29 Aug 23
By Sarah Daniels
Less than six months after FDA’s approval of Syfovre (pegcetacoplan) earlier this year, seven cases of retinal vasculitis have been reported (1).
Syfovre (pegcetacoplan) is the first drug to treat geographic atrophy (GA), a sight-robbing condition that accompanies age-related macular degeneration (AMD).
No cases of retinal vasculitis were reported by the marketing authorisation holder (MAH) Apellis following the administration of 23,000 injections of the drug during clinical trials.
Whilst this is not the first – and certainly won’t be the last – time a new safety concern, never seen in clinical trials, appears when a product is first marketed (see Table 1), it is helpful to understand why this happens and how to manage the situation when it arises to minimise risks to patients.
|Safety concern||Drug/MAH||Mode of action||Indication||1st approval||ADR mechanism||Action|
|Churg -Strauss syndrome||Zafirlukast (Accolate)/||Leukotriene antagonist (1st in class)||Asthma||USA 1996||Not drug related||1997 warning|
|Severe liver damage||Troglitazone (Rezulin)/ Pfizer||thiazolidinediones.||Type 2 Diabetes mellitus||USA March1997||Idiosyncratic||2000 withdrawn|
|Progressive multifocal leukoencephalopathy (PML)||Natalizumab (Tysabri)/ Biogen Idec||Humanised monoclonal antibody||Multiple sclerosis||USA November 2004||Pharmacodynamic||Feb 2005 Withdrawn, reintroduced 2006 with restrictions|
|Thrombosis with thrombocytopenia syndrome (TTS)||COVID-19 vaccines (AZ, J&J)||Adenoviral vaccine||COVID-19 prevention||GB June 2021||Under investigation||Distribution paused prior to updated product information 2022|
The first thing to note, is that premarketing clinical trials are conducted in relatively small, homogeneous patient populations, and thus even a large phase 3 trial is unlikely to run into 10s of thousands of patients. Therefore, once a drug is approved for marketing and is used to treat a bigger and broader patient population, rare adverse events (AEs) may now be identified. This is why post marketing surveillance is absolutely critical.
Consider the Rule of 3: This states that you need three times as many subjects to observe an event when you assume that the AE of interest does not normally occur in the absence of the medication (2). According to this rule, if an AE occurs at a frequency of 1 in 10,000, 30,000 patients would have to be exposed premarketing to have a 95% chance of detecting a single case (see Table 2). Even then, chance dictates that some AEs may still not be detected.
Furthermore, the rule does not hold if the population after launch is wider than the clinical trial populations, when differences in intrinsic and extrinsic factors also affect the risk of the AE. Another consideration is that the AE may have actually occurred premarketing, but was not reported, or mis-identified, or not correctly ascribed to study drug.
|Expected incidence of the adverse reaction||Numbers of patients to be observed to detect 1, 2, or 3 events|
|1 ADR||2 ADRs||3 ADRs|
|1 in 100||300||480||650|
|1 in 200||600||960||1300|
|1 in 1000||3000||4800||6500|
|1 in 2000||6000||9600||13000|
|1 in 10,000||30000||48000||65000|
The Rule of 3 may help understand why thrombosis with thrombocytopenia syndrome (TTS) that occurred following the use of certain COVID-19 vaccinations, was not noticed until many more people had been exposed to the vaccine. The prescribing information was subsequently revised to minimise the risk.
Similarly, the Rule of 3 may also help account for progressive multifocal leukoencephalopathy (PML) which was associated with natalizumab post marketing. After a brief product withdrawal and following a review of the data, the FDA approved the reintroduction of natalizumab, with revised labelling and improved safety warnings to highlight the potential risk of PML and in the USA, adherence to a special programme that restricted availability of natalizumab to authorised centres and required ongoing evaluation of patients during treatment to minimise the risks (3).
On the other hand, the adverse reaction may be idiosyncratic, that is one which is not related to dose or predictable from the known primary pharmacology. These are more challenging to manage since if the AE is unpredictable, it can’t be minimised by amending the prescribing information. A classic example is troglitazone, withdrawn within 6 weeks of its launch due to ADRs of severe liver damage, resulting in death in some cases.
Finally, the safety concern may not actually be related to the drug itself. For instance, although the exact aetiology of the development of Churg-Strauss symptoms in proximity to initiating zafirlukast is unknown; it is thought that withdrawal of chronic corticosteroids, which was possible when asthma was better controlled, “unmasks” the previously undetected disease (4).
This is where the science of signal detection comes into its own. The first step is to try to analyse all the data available at the time and assess causality using a systematic approach. As a starting point, use of the nine tried and tested Bradford-Hill criteria may be helpful (see Table 3).
|Strength (effect size)||A small association does not mean that there is not a causal effect, though the larger the association, the more likely that it is causal.|
|Consistency (reproducibility)||Consistent findings observed by different persons in different places with different samples strengthens the likelihood of an effect.|
|Specificity||Causation is likely if there is a very specific population at a specific site and disease with no other likely explanation. The more specific an association between a factor and an effect is, the bigger the probability of a causal relationship|
|Temporality||The effect has to occur after the cause (and if there is an expected delay between the cause and expected effect, then the effect must occur after that delay).|
|Biological gradient (dose–response relationship)||Greater exposure should generally lead to greater incidence of the effect. However, in some cases, the mere presence of the factor can trigger the effect. In other cases, an inverse proportion is observed: greater exposure leads to lower incidence|
|Plausibility||A plausible mechanism between cause and effect is helpful (note that knowledge of the mechanism is limited by current knowledge).|
|Coherence||Coherence between epidemiological and laboratory findings increases the likelihood of an effect. However, “lack of such [laboratory] evidence cannot nullify the epidemiological effect on associations”.|
|Experiment||“Occasionally it is possible to appeal to experimental evidence”.|
|Analogy||The use of analogies or similarities between the observed association and any other associations|
In addition, there is one final consideration “Reversibility”; depending on the nature of the event, removing the cause may lead to resolution of the event.
Other steps to take may include generation of new data using a post authorisation safety study (PASS). The European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCePP) is a useful resource in this regard (5).
Once a potential new safety concern has been identified, please do not keep it to yourself. It must be communicated to all relevant stakeholders as soon as possible: Healthcare providers should be encouraged to report any unexpected effects to the MAH as soon as possible, and they in turn must keep the relevant competent authorities informed. Following data review, an update or indeed several updates to the prescribing information may be required, as the event becomes better characterised. In rare instances e.g., trogiltazone, it may lead to product withdrawal.
This short paper aims to touch on the most important aspects of new safety concerns. Each step in attempting to understand the causality and impact on the new drug takes time and effort. tranScrip is ideally placed to help.
We have teams of life science and R&D experts to help achieve your goals and seize opportunities. Whatever your project requires, we have the people to meet your needs.