Systematic review: effectiveness and safety of switching between originator infliximab and biosimilar infliximab in patients with inflammatory bowel disease

ABSTRACT

Introduction

Infliximab (IFX) biosimilars are available to treat inflammatory bowel disease (IBD), offering cost reductions versus originator IFX in some jurisdictions. However, concerns remain regarding the efficacy and safety of originator-to-biosimilar switching. This systematic literature review evaluated safety and effectiveness of switching between IFX products in patients with IBD, including multiple switchers.

Methods

Embase, PubMed, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials were searched to capture studies (2012–2022) including patients with IBD who switched between approved IFX products. Effectiveness outcomes: disease activity; disease severity; response to treatment; patient-reported outcomes (PROs). Safety outcomes: incidence and rate of adverse events (AEs); discontinuations due to AEs, failure rate; hospitalizations; surgeries. Immunogenicity outcomes (n, %): anti-drug antibodies; patients receiving concomitant immunomodulatory medication.

Results

Data from 85 publications (81 observational, two randomized controlled trials) were included. Clinical effectiveness outcomes were consistent with the known profile of originator IFX with no difference after switching. There were no unexpected/serious AEs after switching, and rates of AEs were generally consistent with the known profile of IFX.

Conclusions

Most studies reported that clinical, PROs, and safety outcomes for originator-to-biosimilar switching were clinically equivalent to originator responses. Limited data are available regarding multiple switches.

Protocol registration

www.crd.york.ac.uk/prospero identifier is CRD42021289144.

KEYWORDS:

• Biologics
• biosimilars
• Crohn’s disease
• efficacy
• inflammatory bowel disease
• infliximab
• safety, switching
• ulcerative colitis

1. Introduction

The introduction of tumor necrosis factor (TNF) antagonists represented a landmark in the clinical management of chronic conditions like inflammatory bowel disease (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC) [1]. However, the high costs of biologics have had an important impact on healthcare expenditure, especially in chronic conditions, and consequently, patients have had limited access to these effective treatments [2].

Infliximab (IFX; Remicade), a chimeric monoclonal IgG1 anti-TNF antibody, was the first US Food and Drug Administration (FDA) – approved biologic for the treatment of moderate-to-severe IBD, which first gained US regulatory approval in 1998 [3]. Since 2013, several IFX biosimilars have been approved by the US FDA and/or European Medicines Agency (EMA) for the clinical management of IBD [4]. These include (EU and US, respectively) CT-P13 (Inflectra/Remsima; first approval [EMA]: 2013) and infliximab-dyyb (Inflectra)) [5,6], SB2 (Flixabi; first approval [EMA]: 2016) and infliximab-abda (Renflexis) [7,8], GP1111 (Zessly) and infliximab-qbtx (Ixifi; first approval [US]: 2017) [9,10] and ABP 710 (infliximab-axxq; Avsola; first approval [US]: 2019) [11].

Biosimilars are approved following a rigorous process involving detailed analytical and functional assessments and comparative clinical studies showing that they are similar to the originator biologic in terms of safety, pharmacokinetics, immunogenicity, purity and potency [12]. As the development pathway for biosimilars is relatively truncated compared with originator biologics, biosimilars offer the potential for cost savings. By providing competition, biosimilars can improve patients’ access to these medications, allowing more patients the possibility of receiving a potentially more optimal treatment for their disease through additional services or access to newer medicines for those who do not or no longer respond to infliximab or other TNF inhibitors (TNFis) [13,14]. Indeed, biosimilars have alleviated cost-related barriers in some countries and regions, and they have enabled other countries and regions that had minimal access restrictions to channel savings into additional patient services or newer medicines for those who do not or no longer respond to TNFis.

Despite the growing availability of biosimilars and increasing published data on their highly similar clinical profiles (to their respective originators), barriers to biosimilar use exist due to issues around awareness [15], incentive, education [16,17], experience [17], insurance coverage, healthcare system reimbursement, formulary inclusion, drug availability and out‐of‐pocket costs [18–22]. Moreover, many healthcare providers (HCPs) and patients remain skeptical about using biosimilars instead of the originator in bio-naïve patients, as well as switching patients who have responded to an originator to a biosimilar. Lack of familiarity with biosimilars is likely to contribute to persisting doubts regarding the quality, safety and efficacy of biosimilars [23–26].

A prior systematic literature review (Feagan et al.) [27] published in 2019 summarized the evidence on switching between IFX biosimilars and originator in patients with a range of inflammatory disorders, including CD and UC. The 2019 analysis revealed no differences in terms of efficacy or safety associated with single switching [27]. However, the data available were relatively sparse, with only six randomized controlled trials (RCTs) examined. Since that initial systematic review, numerous publications have emanated in the area focusing on IBD alone. In addition, at the time of the 2019 publication, data were only available for single switches (i.e. originator IFX [oIFX] to IFX biosimilar), while relatively limited information is available regarding switching between IFX biosimilars or multiple switching [28]. Moreover, few data were available in 2019 regarding the influence of switching on mucosal healing.

The aim of the present systematic literature review is to evaluate the updated and existing evidence on the safety and effectiveness of switching between IFX products (including biosimilars) in patients with IBD.

2. Materials and methods

2.1. Study design and objectives

The main objective of this systematic literature review is to assess the safety and effectiveness of switching between IFX products in patients with IBD. Patient-reported outcomes (PROs) and immunogenic response are also assessed.

This systematic literature review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and is registered on PROSPERO (Prospective Register of Systematic Reviews) to avoid duplication and reduce potential reporting bias (registration number: CRD42021289144).

2.2. Search strategy

PubMed, Embase via Ovid, the Cochrane Database of Systematic Reviews via Ovid, and the Cochrane Central Register of Controlled Trials (CENTRAL) via Ovid were searched to identify English-language articles (excluding conference abstracts) published between 1 January 2012 and 17 November 2021. Hand searches of abstracts from key IBD congresses were also conducted. The search was updated to include studies published between 17 November 2021 and 4 April 2022.

The search strings included ulcerative colitis, Crohn’s disease and inflammatory bowel disease, as well as the following terms for IFX biosimilars and biosimilar candidates: GP 2018, ABP 710, BCD-055, GP1111, PF-06438179, GS 071, NI 071, SB2, Flixabi, CT-P13, Emisima, Flixceli, Infliximab BS, Remsima, infliximab-dyyb and Inflectra. The search string also contained known synonyms for these molecules (see Supplementary file 1 for full search string).

2.3. Eligibility criteria

Publications of interest measured efficacy/effectiveness, safety and immunogenicity after switching (Supplementary file 2). The effectiveness outcomes measured included disease activity, disease severity and response to treatment, including PROs. Safety outcomes included the incidence and rate of adverse events (AEs), discontinuations due to AEs, failure rate, hospitalizations and surgeries. Immunogenicity outcomes included anti-drug antibodies (ADAs; n, %) and number of patients receiving concomitant immunomodulatory medication (n, %).

Studies were selected for inclusion according to the prespecified Population, Intervention, Comparison, Outcomes, Study framework (see Supplementary file 3 for full eligibility criteria). Briefly, publications were included if they reported real-world evidence/observational studies or Phase II – IV clinical trials (including open-label extensions [OLEs]) involving patients with IBD (CD and UC) who switched between approved IFX products (i.e. from oIFX to biosimilar IFX or vice versa, or from one IFX biosimilar to another). Publications reporting preclinical, pilot or Phase I clinical studies, protocols, case reports/studies, notes, commentaries, letters, editorials, opinions, economic model studies, meta-analyses and reviews were excluded from the analysis. Also excluded were studies reporting outcomes that were not specific to patients switching between approved IFX products or reporting outcomes for mixed treatments and not specific to a given IFX product, as well as studies in pregnant women and studies that did not report effectiveness, safety or immunogenicity following a switch between IFX products. In addition, studies in patients who received a single version of IFX (i.e. originator or biosimilar) without a switch were excluded.

