Report from the DIA’s Pharmacovigilance and Risk Management Strategies Conference, Washington, DC, 22-24 January 2018
Despite the US government shutdown resulting in a noticeable absence of FDA personnel, the DIA’s Pharmacovigilance and Risk Management Strategies conference was a well-attended event. The conference was hosted in Washington, DC, from 22-24th January 2018. As usual the programme was excellent, extensive and covered many hot topics; a recurring theme was how quickly some players in the PV industry are adapting to modern-day challenges including newer technologies for treatment, so-called big data, seemingly incessant social media and the rise of artificial intelligence.
Although not able to attend in person, we were still presented with the updates from the FDA representatives thanks to the wonders of technology. I was particularly impressed with the extent of FDA’s recent achievements and developments showcased, which included application of the Sentinel system, enhanced public access to FAERS data, analysis of social media as a pharmacovigilance source, and evaluation of machine learning and other technologies. Also remarkable was the 1.8 million adverse events reported to FAERS in 2017, bringing the total number of reports to 14 million. No wonder initiatives such as automation and the Sentinel system are required to help get a handle on the vast volumes of data in order to find the proverbial needle.
The session by John Peters and Howard Chazin highlighted some of the potential safety issues unique to generic drugs; they explained how not all drugs are equal in this regard even when shown to be bioequivalent, therapeutically equivalent and pharmaceutically equivalent. In fact there is a specific department at the FDA dedicated to generic drug surveillance, which focuses on issues such as lack of efficacy, product quality, packaging, off-label use and new safety concerns not previously seen with the reference listed drug.
The new ICH E19 guideline on optimising safety data collection was outlined by Ellis Unger. Interestingly, this guideline may seem like a paradigm in the world of ‘big data’ that we live in, as the proposed concept is that of a targeted approach to collecting safety data rather than collecting all and every piece of information that might one day be relevant. Not only would this reduce the burden for industry, but also the burden for patients; it was postulated that such an approach could encourage patient participation in clinical research. It was encouraging to hear that the FDA already seem to be backing this concept, and by the looks of the Expert Working Group members, so are many other regulatory agencies. As with all ICH guidance, adoption across multiple regions will make for an optimal effect.
Even if ICH E19 is widely adopted in the future, the volume of ICSRs reported is still growing year on year. What seems to be a growing trend to address with this challenge is to supplement the human workforce with a digital workforce. The session on machine learning and artificial intelligence set out to provide perspectives from industry, regulators and technology companies on the use of these technologies in pharmacovigilance. Although there was a distinct difference between the FDA who are embracing AI and the MHRA who admitted they are still on the fence, the overall audience expectations were that some degree of artificial intelligence was inevitable in the next 2-10 years. Encouragingly for pharmacovigilance professionals, there seemed to be general agreement that machines would not completely replace humans in the pharmacovigilance world, and that it would be ideal for regulators and industry to collaborate as we head down this path.
Other highlights from the conference were the regional updates. The UK MHRA discussed the implications of Brexit on PV processes, including the possible need for a UK QPPV. It seems that pharma are still waiting to see what they will need to adapt and to what extent, similar to many industries outside of pharma. The ideal scenario would be to have a co-operative agreement rather than a completely independent regulatory system, but there are still a lot of unknowns.
Both the Japanese PMDA and Health Canada emphasised the need to increase the use of real world data as a source of safety information, and this will be addressed as change to PV guidelines (Japanese GPSP) and regulations (e.g. Vanessa’s Law) are implemented. And the EurAsian Economic Union announced the publication of Good Pharmacovigilance Practice guidelines.
So I left the conference armed with broader insights into some exciting regional developments and achievements, but also with the mind-set that we need to keep on top of these many changes to ensure we are not the ones left behind on the porch.
Image credit: Patrick Hendry