Current state of biologic pharmacovigilance in the European Union: improvements are needed

ABSTRACT

Introduction: Pharmacovigilance is essential to monitoring the safety profiles of authorized medicines. Compared with small-molecule drugs, biological drugs are more complex, more susceptible to structural variability due to manufacturing processes, and have the potential to induce immune-related reactions, underscoring the importance of safety monitoring for these products. Although highly similar to reference products, biosimilars are not expected to be structurally identical. For these reasons, proper reporting of potential adverse drug reactions (ADRs) using distinguishable names and batch numbers is essential for accurate tracing of all biological drugs. To address the need for robust pharmacovigilance, the European Parliament and Council of the European Union provided legislation regarding pharmacovigilance of biologics in 2010.

Areas covered: This narrative review examines the current state of pharmacovigilance for biologics in the European Union (EU) and discusses relevant information on pharmacovigilance of biosimilars, the current EU pharmacovigilance system, and areas that could be improved.

Expert opinion: Although steps have been taken to improve pharmacovigilance of biologics in the EU, several enhancements can still be made, including additional training for healthcare professionals on ADR reporting, the use of 2D barcodes that enhance traceability, and an open discussion of potentially missed opportunities in the pharmacovigilance of biosimilars.

KEYWORDS:

• Adverse drug reactions
• biopharmaceuticals
• biosimilars
• pharmacovigilance

1. Introduction

Pharmacovigilance is the cornerstone in monitoring the safety profiles of medicines once they are in clinical use [1]. The information collected by routine pharmacovigilance practices allows risk communication and minimization measures to occur in a timely manner when new safety signals are identified for a given indication or product, and this information can also prompt remedial actions if changes in a safety profile are noted during the life cycle of the product [1,2]. In the era of multisource biologics, routine pharmacovigilance becomes particularly important in ensuring the safe and effective use of these products, as rare events such as immune-mediated reactions in some cases may be detected only after the products are marketed [1,3]. Moreover, biologics are more susceptible to physicochemical differences (i.e. structural and functional) due to differences in manufacturing, their environment, or handling conditions relative to small-molecule drugs [2].

The European Medicines Agency (EMA) defines a biosimilar as ‘a biological medicine highly similar to another biological medicine already approved in the EU’ [4]. The first biosimilar was approved in the European Union (EU) in 2006 [5]. Although approved biosimilars are expected to have equivalent safety and efficacy compared with their reference products and the EU monitoring system has not identified new safety concerns related to biosimilars, there is the potential for differences in safety signals in the postmarketing setting, as with all biologics [2,4]. In addition to encouraging submission of adverse drug reaction (ADR) reports, the challenge with pharmacovigilance for biologics is to ensure that the data collected are accurate to the brand and batch of product used so that meaningful conclusions can be drawn from these reported data for all biologics, including biosimilars [1].

An example of the importance of accurate, longitudinal data collection for biologics is demonstrated by the antibody-mediated pure red cell aplasia (PRCA) outbreak that occurred in patients with chronic kidney disease that was traced to a specific preparation of epoetin alfa (Eprex®; Janssen-Cilag GmbH, Neuss, Germany). This ADR was an immune response that resulted in severe anemia from the formation of anti-epoetin antibodies due to changes in the product’s manufacturing, formulation, and primary container [6–8]. Identification of the causes of PRCA was delayed, potentially related to late, incomplete, or inaccurate product-specific reporting, despite the limited number of erythropoietin-stimulating agents available at that time. This example highlights the possibility of ADRs occurring because of manufacturing changes over the life cycle of a product. Similarly, a clinical trial evaluating the safety and efficacy of subcutaneous administration of the biosimilar Binocrit® (epoetin alfa [HX575]; Sandoz GmbH, Kundl, Austria) compared with its reference product (Eprex®) was suspended because two of 174 patients receiving the biosimilar developed erythropoietin-neutralizing antibodies, one of whom developed PRCA [9]. This rate of neutralizing antibodies was higher than expected and was believed to have resulted from denaturation of epoetin alfa and formation of immunogenic aggregates due to elevated levels of tungsten, which leached during the manufacturing of prefilled syringes [6,9,10]. The spike in cases of PRCA associated with subcutaneous administration of epoetin alfa resulted in regulatory agencies in many countries contraindicating subcutaneous administration, resulting in a decrease in the number of cases. Notably, between 2006 and 2010, after the subcutaneous presentation of epoetin alfa was changed to use coated stoppers and the subcutaneous formulation was reintroduced, PRCA rates were low and no significant differences between Eprex® and its comparators were observed [11]. Binocrit was approved by the EMA in 2007, and no unexpected safety signals have been observed during routine pharmacovigilance [12,13].

