Experimental Medicines For COVID-19 Could Help Someday, But Home Runs Not Guaranteed
If the coronavirus vaccines currently being tested don't pan out, don't expect new drugs to fill the gap any time soon.
Many drugs are in the works, and those that succeed could play a role in reducing symptoms and sometimes saving lives. But, given the way drugs are developed, it's unlikely that any single medicine will be anywhere as potent against the coronavirus as a successful vaccine.
So far, scientists have only identified one kind of drug that has been shown to save lives: steroids. A key study found that the steroid dexamethasone reduced death rates among people on ventilators from 41% to 29%. That's a substantial improvement, but still far short of a cure.
The other much-touted drug to treat COVID-19 is remdesivir. The Food and Drug Administration granted emergency use authorization for this drug even though it has not been shown to save lives. It appears to shorten hospital stays.
There's a lot more in the pipeline. Some drugs try to destroy the virus. Others try to control the body's immune system so it doesn't overreact to the infection. Another approach is to prevent the virus from causing an infection.
This week drugmaker Eli Lilly & Co. said it is making progress on its virus-blocking approach. Its experimental product is derived from a person who had successfully fought off infection.
"We found the one antibody in their body that was most potent," says Dr. Dan Skovronsky, Lilly's chief scientific officer. "We engineered it in our laboratories, turned it into a medicine, made it in our factory and started testing in patients" – all within a remarkably short six-month time span.
This proof-of concept study suggests that the strategy of blocking virus with antibodies holds promise. Inmaculada "Inma" Hernandez, at the University of Pittsburgh school of pharmacy, is hopeful, but she doubts that these drugs will be a game-changer.
"These drugs are so complex to produce, probably we will not have antibodies available for treating everybody who gets coronavirus," she says. "They're probably going to be considerably expensive."
Antibodies as a class are among the highest-priced drugs. And Derek Lowe, a drug developer and author of a blog about the pharmaceutical industry called In the Pipeline, notes that Lilly used hefty doses of these antibodies. The most effective dose appeared to be 2.8 grams — that's heavier than a penny. Making enough to treat the tens of thousands of Americans who get sick every day would require "truckloads of antibody," Lowe said.
Lilly executive Skovronsky says he's hopeful a dose one-fourth of that amount will be effective. And he notes that, unlike other treatments involving antibodies, this one would require just a single dose, which would help extend supplies and reduce the cost per patient.
"We're confident that this will be an affordable drug for society — for governments," he says. "Of course, our expectation would be that governments wouldn't pass any expenses along to patients."
Lilly is not the only company trying this approach. Regeneron has a cocktail of two potent antibodies that it is now testing. A related approach, called convalescent plasma, involves transfusing blood plasma that contains antibodies. The antibody levels are much less concentrated in the serum than in the experimental drugs.
Blocking infection could be especially important if vaccines now in development prove to be ineffective. Even if there is a mostly effective vaccine, some people are still likely to fall ill. So, to minimize the effects of that, many drug companies are trying to develop drugs that will tamp down the body's overreactive immune system. That leads to potentially deadly inflammation. This is what the steroids do, but the hunt continues for other options.
"The problem is inflammation and the immune system are so just ridiculously complicated, that you have to be empirical," says drug researcher Lowe. "You have to say, 'Well, that looks like a good idea, let's go find out if it's real.' And a lot of the times it's not." That exploration consumes time and money, and no single drug is likely to do everything that's needed to control the runaway immune system.
A true "cure" for COVID-19 would be a drug, or drugs, that actually knock out the coronavirus itself. There are a few actual cures for viral diseases, including hepatitis C. HIV, which causes AIDS, can also be kept in check for many years with a drug cocktail. But this approach requires a combination of drugs, and finding not just one but multiple drugs that work well together is no easy task. "If you wanted to do that for the coronavirus, you'd still be looking at quite a few years," Lowe says.
Most efforts to find curative drugs for viruses have failed. In fact, the reality in drug development is that most ideas that look great on paper actually fail somewhere along the way. They are either too toxic, or they simply don't work.
Dr. Vinay Prasad, a cancer researcher at University of California San Francisco and frequent critic of the drug-approval process, says even when a medication does work, it is rarely revolutionary.
"Most of our effect sizes are modest benefits," he says. "Five percent benefits. Two percent benefits. One percent benefits."
When those drugs are combined, the benefits can add up — sometimes to quite effective therapies. But that takes years of trial and error. Prasad uses a baseball analogy for how he thinks coronavirus drug development might play out.
"It's possible that we're going to get singles and doubles," he says. "We might not get the home run, and we might have to rely on the core things that we know control viral spread."
Masks and social distancing are, in the short run at least, more potent than pills.
Skovronsky at Lilly isn't discouraged, though. He says a cure usually doesn't require any medicine.
"The truth is that for most people who get COVID-19 probably they cure themselves," he says. The primary task for drug developers is to find medicines effective for the 5% to 10% of people who get seriously ill.
You can reach NPR Science Correspondent Richard Harris at email@example.com.