Safety used to be widely seen as a 'sales prevention' tool. Thankfully that has changed recently. But is it still viewed the same way in clinical trials? The Switzerland-based working group Council for International Organisations of Medical Sciences (CIOMS) recently produced a paper on this subject, discussing what pharmaceutical companies can do to embrace drug safety in clinical trials. CIOMS says that there is a need to incorporate newer approaches for managing safety information in the clinical trial setting, and to adapt the methods and tools used in post-approval pharmacovigilance to the early and late stages of pre-approval development of medicinal products.

The implications of a new approach to safety could be seen as onerous for a lot of stakeholders, including investigators, ethics review committees and, most importantly, pharmaceutical sponsor companies and their CROs. But they could also present an opportunity to streamline business, and more importantly, safety practices. CIOMS suggests that sponsor companies and CROs use the same methodology and tools in early and pre-approval development as in post-approval pharmacovigilance. On paper this is an excellent suggestion, but how easy is it in practice?

Collection of safety data
The collection of the correct data is obviously paramount and it has always been the practice that "more is better." CIOMS suggests that during development trials all adverse events should be collected; in the post-approval period, however, when the safety profile of a drug is better understood, it could be acceptable to collect less data. Thus, if study monitors were trained to look not only for adverse events but also at the type of data that is useful, without making the ultimate decision of whether to report all data elements, the amount of data analysed could be reduced and the data available would be of more use. This refined training also would allow the investigator to understand better what constitutes 'good data' rather than just data.

Investigators obviously need to inform a sponsor immediately of a serious adverse event (SAE). However, CIOMS suggests that investigators make a simple judgement on whether or not an SAE is related to the study drug. If however there are incidents of minor ailments such as headache and a subject is hospitalised because of a very bad headache (hospitalisation is a SAE) then it is reasonable to assume that it could be related to the study drug.

Whilst the investigator cannot be the arbiter of what caused the SAE, this is a step forward and can help further investigation.

Safety data management
As mentioned previously the collection of every piece of data imaginable is of no use unless it is well documented in a format that allows proper reporting and analysis. CIOMS makes several suggestions, which, in essence, are easily followed; however, how a sponsor implements these suggestions is more challenging. Consideration should be given to all or most of the following:

• How is data currently collected? There are several proven systems available, and there is also the option to build one's own.
  
• Who collects this data and inputs it into the chosen system?
• Do sponsors outsource the collection and or processing of clinical trials safety data, does their own safety department collect it themselves or does another arm of drug safety collect the data?

Obviously if harmonisation and the collection of better data is the desired result then it would be better to have both clinical and post-marketing safety under the same wing. This is not always easy or practicable, so does that point to outsourcing? Again, this is not always easy or practicable but maybe a combination of the two is more easily managed.

Regulatory reporting
Guidelines exist for the reporting of safety information from clinical trials, as they do for all safety reporting, and are generally adhered to whatever interpretation is made of them. CIOMS has made several suggestions, none of which if adopted should cause concern to stakeholders. In fact they will produce better, cleaner data that will help significantly in the promotion of safe drugs.

Until the EU Clinical Trial Directive, few authorities (EU, Switzerland and the US) required the periodic communication of safety information from clinical trials. Even now there are different formats and indeed different sorts of information required. CIOMS recommends defining a single Development Safety Update Report (DSUR) for submission to regulators annually. Moreover, they recommend that these reports be based on a whole development programme rather than per protocol. They also suggest for products with a well-established safety profile that the DSUR is replaced by a PSUR (Periodic Safety Update Report). The question remaining is who writes the report? Does it remain in house, or because of the collection of clean relevant data, does it lend itself to outsourcing, which could be a more cost effective solution for what is considered an onerous task.

Good pharmacovigilance practice
Whilst pharmacovigilance has generally been used in reference to post-approval drugs, CIOMS suggests that it be used to cover drug development too.

However, good practice is not only there to be followed, but also to be used as guidance for better practices to be developed. The better and more refined the practice, the better the conduct of the trial, the better the data collected and ultimately, the greater the knowledge a sponsor will have about its products.

So how do you ensure a flexible programme that encompasses the diversity of products but is also able to deliver the end result required? SOPs (standard operating procedures) should be broad enough to allow the flexibility but still gather the information required. CIOMS recommends forming a dedicated safety management team that will review all available (not just post-marketed or clinical) safety data. This team can be adjusted in size relevant to the size of the organisation: a start-up with one development product can use this approach, with outsourced help, just as well as a large pharmaceutical company. In many cases, it will be easier for the smaller company, as people have many different roles to play, whereas with large pharmaceutical companies it is necessary to achieve the right balance.

Conclusions
Generally, the recommendations from CIOMS are practicable. With some adaptation from stakeholders, they could result in a more streamlined approach, and the ability to choose where to collect and store data, when to use internal or external resources and, most importantly, how to use that data to best serve the population the drug is in line with the needs of the pharmaceutical company whose time and effort has gone into developing the drug in the first place. The CIOMS recommendations should not be viewed as adding to the workload of drug safety, but as a step towards better and safer promotion of drugs.

About the author
Michael Reynolds is head of Sales and Marketing at APCER Pharma, a global provider of integrated marketing, regulatory and pharmacovigilance services to pharmaceutical companies.