Although new safety events may be rare, robust pharmacovigilance measures should be in place to provide clear and accurate product traceability by identifying the specific product, manufacturer, and batch of a biologic so that ADRs can be readily and accurately attributed and effectively mitigated. In an attempt to accomplish this aim, regulators around the world have taken different approaches to facilitate product-specific pharmacovigilance [14]. The EU, the first region to adopt biosimilars, has taken a legal approach, passing a legislative requirement to be implemented by Member States requiring the reporting of ADRs for biologics using brand name and batch number [15]. In contrast, the US Food and Drug Administration (FDA) has adopted the use of distinguishable nonproprietary names for biological drugs and biosimilars through the addition of a unique identifier as part of the proper name [16]. Japan has a similar system that uses the Japanese Accepted Name followed by an identifier indicating that it is a biosimilar and a number indicating the order of approval [17].

In particular, this narrative review examines current work and ongoing initiatives in the EU to enhance and improve pharmacovigilance practices and to provide recommendations that could improve the safety monitoring of all biological products.

2. Why additional measures are needed for pharmacovigilance of biologics, including biosimilars

Biological drugs are large protein structures, often hundreds of times larger than small-molecule drugs [2,4]. Biologics are also structurally complex, with multiple posttranslational modifications such as glycosylation that may affect efficacy or safety [4]. Manufacturing changes also occur during the life cycle of a biologic and have the potential to have a significant impact on safety and efficacy. Although new safety signals resulting from changes in the manufacturing processes of biologics are exceptionally rare, all biologics necessitate safety monitoring over their entire life cycle [3,18].

Pharmacovigilance of biologics is further complicated by the potential for delayed immunologic reactions where there may be a lag between administration of the biologic and the appearance of a serious reaction (e.g. neutralizing drug antibody) [19]. This time lag may hamper the ability to attribute a suspected ADR to a specific product when a patient has been switched between related biologics. The experience with antibody-mediated PRCA demonstrates the importance of traceability of biologics to ensure that ADRs are assigned to the correct product so that appropriate action can be taken [20].

A biosimilar is a biological medicine highly similar to but not necessarily identical to a currently approved biological medicine. Because the manufacturing of biosimilars uses unique, proprietary cell lines as opposed to well-defined chemical synthesis as for small-molecule drugs, biosimilars, unlike small-molecule drugs, are not exact copies of their reference products [2,21,22]. Differences in manufacturing processes between a biosimilar and its reference product can result in minor structural differences [20,21]. Such differences are expected and permitted as long as they do not affect safety and efficacy [4,23]. However, it is possible that minor structural differences may result in rare or delayed safety events, such as immune-related events, that may not be identified in registrational clinical trials, and thus may not be detected until after marketing authorization when there is longer and larger patient exposure [2].

Postapproval manufacturing changes occur for all biologics, including biosimilars. When such changes occur, the manufacturer must have an appropriate system in place to ensure that product quality, patient safety, and supply are maintained [24]. Following a change in the manufacturing process of any biologic, the manufacturer needs to provide evidence that the active molecule is ‘comparable’ to the original one after the manufacturing change [25,26]. This is different than establishing biosimilarity, in which a biosimilar made by a different manufacturer is compared to a previously approved reference product from a different manufacturer [26].

Substitution of a generic product for a branded chemical drug is commonly practiced by pharmacists and often occurs without prescriber knowledge [27]. These generic drugs share the same nonproprietary name as their reference product. Thus, if brand name, manufacturer, or batch number details are not recorded, the practice of substitution may lead to the capture of ambiguous product information, which can contaminate reference product safety databases, thereby preventing safety events from being accurately tracked and associated with the correct product. This consideration has been exemplified by ADR reporting patterns for enoxaparin following the launch of generic enoxaparin, which is regulated as a traditional small-molecule pharmaceutical in the US under the Federal Food, Drug, and Cosmetic Act but is considered a biologic by the EMA [28]. After the originator enoxaparin (Lovenox®; Sanofi-Aventis, Bridgewater, NJ) lost exclusivity and a generic version entered the market, its market share declined rapidly; however, at a time when 50% of the market was being supplied by generics, only 5% of spontaneous FDA Adverse Event Reporting System (FAERS) reports were processed by manufacturers of generic enoxaparin [29]. Based on market share of newly launched generic enoxaparin products, ADR reports attributable to specific generics were ninefold lower than expected. The consequence of this level of misattribution is that a safety signal for the generic would need to be at least nine times more frequent to be detected [29]. Misattribution of ADRs from generics to the branded originator is frequent and has been observed for commonly used drugs [30].