2.4. Screening and data extraction

Results from each database were downloaded to EndNote bibliographic software, with duplicates removed using EndNote algorithms and by hand. To ensure transparency, duplicate citations were retained. A pre-screen of the EndNote database was conducted by a single researcher to exclude records that were noticeably irrelevant (e.g. relating to animal studies or preclinical trials). The excluded records were evaluated by a second researcher and retained in an EndNote folder for transparency. References for which there was a disagreement were retained for title/abstract screening. Titles and abstracts, followed by full texts, were screened for eligibility in parallel by two independent researchers. Reasons for exclusion were provided and cross-checked by both reviewers. Data from full-text articles were extracted by one reviewer into a prespecified Microsoft Excel grid, which was assessed for accuracy by a second reviewer. Reference lists from eligible reviews and manuscripts were also examined to identify any publications that may not have appeared in the database searches.

2.5. Quality/bias assessment

RCTs were assessed for quality using the Cochrane risk-of-bias (RoB) 2.0 tool [29]. Observational longitudinal studies (prospective or retrospective) were assessed for quality/risk of bias using the Newcastle – Ottawa Scale (NOS) [30], which was also used for any non-randomized studies. Observational cross-sectional studies were assessed for quality using the adapted NOS [31].

3. Results

3.1. Study selection

A total of 622 records (Embase, n = 237; Cochrane, n = 42; PubMed, n = 343) were identified from database searches between 1 January 2012 and 17 November 2021 (Figure 1). An additional 55 records were identified in the updated search from 17 November 2021 to 4 April 2022. There was a 94% level of agreement in screening between researchers. After removal of duplicates, screening and application of eligibility criteria (based on interventions, outcomes, patient population and study design), a total of 85 publications (65 full-text articles, 20 conference abstracts) were included in this analysis which comprised 84 studies in total (81 observational studies; and two clinical trials, covered by four publications).

Figure 1. PRISMA diagram.

ACG, American College of Gastroenterology; DDW, Digestive Disease Week®; ECCO, European Crohn’s and Colitis Organisation; L1, title and abstract eligibility screening according to inclusion and exclusion criteria; L2, full-text screening; UEG, United European Gastroenterology.

3.2. Study characteristics

Most studies were conducted in Europe (n = 63) [32–40]; remaining studies were reported in the United States (n = 8) [41–48], Australia (n = 1 longitudinal study) [11], or Asia (n = 6) [49–54], and four were multinational studies [55–58].

The studies reported multiple switching (n = 7) [34,43,47,59–61], biosimilar-to-biosimilar switching (n = 7) [34,47,59,60,62–64], biosimilar-to-originator switching (n = 3) [35,55,65], and single switch from originator to biosimilar (n = 63) (Figure 2). Sample sizes for these studies ranged from 9–1409 participants.

Figure 2. Switch pattern in the included studies.

Among the longitudinal studies assessed for risk of bias using NOS (n = 52), 18 (34.6%), 13 (25.0%), and 16 (30.8%) were considered to have a high, medium, and low risk of bias, respectively. The RCTs were judged to have a low risk of bias based on the Cochrane RoB tool.

3.3. Clinical effectiveness outcomes: IFX originator-to-biosimilar

3.3.1. Randomized controlled studies (n = 2)

Four publications presenting results from two RCTs were identified, including three reporting data from the NOR-SWITCH trial in Norway [66–68] and one reporting the multinational PLANET-CD trial [55] (Table 1). Both included patients with IBD (CD and/or UC) who switched from oIFX to CT-P13.

Table 1. RCTs evaluating switching from IFX originator (oIFX) to biosimilar (N = 2), IFX biosimilar to oIFX (N = 1) or IFX biosimilar maintenance (N = 1).

Author, yearCountry	Reference	Study design Sample size Study period	Population	Transition Timing of switch	Age (years) Female, n (%) Disease duration (months)	Clinical outcomes assessed	PROs assessed	Safety outcomes assessed	Summary
Jørgensen, 2017; Goll, 2019; Jørgensen, 2020 Norway	[66–68]	RCT N = 77 OLE N = 65 Oct 2014 to Jul 2016	CD	oIFX → CT-P13 after ≥6 months’ oIFX treatment	Age: 39.5 (14.2)^a F: 30 (39) Disease duration: 14.3 (8.5)^a	HBI, PGA, CRP, Fc, Remission, Disease worsening	EQ-5D index score, IBDQ, SF-36, WPAI	Overall AEs, TEAEs after switch, SAEs, Discontinuation due to AE, ADAs, All treatment discontinuation	Efficacy, safety and immunogenicity of both the originator and biosimilar IFX were comparable in patients with CD and UC in the NOR-SWITCH main and extension trials. These explorative subgroup analyses confirm that there are no significant concerns related to switching from oIFX to CT-P13 in patients with CD and UC
		RCT N = 46 OLE N = 42 Oct 2014 to Jul 2016	UC	oIFX → CT-P13 after ≥6 months’ oIFX treatment	Age: 44.4 (14.8)^a F: 14 (30) Disease duration: 11.5 (7.5)^a	PMS, PGA, CRP, Fc, Remission, Disease worsening
		OLE of RCT N = 62 Oct 2014 to Jul 2016	CD	oIFX for 52 weeks in RCT → CTP13 in OLE	Age: 40.2 (13.1)^a F: 27 (44) Disease duration: 15.2 (9.1)^a	PGA, CRP, Fc, Remission, Disease worsening	IBDQ	Overall AEs, TEAEs after switch, SAEs, Discontinuation due to AE, ADAs, All treatment discontinuation
		OLE of RCT N = 38 Oct 2014 to Jul 2016	UC	oIFX for 52 weeks in RCT → CTP13 in OLE	Age: 45.9 (13.4)^a F: 14 (37) Disease duration: 13.7 (9.5)^a	PMS, Fc, Remission, Disease worsening
Ye, 2019 (PLANET-CD) Multinational	[55]	RCT N = 220 Aug 2014 to Feb 2017	CD (all-randomized population)	CT-P13 → CT-P13 CONTINUERS at week 30 n = 56	Age: 39.0 (19.0–68.0)^b F: 27 (48.2) Disease duration <3 years: 34 (60.7)^c ≥3 years: 22 (39.3)^c	CDAI, CRP, Fc, Remission, Mucosal healing	PRO-2 score, SIBDQ score	TEAEs, SAEs, Discontinuation due to AE, ADAs, All treatment discontinuation	On the whole, prespecified non-inferiority criteria were met for both populations, and CT-P13 was considered non-inferior to oIFX. CDAI-70 or CDAI-100 response rates and clinical remission rates between groups were similar at weeks 6, 14, 30 or 54. PRO-2 scores improved after study drug infusion, with no notable difference in PRO-2 scores between groups at any time-point. There were no notable differences between the CT-P13 and oIFX groups. Mean Fc concentrations decreased from baseline to week 6 and then remained stable, with no notable differences between groups at any visit. Mean CRP concentrations decreased from baseline to week 6 and then remained stable, with no notable differences between groups
				CT-P13 → oIFX with switching at week 30 n = 55	Age: 31.0 (18.0–69.0)^b F: 21 (38.2) Disease duration <3 years: 27 (49.1)^c ≥3 years: 28 (50.9)^c
				oIFX → oIFX continuers at week 30 n = 54	Age: 31.0 (18.0–69.0)^b F: 28 (51.9) Disease duration <3 years: 30 (55.6)^c ≥3 years: 24 (44.4)^c
				oIFX → CT-P13 with switching at week 30 n = 55	Age: 35.0 (19.0–63.0)^b F: 21 (38.2) Disease duration <3 years: 29 (52.7)^c ≥3 years: 26 (47.3)^c

Abbreviations: ADA, anti-drug antibody; AE, adverse event; CD, Crohn’s disease; CDAI, Crohn’s Disease Activity Index; CRP, C-reactive protein; CT-P13, infliximab biosimilar; F, female; Fc, fecal calprotectin; HBI, Harvey – Bradshaw Index; IBDQ, Inflammatory Bowel Disease Questionnaire; IFX, infliximab; N, sample size at baseline; n, subset of sample; oIFX, originator infliximab; OLE, open-label extension; PGA, Physician’s Global Assessment; PMS, partial Mayo Score; PRO, patient-reported outcome; RCT, randomized controlled trial; SAE, serious adverse event; SF-36, 36-Item Short Form Health Survey; SIBDQ, Short Inflammatory Bowel Disease Questionnaire; TEAE, treatment-emergent adverse event; UC, ulcerative colitis; WPAI, Work Productivity and Activity Impairment Questionnaire.

