In the discovery and development of new medicines and vaccines, researchers are always looking to maximise the benefit to the patient, while minimising side-effects.
The Covid-19 pandemic has put the spotlight on the issue of safety assessments amid the urgent need to develop vaccines.
The speed at which the pandemic has evolved - plus occasional bureaucratic missteps, communication gaps, actions of malicious players, and not to mention that lives and livelihoods are at stake - has created a fertile ground for misinformation to proliferate, especially amid the explosion of social media platforms and commentators.
Against that, it is not surprising that ill-informed, incendiary letters and petitions on
adverse effects of vaccines have stirred up controversy.
In Singapore, an open letter by 12 doctors last month, later retracted by 11 of them, urged the Government to give children here the China-made vaccine, Sinovac, which uses the traditional "killed vaccine" technology, and not the new-technology mRNA vaccines from Pfizer-BioNTech and Moderna deployed in the public vaccination programme. The letter claimed that it was not known what side-effects might surface in 10 to 20 years.
The Straits Times reported earlier this week that a senior infectious diseases specialist, Associate Professor David Lye, had spoken up against the misinformation.
So this prompts the question: How have scientists, the industry and regulators approached the thorny issue of understanding and managing adverse effects from medications?
The process behind safety assessment
As new therapies are being developed, scientists start to build an understanding of possible untoward drug effects right from the start of the discovery process.
There are some adverse effects that will be predictable in clinical usage. A new insulin for diabetes is expected to lower blood sugar levels, but by the same measure may cause severely low glucose.
A sedative medication may provide relief from anxiety, but cause drowsiness.
These are sometimes referred to as "on-target" effects. They are an extension of the medication's primary pharmacology, are generally predictable and often related to the dosage.
On the other hand, there may be adverse effects that are not so predictable. A painkiller like paracetamol may cause severe liver damage, drugs that lower cholesterol may cause muscle injury, and some antibiotics cause allergic reactions.
The mechanisms by which these drugs cause these side effects are unrelated to their primary pharmacology, and sometimes termed "off-target" effects.
How to mitigate these? Through experience, algorithms, and increasingly supported by computational power, scientists try to improve the potency and selectivity of these drugs, and design the molecules to minimise their potential for complications.
In early toxicological testing in the laboratory and on animals, tests are done to detect deleterious effects, including potential for causing gene toxicity, cancer, reproductive and in-utero injury, or harm to other organ systems.
In clinical testing, the relatedness of an adverse event is assessed by biological plausibility, time relationship to dosing, resolution on stopping the drug, re-occurrence on reintroduction, and dose response.
This process helps investigators link adverse effects with drugs, and the doses at which they occur. Quite often, events may seem at first to be related, but turn out to be incidental.
In late pivotal phase 3 registration studies, the power of detection lies in large numbers.
The Pfizer-BioNTech Covid-19 vaccine - known as Comirnaty, or BNT162b2 - phase 3 trial had more than 43,000 participants, of whom half received two doses, and the other half were assigned to the untreated placebo group. Placebo control is an extremely important scientific feature of late phase clinical trials, to both elucidate effectiveness of drugs, and draw comparisons to background rates of events that may otherwise be erroneously attributed to the investigational agent.
In the Comirnaty phase 3 study, there were six deaths at the time of data submission, two in the active treatment group, and four in the placebo group. Of the two deaths in the treatment group, one was the result of cardiac arrest, and the other of heart vessel disease.
Of the four deaths in the placebo group, one was from a heart attack, another from a stroke, and the causes of the other two were unknown.
This clearly demonstrated that in a very large trial, there was a balance of cardiovascular events between the treated and untreated groups, and cardiovascular events could not be solely attributed to the vaccine.
Once a treatment enters broad clinical usage, it does become more challenging to compare adverse events to control groups. One good, if imperfect, way is to benchmark against a baseline of various disease rates, through registries or databases.
According to the Singapore Heart Foundation, in 2019 there were 6,291 deaths due to cardiovascular illness, inclusive of stroke, heart vessel disease and hypertension, or roughly 500 deaths a month, most in the elderly group.
At the height of the vaccination campaign for the elderly at the turn of the year in Singapore, thousands of people aged 75 and above were being vaccinated daily. With 500 deaths a month expected from cardiovascular disease, one could surmise that a number of these would have occurred within days of having received a vaccine.
To be clear, families and loved ones are understandably distraught if their elderly relatives suffer from disability or death following a vaccination, and will seek answers. However, in the search for the truth, it is important that all parties concerned do not mistake the timing of an event with its actual cause.
