Over the past 100 years, the treatment for rheumatoid arthritis (RA) definitely changed. The concept that pharmacological agents should be used to reduce damage caused by the disease, rather than simply controlling symptoms, resulted in the introduction of several agents initially called “slow-acting anti-rheumatic drugs” (SAARDS). These included drugs such as intramuscular or oral gold salts, D-penicillamine, antimalarial agents, sulphasalazine, and methotrexate (MTX). ¹ Later on, the acronym SAARDS was replaced by “disease modifying antiheumatic drug” (DMARD), which became the cor- nerstone of the therapy. ² Then, the advent of biologic therapy, including the several anti-cytokine therapies, T cell-modulating therapies, and rituximab changed the management of RA. ^2,3

Rituximab is a chimeric, monoclonal antibody (mAb) that specifically joins to the antigen CD20 expressed on the surface of the majority of the human B lymphocytes. ^1,4 It acts in some autoimmune diseases based on antibodies and has become the hallmark of anticancer mAbs, which offers a new approach to the treatment of B-cell lymphoproliferative disorders. ⁴

B cells may promote synovial inflammation for cytokines, chemokines, and auto-antibodies since they are antigen-presenting cells and activators of T cells. Thus, depletion of B cells could interfere with different mechanisms involved in the pathogenesis. ^2,5,6 Rituximab achieves selective B lymphocyte depletion. ^7,8 Hence, the development of biologic agents represented a major step forward in the treatment of RA; but, some of these products may induce an immune response, especially when administered as multiple doses in the long-term. ^9,10

However, rituximab can trigger acute reactions associated with the first infusion such as pruritus, fever, urticaria / rash, chills, rigors, sneezing, angioneurotic edema, throat irritation, cough and bronchospasm, hypotension, or hypertension. ¹¹

The reason why rituximab may cause infusion-related reactions (IRRs) during the first infusion may be partially a response to cytokine release. ¹² It has been seen that the levels of inflammatory cytokine factor alpha necrosis (TNF-α), interleukin-6 (IL-6), interleukin-8 (IL-8), and interferon-gamma (IFN-γ) are significantly increased during the administration of rituximab compared with baseline measurements. ¹²

In Mexico, pharmacovigilance became a mandatory practice 10 years ago (NOM 220-SSA1-2002). ¹³ Our group has previously performed a pharmacovigilance study of rituximab on Mexican patients with lymphoma. Our results show that the peri-infusional reactions observed are the same as in other populations, but the incidence is lower. ¹⁴

To our knowledge, there is no pharmacovigilance information for rituximab in Mexican patients with RA. Hence, our aim was to evaluate the frequency and characteristics of adverse reactions to rituximab in Mexican patients with RA.

A prospective, observational, descriptive open-label, multicenter cohort study of intensive pharmacovigilance was conducted between September 2008 and February 2010 in 10 hospitals throughout our country.

This work complies with the principles of the 2008 version of the Declaration of Helsinki. The study was approved by Centro Nacional de Farmacovigilancia and evaluated by an independent bioethics committee. For the sample calculation, Levin’s study ¹⁵ was taken as a reference, resulting in our sample in 227 patients.

Rituximab adverse reactions were collected in a checklist. The same checklist was used in all centers in order to standardize the collection of adverse events. Data were gathered by the nurse in charge, under the supervision of the treating physician. The treating physician was also responsible for grading the observed events according to the NIH and Naranjo Scales.

We obtained patients data from their clinical records, including age, sex, diagnosis, premedication, rituximab dosage, infusion duration, adverse reactions, use of simultaneous drugs, and, if required, causes of withdrawal of the rituximab infusion. We included male and female adult patients with RA treated with rituximab (MabThera ^® , F. Hoffmann-La Roche, Basel, Switzerland) in combination with methotrexate. Rituximab was indicated as an intravenous infusion in two doses of 1000 mg on days 1 and 15. ¹⁶ Pregnant women were excluded. All records with incomplete information were excluded from the analysis.

Adverse reactions were classified according to the National Cancer Institute Scale in five grades (mild, moderate, severe, life-threatening or disabling, death). ¹⁷ Causality was established using the Naranjo algorithm and considered as possible or probable. ¹⁸ Premedication protocol for all the centers consisted of hydrocortisone 100 mg IV, chlorpheniramine 100 mg PO, paracetamol 1000 g PO.

