Clinical trials (any kind of clinical trial) including those versus placebo and those comparing rituximab with any other MS therapy alone or in combination. English and Spanish were considered as possible languages.

Adult patients (≥18 years old) diagnosed with MS McDonald or Poser criteria [11], [12], [13] affected by any type of MS (RR, SP, PP or PR).

Although rituximab is used to treat neuromyelitis optica, this demyelinating disorder was excluded from the review due to the different pathogenic processes underlying this condition compared to MS.

We considered three types of interventions: i) rituximab alone (comparing disease activity on treatment with baseline activity), ii) rituximab versus placebo or versus any other MS therapy and iii) combination of rituximab with steroids versus placebo or steroids alone.

As primary outcomes we considered: i) the proportion of patients with confirmed disability progression at one year, defined as an increase of at least 1 point in the baseline Expanded Disability Status Scale (EDSS) [14] if the baseline is <6 or an increase of 0.5 point in the EDSS if the baseline EDSS is ≥6 and ii) the proportion of patients who withdrew from the study due to major side effects of the drug (causing death or hospitalization of the patient), and patients with grade 4 adverse events (AEs).

As secondary outcomes we considered: i) the annualized relapse rate at the end of the study period and ii) the number of gadolinium-enhancing (Gad-enhancing) lesions on the T1 brain MRI scans and number of new T2 bright lesions on the brain MRI scans.

Studies 1 [16] and 2 [17] recruited adult patients (18–55 years old) with a diagnosis of RRMS with at least one relapse in the preceding year and with an EDSS at entry between 0 and 5.0 (both inclusive).

For study 3 [18] adult patients between 18 and 65 were included, with a diagnosis of PPMS and a disease duration ≥1 year. Baseline EDSS was between 2.0 and 6.5, inclusive, with a functional system scale score of ≥2 for the pyramidal system or gait impairment due to lower extremity dysfunction and presence of IgG oligoclonal bands or elevated CSF IgG or both.

Study 4 [19] is an add-on therapy study that included adult patients aged 18–65 years old diagnosed with RRMS, with a baseline EDSS ≤6.5, treated with an injectable first-line disease-modifying therapy (DMT) for at least 6 months, with breakthrough disease defined as having a clinical relapse in the prior 18 months while taking the DMT, and at least one Gad enhancing lesion on any of 3 monthly pretreatment brain MRI scans.

Patients received brain MRI scans with and without gadolinium at baseline for the four studies and at weeks 4,8,12,24,36,48,60 and 72 for study 1 [16], at weeks 4,12,16,20,24,28,36 and 48 for study 2 [17], at weeks 4,48 and 96 for study 3 [18] and at weeks −8, −4, 12, 16 and 20 for study 4 [19].

Exclusion criteria for study 2 [17] included neuromyelitis optica, PPMS, SPMS or progressive relapsing forms, relapse within 30 days, prior treatment with cyclophosphamide or mitoxantrone within 12 months, systemic corticosteroid therapy within 30 days, treatment with interferon beta, glatiramer acetate, natalizumab, plasmapheresis or intravenous immune globulin within 60 days or non-lymphocyte-depleting immunosuppressive therapies within 90 days. For study 3 [18] the exclusion criteria included history of MS exacerbation or neuromyelitis optica, history of myelopathy or neurodegenerative central nervous system conditions, systemic autoimmune disorders, recurrent or chronic infections, recent treatment with immunomodulating or immunosuppressant therapies and metabolic, hematologic or immunologic laboratory abnormalities. For study 4 [19] patients with other medical illness, aspartate aminotransferase, alanine aminotransferase or creatinine twice normal upper limit, prior use of a major immunosuppressive agent (cyclophosphamide, mitoxantrone, cladribine, natalizumab or other monoclonal therapeutics) or the use of methotrexate or azathioprine within 6 months were excluded. Exclusion criteria were not specified for study 1 [16].

Study 1 [16] was an open-label trial in which patients received 1,000 mg of IV rituximab on days 1 and 15 and a repeat course on weeks 24 and 26. In studies 2 [17] and 3 [18] patients were randomly assigned 2∶1 to rituximab (same regimen than for study 1) or placebo at weeks 0,2 for study 2 and at weeks 0,2,24,26,48,50,72,74 for study 3. Study 4 [19] was an add-on therapy, and subjects continued their ongoing first-line DMT throughout the study and rituximab was added at weeks 1,2,3 and 4. The dose was 375 mg/m² IV weekly x 4.

Studies 1 [16], 2 [17] and 4 [19] administered premedication (acetaminophen and diphenhydramine hydrochloride) to prevent adverse events before rituximab infusion.

The primary outcome for study 1 [16] was the safety of rituximab, determined by adverse events (AEs) and serious AEs, including worsening MS, number and severity of infusion-associated events (defined as those reported during or within 24 hours of an infusion), number and severity of infectious AEs, any clinically significant changes in laboratory or vital sign measurements, the incidence of human anti-chimeric antibodies (HACA) and the total number of Gad-enhancing T1 lesions over the 72-week trial.

