Requirements of an EU PV system – Signal Management

Challenges Faced by Marketing Authorization Holders in Signal Management for Pharmacovigilance Systems

A critical part of any pharmacovigilance (PV) system is the process of signal management. As per article 19 of the Commission Implementing Regulation (EU) No 520/2012, Marketing Authorization Holders (MAHs) are legally required to identify new or changing adverse events related to their medicinal product or Active Pharmaceutical Ingredient (API). 

The challenge comes when deciding on how to identify, validate and act on these new adverse events in an appropriately scientific manner.

Current MAHs Issues

Many MAHs are faced with the following issues: 

• Huge quantities of safety data, often in the form of spontaneous Adverse Drug Reports (ADRs) or as Individual Case Safety Reports (ICSRs), are collected or available that can vary greatly in quality.
• The data being collected is decentralized and does not have the luxury of being collected from a clinical trial where variables can be tightly monitored and compared to a control.
• EU guidelines on signal management leave room for interpretation and innovation, as they often do. This can be difficult for untrained personnel who are looking to transpose these guidelines into their own functioning PV and signal management systems.

With this in mind, this article aims to briefly demonstrate signal management flow within the EU by using the example of a MAH with a nationally authorized product on the EU market who was previously out of compliance because they did not carry out any signal management obligations, nor were changes made to their pharmacovigilance system to rectify this.

The general flow of signal management is as follows:

Signal detection 

Signal detection encompasses the processes related to the initial capturing of trends in safety data that indicate a particular drug is causally related to an adverse event. Methods for signal detection can be divided into two broad subcategories: disproportionality methods that rely on statistical detection algorithms, and the clinical review method which is more commonly used when few reports have been received for a product. Signal sources can be varied and include spontaneous reports, periodic safety update reports (PSURs), pre-clinical and clinical studies and medical literature. Within the EU, signal detection is greatly aided by the presence of EudraVigilance (EV), the EUs safety data processing system, and more specifically, the EV Data Analysis System (EVDAS). Registration for the EV platform is a legal requirement for EU MAHs, as all adverse drug reactions (ADRs) received must be reported as ICSRs through the EVWEB platform, where they are made accessible to other MAHs through the aforementioned EVDAS platform. 

In the case of the MAH that forms the subject of this article, safety data was being collected in a manner non-compliant with EU GVP guidelines. 24/7 ADR collection systems were not in place and medical literature monitoring was rudimentary and only extended to local literature. Critically, the MAH was registered on the EV platform, however the MAH was not utilizing the signal management tools available to them that were granted through it. EVDAS registration had not been carried out, therefore statistical options for signal detection were extremely limited, and only a very small safety dataset composed of directly received spontaneous ADRs were applied. 

In EV, safety data reported by MAHs across the EU can be found in one centralized location, and through EVDAS, can be analysed with much greater ease considering the built-in statistical tools and wide pool of validated data. Users of EVDAS are able to search for drug-event combinations (DECs) for potential signals, and then apply these statistical tools to the results. Among these statistical tools are: 

• Relative Odds Ratio (ROR) – The ROR is a disproportionality measure applied to DECs. Simply put, ROR is the odds of a certain medical event occurring with a medicinal product, compared to the odds of the same event occurring with all other medicinal products in the database. For example, if the ROR is equal to three, the odds of reports of this event being associated with your medicinal product are three times higher than the odds of reports of this event among all other reports in the database.
• Significant Drug Reaction (SDR) – For any particular DEC, if any one of a number of criteria are met, then it is classified as an SDR within the EV platform and recommended for further analysis.

These criteria are: 

• The lower end of the 95% confidence interval greater than one; 
• A total number of cases equal or above 5; 
• Being classified as an important medical event.  

In order to ensure compliance to PV standards, new procedures were put in place for the MAH to fully utilize the power of EV as a signal management tool. Registration on the EVDAS platform was carried out, and a master procedure for signal management was drafted that described the proper usage of the described statistical tools for signal detection purposes, including specified target values for DECs to be passed on to the next stage of signal management, such as number of reports received, seriousness, ROR value and SDR classification. More specific procedures were drafted to cover processes such as submission and export of ADRs as ICSRs through EV, global medical literature monitoring.

Signal Validation 

Once a DEC has been detected as a potential signal, it moves onto the signal validation stage. Signal validation strategies can be varied. In order to fulfil the example MAHs requirements towards signal validation, ICSR reviews derived from the EV database were added as a step for DECs identified as SDRs. Criteria for this review included an evaluation of the medical terminologies used in the report, the source of the report, the quality of the information within (duplicate reports, miscoding, etc.), whether a credible temporal relationship was included in the report, challenge/dechallenge/rechallenge attempts, and many more. Potential signals were then further validated through extended searches for DECs on PV databases other than EV. As a final validation strategy, any DEC identified as significant was queried on online medical literature databases using pre-defined search terms; medical literature is a high quality source of safety information. 

Communication of safety information 

Once a validated signal has been identified, the MAH is then obligated to inform both the European Medicines Agency and the relevant National Competent Authority where their product is authorized. Reporting timelines for the sample MAH were specified in the procedure: validated signals detected for particular special substances and products designated by the EU must be reported as standalone notifications within 30 days of signal assessment, as well as within PSURs submitted according to the EURD list and/or as safety variations. Safety issues that are considered urgent and may require prompt regulatory action must be reported as emerging safety issues within 3 days.

In conclusion 

Signal management can be a dense and complex subject that is often difficult to navigate. There is no one-size-fits-all approach:  signal management strategies are highly dependant on the nature of the product/substance that they will apply to. Quintian Pharma, as demonstrated in this article, is able to tailor-make and implement signal management strategies for all kinds of MAHs and products!