Analysis of clinical trials of biosimilar infliximab (CT-P13) and comparison against historical clinical studies with the infliximab reference medicinal product

Abstract

Objective: To examine whether efficacy, safety and pharmacokinetic (PK) data observed with CT-P13 (Remsima^®; Inflectra^®), an infliximab biosimilar, are similar to those from published reports with the reference medicinal product (RMP; Remicade^®) in patients with rheumatoid arthritis (RA) or ankylosing spondylitis (AS). Methods: Literature searches were performed to identify clinical studies with infliximab RMP. Efficacy, safety and PK data were indirectly compared with data from head-to-head clinical trials of CT-P13 and RMP. Results: CT-P13 and RMP produce similar efficacy in patients with RA and AS when compared across clinical studies. There are no substantial differences in the incidence of infusion-related reactions, infections, serious infections, malignancy or lymphoma. PK data in patients with RA are similar in direct comparisons and comparisons with historical data. Conclusion: Efficacy, safety and PK data are highly comparable between CT-P13 and RMP, both in head-to-head clinical studies, and indirect comparisons with historical clinical data for RMP.

Keywords:

• ankylosing spondylitis
• biosimilar
• CT-P13
• efficacy
• infliximab
• literature search
• pharmacokinetics
• rheumatoid arthritis
• safety

Rheumatoid arthritis (RA) is an autoimmune disease of unknown etiology affecting approximately 0.5–1% of the general population. It is characterized by a chronic inflammatory reaction in the synovium of joints leading to destruction of cartilage and erosion of juxta-articular bone [1–4]. A variety of leukocytes including macrophages, T cells, B cells, plasma cells and mesenchymal cells, including synovial fibroblasts and their cytokines, contribute to the pathogenesis of RA [1]. The cytokines known to be associated with the development of RA include TNF, IL-1, IL-6, IL-7, IL-15, IL-17A, IL-17F, IL-18, IL-21, IL-23, IL-32, IL-33 and granulocyte-macrophage colony-stimulating factor [2]. Among these, TNF has been consistently shown to be one of the cytokines most directly implicated in the disease processes involved in RA [5]. TNF is produced by various cells in the body including those in the synovial membrane and synovial fluid of patients with RA, and particularly within the cartilage-pannus junction of joints [6,7]. Furthermore, experiments in murine models of RA demonstrate that blockade of TNF is associated with anti-inflammatory and protective effects on joints [8].

It is estimated that ankylosing spondylitis (AS) affects approximately 0.1–1.4% of the population globally with inflammation of the axial skeleton leading to back pain and ankyloses [9,10]. The cause of AS has not yet been fully elucidated; however, the two cardinal features of AS are inflammation and new bone formation [11,12]. Infiltration of macrophages, T cells and increased levels of some cytokines have been observed in the synovial portion of the sacroiliac joints of patients with active AS [13]. TNF appears to play an important role in the range of inflammatory cytokines investigated in AS pathogenesis [14]. Studies in patients with AS have found high TNF messenger RNA expression near the site of new bone formation, as well as increased serum levels of TNF [13,15]. It has also been reported that mice overexpressing TNF show axial ankylosis and enthesopathies resembling human AS [16].

Given the involvement of TNF in the pathogenesis of both RA and AS, anti-TNF biologics have been investigated and developed for the treatment of these disorders [17]. Infliximab (a chimeric mAb) was the first anti-TNF biologic introduced, and has been in clinical use for the treatment of RA, AS and other inflammatory diseases since the late 1990s. Since then, other anti-TNF biologic drugs have become available, including adalimumab (a human mAb), etanercept (a fusion protein of recombinant TNF receptor and Fc region of Ig), golimumab (a human mAb) and certolizumab pegol (a PEGylated Fab′ fragment from humanized mAb) [17].

Infliximab was initially developed by Centocor (now Janssen Biotech, Inc.) and marketed by Janssen Biotech, Inc. under the brand name Remicade^® (hereafter referred to as the reference medicinal product [RMP]). In 2013, CT-P13 (Remsima^®, Inflectra^®), the first biosimilar infliximab, was approved by the EMA in all the indications granted to the RMP. Approval was based on an extensive and comprehensive comparative program, including two pivotal randomized controlled trials (RCTs) for RA and AS that involved a total of 856 patients (PLANETRA [Programme evaLuating the Autoimmune disease iNvEstigational drug cT-p13 in RA patients] and PLANETAS [Programme evaLuating the Autoimmune disease iNvEstigational drug cT-p13 in AS patients]) [18–20]. These two studies demonstrated equivalence in efficacy and pharmacokinetics (PK) between CT-P13 and RMP and confirmed that the two agents are highly comparable in terms of safety.

