Panaceas, Percentages, Paths Ahead

Donald J. Trump said on January 23^rd, 2016: “I could stand in the middle of 5th Avenue and shoot somebody and I wouldn’t lose voters.” Well, last Thursday he did it, with words rather than a pistol. Fortunately, his aim was bad – so far none of the people who drank household disinfectants on his advice have died.

Treatment update

Remdesivir: Oy vey iz mir. Two randomized placebo-controlled trials of Gilead Science’s experimental antiviral within days, with opposite results. 

First a study in China found zero difference in time to recovery, viral clearance, or death rate between treated severely ill patients and controls. This was terrible news. 

Then very preliminary reporting of a similar but larger American trial suggested that, on the contrary, the drug shortens recovery time and may reduce death rates. This would be great news.

Anthony Fauci and the US Food and Drug Administration are walking on the sunny side. I hope they’re right, but my trust in Gilead is limited. They’re the guys who brought you a $1000 per pill treatment for hepatitis C, they’ve been disgustingly venal about remdesivir, and they quashed an initial one-paragraph report of the negative Chinese study as soon as it was posted online.

So why have two trials given such different results, and which one is right?

Anybody’s guess, especially since it seems details of the American trial won’t be available for some time. But I do have three thoughts to contribute: 

1)    The timing may have been different. With antiviral drugs – whether Tamiflu against influenza or acyclovir against herpes – the earlier you start taking them, the better. The Chinese study did show a hint of benefit among patients who began remdesivir 9-10 days after falling ill, as opposed to 10-12 days. Perhaps the American patients were treated earlier.

2)    There may have been other differences in the patient populations. In China the remdesivir group included more women than the placebo group (good), but had slightly more severe disease, preexisting conditions, and late enrollees (bad). These relatively minor imbalances could have been enough to influence the results.

3)    Other drugs were permitted in both trials. We have no information about the American one, but we know most Chinese patients received corticosteroids, 90% received antibiotics, and many received either lopinavir–ritonavir or injected interferon. Medication interactions could have affected patient outcomes. 

Conclusion: if the American results pan out, it will be the first time a drug shows any real effect against this disease, and starting patients on it earlier might yield even better results. But only a few days are said to be shaved off recovery time, and if there’s any effect on the death rate it’s relatively minor. We still don’t have a miracle cure, and none is really likely to show up.

Favipiravir: Millions of doses of the only other antiviral still in the running are being stockpiled by the Japanese government.

Tocilizumab and sarilumab: These related arthritis drugs, which were both touted as possible last-ditch miracle cures, are looking much less promising after a randomized placebo-controlled trial of the latter found no effect at all.

Hydroxychloroquine: The little glimmer remaining in Plaquenil’s star seems to be fading fast, now that the first controlled research results proved so discouraging – no efficacy, deadly side effects ­– that the American Food and Drug Administration has had to issue a warning against using the drug in COVID-19 outside of clinical trials. You can watch Fox News for hours now without once hearing the word “hydroxychloroquine.” 

Convalescent serum: The Mayo Clinic-led Convalescent Plasma Project is being expanded to include hundreds of American hospitals, but unfortunately it is conceived as a compassionate use program rather than a clinical trial, so it won’t tell us much about efficacy. Fortunately, several New York hospitals have begun a proper placebo-controlled trial. Researchers in the Italian regions of Marche, Lazio, Campania, Umbria, Puglia, and Tuscany have been contacting known COVID-19 survivors as potential plasma donors, but they seem to be proceeding in slow motion. One informant tells me he received a recruitment email two weeks ago, eagerly accepted, and had a winning supply of antibodies confirmed in his blood, but still hasn’t been called on to donate. The UK is also seeking donors, but they’ve had such poor luck that their planned research is being curtailed.

Regeneron’s antibody cocktail: They should soon start testing a mixture of anti-COVID-19 immunoglobulin produced by tricking mice into producing human-type antibodies. The effect should be similar to convalescent serum, but since it’s easier to round up mice than human donors the technology ought to be easier to scale up.

By now so many once-promising drugs have hit the dust that the rest of what’s left coming down the research pipeline is largely fluff (Viagra? nitric oxide? famotidine? losartan? BCG?) or Hail-Mary players (colchicine, heparin, interferon, ruxolitinib).

Vaccines 

Months after the race began, candidate vaccines now number over 100 and use wildly varying methodologies. 

