The history of the treatment of proliferative lupus nephritis is a long journey that spans over half a century and travels down many often-divergent roads. This past tells the story of the renal biopsy, as well as the paths to the use of corticosteroids, azathioprine, cyclophosphamide, mycophenolate mofetil, calcineurin inhibitors, and even rituximab. There is no shortage of treatment options for lupus nephritis, but it is important to know this history to understand the strength of the evidence (or lack thereof) on which we base our treatment decisions.
Your standard induction treatment may be high dose oral steroids and mycophenolate mofetil, which is a sensible option, but other very reasonable options exist. Evidence strongly supporting one strategy over the other is rarely available and the tie-breakers may often be related to considerations of compliance, side effects, availability, route of administration, and cost. Lupus nephritis was once a terminal disease but this changed with the availability of steroids, the workhorse of essentially all induction protocols for proliferative lupus nephritis. This is where the story begins.
In 1930, an adrenal extract was found to prolong the life of patients with Addison’s disease. It took 20 years of painstaking laboratory work to develop cortisone, which had miraculous effects on patients with debilitating arthritis, and led to a Nobel Prize in Medicine in 1950! Prednisone became available 5 years later and had the advantage of being a potent oral glucocorticoid with essentially no mineralocorticoid effect.
In 1957, lupus giants Muehrcke, Pollack, Pirani, and Kark reported 10 patients in the new field of renal biopsy proven lupus nephritis who received 50 mg of cortisone a day (equivalent to about 10 mg of prednisone/d) with abysmal outcomes, watching all 10 die, 8 of renal failure. Then in 1961, they reported their results using “large doses” (47.5 mg/d) of the now readily available and much stronger prednisone in 16 patients with lupus nephritis. Only 7 died, 5 in renal failure, but 9 survived. This was confirmed in Pollack’s 1964 report comparing histology and clinical outcomes in patients receiving high v low dose prednisone. Prednisone at 1 mg/kg/day eventually became a standard of care for baseline induction therapy for patients with proliferative lupus nephritis.
While steroids became the critical element in any lupus nephritis induction protocol, the future introduced many immunosuppressive agents that were either added to steroid induction protocols or found a role as “steroid sparing” for the maintenance phase of the treatment. Although the alkylating agent mechlorethamine (nitrogen mustard) and the antimetabolites 6-mercaptopurine and 6-thioguanine (precursors to azathioprine) were starting to be used in the early 1960s for autoimmune diseases including lupus nephritis, it wasn’t until the 1970s when azathioprine and cyclophosphamide (another alkylating agent) were added to steroid protocols for lupus nephritis. The evidence that these combinations were superior to prednisone alone was not strong, with most studies flawed by small patient numbers, inconsistent histologic classification, and short term follow-up.
In 1984, a meta-analysis by Felson and Anderson published in the NEJM determined that the addition of azathioprine or oral cyclophosphamide (CycP) to prednisone provided superior renal outcomes than when using prednisone alone. In 1986, also in the NEJM, the NIH group led by Balow and Austin reported their experiences with immunosuppressive agents for lupus nephritis in an article entitled “Therapy of lupus nephritis: Controlled trial of prednisone and cytotoxic drugs” which was a pooling of their experiences treating 107 patients with lupus nephritis from 1969-1981, and thus, was not at all controlled. This report was novel, however, in that they reported the use of CycP administered as monthly intravenous boluses (IV-CycP). By 1987, they were using the IV bolus route exclusively. Simultaneously, the Collaborative Study Group (CSG) led by Lewis et al was studying the effect of plasmapheresis added to daily oral CycP and prednisone in 86 patients with severe lupus nephritis (the addition of pheresis did not improve clinical outcome).
However, as the NIH group’s data had wide exposure by nature of being the lead article in the NEJM, the intermittent method using IV-CycP bolus became a popular treatment option. The two different routes of CycP administration have never been compared in a randomized clinical trial, but a meta-analysis by Bansal and Beto in 1997, as well as a 2006 retrospective comparison by Mok et al of 212 patients (½ receiving IV and the other ½ oral CycP) failed to find one route superior to the other. In that regard, McKinley et al reported their experiences with oral and IV-CycP were similar, but that oral CycP was considerably cheaper and more convenient than the IV route and considered oral CycP “an underused therapeutic option.”