Misattribution or ambiguous attribution has also been observed for biologics. In pharmacovigilance systems in the United States, the EU, Australia, and the Netherlands, accurate brand name attribution for filgrastim ranged from 58% in Australia to 90% in the EU [31,32]. Given the inherent, but accepted, variability in biological medicines and their potential to induce unwanted side effects (such as immune reactions), the situations noted above reinforce the notion that biosimilars must not be treated as generic medicines and, as such, require more robust product-specific pharmacovigilance capabilities.

3. Pharmacovigilance in the EU

3.1. Overview of pharmacovigilance systems in the EU

Recognizing that biosimilars require product-specific ADR traceability and must not be treated as generic medicines, the EU introduced new legislation in 2010 with respect to pharmacovigilance of biologics [15,33]. The European Parliament and the Council of the European Union issued a directive emphasizing the requirement ‘to identify clearly any biological medicinal product prescribed, dispensed, or sold in their [Member States’] territory which is the subject of a suspected adverse reaction report, with due regard to the name of the medicinal product…and the batch number’ [15]. The legislation also describes responsibilities for each Member State, which includes the operation of a pharmacovigilance system useful for monitoring adverse reactions from medicinal products [15]. Further adjustments were made to the 2010 legislative framework to improve patient health and safety by strengthening measures for monitoring safety and conducting reviews at a European level [34,35]. The 2010 legislation came into effect in July 2012 [36].

The EMA is responsible for ‘implementing measures for continuously supervising the quality, safety, and efficacy of authorised medicines to ensure that their benefits outweigh their risks’ and ‘developing best practice for medicines evaluation and supervision in Europe, and contributing alongside the Member States and the European Commission to the harmonisation of regulatory standards at the international level’ [37]. The EMA has provided good pharmacovigilance practices for biologics as well as a definition of roles and responsibilities for all stakeholders [38]. In good pharmacovigilance practices guidelines, the EMA distinguishes biologics from chemically synthesized medicines, noting their complexity and how changes in the manufacturing process can affect product quality. Importantly, the EMA notes that traceability is a key requirement for biologics, including both the reference products and biosimilars, because of the greater inherent variability in product characteristics [38]. In another effort to improve pharmacovigilance in the EU, the labels of all new biologics, including biosimilars, should have a standardized explanatory sentence and a ‘black triangle’ (▼), which indicates that they are subject to additional monitoring [2,15,33,39]. Individual Member States have also taken steps to improve pharmacovigilance, including raising awareness through social media and other campaigns, improving communication with and education of healthcare professionals, and refining electronic systems for ADR reporting [40].

Various groups within the EMA exist to collect, monitor, assess, and react to safety signals across the EU [37]. As part of the new EU pharmacovigilance legislation, the Pharmacovigilance Risk Assessment Committee (PRAC) was established and is responsible for assessing and monitoring safety issues for medicines at the EU level [36]. The PRAC provides recommendations to the EMA with respect to opinions for centrally authorized medicines and ‘on the development of guidelines and standards and advises on operational aspects of EU pharmacovigilance’ [37]. The PRAC also provides recommendations to the Committee for Medicinal Products for Human Use (CHMP) when it adopts opinions for centrally authorized medicines and referral procedures [37]. The EMA supports various databases for the collection of pharmacovigilance data; in particular, EudraVigilance is an electronic database used to collate worldwide reports of suspected ADRs [36,41]. Marketing authorization holders and Member States are required to submit suspected ADR reports to EudraVigilance, making this database integral to detecting safety signals [15,36]. Even with these advancements and strengths, there is room to further improve the efficiency and effectiveness of pharmacovigilance practices within the EU [36].

3.2. Implementation and experience

The implementation status of the pharmacovigilance directive across the European Union varies among Member States and ranges from simple incorporation of the legislation into national law to full implementation of the legislation reflected by changes in relevant national systems. Such differences can be observed in discrepancies among Member States at a basic level with respect to appropriate classification of biological drugs as being biologics and not small-molecule drugs within their local pharmacovigilance databases. Klein et al. examined five Member States’ pharmacovigilance data and found considerable discrepancies for the classification of various biological drugs, including somatropins, epoetins, filgrastims, and monoclonal antibodies [42]. Correct classification as biologics ranged from 0% to 100%, depending on the Member State and the biologic examined [42].

