Introduction

Multiple sclerosis (MS) is a chronic disease with a profound alteration of the immune system, classically treated with strategies directed against cellular immunity. Although antibodies against the central nervous system (CNS) or myelin-related glycoproteins might play a role in the pathogenesis of the disease, as demonstrated by the intrathecal synthesis of immunoglobulins restricted to the CNS [1,2,3,4,5], only in the last decades, the focus of MS treatment has been directed to the humoral immune system.

With the premise that humoral immunity could be involved in MS, anti-CD20 monoclonal antibodies (MoAb) capable of destroying B cells (via complement and antibody-dependent cellular cytotoxicity) and, consequently, acting through modulation of T-cell activation mediated by B cells, are being increasingly considered for the treatment of MS. Rituximab (RTX) [6] was tested by several groups in active relapsing-remitting MS (RRMS) [7, 8] and progressive MS (PMS: primary-progressive MS-PPMS and secondary-progressive MS-SPMS) [8]. This drug has shown a high efficacy over relapses in RRMS [9,10,11], and a reduction of disability progression in a specific subgroup of PPMS patients, younger patients (less than 51 years old) with at least one gadolinium enhancing lesion (GEL) [12]. Based on these benefits, RTX has been largely administrated out of label in RRMS patients who experienced disease activity on the standard therapies, and also in PMS [13, 14].

The aim of our work was to describe the immediate and long-term efficacy and safety concerns of the treatment with RTX in MS, both in patients with RRMS and PMS. This report gives class IIb evidence about effectiveness and safety of RTX in MS.

Patients and methods

The patients with MS selected in this study comprised both RRMS and PMS who fulfilled the McDonald 2010 diagnosis criteria, and received treatment with RTX in two University Hospital MS Units from the Hospital Universitari i Politècnic La Fe (HUPLF) and the Hospital Clínic Universitari (HCU), two tertiary hospitals located in Valencia, Spain. Only patients with a minimal follow-up period of 6 months were included. In all cases, approval from both the Spanish Medicines Agency of the Spanish Ministry of Health (Agencia Española de Medicamentos y Productos Sanitarios–AEMPS) and the local Ethical Committee from each hospital was obtained, as per the Spanish Royal Decree for compassionate use of medicines (RD 1015/2009, June 19th). All patients signed a specific informed consent explaining potential risks and benefits and, if it was the case, potential treatment alternatives.

Study design and procedures

An observational, retrospective study of the clinical and radiological characteristics of RRMS and PMS patients treated with RTX was carried out, analysing data included in the GITEM Registry of patients (Grup d’Investigació i Tractament de l’Esclerosi Múltiple–Group for the investigation and treatment of Multiple Sclerosis), approved by the Health Department of the Valencian Autonomous Region (Health Regional Authority–Conselleria de Sanitat-), and the data collected by both MS Units during clinical visits in the follow-up period.

The decision to treat with RTX was agreed among all the neurologists of each MS unit. The criteria to initiate RTX treatment were: (1) in RRMS, patients with suboptimal response (see definition of treatment failure below) to disease-modifying treatments (DMTs) or aggressive disease (≥ 2 relapses in less than 1 year and score > 2.0 in the Expanded Disability Status Scale–EDSS), in whom currently approved the second-line DMTs were not considered as a safe option due to progressive multifocal leukoencephalopathy (PML) risk or to other medical condition; (2) in PMS, if one or more of these criteria were fulfilled: (a) increase of disability since the last year measured by the EDSS with the presence of IgM oligoclonal bands (IgM-OCB, considered a biomarker for inflammation and for a worse prognosis of the disease) in the cerebrospinal fluid (CSF) [15,16,17,18]; (b) presence of GEL; (c) relapse.

Other treatments for MS like alemtuzumab or ocrelizumab were not available in the period the treatment was offered.

For safety concerns related to others treatment previously used for PMS like mitoxantrone [19], nowadays, RTX is the only one option considered to treat progressive MS in our units.

All patients were assessed in routine clinical practice. At the time that the decision to start RTX was made (considered as baseline), demography data (age at the first attack of MS and gender) and retrospective clinical data (presence of IgG or IgM-OCB, previous DMT use, EDSS score, and annualized relapse rate-ARR), were collected. Both EDSS score and ARR were obtained from the year before RTX and at baseline. The previous radiological activity was collected in most patients. At baseline, a complete blood analysis including platelets, red and white cell blood recount, hepatic and renal function, viral serologic assessment (hepatitis B and C virus, and human immunodeficiency virus), and screening for latent tuberculosis infection, were performed in all patients.

