Is Newer Safer? Adverse Events Associated with First-Line Therapies for ANCA-Associated Vasculitis and Lupus Nephritis

Abstract

Clinical outcomes in ANCA-associated vasculitis (AAV) and lupus nephritis have improved greatly with treatment regimens containing high-dose glucocorticoids and cyclophosphamide. However, with the use of these medications come significant adverse events, most notably infections, cytopenias, malignancies, and reproductive abnormalities. Multiple recent randomized controlled trials in AAV and lupus nephritis have compared cyclophosphamide-based regimens with agents such as rituximab, mycophenolate mofetil, and azathioprine, with the hope of providing better clinical outcomes with improved safety profiles. Although some of these newer regimens are now considered first-line treatments of these diseases, their adverse event profiles have been disappointingly similar to those of cyclophosphamide-based protocols. Physicians and patients should consider the adverse event profiles generated by these trials in the context of their extensive use in other patient populations, as well as available measures to prevent such events, when choosing the ideal regimen for an individual patient.

Introduction

The number of therapeutic options for the treatment of ANCA-associated vasculitis (AAV) and lupus nephritis has expanded greatly in the last 30 years. Before this time, treatment regimens were mostly limited to glucocorticoids and alkylating agents. While these agents have no doubt improved clinical outcomes, their use has come with substantial adverse events, particularly given the need for repeated and prolonged doses because of the chronic and relapsing nature of these diseases. Cyclophosphamide (CYC) alone is associated with a variety of worrisome adverse events, including infections, malignancies, alopecia, and reproductive abnormalities, the latter of which is particularly pertinent in diseases such as lupus nephritis, which affect a young patient population.

With the advent of protocols with lower doses and newer immunosuppressive medications, the nephrologist can now individualize prescriptions with both disease response rates and adverse event profiles in mind. However, the experience using these immunosuppressive medications in glomerular disease is much more limited compared with their use in the oncology, rheumatology, and transplant populations. Just as data on efficacy of these medications is limited in glomerular diseases, so too are data on their adverse event profiles. The aim of this article is to review the recent randomized controlled trials in induction therapy for AAV, as well as induction and maintenance treatment of lupus nephritis. We discuss the efficacy of these regimens while highlighting their adverse event profiles. It is this informal risk/benefit ratio that can be considered by nephrologists and patients when individualizing treatment of these glomerular diseases.

Recent Randomized Controlled Trials

Induction Therapy for ANCA-Associated Vasculitis: CYC and Rituximab

First-line induction therapy for AAV has consisted of CYC plus glucocorticoids for many years. Use of this therapy is based on extensive uncontrolled and randomized controlled trials that showed improved remission rates with CYC-based regimens compared with steroid monotherapy and other immunosuppressive agents (1). Multiple randomized controlled trials have compared oral with intravenous pulse CYC therapy, the largest of which found similar median time to remission (3 months) and 9-month remission rates (88%) with oral (2 mg/kg per day) and intravenous (15 mg/kg) regimens in 149 patients with newly diagnosed generalized AAV and renal involvement but not immediately life-threatening disease (2). Leukopenia occurred earlier and more commonly in the oral CYC group (33 episodes versus 20 episodes in the intravenous group), but the numbers of infections were similar in both groups. Hemorrhagic cystitis, malignancy, and amenorrhea were rare in this trial (Table 1). Despite the higher incidence of leukopenia with oral therapy, some experts prefer oral CYC because of the lower rates of relapse observed in other studies (3,4).

Table 1.

Adverse events in recent trials for induction treatment of severe ANCA-associated vasculitis

Study (Reference)	End Points		Adverse Events		Infection		Hematologic Events				Hemorrhagic Cystitis	Cancer
de Groot et al. (50)a	Median time to remission (mo)	Remission rate (%)	Patients experiencing≥1 adverse event (%)	Adverse events (n)	Infections (n)		Leukopenia episodes (n)				Events (n)	Patients with cancer (%)
Oral CYC (n=76)	3	88	77	132	21		33				1	0
iv CYC (n=73)	3	88	77	96	20		20				2	1
RAVE induction (5)	BVAS/WG score 0+successful steroid taper (%)		Patients experiencing≥1 selected adverse event (%)	Adverse events (n)	Severe infection (%)		Leukopenia (%)		Thrombocytopenia (%)		Patients with event (%)	Patients with cancer (%)
RTX (n=99)	64		22	1035	7		3		3		1	1
Oral CYC (n=98)	53		33	1016	7		10		1		1	1
RAVE 18-mo follow-up (6)	Remission at 18 mo (%)		Patients experiencing≥ 1 adverse event (%)	Adverse events (n)	Severe infection (%)		Severe leukopenia (%)				Patients with event	Cancers (n)
RTX (n=99)	39		99	1399	12		5				NN	5
Oral CYC (n=98)	33		100	1420	11		23				NN	2
RITUXVAS (7)	Sustained remission (%)		Patients experiencing severe adverse event (%)		Severe (%)	All (%)	Neutropenia (%)	Anemia (%)		Thrombocytopenia (%)	Patients with event	Cancers (n)
RTX (n=33)	76		42		18	36	6	6		3	NN	2
Control (n=11)	82		36		18	27	9	18		0	NN	0

