Infliximab in IBD:9-year Experience in Clinical Practice

Alimentary Pharmacology & Therapeutics. 2010;31(5):553-560. © 2010 Blackwell Publishing

Infliximab Safety Profile and Long-term Applicability in Inflammatory Bowel Disease: 9-year Experience in Clinical Practice

Abstract
Background Most available data on infliximab therapy come from large, short-term, pivotal RCTs and concerns about long-term safety profile still remain.
Aim To evaluate the long-term safety profile of infliximab in inflammatory bowel disease (IBD) in a clinical practice setting.
Methods Since 1999, all IBD patients treated with infliximab were registered and clinical outcomes prospectively recorded up to March 2008, loss of follow-up or patient's death. Infliximab regimens and preventive measures were in accordance with the prevalent guidelines or with the manufacturer's recommendations.
Results One hundred fifty-two patients were included (121 Crohn's disease, 24 ulcerative colitis, 7 indeterminate colitis), with a median of 5 infliximab infusions (IQR 3–8) and 87% of patients received at least three infusions. Seventy-nine per cent of them received concomitant immunomodulators and 70% were pre-medicated with hydrocortisone from the first infusion. After a median follow-up of 142 weeks, 13% presented infusion reactions, 13% viral or bacterial infections and two patients developed neoplasia. The mortality rate was 2.6% (four patients).
Conclusions Infliximab therapy is safe when the recommended preventive measures are implemented, with a rate of serious adverse events less than 10%. No new safety signals were found.

Introduction
Infliximab (IFX) is a chimeric IgG antibody against tumour necrosis alpha (TNF) with proven efficacy in both Crohn's disease (CD) and ulcerative colitis (UC) for inducing and maintaining remission.[1–6] Despite its widespread use in the last years, concerns about its safety profile still remain. Side effects are mainly related to immunosuppression (tumours and infections) or drug-related immunogenicity (acute infusion reactions – AIR, delayed hypersensitivity reactions – DHR, secondary loss of response). While the real effect of anti-TNF agents on cancer development still remains controversial (at least when these agents are used in combination with thiopurines or methotrexate), a variety of opportunistic infections during anti-TNF therapies led to the implementation of several preventive measures, such as an appropriate latent tuberculosis[7, 8] and hepatitis B infection work-up[9] or Pneumocystis jiroveci (carinii) prophylaxis when IFX is to be used, and especially in the setting of combined immunosuppressive treatment.[10] On the other hand, although immunogenic-related adverse effects used to be mild and can be easily managed, they can lead to treatment discontinuation. The use of concomitant immunomodulators (IMM),[11] pre-medication with i.v. hydrocortisone,[12] the scheduled administration of IFX infusions every 8 weeks or even induction regimens with three infusions,[13, 14] should be considered as appropriate strategies to reduce drug-induced immunogenicity.[15]

Most available data on IFX therapy come from large pivotal RCTs, but both efficacy and safety profile may differ from them in clinical practice. Recently, emerging 'top-down' treatment strategies propose to introduce both biological agents and/or immunomodulators earlier in the course of the disease,[16] exposing a greater proportion of patients to a potentially more toxic therapy. From this point of view, long-term follow-up safety data coming from clinical practice remain critical. The aim of this study was to evaluate the safety profile of IFX in inflammatory bowel disease (IBD) patients reported as the experience of 9 years in a Catalonian tertiary referral centre cohort.

Patients and Methods
Study Population and Data Collection
Since 1999, when IFX was first used in a compassionate manner for patients with CD in our centre, clinical and epidemiological characteristics, concomitant treatment, and any adverse event that occurred in the setting of IFX therapy for the treatment of IBD were prospectively registered in our IBD database.

Data of every IBD patient using IFX was collected from diagnosis until March 2008, loss of follow-up or patient's death. Patients who did not visit the outpatient clinic within 6 months of data compilation were contacted over the telephone.

Infliximab treatment regimens, as well as preventive measures, were in accordance with the prevalent guidelines proposed by scientific committees or the manufacturer at each moment. Therefore, IFX dosing was 5 mg/kg body weight as a 2 h i.v. infusion in all patients. Concomitant IMM were maintained when possible, and from 2004, all patients were pre-treated with 200 mg of i.v. hydrocortisone before every IFX infusion. Until 2006, patients were admitted to receive their IFX infusions and from January 2007, they came to the out-patient infusion unit for drug administration. Patients were monitored at each infusion by an expert physician and asked about symptoms of infection or any other adverse event; in addition, full blood count and serum C-reactive protein and alanine-aminotransferase levels were determined.

