By Jonathan Kay
Arthritis & Rheumatology
Volume 68, Issue 5, pages 1049–1052, May 2016
Since the US Food and Drug Administration (FDA) first approved the anti–tumor necrosis factor (anti-TNF) monoclonal antibody infliximab in 1998 to treat patients with inflammatory bowel disease, targeted biologic therapies have revolutionized the treatment of inflammatory diseases. However, their high cost has limited access to these very effective agents for many patients. Thus, biosimilar agents are being developed so that these biopharmaceuticals can be made available at significantly lower prices. CT-P13, a biosimilar infliximab developed in South Korea by Celltrion, was the first biosimilar therapeutic monoclonal antibody approved by a regulatory agency, and it was approved by the US FDA on April 5, 2016. In this issue of Arthritis & Rheumatology, Kim and colleagues report on the use of CT-P13 in South Korea during the first 16 months after its introduction in November 2012 .
A biosimilar is a copy of a biopharmaceutical that no longer is protected by patent, which has the same primary amino acid sequence as its reference product, has successfully undergone rigorous analytical and clinical assessment in comparison to its reference product, and has been approved by a regulatory agency according to a specific pathway for biosimilar evaluation . It is important to differentiate a biosimilar from a biomimic (or “intended copy”), which is a replica of a biopharmaceutical that has not been developed, assessed, or approved according to regulatory guidelines for biosimilars. Similarity of a biomimic to the originator biopharmaceutical has not yet been demonstrated by a stepwise and comprehensive comparability exercise. Whereas the primary amino acid sequence of a biosimilar must be identical to that of its reference product, that of a biomimic may differ from the originator biopharmaceutical. A biomimic may also be dissimilar from the originator biopharmaceutical in formulation, dose, dosing regimen, efficacy, safety, and immunogenicity, any of which could result in a clinically significant difference in efficacy or toxicity in comparison to the original. Examples of biomimics are Reditux, a rituximab biomimic produced and marketed by Dr. Reddy’s Laboratories, Ltd. (Hyderabad, Telangana, India) since 2007 and also in Bolivia, Chile, Ecuador, and Peru, and Yisaipu, an etanercept biomimic manufactured and distributed by Shanghai CP Guojian Pharmaceutical Company, Ltd. (Shanghai, China) since 2005 and also under different brand names in Colombia, India, and Mexico. In contrast to biomimics, approved biosimilars have undergone rigorous comparison to their reference products, and have shown no clinically meaningful differences.
The process by which a biopharmaceutical is manufactured typically undergoes a number of modifications during its lifecycle, and these modifications are registered with the regulatory agencies. For example, the reference product for infliximab (Remicade) has undergone more than 35 such registered manufacturing changes since its initial regulatory approval . With such changes, subtle variations in posttranslational modification occur . As long as these changes fall within a prespecified quality range, this “drift” in composition is considered to be acceptable, and the biopharmaceutical can be marketed without additional clinical studies or any change in labeling. The subtle structural differences that exist between a biosimilar, which has been approved by a regulatory agency, and its reference product are typically of a similar magnitude. Thus, after having undergone extensive evaluation in direct comparison to its reference product, an approved biosimilar can be considered in the same way as another lot of the reference product. However, without being compared directly, 2 approved biosimilars of the same reference product that are produced by different manufacturers cannot necessarily be considered to be as similar to each other as each is to its reference product.
The regulatory pathway for approval of biosimilars in the US, Canada, the European Union, and most other countries is abbreviated, in contrast to that for originator biopharmaceuticals. Approval of a biosimilar can rely upon data generated from approval of the originator biopharmaceutical upon which it is based, after extensive comparative assessment demonstrating that the biosimilar is “highly similar” to its reference product. In 2005, the European Medicines Agency (EMA) was the first regulatory body to issue guidance regarding the evaluation of biosimilars. It proposed a stepwise approach to compare a biosimilar to its reference product. Extensive analyses comparing their structure and function serve as the foundation of this evaluation. Subsequently, their pharmacokinetic and pharmacodynamic properties are compared, ideally in healthy volunteers. Finally, at least one clinical trial is carried out in patients with a disease for which the reference product is licensed, in order to evaluate any potential clinically meaningful differences between the biosimilar and its reference product. The immunogenicity and safety profiles of the 2 drugs are assessed throughout all of the clinical studies.
In the US, the Biologics Price Competition and Innovation Act of 2009, which is a component of the Patient Protection and Affordable Care Act of 2010 (Public Law 111-148), amended Section 351 of the Public Health Service Act (42 U.S.C. 262) to establish a pathway by which a biosimilar can be evaluated in comparison to a single reference product. To be considered for this abbreviated biologic license application, the biosimilar must be “highly similar” in molecular composition to its reference product and have the same primary amino acid sequence, presumed mechanism of action, route of administration, dosage form, and potency as its reference biopharmaceutical. The biosimilar may only be reviewed and approved for those indications for which its reference product has been licensed. The US FDA follows an approach similar to that of the EMA, but evaluates the “totality-of-the-evidence” when comparing a biosimilar to its reference product, considering all aspects of the comparative evaluation together and with equal importance .
