Long-term effectiveness and pharmacokinetics of the infliximab biosimilar CT-P13 after switching from the originator during the treatment of inflammatory bowel disease

Abstract

Objective

Switching patients from the originator infliximab to a biosimilar is a measure to expand access to treatments and counteract its negative impact on healthcare budgets. However, industry-independent long-term studies on the effect of switching in real life to support the lack of switch-related problems in inflammatory bowel disease (IBD) patients are sparse, as are studies addressing infliximab pharmacokinetic behaviour. The objectives were to investigate the effectiveness and the pharmacokinetics of CT-P13 after switching from originator infliximab in a real-world population of IBD patients with a follow-up of 2 years.

Method

Prospective, single-centre, observational 2 year study conducted in IBD adult patients with stable disease treated with the originator infliximab who were switched to CT-P13. Four time points were defined for follow-up: prior to the switch, 4–8 weeks after the switch, 8 months later, and 2 years later. Outcome measures were the proportion of patients with clinical, endoscopic and biochemical remission, and changes in biochemical inflammation markers (albumin, C-reactive protein, faecal calprotectin) and infliximab clearance.

Results

42 IBD patients were switched, of which 36 (85.7%) remained on CT-P13 throughout the 2 year study period. Only two patients discontinued CT-P13 due to loss of response. The proportion of patients who displayed clinical, endoscopic and biochemical remission were unchanged during the follow-up (p<0.05) and no statistically significant changes were observed in the biochemical markers of disease activity. The median (IQR) clearance estimated for the infliximab originator before the change was 0.364 (0.321–0.415) L/day, and for the CT-P13 biosimilar it was 0.361 (0.323–0.415) L/day 4–8 weeks after the change, and 0.370 (0.334–0.419) L/day 2 years after (p=0.395).

Conclusion

Switching from originator infliximab to biosimilar CT-P13 did not affect the long-term clinical outcomes or the pharmacokinetic behaviour. This information provides the clinician more evidence for the success of switching and supports non-medical switching in adult IBD patients.

Introduction

Infliximab is a chimeric IgG1 antibody that inhibits tumour necrosis factor α (TNFα). It is approved for treatment for a number of chronic immune-mediated inflammatory disorders, including inflammatory bowel diseases (IBD) such as ulcerative colitis (UC) and Crohn’s disease (CD). Due to the large proportion of patients who respond to infliximab and other TNFα inhibitors and their favourable safety profile, their use has substantially improved the management of these pathologies.^{1 2} However, the cost of these biological agents is very high and the prevalence of IBD is increasing worldwide,³ factors which pose a substantial economic burden on healthcare systems.

After expiration of the patent of the originator infliximab, the first biosimilar of infliximab, CT-P13, at a substantially lower price, was approved by the European Medicines Agency (EMA) and the US Food and Drug Administration (FDA) in 2013 and 2016, respectively.^{4 5} The availability of this biosimilar was considered an opportunity to control treatment costs, thereby improving access to biological drugs.

Nevertheless, despite the cost savings of using biosimilar infliximab, its uptake by many gastroenterologists for IBD patients was initially slow.⁶ This was because approval of CT-P13 by the EMA and FDA was based on two randomised controlled trials carried out in patients with rheumatoid arthritis and ankylosing spondylitis,^{7 8} and approval for IBD and other indications was based on extrapolation—that is, no studies were available with the biosimilar CT-P13 in IBD patients. Data extrapolation raises criticism due to the different mechanisms of action of infliximab between IBD and rheumatological diseases (ie, induction of apoptosis of activated leucocytes vs neutralisation of soluble TNFα).⁹

Additionally, physicians have expressed concern regarding the efficacy, safety and drug antibody formation when switching stable patients treated with originator infliximab to the biosimilar, due to the unavailability of scientific evidence about switching. For this reason, CT-P13 was mainly prescribed for new patients starting treatment with biological therapy, while the decrease in infliximab expenditure would have been more noticeable if the treatments with the originator infliximab were switched to the biosimilar. Subsequently, a few studies of different types in IBD patients have been published which have provided evidence on the efficacy of the biosimilar,¹⁰ and results have demonstrated a lack of switch-related changes in safety, efficacy or immunogenicity.^{9 11–14} These studies have increased clinician confidence in using and switching to the biosimilar.⁶

However, industry-independent long-term studies on the effects of switching in real life are sparse, as are studies addressing possible differences in infliximab pharmacokinetic behaviour after switching from the originator drug to the biosimilar in IBD patients.