^aMean (SD).

^bMedian (IQR).

^cn (%).

The NOR-SWITCH trial included Norwegian patients with UC (n = 93) and CD (n = 155), and other inflammatory conditions (n = 482) who had received ≥6 months’ treatment with oIFX. Patients were randomized 1:1 to continue oIFX (maintenance group) or switch to CT-P13 treatment (switch group) for 52 weeks of follow-up in the main study [68], followed by a 26-week OLE trial where all patients received CT-P13 treatment [66,67]. Similar clinical outcomes (Harvey – Bradshaw Index [HBI], Physician’s Global Assessment [PGA], C-reactive protein [CRP], fecal calprotectin [Fc], remission, disease worsening) were reported after switching in the main and OLE trials. At baseline of the main study, the mean (standard deviation [SD]) disease duration in patients with CD and UC was 14.3 (8.5) and 11.5 (7.5) years, respectively, in the switch group [67]. NOR-SWITCH reported a decrease in partial Mayo Score (PMS) at 12 months versus baseline; and in the OLE, PMS was comparable between the maintenance and switch groups [66–68]. This was also true for HBI, CRP, and Fc. HBI increased at 12 months from baseline in the maintenance and switch groups, and decreased at 18 months in the switch group [68]. PGA did not show a significant mean difference at 12 months among 77 patients with CD after switching to CT-P13 (mean [SD], 0.93 [1.4]) versus baseline (0.88 [1.56]), suggesting no worsening of disease activity after transitioning to biosimilar treatment [66–68].

In the multinational PLANET-CD study reported by Ye et al. (2019), patients with CD (n = 220) who had not responded to or were intolerant to non-biological treatments were randomized to initiate CT-P13 (n = 111) or oIFX (n = 109) as follows: CT-P13 → CT-P13 (n = 56), CT-P13 → oIFX (n = 55), oIFX → oIFX (n = 54) or oIFX → CT-P13 (n = 55), with switching at week 30 [55]. Among 109 patients who initiated oIFX, both the oIFX → oIFX and oIFX → CT-P13 switch groups showed similar Crohn’s Disease Activity Index (CDAI)-70 and CDAI-100 response rates and clinical remission rates at weeks 6, 14, 30 and 54. Mean Fc and CRP concentrations decreased from baseline to week 6 and then remained stable, with no notable differences between groups at any visit.

Ye et al. (2019) also assessed endoscopy findings in PLANET-CD to provide objective information about the degree of inflammation [55]. The presence and level of improvement of mucosal damage were evaluated by colonoscopy at baseline and week 54, and mucosal healing was assessed using the Simple Endoscopic Score for Crohn’s Disease (SES-CD); there was no significant change in SES-CD scores between the various switching patterns (Table 2) [55].

Table 2. Mucosal healing associated with switching from (A) IFX originator (oIFX) to biosimilar (N = 4) or (B) IFX biosimilar to oIFX or biosimilar maintenance (N = 1) in patients with CD in RCTs and longitudinal studies.

Author, yearCountry	Reference	Study design	Sample size	Follow-up time	Transition	Outcome results
(A) IFX originator to biosimilar
Ye, 2019 (PLANET-CD) Multinational	[55]	RCT	N = 55	Absolute: 12.5 months	oIFX → CT-P13	SES-CD score: mean (SD), 7.8 (6.70) 12.5 months: n (%), 8 (21.1); N = 38
Armuzzi, 2019; Fiorino, 2017 Italy	[69,70]	Observational prospective	N = 87	Mean (SD): 12.9 (7.6) months Median (IQR): 11.7 (6.04–20.7) months	oIFX → CT-P13	SES-CD score: mean (SD) Baseline: 3.7 (4.5), n = 87 6 months (a): NR, n = 78 12 months (b): 1.5 (2), n = 48; p = 0.06
Padival, 2021; Padival, 2020 USA	[45,46]	Observational prospective	NR	NR	oIFX → SB2	SES-CD score: median (IQR), improvement 0 (−3, 0), N = 7; p = 0.57
Tursi, 2019;Tursi, 2017Italy	[71,72]	Observational retrospective, Observational prospective	N = 45	Absolute: 12 months	oIFX → CT-P13	Mucosal healing during folow-up, n (%): 16 (84.2); N = 19
(B) IFX biosimilar to originator or biosimilar maintenance
Ye, 2019 (PLANET-CD) Multinational	[55]	RCT	N = 56	Absolute: 12.5 months	CT-P13 → CT-P13	SES-CD score: mean (SD), 10.3 (9.09) 12.5 months: n (%), 16 (33.3); N = 48
	[55]	RCT	N = 55	Absolute: 12.5 months	CT-P13 → oIFX	SES-CD score: mean (SD), 10.4 (7.79) 12.5 months: n (%), 12 (26.1); N = 46

Abbreviations: CD, Crohn’s disease; CT-P13, infliximab biosimilar; IFX, infliximab; IQR, interquartile range; n, sample with outcome; N, total sample size; NR, not reported; oIFX, originator infliximab; RCT, randomized controlled trial; SB2, infliximab-abda biosimilar; SD, standard deviation; SES-CD, Simple Endoscopic Score for Crohn’s Disease.

Both RCTs included PRO assessments. In NOR-SWITCH, there was no substantial change in the EQ-5D index score, 36-Item Short Form Health Survey (SF-36) physical component summary score or Work Productivity and Activity Impairment Questionnaire at 12 months, or in the IBD Questionnaire (IBDQ) score at 12 and 18 months from baseline after switching, indicating that CT-P13 has similar efficacy to oIFX [66–68]. In PLANET-CD, a decrease from baseline in 2-item patient-reported outcome (PRO-2; mean [SD]) scores corresponding with improvement (−10.7 [5.94], n = 47) and an increase from baseline in the Short IBDQ (SIBDQ) score (17.0 [10.54], n = 55) were observed at 12.5 months post switch [55]. There were no notable differences in PRO-2 scores between the CT-P13 and oIFX groups at any time-point in the study.

3.3.2. Longitudinal studies (n = 60)

Most longitudinal studies included in this analysis were observational prospective studies. The included publications commonly reported clinical outcomes of non-cardiac CRP, clinical response, remission and disease worsening (Table S1) [11,32–42,44–46,49–51,53,54,57,59–61,69–112].

3.3.2.1. Clinical response (endoscopic)

Four studies reported Mayo Score as an outcome in patients with UC [45,69,72,77]. An Italian observational prospective study conducted by Armuzzi et al. reported a significant decrease (n = 87, p<0.001) in Mayo Score at 6 months versus baseline, which remained stable at 12 months in patients who switched from oIFX to CT-P13 [69]. Furthermore, in the subgroup of patients with UC who had been previously exposed to IFX and were already responders to therapy and likely under remission at the time of switching, there was a significant further improvement in total Mayo Score and endoscopic Mayo subscore at 12 months [69]. Padival et al. reported a similar improvement from baseline in Mayo Score, with no significant change after switching from oIFX to SB2 [113]. Endoscopic Mayo subscore was used as an outcome for endoscopic severity in four further studies [71,72,81,102].

PMS was reported in 21 studies involving patients with UC. An observational prospective study reported a significant response (p<0.001) in PMS among patients switching from oIFX to SB2 (n = 17) compared with patients switching from CT-P13 to SB2 (n = 43) and multiple switchers (n = 24), indicating better clinical response in the single-switch patients [60,94]. Massimi et al. (2021) reported significant changes in PMS associated with clinical remission (p = 0.02) at first follow-up (135 days) versus baseline in patients (n = 85) switching from oIFX to SB2 [96]. Another observational prospective study (n = 158) reported no mean change in PMS after 12.5 months of follow-up in patients switching from oIFX to CT-P13 when compared with baseline [59]. A Finnish observational study (n = 30) reported a non-significant decline in PMS (p = 0.07) after switching from oIFX to CT-P13 compared with baseline [82]. Another observational prospective study (n = 31) by Guerra Veloz et al. reported a non-significant change (p = 0.058) in median PMS at 12 months after switching to CT-P13 [114]. Similarly, Padival et al. (2020, 2021) reported a non-significant decline (p = 0.8) in PMS after switching [45,46].

