Many of the synthetic and manufacturing processes involve proprietary techniques to produce a biosimilar of the reference product. Specific manufacturing features are summarized in Table 15.1. A large number of variables exist, among other features, in the choice of the cell vector and cell expression system and cell line and master cell banks. Likewise, conditions for expansion of the cell lines are proprietary, and variables which may be highly controlled are not known to the biosimilar manufacturer and may play into intellectual property concerns after drug approval (Fig.15.1). Differences in synthesis can potentially result in different posttranslational modifications, possiblyaffecting efficacy, safety, and immunogenicity of the product [8]. For both the biosimilars for IFX and ADA, the FDA closely considered whether differences in posttranslational afucosylation may yield to differences in pharmacokinetics and determined that no meaningful differences were present. Differences also can emerge in the design and construction of cell production bioreactors, filtration, and chromatographic purification steps. Notably, theFDA does not require an approach to “independently establish the efficacy and safety of the biosimilar,” but rather “a demonstration of the biosimilarity between the proposed product and a reference product.” Figure 15.2 depicts the relative weights it places on the different steps in determining biosimilarity.

As a first step, structural analyses are required that the proposed product will encode for the same primary amino acid sequence as the reference product as well as an analysis ofsecondary, tertiary, and quaternary structures. A detailed physiochemical analysis of CT-P13 has been carried out to the reference infliximab. Detailed biochemical techniques used to compare the two are beyond the scope of this chapter and are reviewed in detail elsewhere [9]. Higher-order structures were found to be indistinguishable by multiple assays between the reference infliximab andCT-P13. Importantly,comparable biologic activity of CT-P13 and the reference drug was demonstrated based on its mechanism of action, including invitro TNF neutralization activity, TNF-binding affinity based on ELISA, and cell-based TNF-binding affinity (Fig.15.3a, b). Bridging assays for reference USinfliximab and reference European Union (EU) infliximab demonstrated similar findings between the two products [10].

The pharmacokinetic (PK) properties of CT-P13 were studied in a double-blind, three-arm, parallel-group study of the biosimilar CT-P13 and two formulations of Remicade^® using healthy subjects receiving a single infusion dosed at 5 mg/kg of the biosimilar (n = 70), Remicade^® from Europe (n = 71) or Remicade^® from the USA (n = 70). All three formulations were essentially equivalent in terms of maximal infliximab concentration (C _max), area under the serum concentration (AUC) time curves, and no differences in treatment-emergent adverse events among the 211 study subjects [11].

The PLANETAS trial was a phase 1, double-blind, multicenter study of 250 anti-TNF naïve patients with active ankylosing spondylitis (AS) randomized to receive CT-P13 (n = 125) or Remicade^® (n = 125) dosed at 5 mg/kg at weeks 0, 2, and 6 then every 8 weeks up to 30weeks [5]. The AS patient population was deemed the immune-mediated inflammatory disorder closest in approximation to healthy volunteers in order to study pharmacokinetics and medication-related safety and efficacy as the goal was to identify differences primarily related to the treatment, not due to disease state [12]. Of note, AS patients with inflammatory or rheumatic diseases were excluded from the study presumably including coexisting IBD diagnoses, present in approximately 5–10% of AS patients [13].Steady-state PK data, based on AUC and C _max values, trough levels, and medication half-life were essentially equivalent for CT-P13 and Remicade^®-treated patients at all measured timepoint post-infusions. Clinical response rates at weeks 14 and 30 were 63% and 71% for CT-P13 versus 65% and 72% for Remicade^®, with similar changes in baseline activity scores and quality-of-life scores at weeks 14 and 30 [5]. In the PLANETAS study, anti-drug antibodies (ADA) occurred in 9% and 27% of CT-P13-treated patients comparable to 11% and 23% of Remicade^®-treated patients at weeks 14 and 30, respectively, with the presence of ADA negatively influencing the PK of both agents [14].Treatment-emergent adverse event rates at week 30 were 65% for CT-P13 versus 64% for Remicade^®, including infusion reactions [5]. Partial remission rates, adverse events, and pharmacokinetic profiles for CT-P13 and Remicade^® (AUC and C _max) remained equivalent at week 54 [14].

