Signal Management in Drug Safety

OVERVIEW

As per GVP Module Ⅸ, a signal is defined as “Information arising from one or multiple sources, including observations and experiments, which suggests a new potentially causal association, or a new aspect of a known association between an intervention and an event or set of related events, either adverse or beneficial, that is judged to be of sufficient likelihood to justify verificatory action”

To put in simple terms signal in pharmacovigilance (PV) is any new information that

• Suggests a new causal association or
• A new aspect of a known association

Signal can be either beneficial or adverse. Adverse signals, often called as safety signals are an important part of pharmacovigilance system.

Signal management is considered as a critical process in pharmacovigilance and to ensure the up-to-date safety profile of a medicinal product.

Source of signals

Signals can arise from any source of data that a PV system encompasses of. The common sources of signals are:

• Spontaneous reporting systems – Individual case study report (ICSRs) that are databased,
• Active surveillance systems – Sentinel sites, Registries,
• Quality defects/complaints,
• Clinical data including pharmacoepidemiology data,
• Scientific literature,
• Digital media, etc.

Signal management process

As a Marketing authorization holder (MAH) of pharmaceutical products, companies hold the responsibility to have an organized signal management process within their PV systems.

It includes various processes starting from signal identification to its closure. These are:

• Signal detection
• Signal validation
• Signal prioritization
• Signal assessment
• Recommendations for action

Individual organisations may follow alternative signal management processes and terminology but should adhere to the general principles of signal management.

Signal detection

This involves the identification of signals through one or more of the below methods.

• Traditional – Case and case series review – This involves review of spontaneous reports and other post marketing adverse events (AEs). The ‘index case’ or a ‘sentinel case’ which strongly supports causality of an AE with the drug can lead to the identification of a signal.
• Traditional – Analysis of larger datasets – Involves analysis of line listings, cumulative tables or periodic safety reports. Changes in frequency, distribution, duration, severity or outcome of AEs are looked into.
• Statistical methods – These methods are generally employed when the database to be analyzed is too large for inspection of individual cases. These methods focus on groups of cases to identify high proportions of specific AE with a specific medicinal product as compared to the reporting of this event for all other medicinal products (disproportionate reporting).
• Disproportionate reporting – Ratio of proportion of ICSRs of a specific AE in the presence and absence of a specific drug. A high value of the ratio in a particular Drug-Event combination (DEC) suggests further investigation.
• Methods aimed at specific groups of AEs – A best example of such groups is Designated Medical events (DMEs). Some AEs are known to be caused due to inherent nature of the drug and draw attention of signal assessors.

Signal detection should follow a methodology which takes into account the nature of data and the characteristics (e.g., time on market, patient exposure, target population) as well as the type of medicinal product concerned (e.g., vaccines and biological medicinal products).

Signal validation

Also known as signal evaluation, aims at determining whether there is sufficient evidence pointing towards a new possible causal association or new aspect of a known association (e.g., increase in frequency or severity) in the detected signal, which justifies further analysis of the signal. Evaluation of a signal is done based on following parameters:

• Previous awareness – Information that is already knowns regarding the detected signal and which is included in the product information (summary of product characteristics (SmPC) and package leaflet) and information regarding any previous assessment of the association.
• Strength of evidence – To identify the strength of the available data pointing towards possible causal association based on: Quality of the data, consistency of evidence across cases, disproportionality of reporting, number of cases showing temporal association with positive dechallenge/rechallege, reporting rate etc.
• Clinical relevance – To identify the clinical importance of the association e.g., seriousness and severity of AE, outcome and reversibility of an AE, events occurring in vulnerable population etc.

Possible outcomes of a signal validation can be:

• Validated – sufficient evidence to proceed to further assessment
• Non-validated – available data does not point towards a possible causal association; hence, no further assessment is warranted
• Inconclusive – Existing data is not sufficient to take a further decision. In these scenarios, the MAH/regulator actively monitor the signal to see if any new information comes up in future warranting further assessment.

Signal assessment

Signal assessment aims at further evaluation of the signal to identify the need for a regulatory action or risk minimization action.

This includes comprehensive review of data from various additional sources to obtain further information on the validated signal and to establish causality of the AE with the drug. This takes into consideration various other sources such as

• clinical trial data
• Findings from scientific literature
• Non-clinical findings
• Other publicly available safety databases e.g., Eudravigilance, FAERS, VigiBase etc.
• healthcare databases that may provide information on characteristics of exposed patients and medicines utilization patterns;
• information from other regulatory authorities worldwide

Possible outcomes of a signal assessment can be:

• New or changed Risk – for signals where a causal association or new aspect of known association is established.
• Refuted– for signals where a causal association cannot be established based on the review of available data.
• Inconclusive – Existing data is not sufficient to confirm the causal association; however, event is clinically important and does not allow to refute the signal at that point of time. In these scenarios, the MAH/regulator actively monitor the signal to see if any new information comes up in future warranting further assessment.

Signal prioritization

Signals with significant impact on public health or benefit-risk profile of the drug are prioritized. The timeframe for the further management of signal will depend on prioritization. Signal prioritization should be done at all stages of the signal management after detection.

Signal prioritization takes into consideration various aspects of the AE and drug involved in a signal to identify the impact. Such as:

• Severity, seriousness, outcome, reversibility and preventability of an AE
• Patient exposure and frequent of the AE
• Potential of the signal to be applicable to other drugs of same class

Recommendations for actions

Recommendations for actions are made for signals for which a causal association is established after the signal assessment. These recommendations include but not limited to:

• Update of product information
• Further studies – preclinical, clinical or epidemiological
• Additional risk minimization measures
• Periodic review of the safety issue (as an identified or potential risk)
• Regulatory actions (notifying authorities, recall of products etc.)

The timelines for the initiation and completion of the activities are then determined and the actions are executed.

Exchange of information

The exchange of the information regarding the validated signals or the recommendations for actions occurs at various levels such as

• Regulatory authorities
• Health care professionals and patients
• License partners
• Other MAHs of same drug or same class of drug

A successful signal management process is key for the timely identification of the safety risks and mitigating them resulting in minimizing the public health impact and maintaining the positive benefit-risk profile of the drug. This can be achieved by following the practices such as

• Documentation – All the process of signal management should be well documented by the MAH. This includes proper tracking and documentation of the strategy, actions and decisions. As per GVP module IX, an MAH is legally bound to put in place a record management system for all documents used for pharmacovigilance activities that ensures the retrievability of those documents as well as the traceability of the measures taken to investigate safety concerns, of the timelines for those investigations and of decisions on safety concerns, including their date and the decision-making process.
• Transparency – Transparency in terms of signal detection and assessment
• Effective communication – Timely communication of the information to regulatory authorities, HCPs and other MAHs.

References

1. EMA Guideline on good pharmacovigilance practices, GVP Module IX and Module IX Addendum I
2. FDA Guidance for Industry: good pharmacovigilance practices and pharmacoepidemiologic assessment