At first, a dose of 1000 mg intravenous RTX was administered twice, on day 1 and day 15. Intravenous premedication to prevent allergic reaction to the infusion consisted of 1 g of paracetamol, 100 mg of prednisolone, and 5 mg of dexchlorpheniramine. RTX was administered at the outpatient facilities by trained nurses. The attending neurologist collected all infusion-related adverse effects. After RTX administration, scheduled clinical visits were planned every 3 months. During these visits, the neurologist registered the presence of new symptoms, EDSS score, and potential side effects, as well as results from a complete blood test, including lymphocyte subpopulations counts. An MRI scan was performed 6–12 months from the first RTX administration.

For maintenance, an isolated dose of 1000 mg was administrated when the percentage of total CD19 + cells was 2% or more. This criterion was decided taking into account that RTX is an off-label treatment for MS and because of the absence of established recommendations for re-treatment, so the safety of patients has been the priority in our clinical practice.

Definitions

A relapse was defined by the presence of new or worsening of the previous neurological symptoms, maintained for more than 24 h, in the absence of fever or significant infectious processes and accompanied by objective changes in the neurological examination.

Confirmed improvement (CID) and confirmed worsening of disability (CWD) were defined in PMS by a decrease or increase, respectively, in one point in EDSS (if EDSS was less than 6) or in 0.5 point (if EDSS was 6 or more) persisting after 6 months. In RRMS, confirmed, worsening, or improvement, was evaluated similarly.

Clinical activity was defined as the presence of relapses and/or sustained progression of disability and radiological activity was defined as the presence of new T2 and/or GEL in MRI.

Treatment failure was defined as the presence of one of the following: (1) two relapses in 1 year, (2) one mild relapse and a new GEL in an MRI acquired minimum 3 months after relapse, and (3) one disabling relapse with residual EDSS of at least two points.

No evidence of disease activity (NEDA) was considered if the absence of clinical and radiological activity was demonstrated.

Statistic analysis

SPSS (Statistical Package for the Social Sciences, Chicago, IL, USA) 21.0.v and GraphPad Prism v5.01 were used. Kaplan–Meier survival analysis for the median time to a relapse and to increase of disability was run. Univariate and multivariate Cox regression analyses to explore potential predictive variables for CWD status after RTX were performed. As covariates, gender, age, disease duration, EDSS at baseline, and presence of confirmed worsening of disability, ARR and radiological activity from the previous year to RTX initiation were considered.

Results

Baseline characteristics

At May 2017, from a total of 95 patients, five were excluded, because they did not have 6 months of follow-up since the initiation of RTX. Finally, 90 patients were considered, 65 from the HUPLF, and 25 from the HCU. Fifty six were women (62.2%) and the median age at first attack of MS was 29.6-year-old (standard deviation-SD 9.8). Regarding the clinical form of MS, 31 patients had RRMS (34.4%) and 59 had PMS (63.6%): 45 SPMS (50.0%) and 14 PPMS (15.6%). Most patients had IgG-OCB in the CSF (74/84, 88.1%) and more than a half had IgM-OCB (46/75, 61.3%). Forty-nine patients (55.4%) had one or more GEL in the baseline MRI.

From all patients, 76 (84.4%) received at least one DMT before RTX; most of RRMS patients received the second-line drugs and most of PMS received mitoxantrone or cyclophosphamide (Table 1).

Table 1 Baseline clinical and demographic characteristics of the total of the series and of RRMS, SPMS, and PPMS separately
Full size table

Twenty-six RRMS patients (83.9%) had received the first- and second-line treatment (fingolimod and natalizumab) with suboptimal response. Most of them had received natalizumab and were switched to fingolimod due to PML risk. If fingolimod was not enough to control inflammatory activity after natalizumab, rituximab was initiated. In three RRMS patients, rituximab was initiated after the first-line treatment failure; in two cases, RTX was the first MS treatment ever received. One of these latter patients had an aggressive disease course and also had concomitantly an ankylosing spondylitis, so rituximab was chosen as the first option to treat both conditions. The rest of these patients (four patients in total) also had an aggressive disease, with disabling relapses that needed plasma exchange (PLEX) treatment because of lack of effect of methylprednisolone. After having a good response to PLEX (treatment against humoral immunity), maintenance treatment with rituximab was indicated. All these patients had positive serology to JC viruses, so natalizumab was not considered as the first option.