Mucosal healing (SES-CD) was reported in three longitudinal studies of patients with CD who switched from oIFX to biosimilars (Table 2) [46,69–72]. None of these studies showed a significant change in mean SES-CD score at 6 and 12 months after switching. Moreover, in a 12-month follow-up of the prospective study by Tursi et al. (2019), 16/19 (84%) patients achieved mucosal healing, suggesting longer-term maintenance of disease stability in these patients [72].

Most of the studies that used colonoscopy to assess endoscopic improvement did not describe a standardized formal reporting system for endoscopic assessments [41,48,55,58,72,81,86,95,102,115]. Chaparro (2019) used global endoscopic criteria of mild, moderate or severe activity for patients with CD [81]. For post-surgical patients with CD, endoscopic activity was rated according to Rutgeerts score, and for those with ileal involvement, only complete colonoscopies with ileoscopy were assessed.

3.3.2.2. Clinical disease remission

A total of seven studies reported CDAI scores in patients with CD [36,49,50,53,56,57,91]. An observational longitudinal study (n = 23) in South Korea and the EU reported similar mean CDAI scores from day 0 (time of switch) to 60 months, indicating relative disease stability after switching from oIFX to CT-P13 [57]. Another publication by the same authors reported that a substantial proportion of patients had CDAI scores of<150 points within the same time intervals [56]. Schmitz et al. (2018) also observed that the numbers of patients with CD reporting moderate disease activity (CDAI 150–400) after switching from oIFX to biosimilar were comparable between the four time points measured, with no significant difference in activity scores before and after switching [36]. HBI was reported in 26 studies, three of which reported a significant change in HBI following switching from oIFX to biosimilar IFX [69,96,116]. One observational prospective study (n = 87) noted a significant decline in HBI (p = 0.01) at 6 and 12 months after switching [69,70]. Similarly, Guerra Veloz et al. (n = 116) observed a significant difference (p = 0.001) in HBI (median [interquartile range (IQR)]) when compared between 0 and 12 months following switching [115]. Macaluso et al. (n = 43) observed significantly improved HBI (p<0.001) in the multiple-switch group when compared with single-switch groups (oIFX → SB2 and CT-P13 → SB2) [60,94]. In most of the remaining studies, HBI scores remained stable after switching from oIFX to biosimilar.

Improved pediatric CDAI (PCDAI) scores (<10 points) in a pediatric IBD population (n = 42) were observed 6 months after switching from oIFX to biosimilar CT-P13 versus baseline (p = 0.005), with retention of overall patients’ stability in both CD and UC populations (van Hoeve et al., 2019) [107]. The proportion of patients in clinical remission did not change after switching. Other studies that evaluated PCDAI score as an outcome showed stability or no change in PCDAI scores at 12 months following switching [51,85]. Cheon et al. (2021) reported that PCDAI scores were stable at 12 months and maintained for up to 60 months after switching [56,57].

Two observational studies reported PGA in patients with IBD who switched from oIFX to CT-P13 [85,101]. Both studies reported non-significant changes in PGA score over 12 months suggesting maintenance of clinical activity after switching.

A total of 23 studies reported remission outcomes in patients with IBD. In a prospective observational study (n = 361) by Bronswijk et al. (2020), in which over 60% of patients experienced clinical remission at baseline, a high hemoglobin level and low PRO-2 were significantly associated with clinical remission at 6 months post switch [79]. Disease-worsening end points were reported in five studies among patients with IBD switching from oIFX to biosimilar [11,44,79,102]. Ho et al. (2020) reported that 347 (25%) patients in the CT-P13 group versus 375 (27%) patients in the oIFX group had disease worsening requiring acute care or treatment failure, suggesting better (p<0.01 for non-inferiority) tolerability of CT-P13 versus oIFX [42]. However, the authors concluded that there was no increased risk of disease worsening for patients who switched to infliximab-dyyb compared with patients who remained on oIFX [42]. A total of 114 (10%) patients in the CT-P13 group required acute care due to disease worsening compared with 174 (15%) patients in the oIFX group (p<0.01) [42]. A separate study reported endoscopic remission in 38/119 (55%) patients before switch and 45 (65%) (p > 0.05) after switching from oIFX to CT-P13, with endoscopic worsening observed in 10/119 (15%) patients ~8 months after switching [102].

Nine longitudinal studies in patients with IBD reported PROs after switching from oIFX to biosimilar IFX [37–39,59,79,88,91,99,104,112,116]. In one prospective study (n = 361), the median PRO-2 decreased significantly from baseline to first infusion and at 6 months after infusion in patients who switched from oIFX to biosimilar [79]. Another study by Razanskaite et al. (2017) showed a significant increase in median IBD-Control-8 subscore, which improved during switch and at the third dose [104]; however, IBD-Control-Visual Analogue Scale showed no significant difference (p = 0.65) in mean value after the third dose of CT-P13 compared with before the switch.

Three studies reported PROs among patients with CD who transitioned from oIFX to biosimilar [76,91,106]. No change was observed in the Short Health Scale mean composite score at 12 months after switching to biosimilar in a prospective study (n = 195) conducted by Bergqvist et al. (2018), indicating similar therapeutic effectiveness of biosimilar to oIFX [76]. In a small observational prospective study (n = 26), Huoponen et al. (2020) [91] identified a significant increase (p = 0.018) in median (IQR) SIBDQ score at 3 months (193 [190–206]; n = 20) compared with baseline (188 [170–198]; n = 26), correlating with improved clinical outcomes after switching. Across studies, there were generally no notable differences between the biosimilar IFX and oIFX groups in terms of change from baseline in SIBDQ score among patients with IBD. Similarly, there were no notable changes from baseline in EQ-5D [58] score post switch.

3.3.2.3. Biological disease remission

A total of 12 and 27 longitudinal studies reported Fc and CRP outcomes, respectively, for patients with IBD who switched from oIFX to biosimilar (Tables S2 and S3). In general, mean Fc and CRP concentrations decreased from baseline (start of oIFX therapy) to week 6, then remained stable with no notable differences following switching. Across studies, median Fc concentrations post switch showed no significant change versus Fc concentrations at baseline, which corresponded to the achievement of biochemical clinical remission among patients (Table S2) [32,34,37–39,45,46,51,64,80,82,85,89,91,92,95,96,106]. In a retrospective study by Binkhorst et al. (2018) involving 197 Dutch patients, there was a slight decline in median Fc concentration 3.7 months following switch from oIFX to CT-P13, although still within the normal range as was before switch [32]. In a UK study by Gervais et al. (2018) (n = 27), despite some fluctuations in median Fc concentration from baseline (113.5 mg/kg) to month 6 (148 mg/kg) and month 12 (99 mg/kg) after switching from oIFX to CT-P13, there was no significant difference between time-points (p = 0.98) [85]. Similarly, CRP concentrations were generally maintained following switching across the various observational studies (Table S3) [11,32,34,36–39,41,45,46,51,59,60,62–64,79–81,83–85,87,89,92,94–96,98,99,102–104,107,108,117]. For example, no significant change in median CRP was observed between time of switch (1.4 mg/L) to 6 months after infusion (1.4 mg/L; p = 0.39) in a Belgian study by Bronswijk et al. (2020) involving 361 patients with IBD switching from oIFX to CT-P13 [79]. In another prospective observational study involving 148 German patients switching from oIFX to SB2, Fischer et al. (2021) observed little change in CRP concentrations at 18 months compared with baseline, with a slight decline at 6 months and elevation at 12 months [83].

3.4. Clinical effectiveness outcomes: biosimilar-to-IFX originator, biosimilar-to-biosimilar, biosimilar maintenance, or multiple switching

3.4.1. Randomized controlled studies (n = 1)

The PLANET-CD study [55] evaluated reverse switching from CT-P13 to oIFX versus continued CT-P13 treatment in patients with CD (Table 1). Baseline (i.e. week 30) CDAI scores were higher in reverse-switch patients (n = 55) than in patients who continued CT-P13 treatment (n = 56). At 12.5 months, fewer reverse-switch patients achieved CDAI-70 (n = 39, 71%) and CDAI-100 (n = 38, 69%) change scores compared with those who continued CT-P13 (n = 44, 79% and n = 43, 77%, respectively) [55]. Mucosal healing occurred in 12 (26.1%, n = 46) reverse-switch group, where the mean SES-CD score was 10.4 (7.79); there was no significant change in SES-CD score in either maintenance or reverse switch group (Table 2).