Further limitations have been observed in ADR reports submitted to EudraVigilance. Vermeer et al. analyzed reports from EudraVigilance over the period of 2004 to 2010 to study the traceability of biologics and biosimilars in this database [43]. The analysis included biopharmaceuticals for which a biosimilar had been approved in the EU (at the time, only a few biosimilars had been approved: epoetin, filgrastim, and somatropin) [43]. Ninety-six percent of the reports for the three product classes included a clearly identifiable product name (e.g. brand name or INN plus name of marketing authorization holder) of the suspected drug [43]. By contrast, only 5.3% of the reports included the batch number for the three product classes with approved biosimilars (Table 1) compared with 21.1% for reports for all biopharmaceuticals [43]. The reports were heavily weighted toward epoetins, with reports on epoetin representing 70.7% of the total suspected ADR reports [43], likely due to heightened awareness of safety concerns with epoetins associated with the PRCA outbreak. Reports for epoetins had better name attribution (i.e. identifiable product name in 98.9% of reports for epoetins) compared with reports for the other drug classes, thus exaggerating the accuracy of product identification overall [43]. Excluding reports for epoetins, 89.7% of the ADRs reported to EudraVigilance included a clearly identifiable product name [43]. It should also be noted that this study examined a time period before the requirement for brand name and batch number reporting.

Table 1. Identifiability of biosimilars by product name in EudraVigilance between 2004 and 2010 [43].

Product class	Suspected ADR reports, n	Weighted suspected ADR reports, n/N (%)	Suspected ADR reports with identifiable product name, n (%)	Suspected ADR reports with identifiable product name and traceable batch number, n (%)
Epoetin alfa	6903	6903/9759 (70.7)	6829 (98.9)	318 (4.6)
Filgrastims	706	706/9759 (7.2)	600 (85.0)	72 (10.2)
Somatropins	2150	2150/9759 (22.0)	1963 (91.3)	128 (6.0)
Total reports	9759	NA	9392 (96.2)	518 (5.3)
Total excluding epoetins	2856	NA	2563 (89.7)	200 (7.0)

ADR: adverse drug reaction; NA: not applicable.

Vermeer et al. recently published a similar analysis of ADR reports from EudraVigilance from January 2011 through June 2016 [32], a period that covers the introduction of the 2010 EU pharmacovigilance legislation on ADR reporting [15,33]. The study included 10 classes of biologics for which biosimilars (i.e. etanercept, infliximab, insulin glargine, filgrastim, epoetin alfa, somatropin, and follitropin alfa) or related products (i.e. human normal immunoglobulin, interferon beta-1a, and octocog alfa) have been approved in Europe [32]. Overall, 96.7% of the suspected biologics and 86.3% of the biologics categorized as concomitant or interacting medications included adequate identifiers in ADR reports in this time frame. When narrative and reporter comments were included, the identifiability of suspected biologics increased to 97.2% [32]; however, data reported to the Amgen safety database (detailed below) [44] suggest that more work still needs to be done to improve identification accuracy. Identifiability varied by class and ranged from 89.5% for filgrastim to almost 100% for interferon beta-1a [32]. The most commonly reported suspected biologic was etanercept, representing 37.5% of the overall reports [32]. There were notable improvements in the proportion of ADR reports with identifiable products between the first analysis (2004 to 2010) [43] and most recent analysis (2011 to 2016) [32], with reports with identifiable products increasing from 85.0% to 85.9% for filgrastim and from 91.3% to 96.2% for somatropin. Reports with identifiable products decreased slightly from 98.9% to 96.4% for epoetins [43]. Batch traceability was 20.5% in the 2011 to 2016 time frame for the 10 classes of biologics analyzed [32], which is similar to the 21.1% reported for all biopharmaceuticals in the 2004 to 2010 time frame, and ranged from 1.8% for follitropin alfa products and 69.3% for products based on human normal immunoglobulins [32]. The authors suggest that the low rate of batch traceability, and lack of a trend for improved batch traceability in their recent analysis, indicates that the EU pharmacovigilance legislation has not yet had an effect on the inclusion of batch numbers when reporting suspected ADRs for biologics [32].

