Surveying adverse event rates: lessons from a virosomal hepatitis A vaccine

Keywords::

• adverse events
• frequency rates
• postmarketing
• risk perception
• safety assesment
• virosomal hepatitis A vaccine

Methods of surveying, assessing and communicating adverse events (AEs) of medicinal products have undergone tremendous changes during the last decades. The vast majority of these mainly regulatory-driven changes were improvements for the benefit and safety of the users of these products [1]. Vaccines are under special scrutiny as they are mainly given to healthy subjects. It is of great importance that the communication of AE risks related to vaccine usage is medically sound and comprehensible so that compliance with using vaccines is not jeopardized.

During over 15 years of involvement in the development of vaccines for a European company, I came across some regulatory safety aspects, which in my opinion are in conflict with common sense. Based on the clinical pre- and postlicensing experience with a virosomal hepatitis A vaccine (HAV; Epaxal^®, Crucell BV, Leiden, The Netherlands), I would like to call attention to two such regulatory aspects: the safety profile imposed for the product's life span based on limited, mainly solicited prelicensing safety experiences and the AE risk frequency categories to be communicated by law in the product information leaflets.

Virosomal hepatitis A vaccine

Epaxal is an aluminum-free hepatitis A virus (HAV) vaccine, and consists of inactivated HAV bound to influenza virosomes [2]. It has proved to be efficacious [2], highly immunogenic and well tolerated in both adults and children [3–5], and to be associated with significantly fewer local AEs when compared to aluminum-adsorbed HAV vaccines [4–8].

Licensed safety profiles based on limited experience & cut in stone for the product's life span

The safety profile of a vaccine, as published at the time of marketing, is based on a limited set of prelicensing data. There is no question that during its life cycle this profile should be revised as soon as new safety concerns arise, either from new clinical studies or from postmarketing surveillance [1].

The main ‘bulk’ of the original safety profile usually consists of solicited, unspecific systemic AEs (headache, malaise, anorexia and so on) and solicited injection site reactions, as reported in the clinical development studies. What disturbs me is that this first profile is later never fundamentally reassessed and revised based on the postlicensing experience. Instead, new AE terms are added every now and then to the package insert, but the ‘20% headache’ and ‘30% fatigue’ from the Phase II studies remain forever associated with the product.

After licensing the virosomal HAV in Europe in the late 1990s, I was rather discontent with the imposed initial safety profile, which attested this well-tolerated vaccine in the summary of product characteristics (SPC) as not more than an ‘acceptable’ tolerability based on the prelicensing AE risk frequencies (see the following discussion on qualitative description of AE risks). In addition, the SPC information made Epaxal look worse regarding injection site reactions in comparison to an aluminium-adjuvanted competitor, although direct comparisons prelicensing had shown the opposite [6], and this was confirmed later on repeatedly [4,5,7,8]. Consequently, I initiated a large multicenter postmarketing safety study in over 2500 travelers who were aged 2–79 years, aiming to challenge the validity of the incidence rates of each single AE term listed in the SPC: anorexia, diarrhea, dizziness, fever (>37.5°C), headache, malaise, nausea, skin rash, vomiting and injection site pain, swelling and redness [9]. The objective was to assess the safety of a single dose of Epaxal based on the comparison of AE reporting rates (up to 4 days after vaccination) generated by solicited (enquiring about all the above-mentioned AEs) versus unsolicited (open) questioning. These postvaccination AE rates were also compared with the baseline systemic AE rates (within 24 h before vaccination) regarding the same AE terms. As expected, soliciting AEs ‘produced’ significantly higher incidences of systemic AE reports, for example, 6.4 versus 0.1% anorexia or 15.6 versus 4.6% headache, the corresponding baseline rates being 1.5 versus 0.0% and 10.0 versus 0.9%, respectively. Similar to the subjective systemic AEs, soliciting of the physical signs of injection site swelling and redness resulted in considerable reporting bias, that is, 9.5 versus 1.7% and 7.2 versus 1.0%, respectively.

The conclusions from the study were – among others – that there is a need to develop adequate methods to separate background incidence of health events from the attributable postvaccination events, that it is advisable to list in publications both solicited and unsolicited reporting rates and disclose the AE collection methods in more detail and that the safety information from Phase IV studies should (by the licensing authorities) be considered part of the regular SPC updates and revisions [9].

