Every drug ever discovered has side effects. This is commonplace, or at least it should be. However, important questions immediately arise about how serious these side effects can be and how often they occur with normal use. Drug developers speak of a therapeutic window between the effective dose and the higher doses that are more likely to cause problems. With a high enough dose, of course, everything is problematic: water itself is deadly if you drink insane amounts of it in a short period of time. But ideally, you want a nice big gap between what a patient is likely to ingest and any type of toxic dosage. In case you’re wondering, acetaminophen is the most common drug with a narrow window. Overdoses are responsible for most of the acute liver failure in the western world, especially when combined with alcohol.
But what about side effects at the normal doses themselves? This becomes a task for statistics, because the more people take a particular medicine, the more likely it is that someone will react badly to it. These can range from a headache or a short-lived skin rash to extremely serious conditions like anaphylactic shock. As you can imagine, drug companies are very interested in knowing how common these things can be. But in many cases that is hard-earned knowledge. The word pharmacovigilance includes the process of observing such things after a new therapy has been approved and it can put stress on the nerves.
Clinical trials of 10,000 or 20,000 patients weren’t enough to solve this problem, and these are roughly the largest ever done
The recent headlines about vascular side effects with some coronavirus vaccines are a good example of this. As I write this is a very evolving story. However, it seems clear that the vaccine clinical trials of 10,000 or 20,000 patients were insufficient to solve this problem, and they are frankly the largest clinical trials ever conducted. If only one in 50,000 or one in 100,000 patients has a specific side effect, then you are really out of the range of your statistical performance during development.
What brings about such rare occurrences and why are they so rare in the first place? The vaccine story provides another answer: the extraordinarily personal nature of the immune system. First of all, our gene sequences are most variable when they code for different proteins of the immune system. Things are further personalized by recombinant DNA events in our B cells, giving each of us a huge combinatorial library of antibodies. Then our individual life experiences each give us a further individual immune landscape in our memory B and memory T cells. There weren’t enough people on earth to produce a duplicate.
With a therapy like vaccination that directly targets immune function, it’s clear how you can come across a rare combination of cross-reactive antibodies. However, this also applies to completely independent low molecular weight active ingredients. This is especially true for those who may form covalent bonds to reactive protein residues (potentially nearly a third of all drugs marketed). The modified proteins that are occasionally produced trigger an immune response with one of the myriad trillions of possible antibodies, and a side effect arises.
Nothing near a complete antibody profile has ever been done on a human
So far it is basically impossible to predict such things. Nothing near a complete antibody profile has ever been done on a human. The number of possibilities (all these antibodies, all these theoretical binding events and downstream effects) is not experimental and cannot be calculated and will remain so for the foreseeable future. What does that leave us?
Pharmacovigilance. We need to watch out for such things and collect as much reliable data as possible as soon as possible. If there are certain classes of patients at higher risk, we need to learn who they are so that they can be warned while others can continue to benefit from the therapy itself. Hurting a few people can be unacceptable, but so can a much larger number of patients in need. It all depends on the numbers, the damage, and the need. It is a judgment and always will be – and if it isn’t your decision, be glad.