>> Within the European Union (EU), biosimilars are approved by the European Medicines Agency (EMA). Here, the biosimilarity with the reference drug is checked on the basis of comparability studies. During the approval process, however, there is no review or confirmation of any interchangeability of biosimilars with the reference product. In the guide for healthcare professionals on biosimilars, prepared jointly by the EMA and the European Commission, the term “interchangeability” distinguishes between switching carried out under medical supervision and substitution carried out at the pharmacy level without consultation with the prescribing doctor [5 ]. The Guide points out that “the decision as to whether an exchange or substitution of the reference medicinal product with the biosimilar is permitted […] is taken at national level” since “responsibility for prescribing practices […] lies within the competence of the Member States”. lies [5].
The Paul Ehrlich Institute (PEI), which is the national regulatory authority for biologically produced medicines in Germany, points out that it is not responsible for this issue and that a decision must be made by the national health system. With the exception of the substitution of bioidenticals permitted in Germany, (non-medical) substitution is not permitted in any of the EU member states. However, the change (made by a doctor) is viewed by the PEI as unproblematic [6].
In its guidelines on biosimilars, the Medicines Commission of the German Medical Association (AkdÄ) presented and evaluated the evidence available on the exchange of biosimilars as of August 15, 2017, based on a literature search. While the change is assessed as unproblematic based on the available data, the AkdÄ rejects the automatic substitution of a reference drug by biosimilars [7].
In the AkdÄ's opinion, the change made by a doctor is sufficiently evidence-based in the approval process, as it is based on evidence of the same quality, effectiveness and safety compared to the reference drug. Regarding automatic substitution, the AkdÄ concludes that “valid data to assess the effects of substitution on adherence and therapy outcomes are missing” [7]. With regard to the question of substitutability, the AkdÄ guidelines were updated and reassessed as part of this review.
The FDA concept
Unlike the EMA, the US Food and Drug Administration (FDA) has an independent two-stage approval process for biosimilars. The actual approval takes place through proof of biosimilarity. However, in a second independent step, the status of “interchangeability” can also be granted, which, if the individual states regulate accordingly, includes the permission for substitution at the pharmacy level.
In May 2019, the FDA adopted a guideline on the requirements for the evidence to be submitted for the automatic substitution of biologicals [8]. As of October 2019, however, interchangeability status has not yet been granted to any biosimilar.
The methodological requirements for the evidence to be presented are based on three points, the first two of which have already been answered regarding approval as a biosimilar:
• The biological product is biosimilar to the reference product
• It is expected that the same clinical outcome as the reference product will be achieved in every patient
• For a product that is administered multiple times to one person, the risk of safety or reduced effectiveness is not increased by switching the product and the reference product

As a study design, the Food and Drug Administration requires randomization into two study arms after an introductory phase in which treatment with the reference product takes place. Here, further treatment with the reference product should be compared with repeated changes. It should be noted that a sufficient duration of the introductory phase is ensured with regard to a solid baseline depending on the study objectives. At least three switches must be included in the switch arm, with the last one leading from the reference product to the proposed replaceable product.
The endpoint categories to be examined are pharmacokinetics/dynamics (PK/PD), immunogenicity and safety. In addition, efficacy endpoints can be reported. Validated assays should be used for pharmacokinetic data. The study on which the proof of biosimilarity is based serves as a validation study with comparable results for the biosimilar and reference product.
The comparative assessment should then occur during the final exposure period, after an appropriate washout period of at least three or more half-lives [8].
Transferability to Germany
As part of the evaluation of interventions by self-administration, the principles of evidence-based medicine (EbM) are the agreed standard. The application of the international EbM standards, which has been practiced by IQWiG and the G-BA for many years, has always been argued to protect patients from potential harm or lack of effectiveness of an intervention. The importance of this principle is also made clear by the fact that the G-BA has committed itself in its rules of procedure to determining “the generally recognized state of medical knowledge on the basis of evidence-based medicine”. There are no apparent reasons for not applying these principles when deciding on the substitutability of biologicals. Rather, when checking the potential harmfulness of a planned measure, the same strict requirements must be met as, for example, B. be used as part of the benefit assessment.