Infusions were performed following the protocols of each center. They were considered as fast (0 to 180 min), or slow (> 181 min). The infusion rate was calculated as follows: mean infusion dose/mean infusion time (mg/min ^-1 ). The first and second infusions were compared to estimate the relative risk of rituximab-induced adverse reactions.

Data analysis was limited to a description of adverse events, as required by the Mexican Official Norm on Pharmacovigilance NOM-220-SSA1-2016. ¹³

Patients were treated with two infusions of 1000 mg of rituximab. The first infusion was given on day 1 and the second on day 15.

We collected the information from 383 patients (350 female, 33 male) with RA who received 740 infusions of rituximab. After discarding all incomplete clinical records, there were 335 patients left: 26 male and 309 female (table 1), with a mean age of 42 years (range = 18 to 81).

Rituximab was prescribed along with methotrexate. Patients received a total of 631 infusions of rituximab and 36 adverse reactions (5.7%) were informed in 24 patients (7.1%). Mean age for the group with adverse reactions was 49 years (range = 20 to 75). All 24 patients with adverse reactions were female.

General characteristics of the group of patients with adverse reactions are shown in table 2. The mean dose of rituximab for the whole group of patients was 1000 mg/ infusion (range = 1000); thus, the mean dose of rituximab in the group of patients experiencing an adverse effect was 1000 mg/infusion.

Twenty patients had an adverse reaction during the first infusion, 4 in the second. Likewise, 2 patients had more than 1 adverse reaction during the first and second infusions. Therefore, 20 out of 36 adverse reactions (55.5%) occurred during the first infusion. Mean infusion duration for the whole group was 260.6 min (range = 60 to 669), while it was 257 min (range = 153 to 387) for the group with adverse reactions. Only 2 out of 36 adverse reactions (5.5%) were associated with a fast infusion. The most frequent adverse reactions were rash (n = 8), dyspnea (n = 5), and headache (n = 4). All adverse effects were considered as mild to moderate according to the NCI Scale (table 2). Thirty-three adverse effects were considered as possible (91.6%) and only three as probable (7.7%).

As previously described and consistent with the literature, we also found that most of the adverse reactions in this study were associated with the first infusion. There were 631 distributed infusions: first infusion 335, second infusion 296 patients.

The frequencies of the adverse reactions in the first and second infusion were recorded. Therefore, we analyzed the characteristics of the first (335) and second (296) infusions in relation to the adverse reactions using an F-test and we did not find a difference between variances.

From the total of 631 infusions, only 18 were rapid (2.9%) and 613 infusions were slow (97.14%); in the rapid infusions there were only 2 adverse reactions (0.31%), so it was decided to compare the first with the second infusion. Mean infusion time for the first infusion was 262.58 min (range = 81 to 669). Mean times for slow and fast first infusions were 260.6 min (range = 81 to 669) and 132.66 min (range = 60 to 153), respectively. Mean rate of infusion for the first infusion was 3.8 mg/min -1 (range = 1.5 to12.34), while for slow and fast infusions it was 3.837 mg/min ^-1 (range = 1.49 to 5.55), and 7.5 mg/min -1 (range = 16.66 to 6.53), respectively. The relative risk of suffering an adverse reaction was 7.16% with the first infusion and 4.05% with the second infusion. The relative risk of suffering an adverse reaction in the first infusion was 7.69% with the rapid infusion and 7.11% with the slow infusion. The relative risk of suffering an adverse reaction in the second infusion was 0% with the rapid infusion and 4.28% with the slow infusion.

In this study, adverse reactions to rituximab were presented according to the patterns reported worldwide for rituximab, they cannot be distinguished from a hypersensitivity reaction and, according to the new classification for adverse reactions by biologics, they are type α.

All patients received premedication with intravenous corticosteroids (hydrocortisone 100 mg IV), chlorpheniramine 100 mg PO and paracetamol 1,000 mg PO. However, 10 patients did not receive an antihistamine and none of them presented an adverse reaction.