The objectives of this study were to investigate prospectively the effectiveness of CT-P13 and to analyse its pharmacokinetics after switching from the originator infliximab in a real-world population of IBD patients with a follow-up of 2 years.

Method

This was a 2 year, single-centre, prospective and observational study conducted in all adult IBD patients treated at our tertiary referral centre with the originator infliximab and switched since April 2017 to the biosimilar CT-P13. In our centre, proactive therapeutic drug monitoring of infliximab levels and other biological drugs is routinely carried out in IBD patients to guide dose adjustments and improve clinical outcomes.¹⁵ In 2017, after CT-P13 was commercialised, the hospital promoted not only starting treatment with this biosimilar in naive patients, but also switching all patients taking the originator infliximab to the biosimilar as a measure to expand access to innovative treatments and counteract negative impacts on the hospital’s financial budget. This study was then designed to follow-up the effectiveness and pharmacokinetics of the biosimilar, helping to monitor successful switching from the originator. The study protocol was authorised by the Ethics Committee of Clinical Research of our university hospital (CEIC number: PI41/03/2017), and all patients signed informed consent regarding their willingness to participate in the study.

Eligible patients were aged 18 years or older with a diagnosis of CD or UC, and treated with the originator infliximab (Remicade). All patients had initially received a 5 mg/kg dose of infliximab by intravenous infusion at weeks 0, 2, 6 and 14 (induction phase) as indicated in the product information. Serum infliximab concentrations were measured immediately before intravenous infusion of the drug (trough concentrations) and, subsequently, dose adjustments were made following our centre’s proactive drug monitoring protocol.¹⁵ Target therapeutic range was established at 3–10 µg/mL according to the available literature,^16–18 except in patients with perianal fistulising disease whose target concentrations were higher (10–15 µg/mL).¹⁹ The use of immunomodulators was permitted. At the time of the switch to the biosimilar, all patients presented stable disease with no changes in their levels of disease activity in the last 4 months, no flares, and with serum infliximab levels within the therapeutic range.

Patient characteristics recorded at baseline included demographic data (sex, age, age at diagnosis, and age at start of infliximab), clinical characteristics (disease extent and behaviour according to the Montreal classification,²⁰ presence or absence of perianal fistulising disease), and concomitant use of immunomodulators.

The switch was done in a controlled setting and four time points were defined for follow-up according to the protocol implemented at our centre: prior to the switch, 4–8 weeks after (before the administration of the second dose of biosimilar infliximab), and 8 months and 2 years after switching. During follow-up, data collection was carried out coinciding with an infliximab infusion and included the following information: infliximab dose and infusion frequency, serum infliximab trough concentration, biochemical disease markers (albumin, C-reactive protein, and faecal calprotectin), disease activity (Harvey Bradshaw Index (HBI) for CD and Partial Mayo Score (PMS) for UC), and colonoscopy and ultrasound findings. Adverse events suspected of being related to the drug, including infusion reactions, were also collected.

Infliximab samples were taken immediately before the intravenous infusion of the drug (trough levels). Serum infliximab concentrations and antibodies‐to‐infliximab (ATIs) were determined at our centre with a capture enzyme-linked immunosorbent assay (ELISA; Promonitor-IFX), which has a lower quantification limit of 0.03 µg/mL for infliximab concentrations. In this technique, serum infliximab binds to TNF and is detected by an anti F (ab0)2-infliximab, horseradish peroxidase-labelled antibody.²¹ This assay equally well quantifies originator and CT-P13 infliximab. ATIs were defined as positive when titres were >3.5 AU/mL according to laboratory assay.

