Six medical stories that illustrate the vital importance of clinical trials

It's the emergency. We are "at war". We must treat and quickly, without makingclinical tests. This is the rhetoric of those who wish to generalize the treatment of Covid-19 through chloroquine. A petition has even collected more than 200,000 signatures on this subject. A survey reports that 59% of French people believe in the effectiveness of chloroquine in the treatment of Covid-19. Should we remember that a scientific fact cannot be decided? It imposes itself on us. " Can the French tell me if bag6 (a uncomfortable coding for a protein) has a real role in thermotolerance? This would save me long months of research. Thank you ", ironically tweeted Anthony Guihur, biology researcher molecular at the University of Lausanne.

But so be it. A review of some medical stories will show that often taking a little longer, even in an emergency, even with molecules well known, it is necessary to avoid killing people. Yes, ignorance and overconfidence kill.

Once upon a time there were clinical trials …

Before moving on to the crisp but nevertheless tragic stories, we must briefly review what a clinical trial is and its history. A clinical trial, according to the International Committee of Medical Journal Editors, is defined as " Any research project that prospectively assigns human subjects to intervention and comparison groups in order to study the cause and effect relationship between a medical procedure and the evolution of a state of health. " For that, a rigorous methodology is implemented to apprehend only the effectiveness of the treatment and nothing else. A clinical trial consists of several phases which all take more or less time. This is why the arrival of a new drug in the market takes a considerable time.

First, paths must be discovered by basic research. Then, these same tracks must be confirmed by studies on cell cultures and on animals. Only after all this long process, human trials can begin. These take place in four phases. The first two phases aim to confirm or deny the activity detected in vitro and in vivo, check the tolerance and adapt the dosage optimal. The third measures the effectiveness of a drug compared to a group of patients treated with a sham. The fourth phase is pharmacovigilance. The drug can be prescribed by doctors, and patients report side effects that have escaped the studies, which cannot capture the complex reality (long term, host-drug-environment interaction, etc.).

The clinical trial is a relatively old concept. Already in the year 1000, Avicenna, physicist and philosopher of the Arab world, described in his encyclopedia what can be compared to the beginnings of clinical trials. However, it was still another seven centuries before the first clinical trial was born: that of James Lind to determine the effectiveness of citrus against scurvy (even if he was fiercely convinced that it did not come from a substance contained by citrus fruits …). It is the first clinical trial with control groups. Over the centuries, the method has been perfected (use of a placebo, randomization, replications, etc.) alongside evolution of the scientific method. However, clinical trials were not always systematic when doctors were relatively confident; they were practiced late for some therapies who were believed to save lives.

When medicine was killing patients!

Here is the story of these treatments which, for several years, killed patients instead of treating them, either because the hypothesis was of an unstoppable biological logic, or because the treatment was effective on certain biological criteria. Without further ado, let's get started.

The Crash Review … lives up to its name

This clinical Review was published in the prestigious medical journal The Lancet in 2005 and included more than 10,000 patients. Its purpose was simple: to determine whether the use corticosteroids (medication anti-inflammatory powerful) was beneficial in patients with head trauma. Before its publication, doctors sometimes put people who have suffered a violent shock to the head and who are not very conscious (with a Glasgow score less than 15) under infusion of corticosteroids. Why ? Biologically, the reasoning is unstoppable. The shock induces a inflammation bad prognosis which induces death in the worst case. In other words, to formalize: A induces B which induces (or can induce) C. If we suppress B, we suppress (or reduce the risk of) C. But medicine, far from being angry with Aristotelian logic, n does not have all the information in its possession. She does not know everything (this argument is not valid to defend fake medicines such as homeopathy). In fact, the Crash trial concludes, after six months of study, that patients on corticosteroids die 7 to 24% more than the control group and suffer 5 to 10% more severe disability than this same group. Until publication of this clinical trial, it is estimated that doctors "killed" 40 people each year thinking of serving the oath ofHippocrates.

Cast Review

This was mentioned in our Live Facebook regarding chloroquine. It is the same principle as the previous story. This essay aimed to assess the benefit of certain antiarrhythmic treatments in the prevention of the sudden death in people who have been victims of myocardial infarction. In fact, we know that after a heart attack, patients who get out of it have more heart rhythm problems, especially ventricular extra-systoles and that this considerably increases the risk of dying of sudden death. We have witnessed the same reasoning. Heart attack induces rhythm disturbances that increase the risk of dying from sudden death. If you treat rhythm disturbances, you logically reduce the risk. The trial enrolled more than 1,400 patients and came to the following conclusion: anti-arrhythmics, hitherto widely prescribed by doctors, caused twice as many deaths as in the placebo group. The authors attribute the excess mortality to the occurrence ofischemia (a bloodless organ) and pro-arrhythmia. Some estimates in the scientific literature suggest that prescriptions before clinical trials have killed around 60,000 people.

