Transmission electron micrograph of SARS-CoV-2 virus particles Photo credit: National Institute of Allergy and Infectious Diseases


We are now threatened by Omicron, a new Covid variant with a name like a villain from a Transformers movie. Although on paper its genes look scary, we still know very little about its real-life behavior.

So, while travel precautions make sense until we know more (and perhaps we could finally get serious about quarantine the way nations from China to New Zealand did), panic seems premature.

Some respected South African doctors believe it is less dangerous than earlier variants. Although cases are rising fast in Gauteng, the province where it originated, hospitalizations and deaths are not—or not yet. Other scary variants, like the Brazil one and the previous South African one, did less damage that we first feared.

Unicron, deadliest foe of the Transformers
Credit: Wikipedia Fair Use

Nonetheless, even if it does prove dangerous, we also have good news that we lacked when previous variants emerged: we now have two pills that probably will work against all new variants.

How? Because, unlike antibodies, the drugs don’t just glom onto the virus’s outer spikes, which constantly shift their shapes. They enter our cells, as the virus does, and gum up the virus’s replication process, which is thought to be the same for all variants. The two are the Merck pill, molnupiravir (named after Mjölnir, Thor’s hammer), and the Pfizer pill, Paxlovid.

There is now a debate about which is best. The latest data on molnupiravir found it only 30 percent effective at preventing hospitalization, rather than the 50 percent Merck claimed in October. Three weeks ago, Pfizer claimed Paxlovid was 89 percent effective; we still await that trial’s final data. Deaths while using either were very rare: only one so far.

I think this “I only want the best, which is it?” debate—an echo of the one over Pfizer, Moderna, J&J and AstraZeneca vaccines—misses the point.

I suspect the best way to use these two powerful drugs—and others in the pipeline if they work out—will be to combine them. If we don’t, we might lose all of them, one by one.

Here’s why:

One-shot magic bullets like penicillin start off like gangbusters. Headline writers love them. So do Pharma C.E.O.’s—they sell millions of doses.

But as the pathogens mutate and become resistant, they lose their magic.

For decades, we have had this lesson beaten into us again and again and again. But it never sticks.

In the 1980s, H.I.V. rapidly showed us the folly of the magic-bullet approach. It mutated so fast that AZT, the first bullet, quickly failed. That led to the concoction of triple-therapy cocktails, and in 1996, Dr. David Ho, a triple-therapy pioneer, ended up on the cover of Time magazine.

Hepatitis C reinforced the lesson: since antivirals for them emerged in 2013, millions of chronic carriers have been cured. There are now eight hep C regimens—all involving two, three or even four drugs.

But we forgot that we had actually learned this lesson before, back when diseases took longer to cross borders.

In the 1940s, tuberculosis could be beaten by streptomycin alone, and many Americans were cured before resistant strains proliferated. By the 1960s, however, success took at least four drugs; until recently, the only hope for someone with a highly resistant strain was seven drugs—taken in up to 40 pills a day for a year or more. (The consequences could include deafness, death of foot and hand nerves, and psychosis.)

In the Vietnam War of the 1960s, synthetic quinine (chloroquine) stopped working against malaria. After fumbling with several magic bullets, we now use cocktails that combine artemisinin—a fast-acting killer that won the 2015 Nobel Prize for a Chinese scientist—with drugs that linger longer in the blood to mop up survivors.

Why do cocktails work when single therapy fails?

Imagine fighting an opponent who has an ax and can only do one stroke—straight down on your head. If he lands it, you’re dead. But if you learn to step aside at the last second, you live, and he’s useless. That is the equivalent of a virus evolving resistance to a drug’s “mechanism of action.”

Single-mechanism adherent Credit: Tylwyth Eldar / Wikimedia Commons

But suppose you instead face three opponents with three different mechanisms: One has the ax. One has a scimitar and swipes at your neck, which you can dodge by ducking. One has a spear, which he thrusts at your stomach; you can avoid it by jumping back.

But no one can step aside, duck, and jump backwards all at the same time. The “cocktail attack” will get you.

Molnupiravir and Paxlovid have different mechanisms. The first, a nucleoside analog, swaps a new, misshapen rung into the RNA ladder. (More detailed explanations here and here.) That triggers random mutations until, as one scientist described it, the virus “mutates itself to death.”

Paxlovid, by contrast, is a protease inhibitor. Proteases—of which there are hundreds—are tiny protein scissors that snip up bigger proteins. A virus is simply a protein shell with instructions for making new proteins. Paxlovid binds to one crucial protease, blocking it as a paper clip might block the cutter on a pasta machine. The whole kitchen fills with one unwieldy strand of fettucine—and dinner is cancelled.

Molnupiravir has another reason for caution: some scientists fear it could also trigger mutations in the DNA in our own cells. That could be lethal to embryos and cause cancer in adults. Merck insists it is safe at the low, short dosages that work against Covid. But it was worried enough to exclude from its testing all pregnant woman and all women and men who were not using birth control.

Wouldn’t it be safer to use both drugs in combination?

A cocktail could not only delay or prevent the rise of drug-resistance mutations but it should also permit doctors to use less of each drug—lowering the risk of dangerous side effects.

So why do we always first go for the bullet instead of the cocktail?

Because that’s how the patent system has worked since the 19th century. No one needs a cocktail sewing machine, light bulb, or cotton gin. You only want one.

Then you want the inventors to compete on price and improvements, so you forbid them from colluding with each other. That’s the core of the 1890 anti-trust laws.

Which is fine for cotton gins. Not so fine for dying humans. Dying humans often benefit from a team approach: surgeon and oncologist, stent and statins, chemotherapy and immunotherapy, and so on.

So I’d say: let’s draw a distinction between companies that make toasters and companies that save lives.

Let’s modify the old “winner take all” model to instead encourage—or even require—pharmaceutical companies to cooperate when they invent life-saving cures. (I don’t care if they still want to battle over market share in the erectile dysfunction field.)

They should be compelled to run clinical trials of cocktails and be permitted to mix several companies’ drugs into one pill when it’s chemically possible.

I’m not arguing in favor of just taking the brakes off. Drug companies have been absolutely rapacious about maximizing their profits—often at the expense of sick Americans and tax-paying Americans (to say nothing about how they once behaved at the expense of sick Africans.)

That balance between cooperation and collusion must be managed by the F.D.A. or a new agency with the overall goal of saving lives.

Since drug companies are global, it should be coordinated with other countries.

On rare occasions, the drug industry has done this by itself. In 2009, Pfizer and GSK melded and spun off their H.I.V. divisions to create ViiV Healthcare, a new company that could make cocktails to compete with Gilead’s three-drug pill. (They also invented H.I.V. formulations for children.)

Weirdly, what I’m describing often happens first in ways that benefit only poor countries. Cipla, an Indian company was making three-in-one H.I.V. pills for Africa almost a decade before any were available here . . . because of our antitrust laws.

That could be fixed—and should be before any of our magic Covid bullets turn into Nerf gun darts.

Disclosure: Since I no longer work for The New York Times, I’m no longer bound by its rules against owning stocks in medical companies. I own some stock in Codexis, which makes enzymes used by both Pfizer and Merck. It will certainly do better if they sell more pills; I can’t predict what effect easing anti-trust laws would have. That’s not why I wrote this — but you be the judge.


Donald G. McNeil Jr. was a New York Times journalist from 1976–2021. His last beat was being the lead reporter on the pandemic. He received the 2020 Chancellor Award and the 2021 NYT Pulitzer. This article originally ran on his Medium page on November 29, 2021.