key: cord-0879070-i2emky4c
authors: Broadbent, Alex
title: Better the drug you know: Commentary on “Daughton 2020, Natural experiment concept to accelerate the re-purposing of existing therapeutics for Covid-19”
date: 2020-05-18
journal: Glob Epidemiol
DOI: 10.1016/j.gloepi.2020.100027
sha: 8902b0a293a5f7e0ee84fb7dc4fd4977305f5d37
doc_id: 879070
cord_uid: i2emky4c

nan

The history of drug discovery is a history of happy accident. The chance floating of a spore from one research laboratory to another would have made no difference to human history had it not been for the presence, in a single building, of research programmes into what, at the time, seemed the quite different topics. Nor would the spore have developed in the culture in which it landed, had the researcher whose Petri dish it was not been on vacation, and the weather in his absence been unusually warm. And none of this would have made any difference had the researchers in question not been friends, who, between them, were able to detect a signal of potential medical significance. This signal had been undetected over a decade earlier, when an anti-bacterial effect of the penicillin mould had been observed as nothing more than a curiosity. The risky decision on the part of a single individual to devote all the resources of a single make-shift laboratory in wartime Britain, with German invasion imminent, was also remarkable, and crowning it all is modern medicine's 'mystery of mysteries': "the question of how, and more particularly why, a handful of the simplest of micro-organisms should have the ability to create these complex chemicals, or why they should exist at all"(Le Fanu, 2012, p. 26) . While the development of antibiotics did require considerable human perseverance, it also required remarkable luck.

Antibiotics are perhaps the most famous example of a common phenomenon. In contrast to the milestone surgical developments, which for the most part are more direct results of creative thinking by determined individuals persevering on a stubborn problem (e.g. open heart surgery, hip replacement, organ transplant), the major pharmaceutical discoveries have been largely accidental. Several other major pharmaceutical discoveries have concerned substances already known and being deployed, or investigation, for some purpose other than the one for which they were to become famous, for example cortisone and sildenafil citrate (Viagra).

Drug discovery is chancy in a less benign way too: it can be dangerous. There is a horrible history of pharmaceutical interventions reaching market without the detection of major side effects. Thalydomide is the most famous of these, and the birth defects to which it gave rise led to much more stringent regulations on drug testing. However, no testing process is without risk, and when there is a strong incentive to achieve a positive outcome, these processes are under pressure. The huge social and, perhaps, financial benefits of developing an effective preventive or curative intervention for COVID-19 create pressures of this kind.

The advantages of "repurposing" or "repositioning" existing drugs are obvious, and especially so in this context. Safety and side effect profile of a drug already in use are likewise already known. Moreover, production capabilities already exist, and the risk of a complicated patent-war -or, worse, international political games -is reduced.

Poorer countries are more likely to be able to afford a drug that is already on the market than one is newly developed. It goes without saying that even if a drug in our current arsenal appears promising against COVID-19, trials on the efficacy of existing drugs for the new usage, for their safety in this context, for significance any differences in required dosage, are still necessary. But the testing protocol required for an entirely new substance would be longer, and the discovery process itself is also slow and expensive when it begins at the molecular level.

J o u r n a l P r e -p r o o f tests)" (Daughton, 2020, p. 00) . These groups are then to be compared for drug use, and common drugs among patients apparently doing better against COVID -19 (the disease caused by SARS-CoV-2 infection) can then referred for further testing. This is a simple, practical idea, and one with an excellent epidemiological pedigree, besides the parallel with natural experiment that the author highlights. The method shares a common structure with the epidemiologically seminal case-control study, developed for detecting causal effects of diseases whose course runs too long for impatient mortal investigators to observe with sufficient frequency to draw results (and whose suspected cause, smoking cigarettes, had become widespread only relatively recently when the connection became of interest and concern). Speed is thus a central purpose of the method, explaining the utility of its application to accelerate the process of drug discovery.

Case-control studies are not final arbiters of causality, but as discovery tools, they are practical, cheap, and fast. Daughton does not present the proposal in terms of the case-control methodology, but in terms of natural experiment. However, Daughton does clearly acknowledge that the method is not the final word on causality, but rather a way of providing promising leads for rigorous follow-up. Used this way, as it is proposed, Daughton's method inherits the virtues of the case-control design made famous by the discovery of the cause of almost all lung cancer.

Daughton does an excellent job of dealing with the devil in the detail, considering numerous potential drawbacks of the method, as well as detailing its advantages.

Perhaps the over-riding advantage, however, is that the Daughton Method can be implemented cheaply and without risk. It is hard to see what potential drawback could override the case for at least trying the method in the present situation. Given the resources currently directed (and often diverted) to drug discovery for COVID-19, a Journal Pre-proof J o u r n a l P r e -p r o o f relatively small investment could yield a significant results. Even if nothing shows up for COVID-19, this method could be applied for other purposes. Who knows, perhaps even the application to COVID-19 drug search might result in happy accidents for other interventions on other diseases.

It is to be hoped that Daughton's paper is widely read, and that the Daughton Method is tried in the rush to find effective pharmaceutical interventions against COVID -

No competing interests

Natural Experiment Concept to Accelerate the Re-purposing of Existing Therapeutics for Covid-19. Global Epidemiology. In press

The Rise and Fall of Modern Medicine