Numerous RCTs have also been conducted with RMP and the outcomes reported in the literature [21–34]. Consequently, there is also a need to compare the data observed with CT-P13 and RMP in PLANETRA and PLANETAS indirectly to the published data obtained with infliximab RMP from other RCTs. The objective of this article is, therefore, to examine whether clinical outcomes (efficacy, safety and PK) observed with CT-P13 in PLANETRA and PLANETAS are similar to the historical data from RCTs with RMP identified through literature searches.

Methods

Literature searches

Literature searches were performed to identify articles that reported the findings from clinical studies with infliximab RMP (with a cutoff for the publication date at the end of 2014). The databases searched were MEDLINE, EMBASE, Cochrane Central Register of Clinical Trials, DERWENT DRUG FILE, BIOSIS PREVIEWS and SCISEARCH. The searches focused primarily on RCTs testing the same posology used in PLANETRA and PLANETAS, but open-label and observational studies were also considered, where relevant.

Selection of studies for analysis

In addition to PLANETRA and PLANETAS, the literature searches identified eight studies with RMP in patients with RA and six in patients with AS [21–34].

Selection of studies for efficacy analysis

Because disease baseline characteristics and concomitant treatment may have an impact on efficacy parameters in patients with RA and AS, data from studies with different inclusion criteria from PLANETRA and PLANETAS, or those that allowed concomitant methotrexate (MTX) treatment in studies of AS, were excluded from the efficacy analysis (but included in the safety analysis) to aid appropriate comparisons across studies.

Selection of studies & data for RA efficacy analysis

Since the patient population studied in PLANETRA did not include MTX naïve and early RA, two of the eight studies identified were excluded from the efficacy analysis because one included MTX-naïve patients only and the other included newly diagnosed patients with less than 1 year of disease duration [27,28]. Additionally, because the dosing schedules in PLANETRA were 3 mg/kg at weeks 0, 2, 6 and then every 8 weeks, data from study arms with different dosages or schedules were excluded from the efficacy comparison for RCTs of RMP.

Across the studies in RA, the American College of Rheumatology (ACR) response was the most common efficacy index to be investigated [19,21–26]. Consequently, the ACR responses observed reported from the studies were compared in the RA efficacy analysis.

Selection of studies & data for AS efficacy analysis

From six RCTs with RMP in AS, two were excluded due to co-treatment with MTX and considerably different inclusion criteria to PLANETAS, respectively [33,34].

For the studies in patients with AS, the most common efficacy indices to be reported were 20% improvement in response according to Assessment in AS International Working Group criteria (ASAS20), Bath AS Disease Activity Index (BASDAI) and Bath AS Functional Index (BASFI). Therefore, these efficacy assessments are compared further for CT-P13 and RMP in the analysis presented here.

Selection of studies for safety analysis

Data from all studies identified for RMP in RA and AS were analyzed for treatment-emergent adverse events (TEAEs) of special interest. These included infusion-related reactions, all infections, all serious infections, malignancy and lymphoma. Incidence rates were directly compared (up to week 54) between CT-P13 and RMP in PLANETRA and PLANETAS [19,20,35,36] and indirectly to the rates seen with RMP in other relevant published studies [21–34].

Selection of studies for PK analysis

There was only one published report identified from RCTs of RMP that included PK data [21]. This involved patients with RA and so the data were compared with PK data obtained from PLANETRA.

Results

Efficacy analysis

Efficacy analysis for RA studies

A total of 1353 patients with active RA treated with CT-P13 or RMP were included in the efficacy analysis (from the PLANETRA study and five other published RCTs) (Table 1). Although there are inherent limitations associated with comparing data across clinical studies, it is important to note that in this case the inclusion criteria, MTX dose, infliximab dosing schedules and baseline disease characteristics (such as number of tender and swollen joints, and C-reactive protein levels) were comparable between PLANETRA and the five studies with RMP included in the analysis.

Table 1. Comparison of design and baseline characteristics of clinical studies using RMP or CT-P13 for RA (for efficacy comparison).

