)
Aurinia was involved in all data collection, analysis, and interpretation.
Aurinia reviewed the manuscript prior to submission, but study authors maintained editorial control.
Results
During the study period (April 13, 2017 - October 10, 2019), 357 patients were enrolled with 179 being randomized to the voclosporin group and 178 to the placebo group. In the voclosporin group, 8% of patients did not complete the study compared to 17% in the placebo group. The reasons for study discontinuation were similar between the two groups (Figure 1, below). Median medication adherence for voclosporin and placebo was similar for both groups at 99%. Additionally, the mean daily dose of MMF [2 g per day] and duration of exposure were similar for both groups.
Figure 1 Trial profile from Rovin et al., Lancet 2021
Baseline patient characteristics
Approximately 90% of the patients in the study were female. The study also included a mix of races with the majority of patients having a normal GFR and the majority of patients having biopsy-proven class IV lupus nephritis. The mean time to lupus diagnosis was about 4.7 years with a mean baseline UPCR was about 4 mg/mg. Additional baseline characteristics of patients can be seen in table 1. In the voclosporin group, 161 (90%) of patients were female compared to 152 (85%) female patients in the placebo group. The remainder of the baseline characteristics are summarized in table 1A & 1B.
Table 1A: Demographic and baseline data (Rovin et al., Lancet 2021)
Table 1B: Demographic and baseline data (continued)
Primary endpoint
Approximately 18% more patients in the voclosporin group achieved complete renal response compared to the placebo group at 52 weeks(41% vs. 23%; OR 2.65 [CI 1.64-4.27]; p <0.0001) (Table 2 & Figure 3). In terms of the four composites of renal response, all favoured voclosporin. However, UPCR ≤ 0.5 mg/mg was the only composite that was statistically significant (45% vs. 23%; OR 3.11 [CI 1.93-5.00]; p<0.001) (see page 11 of appendix).
Figure 2: Complete and partial response endpoints. (Rovin et al., Lancet 2021)
Table 2: Summary of complete and partial responses. (Rovin et al., Lancet 2021)
The subgroup analyses of the primary endpoint at week 52 favoured voclosporin across all subgroups with statistical significance reported in the following subgroups (Figure 4):
Any sex
Age
Ethnicity subgroups (Asian, Black and other race)
Regions - Asia-Pacific & Latin America,
Biopsy class - (Other)
MMF used at screening
Max daily MMF dose during study period (≤ 2 g/day)
However, one should note that what matters in subgroup analyses are the interaction p values (unfortunately not presented here) and most subgroups here do not appear that different to the eye (perhaps MMF at screening being the exception).
Figure 4: Subgroup analysis of primary endpoint. (Rovin et al., Lancet 2021)
Secondary endpoints
All of the following key secondary endpoints in the study favoured voclosporin with statistical significance (Table 3):
Complete renal response at 24 weeks
Partial renal response at 24 weeks
Partial renal response at 52 weeks
Time to UPCR ≤ 0.5 mg/mg, days
Time to 50% reduction in UPCR, days
The probability of achieving a UPCR ≤ 0.5 mg/mg and >50% reduction in UPCR from baseline was significant in the voclosporin group. This was demonstrated with an early separation in the Kaplan-Meier curves comparing the two groups showing signs of an early treatment response with voclosporin and being maintained throughout the study (Figure 3).
Figure 6: Probability of (A) UPCR ≤ 0.5 mg/mg and (B) ≥ 50% reduction from baseline UPCR (Rovin et al., Lancet 2021)
All other secondary endpoints assessed during the study including SELENA-SLEDAI (Systemic Lupus Erythematosus Disease Activity Index) scores and immunology parameters (C3, C4 and anti-dsDNA levels) did not show statistically significant differences between the two groups.
Adverse events
The most common adverse events in both groups were infections; pneumonia was the most serious and frequent encountered infection in both groups. Although a total of 6 patients died during the study and follow up period, the investigators deemed that none of the deaths were related to any treatments from the study (Table 3).