Frequency categories for communication of AE risks in conflict with popular perception

To make informed decisions about accepting a vaccination for themselves or their children, people need accurate and easily understandable information about possible vaccine-associated AEs and the likelihood of their occurrence, because the perception of communicated health risks may affect medical decisions and cause health behavior changes [10,11].

In 1998, the European Commission introduced a set of qualitative frequency descriptions for the AE section of the patient information leaflets of medicinal products that range from very rare (<0.01%), rare (0.01 to <0.1%), uncommon (0.1 to <1.0%), common (1.0 to <10.0%) to very common (≥10%) [101]. They did so presumably with the intention of helping lay persons to better understand the possible health risk, assuming that such vague verbal labels are better understood than numerical information. These proposed frequency descriptions are, however, in conflict with popular understanding [12]. For example, are headache and redness at the injection site occurring at rates of 4.6 and 1.0%, respectively (unsolicited AE reporting rates) [9] really ‘common’ events? Certainly not in the mind of a lay person. These inadequate qualitative descriptions of AE risk frequencies were one of the reasons for me to initiate the above-mentioned Epaxal postmarketing safety study [9].

Attentive investigators assessed the general public's interpretation of the verbal descriptors for AE frequency as officially recommended for use in medicine information leaflets [101] and examined whether differences in interpretation affect people's perception of risk and their judgments of intention to comply with the prescribed treatment [10,13]. The results showed that qualitative descriptions of AE health risks led to gross overestimation of risk [12] and that “the ‘overestimation’ resulted in significantly increased ratings of perceived severity of side effects and risk to health, as well as significantly reduced ratings of intention to comply, compared with those for people who received the probability information in numerical form” [13]. The researchers urged that “the European and national authorities should suspend the use of the EU recommended terms until further research is available to allow the use of an evidence-based approach” [13]. As far as I am aware, nothing changed at the licensing authority level regarding the AE frequency descriptors. Although the 1998 European Commission guideline [101] was revised in January 2009 [102], the same qualitative frequency descriptions are still used, now under the heading ‘MedDRA (Medical Dictionary for Regulatory Activities) frequency conventions’ [103].

However, and this is a ray of hope, the US FDA and the US Department of Health and Human Services published a very well-referenced document on the scientific foundations for effective communication last year that was targeted at nonexperts, Communicating Risks and Benefits: An Evidence-Based User's Guide [104]. One of the key statements in the section on ‘quantitative information’ stresses: “Provide numeric likelihoods of risks and benefits. Describing risks solely with words, such as ‘you have a low chance of experiencing a side effect’ is ineffective. It does not provide patients with the details needed to make an informed decision; it increases risk perceptions, and patients vary in their interpretations of what low and high risks are. Thus, it is imperative to provide patients with numerical estimates of the risks and benefits associated with treatment options”. Furthermore, the authors comment in the introduction that “sound communications must be evidence based in two related ways”. One is that “communications should be consistent with the science…” and the other is “communications should be evaluated, because even the best science cannot guarantee results”, something the European Commission apparently did not do before introducing the 1998 guideline containing obviously unvalidated qualitative descriptors of AE frequencies.

Conclusion

When reading for the preparation of this editorial from a recent review on “Strengthening and rationalizing pharmacovigilance in the EU: where is Europe heading to?” [14], I realized that the two regulatory safety concerns I am voicing in this editorial are only tiny aspects of the huge European pharmacovigilance legislative framework. The authors stress that the new legislations will improve numerous uncertainties in the current legislative framework and will provide – among others – for “strengthening medicine safety transparency and communication so that the understanding and trust of patients and health professionals in the safety of medicines will improve”.

The above-mentioned FDA document will help to revive the discussion on risk communication at the European Medicine Agency, MedDRA and other relevant bodies [104]. I hope that the leaflet information guidelines [101,102] will be fundamentally revised regarding the AE risk communication subsequently and, of course, also all the safety information in the SPC and patient information leaflets.

Financial & competing interests disclosure

The author worked for Crucell Switzerland AG (formerly Berna Biotech/Swiss Serum and Vaccine Institute from 1994 to 2011), first as Medical Director and later as Medical Advisor. He currently works as an independent biopharmaceutical consultant. He does not hold any shares in Crucell and this editorial is written independently from the company. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.