Adverse drug reactions of Rituximab in patients suffering from autoimmune neurological diseases

Abstract

Background

Rituximab, a chimeric human/mouse monoclonal antibody targeting CD-20 antigens, has been used recently for various rheumatological and autoimmune diseases, including autoimmune neurological disorders.

Objectives

We aimed to study the frequency, seriousness, causality, and preventability of adverse drug reactions (ADRs) of rituximab in Iranian patients with autoimmune neurological diseases.

Methods

In this cross-sectional observational study, patients with autoimmune neurological diseases who had an indication for rituximab treatment were enrolled. Naranjo adverse drug reaction probability scale was used to assess the causality of ADRs, and the preventability of the ADRs was determined by P-Method. The seriousness of ADRs was also determined.

Results

A total of 264 ADRs were recorded from 97 patients. The Median (min-max) number of ADRs experienced by patients was 3 (1–7) events. 11.3% of patients experienced serious ADRs. 18.2% and 26.9% of ADRs were Definite and Probable, respectively. Only 5% of the ADRs were ‘’preventable”. The most frequent ADRs were rituximab infusion-related reactions.

Conclusion

Rituximab had an acceptable safety profile in our study patients. However, there must be certain cautions regarding the use of the medication for the elderly or patients with a compromised immune system. Timely detection and management of ADRs would also be crucial to prevent severe and permanent damages. Moreover, considering that rituximab is used as an off-label treatment for autoimmune neurological diseases, a risk-benefit assessment would be necessary before deciding on the treatment choice.

Graphical abstract

Introduction

Different immunomodulating and immunosuppressive agents have been used in the treatment of autoimmune neurological diseases [1]. Monoclonal antibodies have been used in the treatment of different neurological disorders such as Multiple Sclerosis (MS), Myasthenia Gravis (MG), and Neuromyelitis Optica (NMO) [2].

Rituximab, a chimeric human/mouse monoclonal antibody targeting CD-20 antigens, has been used recently for various rheumatological and autoimmune diseases as an appealing off-label treatment [3]. In the case of MS, rituximab showed acceptable clinical efficacy outcomes and revealed better outcomes than commonly used disease-modifying agents, specifically in patients with the relapsing-remitting disease [4]. In addition to efficacy, the safety of rituximab in the treatment of autoimmune neurological diseases is a major concern.

The safety profile of rituximab has been described well in rheumatological diseases. In a 6-year-long study regarding the efficacy and safety of rituximab in 2484 rheumatoid arthritis patients, 21.4% of the patients reported adverse drug reactions (ADRs). The most common ADRs included pharyngitis (8.1%), paresthesia (9.1%), and dermatological problems (5.5%). There were also 3 cases of general septicemia [5]. In an 18-month cohort study of rituximab in ' ‘Wegener’s granulomatosis”, out of 188 patients, 21 have reported ADRs, including serious infections (13.6%) and a decrease in immunoglobulin levels (especially IgG). During the study, 14 patients discontinued their medication due to pregnancy, serious infections, hypogammaglobulinemia, serious reactions during drug injection, late neutropenia, and lack of compliance [6].

Observed ADRs of rituximab in treating neurological diseases have been reported in different studies, including clinical trials, retrospective studies, and cohort studies [7–11]. However, the causality and preventability of the ADRs were not reported.

This study aimed to describe the frequency and seriousness of rituximab ADRs in Iranian patients with autoimmune neurological diseases. Additionally, the causality and preventability of detected ADRs have also been studied.

Methods

This cross-sectional observational study was conducted from July 2018 to January 2019. The ethical evaluation of this study was performed by the Tehran University of Medical Sciences’ (TUMS) ethics committee. Patients who were referred to the day-care clinic of Shariati hospital to receive rituximab infusion for their autoimmune neurological disorders were enrolled in the study. Patients with an active infection and serious hepatic impairment were excluded. Written consent was taken upon patient enrollment.