The most promising American contender seems to be from Moderna, a Cambridge, Mass. company that hopes to move forward soonest to large Phase 2 studies, aided and abetted by the Trump administration. But they still need to wait for Phase 1 safety results on the handful of volunteers who received the vaccine in March.

Researchers at the Jenner Institute at Oxford University are farther ahead. They have already shown that their candidate vaccine works in macaque monkeys, and are leapfrogging over some other preliminaries to head straight for large-scale Phase 2-3 tests of efficacy. Theirs is the only vaccine with any possibility of being in pharmacies before the end of this year. One Indian company is betting on it big-time, planning to churn out millions of doses a month even before there’s any evidence that it works in human beings.

Only one other group has tried its vaccine in primates, the Chinese firm CanSino, which is already into Phase 2 trials.

Forbes magazine has published guides to investors as to where to place their wagers.

Do Swedes do it better?

The Open-It-Up gang are celebrating Sweden’s supposed success in controlling its COVID-19 without a strict lockdown, accompanied by the cheerful chirping of one influential epidemiologist and unlikely predictions that the country will reach herd immunity levels in May. There are two big problems, though. 

The first is that Sweden has not done all that brilliantly. It has seen a total of 2274 deaths, for a rate of 225 per million, about 8 times higher than Iceland, Denmark, or Norway, all of whom had more of a shutdown – and its epidemic is still on the upswing. 

The second problem is that Sweden may not have shuttered its restaurants or put its population under house arrest, but that has not meant life as usual. The government issued social distancing and hand-washing guidelines, advised people to work from home if possible, and suggested the elderly stay home, fining or even shutting down bars and restaurants that set their tables too close together. The Swedes have apparently followed those guidelines fairly scrupulously, at least until a recent warm spell. 

Do Italians do it worse?

The Italian response to COVID-19 has been blasted by much of the American press, with every player from politicians to “socialized medicine” blamed for the holocaust in northern Italy. Along with some other observers, I don’t see it that way. Underfunding was crucial in that catastrophe, but it was ordained by the European Union, not by Italian social policy. Yes, factories were left open too long, but on the whole Lombardy actually lost relatively little time before shutting down, and correctly went overboard in, say, setting up field hospitals rather than risking a shortfall. Early on nobody really understood the need for PPE, and later it was in short supply, leading to the COVID-19 deaths of 152 doctors. Nursing homes have been a disaster in Italy – but no more so than in the USA, which has had at least 14,000COVID-19 deaths in long-care facilities. And things would have been much worse, not better, without the National Health Service. 

Stats of  two cities

New York City: Counting all causes of death, 27,600 New Yorkers succumbed between March 11^th and April 25^th, compared with an average of 6500 for the same period in past years. We can conclude that the pandemic actually cost at least 21,000 lives (4000 more than the official death toll). Nearly half of New Yorkers are personally acquainted with someone who is known to have died of COVID-19.

Rome: Again counting all causes of death, 2753 people died in the city of Rome during March 2020. The expected number based on 2018-19 was… 2752. So the lives the lockdown saved in car accidents, non-covid infections, etc. fully compensated for those lost to COVID-19. (This is of course not true for northern Italy, nor for England, France, Spain, Holland, etc.)

Burn, baby, burn 

Some plausible characters, such as a physician who gets plastered all over television and a former Rockefeller University biostatistician who’s now been disavowed by his previous employers, have been suggesting we just get this COVID-19 thing over with fast by allowing the coronavirus to rampage its way through the population until herd immunity is reached. Assuming infection does lead to immunity (still unproven) they’ve at least got the time angle right – the pandemic would end much sooner if all restrictions were lifted except keeping the elderly at home and 80-90% of the population were allowed to get the disease through natural transmission. 

But the name of the game is not to end the pandemic as fast as possible but to have as few people die as possible. If the epidemic were left to burn itself out, it would kill well over a million in the USA alone. As for protecting only the elderly, about 20% of COVID-19 deaths in the United States are in people younger than 65, i.e. 12,000 or so thus far. If you’re cool with those kind of figures, good for you. I’m too soft-hearted myself.

How open is reopen?

From here in Berkeley some of the “reopenings” around the country are just bringing other states around to the sensible approach California has taken ever since it pioneered the stay-at-home concept in the United States. Alaska has allowed limited in-store shopping at retail stores – I’ve already scored slippers, pillows, and computer cables at local shops. Oklahoma reopened its state parks – my husband and I meet one friend or another about once a week in Tilden to hike and chat. Iowa has allowed farmers’ markets to reopen – we bought bread, watercress, and halibut at one last Saturday. In Miami-Dade County kids can now shoot hoops individually – we enjoy watching those solo performances at all our local parks. All with appropriate social distancing, of course. But California isn’t stupid enough to reopen restaurants or movie theaters yet, like Texas, or beauty salons and tattoo parlors, like Georgia. In a movie theater patrons can at least space themselves, but can you get a tattoo or a haircut from six feet away?