For study 2 [17], the primary outcome was the cumulative number of Gad-enhancing lesions on T1-weighted MRI brain scans at weeks 12, 16, 20 and 24.

In study 3 [18] the primary outcome was the time to confirmed disease progression (CDP), defined as a sustained EDSS increase of ≥1 point from baseline EDSS if the baseline EDSS was between 2.0 and 5.5 or an EDSS increase of ≥0.5 points if the baseline EDSS was >5.5 points (if change not attributable to another etiology) sustained for ≥12 weeks.

Study 4 [19] considered the reduction in the sum of Gad-enhancing lesions on T1 MRI brain scans at weeks 12, 16 and 20, compared to the MRI scans obtained at −8, −4 and 0 weeks.

The secondary outcomes for study 1 [16] were proportion of patients experiencing a confirmed relapse and number of relapses per patient during the study, change from baseline in the total number of Gad-enhancing T1 lesions, total number of new T2 lesions and cumulative volume of T2 brain lesions. Finally another outcome included the CD19+ lymphocyte counts to evaluate the pharmacokinetic and pharmacodynamic profile of rituximab.

In study 2 [17], the secondary outcomes were the proportion of patients with relapses, the annualized relapse rate, the total number of new Gad-enhancing lesions on T1-weighted brain MRI scans at weeks 12,16,20 and 24 and the change from the baseline lesion volume on T2-weighted MRI scans.

In study 3 [18], the secondary outcomes included change from baseline to week 96 in the volume of T2 lesions and change in brain volume (brain parenchymal fraction) on brain MRI scans. Exploratory outcomes included time to CDP sustained for ≥24 weeks, change in Multiple Sclerosis Functional Composite (MSFC) summary and component scale scores, CD19+ B-cells counts and immunoglobulin levels and pre-specified subgroup analysis by age, gender, race/ethnicity, use of prior therapy, EDSS at baseline, duration since MS symptom onset and presence of Gad-enhancing lesions on brain MRI at baseline.

Finally, in study 4 [19], Multiple Sclerosis Functional Composite Scale (MSFC) was performed and mean MSFC at weeks −4, 0 was compared to mean MSFC at weeks 24, 28 and 32. Change in EDSS was also explored comparing EDSS at week 32 to baseline. The correlation between CSF B and T cell counts and MRI response and the impact of rituximab on neutralizing antibodies to interferon beta were also explored. Finally, adverse events were recorded.

For both study 2 [17] and 3 [18] there was a neurologist clinically evaluating the patient, blinded to treatment assignment in addition to a treating investigator handling tolerability issues to study drug. For studies 2 [17], 3 [18] and 4 [19], MRI scans were blindly read.

Two different clinical outcomes were considered in these studies: disability progression and the relapse frequency or relapse rate.

In studies 1 [16] and 2 [17] disability progression is not provided. In study 3 [18] disability progression is defined as a sustained EDSS increase of ≥1 point from baseline EDSS if the baseline EDSS was between 2.0 and 5.5 points (inclusive) or an EDSS increase of ≥0.5 if baseline EDSS was >5.5 points. This change could not be attributable to other etiology such as fever, concurrent illness, injury, adverse reactions to concurrent medications or recent relapse, and had to be sustained for ≥12 weeks. Study 4 [19] compared baseline EDSS to week 32 and confirmation at week 52. Sustained change in EDSS was defined as in study 3 [18].

Relapses were defined in study 1 [16] as the occurrence of new or worsening neurological symptoms consistent with MS manifestations, evolving over less than 3 months and accompanied by objective neurological worsening, consistent with an increase of at least half a step on the EDSS, two points on one of the appropriate functional system scales (FSSs) or one point on two or more of the appropriate FSSs. The change had to be verified by the investigator and must have affected the selected FSSs. Symptoms had to persist for more than 24 hours and could not be attributable to confounding clinical factors (fever, infection, injury, adverse reactions). Other symptoms that were not accompanied by changes on clinical examination were not considered relapse.

In study 2 [17] relapse was defined as new or recurrent neurologic symptoms consistent with MS that lasted for at least 48 hours, preceded by a relatively stable or improving neurologic status for at least 30 days.

In study 3 [18] relapses were not considered as outcome as all patients had a primary progressive form of the disease (patients do not have relapses by definition). Study 4 [19] did not provide a definition of relapse despite of including RRMS patients and defining breakthrough disease according to clinical relapses and radiological activity. Although this study was not designed to analyze effects on the relapse rate, authors describe the relapse rate at baseline and at the end of the study.

Study 1 [16], study 2 [17] and study 4 [19] included the number of Gad-enhancing lesions on T1-weighted brain MRI sequences during the study. Study 1 [16] also included the total number of new T2-bright lesions and cumulative volume of T2-bright brain lesions. Study 2 [17] included the total number of new Gad-enhancing lesions on the T1-weigthed sequences. In addition, studies 2 [17], 3 [18] and 4 [19] included the change from the baseline lesion volume on T2-weighted MRI scans. In study 3 [18] change in brain volume was measured by brain parenchymal fraction. Study 4 [19] also included black hole number and volume, and number of T2-bright lesions on brain MRI.