One hundred and two patients with autoimmune neurological diseases who had an indication for rituximab treatment, received 1000 mg in one or two doses, two weeks apart as the starting dose and 500–1000 mg every six months as a maintenance dose. The premedication for rituximab infusion included oral diphenhydramine, 37.5 mg, oral acetaminophen, 1 g, and intravenous hydrocortisone, 100 mg.

A trained pharmacist visited the infusion clinic patients and detected patients who experienced any ADRs of rituximab. Data regarding ADRs were gathered through history taking and using ‘patients’ medical records. The patients were carefully observed for two hours after the rituximab infusion, and in case of a serious ADR, they received proper management and were observed for a longer period. Naranjo adverse drug reaction probability scale was used to assess the causal relationship between the ADRs and rituximab [12]. To evaluate the preventability of the ADRs, the P-Method was used [13]. The seriousness of ADRs was also determined based on a previously published guide[14].

All the patients’ data and assigned scores were then rechecked by a clinical pharmacist.

Statistical analysis

Data were analyzed using SPSS software (version 24 for Windows). Descriptive statistics were computed for all study variables. Numerical variables were described by mean (SD) or median (min.-max.), and categorical variables were described by frequency (Frequency, %). The nonparametric Mann-Whitney test was used to compare the ADR occurrence rate between males and females. The targeted organ’s ADRs were compared between males and females using Chi-Square and Fisher’s exact tests. Spearman correlation coefficient was used to assess the association between the occurrence of ADRs and quantitative variables such as age and Body Mass Index (BMI). P-value < 0.05 was considered as a significant level.

Results

Five of the 102 enrolled patients didn’t report any ADRs. A total of 264 ADRs were recorded from 97 patients. The demographic and clinical characteristics of the patients are summarized in Table 1. Fifty-five patients aged 25 to 50 years old, and 72% were females. More than half of the patients had multiple sclerosis (Table 1).

Table 1.

‘Patients’ clinical and demographic characteristics

Sex, Female (No. (%))	70 (72.2)
Age, Years (Mean (SD))	41.43 (13.12)
BMI, Kg/m 2 (No. (%))
< 18	3 (3.1)
18-<25	45 (38.5)
> 25	45 (57.7)
BMI, Kg/m 2 (Mean (SD))	25.76 (4.25)
Indication of Rituximab, (No. (%))
MS	52(53.6)
MG	27(27.8)
NMOSD	10(10.3)
Immune-mediated Myopathy	6(6.2)
CIDP	2(2.1)
Number of infusion sessions, (Median (Min.-Max.))	4 (1–9)

BMI: Body Mass Index; CIDP: Chronic Inflammatory Demyelinating Polyradiculoneuropathy; MG: Myasthenia Gravis; MS: Multiple Sclerosis; NMOSD: Neuromyelitis Optica Spectrum Disorder;

The Median (min-max) number of ADRs experienced by patients was 3 (1–7) events. Twenty-five patients (25.7%) experienced only one ADR. Out of 97 patients, 11 (11.3%) patients experienced serious ADRs, including 2 cases of death following sepsis, 6 cases of hospitalization, one case of lung tumor, and 2 cases of persistent tinnitus. One of the patients who died following infection and sepsis was a 74-year-old man. About three weeks after the second rituximab dose, the patient felt fatigued and had general sickness, and a week later was admitted to the hospital with tachycardia, fever, and malaise. He was then diagnosed with pneumo-sepsis. After ten days of using broad-spectrum antibiotics, there was a decrease in the level of consciousness. The bronchoalveolar lavage (BAL) confirmed pneumocystis pneumonia (PCP). The patient eventually died because of multi-organ failure and disseminated intravascular coagulation (DIC). The other patient was a 60-year-old male who also died following sepsis and DIC 11 days after receiving rituximab.

The frequency of ADRs based on the involved organ/system and a brief description of observed ADRs are reported in Table 2. The most frequent ADRs among study patients were rituximab infusion-related reactions. Among patients who experienced only one ADR, 60% reported reactions following the infusion of the drug.

