MAINRITSAN is the first and only randomized study to compare RTX and AZA as maintenance therapy in patients with GPA (n=87), MPA (n=23) or renal-limited vasculitis (n=5).31 The study (controlled, open and multicenter) included 115 patients (80% with new diagnosis, 20% with recurrent disease) in complete remission after induction therapy with GC and CPM. The maintenance protocol was based on the administration of prednisone (approximately 5mg/day for at least 18 months) in combination with AZA (n=58, 2mg/kg/day for 1 year, 1.5mg/kg/day for 6 months and 1.0mg/kg/day for the last 4 months) or RTX (n=57, 500mg initially on days 0 and 14 and, thereafter, every 6 months, for a total of 5 applications). The percentage of patients with severe relapses (recurrence or worsening of the vasculitis with manifestations that were life-threatening or affected an essential organ, like the brain or heart) after 28 months of follow-up (primary endpoint) was 5% (n=3) among patients treated with RTX and 29% (n=17) among those treated with AZA, demonstrating a clear superiority of the monoclonal antibody (P=.002). Among the patients who relapsed with AZA, in 8 it occurred during the first year; in 2, between months 12 and 22; and in 7, after the complete discontinuation of the medication (months 24–28). In the case of RTX, there was 1 relapse per month in months 8, 22 and 24. With respect to minor relapses, there were no significant differences between the two groups in terms of the rate of adverse events or the incidence of infections.

Although the results of the study with RTX are very promising, with one of the lowest rates of relapse reported in these vasculitides, it strikes us as interesting that 41% of the flares recorded in the AZA group occurred when the patients were no longer receiving the medication, possibly causing the therapeutic effect of the drug to be underestimated. However, the protocol for maintenance with AZA is similar to that utilized previously in two large-scale clinical trials involving the same type of patients, and the results of which are similar to those observed in the MAINRITSAN trial.5,8 As occurs with other IS,32 treatment with AZA or RTX for a longer period of time than that evaluated in the MAINRITSAN trial could be related to a higher rate of sustained remissions, a possibility that should be examined in clinical trials with longer follow-up times.

In short, the MAINRITSAN trial showed that the combination of low doses of GC and RTX at fixed intervals (every 6 months, which is the average duration of the therapeutic effect33) can be considered a valid, effective and safe alternative for maintenance of remission in patients with AAV, mainly those with newly diagnosed GPA. At the present time, two clinical trials to evaluate RTX as maintenance therapy are being developed. The first, MAINRITSAN 2 (clinicaltrials.gov NCT01731561), will compare two strategies for RTX administration (500mg at fixed intervals of 6 months vs 500mg whenever increases in CD19+ lymphocytes or ANCA are detected). The other study, RITAZAREM (Rituximab Vasculitis Maintenance Study) (NCT01697267), will evaluate patients with refractory AAV receiving RTX (1g every 4 months) or AZA for 2 years.

The studies that are to be reviewed below disclose certain differences and limitations that should be taken into account for a correct interpretation of the conclusions derived from them. These series evaluated heterogeneous populations, the dose and regimen of RTX administration varied, and there are differences in the reported maintenance protocols, as the drug was utilized: (1) at fixed intervals in a so-called prevention strategy; (2) depending on serological markers, such as an increase in the ANCA level or B lymphocyte reconstitution; or (3) only at the time of relapses.

In the vast majority of the reported series, it has not been possible to establish a relationship between the change in ANCA levels and the development of recurrence.5,34,36,37,39,41,44,45,50,54–56 In these studies, the percentages of patients who became ANCA-positive or in whom the levels increased, and who had relapses or remained in remission ranged from 33% to 77% and from 38% to 66%, respectively, which shows that the sensitivity of this variable is too low to serve as the basis for decisions on maintenance therapy. The only study that found an association between ANCA levels and recurrences reported that seroconversion to ANCA positivity during the post-RTX period was associated with a 7-fold higher risk of having severe relapses.43

With respect to the specificity of ANCA, it has recently been reported that the presence of PR3-ANCA, but not a change in the levels of these antibodies, was related to higher long-term risk of relapses.37,43,46,50

One important conclusion derived from the RAVE extension studies is that the possibility of RTX-treated patients having a relapse is extremely low as long as the assays for ANCA are negative and B lymphocytes remain undetectable in peripheral blood.46,50 In fact, only 1 of the 76 patients included in these studies had a flare while these two markers were below the level of detection.

As in the case of ANCA, the levels of these cells in peripheral blood have not been shown to be an accurate enough test for the prediction of flares.31,37,39,41,44,56 Depending on the study analyzed, the reported finding is that approximately 50% of the relapses occur prior to B cell repopulation, that 44%–88% of the patients who relapse have detectable B lymphocytes, and that between 25% and 66% of those patients in whom normal levels of B lymphocytes are detected do not have relapses.5,34,36,37,45,46 Moreover, the time between B cell reconstitution and the development of relapses is not constant either; some patients relapse coinciding with lymphocyte repopulation, whereas in others, it may take some time (average 80±85 days).57

One of the few series in which a good correlation was found between AAV reactivation and B lymphocyte repopulation reported that the development of flares (n=32) was always preceded by the detection of these cells in peripheral blood, with a median between the two events of 8.5 months.38 Interestingly, in this report, all the recurrences were preceded by or associated with an increase in PR3-ANCA.