Gastrointestinal adverse effects not noted in any of the above trials. CYC, cyclophosphamide; iv, intravenous; RAVE, Rituximab in ANCA-Associated Vasculitis; BVAS/WG- Birmingham Vasculitis Activity Score for Wegener’s Granulomatosis; RTX, rituximab; NN, adverse event not noted to have occurred, but not specifically mentioned in the manuscript; RITUXVAS, Rituximab Versus Cyclophosphamide in ANCA-associated Vasculitis.

^aIn the de Groot et al. trial, two episodes of alopecia were recorded in the oral CYC group and amenorrhea was noted in 1% of the intravenous CYC group. Alopecia and reproductive abnormalities were not otherwise noted in the other trials.

Two multicenter randomized controlled trials published in 2010 explored the use of rituximab, a monoclonal chimeric anti-CD20 antibody, in AAV. The Rituximab in ANCA-Associated Vasculitis (RAVE) trial (5) was a noninferiority trial that randomly assigned 197 patients with newly diagnosed or relapsed AAV and manifestations of severe disease to intravenous rituximab (375 mg/m² weekly for 4 weeks) or control (oral daily CYC, 2.0 mg/kg per day, followed by azathioprine once remission was achieved at 3–6 months). Both groups received pulse methylprednisolone (1000 mg intravenously for three doses) followed by prednisone (1 mg/kg per day) tapered to off by 5 months in patients in remission. The primary endpoint was a Birmingham Vasculitis Activity Score for Wegener granulomatosis of 0 and successful taper of prednisone at 6 months. Rituximab met criterion for noninferiority; 64% of patients reached the primary endpoint compared with 53% in the CYC group (P<0.001). Longer follow-up of the RAVE trial found sustained remission rates of 48% and 39% with rituximab at 12 and 18 months, respectively, compared with 39% and 33% in the control group (P<0.001) (6). Similar remission rates were observed between groups in patients with major renal disease (75% complete remission in the rituximab group versus 76% in the CYC group).

The second trial was the Rituximab Versus Cyclophosphamide in ANCA-associated Vasculitis (RITUXVAS) study (7), an open-label trial that randomly assigned 44 patients with AAV and renal involvement to standard therapy (intravenous CYC, 15 mg/kg every 2 weeks for three doses, then every 3 weeks thereafter until remission, followed by azathioprine, 2 mg/kg per day, when in remission) or the addition of rituximab (375 mg/m² weekly for four doses) to a CYC-containing regimen (intravenous CYC, 15 mg/kg, with first and third rituximab doses). Both groups received pulse methylprednisolone followed by oral prednisolone (1 mg/kg per day) with reduction to 5 mg/d by 6 months. Of note, patients in the RITUXVAS were older (average age, 67–68 years) and had more severe renal disease (median eGFR, 20 ml/min per 1.73 m² in the rituximab group and 12 ml/min per 1.73 m² in the control group, with 9 patients requiring dialysis at entry) than those in the RAVE trial (average age, 52–54 years; creatinine clearance, 54 ml/min in the rituximab group and 69 ml/min in the control group in RAVE), and 11 RITUXVAS patients also underwent plasma exchange therapy. The first primary endpoint, sustained remission, was observed in 76% of the patients in the rituximab group and 82% in the control group (P=0.68).