Pharmacokinetic parameters of infliximab for each patient were estimated with a Bayesian approach based on a non-linear mixed population pharmacokinetic model of infliximab developed at our centre for IBD adult patients using the programme NONMEM, version 7.3.0 (Icon Development Solutions, Ellicortt City, MD, USA). Body weight, faecal calprotectin and ATIs proved to be the variables with the most significant impacts on infliximab pharmacokinetics and were the patient parameters included in the model.¹⁵ Briefly, this approach allows for estimating individual pharmacokinetic parameters using the a priori pharmacokinetic parameters of the population model as the starting estimate for an individual, and then adjusting these estimates for the individual based on measured infliximab serum levels, taking into consideration the variability of the population parameters. In our validated bi-compartmental population kinetic model, clearance (CL) is the main parameter that characterises the pharmacokinetic behaviour of infliximab, because the volume of distribution does not present individual variability and was set at a value of 4.8 L for all patients.

Effectiveness variables evaluated were the biochemical inflammation markers and the following clinical outcomes: clinical remission, endoscopic remission, and biochemical remission. Clinical remission was defined as HBI <5 in CD patients and PMS <3 in UC patients. Endoscopic remission was considered as radiological remission with no appearance of echography activity, and biochemical remission was considered achieved when faecal calprotectin was <100 mg/kg. Faecal calprotectin was measured using a capture enzyme-linked immunosorbent assay (ELISA; LIAISON XL, Diasorin).

Descriptive statistics were provided using the median (range or IQR) or mean (SD) for the continuous variables, and frequency and percentage for the categorical variables. Biochemical markers for disease activity and Bayesian estimated clearance at the different time points were compared using the correlated-samples analysis of variance (ANOVA) or the Friedman test as appropriate. Clinical outcomes were compared using Cochran’s Q test. All p values were based on a two-sided hypothesis, and those values of p<0.05 were considered statistically significant. Data visualisation and statistical analyses were carried out in R version 3.3.1 (Comprehensive R Network, http://cran.r-project.org).

Results

A total of 42 IBD patients who had stable disease and had received infliximab originator for at least 4 months and then were switched to the biosimilar CT-P13 were included in the study. Patient characteristics at the time of switching are shown in table 1. CD had been diagnosed in 33 (78.6%) patients and UC in nine (21.4%). Thirty-five (83.3%) patients received thiopurines (n=31) or methotrexate (n=4).

Table 1.

Baseline characteristics of patients included in the study

Number of patients		42
Sex, male (%)		54.8%
Age, years, median (range)		42 (18–70)
Age at diagnosis, years, median (range)		26 (9–52)
Age at start of infliximab, years, median (range)		38 (11–64)
Weight, kg, median (range)		66.5 (43–105)
Height, cm, median (range)		168 (152–185)
IBD type, n (%)	CD	33 (78.6%)
	UC	9 (21.4 %)
CD location, n (%)*†	L1. Ileal	16 (48.5 %)
	L2. Colonic	9 (27.3 %)
	L3. Ileocolonic	8 (24.2 %)
	L4. Upper GI involvement	1 (3.1%)
CD behaviour, n (%)	B1. Non-stricturing, non-penetrating	12 (36.4%)
	B2. Stricturing	16 (48.5%)
	B3. Penetrating	5 (15.1%)
Perianal fistulising disease, n (%)		9 (27.3 %)
UC extent, n (%)*	E1. Proctitis	3 (33.3 %)
	E2. Left-side colitis	–
	E3. Pancolitis	6 (66.7%)
Concomitant immunomodulators, n (%)		35 (83.3 %)
	Thiopurines (azathioprine, mercaptopurine), n (%)	31 (88.6 %)
	Methotrexate, n (%)	4 (11.4 %)

*Montreal classification for CD and UC at diagnosis.

†Patients could present several locations of the CD lesion.

CD, Crohn's disease; GI, gastrointestinal; IBD, inflammatory bowel disease; UC, ulcerative colitis.

Altogether, 36 (85.7%) of patients remained on CT-P13 at the end of the 2 year follow-up and were included in the analysis: 29 patients with CD and seven with UC. During the follow-up period, two patients discontinued CT-P13 due to a loss of response despite adequate serum infliximab concentrations (4.8% of the total at the beginning), and another four patients did not complete the full 2 year follow-up post-switch due to a lack of adherence to treatment (n=2), pregnancy (n=1) or transfer to another health centre (n=1). No patients stopped the treatment due to longstanding remission. No infliximab infusion reactions or other adverse effects associated with treatment were recorded in the follow-up period.