Studies have examined the impact of the 2010 EU pharmacovigilance legislation on ADR reporting. ADR reports on filgrastim in the EU submitted directly to Amgen or forwarded by health authorities from April 2012 to December 2014 were assessed for inclusion of brand names to allow attribution to the correct manufacturer [44]. This represents a timeframe when biosimilar filgrastims were available and after passage of the EU 2010 pharmacovigilance legislation requiring Member States to encourage use of trade names in health records and ADR reporting [44]. Twenty-one percent of filgrastim-related spontaneous reports in the EU received directly by Amgen from 2012 to 2014 were not clearly identified to Amgen brands and were coded as ‘filgrastim’ (i.e., brand not specified; Table 2), and 10.8% of the reports forwarded to Amgen by health authorities were coded as ‘filgrastim’ [44]. The finding of 21% unattributed filgrastim reports received by Amgen’s safety department after 2011 and the passage of the EU 2010 pharmacovigilance legislation is consistent with the 85% rate of attribution reported specifically for filgrastims before 2011, suggesting the inadequate implementation of this aspect of the directive [43,44].

Table 2. Summary of reports for filgrastim in Amgen’s safety database (European sourced reports from April 2012 to December 2014 [44]).

Product name coded in Amgen database	Received from health authorities	Spontaneous reports received by Amgen	Overall
Filgrastim, n	8	38	46
Granulokine®, n	1	16	17
Neupogen®, n	65	129	194
Total reports, n	74	183	257
Filgrastim reports of total, %	10.8	20.8	17.9

*Amgen internal analysis of data in Amgen’s safety database. This analysis does not include spontaneous reports accepted by other manufacturers and forwarded to health authorities.

One consequence of the lack of identifiable product name in pharmacovigilance databases is a delay in time to detect the safety signal. For example, more recently, Vermeer et al. (2016) conducted a simulation study to examine the effect of ADR misattribution [45]. The authors found that misattribution could lead to a delay in identification of ADRs associated with the drug. For hypersensitivity reactions from infliximab, a large impact of misclassification was observed; the scenario of 34% misclassification resulted in a doubling in number of cases and time to detection (from 6 months to 1 year) of the risk [45]. The effect of misclassification was primarily dependent on the market share of the drug. For example, for drugs with low market share, a 50% misattribution would approximately double the number of cases and time to detection of new safety signals, whereas for drugs with a large market share, only a 5% rate of misattribution would double the number of cases and time to detection [45]. Application of this model to ADR reports for infliximab in 27 European countries taken from VigiBase indicated that ambiguous ADR reports would lead to a delay in product-specific safety signal detection (Figure 1(a)) [45,46]. Before the launch of biosimilar infliximab in late 2013, roughly 22% of infliximab ADR reports were filed using only the international nonproprietary name (INN; i.e. no brand name). A spike in ADRs and higher brand-level reporting (92%) occurred the year after launch, while INN-only usage fell to 8%. This spike may have been due to the Weber effect, a phenomenon characterized by ADR reporting that peaks one to two years after regulatory authorities approve a drug [47]. As of 2016, reporting by INN alone was at approximately 18%, with higher levels of reporting by brand name. Using the assumptions posed by Vermeer et al. [45] with respect to incidence of events, level of reporting, the observed rate of INN-only reporting above, and a biosimilar market share of 30%, signals could take an additional two to three months to detect, for a total time of up to 9.6 months (Figure 1(b)) [46]. Additionally, a lack of identifiable product names for biosimilars was observed in a pilot study conducted in the Netherlands [48,49]. The Netherlands Pharmacovigilance Centre Lareb received 1523 spontaneous ADR reports for recombinant biologics between 1 January 2009, and 31 December 2014. Of these, 76% had an identifiable brand name but only 5% of the reports contained a batch number [48,49].

Figure 1. Analysis of ADRs for infliximab in 27 European countries. (a) Accuracy of ADRs for infliximab branded and biosimilar products submitted to VigiBase between 2012 and 2016. Data were binned as total ADR reports and as those that specified only the INN infliximab (i.e. not manufacturer specific). (b) Plots showing the estimated impact of misclassified ADR reports on time to detect safety signals based on the assumptions noted in the text and utilizing the exposure-misclassification tool developed by Vermeer et al. [45].

*In absence of exposure misclassification, it takes 956 cases/0.5 years (assuming 90% underreporting) to detect the product-specific risk. ADR = adverse drug reaction; INN = international nonproprietary name.