The adverse event profiles reported in these trials have tempered some of the enthusiasm these trials generated for the use of rituximab in AAV. In the RAVE trial, although more patients in the CYC group had one or more selected adverse events (driven mostly by leukopenia), the observed number of selected adverse events, the percentage of patients with a selected adverse event (5% versus 6%), and number of total or severe adverse events were similar in both groups (Table 1). One patient in the rituximab group and two patients in the CYC group died during the follow-up period. Fourteen percent of patients in the rituximab group and 17% of patients in the CYC group had treatment- or disease-related events leading to discontinuation of therapy. Notably, more malignancies occurred in the rituximab group (cumulative malignancy incidence of 5% in patients exposed to rituximab versus 1% of those without rituximab exposure), but this difference was not statistically significant (P=0.26); approximately half of the patients had relapsing disease, with most of these patients previously receiving cyclophosphamide (Table 1). During 18-month follow-up of these patients, the number or rates of total adverse events, serious adverse events, and non–disease-related adverse events did not significantly differ. The previously observed difference in number of patients experiencing at least one serious adverse event was not found at 18 months. Leukopenia was again more common in the CYC group (23% versus 5%), but serious infections were similar in both groups (12% with rituximab versus 11% with control). Of note, the number of pneumonia-related adverse events was significantly higher in the CYC group than in the rituximab group (11% versus 3%; P=0.03).

The rate of serious adverse events was a second primary endpoint in the RITUXVAS trial. At 12 months, the percentage of patients experiencing severe adverse events (42% with rituximab versus 36% with control), the incidence of severe adverse events (1.00 per patient-year in the rituximab group versus 1.10 per patient-year in the control group; P=0.77), and the incidence of infections (0.66 per patient-year in the rituximab group versus 0.60 in the control group) were similar between the two treatment groups (Table 1). At last follow-up, death rates were identical in both groups (18%), and the only two cancers occurred in the rituximab group. Taken together, these data have prevented the designation of rituximab as a clearly safer alternative to CYC as first-line therapy in AAV, but it may be a preferred first-line agent in some patients.

Induction Therapy for Lupus Nephritis: CYC and Mycophenolate Mofetil

Cyclophosphamide-based induction regimens were also considered the definitive first-line therapy for lupus nephritis for many years. This era was ushered by trials at the National Institutes of Health in the 1970s and 1980s, which concluded that treatment with high-dose intravenous CYC (0.5–1.0 g/m² monthly for 6 months, followed by repeat dosing every 3 months) with corticosteroids led to improved renal survival versus steroids alone in the treatment of lupus nephritis (8–10). However, high cumulative CYC doses were often required because of high rates of relapse and treatment-resistant disease. The toxicities associated with these high CYC doses necessitated novel approaches for treating lupus nephritis.

The 2002 Euro-Lupus Nephritis Trial (ELNT) attempted to address this issue by randomly assigning 90 patients with lupus nephritis and preserved renal function (creatinine<1.3 mg/dl) to treatment with “low-dose” intravenous CYC (500 mg biweekly for 6 doses) versus the standard “high-dose” National Institutes of Health regimen (0.5–1.0 g/m² monthly for 6 months, then quarterly for two doses), with all patients receiving induction steroids and azathioprine as maintenance therapy after completing CYC therapy (11). The percentages of patients reaching the primary endpoint of treatment failure (low-dose versus high-dose CYC: 16% versus 20%, respectively), renal remission (71% versus 54%), and renal flares during follow-up (27% versus 29%) were similar between treatment groups. Ten-year follow-up of these patients showed no significant difference in the endpoints of death (4 in the low-dose group versus 2 in the high-dose group), doubling of serum creatinine (6 in the low-dose group versus 5 in the high-dose group) or ESRD (2 in the low-dose group versus 4 in the high-dose group) (12).

Although the mean cumulative CYC dose was lower in the low-dose group during initial (3.0 g versus 8.5 g) and 10-year (5.5 g versus 9.5 g) follow-up, this reduced-dosing regimen was not associated with a clear-cut improvement in adverse events (Table 2). Although fewer severe infections occurred in the low-dose group during the induction period (7 infections in 5 patients versus 17 infections in 10 high-dose recipients), neither this difference nor the cumulative probability of severe infection or number of severe infections per patient-year was statistically significant. The number of patients with leukopenia (11%) and evidence of gonadal toxicity (7%) were identical between treatment groups, and no episodes of hemorrhagic cystitis occurred. Surprisingly, the only two patients who died were in the low-dose group; six cancers were observed in the low-dose group versus only one in the high-dose group. Reproductive rates were similar between the two treatment groups: six patients in the high-dose group experienced 9 pregnancies and 9 patients in the low-dose group experienced 10 pregnancies. The lack of clear-cut safety benefits despite lower CYC dosing propelled the search for novel induction therapies in lupus nephritis.