One major difference between the two routes is the cumulative dose (CD) of CycP. While the NIH protocol originally recommended 1.0 gm/m2 q 3 months for 4 years (CD ~ 28 grams), by 2004 the IV route morphed into 6 monthly IV-CycP boluses of 0.5-1.0 g/m2, followed by maintenance boluses every three months for 1-3 years. If this were done for 2 years, the CD of IV-CycP would be 24 grams. The CSG protocol with 6-8 weeks of oral CycP at 150 mg/day would have a maximum CD of 9 grams. Thus, the main difference between the two routes, besides the potential cost considerations, is that the CSG protocol has a lower CD that is administered in a much more “up-front” manner, with completion of the cytotoxic element of therapy in only 8 weeks as opposed to 1-3 years as seen in the typical IV protocols. And just when you think you might have the CycP story into perspective, along comes “Eurolupus” (n=90), in which 3 grams (the lowest CD yet of 3 grams) of IV-CycP (500 mg q2 weeks for 6 doses over 3 months) was found to be non-inferior to 8.5 grams over a year using a monthly IV bolus regimen.
The next big addition to the lupus armamentarium came with the FDA’s approval of mycophenolate mofetil (MMF) for its use in transplantation in 1995. MMF inhibits inosine monophosphate dehydrogenase (IMP) and, as such, has effects similar to the other antimetabolites. It was originally found to have antibiotic activity but was eventually found to also have antiviral, antifungal, and anti-cancer properties. However, it wasn’t until the 1980s that MMF was determined to have potent immunosuppressive effects that led to studies in organ transplantation.
It did not take long before this exciting new drug was being used in lupus nephritis. By 2000, Chan et al from Hong Kong (n=42) reported MMF to be equivalent to oral CycP (12 months MMF v 6 months CycP followed by 6 months azathioprine) for induction therapy. In 2005, Ginzler et al reported a US only study in which the majority of the participants were African-American. They concluded that MMF was superior to and with fewer side effects than IV-CycP for induction (n=140). Finally, in 2009, Appel et al reported the ALMs trial (n=370) that essentially repeated their earlier MMV v IV-CycP comparison but on an international scale; however, this time they determined MMF to be similar to IV-CycP in renal and mortality outcomes, although MMF led to improved renal outcomes in the African-American subset. It is worth considering that “the study did not meet its primary objective of showing that MMF was superior to IVC as induction treatment for lupus nephritis,” and also that the follow-up maintenance phase of ALMs reported that patients in the MMF induction arm of the trial were twice as likely to fail maintenance (renal relapse) than those that received induction with IV-CycP (21% v 11%). While this difference was not statistically significant, it may be a signal that represents a tangible difference between these induction options.
Even though azathioprine (MMF’s little brother) was left behind as an induction therapy for the last 20 years, a report from Grootscholten et al in 2006 compared azathioprine to IV-CycP pulse therapy (n=87) and found no difference between the groups in the end-point of doubling of serum creatinine. However, with the large and well-publicized success of MMF, azathioprine didn’t stand a chance.
Calcineurin inhibitors (MMF’s cousins?) have also been studied for a host of renal lupus presentations. While there are efficacy signals with the use of cyclosporine and tacrolimus for induction therapy (especially in the Asian population), they are still not generally considered a first-line induction option for proliferative lupus nephritis. This uncertainty is underscored in the 5-part “Pro and Con” section on this topic in NDT in 2016. A newer CNI, voclosporin, is presently undergoing study in patients with active lupus nephritis.
If MMF was the “new kid on the block”, rituximab became the “golden boy” with FDA approval in 1997 for the treatment of B-cell non-Hodgkin lymphomas. Rituximab, an antibody against CD20 that is expressed on most early and differentiated B-cells, alters the cell surface and allows natural killer cells to attack and kill B-cells. With an appreciation for the role of B-cells in autoimmune diseases, rituximab has been used in many of these conditions, including lupus nephritis. The LUNAR trial reported by Rovin et al (n=144) compared MMF and prednisone to MMF and prednisone with the addition of rituximab. The results were similar in both study arms but since rituximab was added to standard therapy, the question of rituximab as a stand-alone therapy remained. The nonrandomized “Rituxilup” trial (n=50) used rituximab (two 1 gram rituximab doses and two 500mg doses of methylprednisolone both on days 1 and 15), followed by MMF maintenance (but without further induction or maintenance period steroids), and reported 90% of patients achieving a partial or complete remission by 37 weeks. Rituxilup specifically represents the first induction protocol that is essentially steroid-free. A recent study in 44 pediatric (< 14 years of age) patients comparing rituximab to MMF and IV-CycP found that flare-free survival was significantly higher and steroid requirements were significantly less at 3 years in the patients receiving rituximab compared to either MMF or IV-CycP. While the experience with rituximab as an induction therapy for lupus nephritis is still significantly less than with either MMF or IV-CycP, these data are encouraging that rituximab’s future looks bright.