Among reverse-switch patients (n = 55), mean [SD] PRO-2 score (−12.5 [5.92], n = 45) decreased and mean [SD] SIBDQ score (18.9 [10.94]) increased at 12.5 months from baseline after switching. Similar results were obtained at the same time-point in both the mean [SD] PRO-2 score (−12.8 [4.9], n = 40) and mean [SD] SIBDQ score (20.2 [13.1]) in the reverse-switch group [55]. Hence, biosimilar switch, biosimilar maintenance and reverse switch all appear to be associated with amelioration in clinical outcomes.

3.4.2. Longitudinal studies (biosimilar IFX to oIFX, n = 3; biosimilar-to-biosimilar, n = 6; or multiple switching, n = 7)

Longitudinal studies that evaluated biosimilar-to-oIFX, biosimilar-to-biosimilar, or multiple switching are summarized in Table S4 [34,35,40,43,47,59–65,75,94,97,117]. Two longitudinal studies of patients with IBD who underwent biosimilar-to-biosimilar switching (CT-P13 → SB2 [34] or CT-P13 → GP1111 [64]) or multiple switching reported no detectable effect of switching on Fc concentrations (Table S2).

Six studies in patients with IBD who underwent biosimilar-to-biosimilar or multiple switching reported similar CRP concentrations between groups after switching (Table S3) [34,59,60,62–64,94,117]. A study by Hanzel et al. (2021) [34], in patients with IBD who underwent biosimilar-to-biosimilar (n = 80) or multiple switching (n = 69) showed significant declines (p = 0.038) in CRP concentrations (mg/L), but no significant increase (p = 0.582) in the percentage of patients with CRP <5 mg/L at 12 months [34]. A retrospective observational study by Khan et al. (2021) reported that in patients (n = 271) switching from oIFX to SB2 (single or double) 19.2% (n = 52) did not achieve remission at 12 months after switching [43]. Furthermore, compared with a single switch, double switching was not associated with worse outcomes [43]. Another prospective study (Lovero et al., 2021) reported no change in clinical remission rate from 6 months before to 3 months after switching (58% at both time-points) from CT-P13 to SB2 in patients with IBD (n = 36); the rate of mild activity varied from 28% to 33% (p = 0.68) [62]. Mazza et al. (2021) (n = 52) also observed no difference in clinical remission or loss of clinical response after 5.5 months between single-switch (oIFX → CT-P13) and double-switch (oIFX → CT-P13 → SB2) cohorts [61,97].

3.5. Safety and immunogenicity outcomes: IFX originator-to-biosimilar

3.5.1. Randomized controlled studies (n = 2)

Two RCTs reported numbers and frequencies of treatment-emergent AEs (TEAEs) among patients with IBD (Table 3) [55,66–68]. In the PLANET-CD study, 21/55 (38%) patients with CD who switched from oIFX to CT-P13 experienced TEAEs after switching, while 14/54 (26%) patients who continued oIFX experienced TEAEs [55]. The greatest number of TEAEs experienced by the highest percentage of patients with CD was observed in the NOR-SWITCH study: 130 AEs were reported in 57/77 (74%) patients after 52 weeks among patients who switched from oIFX to biosimilar at randomization, whereas 18/62 (29%) patients who switched at the start of the OLE experienced AEs during the 26-week extension [68]. CT-P13 maintenance was associated with 42 TEAEs among 24/65 (37%) patients during the OLE [66–68]. Among patients with UC, 65 events were reported in 29/46 (63%) patients who switched from oIFX to CT-P13 at randomization during the 52-week main NOR-SWITCH study and 20 events in 14/38 (37%) patients who switched to CT-P13 at the start of the 26-week OLE, corresponding to non-inferiority of biosimilar over oIFX in the extension phase [67].

Table 3. Summary of TEAEs in RCTs, and overall AEs and SAEs in longitudinal studies in patients with IBD switching from IFX originator (oIFX) to biosimilar.