Detailed FDA suggestions for study design
In principle, the FDA requirements are also suitable for assessing the interchangeability of biologicals with the reference preparation in Germany. The required study design is sufficient and necessary to answer the question of the safety of multiple switches between reference products and biosimilars.
Deviating from the usual requirements in Germany for the patient relevance of endpoints, pharmacokinetic or pharmacodynamic endpoints are considered when checking automatic interchangeability, which are usually classified here as surrogate endpoints. However, the background is the high sensitivity of pharmacokinetic/dynamic parameters for the exchange, which is also crucial when checking biosimilarity. Patient-relevant effectiveness endpoints, on the other hand, are assessed as less sensitive and only have a supporting function in the evaluation.
Despite the proven safety of multiple switching between reference preparation and biosimilar, the effects on adherence and therapeutic success of automatic substitution at the pharmacy level cannot be meaningfully assessed under controlled study conditions. Nocebo effects are also conceivable, i.e. H. Adverse events due solely to the patient's negative expectations should be taken into account. Ultimately, what is necessary for this is data from everyday care. These in turn would only be collected when the aut-idem substitution was implemented. A demand for appropriate evidence before implementing the proposed regulations for automatic substitution would counteract the legal intention. Against this background alone, such a demand must be refrained from. Nevertheless, it is necessary to demand that the implementation of the aut-idem regulation for suitable biosimilars is accompanied by appropriate health care research studies. To this end, it would be expedient to initially introduce and evaluate the substitution on a regionally limited basis.
Transferability to other indication areas
For the FDA “Interchangeability” status, extrapolation to other indications is generally permitted, but must be scientifically justified. This requirement essentially corresponds to the approval requirements for biosimilars. There is a waiver of additional studies for other indications if
1. it has been demonstrated based on safety, efficacy and immunogenicity data in one of the approved indications that there are no clinically relevant differences and
2. the relevant mechanism of action or the receptors involved in the extrapolated indication are the same.
Provided that the extrapolation to other indications has been checked and confirmed as part of the approval, no separate evidence should be required for interchangeability.
This results in the following basic requirements for the evidence for the automatic interchangeability of biosimilars:
1. Randomized, controlled study design
2. The question of the study must correspond to the planned intervention
3. Transferability of the results to other indications covered by the approval may be based on the decisions of the approval authority for extrapolation
4. In justified cases, the impact on patient-relevant endpoints can be assessed using suitable surrogate parameters due to the higher sensitivity
5. Studies should be supplemented with real-word evidence (RWE) to assess any impact on adherence
Research and assessment of existing evidence
Based on the AkdÄ guidelines, the update and reassessment of this work is presented below. For this purpose, the following question was examined for the systematic literature search: What effect does the substitution of reference medicinal products with biosimilars or the substitution of biosimilars with other biosimilars have in terms of therapeutic effectiveness, drug safety, pharmacokinetics and pharmacodynamics? The research question was deliberately broadened in order to identify all available studies with medication changes from biologics in the database search and to narrow them down by applying the FDA criteria as part of the evidence assessment.

Search process, selection and data extraction
The search strategy was based on the principles of PRISMA
statements [9]. The PubMed database was systematically searched for studies published after August 15, 2017. The search syntax used, among other things, the MeSH terms “biosimilar pharmaceuticals”, “drug substitution” and “therapeutic equivalency” and keywords that indicated a change in medication in the study design (cross-over, switch). Human studies with a comparative design that examined single or multiple switching between reference product and biosimilar or between different biosimilars were selected. Relevant endpoints from the selected studies were then extracted (efficacy, safety, immunogenicity, pharmacokinetics, pharmacodynamics, adverse events).

