Sunday, March 29, 2020

Treating the dread corona: Myths and promise

Salvarsan, the original "magic bullet": an arsenic derivative to cure syphilis
Scientists all around the globe are scrambling to find a pharmacological fix for covid-19. There’s been so much confusion, and so much ballyhooing of false leads, that I felt I should dedicate a slightly wonkish and unusually serious blog post running through the candidate medications, with help from three authoritative and up-to-datesources.
Mostly hype 
1)    Chloroquine
Doctors are hoarding this ancient antimalarial and its cousin, hydroxychloroquine, for themselves and their families. Hydroxychloroquine is now in such short supply that patients who really need it, for autoimmune diseases such as lupus and Sjögren’s syndrome, can’t fill their prescriptions. And a few poor fools are killing themselves by self-dosing with fish bowl cleaner made from chloroquine phosphate.
They’re all convinced we have found the magic bullet. But how come? 
First of all, of course, there’s Donald Trump swearing “It’s going to be great,” while Anthony Fauci tries to hold him back. The National Review has slammed skeptical articles in the press as politically-motivated.
Second, hydroxychloroquine is already FDA-approved, so it can be given off-label immediately by any doctor for any disease, and it is known not to be horribly toxic.
Third, the drug does inhibit coronavirus in test tubes
Believe it or not, that’s all there is on the positive side. In research on actual covid-19 patients the results for hydroxychloroquine have been a wash. 
Studies in China were said to have found it effective, but for a long time the original research reports weren’t available. When Western physicians finally managed to see them a few days ago, the drug turned out to have worked no better than a placebo. 
The only other study* of hydroxychloroquine, in France, made a splash by claiming that 20 patients treated with hydroxychloroquine cleared the coronavirus in their noses faster than 16 patients who received standard treatment alone. 
There are serious problems with this study, even beyond the tiny number of patients. I would suggest you consider looking at the paper for yourselves. I did, and I noticed that their Results section was oddly missing any subsection reporting patients’ clinical status. But I didn’t understand why, until a more careful reader than myself noticed that the paper did in fact mention clinical results – hidden where you wouldn’t notice them, in a subsection entitled “Demographics.” The French researchers, it turns out, had started out giving hydroxychloroquine not to 20 but to 26 patients, and the patients who received the drug actually did worse than those receiving standard treatment: three of them needed to be transferred to the intensive care unit and a fourth died. (Two others dropped out voluntarily.)  No control patient had these outcomes. These damning results suggest that the researchers may have buried their clinical findings on purpose.
One expert has commented, “Researchers have tried this drug on virus after virus, and it never works out in humans.”
As Gertrude Stein said of her native Oakland, “There is no there there.”
2)    Existing antiviral drugs
a.     Lopinavir plus ritonavir: This two-drug combination used in HIV patients performed dismally among Chinese patients with mild-to-moderate disease. 
b.     Umifenovir: Used in Russia for the flu, it was tried in China for covid-19 but similarly flopped.
c.     Oseltamivir (Tamiflu): This widely-used influenza drug has been tried in China and is undergoing some “Why not?” clinical trials, but there is no reason to expect it to be effective.
d.     Baloxavir: A brand-new antiinfluenza drug similarly set to begin clinical trials in China soon but with little hope.
e.     Niclosamide: A deworming medicine that was found long ago to have some effect against the SARS virus in test tubes but never went on to clinical trials.
3)    Corticosteroids (dexamethasone, methylprednisolone….)
Steroids have been used for years as a kind of hail-Mary pass in patients with uncontrollable bacterial infection (septic shock), so it is reasonable to try them in end-stage covid-19. One Chinese study of methylprednisolone suggested a positive effect in desperately ill patients, but its methodology was weak, and the international expert consensus at the moment is that corticosteroids are more likely to do harm than good.
4)    Azithromycin (Z-Pak)
This anti-bacterial (not antiviral) antibiotic is sometimes used to counter bacterial superinfection in serious viral illness, and has been given to some covid-19 patients in combination with hydroxychloroquine. It is known to be safe, but there is no evidence it has any benefit, and some suggestion it may do harm.
5)    Other medications
A panoply of FDA-approved drugs are currently under study in covid-19 patients on theoretical grounds, with no evidence and little hope of a clinical effect: nitric oxide (an inhaled agent used for acute respiratory distress syndrome), vitamin C, sirolimus (another kind of imunosuppressant), and losartan (an anti-hypertensive drug that many experts fear may make covid-19 worse instead of better).
Where I’d put my money
1)    Remdesivir
This antiviral, effective against a variety of coronavirus diseases, is considered by many experts to be the most promising antiviral for covid-19 and is already undergoing numerous trials. It has been used in northern Italian hospitals for weeks, when they can obtain it, and one manufacturer, Gilead, is currently expanding its availability in the US for compassionate care.
2)    Favipiravir
A Japanese antiviral that reduced recovery time and improved chest x-ray abnormalities in two trials completed in Wuhan and Shenzhen involving a total of 340 patients. It is approved for general use in China and for compassionate use in Italy, and is now formally under study, but is not available in the United States.
3)    Convalescent serum
Most people who recover from infections develop antibodies that protect them from reinfection. Products derived from their blood – either whole blood, or the antibody-rich serum that remains when you remove the red blood cells, or a more refined immunoglobulin product – can be effective drugs. For decades, a century ago before we had antibiotics, they was the mainstay of treatment for infection, and are still potentially invaluable. Such products werethe only decent drug treatment for Ebola, and there is reason to hope not only that they might keep moderate covid-19 from deteriorating, but that small daily doses might protect people at very high risk, such as health care workers on the front lines, from becoming infected. The Chinese brought some covid-19 immune serum to Italy, but soon Italy – and eventually the US – will be able to produce their own, now that there is a large enough pool of patients in convalescence. Researchers, including notably a group at my alma mater, New York’s Mount Sinai Hospital, are working to develop blood tests that will be able to identify people who have antibodies against covid-19 and therefore could usefully donate blood for this purpose.
4)    Tocilizumab (Actemra)
This “biologic,” most often used for inflammatory diseases such as rheumatoid arthritis, is an antibody against the proinflammatory cytokine interleukin-6. It has no direct action against viruses, but might help tamp down an overactive inflammatory response in the lungs of patients with covid-19 pneumonia. Preliminary results among 20 Chinese patients suggested improvement over several days, following one or two doses. Unfortunately this drug may also have the potential to promote viral replication, and long-term complications include tuberculosis and other serious infections, so it is only appropriate to try in the sickest patients. A large study has been started by Roche, and sarilumab (a similar antibody) is also being examined.
5)    Interferon
Interferon alpha, beta, and gamma (Avonex and others), like tocilizumab, have complex effects on the immune system, and they may in addition suppress viral multiplication. They are usually prescribed for hepatitis C and multiple sclerosis, and have shown promise in non-covid-19 coronavirus disease. Interferon is quite toxic, however, and like tocilizumab risks making patients worse rather than better.
Injected interferon beta is undergoing clinical trials, but it is hoped that an inhaled formulation of the same drug, code named SNG001, might give similar benefits with fewer side effects. It has been found to improve the recovery of asthma and COPD (chronic obstructive pulmonary disease) patients with lung infections, and is due to begin testing in the United Kingdom in critically ill COVID-19 patients. 
Both antiviral drugs and convalescent serum seem to be more effective early in the course of disease, while the powerful, toxic antiinflammatory medications such as tocilizumab and inhaled beta-interferon are more appropriately used late in the disease, in severely ill patients on respirators, in the hope of warding off off total body shutdown.
The final solution
Our hope for social distancing and even anti-covid-19 medications is basically that they will get us into a holding pattern until an effective vaccine can knock out the pandemic entirely. Groups in a half-dozen countries have already developed candidate vaccines, and a few volunteers have received a first dose. 
Politics has been involved even here: President Trump really really wants the best vaccine to be produced on American soil, for Americans. He apparently tried to lure a German company, CureVac, to do its research and production, if it comes to that, in the United States, and to guarantee the US monopoly rights. The company’s lead investor has confirmed that some such approach was made, and the in the end it seems the German governmentintervened.
Most vaccine research, on the contrary, has been proceeding in the spirit of international cooperation, starting with the Chinese researchers who rapidly sequenced the virus and shared their results.