A masquerade P.S.: until a few days ago, when masks became obligatory in stores, about one in 15 Berkeley pedestrians by my count covered their faces. Alameda County has had only 27 deaths per million, the same as Iceland – which is considered a model for pandemic containment. More evidence that the measures that really count are social distancing (including working from home) and hand-washing.

Italy moves forward

In Italy the epidemic is winding down. The number of COVID-19 patients in intensive care units has fallen by 58% from its peak on April 4^th, and the death rate (282 today) has been meandering downward, while it meanders upward in the United States. 

So Italy can dare to start its economy rolling again – in stages, with strict social distancing, monitoring epidemic curves and hospital capacity, and with precise criteria for when to shut things down again. On May 4^thfactories will reopen, construction restart, restaurants will open for takeaway (until now it’s been delivery only), and solo sport athletes will be allowed to train. Parks will unlock their gates, walks beyond 200 meters of your home will no longer be taboo, and people will be allowed to visit close relatives and to hold funerals – outdoors, and with fewer than 15 mourners. Buses and subways will run more often to allow for a meter between passengers. But you still can’t visit friends (though I’d imagine those nice long walks might bring a few “chance” encounters), and you’ll still need to carry a written justification every time you leave the house. On the 18^th it’s stores, museums, galleries, and supposedly team sports – with two meters between soccer players??? On June 1^st bars, restaurants, and hairdressers. So by the time Alvin and I get released from our obligatory two-week quarantine in mid-June, we should find a Rome much more liveable than the one our friends have been cooped in for more than six weeks now.

They’ve all got it?

Two weeks ago a study out of Stanford University shook up all our ideas about the virulence of the novel coronavirus. The researchers tested the blood of a supposedly representative sample of Santa Clara residents for antibodies to the virus and estimated, in a manuscript posted on the web without peer review, that between 2.5% and 4.2% had had COVID-19, an astonishing 50 to 85 times higher than the number of reported cases. This study, with its implication that the infection is much less virulent and less often fatal than usually thought, was immediately splashed across the press from The Guardian to the Wall Street Journal, and promptly picked up by right-wing websites and Fox News, offering fuel to the protesters clamoring to reopen the country for business. When I looked at the self-published paper myself, what I found was horrifying. Venial sin: they used rapid antibody tests, which are notoriously unreliable. Mortal sin: they obtained their subjects by posting ads on Facebook inviting people to have their blood tested free for COVID-19 antibodies. Now you tell me, who will volunteer to be tested for antibodies? People who think they’ve had COVID-19, of course! The researchers tossed in a bunch of statistical gobbledegook but couldn’t fix this fatal flaw. In computer science there’s a term for it: “Garbage in, garbage out.”

The senior researcher in that Stanford team, John Ioannidis, is famous for his genius at picking holes in other people’s methodology, so we can’t blame ignorance for the awfulness of this paper. Could there be some of the “financial and other interests and prejudices” – such as hunger for headlines – he himself has named as major sources of bias?

They’ve all got it

A similar study in New York City found that 24.7% of adults – have antibodies to the COVID-19 coronavirus. This study, unlike the California ones, I’m willing to believe even without access to details. The magic word is subways. There are about five and a half million rides on an average weekday, by eight million city dwellers. So well over half of all New Yorkers must have been on the subways at least once during the crucial weeks. On March 3^rd, when the coronavirus had already arrived in town and subway ridership was still at normal pre-pandemic levels, the city instituted what it quaintly thought was radical action against the coronavirus: each subway car would be sanitized once every three days. Between packed-in riders and a very contagious virus, even if the sole source of exposure were the subway system an infection rate of 25% would be no surprise.

A stable genius hires only geniuses

April 20: “We did the right thing, because if we didn’t do it, you would have had a million people, a million and a half people, maybe 2 million people dead. Now, we’re going toward 50, I’m hearing, or 60,000 people.” – Donald J. Trump, based on his scientists’ models

April 23: The COVID-19 death toll in the United States passes 50,000.

April 29: The COVID-19 death toll in the United States hits 61,655. And climbing.