Following the identification of these limitations in brand name and batch number reporting, a meeting among Dutch healthcare stakeholders and Escher – an independent platform for regulatory innovation [50] – was held and a number of recommendations were made. First, the group recommended taking steps to improve traceability of biologics through education of reporters, simplification of current technology for reporting (e.g. barcodes), and harmonization among public and marketing authorization holders [48]. They also recommended steps to improve quality of reporting and signal detection, such as improved product labeling, information systems to collect such information, and the development of best practices from other types of biologics, such as those used for blood products or vaccines, that could help direct improvements [48]. To further address challenges in correctly identifying and recording product and batch data for biosimilars, Escher formed a collaboration with the Drug Safety Research Unit in the United Kingdom (UK) to initiate the UK BIO-TRAC study [51].

In response to the above findings, the Dutch Ministry of Health set up a national system to monitor the safety of biologics in Lareb for potential future national expansion [52]. The Ministry also expressed the opinion that ‘biosimilars have been proven to have no relevant differences compared to an innovator biological medicinal product as far as quality, safety and efficacy are concerned’ [52]. They further stated that switching between biologics, either reference biologics or biosimilars, ‘is permitted, but only if adequate clinical monitoring is performed and the patient is properly informed’ [52]. Although data on switching trials are available for some products, the data are dominated by only a few products [53–56]. Thus, additional data from switching studies as well as ongoing pharmacovigilance would be helpful to further assess the safety of switching from a biologic to its related biosimilar or among biosimilars of the same reference biologic.

Recommendations were also recently made in Denmark to improve pharmacovigilance for biologics and biosimilars. New requirements in Danish executive orders as of January 2016 require that physicians record brand name and batch number in patient records and when reporting ADRs [57]. Other recommendations to improve pharmacovigilance of biologics, including biosimilars, included improving patient education and engagement on the topic of biosimilars, informing hospitals and physicians of new reporting regulations, and making hospitals and physicians aware that patients need information on biosimilars and assurances of their safety and efficacy [57].

The experience in the UK from the Medicines and Healthcare products Regulatory Agency (MHRA) has demonstrated how the pharmacovigilance legislation may be implemented. For example, the MHRA uses the Yellow Card Scheme to ‘monitor the safety of all healthcare products in the UK’ and collects information including ADRs, medicines of unacceptable quality, and counterfeit or fake medicines [58]. ADRs reported on Yellow Cards are ‘evaluated, together with additional sources of information such as clinical trial data, medical literature or data from international medicines regulators, to identify previously unknown safety issues’ [58]. As a response to the 2010 EU legislation, the MHRA has been proactively encouraging prescribers to provide the brand name and batch number to prevent any uncertainty when reporting ADRs [59,60]. As such, the Yellow Cards have been modified to include space for brand name and batch number and a notification to explain why this information is important [61].

3.3. Current plans and actions for improving pharmacovigilance

Various agencies have provided recommendations to improve pharmacovigilance. The World Health Organization INN Expert Group has provided guidance on the implementation of the Biological Qualifier (BQ) to assist in product identification and improve traceability [62]. The proposal for the BQ consists of four random consonants and an optional two numerical digits as a unique identifier that follows the INN [62]. The BQ would be available for all biological drugs that have INNs, and while it would not formally become a part of the INN, its voluntary adoption would provide uniform naming of biological substances across different regions’ regulatory authorities [62]. It should be noted that the BQ scheme is currently suspended; pending further studies by the World Health Organization, the future of the project is unknown at this time. In addition, the FDA and EMA have agreed to jointly form an international pharmacovigilance cluster whose primary goal is to support regional risk assessment to help parties with decision-making and timely communication to the public [63]. These goals will be achieved primarily through regularly scheduled teleconferences to promote the exchange of time-sensitive information [63].

The Falsified Medicines Directive (FMD) is designed to minimize the risk of counterfeit medicines entering the drug supply chain across Europe [64,65]. Drugs will be tracked from manufacturers to patients, and each pack will have a tamper-proof security seal and a 2D barcode. The barcode will allow each pack to have its own unique randomized number, which will be stored in a database managed by the European Medicines Verification Organisation. The packs will be checked into the database by manufacturers and then checked out and verified for authenticity by pharmacies when dispensed. The final phases of this scheme are planned for 2019 [64]. At this time, the FMD does not require that product-specific information transferred between trade partners be entered into a patient medical record [64].

Strengthening Collaboration for Operating Pharmacovigilance in Europe (SCOPE) is intended to help medicine regulators operate pharmacovigilance systems pertaining to the EU legislative requirements [66]. Its aims are to help Member States develop best practice in reporting ADRs, managing signals, communicating risks, and enabling Member States to develop, understand, and assess quality management systems for pharmacovigilance, developing standards for benefit-risk assessments, and creating a forum for interaction among European national competent authorities to strengthen regulatory collaboration.