Study Outcomes
For the study purpose, any adverse event (infections, drug-induced immunogenicity, neoplasia, death) was recorded as 'during IFX' or 'post-IFX treatment' whether it occurred within or after 12 weeks from last dose, respectively.

Acute infusion reaction was defined by the occurrence of any adverse event during or within the first 24 h after IFX infusion. DHR was considered as arthralgia, myalgia, skin rash, fever and/or fatigue occurring 24 h to 14 days after IFX administration.

This was an observational study according to the Declaration of Helsinki and approved by the Institutional Review Board of our centre.

Statistical Analysis
Descriptive statistics were calculated as percentages for discrete data and medians with interquartile ranges (IQR) for continuous data. Univariate analysis (Chi-square test or Fisher's exact test -as needed-, and Mann–Whitney tests) was performed to assess the associations between possible risk factors and the appearance of any adverse event. All statistical analyses were performed using the SPSS 12.0 for Windows package (Inc. Chicago, IL, USA).

Results
One hundred and fifty-two patients were included, 121 had CD, 24 UC and seven indeterminate colitis. Baseline clinical characteristics are shown in Table 1. The median number of IFX infusions was five (IQR 3–8) with 87% of patients receiving at least three infusions. A total of 68 patients (45%) received IFX just for induction of remission (1–3 infusions), sixty-six patients (43%) received an induction regimen followed by scheduled maintenance infusions every 8 weeks for 1 year, and only 18 patients (12%) followed IFX maintenance therapy for longer than 1 year. Seventy per cent of patients received hydrocortisone pre-medication from the first IFX infusion. The median follow-up from the first IFX dose was 142 weeks (IQR 52–252) and none of the patients was lost to follow-up while on IFX therapy. A total of 64 adverse events occurred in 49 patients (32%), leading to transient discontinuation of treatment in nine, and definitive discontinuation of the drug in 14 of them.

Infectious Complications
Infectious complications are listed in Table 2. Twelve viral infections (8%) were registered. At the time of viral infections, all patients were on IMM (azathioprine, mercaptopurine, or methotrexate), eight out of these twelve viral infections (67%) occurred during IFX therapy (all of them caused by herpes virus). Of interest, none of these patients was on systemic steroids at the time of infection. No irreversible consequences or deaths were related to viral infections, but two patients required hospitalization. One patient had 1 week of odynophagy, fever, multiple adenopathies and mild elevation of alteration of liver function tests and was finally diagnosed with infectious mononucleosis. The other patient had lymphocytic meningitis secondary to herpes virus infection.

Eleven bacterial infections were found in 10 patients (6.5%). All patients required hospitalization or prolonged their current hospital stay as a consequence of the infectious process. Sixty percent of the patients were on steroids at the moment of bacterial infections. Interestingly, when taking into account only those patients who developed bacterial infections during IFX therapy, all of them were on steroids at that time. A 65-year-old man with a past history or cerebral stroke, permanent aphasia and right hemiparesia, presented with bronchoaspiration with a secondary pneumoniae 20 months after the last IFX infusion. This patient died as a consequence of a severe bacterial infection.

Malignancies
Two patients treated with IFX developed malignancies (1.3%). The first patient had a recent diagnosis of ileal CD and presented a dramatic worsening immediately after the first IFX infusion. The surgical ileal specimen concluded the diagnosis of Burkitt's lymphoma. He died only a few months after. Another patient was diagnosed with a testicular seminoma, more than 5 years after IFX the last infusion, requiring a curative orchiectomy.

Immunogenic-related Side Effects
Twenty patients (13%) had 22 episodes of AIR. Clinical manifestations of AIR are listed in Table 3. AIR occurred after a median of three infusions (IQR 2–5). Of interest, 75% of the patients with AIR were on concomitant IMM and 55% of these patients were being pre-treated with i.v. hydrocortisone since their first IFX infusion. In total, the occurrence of AIR led to IFX discontinuation in eight patients, four required no treatment and 10 achieved successful IFX reinfusion after reducing the infusion rate, or i.v. antihistaminics administration. No association of clinical or therapeutic factors with AIR was found. None of the AIR episodes was life-threatening. Only one delayed hypersensitivity reaction (serum-sickness like disease) was recorded.