All the studies reported adverse events and classified most of them as infusion- or infection-related. Authors described in the studies the cause of death if any occurred.

In terms of safety as an outcome, Table 2 summarizes the main adverse events.

In study 1 [16] there were no reported grade 4 AEs. The majority of the 26 patients (77%) experienced mild to moderate (grade 1–2) AEs while six patients reported grade 3 (severe) AEs (including fatigue, tooth fracture, muscle weakness and headache). One of 26 patients discontinued study drug before the end of the study, not receiving the 26 week infusion, due to an infusion-related AE. Infection-associated events were reported by 61.5% of the patients and all were mild-to-moderate (grade 1–2) in severity.

In study 2 [17], 62.3% of the patients in the rituximab group presented grade1–2 AEs (74.3% in the placebo), 30.4% grade 3 AEs (25.7%) and 4.3% grade 4 (none in the placebo group). There were 3 patients who reported grade 4 events in the rituximab group (ischemic coronary-artery syndrome, malignant thyroid neoplasm and symptoms of acute and progressive MS). A total of 5.7% of patients in the placebo group and 4.3% in the rituximab group withdrew from the study because of AEs. There was a death in the rituximab group due to homicide.

In study 3 [18], at week 74 infusion-related reactions were less frequent in the rituximab (4.9%) versus placebo (7.2%) recipients. No grade 4 infusion-related AEs were reported. Infection-related serious AEs (SAEs) were <1% in placebo group and 4.5% in the rituximab treatment. Three patients died during the study. One patient (rituximab group) had a history of brainstem lesions and aspiration and withdrew early from the study. The other two were in the placebo group. One of them died from cardiopulmonary failure during the study and the other contracted pneumonia and died after withdrawing from the study.

In study 4 [19], the infusion-related events led to discontinuation in 2 patients. Eleven patients out of 30 who completed the study had minor reactions that did not preclude additional infusions.

As the main goal of this work was to evaluate the use of rituximab in relapsing and progressive forms of MS, we chose disability as primary outcome, as relapses do not occur in primary progressive forms.

Disability progression as a primary outcome was available for patients in study 3 [18] and study 4 [19]. Although in study 3 rituximab tended to delay the time to CDP at week 96 compared to placebo, with the proportion of patients with CDP of 38.5% and 30.2% respectively (HR = 0.77), the difference was not statistically significant (p = 0.14). In study 4 [19] EDSS improved after rituximab therapy in 7 out of 30 patients, remained stable in 21 and worsened in 2.

The effect of rituximab on the relapse rate was one of the secondary outcomes in studies 1 [16] and 2 [17]. In study 1, 80.8% (21/26) of the patients (all of them received rituximab) remained relapse free at week 72. The unadjusted annualized relapse rate was 0.25 from baseline to week 24 and 0.18 from baseline to week 72. In study 2, the proportion of patients with relapses was reduced in the rituximab group (n = 69) at week 24 compared to the placebo group (n = 35) (14.5% vs. 34.3% respectively, p = 0.02) and at week 48 (20.3% vs. 40%, p = 0.04). The unadjusted and adjusted annualized relapse rates were similar. The adjusted annualized relapse rate at week 24 was 0.37 for the rituximab group and 0.84 for the placebo group (p = 0.04) while at week 48 relapse rates were 0.37 and 0.72 (p = 0.08) respectively. Although study 4 [19] was not designed to examine relapse rate reduction authors found a relapse rate before treatment of 1.27 compared to 0.23 during treatment.

In terms of the radiological outcomes, three out of four studies included the number of Gad-enhancing lesions. For study 1 [16], the mean number of Gad-enhancing lesions was 1.31 at trial entry and decreased to 0.73 at week 4, 0.05 at week 48 and 0 at week 72. In study 2 [17], the mean number of Gad-enhancing lesions at weeks 12,16,20 and 24 was 5.5 in the placebo group and 0.5 in the rituximab group, with a reduction at each study time point beginning at week 12 (with p-values from 0.003 to <0.001). Considering the brain MRI scans obtained at weeks 12, 16, 20 and 24, 80.3% of the rituximab-treated subjects had no Gad-enhancing lesions compared to 51.4% in the placebo group. In study 3 [18], new Gad-enhancing lesions were not considered as contrast was not administered after baseline. In study 4 [19] 74% of the three post-treatment MRI scans were free of Gad-enhancing lesions. Mean number of Gad-enhancing lesions per month prior to treatment was 2.81 while after treatment it was 0.33.

The selected studies were heterogeneous and their primary and secondary outcomes as well as inclusion criteria were different. Not all studies were placebo-controlled and the characteristics of relapsing-remitting and primary-progressive patients are difficult to interpret together as the epidemiology and disease course vary for both. Considering all these facts, a meta-analysis was not considered appropriate for the scope of this systematic review.