Lastly, sometimes causality may be linked if events are unusual or bizarre. An example relates to the erectile dysfunction drug Viagra. Soon after its launch, some patients experienced blue-tinged vision. As this was a phenomenon that had almost never been seen before, it was easy to identify the culprit. On the Covid-19 front, some way into the deployment of the Oxford-AstraZeneca vaccine, there were reports of a normally rare event, cerebrovascular venous sinus thrombosis.
The unusual feature here was blood clotting, associated with low platelet levels. Both the Medicines and Healthcare products Regulatory Agency (MHRA) in Britain and the Food and Drug Administration (FDA) in the United States have recognised this as a possible side-effect linked to the vaccine.
Even so, this is still very rare, and the FDA and the MHRA have maintained that the benefits of receiving the Oxford-AstraZeneca vaccine continue to outweigh the risk.
Occasionally though, serendipity strikes. In the 1990s, Viagra was being investigated for the treatment of hypertension and chest pain. However, in male patients it had the effect of inducing and sustaining penile erections, and a legendary drug was thus born. In most cases, medication's adverse effects cause discomfort, disability or death. It is rewarding that they sometimes result in happy endings.
How long is long-term safety data?
In the monitoring of safety data for newly developed therapeutics and vaccines, how long is long enough? The short answer is that among other factors, it depends on the nature of the agent in question, the likely mechanisms of action, the possible adverse effects that may occur, the reversibility of these adverse effects, and the urgency of the medical need.
This typically runs from months to several years.
There is a narrative in circulation claiming that 10-20 years of data is needed before mRNA vaccines could lay claim to long-term safety. This proposition was buried within an obtuse manifesto, and therefore hard to decode, but a casual reader could perhaps have taken away a few implications.
First, that 10-20 years of data is necessary for a claim to long-term safety; second, that longer safety data conferred greater confidence of it being safe; and third, that this duration for safety monitoring was necessary before introducing new agents for use in children and adolescents.
For chronic illnesses, where medications are taken long term, and sometimes for life, longer-term chronic dosing studies may be required, to capture both efficacy and safety. Examples of these range from a one-year phase 3 study of amlodipine for hypertension, to a four-year phase 3 study of empagliflozin for diabetes.
Antibiotics are typically used short term, and trials may dose patients for up to two weeks, and follow up for one to several months. These durations of trials and safety monitoring are sufficient for approval of the drug for safe use, but nothing at all in the range of 10 to 20 years.
Second, it should be noted that just because a drug has been in common usage for longer, it does not mean it is safer.
Some first-generation oral diabetes drugs caused higher incidences of hypoglycaemia and have been superseded by second-generation drugs. Paracetamol is an old and familiar painkiller, but to this day is causing poisonings and liver failure.
In fact, a large part of drug development is the imperative to make new and safer versions of medicines already on the market.
Last, providing early access of therapeutics to underserved populations, such as the elderly and the young, is a key principle of drug development.
It is far too easy to just perform the clinical trials in young, healthy adults and leave a drug's usage restricted to that group. This has long left children in dire situations where a physician may be forced to use a drug off-label to treat an ill child.
The FDA, in fact, mandates and incentivises companies to conduct studies in paediatric populations through several pieces of legislation. Several life-saving medications, like insulin for diabetes, first studied in man almost exactly 100 years ago, and genetically modified cell therapies such as Kymriah for fatal blood cancers, just launched in the last few years, were tested primarily in children and adolescents first.
To suggest that one needs 10 to 20 years of safety data before introducing a new therapy to children is plainly ludicrous, irresponsible, and is a gross misunderstanding of what safety in drug development is.
Who calls the shots on the shots?
It takes a village to raise a child, and so it takes a competent agency to rigorously evaluate drug efficacy, safety and quality. The Singapore Health Sciences Authority is an agency of high international standing and reputation, staffed by hundreds who are steeped in their areas of speciality.
The review of a submission package is a sophisticated exercise, requiring expert input from clinicians, pharmacists, toxicologists, laboratory specialists, biologists, manufacturing quality inspectors, biostatisticians and more - all working collaboratively to piece together the massive puzzle that is a drug approval package.
Lone-wolf observers, no matter how erudite in gleaning information from the press, social media and partially reviewed journal articles, still do not have access to the breadth of information that agencies review, often confidentially, and are therefore ill-equipped to make an informed determination of whether a novel agent passes muster, let alone petition on behalf of the local population based on porous premises.
The types of deliberations on safety, efficacy and quality made by the agency cannot be reconstructed by untrained groups of individuals using open source data.
However, the public can be assured that there is a lot of rigour and science, and some measure of art, to building safe drugs, as well as studying and understanding the inevitable side-effects that follow.
This is ensconced in both the medicines development and regulatory processes that are meant to protect the public good, in Singapore and the world over.
• Dr Danny Soon is chief executive of the Consortium for Clinical Research and Innovation Singapore. He is also executive director of the Singapore Clinical Research Institute.