References/study		PLANETRA [19]		ATTRACT [21,22]	START [23]	Zhang et al. [24]	Abe et al. [25]	ATTEST [26]
Inclusion criteria		≥6 swollen and ≥6 tender joints and plus two of: morning stiffness ≥45 min, CRP >2 mg/dl, ESR >28 mm/h despite MTX therapy for 3 months			6 swollen and 6 tender joints despite MTX therapy for 3 months	≥3 swollen and ≥8 tender joints and plus two of: morning stiffness ≥45 min, CRP >1.5x upper limit of normal, ESR >28 mm/h despite MTX therapy for 3 months	≥6 swollen and ≥6 tender joints and plus two of: morning stiffness ≥45 min, CRP >2 mg/dl, ESR >28 mm/h despite MTX therapy for 3 months	>10 swollen joints, >12 tender joints, and CRP levels >1 mg/dl (upper limit of the normal range, 0.5), inadequate response to MTX
Number of patients		302	304	86	360	87	49	165
Arms		3 mg/kg CT-P13	3 mg/kg RMP
MTX, mean ± SD or med (IQR)		15.6 ± 3.1 mg/week	15.6 ± 3.2 mg/week	16 ± 4 mg/week	15.0 (10, 18) mg/week	N/A	7.1 ± 1.9 mg/week	16.3 ± 3.6 mg/week
Dosing		At week 0, 2, 6 and q8w (up to week 30)		At week 0, 2, 6 and q8w (up to week 30)	At week 0, 2, 6, 14, 22	At week 0, 2, 6, 14	At week 0, 2, 6	On days 1, 15, 43, 85 and every 56 days thereafter
Primary end point		ACR20 (week 30)		ACR20 (week 30)	A serious infection (week 22)	N/A	ACR20 (week 14)	DAS28 (ESR) at 6 months
Baseline characteristics, mean ± SD or med (IQR)	Tender joints	25.6 ± 13.9 22.0 (N/A)	24.0 ± 12.9 21.0 (N/A)	32 (16, 46)	22 (15, 31)	N/A	19.0 ± 11.8	31.7 ± 14.5
	Swollen joints	16.2 ± 8.7 15.0 (N/A)	15.2 ± 8.3 13.0 (N/A)	19 (13, 30)	15 (11, 21)	N/A	15.1 ± 9.0	20.3 ± 8.0
	CRP (mg/dl)	1.9 ± 2.5 1.1 (N/A)	1.9 ± 2.2 1.1 (N/A)	3.1 (1.3, 5.3)	1.6 (1, 3)	N/A	4.2 ± 3.1	3.3 ± 3.2
	ESR (mm/h)	46.6 ± 22.4	48.5 ± 22.6	49 ± 23	N/A	N/A	N/A	47.8 ± 30.4
	Rheumatoid factor positive (%)	74.5	72.7	84	82.8	N/A	N/A	84.8

ACR: American College of Rheumatology; CRP: C-reactive protein; DAS28: Disease activity score in the 28 joints examined; ESR: Erythrocyte sedimentation rate; IQR: Interquartile range; MTX: Methotrexate; N/A: Not available; q8w: Every 8 weeks; RA: Rheumatoid arthritis; RMP: Reference medicinal product; SD: Standard deviation.

The ACR responses reported from the studies are shown in Table 2. Overall, the responses seen for ACR20, ACR50 and ACR70 with CT-P13 in combination with MTX were within the range of responses seen with RMP plus MTX. With regards to ACR20, after 14–30 weeks of RMP treatment combined with MTX, reported responses ranged from approximately 42 to 75.9% across the published studies. During the same time period, ACR20 responses with CT-P13 ranged from 60.9 to 63.6% in PLANETRA. Between weeks 52 and 54, ACR20 response with RMP ranged from 42 to 55.8% in the relevant published studies. The ACR20 with CT-P13 at week 54 was 57% (Table 2).

Table 2. ACR response rate in RA patients: comparison of PLANETRA^† and published randomized controlled studies of RMP.