In Latin America, the pharmacovigilance programs are not common, and Mexico is not an exception. So far, no formal study of pharmacovigilance had been conducted in Mexico.

The available information on the adverse reactions to drugs comes from populations different from the Latin-American one; therefore, pharmacovigilance studies are very desirable because they allow us to obtain extremely relevant data about the behavior of such populations when administered drugs and about the adverse reactions they may trigger.

Hence, a worldwide, current challenge lies in biopharmaceuticals use for the treatment of several disorders, such as neoplasms and autoimmune diseases like RA; as well as in their adverse reactions, which are not present in populations similar to ours. In other words, previous clinical trials of drugs pharmacology were performed in ethnically different populations from ours; thus, the genetic features may be dissimilar. This could lead to pharmacokinetic variations and to diverse responses in our population, such as a lack of response to the study drug or higher toxicity than those reported worldwide.

However, these types of studies are being carried out in our population. Then, we used rituximab despite its well-known adverse reactions and although they have never been explored in our population. On the one hand, the exact triggering mechanism for adverse reactions of monoclonal antibodies is often unclear. Clinically, they cannot be distinguished from a type I IgE-mediated hypersensitivity reaction, but their mechanism is different, apparently due to the cytokines released. According to the new classification of adverse reactions for biological agents, these are type α. ^19,20 On the other, it is important to know the atopic background of the patients since it has been related to adverse reactions to other antibodies. ²¹ Therefore, to reduce the risk of reactions of hypersensitivity during the infusion, it is important to record allergy to drugs, the administration of adequate premedication, careful monitoring of the patients, as well as the early intervention before the onset of an adverse reaction. ¹⁹

The percentage of adverse reactions reported in this study was lower than other authors’. Twenty-four patients (7.1%) out of 335 patients with RA received 631 infusions, only 24 referred mild to moderate adverse reactions.

In patients with RA, Emery P ¹¹ reported adverse reactions moderate to severe in an IIb phase study of 81% in 124 patients who received two infusions of 500 mg, most of them in the first infusion. Other authors reported 36% of adverse reactions mild to moderate during the first infusion, in 121 patients. ²²

It is considered that the incidence of adverse reactions in RA is lower in frequency and severity due to the fact that these patients do not experience cytokine release syndrome caused by tumor cell lysis that occurs in patients with B cell malignancies. ²²

Perhaps, the low frequency of adverse reactions in the studied population might be related to the majority of the patients having received premedication based on corticoids, antihistamine drugs, and paracetamol during the first and subsequent infusions. Even though the genetic polymorphism of the population was not studied in this cohort, it might be influential.

Several pathogenic mechanisms describing the infusion-related syndrome have been postulated, including cytokine production (tumor necrosis factor-α, interleukin-6, interleukin-8, and interferon-γ), complement activation, intravascular coagulation, and tumor lysis syndrome. ^20,23,24 The genetics of the population can also influence the immunogenicity of the biopharmaceutical drugs. ¹⁰

According to their classification, adverse reactions to biopharmaceutical drugs are related to cytokine release syndrome (type α); hypersensitivity by IgE, IgG, and T cells (type ß); immune imbalance syndromes (type γ); cross reactivity with a cells or tumor cells (type δ); and finally, non-immunological reactions that reveal unknown functions of biological agents (type ε). ²⁰

There are many factors influencing the immunogenicity of biopharmaceutical drugs. Some of them are associated with the manufacture and storage of biological agents, or to some other reasons: 1) structural properties (e.g. sequence variation and glycosylation); 2) storage conditions (denaturation or aggregation caused by oxidation); 3) contaminants or impurities in their preparation; 4) dose and length of treatment; 5) route of administration; 6) appropriate formulation; 7) patients’ genetics. All of them can result in adverse reactions. ^10,25

The causality obtained with the Naranjo algorithm during rituximab peri-infusion ranges between possible and probable; thus, we cannot completely assure that the adverse reactions reported in this study were caused by rituximab since there may be other risk factors that trigger adverse reactions. The speed of infusion did not seem to influence the appearance of adverse reactions, which were present mainly in the slow infusions. The speed of infusion was determined according each center protocol and sometimes lasted more than six hours as a safety measure.