In the subsequent open-label extension study, CT-P13 patients were allowed to either continue treatment with CT-P13 (n = 88), and Remicade^®-treated patients were switched to CT-P13 (n = 86) at week 54 and followed for an additional 48 weeks. Notable findings included similar partial remission rates at weeks 78 and 102 between CT-P13-treated patients who continued therapy (70% and 81%) and patients who switched from Remicade^® to CT-P13 at week 54 (77% and 77%). However, treatment-emergent adverse event rates were higher for the switch group (Remicade^® to CT-P13, 71%) compared to the CT-P13 patients with continued treatment (49%). ADA were present in 22% and 25% of continued CT-P13-treated patients at weeks 54 and 102, respectively, compared to 26% at week 54 and 31% at week 102 for the CT-P13-switched group [15].

The PLANETRA trial was a phase 3, randomized, double-blind, multicenter, parallel-group study of CT-P13 in rheumatoid arthritis patients with active disease despite treatment with ≥3 months of methotrexate dosed at 12.5–25 mg weekly [4]. In the PLANETRA trial, eligible RA patients were randomized to receive CT-P13 (n = 302) or Remicade^® (n = 304) dosed at 3 mg/kg at weeks 0, 2, and 6 then every 8 weeks with the primary endpoints assessed at week 30 with continued methotrexate administration. The primary aim of the PLANETRA trial was to demonstrate therapeutic equivalence between the two treatment groups defined as 95% confidence intervals (CI) of treatment response within a margin of ±15% at week 30. Week 30 response rates were similar for CT-P13 (60.9%)- and Remicade^® (58.6%)-treated patients with the 95% CI range of -6–10%, within the prespecified equivalence margin. Adverse event profiles (CT-P13 60.1%, Remicade^® 60.8%) at week 30 and PK data profiles (AUC and C _max values) measured after each infusion were also equivalent between the two treatment groups [4]. At week 30, 25.8% of CT-P13-treated patients and 25.4% of Remicade^®-treated patients developed anti-drug antibodies using ECL-based assays for ADA detection. Among patients continuing with the PLANETRA study to week 54, remission and response rates, PK profiles, and adverse event rates were again comparable between the two treatment groups. ADA positivity at week 54 was substantially higher than reported during the PLANETAS trials for AS with 52.3% of CT-P13-treated and 49.5% of Remicade^®-treated patients having antibodies present by week 54 with lower resultant response rates [16].

In the open-label extension study, beginning at week 54, PLANETRA study patients treated with CT-P13 could continue with scheduled 3 mg/kg dosing every 8 weeks (n = 158), or Remicade^®-treated patients could switch to CT-P13 at the same dosing and interval for an additional 48 weeks (n = 144). Clinical efficacy and adverse event rates were comparable betwee n the continued versus switched groups, with the proportions of CT-P13-treated patients with ADA also similar between the patients who continued CT-P13 (49.1% at week 54, 46.4% at week 102) and the patie nts who switched from Remicade^® to CT-P13 (49.3% at week 54, 49.6% at week 102) [17].

There are limited published data commenting on the safety, efficacy, and bioequivalence of CT-P13 for the inflammatory bowel diseases (IBD), Crohn’s disease (CD), and ulcerative colitis (UC), consi sting mostly of small retrospective studies performed in Korea, Poland, or Hungary (Table 15.2) [18]. One of the larger, prospective observational cohort studies using CT-P13 for IBD included 78 patients (46 CD/32 UC with 28% CD and 16% UC patients having had prior biologic exposure) who received CT-P13 5 mg/kg at weeks 0, 2, and 6, except for three severe UC patients who received either 10 mg/kg due to low albumin/high C-reactive protein or an extra infusion 5–7 days later. Clinical efficacy was assessed at week 14 including clinical activity scores, trough levels, and ADA positivity. At week 14, 79% of CD patients and 56% of UC patients achieved clinical remission with reductions in C-reactive protein and fecal calprotectin values compared to baseline and no unexpected adverse events. Eight patients had undetectable trough levels at week 14, and 7 of these 8 patients had detectable antibodies, but these patients were only treated with CT-P13 monotherapy [19]. Another recently published prospective, nationwide, observational cohort study from Hungary followed 210 IBD patients (126 CD and 84 UC) treated with CT-P13 induction at 5 mg/kg. Reported outcomes included week 14 response and remission rates: CD 81% and 54%, UC 78% and 59% with infusion reactions occurring among 7% of patients, and an adverse event rate of 17% [20].

However, there are no randomized controlled trial data currently available that are equivalent to the PLANETAS or PLANETRA studies to confirm the clinical efficacy, safety, and pharmacokinetic profiles of CT-P13 for the IBD patient population. The influence of IBD-specific diseasestate-related factors on therapeutic efficacy, safety, and pharmacokinetics when considering extrapolation of indications to include IBD patients remains unexplored based on the currently available data. A primary issue of concern pertains to the potential for immunogenicity with the biosimilar product when used interchangeably with the reference product. There are multiple factors influencing immunogenicity aside from just the biosimilar drug itself, including medication dosing, schedule, disease type and severity for which treatment is indicated, and the use of concomitant medications [21].