Table 2.

Adverse events in recent trials for induction treatment of lupus nephritis

Study (Reference)	End Points			Infection			Hematologic Events			Cancer	GI Events			Alopecia	Reproductive Events
ELNT (11)	Treatment failure (%)			Serious		Other	Leukopenia (%)	Toxic anemia (%)		Patients with cancer, % (n)	Patients with event			Patients with event	Menopause (%)	Transient amenorrhea (%)
Low-dose iv CYC (n=44)	16			7 in 5 patients		10 in 5 patients	11	0		3 (1)	NN			NN	4	4
High-dose iv CYC (n=46)	20			17 in 10 patients		10 in 7 patients	11	2		0	NN			NN	2	2
ELNT 10-yr follow-up (12)	Death (%)	SCr doubling (%)	ESRD (%)	Patients with event			Patients with event			Patients with cancer, % (n)	Patients with event			Patients with event	Successful pregnancies
Low-dose iv CYC (n=44)	11	14	5	NN			NN			14 (6)	NN			NN	10 in 9 patients
High-dose iv CYC (n=46)	4	11	9	NN			NN			2 (1)	NN			NN	9 in 6 patients
Ginzler et al. (13)	Cumulative remission (%)			Severe infection (%)	Other infection (%)		Lymphopenia (%)		Anemia (%)	Patients with cancer	Upper GI symptom s (%)	Diarrhea (%)		Patients with event	Menstrual irregularities (%)	Amenorrhea (%)
MMF (n=83)	52			1	7		22		2	NN	28	18		0	10	0
iv CYC (n=75)	30			4	7		37		3	NN	33	3		10	15	3
ALMS Inductiona (14)	Decrease in UProt/Cr and Stable SCr (%)			Patients with ≥1 infection (%)			Anemia (%)			Patients with cancer	Nausea (%)	Vomiting (%)	Diarrhea (%)	Patients with event (%)	Patients with event
MMF (n=185)	56			69			13			NN	15	14	28	11	NN
iv CYC (n=185)	53			62			7			NN	46	38	13	36	NN

Hemorrhagic cystitis was not noted in any patients in these trials. GI, gastrointestinal; ELNT, Euro-Lupus Nephritis Trial; iv CYC, intravenous cyclophosphamide; NN, adverse event not noted to have occurred, but not specifically mentioned in the manuscript; SCr, serum creatinine; MMF, mycophenolate mofetil; ALMS, Aspreva Lupus Management Study; UProt/Cr, urinary protein-to-creatinine ratio.

^aALMS reported events only when the incidence of that event was >10%.

Many hoped that mycophenolate mofetil (MMF), an inosine monohydrate dehydrogenase inhibitor, would prove to be a safer alternative to CYC in lupus nephritis given its efficacy and safety in kidney transplant recipients and in uncontrolled case series in lupus nephritis. Two large, multicenter, prospective randomized controlled trials have compared CYC with MMF (13,14). These trials randomly assigned 140 and 370 patients, respectively, to 6 months of steroids plus either oral MMF (goal daily dose of 3000 mg/d) or high-dose pulse intravenous CYC. The trial by Ginzler et al. found higher rates of complete remission with MMF (22.5% versus 5.8% in the CYC group) and similar partial remissions in both groups (29.6% in the MMF group versus 24.6% in the CYC group), whereas the Aspreva Lupus Management Study (ALMS) induction trial found no difference in the primary endpoint (a prespecified decreased in urine protein-to-creatinine ratio and stabilization in serum creatinine) in both groups (56% in the MMF group versus 53% in the CYC group).

These two trials supported the use of MMF as an alternative first-line agent to intravenous CYC to induce disease remission in lupus nephritis, but failed to show that MMF was a definitively safer medication. While CYC treatment was associated with a higher incidence of severe infections (4% versus 1%) and leukopenia (37% versus 22%) in Ginzler and colleagues’ study, this improved safety profile was not replicated in the ALMS induction trial (Table 2). As expected, more reproductive irregularities and alopecia were observed with CYC, whereas diarrhea was more common in the MMF group. Surprisingly, in the ALMS trial, 13% of patients in the MMF group withdrew from therapy because of adverse events versus 7.2% in the CYC group (P=0.07), and more patients died in the MMF group than in the CYC group (9 versus 5; P=0.29); however, neither of these results statistically significantly differed between treatment groups. The higher mortality rate in the MMF group could not be fully explained, but most deaths were from infections and occurred in patients who had severe renal disease and respiratory signs and symptoms that rapidly worsened soon after enrollment. The use of high-dose corticosteroids and MMF could have contributed to deaths in this group of very ill patients. No malignancies were reported in either study. Although these trials ushered in a new era in the treatment of lupus nephritis, they failed to establish MMF as a reproducibly more effective and safe alternative to CYC.