Table 2.

ADRs categorized by the targeted organ or system

Type of ADR	No. of ADRs (%), Total N = 264	Percent of ADRs in Naranjo categories				Description of reported ADRs	No. of patients reported the ADR (%)
		Definite	Probable	Possible	Unlikely		All patients (No.=97)	Males, (No.=27)	Females, (No.=70)	P-value†
Infusion related	66(25.0)	69.7	24.2	6.1	0	Mostly observed in the first or second doses of the drug, and its rate was reduced after the third injection. The most frequent ADRs in this category included flushing and dyspnea. The most important infusion-related ADRs were 2 cases of angioedema in the face and lips followed by fever and chills for two days after infusion, one severe fatigue leading to hospitalization, and one hypertensive crisis resulted in discontinuation of the infusion.	66(68)	17(63.0)	49(70)	0.505‡
Cardiovascular	23(8.7)	4.3	34.8	60.9	0	Included peripheral edema, flushing and hypertension.	23(23.5)	5(18.5	18(25.7)	0.455‡
Neurological	53(20)	0	22.6	73.6	3.8	Included sleep disorders, anxiety, phonophobia leading to aggressiveness, dizziness and attention deficit.	53(54.6)	14(51.9)	39(55.7)	0.732‡
Dermatological-Mucosal	26(9.8)	0	46.2	50.0	3.8	The most frequent ADRs in this category were hyperhidrosis and dry skin. The most important ADRs in this category included hair loss. In one patient, the itching sensation during the infusion lasted for several months after receiving the drug and a case of hypersensitivity to the drug was also observed.	26(26.8)	5(18.5)	21(30)	0.235‡
Endocrine-Metabolic	16(6)	0	18.8	81.3	0	There have been 2 cases of a significant increase in appetite and 5 cases of hyperglycemia.	16(16.5)	1(3.7)	15(21.4)	0.036§
Gastrointestinal	14(5.3)	0	14.3	85.7	0	The most important ADR of this type was nausea and vomiting.	14(14.4)	2(7.4)	12(17.1)	0.337§
Hematologic	1(0.4)	0	0	100	0	One case of anemia.	1(1)	0(0)	1(1.4)	1.000§
Respiratory	15(5.7)	6.7	53.3	40.0	0	Included sore throat, coughs and two cases of dyspnea after a workout.	15(15.4)	7(25.9)	8(11.4)	0.114§
Musculoskeletal	11(4.1)	0	27.3	72.7	0	Included back pain and pain and /or swelling in joints.	11.(11.3)	5(18.5)	6(8.6)	0.174§
Renal /Hepatic	2(0.7)	0	0	100	0	Two cases of a rise in alanine transaminase (ALT)	2(2)	0(0)	2(2.9)	1.000§
Infection	27(10.2)	0	18.5	81.5	0	Twenty-seven patients experienced a bacterial, viral or fungal infection. In four patients infection led to hospitalization, two patients developed secondary sepsis and died (in one of them, pneumocystis pneumonia was confirmed) and in one patient permanent visual damage due to ocular candidiasis occurred. There were also 12 upper respiratory tract infections, five urinary tract infections, two fungal dermatitis and one otitis.	27(27.8)	3(11.1)	24(34.3)	0.022‡
Ocular	10(3.8)	0	20.0	80.0	0	Four patients reported blurred vision following rituximab infusion, which was resolved in 1–3 days in all of the patients except for one who turned out to have a blood clot in the left eye. There was four cases of myopia at various degrees where the patients had difficulty in seeing far objects.One patient reported eye strains following reading a book or driving. There was also a case of cataract.	10(10.3)	2(7.4)	8(11.4)	0.721§

† For comparisons made on the frequency of ADRs between males and females

‡ Pearson Chi-Square

§ Fisher’s Exact Test

ADR occurrence rate was compared among males and females. Median (min-max) ADR events were 3 (1–7) and 2 (1–5) in female and male patients, respectively. However, the difference was not statistically significant (P = 0.064). The frequency of different types of ADRs based on the involved organ/system was also compared among males and females (Table 2). Females were more likely to develop endocrine/metabolic ADRs (21.4% in females versus 3.7% in males, P = 0.036). Infections were more frequent in men than women (34.3% versus 11.1%, P = 0.022).