Evidence assessment of search results
Both the studies found by the AkdÄ as part of the literature search and the studies found as part of the update were checked for their suitability for assessing the safety of the exchange between reference product and biosimilar using an evidence grid based on the requirements of the FDA. The criteria of study design, switch versus non-switch, definition of the control group and number of medication changes were applied to the search results in order to identify studies that had the greatest congruence with FDA guidelines.
Updated literature search
The 35 studies included in the update (Fig. 1) included 15 randomized controlled trials (Fig. 2). This study design meets the basic requirement of the FDA evidence requirements for automatic interchangeability of biological
similars.
Another FDA requirement to demonstrate interchangeability is to examine a drug switch from reference product to biosimilar with respect to safety, altered efficacy, immunogenicity, and PK/PD parameters compared to constant treatment with the reference product. To this end, the FDA recommends the use of studies with a switch design, consisting of a control study arm in which study participants are continuously treated with the reference drug and at least one other study arm that includes multiple switches between treatment with the reference product and biosimilar [8] .
Among the 15 RCT studies, seven studies were identified that contained both a switch arm and a study arm with constant treatment [10-16].
According to the FDA recommendation, the control study arm should use constant treatment with the reference product. Six studies met this criterion and examined the switch arm compared to long-term treatment with the reference product.
Furthermore, the FDA recommends at least two exposure periods for both the reference product and the biosimilar (i.e., at least three switches between the two products). Of the seven studies, three studies showed such a multiple switch (Blackwell et al. [10], Blauvelt [11], Gerdes et al. [13]). The study by Blackwell et al., Blauvelt et al. and those of Gerdes et al. analyzed EGALITY study are explained in more detail below.

Blackwell et al. study
In a four-arm RCT study, Blackwell et al. the effects of alternating treatment with the filgrastim biosimilar EP2006 “Zarzio” and the reference product “Neupogen” [10]. 218 patients with breast cancer undergoing neoadjuvant myelosuppressive chemotherapy were included. These were randomized into four study arms in a ratio of 1:1:1:1 (see Fig. 3). Two study arms received either the biosimilar “Zarzio” or the reference product “Neupogen” for the entire duration of the study. Study participants in the other two arms received alternating treatment with reference product and biosimilar (starting with the reference product or with the biosimilar) per period.
The primary endpoints of the study compared the study arms in terms of efficacy, PD/PK (absolute neutrophil count), safety (adverse events) and immunogenicity. The study included six cycles, each with five medication changes in the switch study arms.
The study results demonstrated that although a total of 3.4% (n=3) more febrile neutropenias (FN) occurred in both switch groups than in the comparison arm with reference product treatment (n=0), overall there was non-inferiority of the biosimilar EP2006 for the occurrence of FN was detected. Overall, only one study participant in the switch study arm was hospitalized due to FN in the sixth cycle. Side effects of a similar magnitude were reported by both study participants in the reference group (39.2%) and by study participants in the switch arms (42.1%).
Muscoskeletal disorders after filgrastim therapy were reported by 39.2% (reference product) and 35.5 % (biosimilar) of study participants. 33.3% and 30.8% of the study participants (reference product and switch study arm, respectively) reported suffering from bone pain. In addition, no clinically relevant differences could be determined between the study participants in both switch arms and the study participants in constant treatment, neither for safety endpoints nor for other efficacy and immunogenicity parameters.

ADACCESS study (Blauvelt et al.)
The second study that meets the FDA evidence requirements
The adalimumab biosimilar approval study by Blauvelt et al. This randomized, double-blind, multicenter study depicts the four-fold switch between the adalimumab originator (“Humira”) and the biosimilar GP2017 (“Hyrimoz”).
