Success Using Tacrolimus in Patients with Proliferative and Membranous Lupus Nephritis and Refractory Proteinuria

Abstract

Standard treatment for class III, IV, and V lupus nephritis (LN) is a combination of oral corticosteroids and mycophenolate mofetil (MMF). There is an estimated failure rate of 16%. Several small studies have looked at the use of tacrolimus in class III, IV, and V LN, both as induction treatment and as maintenance in patients refractory to other treatments. The majority of these studies were conducted in Asian patients. We discuss a cohort of eight female patients of various ethnicities with biopsy proven LN. All patients were evaluated retrospectively. Six were started on tacrolimus after failing to respond to MMF and corticosteroids, and one was started on tacrolimus alone because treatment options were limited by pregnancy. Five were Caucasian, one African American, one Hispanic, and one Vietnamese. Mean tacrolimus dose was 3.3mg daily (range 2–5 mg) titrated to a mean level of 3–6 ng/dl (range 3–6.6 ng/dl) for a mean of duration of 16 months (range 2–54 months). Six patients experienced complete remission (proteinuria <0.33g/day), and two patients had a partial remission (minimum of 50% reduction in baseline proteinuria). Albumin increased an average of 32%. Average C3/C4 levels were 64/15 mg/dL, respectively, prior to treatment, and 95/25 mg/dL following treatment. No treatment-limiting adverse effects were reported. Our case series supports the growing body of evidence that tacrolimus is an effective therapy in LN patients with refractory proteinuria. Further studies are required to establish the long-term safety and efficacy of tacrolimus.

Introduction

Renal complications are a major source of morbidity and mortality in patients with systemic lupus erythematosus (SLE) and are present in 25–75% of patients.¹ Deposition of circulating immune complexes and recurrent flares lead to chronic glomerular scarring, and poorly controlled glomerulonephritis is a major risk factor for renal deterioration and poor long-term outcome.² Histopathologic diagnosis and staging by International Society of Nephrology/ Renal Pathology Society (ISN/RPS) classification, guides therapy with a goal of normalizing renal function and preventing progressive loss by inducing and maintaining remission.³ Treatment of classes III, IV, and V LN remains a management challenge in many cases. Established therapies include glucocorticoids plus pulsed intravenous cyclophosphamide (IVC) or oral mycofenolate mofetil (MMF).

Cyclophosphamide has been commonly accepted as standard therapy since studies done through the National Institutes of Health in the 1980s reported that pulse therapy with cyclophosphamide improves the prognosis of LN compared with glucocorticoid therapy alone.⁴ Unfortunately it has an unfavorable side effect profile, with adverse effects including hemorrhagic cystitis, carcinoma of the bladder, and bone marrow suppression, some of which can be fatal. It can also result in permanent sterility in young female patients.^5,6

Mycophenolate mofetil (MMF) is an alternative to cyclophosphamide, and studies have demonstrated non-inferiority as well as a lower incidence of adverse effects when used as maintenance therapy.⁷ Azathioprine has also been explored, but was shown to be less efficacious in treatment of LN when compared to MMF with a similar side effect profile.^8,9

Evidence has accumulated to support MMF as the preferred treatment in conjunction with steroids. However, a significant proportion of patients have persistent proteinuria despite this regimen, with estimated failure rates of 16–20%.^9–11 Continued proteinuria is predictive of poor outcome in most studies with no remission leading to high rates of chronic kidney disease and eventual progression to end stage renal disease. Even a partial remission in lupus nephritis is associated with a significantly better patient and renal survival compared with no remission. We defined partial remission as decrease in proteinuria by greater than 50%, and complete remission as reduction in proteinuria to <0.33g/day based on spot protein/creatinine ratio.¹²

More recently a group of small studies, mostly performed in Asian patient populations, have looked at tacrolimus in patients with class III, IV, and V LN. Tacrolimus is currently employed in transplant medicine to control rejection after kidney, liver, heart, and bone marrow transplantation, and is known as a safe and effective immunosuppressant. It is a macrolide calcineurin inhibitor and interferes with both T-lymphocyte signal transduction and IL-2 transcription. Known side effects and complications of tacrolimus use include tacrolimus nephrotoxicity, infections, dyslipidemias, tremor, low magnesium, new onset hypertension, and new onset diabetes mellitus, which can manifest as diabetic ketoacidosis.¹³ However, unlike many other immunosuppressives, tacrolimus has been shown to be relatively safe in pregnancy.¹⁴ MMF is associated with increased risk of facial malformations and first-trimester pregnancy loss.