The first biosimilar approved in the US using this new Section 351(k)–defined pathway was filgrastim-sndz (Zarxio) on March 6, 2015 . Filgrastim-sndz is a granulocyte colony-stimulating factor (G-CSF) manufactured by Sandoz. In seeking approval for its biosimilar filgrastim, Sandoz conducted extensive analytical testing, which demonstrated that this biosimilar had an identical primary amino acid sequence and was highly similar in its biologic activity and other structural aspects to multiple lots of the originator filgrastim (Neupogen), produced by Amgen both in the US and in the European Union. Rodent toxicology studies yielded similar results for both the biosimilar and the originator filgrastim. Single- and multiple-dose pharmacokinetic and pharmacodynamic studies conducted in healthy subjects, using the absolute neutrophil count as the pharmacodynamic marker, supported the demonstration of biosimilarity . In addition, a single randomized, double-blind trial comparing filgrastim-sndz to the originator filgrastim, conducted in 218 patients receiving chemotherapy for breast cancer, demonstrated equivalence in the number of consecutive days with absolute neutropenia and a similar safety profile between the 2 agents . Neither the biosimilar nor the originator filgrastim was immunogenic, as no subjects in any of the clinical studies developed antidrug antibodies. Based upon this “totality-of-the-evidence,” the US FDA granted filgrastim-sndz marketing approval for all 5 indications for which the originator filgrastim is licensed in the US.
Developing biosimilars of complex proteins, such as immunoglobulins, presents additional challenges. Whereas G-CSF is a relatively small, single-chain peptide of 175 amino acids and is not glycosylated when produced in Escherichia coli, monoclonal IgG antibodies are large, glycosylated proteins (with a molecular weight of 150 kd) that consist of 4 subunit chains. The biosimilar infliximab, CT-P13, is highly similar to originator infliximab in its physicochemical characteristics . In the phase I PLANETAS study, 250 patients with ankylosing spondylitis were treated with multiple doses of either originator or biosimilar infliximab at a dose of 5 mg/kg intravenously as monotherapy [10, 11]. The pharmacokinetics of CT-P13 were equivalent to those of originator infliximab, since the 90% confidence intervals of the ratios of geometric means of the pharmacokinetic parameters at steady state (area under the concentration–time curve and observed maximum serum concentration) for CT-P13 and originator infliximab between weeks 22 and 30 both fell within the prespecified margins of ±20%. The efficacy and safety of CT-P13 and originator infliximab were comparable, in that the proportions of treatment responders according to the Assessment of SpondyloArthritis international Society criteria for 20% and 40% improvement at weeks 14, 30, and 54 were similar, and also the proportions of subjects who experienced adverse events, infections, or infusion reactions were similar.
In the phase III PLANETRA study, 606 patients with rheumatoid arthritis whose disease remained active despite methotrexate therapy were treated with multiple doses of either originator or biosimilar infliximab at 3 mg/kg intravenously in combination with methotrexate [12, 13]. The efficacy of CT-P13 and that of originator infliximab were equivalent at the primary end point of 30 weeks: the difference between the proportions of responders, according to the American College of Rheumatology 20% improvement criteria , was 2.3%, with 95% confidence intervals that fell within the prespecified margin of ±15%. Other efficacy parameters were also comparable between the 2 treatment arms at weeks 14, 30, and 54 and there was little difference between the proportions of subjects who experienced adverse events, infections, or infusion reactions through week 54. Based on these data, CT-P13 was granted marketing approval by regulatory agencies as the first biosimilar therapeutic monoclonal antibody, with initial approval granted by the South Korean Ministry of Food & Drug Safety on July 23, 2012, and subsequent approval by the EMA, the Colombian Instituto Nacional de Vigilancia de Medicamentos y Alimentos, Health Canada, the Japanese Pharmaceuticals Medical Devices Agency, the Turkish Ministry of Health, and the Australian Therapeutic Goods Administration.
Using claims data, Kim and colleagues found that use of infliximab, both the originator and the biosimilar versions, increased slightly after the biosimilar became available in South Korea . In contrast, use of the other marketed TNF inhibitors decreased after the biosimilar infliximab was marketed. As might be expected, given the caution with which physicians might approach a new type of medication, the biosimilar was prescribed for only a small proportion of infliximab-treated patients during the first several months after it became available. However, by the end of this 16-month period, biosimilar infliximab accounted for nearly one-fifth of all infliximab prescriptions in South Korea. Had the price of originator infliximab not been reduced by 30% immediately after the biosimilar became available in South Korea, a price differential between the biosimilar and the originator might have resulted in a larger market share for the lower-priced biosimilar.