After a six-month washout period, patients were randomized 1:1 to receive a loading dose of either the reference product or the biosimilar subcutaneously. In the first phase of the study, a 40 mg dose was given every two weeks until the 15th week (see Fig. 4). Patients who demonstrated ≥ 50% improvement in the Psoriasis Area and Severity Index (PASI) scale during the initial study period were re-randomized in a 2:1 ratio. In the second study period (weeks 17-35), these patients either continued treatment with the product from the first study period (control group) or switched between reference product and biosimilar three times for six weeks each. In the last study phase of the extension phase, all patients again received the medication assigned in the first study phase.
The study confirmed the biosimilarity of GP2017 to the adalimumab reference product through comparable safety, efficacy and PK/PD parameters in the treatment of patients with moderate to severe plaque psoriasis. No clinically relevant differences in immunogenicity and safety between reference product and biosimilar could be determined, neither among study participants with multiple medication changes nor among study participants with constant reference product or biosimilar treatment.

EGALITY study (Gerdes et al.)
The third study that fully meets the FDA's requirements for “interchangeability” (with the exception of the investigation of PD/PK parameters) is the EGALITY study, which was conducted by Griffith et al. [17] and by Gerdes et al. was analyzed for the effects of multiple medication changes. This efficacy and safety study investigated the multiple switching between the etanercept biosimilar GP2015 and the reference product ETN “Enbrel”. The study was conducted multicenter in 11 European countries and South Africa and included 531 patients with chronic psoriasis who had already received light therapy or systemic psoriasis therapy. During the first study period, the study participants independently administered either the reference drug or the biosimilar twice a week after prior randomization. Subjects who demonstrated ≥50% improvement in psoriasis-affected body area after the first 12 weeks of the study were re-randomized. Further treatment in the second study period took place either with the drug that was administered in the first study period or with a three-time medication change between biosimilar and reference product (see Fig. 5).
The study results showed a similar treatment course between study participants with constant treatment and those with multiple biosimilar reference product changes. The incidence of side effects was approximately the same (36.7% switch study arms (aggregated) vs. 34.9% constant treatment). This study exemplifies that multiple medication switching between etanercept
Biosimilar and the etanercept reference product have no significant effects on clinical parameters in the medium term.
Both the studies by Blackwell et al. and Blauvelt et al. and the EGALITY study comply with the standard postulated by the FDA in its guidelines due to the study design used with multiple medication changes.
Evidence assessment of the studies
AkdÄ guidelines
Of the 59 switch studies included in the AkdÄ guide, 44 studies had an RCT study design (see Fig. 6).
When controlling for the other FDA criteria for safe biosimilar substitution, only eight RCTs had a design with a control study arm and a switch arm. In four of these studies, the control arm consisted of biosimilar treatment. The other four studies met the FDA's evidence requirements by treating study participants in the control arm with the reference product.
In comparison to the updated literature search, the AkdÄ guidelines did not report any studies that fully met the evidence requirements required by the FDA. Although four studies met essential FDA criteria for a safe exchange, they only examined a one-time switch between reference product and biosimilar.
Real-world evidence
Although one of the studies listed in the AkdÄ guidelines does not meet the FDA's evidence requirements, it does provide evidence from a real-world setting that cannot be represented in the RCT design. The Danish national DANBIO cohort study examined the automatic, non-medical substitution of the infliximab reference product with a corresponding biosimilar [18]. The study included 802 participants with arthritic diseases and collected efficacy parameters as endpoints. Adherence parameters were not examined during the study. Using the DANBIO registry, study participants were regularly monitored for disease activity before, during and after switching to the biosimilar. This study found comparable efficacy and safety of the biosimilar as well as a comparable drop-out rate to the reference drug.
In addition, the authors published a similar follow-up study in 2019 based on the DANBIO registry data [19]. The study was similar in design and intervention to the study from 2017. In total, the study included 2,091 study participants and examined clinical parameters and the treatment success of study participants with biosimilar switching and constant reference product treatment. The authors cited patient-related factors and non-specific drug effects as the reason for switching medication in the switch group to the reference product.