Studies in murine models have demonstrated a decrease in intraglomerular cellular proliferation and normalization of affected podocytes.¹⁵ This effect has been confirmed in humans as well, Nonaka reports a case of membranous LN in which a young female patient showed histologic improvement of subepithelial deposits after tacrolimus therapy.¹⁶ This mechanism is thought to be independent of its immunosuppressive function, a theory supported by similar findings when it is used in other proteinuria inducing disease states, such as IgA nephropathy.¹⁷ Tacrolimus is postulated to alter levels of the protein synaptopodin and thereby stabilize the podocyte cytoskeleton. It has now been used successfully both as multi-target therapy using a combination of steroids, MMF, and tacrolimus, and as primary induction monotherapy with steroids. For the period 2005–2012, there were approximately 10 open label trials, case series, or randomized controlled trials, and two meta-analyses investigating tacrolimus as a treatment for LN.^18–28 Available evidence suggests higher rates of remission when added to MMF, also that tacrolimus alone is an option for induction therapy. The existing evidence reviewing side effects varies, some studies observed severe effects including severe infection and new onset diabetes, while others demonstrated a benign side effect profile for tacrolimus.

Several questions remain unanswered in the small group of existing studies on tacrolimus in LN. Optimal dosing and duration of therapy still needs to be determined. Many studies lack long-term follow up. Also, racial discrepancies are not well addressed. We present data from eight female patients with biopsy proven class III, IV, or V LN and persistent nephrotic range proteinuria in a retrospective case series.

Case Series

We retrospectively discuss eight female patients who met the diagnostic criteria for SLE according to the American College of Rheumatology definition, had a renal biopsy confirming class III, IV, or V LN, and were treated with tacrolimus. Five were Caucasian, one Vietnamese, one Hispanic, and one African American. Mean age was 27 (range 20–44). There were two patients with class IV LN, two with class IV/V LN, two with class III/V LN, and two with pure class V LN. Details regarding patient demographics and treatment history are presented in table 1.

Table 1.

Patient Characteristics prior to addition of tacrolimus

Patient	Age	Race	Sex	Duration of lupus	ISN Class	Time since renal biopsy	Prednisone dose at time of tacrolimusaddition	Time on MMF	Previous immunosuppressive regimens/other immunosuppressive medications	Additional relevant medications
1	27	Caucasian	F	4 years	III/IV	3 years	30mg	3 years	MMF, prednisone	Lisinopril, simvastatin
2	28	Caucasian	F	10 years	IV	8 months	60mg	10 months	hydroxychloroquine, prednisone, cyclophosphamide, rituximab, azathioprine, MMF	Lisinopril
3	28	Caucasian	F	14 years	V	2004	20mg	5 months	MMF, prednisone, hydroxychloroquine	Losartan, metoprolol
4	20	Vietnamese	F	2 years	IV/V	2 months	60mg	2 months	prednisone, MMF	Lisinopril, Mg oxide
5	29	African American	F	11 years	IV/V	11 months	40mg	3 months	Cyclophosphamide, Prednisone, Hydroxychloroquine, MMF	Lisinopril
6	44	Caucasian	F	4 years	III/V	2 months	Prednisone 40mg	2 years	hydroxychloroquine, MMF, prednisone	Lisinopril
7	21	Caucasian	F	8 months	V	N/A - pregnant	50mg daily	N/A - deferred due to pregnancy	Prednisone, Azathioprine	None
8	22	Hispanic	F	3 months	IV	1 month	Methylprenisolone 48mg QD (prednisone allergy)	2 months	hydroxychloroquine, MMF, solumedrol	Amlopdipine, furosemide, Micardis, metoprolol

For each patient described here, the decision to start tacrolimus was based on clinical data at the time and failure to respond to well-established immunosuppressive regimens. All were started at 0.15–0.20 mg/kg body weight taken either once daily or broken into twice daily dosing, depending on patient preference, and titrated to a goal trough of 4–6 ng/ml, the established therapeutic dose studied in transplant medicine. Troughs were measured anywhere from monthly to every three months, depending on stability of the dose. Patients were followed on a monthly basis with serial measurements of spot protein/creatinine ratio to quantify proteinuria, serum albumin, serum creatinine, and serum complement levels. Serum electrolytes, blood pressure, fasting lipid profiles, and random glucoses were monitored as well. We also report percent reduction in proteinuria and percent increase in albumin for each patient and if they meet the criteria for complete or partial remission (table 2). Any adverse effects were assessed with careful review of systems questioning at monthly visits.

Table 2.