Table 3: Adverse events (Rovin et al., Lancet 2021)
Discussion
The findings from the AURORA 1 phase 3 trial (a follow up to the phase 2 AURA-LV study) demonstrate that the addition of voclosporin to the standard of care for lupus nephritis (mycophenolate mofetil and glucocorticoids) resulted in a superior renal response when compared to placebo. Although the positive results from this study suggest that this novel calcineurin inhibitor may be useful in achieving rapid reduction of proteinuria, there are some important limitations that must be explored in terms of its application in clinical practice.
Strengths
Efficacy & safety profile: Patients in the voclosporin group had significant reduction in proteinuria within the first 6 months of the study and maintained this reduction at 52 weeks with no significant difference in adverse events between two groups.This study demonstrated in patients with lupus nephritis that a rapid tapering regimen of glucocorticoid therapy can be effective. Furthermore, the safety profile for both voclosporin and placebo were similar in this study when used in conjunction with MMF and low-dose glucocorticoid treatment.
Generalizability: A major criticism of previous trials in lupus nephritis is the lack of diversity in the study population. However, this study included a racially diverse population from a broad geographical representation around the globe.The results provide evidence to support the efficacy of calcineurin inhibitors in treating lupus nephritis within diverse populations.
Less steroid use: The maintenance dose of prednisone was 10 mg daily in another study with tacrolimus used in combination with steroids and MMF. Whereas in this study, the dose of prednisone was reduced down to 2.5 mg per day by week 16. Patients in AURORA 1 only received 2 days of induction therapy that varied based on weight (0.5 g/day >45 kg, 0.25 g/day <45 kg) whereas in the aforementioned comparison study, patients received a standard dose (0.5 g/day) for 3 days.
Limitations
Lack of novelty: Although the study demonstrates a reduction in proteinuria; all CNI agents are known to reduce proteinuria by their effect on podocyte stability. Additionally, there was no change in the immunological markers for lupus activity. Data for extra-renal manifestation of lupus activity was not reported and as a result the use of this agent does not offer any new from an immunotherapy perspective.
Cost: Given the similar effects that voclosporin can have on proteinuria reduction much like other, more affordable CNI agents, the estimated cost of $92,000 annually is a major drawback.
Variations in biopsy timing: Despite the majority of patients having renal biopsies done within 6 months of screening, the variations in timing of biopsy was a limitation. A standardized approach of having biopsies done at the start of enrollment would have provided a baseline of disease activity at the start of the study.
Type of disease: This study did not present data on patients with a new diagnosis compared to thoses with relapsing lupus nephritis and their response to treatment. A large number of patients with lupus nephritis who have reductions in the dose of immunosuppressive therapy may experience relapsing disease. Therefore an understanding of the baseline data about the type of disease at the start of the study would have been important in assessing the utility of voclosporin across different phenotypes of lupus nephritis.
Study duration: Given the chronicity of lupus nephritis and variability in the clinical history of the disease, a duration of 52 weeks for the study may not capture long term efficacy and safety profile of voclosporin. However, the AURORA 2 (NCT03597464) study is ongoing and will provide additional safety and efficacy data for an additional 104 weeks.
Conclusions
There have been significant advancements in therapies for lupus nephritis but many standard treatments have shown suboptimal renal response rates along with risk of toxicity to patients. Although the preliminary safety and efficacy data from this study is promising, the information gathered is not revolutionary. CNI agents have been previously well known to reduce proteinuria in patients and this study does not present any new information that would immediately change practice. Furthermore, practitioners should consider whether the change in a single amino acid in voclosporin from the structure of cyclosporin is worth that extremely high cost. Current existing therapies with cyclosporin and tacrolimus may provide the important and similar benefit of proteinuria reduction in patients with lupus nephritis without increasing the burden of cost to patients and our healthcare system.