There was no correlation between the number of ADRs experienced by the patients and patients’ age (P = 0.574) and BMI (P = 0.461).

Causality assessment of ADRs based on the Naranjo algorithm revealed that 48 (18.2%), 71 (26.9%), 142 (53.8%), and 3 (1.1%) ADRs were Definite, Probable, Possible, and Unlikely, respectively. The causality assessment results (Naranjo categories) of different ADRs based on the involved organ/system are also reported in Table 2.

Out of all the 264 recorded ADRs, only 13 (5%) ADRs were preventable, and 251 (95%) ADRs were unpreventable using the P-Method algorithm.

Discussion

This observational study tried to provide a picture of the rituximab safety profile in the treatment of autoimmune neurological diseases in Iranian patients.

Rituximab ADRs occurred commonly in the study patients, and the frequency was remarkably higher than values reported in previous studies on patients with multiple sclerosis (MS), neuromyelitis optica spectrum disorder (NMOSD), and myasthenia gravis (MG). Moreover, according to the Naranjo causality assessment scale, nearly 40% of reported ADRs were definite and probable.

Mild to moderate ADRs have been reported in 26.7% of patients in a systematic review and meta-analysis in which the efficacy and safety of rituximab were studied in the treatment of relapsing-remitting MS[15]. Additionally, a systematic review and meta-analysis on the efficacy and safety of rituximab in NMOSD reported that 26% of patients developed ADRs[10]. However, a more recent study in Iranian patients with NMOSD who were treated with rituximab revealed that 68.2% of patients experienced at least one ADR [16]. The observed variation in the frequency and occurrence rate of rituximab ADRs among studies could have resulted from differences in study designs and patient follow-up durations. Moreover, previous treatment with immunomodulating agents and other antibody therapies (e.g., natalizumab) should be considered in the interpretation of differences.

The Demographic and some of the clinical characteristics of the patients could affect the occurrence of ADRs [17]. Similar to our study patients, most of the patients in the aforementioned studies were middle-aged adults, and females constituted 60 to 70% of the patients. We observed that women experienced more ADRs than men. However, the difference was not significant. Moreover, we could not find any other study on autoimmune neurological diseases that reported an association between patients’ characteristics and the occurrence of ADRs following rituximab administration.

Some of the patients who were recruited in this study were receiving other medications such as anti-diabetic drugs. However, as they were taking those medications for a long period before the start of this study and reported no ADRs, it is unlikely that the ADRs be related to other medications. Also, all the patients’ medication lists were checked by a clinical pharmacist to see if there are any drug interactions that may contribute to ADRs.

Nearly one-third of our patients experienced only one ADR; among them, 60% experienced infusion-related reactions. In other words, 68% of our study patients reported infusion-related ADRs that were generally mild, and four patients developed moderate to severe reactions. In a study by G. ‘D’Arena et al., the frequency of infusion-related ADRs were 25-39.5% in hematologic malignancies and 9.4–17.5% in autoimmune disorders [18]. Infusion-related ADRs due to rituximab administration have been reported in nearly 3 to 30% of patients with autoimmune neurological diseases in different studies[7–11].