Study	Week	CT-P13	RMP	Analysis population	Ref.
ACR20 (%)
PLANETRA	54	57	52	ITT
ATTRACT	54		42	N/A	[22]
ATTEST	52		55.8	ITT	[26]
PLANETRA	30	60.9	58.9	ITT
ATTRACT	30		∼50	ITT	[21]
ATTEST	28		59.4	ITT	[26]
START	22		58	ITT	[23]
ATTRACT	22		∼53	ITT	[21]
Zhang et al.	18		75.9	N/A	[24]
ATTRACT	18		∼45	ITT	[21]
PLANETRA	14	63.6	58.2	ITT
ATTRACT	14		∼42	ITT	[21]
Abe et al.	14		61.2	N/A	[25]
ACR50 (%)
PLANETRA	54	33.1	31.6	ITT
ATTRACT	54		21	N/A	[22]
ATTEST	52		36.4	ITT	[26]
PLANETRA	30	35.4	33.9	ITT
ATTRACT	30		27	ITT	[21]
ATTEST	28		37	ITT	[26]
START	22		32.1	ITT	[23]
Zhang et al.	18		43.7	N/A	[24]
PLANETRA	14	33.1	29.9	ITT
Abe et al.	14		30.6	N/A	[25]
ACR70 (%)
PLANETRA	54	16.2	15.2	ITT
ATTRACT	54		10	N/A	[22]
ATTEST	52		20.6	ITT	[26]
PLANETRA	30	16.6	15.5	ITT
ATTRACT	30		8	ITT	[21]
ATTEST	28		24.2	ITT	[26]
START	22		14	ITT	[23]
Zhang et al.	18		22.99	N/A	[24]
PLANETRA	14	13.9	12.2	ITT
Abe et al.	14		10.2	N/A	[25]

^†PLANETRA values for week 54 are unpublished data.

ACR: American College of Rheumatology; ITT: Intention to treat; N/A: Not available; RA: Rheumatoid arthritis; RMP: Reference medicinal product.

In terms of ACR50, from week 14 to 30, the response ranged from 27 to 43.7% with RMP and, in PLANETRA, from 33.1 to 35.4% with CT-P13. Between weeks 52 and 54, the ACR50 response with RMP ranged from 21 to 36.4% in published studies. At week 54 in PLANETRA, the ACR50 response with CT-P13 was 33.1% (Table 2). Similar findings were seen with RMP and CT-P13 on ACR70 response.

Efficacy analysis for AS studies

A total of 635 patients with AS treated with CT-P13 or RMP were included in the efficacy analysis (from the PLANETAS study and four other published RCTs) (Table 3). Inclusion criteria, baseline characteristics, dosages and dosing schedules were all comparable for PLANETAS and the RCTs reporting data for RMP in AS.

Table 3. Comparison of design and baseline characteristics of clinical studies using RMP or CT-P13 for AS (for efficacy comparison).

References/study		PLANETAS [20]		Braun et al. [29]	ASSERT [30]	Giardina et al. [31]	Breban et al. [32]
Inclusion criteria		BASDAI score of ≥4 and a visual analog scale score (VAS) for spinal pain of ≥4					BASDAI ≥3 axial pain (second item on the BASDAI) ≥3
Number of patients		125	125	35	201	25	124
Arms		5 mg/kg CT-P13	5 mg/kg RMP
Dosing		At week 0, 2, 6 and q8w (up to week 30)		At week 0, 2, 6	At week 0, 2, 6 and q6w (up to week 18)	At week 0, 2, 6 and q6w (up to week 102)	At weeks 4, 6, 10 and q6w (up to week 52)
Primary end point		AUC, Cmax,ss		BASDAI at week 12	ASAS20 at week 24	ASAS20 at week 12, 24, 36, 48, 96, 104	ASAS20 at 58 weeks
Baseline characteristics (mean ± SD) or (median [IQR])	BASDAI	6.7 ± 1.4	6.6 ± 1.6	6.5	6.6 (5.3–7.6)	6.5 ± 1.2	6.2 ± 1.5
	BASFI	6.2 ± 1.9	6.2 ± 2.2	5.5	5.7 (4.5–7.1)	6.1 ± 0.9	5.4 ± 2

AS: Ankylosing spondylitis; ASAS: Assessment of SpondyloArthritis International Society; AUC: Area under the plasma concentration–time curve; BASDAI: Bath ankylosing spondylitis disease activity index; BASFI: Bath ankylosing spondylitis functional index; C_max,ss: Maximum steady state plasma drug concentration during dosage interval; IQR: Interquartile range; q6w: Every 6 weeks; q8w: Every 8 weeks; RMP: Reference medicinal product; SD: Standard deviation.