Author, yearCountry	Reference	Population or subpopulation	Study design	Sample size	Follow-up	Transition	TEAEs	SAEs
Jørgensen, 2017; Goll, 2019; Jørgensen, 2020 Norway	[66–68]	CD	RCT	N = 77	12 months	oIFX → CT-P13	Number of events: 130 Patients: n (%), 57 (74)	Number of events: 9 Patients: n (%), 8 (10)
			OLE of RCT	N = 62	6 months	oIFX → CT-P13	Number of events: 28 Patients: n (%), 18 (29)	Number of events: 4 Patients: n (%), 3 (5)
		UC	RCT	N = 46	12 months	oIFX → CT-P13	Number of events: 65 Patients: n (%), 29 (63)	Number of events: 3 Patients: n (%), 3 (7)
			OLE of RCT	N = 38	6 months	oIFX → CT-P13	Number of events: 20 Patients n (%) 14 (37)	Number of events: 2 Patients: n (%), 2 (5)
Ye, 2019 (PLANET-CD) Multinational	[55]	CD	RCT	N = 55	Absolute: 12.5 months	oIFX → CT-P13	Patients: n (%), 21 (38)	TEAEs: n (%), 2 (4)
Armuzzi, 2019; Fiorino, 2017 Italy	[69,70]	IBD	Observational prospective	N = 155	Mean (SD): 12.9 (7.6) months	oIFX → CT-P13	NR	Total SAEs: n (%), 18 (11.6)
Bergqvist, 2018 Sweden	[76]	IBD	Observational prospective	N = 313	Absolute: 12 months	oIFX → CT-P13	NR	Number of events (%), 7 (2.2)
		CD	Observational prospective	N = 195	Absolute: 12 months	oIFX → CT-P13	Patients: n (%), 47 (24.1)	Patients: n (%), 4 (2.1)
		UC	Observational prospective	N = 118	Absolute: 12 months	oIFX → CT-P13	Patients: n (%), 25 (21.2)	n (%), 3 (2.5)
Binkhorst, 2018 Netherlands	[32]	IBD	Observational retrospective	N = 197	3.7 months	oIFX → CT-P13	Adverse reaction in patients: n (%) after switch, 12 (6.1)	NR
Bronswijk, 2020 Belgium	[79]	IBD	Observational prospective	N = 361	Absolute: 6 months	oIFX → CT-P13	n/N (%), 8/361 (2.2)	NR
Chaparro, 2019 Spain	[81]	IBD	Observational retrospective	N = 199	Median (IQR): 18 (10–22) months	oIFX → CT-P13	During follow-up: n (%), 48 (10.08) Number of events: 61	NR
Cheon, 2021; South Korea, EU	[56]	IBD	Observational retrospective	N = 63	Absolute: 60 months	oIFX → CT-P13	n (%), 42 (66.7)	NR
Eberl, 2017 Finland	[82]	IBD	Observational prospective	N = 62	3.7 months	oIFX → CT-P13	Number of events: 4	NR
Fischer, 2021 Germany	[83]	IBD	Observational prospective	N = 144	Absolute: 18.5 months	oIFX → SB2	n/N (%), 40/144 (27.8)	SAE, number of events: 11 (breast and prostate carcinoma, neuroendocrine malignancy, liver abscess, bronchopulmonary infection, intestinal surgical procedures)
Guerra Veloz, 2019; Argüelles-Arias, 2017, Argüelles-Arias, 2017 Spain	[73,74,86]	IBD	Observational prospective	N = 100	Absolute: 24 months	oIFX → CT-P13	n (%), 14 (14)	n (%), 14 (14)
Guerra Veloz, 2018 Spain	[115]	IBD	Observational prospective	N = 167	Absolute: 12 months	oIFX → CT-P13	n (%), 12 (7.2)	Number of events: 3 (3 AEs were considered as SAEs)
Guerra Veloz, 2018 Spain	[114]	IBD	Observational prospective, Observational retrospective	N = 98	Absolute: 24 months	oIFX → CT-P13	AE: n (%), 11 (11.2)	NR
		CD	Observational	N = 67	Absolute: 12 months	oIFX → CT-P13	NR	Patients: n (%), 1 (1.49)
		UC	Observational prospective, Observational retrospective	N = 31	Absolute: 24 months	oIFX → CT-P13	NR	n (%), 1 (3.25)
Guerrero Puente, 2017 Spain	[87]	IBD	Observational retrospective	N = 36	Mean (SD): 8.4 (3.5) months	oIFX → CT-P13	n (%), 4 (11.11)	NR
		CD	Observational retrospective	N = 23	NR	oIFX → CT-P13	Patients: n (%), 3 (8.3)	NR
Guiotto, 2019 Italy	[108]	IBD	Observational retrospective	N = 66	Absolute: 12 months	oIFX → CT-P13	n/N (%), 7/66 (10.60)	NR
Hoivik, 2018; Buer, 2017 Norway	[80,89]	IBD	Observational prospective	N = 143	Absolute: 18 months	oIFX → CT-P13	n (%), 74 (51.7) Number of events: 113	n/N (%), 5/113 (4.4)
		CD	Observational prospective	N = 99	Absolute: 18 months	oIFX → CT-P13	Patients: n (%), 14 (14.1)	NR
		UC	Observational prospective	N = 44	Absolute: 18 months	oIFX → CT-P13	n (%), 3 (6.81)	NR
Hlavaty, 2016 Slovakia	[88]	CD	Observational retrospective	N = 10	12.92 months	oIFX → CT-P13 maintenance	Patients: n (%), 1 (10)	NR
Kim, 2019 South Korea	[49]	IBD	Observational retrospective	N = 101	Absolute: 60 months	oIFX → CT-P13	n (%), 13 (12.87)	NR
		CD	Observational retrospective	N = 78	Absolute: 60 months	oIFX → CT-P13	Patients: n (%), 12 (15.4)	NR
		UC	Observational retrospective	N = 23	Absolute: 60 months	oIFX → CT-P13	n (%), 1 (4.3)	NR
Kolar, 2017 Czech Republic	[92]	IBD	Observational prospective	N = 74	Absolute: 12.92 months	oIFX → CT-P13	n (%), 35 (47.29)	NR
Macaluso, 2021; Macaluso, 2020 Italy	[60,94]	IBD	Observational prospective	N = 17	NR	oIFX → SB2	NR	Number of events: 3 (IRs and infections) occurred in three patients n (%), 3 (17.6)
Mahmmod, 2021; Mahmmod, 2020 Netherlands	[35,40]	IBD	Observational retrospective	N = 684	Absolute: 12 months	oIFX → CT-P13	NR	IBD-like symptoms: n (%), 176 (25.7), Dermatological symptoms, n (%): 150 (21.9)
Massimi, 2021 Italy	[96]	IBD	Observational prospective	N = 85	Mean (range):first follow-up after switch,4.4 (1.8–66.1) months;Second follow-up after switch, 10.8 (6.2–13.5) months	oIFX → SB2	First follow-up visit, number of events: 4 Second follow-up visit, number of events: 5	NR
Mazza, 2022; Mazza, 2020 Italy	[61,97]	IBD	Observational retrospective	N = 66	Median (range):37 (3–58)	oIFX → CT-P13	Overall AE: N = 66; n (%), 8 (12.1); p = 0.772	NR
Nakagawa, 2019 Japan	[50]	IBD	Observational prospective	N = 219	Absolute: 24 months	oIFX → CT-P13	All ADRs: n (%), 24 (11.0)	NR
		IBD	Observational prospective	N = 38	Absolute: 24 months	oIFX → CT-P13	All ADRs: n (%), 13 (34.2)	NR
Petitdidier, 2019 France	[99]	IBD	Observational prospective	N = 113	Absolute: 12.5 months	oIFX → CT-P13	n (%), 20 (18.0)	n (%), 2 (1.8)
Privitera, 2020 NR	[101]	IBD	Observational prospective	N = 96	Absolute: 12 months	oIFX → CT-P13	n (%), 38 (39.6)	NR
Pugliese, 2021 Italy	[102]	IBD	Observational prospective	N = 119	Median (range): 69.6 (9.6–189.6) months	oIFX → CT-P13	n (%), 24 (20.16)	NR
Ratnakumaran, 2018 UK	[103]	IBD	Observational prospective	N = 191	Absolute: 12 months	oIFX → CT-P13	12 months post switch: n (%), 9 (4.7) 12-month non-switch group: n (%), 0 (0); p = 1.0	NR
		CD	Observational prospective	N = 44	Absolute: 12 months	oIFX → CT-P13	12 months post switch: n (%), 2 (4.5)	NR
Razanskaite, 2017 UK	[104]	IBD	Observational	N = 143	NR	oIFX → CT-P13	Number of events (after 3rd dose of CT-P13): 88; total number of patients reporting AEs: 94 Number of events (before switch): 105; total number of patients reporting AEs: 94	NR
Schmitz, 2018 Netherlands	[36]	IBD	Observational prospective	N = 133	12 months	oIFX → CT-P13	n (%), 13 (9.8)	NR
Sieczkowska, 2016 Poland	[105]	CD	Observational prospective	N = 32	11 months Mean (SD) (range) follow-up: 8 (2.6) (2–11) months	oIFX → CT-P13	AEs during therapy with IFX biosimilar: n = 20	NR
		UC	Observational prospective	N = 7	Mean (SD): 5 (3.6) months	oIFX → CT-P13	AEs during therapy with IFX biosimilar: patients, n = 3	NR
Smits, 2019; Smits, 2017; Smits, 2016 Netherlands	[37–39]	IBD	Observational prospective	N = 83	Absolute: 24 months	oIFX → CT-P13	24 months: n (%), 19 (22.8) Number of events: 22	NR
Strik, 2018 Netherlands and Belgium	[106]	IBD	Observational prospective	N = 118	Absolute: 3.7 months	oIFX → CT-P13	Patients with ≥ 1 AE: n (%), 94 (80)	Patients with ≥ 1 SAE: n (%), 5 (4)
		CD	Observational prospective	N = 60	Absolute: 16 weeks (AEs), 3.7 months (SAEs)	oIFX → CT-P13	Patients with ≥ 1 AE: n (%), 46 (77) Number of events: 132	Patients with ≥ 1 SAE: n (%), 3 (5) Number of SAEs: 6 Patients with ≥ 1 SAE: n (%), 3 (5) SAEs, number of events: 3
		UC	Observational prospective	N = 58	Absolute: 3.7 months	oIFX → CT-P13	Patients with ≥ 1 AE: n (%), 48 (83) Number of events: 118	Patients with ≥ 1 SAE: n (%), 3 (5) Number of SAEs: 2 Patients with ≥ 1 SAE: n (%), 3 (5); number of events: 3
Trystram, 2021 France	[59]	IBD	Observational prospective	N = 158	Absolute: 12.5 months	oIFX → CT-P13	n (%), 63 (39.9) Number of events: 77	n (%), 10 (6.3)
Tursi, 2019; Tursi, 2017 Italy	[71,72]	CD	Observational retrospective, Observational prospective	N = 45	Median (IQR): 24 (6–24) months Absolute: 12 months	oIFX → CT-P13	Patients: n (%), 1 (0.7)	NR
Yoo, 2020 US	[48]	IBD	Observational retrospective	N = 55	Average: 13.9 months	oIFX → CT-P13	Drug-related AEs: n (%), 11 (20)	NR
Van Hoeve, 2019 Belgium	[107]	IBD	Observational prospective in pediatric patients	N = 42	Absolute: 6 months	oIFX → CT-P13	AEs may have occurred more than once in a single patient 6 months prior to switch: total number of events, 65 6 months after switch: total number of events: 45; p = 0.317	Hospitalization (due to disease exacerbation): n (%), 1 (2.38)
Kang, 2015 South Korea	[52]	UC	Observational retrospective	N = 4	Median: 9.6 months	oIFX → CT-P13	AE (arthralgia): n/N (%), 1/4 (25)	NR

Abbreviations: ADR, adverse drug reaction; AE, adverse event; CD, Crohn’s disease; CT-P13, infliximab biosimilar; IBD, inflammatory bowel disease; IFX, infliximab; IQR, interquartile range; IR, infusion reaction; n, sample with outcome; N, total sample size; NR, not reported; oIFX, originator infliximab; OLE, open-label extension; RCT, randomized controlled trial; SAE, serious adverse event; SB2, infliximab-abda biosimilar; SD, standard deviation; TEAE, treatment-emergent adverse event; UC, ulcerative colitis.

