An Update on Therapies for Proliferative Lupus Nephritis: How Certain Can We Be About the Evidence?

The treatment of glomerular diseases has often been considered more art than science. The best glomerular disease clinicians demonstrate a seemingly intuitive skill in choosing therapies and knowing beforehand which drug will work for which patient. When asked to explain their treatment rationale, the experts at managing glomerular diseases often fall back on some version of, “This patient reminds me of that other patient who responded to this specific therapy.” This kind of decision making is necessary in a field in which the bulk of evidence still comes from observational studies and the answers to many clinical questions cannot be found in prospective randomized controlled trials.

Among the glomerular diseases, lupus nephritis has served as a beacon of hope for evidence-based medicine. More than half of all patients with lupus will have clinically evident kidney disease during their disease course, and approximately 10% to 17% of those with lupus nephritis will progress to end-stage kidney disease (ESKD). Although this number is still unacceptably high, it nonetheless represents a major improvement in outcomes compared with kidney survival rates for lupus nephritis as low as 20% before 1980.^1,2 This improvement in turn has been attributed to widespread use of immunosuppression, in large part because of a consistent record of randomized trials of induction and maintenance treatment regimens.³

However, Tunnicliffe et al^4,5 have recently updated a Cochrane Review of immunosuppressive therapies for proliferative lupus nephritis (building on prior Cochrane Reviews of this same subject in 2004⁶ and 2012⁷) that is notable for its inability to provide a definitive answer to the most basic question regarding lupus nephritis. Using the authors’ own words: “We are still not certain which is the best treatment for lupus nephritis to protect against needing dialysis in the longer term.”^4(p3) After adding 21 new studies to their database, which now includes 71 studies involving 5,089 participants, the Cochrane team was not able to identify a single best therapy for proliferative forms of lupus nephritis, class III or IV lesions (either alone or in conjunction with class V, membranous, lesions), which carry the highest risk for progression to ESKD and subsequently premature mortality.

In large part, this is a problem of the sausage’s quality being subject to its ingredients. Even the best and largest clinical trials of lupus nephritis have used relatively short-term measurements of proteinuria (eg, at 6 or 12 months) for primary outcome assessments of complete and partial remission.^8–12 These studies have not been powered or used long-enough follow-up periods to assess for “hard” outcomes such as progression to advanced stages of chronic kidney disease, ESKD, and/or death, outcomes that certainly occur in lupus nephritis but usually take longer than 1 or 2 years. In comparison, when Palmer et al¹³ recently reported results of a network meta-analysis of randomized trials for proliferative lupus nephritis, using 53 studies involving 4,222 participants, they focused on proteinuria remission as an outcome and concluded that complete remission was more likely when an induction phase regimen included mycophenolate mofetil (MMF; odds ratio [OR], 1.44; 95% confidence interval [CI], 1.00–2.06), calcineurin inhibitors (CNIs) alone (OR, 1.74; 95% CI, 1.09–2.79), or MMF in combination with CNIs (OR, 2.69; 95% CI, 1.74–4.16) compared to a traditional regimen of intravenous cyclophosphamide. However, even in this setting, these investigators were forced to concede that complete remission was a heterogeneous outcome with variable definitions across their analyzed studies.¹³

The updated Cochrane Review is more notable for what it could not determine than for what it was able to conclude with any degree of certainty. The authors posed 3 questions as objectives for this review (Are new immunosuppressive agents superior to or as effective as cyclophosphamide-based regimens? Which agents, dosages, routes of administration, and duration of therapy should be used? Which toxicities occur with different treatment regimens?) and were unable to provide conclusive answers for any of them. MMF, in their analysis, may lead to higher rates of remission compared to intravenous cyclophosphamide, with an “acceptable adverse event profile,” but their certainty in the supporting evidence was rated “low” and included the possibility of no difference between treatments. This is particularly disappointing in the realm of adverse events. Most nephrologists, like their patients with lupus nephritis, have embraced MMF despite the largest induction study (Aspreva Lupus Management Study [ALMS]) showing only noninferiority to cyclophosphamide, based on the inherent assumption that MMF is a “less toxic” therapy than cyclophosphamide. However, in their updated review, the Cochrane group was not able to find statistically significant differences between MMF- and cyclophosphamide-based regimens in terms of major infections (relative risk [RR], 1.05; 95% CI, 0.69–1.60), leukopenia (RR, 0.59; 95% CI, 0.33–1.08), or even ovarian failure (RR, 0.36; 95% CI, 0.06–2.18).

So many questions about the optimal treatment of proliferative lupus nephritis remain for which the studies reviewed by Tunnicliffe et al are unable to provide answers. Newer agents, such as rituximab,¹¹ and other biologics, such as abatacept,¹⁴ have not provided enough convincing data to advocate their regular use in this population, although there is certainly persistent interest in using these agents in patients for whom traditional regimens of MMF and/or cyclophosphamide have failed. Likewise, the not-so-new class of drugs, CNIs, have shown remarkably high rates of remission when combined with MMF in 2 large studies from China,^12,15 but these results are limited by questions of: (1) generalizability to non-Asian patients with lupus nephritis and (2) sustainability of CNI-induced remissions given their ability to lower proteinuria independent of an effect on immune function. We also do not have convincing data from the analysis of the included studies on what dose and duration of corticosteroids, always given with induction and variably given with maintenance-phase therapies in these trials, should be used in treating proliferative lupus nephritis, how long maintenance phase therapy with MMF should be continued (including whether this drug can be safely stopped altogether), and how to define and treat resistant forms of lupus nephritis.

However, the news does not have to be this dismal and may only appear that way in the setting of aggregate analyses such as these that are designed to make sweeping statements and come across as “negative” when such pronouncements cannot be made definitively. In 2018, although we do not have the robust evidence base of clinical trials with tens of thousands of participants, we have enough clinical trial data in proliferative lupus nephritis to tell our patients that: (1) MMF is at least equally effective as cyclophosphamide for induction therapy, and (2) MMF is the preferred agent for maintenance therapy. However, beyond that, we must admit the gaps in lupus nephritis that remain to be filled, gaps that are clearly highlighted in the Cochrane group’s analyses. Even the best induction regimens have unacceptably high nonresponse rates, and even the most rigorous maintenance-phase regimens are hampered by unsatisfactory relapse rates. Unfortunately, we do not even have an array of biomarkers or validated clinical parameters with which we can prognosticate such poor outcomes beyond the kidney biopsy because progression from active to chronic lesions on a repeat biopsy remains the most consistent predictor of long-term risk for ESKD.¹⁶

Improving the evidence base is an obvious first step in addressing all the deficits highlighted in this updated Cochrane Review. Unfortunately, some of the largest ongoing clinical trials in lupus nephritis still rely on a proteinuria-based remission end point, exemplified by the current studies of voclosporin, a novel CNI, and belimumab, a monoclonal antibody directed against B lymphocyte stimulator (BLyS; ClinicalTrials.gov identifiers NCT03021499 and NCT01639339, respectively). Lupus nephritis trials should be designed to be large enough and with sufficient duration of follow-up to evaluate non–proteinuria-based outcomes, such as doubling of serum creatinine level, progression to ESKD, and all-cause mortality. Also, these trials should be designed to specifically address questions about doses and duration of therapy, requirements for concomitant corticosteroid use, and toxicity profiles that our patients continue to ask and that, unfortunately, we continue to answer with low-certainty evidence (to borrow a phrase from these Cochrane authors).