The final results after analysing the data excluding these latest five patients have not significantly changed (data not shown).

The mean disease duration from the first relapse to RTX was 12.4 years (SD 9.5). The mean EDSS 1 year before RTX was 4.2 (SD: 1.8), and at baseline, it had increased to 4.9 (SD: 1.7). The ARR 1 year before starting RTX was 0.86 (SD 1.0) (Table 1). In summary, despite standard treatment for MS, all patients included had an active disease during the year previous of starting rituximab, in the form of relapses, progression of disability, radiological activity, or a combination of some of them.

When analysing patients grouped by the MS course, in the RRMS group, 24 patients (85.7%) had experienced at least one relapse, 13 patients (43.3%) had a confirmed worsening of one point in the EDSS, and 22 patients (71%) had shown radiological activity, with a mean GEL number of 3.4 (SD 5.7). Likewise, in the PMS group, 22 (37.2%) had relapses, 48 patients (81.3%) achieved CWD status, and 26 patients (46.4%) had radiological activity, with a mean GEL number of 2.1 (SD 4.8).

Rituximab effectiveness

After RTX, all patients were followed up at least for 6 months; 79 (87.7%) were followed for more than 1 year and 54 (60%) for more than 2 years. The median follow-up time was 31.3 months (SD: 20.0) and the median number of RTX cycles received was 3 (range 1–9). The median time between treatment courses was 10.8 months (range 6.0–29.5) with a median percentage of CD19 + lymphocyte count of 0.3% (range 0–3).

Considering the whole cohort of patients, ARR was reduced an 88.4% (p < 0.001) and there was also a reduction in the number of GEL from 2.56 to 0.06 (p < 0.001) (Fig. 1). Once RTX was initiated, nine patients suffered a relapse (eight RRMS and one PMS patients); in five of them occurred within the first semester after treatment (Fig. 2). Ninety percent of patients remained free of relapses.

Fig. 1
figure 1

Evolution of annualized relapsing rate, before and after rituximab. Data corresponding to the total of the series and in RRMS and PMS separately are shown. A significant reduction of ARR was observed for both groups. RRMS relapsing-remitting multiple sclerosis, PMS progressive multiple sclerosis

Full size image
Fig. 2
figure 2

Graphical representation of relapses before and after rituximab, for both groups: RRMS and PMS. Most of relapses occurred in the first 6 months after starting rituximab, being infrequent after the first year under treatment. RRMS relapsing-remitting multiple sclerosis, PMS progressive multiple sclerosis

Full size image

Respect to the mean EDSS score, an increase from 4.2 to 4.9 was observed in the year before starting rituximab (p < 0.001). After 1 year of treatment, a decrease to 4.6 was observed (p = 0.01), and EDSS remained stable in the second year of therapy with RTX. EDSS variations in each RRMS and PMS are reflected in Fig. 3.

Fig. 3
figure 3

Evolution of disability measured by EDSS, before and after the first and second year under rituximab. A significant increased in EDSS had been observed the year before starting rituximab. For both groups, RRMS and PMS, a significant decrease in EDSS was observed after the first year of treatment, which was maintained during the second year. RRMS relapsing-remitting multiple sclerosis, PMS progressive multiple sclerosis

Full size image

Twenty-two patients (24.4%) experienced a CWD: 1 RRMS patient and 21 PMS patients (15 SPMS − 33.3% and 6 PPMS − 42.9%) (Fig. 4a).

Fig. 4
figure 4

a Kaplan–Meier survival analysis of risk of experience confirmed worsening of disability after rituximab. Seventy-five percent of the total cohort was free of confirmed worsening of disability under rituximab, 96.8% of RRMS, 66.7% of SPMS, and 57.1% of PPMS. b Kaplan–Meier survival analysis of risk of experience confirmed improvement of disability after rituximab. Twenty-three percent of the total cohort experienced a confirmed improvement of disability under rituximab, 31% of RRMS, 26.6% of SPMS, and 16.6% of PPMS. CWD confirmed worsening disability, CID confirmed improvement disability, RRMS relapsing-remitting multiple sclerosis, SPMS secondary-progressive multiple sclerosis, PPMS primary-progressive multiple sclerosis

Full size image

On the other hand, 21 patients (23.3%) experienced a CID: 8 RRMS patients (31.0%) and 13 PMS patients (11 SPMS − 26.6% and 2 PPMS − 16.6%) (Fig. 4b).