The ASAS20, BASDAI and BASFI scores observed with CT-P13 were comparable to the range of scores observed with RMP in relevant published studies (Table 4). Over the entire study periods, ASAS20 response rate ranged from 62.8 to 70.5% with CT-P13 and 64.8 to 75% with RMP. BASDAI scores ranged from 3.7 to 3.9 with CT-P13 and 3.3 to 4.8 with RMP, while BASFI scores ranged from 3.4 to 3.8 and 3.0 to 4.0, respectively.

Table 4. Clinical response in AS patients: comparison of PLANETAS^† and published randomized controlled studies of RMP.

Study	Week	CT-P13	RMP	Analysis population	Ref.
ASAS20 (%)
Breban et al.	58		75	ITT	[32]
PLANETAS	54	67	69.4	ITT
Giardina et al.	48		∼74	N/A	[31]
Giardina et al.	36		∼72	N/A	[31]
PLANETAS	30	70.5	72.4	ITT
ASSERT	24		74.5	ITT	[30]
Giardina et al.	24		∼73	N/A	[31]
PLANETAS	14	62.8	64.8	ITT
Braun et al.	12		∼68	ITT	[29]
Giardina et al.	12		75	N/A	[31]
BASDAI
Breban et al.	58		3.3	ITT	[32]
PLANETAS	54	3.8	3.7	ITT
PLANETAS	30	3.7	3.8	ITT
ASSERT	24		3.7 (median)	ITT	[30]
PLANETAS	14	3.9	3.8	ITT
Braun et al.	12		3.3	ITT	[29]
Giardina et al.	12		4.8	N/A	[31]
BASFI
Breban et al.	58		3.0	ITT	[32]
PLANETAS	54	3.4	3.5	ITT
PLANETAS	30	3.7	3.6	ITT
ASSERT	24		4.0 (median)	ITT	[30]
PLANETAS	14	3.8	3.7	ITT
Braun et al.	12		3.4	ITT	[29]
Giardina et al.	12		3.5	N/A	[31]

Values for BASDAI and BASFI are mean unless otherwise indicated.

^†PLANETAS values for 54 week data are unpublished data.

AS: Ankylosing spondylitis; ASAS: Assessment of SpondyloArthritis International Society; BASDAI: Bath ankylosing spondylitis disease activity index; BASFI: Bath ankylosing spondylitis functional index; ITT: Intention to treat; N/A: Not available; RMP: Reference medicinal product.

Safety analysis

The rates of TEAEs of special interest seen with CT-P13 (in PLANETRA and PLANETAS) and with RMP (in all identified studies) are shown in Table 5. Infections and infusion-related reactions were the TEAEs occurring most frequently with both CT-P13 and RMP in direct comparisons and in historical studies. Moreover, the incidence of TEAEs of special interest was generally comparable between CT-P13 and RMP both in PLANETRA and PLANETAS and in other studies.

Table 5. Incidence rate of treatment-emergent adverse events of special interest with CT-P13 and RMP.

		PLANETRA/PLANETAS
		Incidence rate, n/N (%)		Incidence rate, 100 PY		Historical data of RMP
		CT-P13	RMP	CT-P13	RMP	Incidence		Incidence rate, 100 PY
						%	Range	100 PY	Range
All infections	RA	127/302 (42.05%)	137/300 (45.67%)	39.97	43.62	61.92	(52.81, 68.48)	76.39	(68.48, 91.79)
	AS	55/128 (42.97%)	49/122 (40.16%)	38.76	36.04	48.15	(35.29, 64.17)	74.70	(57.69, 153.36)
All serious infections	RA	13/302 (4.3%)	7/300 (2.33%)	4.09	2.23	4.14	(1.67, 8.48)	5.42	(2.19, 8.48)
	AS	2/128 (1.56%)	3/122 (2.46%)	1.41	2.21	1.88	(0.99, 2.94)	2.56	(2.15, 12.78)
Malignancy and lymphoma	RA	1/302 (0.33%)	4/300 (1.33%)	0.31	1.27	0.66	(0.00, 1,21)	0.53	(0.00, 1.32)
	AS	1/128 (0.78%)	0/122	0.70	0.00	0.76	(0.00, 0.83)	0.36	(0.00, 0.75)
Infusion-related reaction	RA	30/302 (9.93%)	43/300 (14.33%)	9.44	13.69	27.34	(10.00, 67.35)	26.02	(9.66, 97.55)
	AS	11/128 (8.59%)	15/122 (12.3%)	7.75	11.03	10.84	(0.00, 15.00)	16.37	(0.00, 23.66)

AS: Ankylosing spondylitis; PY: Patient year; RA: Rheumatoid arthritis; RMP: Reference medicinal product.

