Reader Comments

Post a new comment on this article

Some comments to the outcome and the conclusions of the analysis

Posted by jalisman on 02 Apr 2013 at 15:10 GMT

The importance of research on medicines regulation cannot be overestimated. Regulatory science can provide important input for the public debate about what requirements are necessary and what level of evidence society requires before a medicine is allowed on the market. Therefore, I read the research article ‘Number of Patients Studied Prior to Approval of New Medicines: A Database Analysis’ by Duijnhoven et al. with great interest. However, although I appreciate the study to a large extend, I feel that when the authors conclude that 80% of medicines approved for chronic use meet ‘the minimal requirements’ (and, conversely, 20% not) an important point is missed and this requires a further discussion of the data.

I have two main comments. Firstly, a subgroup analysis in which the medicinal products are placed in cohorts of the year of authorisation gives an interesting insight. Of the 16 products that not meet ‘the minimal requirements’ six are in the first two years of the eleven year study period (2000-2001), and ten are in the first five (2000-2004). The last three years studied (2008-2010) only saw one product not meeting ‘minimal requirements’ as defined by the authors. This time trend warrants discussion.

Secondly, and more importantly, although the authors refer to the improvements that new legislation on pharmacovigilance might bring to the system, they have not discussed the (new) regulatory instruments the EMA had at its disposal during the research period to deal with applications for which no complete dataset is yet available: the ‘Exceptional circumstances marketing authorisation’ and the ‘Conditional marketing authorisation’. The latter was implemented in 2006.

When approving a medicinal under exceptional circumstances or granting a conditional marketing authorisation this approval is accompanied by a commitment letter of the applicant, stating which clinical trials will be conducted or finalised after the approval of the product. Table 1 (below) lists all 16 medicinal products for chronic use that did not meet ‘minimal requirements’ according to the authors. As the Table shows, for many of the products, appreciation of the specific cases may shed a different light on the conclusions by the authors. For example, 6 of the 16 products were for HIV. In the study, HIV/AIDS infection is considered as a chronic disease. This seems to us to depend on the date the application for the concerned medicinal product was submitted: in the beginning of this century HIV/AIDS was still a life threatening condition. Moreover, several ‘non-complying’ medicinal products were approved under exceptional circumstances or with a conditional market authorization. It is fair to say that in those cases the EMA was aware of the gaps in the clinical dossier, but accepted completion at a later point in time in the light of the medical need. Lastly, two of the identified medicinal products are not or hardly absorbed, which is a factor that decreases the chance that systemic adverse drug reactions occur.

See table here: http://www.plosmedicine.o...

In conclusion, although the research presented in this study is relevant for regulatory science, I believe that the interpretation of the results requires more appreciation of the context in which decision-making about the market authorisation for medicinal products takes place.

John Lisman

No competing interests declared.

RE: Some comments to the outcome and the conclusions of the analysis

mldebruin replied to jalisman on 05 Jun 2013 at 12:20 GMT

We are grateful to John Lisman for reinforcing the importance of research on medicines regulation, and the need for public debate about the level of evidence society requires before a medicine is authorised.

Lisman draws attention to the medicines (intended for chronic use) which did not measure up to the ICH E1 criteria and argues that there are several explanations to justify the limited number of patients included in the premarketing studies. We would like to take the opportunity to respond to his comments.

Lisman states that in our study 80% of products for chronic use meet the ICH E1 criteria ‘and, conversely, 20% not’; this is not the case. In fact only 8% of the products did not meet the criteria, and there is also a large group of drugs for which no data could be obtained. This is presented in table 2 of the article as well as in the text.

As Lisman notes, all of the products in the study, whether meeting guideline criteria or not, were approved and, thus, were considered to have a positive benefit/risk balance in view of the European regulatory agency. Apparently, the limited size of the trial population and the short duration of treatment, e.g. to assess longer-term risk, were offset by well-established beneficial effects and/or medical need. The underlying reasoning, was, however, not studied in our project.

In the article we stress that not only efficacy needs to be studied thoroughly, but that benefit-risk decisions clearly require adverse effects to be studied adequately as well. Deciding on the extent to which this should be done before approval or after approval (when permissible) is a challenge for regulators. Even for medicines which have been studied in large numbers of patients before approval uncertainties will always remain. This is why we chose not to discuss individual products in detail: it would be just as appropriate to discuss those that have been studied in very large numbers of patients. Rather than simplistically categorising the products as compliant or non-compliant with to the ICH E1 guideline criteria we chose to present the actual numbers of patients studied along with the guideline thresholds to provide a comprehensive overview.

We hope to have made clear to all stakeholders that knowledge of a new medicine is inevitably incomplete at the time of approval and therefore proactive pharmacovigilance and epidemiological studies are of crucial importance.

On behalf of all authors
(Ruben G. Duijnhoven, Sabine M. J. M. Straus, June M. Raine, Anthonius de Boer, Arno W. Hoes, Marie L. De Bruin)

Competing interests declared: RGD and MLDB have received financial support from the Medicines Evaluation Board (MEB) under the Regulatory Science collaboration between the MEB and Utrecht University for the submitted work. SMJMS is employed by the MEB. AWH is a member of the Dutch Medicines Evaluation Board but received no funding. AWH was not involved in the decision about the funding. AdB and JMR have no relationship with the MEB. The department of Pharmacoepidemiology and Clinical Pharmacology, Utrecht Institute for Pharmaceutical Sciences, Utrecht University, of which AdB is the chair and RGD and MLDB are employees, has received unrestricted research funding from the Netherlands Organisation for Health Research and Development (ZonMW), the Dutch Health Care Insurance Board (CVZ), the Royal Dutch Pharmacists Association (KNMP), the private-public funded Top Institute Pharma (, includes co-funding from universities, government, and industry), the EU Innovative Medicines Initiative (IMI), EU 7th Framework Program (FP7), the Dutch Medicines Evaluation Board, and the Dutch Ministry of Health and Industry (including GlaxoSmithKline, Pfizer, and others). The authors declare no other competing interests.