3.5.2. Longitudinal studies (n = 60)

AEs were reported in 29 longitudinal studies of patients with IBD who switched from oIFX to biosimilars, ranging from 3.7 to 60 months’ duration (Table 3) [32,36–39,49–51,56,57,59,61,79–83,86,87,89,92,96,97,99,102–104,107,108,110,111,114]. Nine [49,71,72,76,80,87,89,103,105,106,118] and five [49,76,80,89,106] longitudinal studies reported overall AEs for patients with CD and UC, respectively (Table 3). Eleven longitudinal studies reported the proportions of patients with IBD who experienced overall AEs (Figure 3a).

Figure 3. Percentage of patients with (a) AEs and (b) SAEs in patients with IBD after switching from oIFX to biosimilar in longitudinal studies.

AE, adverse event; IBD, inflammatory bowel disease; oIFX, originator infliximab; SAE, serious adverse event.

In a retrospective study by Chaparro et al. (2019) that included 48 (10.8%) patients who experienced a total of 61 AEs during follow-up, the incidence of AEs was 6% versus 13% in switch versus non-switch cohort (p<0.05) [81]. A prospective study by Strik et al. (2018) found that 94/118 (80%) patients with IBD experienced ≥1 AE after switching to CT-P13 after ≥30 weeks of continuous treatment with oIFX [106].

Serious AEs (SAEs) after switching to biosimilars were reported in 12 longitudinal studies [35,40,60,69,70,73,74,76,80,83,86,89,94,99,106,107,114] of patients with IBD (Figure 3b); three studies [76,106,115] each reported separate results for SAEs among patients with CD or UC (Table 3). A prospective study (Fischer et al., 2021) of patients with IBD (n = 144) who switched from oIFX to SB2 reported 11 SAEs in 40 patients (four malignancies, five intestinal surgical procedures, one liver abscess and one bronchopulmonary infection) [83]. Another prospective study (Macaluso et al., 2021) reported three SAEs in 3/17 (17.6%) patients who switched from oIFX to SB2, representing an incidence rate of 18.9 per 100 patient-years’ exposure [60]. A multinational prospective study (Strik et al., 2018) reported ≥ 1 SAE in 6/118 (5%) patients with IBD (CD: 3/60 [5%]; UC: 3/58 [5%]) who switched from oIFX to CT-P13 after 16 weeks of follow-up [106]. Two observational studies of patients with IBD who switched from oIFX to CT-P13 reported SAEs in 4/195 (2.1%) [76] and 1/67 (1.5%) [115] patients with CD, and in 3/118 (2.5%) [76] and 1/31 (3.2%) [115] patients with UC, respectively, after 12 months of follow-up.

Across studies, there were no relevant changes observed in the profile of AEs, SAEs and TEAEs after switching, with moderate numbers of patients reporting certain specific AEs. No treatment-related AEs of grade 4 or higher, malignancies or deaths were reported in any group. The rates of AEs following switching to biosimilar were consistent with the known AE profile of IFX across most studies. However, one study conducted in US veterans showed that patients who switched from oIFX to biosimilar IFX were more likely to discontinue treatment and switch to another originator biologic (468/610 [76.6%]) than patients who remained on oIFX (118/617; 19.2%) [93]. Among the biosimilar recipients, most switched back to oIFX (426/468 [91.0%]). The reasons for discontinuation and switching were unknown.

Only one study (Boone et al., 2018) reported nocebo effects upon switching from oIFX to biosimilars in an IBD population [33]. In that study, nocebo-effect responses, defined as an unexplained unfavorable therapeutic effect subsequent to a non-medical switch from oIFX to biosimilar IFX, with regaining of the beneficial effects after re-initiating oIFX, were reported in 13/101 (12.9%) patients with IBD. Patient education strategies to inform shared decision-making and patient empowerment will likely be needed to lessen the nocebo effect.

3.6. Safety and immunogenicity outcomes: IFX biosimilar-to-originator, biosimilar-to-biosimilar, biosimilar maintenance, or multiple switching

3.6.1. Randomized controlled studies (n = 2)

In the PLANET-CD study, one patient with CD who remained on CT-P13 at week 30 and reverse switched from CT-P13 to oIFX reported SAEs [55]. In the NOR-SWITCH RCT, seven SAEs were reported in six (9%) patients with CD on CT-P13 maintenance [66–68].

3.6.2. Longitudinal studies (n = 3)

Three longitudinal studies of patients with IBD [34,61,62,97], assessed overall AE profiles following biosimilar-to-biosimilar [34] or multiple switching [34,61,97]. One prospective observational study (Hanzel et al. 2021) reported eight (10%) and two (3%) patients who developed AEs after switching from CT-P13 to SB2 (n = 80) or multiple switching (n = 69), respectively [34]. Mazza et al. (2022) reported five TEAEs (all grade ≤ 3; no IRs) in 4/52 double-switched (oIFX → CT-P13 → SB2) patients with IBD after 24 weeks of follow-up, and no differences in safety outcomes compared with a single-switch group of 66 patients with IBD [61].

Only one longitudinal study, reported in two publications (Macaluso et al. 2020 and 2021) [60,94], evaluated SAEs after biosimilar-to-biosimilar (CT-P13 → SB2) or multiple (oIFX → CT-P13 → SB2) switching. Authors reported 11 SAEs in 11/43 (25.6%) patients after 28 patient-months follow-up, and four SAEs in 4/24 (16.7%) patients in the multiple-switch group (oIFX → CT-P13 → SB2) after 16 patient-months follow-up in the biosimilar-to-biosimilar and multiple-switch groups, respectively [60,94].

3.7. Feasibility assessment

83 publications (63 full-text publications and 20 conference abstracts) were considered for qualitative feasibility assessment (see Supplementary file 4 for full details of the feasibility analysis).

4. Discussion

Assessment of the effects of switching between IFX products and multiple switching on safety and effectiveness outcomes can enable HCPs to develop a wider understanding of these therapies when considering switching in patients with IBD.

Of the 85 studies included in this systematic literature review, 65 were longitudinal observational studies; only two were RCTs (one multinational, one conducted in Norway). For most of the identified studies, the clinical, patient-reported and safety outcomes in patients with IBD switching from originator to biosimilar were highly similar to the originator responses, and outcomes following switching were consistent with the known profile of oIFX [32,39,42,55,58,60,75,76,82,88,102]. Of particular note, four studies, including one RCT, reported mucosal healing in patients with CD after switching from oIFX to biosimilar [45,46,55,69–72], and the RCT also identified mucosal healing in patients after reverse switching from IFX biosimilar to oIFX [55]. Interpretation of these findings is difficult because mucosal healing is time-dependent and the conversion of patients from active inflammation to healing status may be reflective of sustained drug exposure, rather than resulting from differences in agent. Across the studies, switching to biosimilar did not have a significant impact on the rates of AEs, SAEs or TEAEs or the proportion of patients with detectable ADAs. In general, switching to biosimilars was associated with stability in AEs, corresponding to non-inferiority of IFX biosimilars over oIFX among patients with IBD. In two studies (NOR-SWITCH and PLANET-CD), the percentage of TEAEs was higher in patients who switched from originator to biosimilar than in patients who continued with originator. However, this difference may be numerical (e.g. higher percentage of AEs in the oIFX group before switching, or higher rate of reporting AEs, both maintained following switching) rather than related to safety signals. Originator-to-biosimilar switching did not increase the percentage of patients with detectable levels of ADAs or increase the rate of discontinuation due to AEs. It is not certain whether therapeutic drug monitoring lessens immunogenicity when switching agents. Additionally, it is not known if the use of concurrent immunomodulators is needed to lessen immunogenicity prior to switching. This strategy may be appropriate in patients undergoing multiple switches.

A few longitudinal studies evaluated multiple switches. In these studies, comparison between single- and multiple-switch cohorts revealed non-inferiority of multiple switches over single switch [61,97]. Only one study demonstrated a significant improvement in clinical remission among a single-switch cohort versus multiple-switch and biosimilar-switch cohorts [60,94].