Pharmacokinetic data

The age, gender and use of concomitant MTX were similar between the relevant studies included in the PK analysis (Table 6) [19,21]. Comparisons of PK data generally support the similarity of CT-P13 and RMP in patients with RA. According to the historical study, the mean concentrations of RMP at the week-30 time point (pre-infusion) were 1.5 ± 1.6 and 8.9 ± 8.1 μg/ml with 3 and 10 mg/kg RMP, respectively (Figure 1). In PLANETRA, the mean concentrations of infliximab at the week-30 time point (pre-infusion) were 3.27 ± 13.5 and 2.48 ± 8.3 μg/ml with CT-P13 and RMP (3 mg/kg doses), respectively.

Table 6. Comparison of populations in studies analyzed for pharmacokinetic data [19,21].

	PLANETRA [19]		ATTRACT [21]
	CT-P13	RMP	RMP
MTX Mean (SD)	15.6 (3.1) mg/week	15.6 (3.2) mg/week	16 (4) mg/week
Age Median (range)	50 (18–75)	50 (21–74)	56 (25–74)
White (%)	72.8	73	93
Female (%)	81.1	84.2	81

MTX: Methotrexate; RMP: Reference medicinal product; SD: Standard deviation.

Figure 1. Mean serum concentration of infliximab. Indicated amount of CT-P13 or RMP were infused at week 0, 2, 6, 14, 22 and the concentration of infliximab was measured before infusion at week 30 [19,21].

Discussion

TNF plays a pivotal role in the pathogenesis of rheumatic diseases, including RA and AS, and the introduction of biologic TNF antagonists in the late 1990s led to significantly improved clinical manifestations and outcomes for patients with rheumatic diseases. However, biologic drugs are expensive because they are complex in structure and difficult to produce consistently. High costs of biologic drugs have led to unequal access to biologic agents according to several factors including a country’s gross domestic product, regulation of the use of biologic agents and healthcare budgets [37]. However, patent expiry of some biologics, including infliximab has opened an opportunity for the development of biosimilar drugs that are designed to be comparable to their RMP, with lower acquisition cost. Regulatory agencies around the world have adopted broadly comparable definitions of a biosimilar drug, and the approval of these agents requires manufacturers to meet strict non-clinical and clinical development guidelines to ensure high similarity to the RMP in terms of efficacy and safety [38].

Recently, CT-P13 – a biosimilar to infliximab RMP – was approved by the EMA based on evidence from a comprehensive program that directly compared CT-P13 and RMP in terms of efficacy, safety and PK, including in two RCTs (PLANETRA and PLANETAS). These two studies showed that efficacy, safety, immunogenicity and PK of CT-P13 and RMP are highly similar.

In this article, we add to the understanding of the similarity of CT-P13 and RMP when used in the treatment of rheumatic diseases. We analyzed data from a total of nine clinical studies in patients with RA and seven in patients with AS to compare the data observed with CT-P13 and RMP in PLANETRA and PLANETAS with historical data with RMP. When comparing efficacy, we focused on studies reporting data for similar patient cohorts and using the same dose, comparable administration schedule and concomitant therapies as in the studies involving CT-P13. Efficacy data were also compared using the most commonly used efficacy assessments in the relevant trials for RA and AS. Considering the totality of the evidence, the results of our analysis suggest that CT-P13 and RMP produce similar efficacy in the RA and AS patient populations, even when comparing across different clinical studies. This is an important confirmation of the findings of another indirect meta-analysis in RA that reported no significant difference between CT-P13 and other biological agents indicated for RA in terms of clinical efficacy and safety [39].