The two pivotal CT-P13 randomized controlled trials investigated biosimilar outcomes as monotherapy dosed as 5 mg/kg for AS and combination therapy dosed as 3 mg/kg for RA in conjunction with methotrexate 12.5–25 mg weekly. Of note, RA patients tend to have lower rates of anti-drug antibody development presumably due to the use of concomitant methotrexate compared to other disease states such as psoriasis and IBD, which tend to have higher rates of i mmunogenicity [22–25]. In accordance to recommendations by the World Health Organization (WHO) and the Food and Drug Administration (FDA), in order to extrapolate across indications with efficacy and safety data, immunogenicity risk should be studied in the highest risk patient population for therapy-related adverse events and anti-drug antibody potential [26, 27]. Thus, further investigation with respect to the clinical efficacy and pharmacokinetics of CT-P13 in the moderate to severe IBD patient population may be valuable to add insight to the true bioequivalence in this different immune-mediated disease state [28, 29].

TheNOR-SWITCH study is a randomized, double-blind, parallel-group study of 155 CD and 93 UC patients to evaluate the efficacy and safety of switching from Remicade^® to the biosimilar across several disease states including CD and UC [30]. After being in a sustained remission for at least 24 weeks on stable dosing of reference infliximab, patients were randomized to continuing the originator infliximab versus switching to CT-P13. The primary endpoint was disease worsening at week 52 defined as an increase in partial Mayo score of at least 3 points with a minimum score of 5 for UC patients and an increase in the Harvey-Bradshaw Index (HBI) score of at least 4 points with a minimum score of 7 for CD patients. Secondary study endpoints included safety and immunogenicity [30].

For CD patients continuing reference infliximab, 21% of patients experienced disease worsening, compared to 37% of patients switching to CT-P13 (95% CI −29.3%, −0.7%). For UC patients, 9% of patients continuing reference infliximab had disease worsening, compared to 12% of patients who switched to CT-P13 (95% CI−15.2%, −10.0%). There were no statistically significantdifferences for PK drug trough levelsbetween reference drug and CT-P13 for either UC or CD patients. Notably, anti-drug antibodies, adverse events, and serious adverse events were similar in both groups.While the strengths of the study included the RCT design, dosing according to standard protocols, and finance of the study by the Norway federal government, limitationswere thatthe study was not powered for non-inferiority within each diagnostic group [30].

There are also two prospective observational studies for CT-P13 in IBD: NCT02539368, CONNECT-IBD, a post-marketing obs ervational cohort of CT-P13 in clinical practice to assess safety, immunogenicity, sustained efficacy, and patient-reported outcomes sponsored by Hospira, and NCT02326155, another observational prospective cohort study of CT-P13 sponsored by Celltrion, the two currently available manufacturers of the infliximab biosimilar [31–33].

Extrapola tion refers to the approval of an approved biosimilar for a condition in which it was not clinically studied, and it is one of the most controversial issues regarding adoption of biosimilars in those countries for which it is approved. The possibility of gaining approval for extrapolation, as well as for interchangeability, is an important motivator for a biotechnology company to embark on a venture in which large financial risks are taken with an uncertain approval process, an unknown landscape of patent battles, and unknown physician, patient, and payer acceptance. Considerations taken by the FDA in granting approval for extrapolation for each disease state consider the “totality of the evidence.” This begins with the establishment of biosimilarity by analysis of primary, secondary, and tertiary structure, posttranslational profile, and in vitro functional characteristics to include TNF binding and neutralization as discussed in the earlier sections of this review. Clinical analysis of the data includes potential differences in mechanism of action (MOA), pharmacokinetic (PK), and pharmacodynamic (PD) data in different patient populations; immunogenicity in different disease states, which may be influenced by different concomitant immunosuppressive agents in the different diseases; and the potential for differences in expected toxicities in different patient populations [26]. Consideration of these same variables and data led to different extrapolation approvals in Canada, the European Union (EU), and the United States (USA).

However, there are a number of potential obstacles in extrapolating from RA and AS to inflammatory bowel disease. Factors which are different between these diseases include different dosages, differences in the use of methotrexate in RA, and the variable patterns of the use of various immunosuppressants in IBD which may affect drug levels, anti-drug antibodies, and resultant differences in clinical efficacy [34–36].