However, due to the less sensitive efficacy endpoints and the unexamined effects on patient adherence, this study does not answer the relevant questions about the impact of substitution at the pharmacy level.
Conclusion and recommendations
With the GSAV, the legislature commissions the G-BA to determine the scientific basis for the interchangeability of biosimilars with regard to their reference product. Building on this, the G-BA should decide in individual cases on the possibility of substituting a reference product with a bio-similar at the pharmacy level. The new Section 129 SGB V expressly regulates “interchangeability with regard to a biological reference medicinal product”, so the question of the interchangeability of biosimilars with one another is not addressed by the GSAV regulation. In contrast to the switch from reference drug to biosimilar, there are currently no studies available for the (multiple) medication switch between different biosimilars that examine the effects in terms of effectiveness, safety and adherence of patients [20]. The differentiation made in the GSAV also makes it clear that the proof of substitutability must be product-related and cannot be drawn from one biosimilar to another. This does not affect the substitutability of bioidenticals with each other.
The principles of evidence-based medicine are the agreed standard for the evaluation of interventions by the self-administration of the statutory health insurance. In its rules of procedure, the G-BA has committed itself to determining “the generally accepted status of medical knowledge on the basis of evidence-based medicine”. There are no apparent reasons for not applying these principles when deciding on the substitutability of biologicals. Rather, when checking the potential harmfulness of a planned measure, the same strict requirements must be met as, for example, B. be used as part of the benefit assessment.
Unlike in Europe, where the decision on possible substitutability is the responsibility of the member states, the question of substitutability is regulated by the US FDA within the framework of a separate approval procedure. The requirements laid down there for the evidence to be provided are essentially also suitable for the question to be investigated in Germany. An essential element here is the need to switch between reference product and organic product multiple times as part of a randomized controlled study.
similar compared to continuous therapy with the reference product. Studies that only examine the one-time switch between reference product and biosimilar treat a different intervention than that intended by the GSAV. Such studies would not correspond to the usual use of the PICO scheme in evidence-based medicine, for example as used in Cochrane reviews. The study design required by the FDA allows a reliable comparative assessment of the two relevant therapy regimes.
The search revealed three studies that meet these requirements, although further FDA requirements can only be verified through a more detailed assessment of the studies. However, the existing studies confirm the feasibility of the requirements of evidence-based medicine for the question at hand. However, possible effects on patient adherence or a conceivable nocebo effect cannot be meaningfully investigated in this design. Any decision on substitution at the pharmacy level should therefore always include a mandate for the self-administration to carry out research into this measure under real supply conditions. Given that there is no relevant international experience with the intervention planned by the legislature, an initial regionally limited introduction of the new regulation would be expedient. <<

background
With the Act for More Safety in the Supply of Medicines (GSAV), the legislature has given the Federal Joint Committee (G-BA) a regulatory mandate for the interchangeability of biosimilars. The G-BA should follow the regulations of the GSAV in the Medicines Directive (AM-RL) in accordance with Section 92 SGB V
decide on the interchangeability of biosimilars within one year of the law coming into force, taking into account their therapeutic comparability. After three years, the G-BA should make regulations in the AM-RL on the interchangeability of biological reference medicinal products by pharmacies. The legislator makes it clear in the justification for the law that “it is about assessing the interchangeability of biosimilars with the respective reference drug” [1]. The review of the interchangeability of biosimilars with one another is therefore not the subject of the mandate to the G-BA. As soon as the G-BA has determined interchangeability with regard to a biological reference medicinal product, the pharmacists' obligation to make a substitution (aut-idem) applies according to Section 129 SGB V, which previously only applied to generics with identical active ingredients and bioidenticals. According to the will of the legislature, “scientific findings on the interchangeability of biosimilars and experiences with the supply practice of biosimilars should be collected and evaluated” within the three years.