Clinical parameters before and after addition of tacrolimus

Patient	1	2	3	4	5	6	7	8
Max tacrolimus dose
	1mg bid	3mg qam/2mg qhs	1.5mg bid	2mg qam/1mg qhs	3mg BID	3mg qam/2mg qhs	3mg BID	2mg QAM/3mg QPM
Spot prot/Cr ratio
Before	5.0	10.7	4.22	10.51	5.15	2.54	7.12	5.13
After	0.65	0.71	0.28	0.10	0.31	0.29	0.23	1.38
Duration until max remission	3 months	6 months	8 months	9 months	11 months	4 months	5 months	5 months - ongoing
% reduction in proteinuria
	87% PR	93% CR	93% CR	99% CR	94% CR	44% CR	97% CR	73% PR
Albumin
Before	2.2	1.76	2.7	1.8	2.9	3.1	2.4	1.7
After	3.0	3.3	3.4	3.5	3.6	3.8	3.5	3.3
Percent change	27%	47%	21%	49%	20%	18%	31%	48%
C3/C4
Before	77/14	62/11	43/13	64/11	74/24	54/8	91/13	45/14
After	92/9	103/25	89/27	107/21	82/25	65/15	93/19	89/21
Serum creatinine
Before	0.73	2.04	0.82	0.75	0.71	0.77	0.55	1.64
After- most recent	0.55	1.09	0.69	0.54	0.75	0.82	0.69	1.06
Average serum tacrolimus level
	6.1	6.48	6.04	6.56	3.6	3.0	6.2	5.6
Current duration of therapy
	4 months - noncompliant	1 year	7 months	9 months	11 months	3 months	6 months	5 months
Adverse effects
	None reported	Otitis externa	Fine hand tremor	hypomag	hypomag	Tooth abscess	None reported	Candidiasis oral hyperglycemia
Final prednisone dose
	20mg	stopped after 10 months	5mg	Stopped after 7 months	Stopped after 6 months	20mg	20mg	Solumderol 24mg

Patients 1, 2, 3, 5, and 6 were similar in that they had a longstanding diagnosis of lupus nephritis with biopsy performed in the past, and had varying levels of proteinuria on a combination of immunosuppressives, but had not yet experienced a clinically significant remission. These four patients had been on relatively high doses of prednisone (30–60mg daily) and MMF for several months. All were on an ACE-inhibitor or angiotensin receptor blocker as well. Following addition of tacrolimus, they experienced 87%, 93%, 93%, 94%, and 44% reductions in proteinuria respectively (one partial remission and four complete remissions, table 2). Patient 1 experienced an increase in proteinuria within one month after she stopped tacrolimus due to noncompliance, and eventually she was lost to follow up (figure 1).

Figure 1.

Spot protein/creatinine ratio (mg/mmol) following addition of tacrolimus

Patient 4 also failed the standard MMF and prednisone but she was relatively early in the LN disease course. She was diagnosed only two months prior to starting tacrolimus. She had nephrotic range proteinuria after two months of MMF and prednisone, even while taking 60mg of prednisone daily. Tacrolimus was added only after a short trial of standard therapy due to severe edema and nephrotic syndrome. She experienced a complete remission after nine months of treatment.

Patient 7 is unique because she was pregnant at time of treatment. At 12 weeks she was started on azathioprine, she had already been placed on oral prednisone, however she had continued nephrotic range proteinuria, the decision was made to start tacrolimus. Tacrolimus 1mg twice daily was started at 14 weeks. She tolerated this regimen well throughout her pregnancy. There were no fetal adverse effects noted on ultrasound, which was performed twice. This patient delivered a healthy term infant by cesarean section. She did not breast feed. She had a complete remission after 5 months of therapy. She was able to stop prednisone after 6 months of treatment, though she did take prednisone throughout her pregnancy. She underwent biopsy after delivery which demonstrated membranous LN. MMF was later added to her regimen following a transient increase in proteinuria.

Patient 8 was started on dual therapy with tacrolimus and MMF in addition to a steroid agent de novo. She had experienced some criteria for SLE in the past but had never been formally diagnosed. She was biopsied shortly after presentation with difficult to control hypertension, nephrotic range proteinuria, hypoalbuminema, and hypocomplementemia. Biopsy revealed class IV LN. Therapy was immediately initiated on steroids, MMF, and tacrolimus 1mg twice a day. Because of a prednisone allergy she was treated with solumedrol instead. She had a partial remission, currently her treatment duration is 5 months. These eight patients did not experience any treatment limiting adverse effects while taking tacrolimus. One had hypertension at baseline and was on two antihypertensives prior to addition of tacrolimus. Two had hyperlipidemia at baseline and were on a statin prior to addition of tacrolimus. One patient developed a fine hand tremor which resolved with slight decrease in tacrolimus dose, two patients developed hypomagnesemia which was corrected by oral magnesium supplementation, one patient developed otitis externa, and one patient had a tooth abscess. Patient 8 had transient hyperglycemia which was managed with dietary modifications, she also had oral candidiasis. Apart from these instances clinical parameters including lipid profile, random serum blood sugar, and blood pressure remained at baseline in all eight of the patients (table 2). No patients experienced a rise in serum creatinine above baseline (figure 2).

Figure 2.