Two female patients were diagnosed with drug-induced lupus erythematosus (DLE) and required hospitalization. One patient had photosensitivity, erythema of both hands, arthritis of fingers and toes and conversion from basal negative towards positive antinuclear antibodies (ANA) (1/640) with negative anti-DNAds, after five IFX doses and 6 weeks from the last infusion. She was treated with steroids and hydroxichloroquine after IFX withdrawal. She maintained CD remission only with IMM subsequent to DLE. The other patient presented, 3 weeks after the last of four IFX infusions, with skin rash and polyarticular arthritis. Despite ANA and aDNAds were persistently negatives, she was diagnosed with DLE and treated likewise with steroids, achieving clinical improvement. As she maintained luminal intestinal activity, she was switched to adalimumab with initial good response and tolerance. Eight additional patients converted anti-DNA to positive within IFX therapy, but none of them developed autoimmune conditions.

One patient developed an autoimmune aortitis after IFX therapy, requiring systemic steroids. The patient was further treated with adalimumab with no worsening of the autoimmune aortitis.

Neurological and Cutaneous Adverse Effects
Two patients developed neurological disorders with complete resolution after specific treatment. Neither of them presented with after-effects. Two weeks after the first IFX dose, one patient presented headache, fever, hemihypoesthesia, nystagmus, and left Babinsky reflex. Magnetic resonance imaging and central nervous system scan ruled out cerebral involvement with findings consistent of pansinusitis. Microbiological and biochemistry study of cerebrospinal fluid revealed no further disease. Neurological symptoms solved after antibiotic treatment. The remaining patient had loss of strength on the right hand and left foot 10 weeks after completing a 3-infusion schedule. Electromyography confirmed diffuse motor involvement of the deep and median peroneal nerves, with negative antigangliosid antibodies. He was diagnosed with multifocal motor neuropathy and received i.v. immunoglobulines for a month with complete clinical resolution.

One patient presented with worsening psoriasis and two patients were diagnosed with inverted psoriasis during IFX therapy, requiring treatment discontinuation in two of them (Table 4).

Mortality
The mortality rate of the whole series was 3% (four patients). The median age of the deceased patients was 51 years (range, 28–65). A 60-year-old man had a surgical ileo-anal reservoir resection and died post-operatively 17 weeks after his last IFX infusion because of a methicillin resistant Staphylococcus aureus pulmonary infection with multiorganic failure and septic shock. Another 28-year-old man had an elective proctectomy and developed abdominal sepsis, abdominal haematoma, leading to hypovolemic and septic shock, and died 15 weeks after the last IFX infusion. As previously mentioned, two additional patients died of bronchoaspirative pneumoniae and Burkitt's lymphoma respectively. Among these deaths, only the latter occurred during IFX therapy and it was considered probably related to IFX treatment.

Infliximab therapy has proven efficient for the treatment of luminal or fistulizing CD and UC resistant to standard therapy.[2, 5, 6] Moreover, its therapeutic effect has been shown to be even better in clinical practice than in the setting of RCTs.[17] To date, two large series assessing the long-term safety profile of IFX in clinical practice were published soon after IFX approval for Crohn's disease[18, 19] and two additional cohorts of IBD have been evaluated for the same purposes recently.[14, 20] Beyond results of large pivotal trials, clinical experience has confined the initial fears on the drug's safety profile to three main groups of adverse effects: infections, immune-related conditions and carcinogenesis. However, each one of these series raised different safety signals such as lymphoma development,[19] neurological disorders and lung cancer[20] or skin eruptions,[14] suggesting the need for more data to reach firmer conclusions.

Although an increased risk of mild urinary and upper respiratory tract infections was reported at the time IFX was approved, the spreading use of the drug was soon associated with a high risk of serious infections such as tuberculosis[21] or hepatitis B reactivation.[9] A wider range of severe opportunistic infections have been reported to occur during anti-TNF therapy,[22] especially when combined with steroids or IMM.[23, 24] The growing fears on infectious complications prompted the recent publication of general recommendations to prevent, suspect and treat them.[25] We found an infection rate of 8% during IFX treatment, very close to those rates most recently reported.[14, 20] Our data suggest that the pattern of infectious complications among patients treated with IFX seems to rely on concomitant treatments. Viral infections appear in patients concomitantly treated with conventional IMM and most of them were because of herpesvirus, in agreement with the largest single-centre IFX series published to date.[14] Although they were mostly diagnosed during IFX therapy, they all occurred in patients treated with IMM; so, it seems that the risk for such infections is mainly related to the use of IMM and that it may increase when using them in combination with IFX. As shown in a recent prospective study, CD patients treated with azathioprine had an increased risk for skin and genital herpes flares as compared with patients not taking this drug;[26] in contrast, it has been reported that IFX does not increase viral load of some latent viral infections such as JCV, Epstein-Barr virus, human herpes virus and cytomegalovirus.[27] Regarding bacterial infections, available data suggest that they often occur when IFX is used concomitantly to systemic steroids,[18, 19] as in our cohort. Recently, a large Belgian study on IFX safety profile found that the risk of severe infections was significantly increased among IFX-treated patients when concomitant steroids (but not IMM) were administered.[14] Thus, concomitant therapies should be taken into account for aetiological suspicion in IBD patients treated with anti-TNF agents when presenting with fever. No cases of tuberculosis were registered in our cohort despite the endemicity of this infection in Spain. This might be attributable to chance, at least in part, because systematic latent TB investigation before starting anti-TNF therapies (tuberculin skin test –with second test in case of an initial negative one in patients under immunosuppressive drugs – chest X-ray, and epidemiological interview) were not implemented until 2003.