Table 2 shows dosage data for infliximab and serum trough concentrations at the different study time points. The mean dose of infliximab originator that was administered to the patients before the switch was not significantly different compared with the biosimilar CT-P13 dose administered 4–8 weeks after, and 8 months and 2 years after switching (p=0.887). The infusion frequency during the study period was not significantly different (p=1.000). In addition, infliximab trough levels remained stable with no evidence of significant differences before the switch and at the different time points established after switching (p=0.801). It should be noted that overall, over the 2 year follow-up period, 61% of patients underwent dose escalation (interval decrease and/or dose increase) and 44% needed de-escalation (interval increase and/or dose decrease) in order to reach the therapeutic range. During this 2 year study period no patient developed ATIs.

Table 2.

Data on dosage and infliximab serum concentrations during follow-up

	Before switch	4–8 weeks after switch	8 months after	2 years after	P value
Infliximab dose administered, mg/kg (mean (SD))	5.20 (0.62)	5.19 (0.63)	5.12 (0.66)	5.11 (0.61)	0.887
Infliximab dose administered, mg/kg/week (mean (SD))	0.73 (0.18)	0.75 (0.19)	0.76 (0.22)	0.77 (0.21)	0.607
Infliximab dosing interval, weeks (median (IQR))	6 (4–8)	6 (4–8)	6 (4–8)	6 (4–8)	1.000
Infliximab dosing interval, weeks (mean (SD))	7.34 (1.22)	7.22 (1.28)	7.05 (1.36)	6.98 (1.49)	0.463
Infliximab trough concentrations, µg/mL (mean (SD))	5.52 (3.10)	5.98 (3.28)	5.67 (3.08)	5.85 (2.64)	0.801

Regarding pharmacokinetic behaviour, figure 1 represents the boxplots of the distribution of the estimated CL of the patients included in the final analysis at each of the time points. The median (IQR) CL estimated for the infliximab originator before the change was 0.364 (0.321–0.415) L/day, and for the CT-P13 biosimilar it was 0.361 (0.323–0.415) L/day 4–8 weeks after the change, and 0.374 (0.341–0.410) L/day and 0.370 (0.334–0.419) L/day 8 months and 2 years after the use of the biosimilar. No statistically significant differences were found in the estimated drug clearance values at the four time points (p=0.395). Due to the apparently high percentage of patients who received concomitant immunosuppressant at the start of infliximab treatment, a subgroup analysis was performed comparing estimated CL between patients with concomitant immunosuppressant and infliximab monotherapy. No significant differences were found between the groups (p=0.980 and p=0.699, respectively).

Figure 1.

Bayesian estimated clearance of infliximab over the two-year follow up. The box represents the interquartile range. The line in the middle of the box represents the median. The whiskers represent the maximum and minimum values.

There were no clinically relevant changes in the biochemical markers for disease activity over the 2 year study period. Median serum albumin, C-reactive protein and faecal calprotectin concentrations 4–8 weeks after switching to the biosimilar, and 8 months and 2 years after, did not differ significantly from those at baseline with the infliximab originator (albumin, p=0.683; C-reactive protein, p=0.773; faecal calprotectin, p=0.885). Figure 2 shows the boxplot of the distribution of the monitored biomarkers for disease activity. Median (IQR) faecal calprotectin concentrations were 15 (15–30) mg/kg before the switch to the CT-P13 biosimilar, and 16 (15–30) mg/kg, 17 (15–48) mg/kg, and 16 (15–40) mg/kg, 4–8 weeks after the change, 8 months and 2 years after, respectively. Median (IQR) C-reactive protein and serum albumin concentrations at these four time points were, respectively: 0.07 (0.04–0.16) mg/dL, 4.50 (4.25–4.75) g/dL, 0.07 (0.04–0.15) mg/dL and 4.50 (4.20–4.80) g/dL; and 0.06 (0.03–0.18) mg/dL, 4.60 (4.40–4.85) g/dL, 0.09 (0.05–0.17) mg/dL and 4.50 (4.30–4.70) g/dL. No significant differences were found in the biochemical markers for disease activity in patients with concomitant immunosuppressant and with infliximab monotherapy (p>0.050 in all cases).