Maintenance Therapy for Lupus Nephritis: MMF and Azathioprine

As occurred with induction therapy, efforts have been made to avoid the prolonged CYC and steroid use in the maintenance treatment of lupus nephritis. A small, randomized trial first evaluated the use of steroids with MMF, azathioprine, or quarterly intravenous CYC as maintenance therapy for lupus nephritis following induction with monthly high-dose intravenous CYC (15). Improvements in patient and renal survival and lower rates of adverse events (hospitalizations, infections, amenorrhea, nausea, and vomiting) were observed with MMF and azathioprine compared with CYC (Table 3). Notably, four of the five deaths in this study were in the CYC maintenance group and were due to sepsis. Subsequent trials therefore focused on non–CYC-based regimens. The Mycophenolate Mofetil Versus Azathioprine for Maintenance Therapy of Lupus Nephritis (MAINTAIN) trial (16) and the ALMS maintenance trial (17) followed, comparing MMF (target dose, 2 g/d) versus azathioprine (target dose, 2 mg/kg per day) head-to-head as maintenance therapy in lupus nephritis for 3-year follow-up periods. MAINTAIN found no significant difference in the incidence of renal flares (19% in the MMF group versus 25% in the azathioprine group) in 105 European patients who had received ELNL intravenous CYC induction therapy, whereas MMF had improved efficacy in preventing the primary endpoint of time to renal failure, death, or doubling of serum creatinine (hazard ratio, 0.44; 95% confidence interval, 0.25 to 0.77; P=0.003) compared with azathioprine in 227 patients in the ALMS maintenance trial.

Table 3.

Adverse events in recent trials for maintenance treatment of lupus nephritis

Study (Reference)	End Points		Infection			Hematologic Events		Cancer	GI Events			Alopecia	Reproductive Events
Contreras et al. (15)a	Patient survival (%)	Renal survival (%)	Total		Severe	Leukopenia		Patients with cancer	Nausea	Vomiting	Diarrhea	Patients with event	Amenorrhea
MMF (n=20)	95	95	32		2	2		0	14	10	12	NN	6
AZA (n=19)	100	80	29		2	6		0	7	4	9	NN	8
iv CYC (n=20)	80	74	77		25	10		0	65	55	12	NN	32
MAINTAIN (16)b	Renal flare (%)		Benign infections (n)	Severe infections (n)		Leukopenia (n)	Anemia (n)	Patients with cancer (n)	Nausea/diarrhea (n)			Patients with event (n)	Transient amenorrhea (n)
MMF (n=53)	19		21	1		2	0	0	8			0	2
AZA (n=52)	25		14	0		11	2	1	8			2	1
ALMS maintenance (17)	Treatment failure (%)		Total (%)	Severe (%)		Leukopenia (%)	Anemia (%)	Patients with cancer (%)	Nausea (%)	Vomiting (%)	Diarrhea (%)	Patients with event (%)	Patients with event
MMF (n=116)	16.4		79	10		23	11	0	17	12	19	10	NN
AZA (n=111)	32.4		78	12		36	13	1	19	16	18	10	NN

GI, gastrointestinal; MMF, mycophenolate mofetil; NN, adverse event not noted to have occurred, but not specifically mentioned in the manuscript; AZA, azathioprine; iv CYC, intravenous cyclophosphamide; MAINTAIN, Mycophenolate Mofetil Versus Azathioprine for Maintenance Therapy of Lupus Nephritis; ALMS, Aspreva Lupus Management Study.

^aAdverse events in the trial by Contreras et al. are expressed as rate of event/patient-year.

^bAdverse events in the MAINTAIN trial are expressed as number of episodes of that event; in the MAINTAIN trial adverse events were not noted once the primary end point was met.

Leukopenia was significantly more common with azathioprine treatment in both trials (Table 3). Infections were not significantly different between treatment groups in either trial, and the three malignancies observed in the two trials were all in patients treated with azathioprine. While the MAINTAIN trial found a similar incidence of transient amenorrhea (two patients in the MMF group versus one in the azathioprine group), no reproductive abnormalities were found in the ALMS trial, and the gastrointestinal adverse effects of both medications were similar (Table 3). These data propelled MMF and azathioprine to be considered first-line maintenance therapies for lupus nephritis, with some data suggesting that MMF has some benefit in safety and efficacy.