Infliximab immune-related side effects include both drug-induced immunogenicity and the development of autoimmune diseases, the most common one being AIR. AIR seem to be related to the formation of anti-IFX antibodies. They are easy to deal with and treat by slowing the infusion rate and the administration of antihistaminics and/or hydrocortisone, but they can also prompt treatment discontinuation. Contrarily to some series in which only those serious adverse events were registered,[18, 19] we collected every symptom that occurred during IFX infusion. Therefore, the incidence of AIR increased from <4% in the Mayo Clinic series to 13% from our series, but it compared well to that of the Leuven cohort.[14] Although reintroduction of IFX after AIR occurrence was not a usual protocol in our institution during the first years of IFX use, AIR led to drug definitive discontinuation in only eight patients, this means 5% of the whole series (3% in the Leuven cohort). Some algorithms for IFX successful re-infusion have been published from 2006 on.[28]

As far as AIR may be a consequence of antibodies against IFX (ATI),[29] it appears reasonable to assume that concomitant immunosuppressive therapy and hydrocortisone pre-treatment may reduce the magnitude of the immunogenic response, and that means not only reducing AIR incidence but also reducing autoimmune disorders, DHR and secondary loss of response, as well.[11, 12] Nevertheless, AIR occurred despite IMM and pre-infusion of hydrocortisone, given that two-thirds of our patients were on IMM at the time of IFX infusions and half of them received hydrocortisone. We did not measure ATI levels because they are not available in our country, but we do search for ANA as they have been related to IFX immunogenicity.[30, 31] Although a retrospective Swedish study showed that patients with baseline positive ANA have an increased risk for AIR,[32] we did not find any relationship between ANA status and AIR development.

The reported rate of autoimmune disorders in our series agrees with those of other large series.[14, 18–20] Eight patients developed positive aDNA, but only one of them developed an immune-related event (DHR), suggesting that aDNA positivization during IFX therapy seems not to be a risk factor to develop autoimmune disorders. We also recorded three cases of dermatological effects related to IFX (inverted psoriasis and psoriasis worsening) accounting for 2%. The aetiology of this paradoxical reaction (IFX is approved for the treatment of psoriasis) remains unclear, but it has been postulated that tissue damage through autoimmune mechanisms is a consequence of excessive TNF inhibition that may lead to activation of autoreactive T cells.[33] A variety of skin eruptions have been reported to occur in almost one-fifth of the patients,[14] but we did not find it at all. Although we do not have a clear explanation for this, it could be argued that differences in environmental (such as weather, diet) or even genetic factors might play a role. The incidence of cutaneous cancers related to IFX seems to be not increased.[34] However, given the high incidence of uncertain cutaneous lesions among IFX users and the concomitant use of thiopurines (that increase the incidence of non-melanoma skin cancers) in a great proportion of these patients, prospective, specifically designed studies in large cohorts are warranted.

Although data related to cancer incidence among IFX-treated patients are still needed, one prospective study suggests that a causal association between IFX and risk of malignant disease is unlikely.[34] In our series, only two tumours were registered, and it could not be ruled out that CD was a misdiagnosis and that the ileal involvement that the patient presented could be also a Burkitt's lymphoma disease.

Mortality in patients with IBD treated with IFX is not increased in comparison with patients not treated with this drug, and mortality is mostly the consequence of complications of the disease.[35] As compared with other retrospective series,[18, 19] our series agrees with this low mortality rate.

In conclusion, our study confirms that the safety profile of IFX, as demonstrated in controlled trials, is reproducible in a clinical-based IBD cohort comprising mainly CD patients. With the implementation of the recommended preventive measures, the rate of serious adverse events is less than 10%, considering IFX therapy safe in the long-term. However, the report of different safety signals from the series published to date warrants large, prospective, long-term studies, based on clinical practice to establish the safety profile of IFX definitively.

Table 1. Baseline clinical characteristics (n = 152)
Expressed as median (range) or frequencies.
IFX, infliximab; IBD, inflammatory bowel disease

Table 2. Infectious complications