Maintenance therapy for lupus nephritis relies on the concept that while induction protocols are effective in inducing a remission, relapses (“flares”) are common. Without a strategy to keep the disease quiescent, you will have cycles of induction followed by drug-free remission, followed by the need for re-induction. Induction takes its toll and while maintenance therapies is not entirely benign, the side effects are considered minimal compared to the risks associated with repeated induction. In the CSG’s lupus nephritis plasmapheresis trial, oral CycP was administered for only two months, but the steroid taper schedule lasted for 3 years. At that point, only if the patient was considered “no disease activity” were steroids stopped.
The paradigm of maintenance therapies subsequently shifted to a more rapid taper of steroids, with the introduction of chronic steroid sparing agents (initially azathioprine and then MMF) that were continued for an indefinite period. As would be expected, these two non-steroid maintenance options were compared. The aforementioned ALMs trial had a follow-up maintenance phase (n=227) in which patients were randomized to receive either MMF or azathioprine for 36-months. The results of this trial found MMF to be the clearly superior maintenance option. On the other hand, the “Eurolupus” trial also had a similar follow-up phase “MAINTAIN Nephritis“ (n=105) in which they found no difference between azathioprine and MMF as maintenance agents for both 4– and 10-year follow-up periods.
In these protocol comparisons, I did not discuss toxicity differences except for the implications of cumulative dosing of CycP. It is also worth pointing out that there were few consistencies between treatment protocols. In these various studies, CycP was used orally for 2-12 months and intravenously for 3-48 months. Some induction trials included high-dose IV steroid “pulses,” others did not. These protocol discrepancies are ubiquitous and make treatment comparisons very difficult. Primary outcomes were typically centered around some measure of remission, but these definitions also varied markedly both in what was being measured (absolute level or % change in proteinuria, serum creatinine, or both) and how long a follow-up time was utilized for that determination (months to years).
To further confuse matters, histologic inclusion criteria were not consistent. All lupus nephritis is not the same. For example, many of the studies included a significant percentage of patients without proliferative lesions (WHO V membranous as the only glomerular pathology). In addition, the mean serum creatinine values at baseline varied from 1-2.0 mg/dl. These are two major predictors of outcome independent of treatment (see table below).
|WHO V alone||Serum Creatinine|
|NIH 1986||15 %||1.0 mg/dl|
|CSG 1992||0%||2.0 mg/dl|
|Eurolupus 2002||8%||1.2 mg/dl|
|ALMs 2009||32%||1.1 mg/dl|
With all the heterogeneity in inclusion criteria, treatment protocols, and outcome metrics, picking a winner is like the game of rock-paper-scissors. You can always determine a winner when you compare one treatment strategy to another, but there is no clear winner when you look at them all.
To summarize, what has the last 50 years taught us about the treatment of proliferative lupus nephritis?
- High dose steroids are better than low dose steroids.
- Adding a cytotoxic agent is probably better than steroids alone but adding plasmapheresis doesn’t help.
- Steroid sparing maintenance protocols are an important adjunct to induction.
However, many questions remain, such as:
- Does it matter if you use an oral or an IV-CycP route?
- Can you get away with the “Eurolupus” protocol using a cumulative IV-CycP dose of only 3 grams, and if so, does that apply only to those of white European descent where it was studied?
- Should MMF be the preferred induction therapy for African-Americans?
- Is there really no role for azathioprine in induction?
- Will rituximab provide improved results or at least be more steroid sparing than MMF or CycP?
- Should CNIs even be seriously considered for induction at this point?
- Is MMF really better than azathioprine for maintenance?
- Should maintenance therapy be continued only for a finite time period?
To fall back on the ubiquitous cop-out of “larger studies are required to answer these questions” is unrealistic. Lupus nephritis is not a common or homogeneous condition and more “definitive” studies are unlikely to occur. And as the past has taught us, just when we think we may have it figured out, a new monoclonal antibody or class of drugs will emerge. It will be compared to the most recent popular therapy in a modestly sized population of patients with diverse histology and renal function. It may be found only to be “non-inferior,” and we will likely be left with as many questions as answers. Fortunately, this long and winding road of therapy options may not be getting any straighter, but it does keep getting wider.