Biosimilar versions of RMPs are anticipated to become important options in established treatment algorithms [40–42]. However, if use of biosimilars is to become more widespread, it is critical to establish that the safety profile of these agents is comparable with what is documented for established original biologics. This is especially important for adverse events of special interest, including the incidence of infusion reactions, infections and malignancies [43,44]. Our analysis of safety compared data across a total of 16 studies in RA and AS, and demonstrated that CT-P13 and RMP are comparable in terms of the incidence of infusion-related reactions, infections, serious infections, malignancy and lymphoma. This similarity in safety profiles held both for head-to-head comparisons in RCTs, and for indirect comparisons with published historical studies.

The PLANETRA and PLANETAS studies have already directly demonstrated that the PK profiles of CT-P13 and RMP are comparable in patients with RA and AS [19,20]. The literature search identified only one historical RCT that reported PK data for RMP in patients with RA [21]. Comparison of data from this historical study with data from PLANETRA generally support the comparability of PK between CT-P13 and RMP in RA. Only small numerical differences in mean concentrations were noted between the two studies. These differences could theoretically have resulted from differences in patient age, gender, ethnicity, BMI, doses of concomitant MTX treatment and batches of RMP, or the quantification methods used in the specific studies reported. However, it is unlikely that age, gender and MTX treatment are causes of this numerical difference since they are comparable between two RCTs. On the other hand, quantification methods used for each RCT may have contributed to the numerical differences, as the conventional ELISA was used in the historical study with RMP but the automated flow-through immunoassay for PLANETRA [19,21].

There are limitations involved in comparing data between clinical trials. Historical studies are all RCTs with placebo arms, whereas the PLANETRA/AS studies are head-to-head trials without placebo arms. Therefore, the actual ACR response rate in the PLANETRA study was a little bit higher than most historical studies, possibly due to the lack of a placebo arm, although the clinical impact of placebo on outcome measures in RCTs is difficult to estimate.

Taken together, clinical outcomes such as efficacy, safety and PK are highly comparable between CT-P13 and RMP, not only in head-to-head comparison RCTs sponsored by the biosimilar developer, but also in indirect comparisons with historical clinical data published for RMP. These findings further emphasize the high degree of clinical similarity between CT-P13 and RMP.

Additional clinical trials, registries and pharmacovigilance studies with CT-P13 are ongoing around the globe and will provide more evidence to guide clinical decision making. In the meantime, indirect comparisons with historical data such as reported here provide important support for the comparability of CT-P13 and RMP.

Key issues

• Since their introduction to clinical practice in the 1990s, biologics have led to better disease management and improved outcomes in patients with rheumatoid arthritis (RA) or ankylosing spondylitis, but they are expensive.

• Patent expiry of infliximab recently led to the development of a biosimilar version (CT-P13) that is expected to be available at a lower cost than the reference medicinal product (RMP), thereby enabling more widespread clinical use.

• CT-P13 was approved by the EMA based on evidence from a comprehensive comparability program, which included two randomized clinical trials (PLANETRA and PLANETAS). These two studies demonstrated equivalence in efficacy and pharmacokinetics between CT-P13 and the RMP.

• Additional confirmation from other analyses would provide increased understanding of the clinical similarity of the two treatment options, so the current analysis indirectly compared outcomes from head-to-head trials with CT-P13 and RMP with those from published studies with RMP.

• This shows that CT-P13 and RMP produce similar efficacy in patients with RA and ankylosing spondylitis when compared across clinical studies. There are no substantial differences in the incidence of TEAEs of special interest, including infusion-related reactions, infections, serious infections, malignancy and lymphoma. Pharmacokinetic data in patients with RA are similar for direct comparisons and comparisons with published data.

• The findings demonstrate that clinical outcomes are highly comparable between CT-P13 and RMP, not only in head-to-head comparison RCTs sponsored by the biosimilar developer, but also in indirect comparisons with historical clinical data of RMP.

• This analysis further emphasizes the high degree of clinical similarity between CT-P13 and RMP.

Acknowledgements

Editorial support (writing assistance, assembling tables and figures, collating author comments, grammatical editing and referencing) was provided by Ryan Woodrow and Mark O’Connor (Aspire Scientific Limited, Bollington, UK) and was funded by Celltrion Healthcare Co., Ltd (Incheon, Republic of Korea).

Financial & competing interests disclosure

DH Yoo received research grants (not related to CT-P13) from Celltrion and is a consultant of Celltrion. W Park received research grants and consultancy fees from Celltrion. C Oh and S Hong are full-time employees of Celltrion. 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.

Notes