Figure 2.

Biochemical markers for disease activity:serum albumin (A), C-reactive protein (B) and fecal calprotectin (C) over the two-year follow-up period. The box represents the interquartile range. The line in the middle of the box represents the median. The whiskers represent the maximum (1·5 times higher than the 75th percentile) and minimum (1·5 times lower than the 25th percentile) values. The circles outside the whiskers represent outliers.

The proportion of patients who displayed the various clinical outcomes were unchanged during the 2 years of follow-up (figure 3). The proportion of patients who were in clinical remission before switching to CT-P13 (88.9%) was not significantly different from that at the first control (91.7%), and 8 months (91.7%) and 2 years after switching (94.4%) (p=0.779). The proportion of patients with endoscopic remission before switching (88.9%) did not significantly differ from that at the first control (88.9%), 8 months (86.1%) and 2 years after switching (88.9%) (p=0.779). No statistically significant changes were found in the proportion of patients in biochemical remission before and throughout the different follow-up time points (p=0.965), and 88.9% of the patients remained in biochemical remission after 2 years. In the 2 year follow-up period, seven patients (19.4%) went from remission to disease activity at some point in the study, and six patients (16.6%) went from disease activity to remission. Finally, in the subgroup analysis, no significant differences were found in the clinical outcomes (p>0.050 in all cases).

Figure 3.

Percentage of patients who achieved clinicalremission (A), endoscopic remission (B) and biochemical remission (C) over the two-year follow-up period.

Discussion

The availability of CT-P13, the first biosimilar of infliximab, was considered a great treatment breakthrough for patients with IBD due to the opportunity to allow earlier and broader access to biologics.^{6 22} However, as was already mentioned in the introduction, switching stable patients treated with the original drug to its biosimilar, which is the strategy that is likely to quickly lead to a significant reduction in healthcare costs, was highly controversial for clinicians. Many felt that non-medical switching driven only by economic concerns did not compensate for the lack of safety evidence for the change, particularly for the possibility of developing immunogenicity. This real world, single-centre study arose from the need to provide support—with objective information about serum concentrations, inflammation markers, and clinical outcomes—for switching to the biosimilar and to verify the lack of switch-related problems in safety and efficacy.

Our study results clearly show that stable IBD patients treated with the originator infliximab who were switched in a real-life setting to CT-P13 did not experience changes in effectiveness or tolerability. Furthermore, the study shows that these results are maintained in the long term and that the kinetic behaviour of CT-P13 is also maintained, an aspect that, to our knowledge, has not been studied before.

Only 4.8% of patients who started the study discontinued it due to a loss of response, a value that is low compared with data published in other switch studies,^{11 23 24} though similar to that reported by Høivik et al.¹² The fact that patients had been closely followed by the proactive monitoring programme at our centre for a long time may have contributed to these results. Therapeutic drug monitoring has been recently identified as a factor associated with increased biological agent survival.²⁵ Additionally, patients received comprehensive information before the switch which may have counteracted any possible nocebo effect. This, together with the lack of adverse events registered and the absence of newly detectable ATIs in the patients, demonstrated that the switch was well tolerated in stable patients.

Trough CT-P13 serum concentrations after switching and during the 2 year observational period were not inferior to those displayed before the switch with the originator infliximab. These results are consistent with other studies carried out in adults and paediatric patients with immune-mediated inflammatory disorders that have compared drug concentrations obtained before and after the change.7 9 13 14 23 24 26 27 Regarding pharmacokinetic behaviour, the PLANETA study,⁸ a phase I randomised trial conducted in patients with ankylosing spondylitis, confirmed the equivalence of CT-P13 and originator infliximab and led to CT-P13 approval by the EMA and the FDA. In this study, the pharmacokinetic parameters calculated for acceptance of bioequivalence were peak concentration (Cmax) and area under the concentration-time curve, both determined for CT-P13 and the originator after single administration. In our study, CT-P13 clearance was estimated by Bayesian adjustment using a population pharmacokinetic model, not only immediately after the switch but also up to 2 years later. No statistically significant differences were found in the kinetic behaviour of the drug after the change to the biosimilar (p>0.05), as in the PLANETA study, but also at long term. We again attribute this finding to the fact that the patients were in a stable situation and that ATIs did not occur in any patient throughout the study which could have altered the CT-P13 clearance.