The source comes of a book whose references we were unable to consult. The information is therefore to be taken with tweezers.

The USSCHFTP trial

In this story, we still let patients die for lack of knowledge but the other way around. Because of a medical contraindication. Until 1996, date of the clinical trial in question, the beta-blockers (a class of drugs that reduce the heart rate) were contraindicated in patients with heart failure. Same Aristotelian logic at work: if beta-blockers calm the heart, they must surely worsen the situation in patients whose hearts are already struggling to pump blood well. Except that the clinical trial showed the exact opposite. Beta blockers reduced the combined risk of dying or being hospitalized by 38%.

Fluoride for the bones of postmenopausal women

With this essay, we return to the stories where experiences have guided medical practice on the basis of substitution criteria. In postmenopausal women who suffered fromosteoporosis, the fluorine was widely prescribed because its action had been demonstrated to strengthen the bone density cancellous bone. But whena clinical trial was undertaken to verify that the action was indeed effective on clinical criteria (because this is what matters at the end), it was noted that the density did indeed increase at the level of the cancellous bone in the group treated but that it decreased on the cortical bone (another part of the bone) and that the number of vertebral fractures was similar. Udder still, non-vertebral fractures were much more common in the treated group. In addition to being ineffective, this treatment, prescribed for years, brought its share of side effects and potentially torpedoed the quality of life of patients.

Rosiglitazone for diabetes

This medicine has been authorized for sale on the market (it is no longer so today in France) thanks to its demonstrated effect on reducinghemoglobin glycated (a biological marker to determine the level of sugar the last few months). Indeed, glycated hemoglobin lower is the sign of a diabetes control. Then comes a day when researchers asked themselves the question about the effect on cardiovascular mortality of this treatment – this is the main risk factor among people diabetics. they have compiled the data from 42 clinical trials and came to the following conclusion: the treatment that was thought to be useful, increases at worst by 274% (the confidence interval is between 0.98 and 2.74) the risk ofaccident cardiac.

Beta carotenes for smokers

This last little story is well known in the nutrition community. It illustrates another type of overconfidence. In fact, in the 1990s, most observational studies showed that the risk of lung cancer in smokers was lower in those who had a higher intake of plants (and therefore beta-carotene); at that time, we didn't know the matrix effect of food. But when clinical trials (especially the Caret essay) Were undertaken to put this hypothesis to the Review of the scientific method (and fortunately they were carried out), most were stopped along the way: supplementation induced more lung cancers than placebo.

To conclude

As in any good story, we need a moral. Here are two.

The first: we are far, very far, from understanding even a tiny part of the complex reality of the world in which we live. We think we understand a mechanism (corticosteroids and inflammation, arrhythmia and sudden death, beta blockers and heart failure), a biological chain relationship (fluoride -> bone density -> fractures / rosiglitazone -> glycated hemoglobin -> cardiovascular system) or have empirical certainty (beta carotene – -> decreases the risk of lung cancer) but we do not measure the extent of our ignorance.

That door a name: "Ce that we want to see and nothing else ”. This formule, taken from Daniel Kanheman's book, System 1, System 2: both gears of thought, suggests to us that human beings are far too focused on what they see and not enough on what they don't see. As part of the mechanisms, we were too focused on the single mechanism under our eyes and we hadn't thought that a multitude of other mechanisms could be in play and bring harmful or beneficial effects. In the context of biological relationships, we only had one parameter in mind without thinking of the plethora of other parameters and metabolic pathways that a treatment can modify. Within the framework of empirical certainty, we intuitively (wrongly) think that our observations (polluted by innumerable confounding factors) are enough to infer causalities between phenomena (only clinical trials allow this to be done in medicine).

The second: doing clinical trials is not a fad of dreamy methodologists and hungry for certain knowledge but of people as pragmatic who know that mistakes can be made. In medicine, making mistakes potentially means killing people. This should make us understand that the real is complex and that we have to be extra careful before doing anything.

Note: This article was largely inspired by the ehealth of Tristan, public health doctor, popularizer and videographer of the Alpha Risk channel. To go even further, this article by Juliette Ferry-Danini, Doctor of philosophy of medicine, on the introduction to the ethics of clinical trials and this ehealth of Lê Nguyen Hoang, doctor of mathematics, on a potential revolution in clinical trial methods.