The EMA conducts meetings with relevant stakeholders to discuss the implementation of pharmacovigilance legislation; there have been 11 stakeholder forums as of September 2017 [67]. Fostering the culture of shared responsibility, the purpose of these meetings is to bring ‘together regulators with patients, healthcare professionals and industry, to take stock of what [has been] achieved and what needs to be the focus over the coming years’ [67].

Since March 2014, the EMA has published summaries of risk management plans (RMPs) and developed a new template effective 31 March 2018 [68–71]. As stated by the EMA, the aim of an RMP is ‘to document the risk management system considered necessary to identify, characterise and minimise the important risks of a medicinal product’ [68]. To achieve this goal, RMPs should possess the following three components: (1) the safety specification, which characterizes the safety profile of the drug, emphasizing important potential risks and where further safety information is needed; (2) the pharmacovigilance plan, which describes planned pharmacovigilance activities to characterize and quantify current and new ADRs; and (3) the risk minimization plan, which describes the planning and implementation of risk minimization measures [68].

Finally, the use of registries for certain diseases may improve pharmacovigilance for biologics, including biosimilars. Registries designed for safety assessments generally will contain substantially more information than reports obtained through spontaneous reporting and have the advantage of systematic data collection on a wide spectrum of patients [72]. Registries may be disease or product focused and may collect data on the specific product, route of administration, dose, duration of use, start and stop dates, manufacturer, batch, lot, and if applicable, device [72]. Data collected in this manner via registries can support accurate tracing of ADRs to the correct product. To this end, the EMA launched the initiative for patient registries in September 2015 with a goal to expand patient registries to support coordination; harmonization of protocols, methods, and data structures; data sharing; and sustainability of existing registries [73]. In 2017, the EMA hosted stakeholder meetings for cystic fibrosis and multiple sclerosis patient registries [73]. As a next step, an EMA task force will work with stakeholders to develop plans to implement recommendations to improve stakeholder collaboration and optimize registries to support regulatory decisions [73].

3.4. Recommendations for stakeholder group improvements

Postmarketing monitoring of biological drugs, including biosimilars, using real-world studies can help improve the understanding of these drugs, increase confidence of clinicians on the use of biologics, and increase access for patients. Although steps have been taken to improve pharmacovigilance of biologics in the EU, additional improvements are possible in several areas. Part of the FMD is the inclusion of 2D barcodes by marketing authorization holders [64]. Although this directive should improve traceability, we suggest further improving the system by providing duplicated removable labels that are easily placed into patient notes after administration of the medicine to capture the brand name and barcode. Furthermore, through this method, manufacturers of biologics could better screen for accuracy when ADR reports are submitted (i.e. call center receipt, subsequent follow-up).

The European Commission is responsible for implementing the decisions of the European Parliament and the Council of the European Union and can enforce and support pharmacovigilance legislation through education of healthcare providers and Member State agencies [36,74]. We recommend professional associations participate in the training of healthcare providers. Such training could include identification of delayed immune-related events and national regulations surrounding switching. Training should also cover key steps and processes on how to report on and what information to document in an ADR report, including necessary unique product identifiers to support accurate tracing. Education of physicians on ways to report ADRs to manufacturers is important to this process. We also recommend that patients and patient advocates receive education on biologics, including biosimilars, and the steps and requirements involved in reporting suspected ADRs. Specifically, patients and their advocates would benefit from understanding the unique features that distinguish biosimilars from generic small-molecule drugs and the importance of pharmacovigilance, including the steps and processes involved in ADR reporting, and what data are important. We recommend that national regulatory agencies for each Member State support education at the national level and ensure that appropriate systems are in place to capture the pharmacovigilance data. Manufacturers have a responsibility to clarify with ADR reporters the specific manufacturer and batch of the product being reported.

Potential ways to improve pharmacovigilance should be considered at future EMA pharmacovigilance stakeholder meetings. Such considerations should include methods to lower rates of ambiguous reporting, including rapid advancement of Member State electronic systems to ensure accurate attribution of ADRs. Because of the potential cost savings offered by biosimilars, payers have good reason to promote their safe and appropriate use. Therefore, we encourage payers to provide pharmacovigilance education to prescribers and patients; such education could come through patient advocacy groups focused on biological medicines. The use of health technology assessments, which provide information on both budget impact and safety [75], is also recommended to encourage payers’ understanding of the importance of pharmacovigilance for biologics, including biosimilars.