Unlike a new medication with a novel mechanism of action, formulation, or dosing regimen, a biosimilar has no attribute other than price to distinguish it from its reference product. In fact, the ideal biosimilar would be identical to its reference product. Thus, a biosimilar does not fill an unmet therapeutic need other than to provide the biopharmaceutical at a more affordable cost. At the time when a biosimilar is introduced to market, the only patients who might be waiting for the biosimilar to fulfill an unmet medical need would be those for whom the originator biopharmaceutical is inaccessible because of cost or restricted market availability, unless they had been treated successfully with an alternative product. Thus, in developed countries in which most medications are widely available, the number of prescriptions for a biosimilar would be expected to be lowest at the time of its release. As health care professionals and patients become more comfortable using the biosimilar, as if it were another lot of its reference product, the number of prescriptions for the biosimilar will increase over time.
Because a biosimilar does not differ from its reference product in any clinically meaningful way, regulatory agencies allow the biosimilar to be approved for indications in which the reference product has been studied and for which the reference product has been licensed, without requiring additional clinical trials of the biosimilar in each indication. This extrapolation of indications is based on a robust package of analytical, in vitro, pharmacokinetic, pharmacodynamic, toxicology, and immunogenicity studies, with at least one clinical trial comparing the biosimilar to its reference product conducted as a “bioassay” in patients with a disease for which the reference product is approved. Although the biosimilar is not studied formally in each of the diseases for which approval is sought, the extensive historical clinical experience with the reference product in each of its licensed indications should serve as predictive evidence of the similar efficacy of the biosimilar. Most regulatory agencies have approved CT-P13 with extrapolation to all indications for which the originator infliximab has been approved. To date, only Health Canada has questioned the validity of extrapolating indications to inflammatory bowel diseases, based on observed differences between CT-P13 and the originator infliximab in the amount of afucosylation, in Fcγ receptor IIIa binding, and in an in vitro assay of antibody-dependent cell-mediated cytotoxicity .
The potential substitution of a biosimilar for its reference product is a controversial issue in the US. According to the Biologics Price Competition and Innovation Act of 2009, if a biosimilar was determined to be “interchangeable,” a pharmacist would be allowed to substitute the biosimilar for a prescribed biologic therapy without involving the prescribing health care provider. Such an interchange might take place on more than one occasion. For example, if an insurer were to mandate that a less expensive version of a biopharmaceutical be used, a patient might be dispensed a biosimilar instead of the reference product that was prescribed by his or her physician. Since an insurer might change which version of a biopharmaceutical is “preferred” over time, it is plausible that a patient might be transitioned from a reference product to its biosimilar, and then to a different biosimilar, and then, perhaps, to one of the biopharmaceuticals that had been dispensed previously. The US FDA has not yet specified what data will be required for a biosimilar to be designated as being interchangeable. However, interchangeability is an issue that is largely limited to the US. Neither the EMA nor Health Canada makes such a designation; instead, interchangeability is regulated by the national agency that regulates drugs in each member country of the European Union and by the pharmacy board in each Canadian province.
A potential concern is that the development of biosimilars might negatively impact the development of novel biopharmaceuticals. Many manufacturers of innovator drugs are now also developing biosimilars of products that have been manufactured by other companies. These resources, which otherwise might be directed toward therapeutic innovation, are being used to develop biosimilars. If the widespread availability of biosimilars significantly drives down the price of all biopharmaceuticals, then pharmaceutical companies might be deterred from developing novel drugs because of a decrease in their return on investment. Moreover, with the advent of very effective medications that are now commercially obtainable to treat rheumatoid arthritis and other inflammatory diseases, the number of patients available to enroll as subjects in clinical trials is limited. A comparator-controlled trial of a biosimilar and its reference product, which already has been proven effective and safe, is likely to be more appealing to potential subjects than would a placebo-controlled trial of a novel biopharmaceutical that has as yet unproven efficacy or safety. Thus, clinical trials of biosimilars compete with trials of novel therapies for enrollment of subjects. However, those patients who have responded inadequately to TNF inhibition would not be eligible for a clinical trial of a biosimilar TNF inhibitor and might still be available to enroll in clinical trials of novel biopharmaceuticals with different mechanisms of action.
The advantage of using a biosimilar, instead of its reference product, is purely economic. The availability of biosimilars should decrease the cost of treating an individual patient, since it is expected that the biosimilar will cost less than the reference biopharmaceutical. In developing markets in which access to biopharmaceuticals is restricted by cost, a lower-priced biosimilar might allow a patient to receive a treatment that previously was difficult to obtain or unavailable. However, in most other countries, why should a patient accept a biosimilar if he or she can receive the originator biopharmaceutical? The justification for this is based upon the Enlightenment moral philosophical concept of the “social contract,” in which an individual patient should accept a lower-cost biosimilar so that comparable medications are more widely available to all members of society. The potential risk to the individual of switching to a lower-cost biosimilar should be outweighed by the potential benefit to society of expanding access to care for all. Although the availability of lower-cost biosimilars might result in a net increased cost of health care, if it allowed more patients to be treated with biopharmaceuticals for their diseases, rather than with relatively inexpensive but less effective small-molecule drugs, this increased global access to more effective treatments for inflammatory diseases should prevent associated morbidity and disability in both the developing and the developed world.