Regarding effectiveness, in our study, no statistically significant differences were found either in the biochemical disease markers or in the clinical outcomes immediately after the switch to the biosimilar infliximab, in the medium term (8 months) or long term (2 years). The good results obtained in clinical and endoscopic remission, which are superior to those observed in previous studies, are truly striking. In the SECURE study, 16 weeks after switching to the biosimilar infliximab, 24% of UC patients and 12% of CD patients were not in clinical remission despite this condition having been one of the inclusion criteria.¹³ In the long-term NOR-SWITCH study, disease worsening occurred in 26% of patients in the originator group and in 30% in the switch group.¹¹ The authors of these two studies pointed out that disease worsening is common in patients treated with both originator and biosimilar infliximab. In our opinion, the lack of disease worsening observed in our patients over the whole 2 year follow-up period could be related to our proactive drug monitoring strategy, which entails determining infliximab concentrations in all patients, even in those who are asymptomatic, to optimise the dose. This strategy improves the efficacy and safety of treatment and reflects decreased development of ATIs.¹⁵ In fact, no ATIs were found over the 2 year follow-up. Van Hoeve et al ²⁶ also used proactive drug monitoring for their switch to the biosimilar infliximab in paediatric patients with IBD, and they found that the proportion of patients in clinical and/or biological remission had not significantly changed 6 months after switching.

At our centre, a proactive drug monitoring strategy was achieved by implementing a multidisciplinary care approach. This approach involves the coordinated efforts of specialists with expertise in their corresponding areas, with the combination of their skills forming a key aspect in helping to improve health outcomes. Hospital pharmacists are capable and uniquely positioned to promote the appropriate utilisation of biosimilar medicines.²⁸ As an expert in clinical pharmacokinetics, the pharmacist contributes to adequately interpreting the results of drug monitoring and, in addition, adequately monitors patients within the framework of the multidisciplinary team.²⁹ This reinforces the position of the European Association of Hospital Pharmacists regarding the role of the hospital pharmacist in this area.²⁸

The main limitation of the study is that it did not have a control group, with patients who continued treatment with the originator infliximab, that would have allowed the results obtained with the biosimilar in the long term to be compared with those of a control group who did not switch. Another important study limitation is the relatively small number of patients, as it was carried out at a single centre. Also, our study population may not be representative of the entire IBD population because only patients who were stable on maintenance therapy were included. Nevertheless, the main strength of this study is that we analysed the effect of switching to the biosimilar in terms of effectiveness, safety and pharmacokinetics, and at different time points, and it is the first study with a follow-up period of 2 years. Furthermore, the data were collected prospectively in routine clinical practice, so our study results reflect the real-life situation. Finally, to our knowledge, this is the first study on the pharmacokinetic behaviour of the change from originator infliximab to the biosimilar CT-P13.

Conclusion

This study shows that switching from the originator infliximab to the biosimilar CT-P13 is well tolerated and does not affect either long-term clinical outcomes or pharmacokinetic behaviour. This information provides the clinician with additional evidence of the success of switching and supports non-medical switching driven by economic reasons in adult IBD patients.

What this paper adds.

What is already known on this subject

• The availability of the infliximab biosimilar CT-P13 is considered an opportunity to control treatment costs and, at the same time, improve access to biological drugs.

• Physicians have expressed concern regarding the efficacy, safety and drug antibody formation when switching stable patients treated with the originator infliximab to the biosimilar, due to the unavailability of scientific evidence about switching.

• Regarding the long-term efficacy of switching, to date only results from studies with a 1 year follow-up period were available.

What this study adds

• Clinical outcomes and pharmacokinetics in stable inflammatory bowel disease patients are maintained 2 years after switching from the originator infliximab to the biosimilar CT-P13.

• This information supports non-medical switching driven by economic reasons in stable inflammatory bowel disease patients.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethics statements

Patient consent for publication

Obtained.