4. Conclusions

The EU recognizes the need for product-specific pharmacovigilance for biologics and implemented a legislative approach to guide the EMA and Member States on requirements for filing ADR reports for biologics. The EMA has been an integral force in facilitating pharmacovigilance of biologics, including the release of guidelines on good pharmacovigilance practices and the convening of stakeholder meetings to review progress and enhance the pharmacovigilance system. Although some Member States have implemented significant changes to comply with the legislation, shortcomings have been observed in various studies of pharmacovigilance databases and in inconsistencies in the implementation across Member States. To resolve such issues, the EMA is actively improving pharmacovigilance in the era of biosimilars, with various initiatives, including the FMD, SCOPE, and improved patient registries for developing best practices across all Member States.

5. Expert opinion

The EU has displayed global leadership in passing legislation and bolstering pharmacovigilance practices for biologics in the era of biosimilars, and we could not catalogue all efforts here. However, Member State implementation of the EU pharmacovigilance legislation may still require greater attention to improve ADR reporting practices to comply with the law, as has been observed through numerous studies. A first key step toward harmonization among Member States is to develop a list of products that should be considered as biologics across all pharmacovigilance organizations. Biologics, including biosimilars, are an important and growing class of therapeutics, and it will be important to be able to accurately trace multisource products. Implementing measures to improve pharmacovigilance, such as the use of distinguishable nonproprietary names, will increase confidence in the safety of biologics, including biosimilars. When properly implemented, pharmacovigilance will improve traceability of ADRs without reducing patient access to care.

Appropriate pharmacovigilance measures should be put in place to ensure that ADRs are correctly attributed to the responsible medicine at national and regional levels. Systems that allow patients, caregivers, and healthcare providers to completely follow the EU pharmacovigilance legislation are needed in many areas. Until such measures are fully implemented, steps should be taken to ensure clear and accurate product traceability through the use and capture of distinguishable brand names, INNs, and batch numbers per current EU legislation. In addition, we encourage further education of stakeholders, including healthcare providers, patients, patient advocate groups, and payers, in the importance of reporting complete and accurate data for ADRs. In the long-term, advances in technology (e.g. use of barcodes at the point of administration and data linkage) hold the promise of establishing high accuracy regarding the specific medication dispensed or administered to patients. The use of a universal barcode system could have the potential to support robust product identification for pharmacovigilance purposes and transfer of information. As healthcare costs in the EU increase, biosimilars will be the mainstay for many governments’ strategies, and it will be important to enact more robust legislation to make it easier to have information on product identification.

Further research is needed to examine how requirements and naming conventions affect safety monitoring by both manufacturers and physicians. For example, how effectively did the amendments to pharmacovigilance legislative framework rationalize and strengthen pharmacovigilance in the EU? Specifically, what are the measurable effects in terms of protection of patients/consumers; clear definition of the roles and obligations for responsible parties; reduction of the administrative burden on manufacturers’ pharmacovigilance systems; and assurance of proper risk management through the accurate reporting of spontaneous ADR? By reviewing the systems in the US, Japan, and EU in the future, a comparison of different regional efforts to promote product-specific pharmacovigilance could serve as a global reference for other regulatory authorities.

Article highlights

• Biosimilars are highly similar to, but not exact copies of their reference products.

• Given the potential for immunogenicity, special considerations are necessary with respect to pharmacovigilance of biologics, including biosimilars.

• The European Commission (EC) has provided legislative guidance to the European Medicines Agency (EMA) and EU Member States on requirements for filing adverse drug reaction (ADR) reports for biologics.

• Inconsistencies and shortcomings exist in several Member States’ pharmacovigilance systems, such as incomplete brand name and batch number reporting for biologics.

• Some EU Member States have implemented changes to their existing pharmacovigilance systems to comply with the legislation and to help address these inconsistencies and shortcomings; however, improvements are still needed.

• The EMA is working to actively improve pharmacovigilance in the era of biosimilars through initiatives such as the falsified medicines directive and SCOPE (Strengthening Collaboration for Operating Pharmacovigilance in Europe).

This box summarizes key points contained in the article.

Declaration of interest

TF, JBJ, and PB are employees of, and own stock in, Amgen Inc. CA was an Amgen employee when this work was initiated. 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. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

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

Acknowledgments

Medical writing support was provided by Miranda Tradewell, PhD, and James Balwit, MS, CMPP (Complete Healthcare Communications, LLC, North Wales, PA, an ICON plc company), whose work was funded by Amgen Inc.

Additional information

Funding

This work was supported by Amgen Inc.