Discussion

These recent prospective and randomized trials for induction treatment of AAV and induction and maintenance treatments of lupus nephritis help to inform physicians and patients about the adverse event profiles of the new generation of first-line immunosuppression for these diseases. Although nephrologists hoped that these new regimens would prove to be safer and more efficacious, the adverse event profiles observed in most of these trials were not clearly superior to those of older strategies. It should be noted that high-dose corticosteroids were used during the induction period in all treatment groups in the trials discussed above, and that steroid treatment may contribute to the similar adverse event profiles observed. Moreover, the frequent monitoring in these trials may have resulted in a decreased incidence of some adverse events (e.g., cytopenias in cyclophosphamide-treated patients in whom a dose-adjustment was made). Here we discuss selected adverse effects of these therapies in the context of older treatment regimens, as well as their use in other disease populations.

Infections

Infections remain one of the most common and concerning adverse events in treating AAV and lupus nephritis. Discussing the risk of infection with different therapies is challenging given the variability in reporting of infections (all infections versus only severe infections, number of patients with an infection versus total number of infections) and duration of follow-up in trials. In one AAV cohort of 158 patients (most of whom were treated with CYC and steroids), 46% of patients experienced a severe infection during a mean follow-up of 8 years (1). CYC with steroids has also been associated with an increased risk of infections in patients with lupus (32%–45% with combination therapy compared with 7%–12% with steroids alone) (10,18,19). In the induction trials for AAV discussed earlier, number and incidence of infections did not differ between treatment groups, (Table 1), and in lupus nephritis induction trials, only the trial by Ginzler et al. demonstrated a difference in severe infections (4% in the CYC group versus 1% in the MMF group), with identical percentages of patients experiencing other infections (Table 2). In the maintenance trials in lupus nephritis, MMF and azathioprine clearly had better infection profiles than did quarterly intravenous CYC in the study by Contreras et al., but neither MMF nor azathioprine demonstrated consistently lower incidences of infection in the MAINTAIN and ALMS maintenance trials (Table 3).

Measures should be implemented to prevent infections when possible. Pneumocystis jiroveci pneumonia (PCP) in immunocompromised patients remains a concern but was not noted in any of the trials discussed here. The incidence of PCP in the rheumatology population has been estimated at 1%–2% based on limited case series data from the 1990s (20). One series reported an incidence of 6% in patients with AAV (21), most of whom were treated with CYC, and a separate review reported an incidence of 1.2% in patients with AAV treated with rituximab (22). Chemoprophylaxis was prescribed (5) or recommended (7) in both AAV trials above but was not noted in the lupus nephritis trials. The use of PCP chemoprophylaxis during CYC treatment is recommended by many experts (23,24) and should be considered with rituximab treatment as well. There are conflicting data about whether MMF may protect against or increase the risk of PCP in renal transplant recipients (25).

Attenuated vaccinations, particularly the seasonal influenza vaccines and pneumococcal vaccines, should be strongly encouraged in these patients, and tuberculosis screening may be considered in selected patients prior to treatment. The reactivation of hepatitis B virus with rituximab use for rheumatoid arthritis and AAV has been published in case reports, but the large clinical trials with rituximab in rheumatoid arthritis required hepatitis B screening before treatment (26). Extrapolating from this experience, hepatitis B screening should be conducted prior to rituximab therapy. Moreover, CYC and repeated courses of rituximab may result in hypogammaglobulinemia, which may in turn predispose patients to increased risk of infections. Intravenous immunoglobulin therapy may be considered in these patients (27).

Cytopenias

The development of cytopenias is common with many immunosuppressive regimens in AAV and lupus nephritis. Monthly pulse intravenous CYC commonly produces leukopenia with a nadir 1–2 weeks after administration, and while oral CYC administration allows more ease for dose adjustment, it also leads to higher cumulative drug exposure. In the trial by de Groot et al., more leukopenic episodes (but not infections) were noted in the oral CYC group (Table 1), but interestingly, the incidence of leukopenia was identical (11%) with low- and high-dose intravenous CYC regimens in the ELNT trial (Table 2).