A recent report found that those patients in whom B lymphocytes were detectable within the first year after treatment with RTX had shorter periods of remission than those in whom repopulation took over 12 months.43

The possible hypotheses to explain the poor correlation between the number of B lymphocytes and the AAV activity include: (1) the existence of niches or “sanctuary sites” for self-reactive B cells in tissues or sites of active inflammation (mainly granulomatous)24; these cell groups would be protected from the pharmacological action of RTX, remaining active before an increase in lymphocytes was observed in peripheral blood21,58; (2) changed dynamics in B cell levels due to a synergic effect between RTX and other conventional IS, such as AZA or CPM, that are frequently administered to patients with resistant or refractory disease59; (3) racial differences between the populations studied, as the elimination of the B lymphocytes depends not only on the RTX dose, but on polymorphisms in the cell surface receptors (FcγRIIIa)60; and (4) differing definitions from one study to another, with varied absolute values for peripheral blood B cells and diverse phenotypes evaluated (CD19+ or CD20+), as a consequence of the use of detection techniques (flow cytometry) with greater or lesser sensitivity.36

A recent publication stated that the detailed phenotypic analysis of subpopulations of B lymphocytes might have the potential to predict relapses. In a study of 54 patients with AAV, it was observed that when <30% of the B lymphocyte repopulation was constituted by CD5+ cells (possibly regulatory B cells), the time to the first recurrence was shorter. This cell subpopulation also correlated negatively with the general disease activity.61 Similarly, it has been reported that, when the B lymphocyte reconstitution is composed mainly of memory cells, there is an increased risk of recurrences.53 This is similar to what occurs in systemic lupus erythematosus, in which repopulation with a higher proportion of CD27+ B lymphocytes (memory cells) has been related to a higher risk of flares.62,63

Mild allergic reactions are the adverse effect most frequently observed (10%–50% of the patients).33,64,51 They usually occur within the first 24h after treatment, and the manifestations are fever, chills, hyper- or hypotension, fatigue, headache, erythema, urticaria, nausea and/or vomiting.33,51

These are the most important complications, and mild or severe infections are reported in 7%–50% of the patients (incidence of 2.6–15.5/100 patient years, similar to that of CPM).5,36,37,42,46,49,50,56,57,65–67 The most common infections are those of the respiratory tract (mainly of viral origin) and the urinary tract, followed by mucocutaneous infections (herpes simplex, herpes zoster and candidiasis).33,36,37,41,46,49,50,65 Although the majority of these episodes are mild, 14%–27% are severe and 3%–29% of the patients develop chronic infections.36,42–45 Infections due to opportunistic microorganisms are rare, and RTX is a safe alternative in patients with a history of tuberculosis.38,39,68

The reported risk factors for the development of severe infections in these patients are high cumulative doses of CPM and the combination of RTX with other cytotoxic drugs.45,69 Although the presence of neutropenia or hypogammaglobulinemia could be associated with increased susceptibility to infectious complications, at present, there is no convincing evidence to prove this.37,38,44

This adverse effect has been reported in 26%–41% of the patients at some point during maintenance with RTX, but is a transient and self-limiting event in more than 50% of the cases.37,43,70

In the majority of the patients, the most marked decrease in the immunoglobulin level was observed during the initial treatment cycles (60% of all the episodes occurring with total doses ≤2g and 70% of those with doses ≤6g).71

Although hypogammaglobulinemia is usually mild to moderate, approximately 7% of the patients have a severe and/or permanent deficiency, requiring discontinuation of the drug and its substitution by intravenous gammaglobulin.43,44,57,51,72 In this respect, this adverse effect has not been directly related to an increase in the development of severe infections,37,38,44 although it could be a contributing factor in patients with other predisposing conditions.73

Among the risk factors that have been associated with the presence of hypogammaglobulinemia, we point out a high cumulative dose of CPM (prior to the initiation of RTX), low baseline immunoglobulin and CD4+ levels, and the concomitant use of the monoclonal antibody with cytotoxic drugs.57,70 Interestingly, some studies have related hypogammaglobulinemia to the number of RTX cycles.74,75 However, this finding has not been corroborated in series with longer follow-up periods.57,51,70,71

This develops in 3%–5% of the patients and is generally transient.5 In the majority of the cases, it has no effect on the patient's clinical status, although there may be cases in which the use of a colony-stimulating factor is required.40,44,76,77

The published literature includes cases of pancytopenia,33,39 macular edema78 and severe pulmonary complications (interstitial disease) in patients with AAV.48 Moreover, there are reports of progressive multifocal leukoencephalopathy in patients with systemic lupus erythematosus and RA,79,80 and a case (unpublished) of this complication in AAV has been detected.48

The authors declare that no experiments were performed on humans or animals for this study.

The authors declare that no patient data appear in this article.

The authors declare that no patient data appear in this article.