On the other hand, SL. Hauser et al. reported that 78.3% of the patients developed this type of ADR in their phase II trial on rituximab in MS [19]. The observed difference between reported frequencies of infusion-related ADRs among retrospective and registered-based studies and prospective, cohort observational studies or clinical trials might result from the difference in nature of the study designs and limitations of retrospective studies and patients’ medical and drug history. However, in a retrospective study that evaluated the efficacy and safety of a branded preparation of rituximab in Iranian patients, a frequency of 70% for infusion-related ADRs has been reported that were generally mild and self-limited [20]. Differences in dosing regimen and premedication might be another reason for observed differences in the frequency of infusion-related ADRs. For example, in the clinical trial by Hauser SL et al., premedication included acetaminophen and H1 receptor antagonists [19]. In our study and most of the aforementioned observational studies, the premedication consisted of a corticosteroid, antihistamine, and acetaminophen. However, the corticosteroid dose considered in the studies is higher than what was used for our patients (i.e., 200–250 mg methylprednisolone or prednisolone in other studies versus 100 mg hydrocortisone in our study). Hence, it should be mentioned that the premedication is considered to reduce the risk or severity of ADRs and therefore, not using any premedication or using a different regimen could affect the rate of ADRs in a similar population.

According to the Naranjo causality assessment scale, about 70% of the reported infusion-related ADRs in our study were definitely due to rituximab. To the extent of our knowledge, no other study described the causality of rituximab ADRs.

One of the challenging and frequent ADRs of rituximab is infection. The overall infection rate ranged from 9 to 40% in previous studies [7–11]. Similarly, nearly 27% of our patients reported infection. Urinary tract and upper respiratory tract infection were the most common ADRs in our patients as well as in other studies.

A nationwide register-based cohort study in Sweden showed that rituximab was associated with a higher rate of serious infections than disease-modifying agents in patients with MS. However, the increased rate was not concerning, as declared by the authors [21].

In the current study, there were 7 cases of severe infections (7.2%). One of these patients developed pneumo-sepsis due to a PCP infection and died. Previous studies have also reported PCP infections in patients treated with rituximab-based regimens [22, 23]. One of our study patients, a 60 years old man, developed sepsis 11 days following rituximab administration and died following DIC. This progressive and fast ADR episode due to rituximab can be alarming when the drug is used for elderly patients.

In our study, a patient developed a lung tumor 22 months after rituximab administration. Scherer et al. has also reported developing different types of tumors following the rituximab administration [24]; on the other hand, another study suggested that malignancies following rituximab administration can be just a coincidence. However, they recommended that there should be close monitoring for secondary malignancies (particularly skin malignancies) in patients treated with the medication [25]. A case of squamous cell carcinoma in a 7-year-long cohort in Germany has put more concern on the development of secondary malignancies following rituximab therapies [26].

Although rituximab ADRs were generally well-tolerated in our study patients and nearly 90% of the ADRs were not serious, 95% of the detected ADRs were not preventable. This could shake the bells for clinicians that no more efforts could be made to reduce the frequency, and close monitoring would be crucial to detect the ADRs and manage them in a timely manner.

This study was limited by its cross-sectional study design. A longitudinal design could have elucidated the pattern of rituximab ADRs, specifically in terms of time-related patterns. Moreover, the small number of study patients has also limited the comprehensiveness and generalisability of the study results.

Conclusion

Considering that rituximab is used as an off-label treatment for autoimmune neurological diseases, a risk-benefit assessment would be necessary before deciding on the treatment choice. Rituximab had an acceptable safety profile in our study patients. However, there must be certain cautions regarding the use of the medication for the elderly or patients with a compromised immune system. Timely detection and management of ADRs would also be crucial to prevent severe and permanent damages. Additionally, close monitoring for secondary malignancies may be beneficial in some cases, but further studies are necessary in this regard.

This is similar to the numbers seen in Lee et al., which reported that about 70% of the patients receiving rituximab for the first time experience ADRs [27] but much higher than some other studies evaluating the safety of rituximab in Rheumatoid arthritis [28, 29].

Declarations

Ethical approval

The ethical evaluation of this study was performed by Tehran University of Medical Sciences’ (TUMS) ethics committee. 

Informed consent

Written consent was taken upon patient enrollment.

Conflict of interest

The authors declare that they have no conflict of interest.