Rituximab use has been associated with the development of late-onset neutropenia in both the hematologic oncology and rheumatology populations. One rheumatology case series noted late-onset neutropenia in 11 of 209 (5%) patients after a median 102 (range, 40–362) days from last rituximab treatment, with higher incidences in patients with AAV and lupus (28). The incidence of neutropenia was 6% in RITUXVAS and 5% in the long-term follow-up of the RAVE trial (specified as ≥grade 2 leukopenia).

Cytopenias are commonly encountered with both MMF and azathioprine use in the transplant populations. In the lupus nephritis induction trials noted earlier, MMF was associated with fewer lymphopenias and similar anemia rates as intravenous CYC (Table 2), and for maintenance therapy had improved cytopenia profiles compared with azathioprine (Table 3). Regardless, patients receiving both MMF and azathioprine require frequent monitoring of their blood counts and may require dose adjustment if cytopenias occur. Importantly, physicians must be aware of medication interactions that can result in severe cytopenias.

Malignancies

Treatment-associated malignancies remain a major concern for patients and physicians with conventional therapies for AAV and lupus nephritis. An analysis of four clinical trials comprising 535 European patients with AAV, 93% of whom received CYC, found a long-term malignancy risk that was 1.6–2.4 times higher than in the general population (29). While much of this risk is driven by nonmelanoma skin cancers, the incidence of other malignancies, such as bladder cancer and acute myeloid leukemia, also increased in many studies. This risk may be related to cumulative CYC dosing. One Danish cohort of 293 patients with AAV found the risks for bladder cancer (n=5) and acute myeloid leukemia (n=3) were increased only in patients who received a cumulative dose of >36 g of CYC (standardized incidence ratio, 9.5 and 59.0, respectively), whereas the risk for nonmelanoma skin cancer was increased at both <36 and >36 g of cumulative CYC use (standardized incidence ratio, 3.9 and 5.2, respectively) (30). Although 2-mercaptoethanesulfanic acid (mesna), a thiol compound that binds the CYC metabolite acrolein, is commonly coadministered with CYC in an attempt to reduce the incidence of hemorrhagic cystitis and bladder cancer, no studies in humans support a decrease in malignancy risk with its use (31).

None of the randomized controlled trials described here demonstrated lower malignancy risks in patients treated with alternative regimens to high-dose intravenous CYC in AAV or lupus nephritis. On the contrary, in the ELNT trial, more malignancies (n=6) were observed during the 10-year follow-up compared with the high-dose CYC group (n=1), and more malignancies were observed in the rituximab treatment groups in the RAVE and RITUXVAS trials. The reason for the latter observation is unclear. Although rituximab has been associated with increased long-term risk of solid malignancies in patients with lymphoma (32), this association has not been shown in the rheumatoid arthritis population (26,33). Patients should be educated about these risks when choosing between treatment regimens.

Although patients with solid organ transplant have higher risks of malignancies (particularly nonmelanoma skin cancers), MMF use has not been shown to be an independent risk factor for the development of malignancies in patients with kidney (34) or heart (35) transplants. Some studies actually suggest that MMF may reduce malignancy risk in these populations. Although some data have associated azathioprine use with malignancies in rheumatoid arthritis (36), these data are controversial (37), and no such link has been established in AAV or lupus. The long-term follow-up of the preceding trials will be important in better defining long-term malignancy risks for the current first-line treatment regimens in AAV and lupus nephritis.

Infertility

Treatment-related infertility is especially relevant in the treatment of glomerular diseases that predominantly affects patients in their reproductive years. The incidence of infertility with different regimens is difficult to estimate due to inconsistencies in the measurement of reproductive abnormalities in these trials, especially in male patients. Moreover, the higher rates of reproductive abnormalities observed with prolonged and higher-dose CYC regimens may significantly differ with its replacement by low-dose CYC and alternative medications.

Concerning female infertility, the percentage of patients in the trials above who developed menopause during induction therapy for AAV of lupus nephritis was consistently <5%. Past trials in patients with SLE and AAV receiving pulse intravenous CYC have demonstrated rates of amenorrhea of 15%–27%, with higher rates of transient amenorrhea in some studies (38,39). This risk may be associated with older age at time of treatment and higher cumulative intravenous CYC doses. The use of gonadotropin-releasing hormone agonists may increase ovarian preservation and fertility rates in women taking CYC for rheumatologic disease or malignancies (40) and should be discussed with women of childbearing age in whom CYC therapy is planned. Rituximab, MMF, and azathioprine are not felt to increase the risk of amenorrhea or infertility in females, and in the trials discussed here, no episodes of amenorrhea were noted in patients treated with rituximab for AAV. Sexually active women in whom CYC or MMF are being considered should be counseled on its potential teratogenic effects, and the use of birth control should be strongly encouraged.