Of the 85 IBD studies included in this systematic literature review, only two RCTs, PLANET-CD [55] and NOR-SWITCH [66–68], assessed the efficacy and safety of IFX biosimilars. Additional well-designed RCTs with clearly defined follow-up periods and outcomes would be useful to provide conclusive evidence regarding the efficacy and safety of biosimilars in this population. In addition, detailed information about the type of endoscopy used is limited, and few studies have reported on hypersensitivity, mucosal healing and fistula closure outcomes in patients with IBD. Only a small number of studies have reported results from multiple switching (n = 5), biosimilar-to-biosimilar switching (n = 6), and reverse switching (n = 4) at this time. Therefore, a comparative analysis of all switching patterns was not possible. While outside the scope of this systematic literature review, RCTs of multiple switching between originator and biosimilar adalimumab have been conducted in patients with psoriasis [119–121] and rheumatoid arthritis [122, 123]. In those studies, multiple switches between reference and biosimilar did not negatively affect pharmacokinetics, immunogenicity, effectiveness, or safety [119,120]. Finally, the follow-up duration for the studies included in this systematic literature review ranged from 3 months to 3 years for most studies, with only three studies reporting a longer follow-up duration (of 5 years). Additional studies with longer follow-up duration may be helpful in guiding decisions regarding maintenance therapy.

The key strength of this review is its use of a standardized, thorough, and transparent approach to identify and appraise the included studies. This review was conducted according to PRISMA guidelines and included a comprehensive literature search in addition to independent selection and quality assessment of the included studies by two researchers. Limitations of this review include those inherent in any systematic review of the published literature, including the potential for missing relevant studies not identified in the systematic literature database searches, studies not published in English or studies published after the date the searches were conducted (4 April 2022). The inherent heterogeneity in study populations, evaluation methods, length of follow-up and outcomes assessed is important to consider when interpreting the findings.

5. Conclusions

This update improves understanding of the safety and effectiveness of biosimilars based on recent publications, which is important to inform the optimal treatment of patients with IBD switching from oIFX to IFX biosimilar for financial or other reasons. This updated systematic review builds on a prior (2019) review of IFX switching in patients with various inflammatory disorders [27]. Notably, this update includes more robust data that have been published in the interim, including the PLANET-CD study and additional analyses from the NOR-SWITCH study [55,66,67]. By performing this systematic review we have identified important areas of interest that have not yet been adequately evaluated.

Evidence from the recent literature, particularly a number of longitudinal real-world studies, indicates that switching from oIFX to biosimilars is associated with similar effectiveness, safety and tolerability. Thus, results continue to support IFX biosimilar use as an alternative to originator for patients with IBD. Over the long term, switching from originator to biosimilars did not impact PROs, quality of life, or immune response in patients with IBD.

Despite the wealth of recent data on biosimilar use in patients with IBD, our systematic literature review highlights some evidence gaps. In particular, high-quality RCTs with longer follow-up periods and clearly defined clinical outcomes using formal reporting procedures for endoscopy outcomes, including mucosal healing and fistula closure, are lacking.

Article highlights

• Despite the growing availability of biosimilars and increasing data on their highly similar clinical profiles (to their respective originators), many healthcare providers still remain skeptical regarding the safety of using biosimilars instead of the originator in bio-naïve patients, as well as switching patients from an originator to a biosimilar.

• A number of longitudinal real-world data indicates that switching from originator IFX to biosimilars is associated with similar effectiveness, safety, and tolerability which continues to support IFX biosimilar use as an alternative to originator for patients with IBD.

• Over the long term, switching from originator to biosimilar IFX had no observed impact on patient-reported outcomes, quality of life or immune response in patients with IBD.

• There is a lack of high-quality randomized controlled trials evaluating biosimilar use in patients with IBD with longer follow-up periods and clearly defined clinical outcomes using formal reporting procedures for endoscopy outcomes, including mucosal healing and fistula closure.

Declaration of interest

GR Lichtenstein has acted as a consultant for AbbVie, American Regent, Cellgene/Bristol Myers Squibb, Eli Lilly, Endo International, Ferring, Fresenius Kabi (Health and Wellness Partners), Gilead, Janssen (now Johnson & Johnson), MedEd Consultants, Pharmacosmos, Pfizer, Prometheus Laboratories, Sandoz, Takeda Pharmaceutical Company, and UCB. Performed CME supported by AbbVie, Allergan, American Gastroenterological Association, American Regent, Annenberg Center for Health Sciences, Boehringer Ingelheim, Bristol Myers Squibb, Eli Lilly, Exact Sciences, Ferring, IBD Horizons, Ironwood, Janssen (now Johnson & Johnson), MedEd Consultants, Pennsylvania Society of Gastroenterology, Sanofi, Salix, Seres Pharmaceuticals, Takeda, Vindico. Other finances received: Eli Lilly for DSMB (Honorarium for Data Safety Monitoring Board), Gastro-Hep Communications (Honorarium for executive editor), Janssen Orthobiotech (Finances to Fund GI IBD Fellowship), Pfizer (Finances to Fund GI IBD Fellowship), Professional Communications, Inc- textbook royalty, Springer Science and Business Media- Editor (Honorarium), Up To Date- Walters Kluwer- Author (Honorarium). Research: Takeda, UCB, Bristol Myers Squibb, Janssen (now Johnson & Johnson). BG Feagan has acted as a consultant for AbbVie, AbolerIS Pharma, Agomab Therapeutics, AllianThera Biopharma, Amgen, AnaptysBio, Applied Molecular Transport Inc, Arena Pharmaceuticals, Avoro Capital Advisors, Atomwise, BioJamp, Biora Therapeutics, Boehringer Ingelheim, Boxer Capital, Celsius Therapeutics, Celgene/Bristol Myers Squibb, Connect Biopharma, Cytoki Pharma, Disc Medicine, Duality Biologics, EcoR1, Eli Lilly, Equillium, Ermium Therapeutics, First Wave BioPharma, FirstWord Pharma, Galapagos NV, GalenAtlantica, Genentech/Roche, Gilead Sciences, Gossamer Bio, GSK, Hinge Bio, HotSpot Therapeutics, InDex Pharmaceuticals, Imhotex, Immunic Therapeutics, JAKAcademy, Janssen (now Johnson & Johnson), Japan Tobacco Inc., Kaleido Biosciences, Landos Biopharma, Leadiant Biosciences, L.E.K. Consulting, Lenczner Slaght, LifeSci Capital, Lument AB, Millennium Pharmaceuticals, MiroBio, Morgan Lewis, Morphic Therapeutic, Mylan, OM Pharma, Origo BioPharma, Orphagen Pharmaceuticals, Pandion Therapeutics, Pendopharm, Pfizer, Prometheus Therapeutics and Diagnostics, Play to Know AG, Progenity Pharmaceuticals/Biora Therapeutics, Protagonist Therapeutics, PTM Therapeutics, Q32 Bio, Rebiotix, REDX Pharma, Roche, Sandoz, Sanofi, Seres Therapeutics, Surrozen, Takeda Pharmaceutical Company, Teva Pharmaceuticals, Thelium Therapeutics, TiGenix, Tillotts Pharma UK, Ventyx Biosciences, VHSquared Ltd., Viatris, Ysios Capital, Ysopia Bioscience, and Zealand Pharma. Speaker for AbbVie, Janssen, and Takeda Pharmaceutical Company. S Singh is an employee of Curo, a division of the Envision Pharma Group, which was a paid consultant to Pfizer in connection with the development of this manuscript. A Soonasra and M Latymer are employees of Pfizer and hold stock or stock options in Pfizer. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

Medical writing support was provided by Sharmila Blows, PhD, of Envision Pharma Group and was funded by Pfizer. The authors thank Karen Smoyer, PhD, of Curo, a division of the Envision Pharma Group, for contributing to the study design, analyzing the data, and organizational contributions to this work, which was funded by Pfizer.

Data availability statement

All data generated or analyzed during this study are included in this published article or as supplementary information files.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14712598.2024.2378090

Additional information

Funding

This study was funded by Pfizer.