Tweetorials: Flu Antivirals & Drug-Drug Interactions
11 March 2023
Late in January 2023 I put out two sets of posts that are sort-of tweetorials. They’re a bit shorter than my usual 30-ish post rambles, but they both touch broadly on aspects of antiviral drugs. So here I’ve bundled them together in a single Substack post with some expanded commentary upfront.
The first topic was flu antivirals, which became front-of-mind when I got the flu in January and started on a course of Tamiflu — which worked like magic by the way, and had me back on my feet in under 48 hours. Although all eyes are on COVID-19 these days, influenza remains a significant worldwide health burden, killing hundreds of thousands of people every year. It’s also a fairly fast mutating virus, and has animal reservoirs of disease that could spill over and cause pandemics — all of which is why our flu vaccines have to be updated annually to reflect what’s in circulation, or likely to be.
One would think we’d have a bit more in the medicine cabinet for flu by now, but the cold reality is that the neuraminidase inhibitors are just about it. There are also some older drugs, the adamantanes (these block the viral M2 ion channel, which prevents viral uncoating), that are not widely used anymore because of the high amount of resistant strains of influenza A in circulation.
Although the tweetorial goes into detail on oseltamivir (Tamiflu), which is in widest use because it’s an oral agent, it was not the first drug in the class. That distinction belongs to zanamivir (Relenza), which was discovered in a collaboration between the Australian CSIRO and Monash University, funded & developed by Biota, and eventually licensed to Glaxo, which became GlaxoSmithKline. It was FDA approved in 1999, around the same time as oseltamivir. Drug discovery geeks like the zanamivir story because it’s perhaps the earliest recorded example of rational structure-based drug design. As a medicine, it has basically the same pros and cons for the treatment of influenza as oseltamivir, but it has one key weakness: it’s not orally bioavailable and must be administered as an inhaled powder.
I was working at GSK in the early 2000s, and I think part of the reason the company in-licensed zanamivir was that they had experience with administration of inhaled agents via their asthma & COPD drug fluticasone/salmeterol (Advair) — indeed both use the same Diskus dispenser technology. I also remember the company going to great lengths to create stockpiles in the mid-2000s when H5N1 avian flu was ripping through bird populations worldwide and people were concerned about the risk of a pandemic from a spillover event.
Which brings us to our second tweetorial topic, itself the outgrowth of a spillover event that (most likely) created the COVID-19 pandemic. Although the entry point for this tweetorial is about nirmatrelvir/ritonavir (Paxlovid), it’s really meant to be an introduction to the larger topic of drug-drug interactions. In the HIV antiviral space, a good number of these medications have metabolism liabilities. Several of the standard highly active antiretroviral therapy (HAART) regimens now incorporate a known CYP3A4 inhibitor like ritonavir or cobicistat to boost the exposure of the other drugs in the combo.
This clinical experience likely gave Pfizer some confidence to proceed with a combination like this when nirmatrelvir (the SARS-CoV-2 main protease inhibitor) showed a similar liability. This decision has not been without consequence, however, as I delve into in the tweetorial: the CYP3A4 inhibition has a knock-on effect to any other co-administered medication metabolized by CYP3A4. It’s such a nuisance that some doctors are avoiding prescribing to their patients who are on a lot of other medications that will require dosing adjustment.
More broadly speaking, however, outside of these examples of deliberately induced drug-drug interactions, DDIs are something that medicinal chemists generally seek to avoid. The downsides of a DDI have been listed above, and it creates an additional burden entering the clinic. Medicinal chemists routinely monitor their compounds for CYP family inhibition, because potent inhibition may translate to a burden to monitor for DDIs in the clinic. DDIs can also swing the other way: some compounds are CYP inducers, meaning they trigger increased CYP expression in the liver, which can decrease the exposure of co-administered victim drugs. So there’s a whole other set of preclinical assays to monitor for CYP induction too. All told: with a few notable and well-managed exceptions, it’s much better to not have a DDI in the first place.
Without further ado:
Tweetorial: Flu Antivirals
Tweetorial: Drug-Drug Interactions