CYC can cause decreased sperm counts and infertility in men. The incidence of these adverse events is not known with contemporary CYC regimens for AAV and lupus nephritis, and it was not clear if detecting male infertility was attempted in any of the above trials. Options for preservation of fertility in men include cryopreservation (sperm banking) and the use of intramuscular testosterone before and during CYC therapy. The latter approach is thought to decrease sperm sensitivity to CYC through inhibition of germinal cell activity, and has shown promising effects on recovery of sperm counts in one small randomized controlled trial and one small case series of glomerular disease treated with CYC (41,42). Currently no data suggest that rituximab, MMF, or azathioprine affects male fertility.

Gastrointestinal Adverse Effects

Nausea and vomiting are perhaps the most well known adverse effects of intravenous CYC therapy for malignancy. Indeed, it is classified as a high-risk chemotherapy for emesis (>90% of patients experience emesis after its use) (43). The incidence of nausea and vomiting with the dosing used in current first-line therapy for AAV and lupus is unknown, but premedication with regimens such as dexamethasone plus ondansetron may eliminate these symptoms (44) and are recommended with intravenous CYC administration. Rituximab is characterized by the American Society of Clinical Oncology as having minimal risk for emesis (<10% experience symptoms) (43). It is assumed, but not specified, that the preceding trials used antiemetic regimens, and gastrointestinal adverse effects were not noted in any patients in either group during short- or long-term follow-up of the AAV trials or the ELNT trial (Table 2).

Gastrointestinal adverse effects with MMF use are well known to nephrologists who care for patients after kidney transplantation. In two large randomized controlled trials comparing MMF with azathioprine (in combination with cyclosporine and prednisone) after kidney transplant, diarrhea occurred in 28%–37% of patients treated with MMF, compared with 17%–24% of patients receiving azathioprine (45,46); more symptoms were associated with higher MMF doses. Nausea (14%–20% in the MMF group versus 20% in the azathioprine group), vomiting (12%–16% versus 6%), and abdominal pain (26%–31% versus 23%) were also common. The incidence of these adverse effects was similar in the lupus nephritis trials above (Tables 2 and 3), with more events during the induction trials (goal MMF dose, 3 g/d; actual achieved dose, 2.6–2.7 g/d) compared with maintenance treatment (goal MMF dose, 2 g/d; actual dose, 1.6–2.0 g/d). Strategies for improving diarrhea include dose reduction, changes in dosing schedule (i.e., changing from 1000 mg of MMF twice daily to 500 mg four times daily), and the use of enteric-coated mycophenolate sodium instead of MMF.

Additional Treatment-Specific Adverse Events

Infusion-related reactions with rituximab occur in >50% of patients treated with rituximab, most commonly during the first infusion. These may be prevented or attenuated using premedication with acetaminophen, antihistamines, and intravenous glucocorticoids. Decreasing the infusion rate may also permit resumption of the infusion after resolution of symptoms. Serum sickness has been reported with subsequent rituximab treatments up to 10 days after the last dose (47). It is unclear whether the development of serum sickness is related to the development of human antichimeric antibodies to rituximab. Rare cases of JC virus–associated progressive multifocal leukoencephalopathy have been reported in rituximab-treated patients (reported incidence <1:20,000 with extensive use in rheumatoid arthritis [26]), but it is not clear whether there is a true association between rituximab and development of progressive multifocal leukoencephalopathy. Noninfectious pulmonary toxicity has also been reported in rituximab-treated patients (three cases reported in patients with rheumatoid arthritis) (48,49).

Conclusion

The many randomized, controlled trials conducted in the last 15 years in AAV and lupus nephritis have resulted in improved outcomes and expanded options for treatment regimens in these diseases. However, the hope that these new regimens would provide evidence for treatments that are both more efficacious and safer than older regimens has not been reproducibly realized by the adverse events observed in these trials. Nonetheless, these trials provide important prospective data on the adverse effects of these medications, allowing physicians and patients to consider an individualized risk/benefit ratio for each patient, a luxury that was not available when CYC plus steroids was the only recommended treatment. However, as a whole, the current data do not support the conclusion that newer treatments are definitively safer in the treatment of AAV and lupus nephritis. Further research is needed toward improving the safety profiles of existing regimens and developing newer and safer treatment protocols.

Disclosures

None.