There was no evidence of disease activity (measured with NEDA) in 70% of the total sample, 74.2% of RRMS patients, and 67% of the PMS patients.

In the multivariate Cox regression, risk for CWD status for all patients after RTX was increased in patients with higher EDSS at baseline (Hazard ratio 1.4 p = 0.01). In patients with PMS, a younger age at baseline was the only variable related to reaching CWD status (Hazard ratio 1.4, p = 0.02).

Tolerability and safety

The main side effect of RTX was the appearance of infusion-related symptoms at first infusion (17 patients, 18.8%) despite of premedication with steroids and antihistamine drugs. Most of these symptoms were solved by decreasing the infusion rate. In three patients, the infusion of RTX needed to be temporarily halted, and in two of these cases, it was well tolerated in a second attempt. Most infusion-related symptoms were classified as mild and lessened with subsequent infusions. Two patients suffered moderated infusion-related symptoms, namely a generalized rash and a symptomatic bradycardia; RTX was definitely discontinued in both cases. Another patient suffered symptoms compatible with serum sickness that were managed successfully with symptomatic treatment, but RTX was no longer administered.

During the follow-up, 14 (16.9%) patients had common infections with good recovery with standard treatment. None of the patients developed severe or opportunistic infections.

One patient presented agranulocytosis 3 months after RTX infusion. When infection and other toxics were ruled out as responsible for the decrease in neutrophils, this side effect was attributed to RTX and the treatment was permanently stopped.

Three patients experienced venous thrombotic events. One developed an isolated deep venous thrombosis (DVT) in one leg, other a DVT with secondary mild pulmonary embolism; both patients continued with the treatment. The second patient was taking concomitant oral contraceptives that were withdrawn. The third case suffered a serious massive pulmonary embolism secondary to a DVT, and died of a sudden death. This latter case had an EDSS score of 8.5, so lack of mobility could have contributed to venous stasis.

The presence of newly formed neoplasms was not observed.

Withdrawal of RTX treatment

Three patients were lost to follow-up 34.6 months after starting RTX (range 21–36) and these patients were no considered to further analysis.

RTX was withdrawn in 22 patients (24.4%). Reaching a CWD status was the reason for withdrawal in 11 cases (50%); all of them PMS. Management of these latter patients after RTX withdrawal included: three received mitoxantrone, one (who also had relapses) an autologous stem cell transplantation (ASCT), one intrathecal rituximab, two patients were treated with monthly pulsed steroids, and four patients did not receive any treatment, because they were no longer ambulatory (EDSS > 6.5) and risk–benefit was not considered appropriate as they had not responded to any previous treatment. In two cases (4.5%), both RRMS, RTX was withdrawn because of new relapses and an ASCT was performed in both patients. In five cases, RTX was discontinued due to side effects: one patient died; in two patients, it was switched to mitoxantrone (both PMS); one RRMS received glatiramer acetate; the other patient rejected any treatment. In two patients, RTX was stopped for desire to become pregnant and both returned to the same treatment after delivery. In two patients, RTX was withdrawn by their own decision: one of them was switched to fingolimod and the other did not receive any treatment.

Discussion

In this cohort of MS patients with an aggressive form of the disease, followed for a mean time of 30 months, RTX was safe and useful for controlling the inflammatory activity and the progression of short-term disability in patients with RRMS. RTX helped to achieve NEDA status in both RRMS and PMS patients. In PMS patients, although this report cannot give evidence about the effect over the progression of MS disability due to the observational design and the lack of a control group, RTX appears to have an anti-inflammatory effect comparable as in RRMS.

Patients with SPMS represented half of our cohort. The presence of IgM-OCB in 55% of this subgroup was considered indicative of an aggressive disease course, though the pathophysiological mechanism underpinning this role remains uncertain. The majority of patients of this subgroup had undergone the previous DMT including—in 40% of case—mitoxantrone and cyclophosphamide adding further evidence, albeit circumstantial, of the aggressiveness of their disease course.