An additional criterion of the FDA scientific guidelines in considering the issue of extrapolation is to “consider whether the tested condition of use is the most sensitive in which to detect clinically meaningful differences and safety and effectiveness [8].” In the case of rh eumatoid arthritis, however, the PLANETRA study of CT-P13 versus infliximab trial was designed as an equivalence trial, and the 95% confidence interval for the treatment difference between CT-P13 and infliximab for the primary endpoint was −6–10%, falling within the range of the equivalence margin selected of −15–15%. These confidence intervals contained the smallest placebo-adjusted response to infliximab, 8%, previously demonstrated in any disease for which infliximab is indicated [4]. Rheumatoid arthritis may therefore be the least sensitive clinical model to detect a potential difference in efficacy between this biosimilar and infliximab in other indications [37]. Another challenge in extrapolating from RA to Crohn’s disease is the divergent efficacy for different anti-TNF agents, as well as other biologics in the two diseases, suggesting the possibility of different mechanisms of inflammatory pathways. For example, while anti-TNF agents are effective in both, anakinra, abatacept, and rituximab are effective in RA, but not in Crohn’s disease [38–41].

In Canada, Health Canada, the national drug regulatory agency, approved CT-P13 for rheumatoid arthritis, ankylosing spondylitis, psoriatic arthritis, and plaque psoriasis,based on the PLANETRA and PLANETAS trials [4, 5, 42]. Health Canada also relied on these trials to demonstrate similarity in pharmacokinetic (PK) parameters, using area under the concentration-time curve over the dosing interval (AUCtau) and maximum serum concentration (Cmax) at steady state.Based on their extrapolation criteria, Health Canada extended approval of CT-P13 for plaque psoriasis and psoriatic arthritis. However, Health Canada, at time of approval, denied extrapolation to adult and pediatric Crohn’s disease and UC. This denial was based on the observed differences in degree of afucosylation and FcγRIIIa receptor binding in addition to differences in some in vitro antibody-dependent cell-mediated cytotoxicity (ADCC) assays [42]. In addition, Health Canada observed that the safety profile in rheumatic diseases is different, specifically citing the risk of hepatosplenic T-cell lymphoma in IBD. In the absence of clinical studies in IBD, it was felt that extrapolation was not warranted for CD or UC in adults or pediatric patients [42].

On the other hand, the European Medicines Agency (EMA) and the FDA Arthritis Advisory Committee (AAC) concluded that extrapolation was warranted to all the diseases for which reference inflix imab had previously received EMAand FDA approval [10, 43]. The agencies reviewed the data supplied by Celltrion and concluded that the issue of diminished afucosylation and ADCC activity occurred only in the most sensitive experimental in vitromodel using NK cells of patients with high-affinity genotypes. In further examination of efficacy and safety of efficacy in IBD, Celltrion had committed to increase enrollment in a post-marketing surveillance study and plans to conduct an additional comparative trial of CT-P13 versus reference infliximab in active CD.

In summary, the totality of evidence analyzed and considered by the FDA AAC (Table 15.3) resulted in a vote of 21 to 3 in favor of extrapolation,based on the results of the RA and AS trials, to all the indications for which reference infliximab had been approved, including adult and pediatric UC and CD resulting in the FDA approval of CT-P13 for all previously approved indications for infliximab [44].Similarly,for ABP 501, theFDA granted extrapolation to all indicated diseases for which adalimumab had been previously approved.

In January 2017, the FDA issued theirguidance document regarding the critical issue of interchangeability , whereby an approved biosimilar can be substituted for a prescribed reference drug without the approval or even the knowledge of the prescribing physician or patient [45]. Beyond demonstrating biosimilarity, the sponsor the biosimilar may request a claim of “interchangeability.” According to FDA draft guidance in May 2015, the approval of interchangeability allows that the biosimilar “may be substituted for the reference product without the intervention of the prescribing healthcare provider” [46]. In theJanuary 2017 FDA guidance document, the FDA defined the weight of evidence to fulfill the “higher-level” requirement that an interchangeable product “can be expected to produce the same clinical result as the reference product, in any given patient and the risk in terms of safety or diminished efficacy of alternating or switching between the use of the biologic product and the reference product is not greater that the risk of using the reference product without such alternation or switch” [46]. This has become a particularly contentious issue that has been legislated in the USA on a state-by-state basis (Fig.15.4). More specifically, the FDA did not grant interchangeability of CT-P13 with reference infliximab when it approved CT-P13 in April 2016, nor did it grant interchangea bility for ABP 501 for adalimumab with the approval in September2016, pending additional studies demonstrating their interchangeability [44].