Serum creatinine (mg/dL) over time following addition of tacrolimus

Overall, five patients experienced a complete remission and three experienced a partial remission as defined previously (figure 1). All patients demonstrated an increase in complement (table 2). Albumin increased an average of 33% (range 18–49%, figure 3). Four patients were able to stop prednisone entirely. Average C3/C4 levels were 64/14, respectively, prior to treatment, and 90/20 following treatment. The mean treatment duration prior to maximum decrease in proteinuria and improvement in serum complement levels was 8 months for the five patients who experienced a complete remission. One patient discontinued early because of medical noncompliance. The other patients described in this case series continue treatment.

Figure 3.

Serum albumin over time (g/dL) following addition of tacrolimus

Discussion

Lupus nephritis has been shown to have significant morbidity and mortality, with 10–30% of patients progressing to end stage renal disease depending on severity. Deposition of circulating immune complexes and recurrent flares lead to chronic glomerular scarring. Persistent proteinuria is directly correlated with worse clinical outcomes. Decreasing proteinuria increases time to dialysis in patients with proliferative and membranous LN. Estimated renal survival at ten years is 94% for complete remission, 45% for partial remission, and only 19% for patients with continued proteinuria.¹² Data from this case series supports the relatively new concept of treating resistant proteinuria with tacrolimus.

Few have performed direct head to head comparisons of tacrolimus and other established alternative treatments for LN. Bao, et al, performed a randomized trial of prednisone, MMF, and tacrolimus versus cyclophosphamide alone. The multi-agent therapy was found superior to cyclophosphamide at achieving complete remission at 6 and 9 months.¹⁸ Chen, et al, looked at efficacy of tacrolimus compared to azathioprine in maintaining remission and found similar rates of relapse but significantly less leucopenia with tacrolimus. Szeto, et al, compared azathioprine to tacrolimus specifically in patients with membranous LN (class V) and found that not only did tacrolimus reduce proteinuria faster, also patients relapsed after it was stopped.¹⁹ In a meta-analysis, Deng, et al, compared cyclophosphamide plus steroids to tacrolimus plus steroids, and determined tacrolimus is more effective as an induction therapy in Chinese patients with class III, IV, and V LN. In our patient population, we observed an added benefit when added to MMF and steroids, similar to Bao's study.

Still, the ideal use of tacrolimus in LN remains unclear. Our patient population involved mostly patients who were refractory to standard treatment, though we did have success using tacrolimus as primary induction therapy in one newly diagnosed patient. Our case series also incorporates one patient who was pregnant during her treatment. Azathioprine has also been shown safe in pregnancy, but, as discussed, is inferior to MMF. Studies in transplant medicine show tacrolimus to be safe and reliable in pregnancy; studies cite a 1% incidence of birth defects, mainly cleft palate.²⁹ This patient had a complete remission within five months. The average time until maximum reduction of proteinuria was 5 months, however at least four of our patients achieved a partial remission within 3 months (figure 1). The rapidity of reduction in proteinuria had a noticeable clinical benefit in these patients, with obvious reduction of edema in some. We lack long-term follow up currently.

Tacrolimus may also be a steroid sparing agent for many LN patients. All of the patients we observed were able to reduce their doses of prednisone. Some were able to stop after treatment with tacrolimus. There have been no studies performed to compare the steroid sparing ability of tacrolimus to other immunosuppressive medications, and this may be an area for future investigations.

There is a wide range of potential adverse effects, including serious hematologic complications such as severe leukopenia, anemias, transaminitis, and neurologic symptoms such as tremor and presyncope. Potential nephrotoxicity is also a concern. Some studies have found side effects to be limiting, with severe infection in particular requiring cessation of therapy prior to achievement of renal remission. Of note this was particularly observed in the Caucasian patients during the few studies that used this patient population.²⁵ Inherently the risk of severe infection with three immunosuppresive agents is higher than with dual or single agent therapy, but we observed only minor infections, none of which required intravenous antibiotics or hospitalization. Patient response to immunosuppressants is variable depending on ethnicity, an observation which has been described extensively for cyclophosphamide and MMF. The Aspreva lupus management study showed better response to MMF among black and Hispanic patients, and a higher prevalence of adverse effects among Asians.^30,31 Similar patterns are not yet well established for tacrolimus.

This case series augments current evidence showing tacrolimus is a viable option in LN patients who are refractory to standard treatment. Use in a pregnant patient with LN in whom other immunsupressants were not effective is a novel concept. Our observations in these patients support the growing evidence looking at tacrolimus, including its use in Caucasian patients and, in the case of one patient, pregnancy. Further exploration is needed to determine the optimal dosing, duration of treatment, and potential long-term side effects of using tacrolimus.

Conflict of Interest

None of the authors identify any conflict of interest.