There are four clinical trials that evaluate the efficacy and safety of rituximab: three of them for RRMS [9,10,11] and one for PPMS [12]. These studies are very heterogeneous in their design, outcomes, and protocol of infusion, but, in RRMS, all of them conclude that rituximab is effective in reducing clinical and radiological activity, and our series is consistent with these results.

Only nine patients suffered relapses after RTX, and in four of them (44.4%) occurred during the first 3 months. Although the presence of early relapses in these patients has been considered for defining treatment failure in our study, it is possible that the effect of RTX is not fully achieved after 3 months after starting the therapy, so this inflammatory activity could not reflect a real lack of efficacy of RTX.

Regarding PPMS, in the OLYMPUS trial [12], RTX marginally reduced the time to CWD status, but the difference did not reach statistical significance except for a pre-planned subgroup of young patients (less than 51 years old) with the presence of GEL in the baseline MRI. Most of our PMS patients remained stable after RTX, without significant changes in the EDSS score. In our series, only the higher EDSS before RTX predict a poor outcome, increasing the risk to achieve CWD status, independent of the previous inflammatory activity.

In a real-world study of 822 off-label rituximab-treated patients (557 RRMS and 198 PMS) with a mean follow-up of 21.8 months, decreases of the ARR to 0.044 for RRMS, 0.038 for SPMS, and 0.015 for PPMS were observed, and in only 4.6% of patients, some radiological activity was observed as in our work [13]. In this study, the EDSS score remained unchanged in patients with RRMS but increased 0.5 and 1.0 for patients with SPMS and PPMS, respectively. However, in our series, a sustained improvement in EDSS was observed for MS patients with the previous relapses (in both the RRMS and the SPMS), while it remained stable for most PPMS patients. In this study, the most common side effects were infections, but, only in 1.7% of patients, infections were severe, as in our sample.

In respect of safety data, available data from the use of rituximab in others autoimmune diseases like rheumatoid arthritis indicated a high tolerability and safety, with low risk of opportunistic infections or secondary malignancies [20]. PML has been reported under rituximab treatment, but in the setting of haematological diseases or with the concomitant use of other immunosuppressive drugs [21, 22]. In our experience, RTX has been generally well tolerated and has shown to be a safe option. The most prevalent side effects were those symptoms related to infusion, but were less frequent than those reported in clinical trials (60–70 versus 18.8%) and mild in most cases. These data show that the potential immunogenicity of RTX is a minor clinical problem. Serious opportunistic infections like PML were not reported. Nevertheless, three patients presented a thrombotic event, one of them with a fatal course. In two of these cases, the presence of other risk factors that could contribute to this side effect was involved: contraceptive treatment and immobility in a patient with advanced disease. In our knowledge, this observation had not previously been described in the literature and there are few reported cases of thrombosis under rituximab treatment [23]. Although we cannot conclude that there is a cause–effect relation between the use of RTX and thrombosis, this observation deserves further considerations when assessing risk–benefit before starting RTX (and probably other anti-CD20 MoAb) in MS patients. In the recently published ORATORIO trial in PPMS patients, one case of fatal pulmonary embolism was described in the ocrelizumab group [24].

The small sample and the absence of a control group are the main limitations of this study. The absence of a control group does not allow for the confirmation of whether clinical and radiological activity reduction observed in this study is actually due to the effect of RTX or if it is due to the natural evolution of the disease or the effect of regression to the mean [25]. However, the observed magnitude of the reduction of clinical and radiological activity, seemly even higher than the observed in clinical trials, deserves not to be belittled.

These data are important in our clinical practice, because RTX could be an effective and safe treatment in aggressive forms of RRMS, in which other standard treatments are not able to control the inflammatory activity. In addition, RTX may be a therapeutic option in some PMS patients, for whom there is currently no any approved effective treatment. In our study, we have not only observed a significant reduction in relapses and enhancing lesions in this subgroup, but even a reduction in the proportion of patients with confirmed disability progression.

More controlled, preferably multicentre, studies designed to assess the efficacy of RTX in aggressive forms of RRMS and PMS are needed, as well as studies to evaluate the predictors of the response to RTX in PMS forms.