But it will probably not be before at least the fall of 2021 that we will have developed a vaccine, proven it effective, and administer it to enough people that herd immunity will protect the uninfected. So for now our best bet lies in sheltering at home and other social distancing measures. Fortunately, there’s double good news on that front. One I’ve mentioned before but it bears repeating: experts say that social distancing measures can be effective even if they are less than draconian and less than constant. And the other is that today, exactly three weeks after northern Italy was locked down, my adoptive country’s rates of death and use of ICU beds are falling convincingly at last (see graphs at end). 

*Interestingly, in the 24 hours since I posted this essay, the French hydroxychloroquine researchers have changed the text of their article, though it is still at the same url. Chiefly, they now specify that 6 patients were - for unexplained reasons -given azithromycin in addition to hydroxychloroquine, and report that those 6 patients cleared the coronavirus particularly well. This changes nothing in the real results of the study: 4 among the 26 patients given active drug and no patients in the control group wound up either in the intensive care unit or dead. Other scientists and a group of sharp psychiatrists have now pointed out these issues as well.


  1. Wonderful information, Susan. Thank you!

    1. Thanks so much! I was afraid I was piling it on too much - but when I finally understood that French hydroxychloroquine article I was so shocked I couldn't stand it.

    2. You didn't understand the French paper. If you did, you would have noticed that there are 3 groups: placebo, hydroxy, hydroxy + azithromycin. Charitably, I might suppose that you were blinded by end stage TDS.
      If you were smarter, it might also occur to you that the clearance speed of hydroxy + azi could be very promising as a prophyactic. Both drugs are cheap and relatively safe.

    3. I'm happy to reply to this one. About azithromycin: the text says "Symptomatic treatment and antibiotics as a measure to prevent bacterial super-infection was provided by investigators based on clinical judgment," but provides no results related to any antibiotic. Not 3 groups, whatever TDS means. About clinical outcomes, let me quote directly from the paper so other readers can judge the legitimacy of the researchers' methods: "Six hydroxychloroquine-treated patients were lost in follow-up during the survey because of early cessation of treatment. Reasons are as follows: three patients were transferred to intensive care unit, including one transferred on day2 post-inclusion who was PCR-positive on day1, one transferred on day3 post-inclusion who was PCR-positive on days1-2 and one transferred on day4 post-inclusion who was PCR- positive on day1 and day3; one patient died on day3 post inclusion..." It may be that real clinical trials will find Plaquenil to be effective despite the absence of anything promising in this paper or in the Chinese studies - that would be wonderful, but I'm not holding my breath.

    4. Addendum: as I have now added to the text, after I posted this essay the French hydroxychloroquine researchers changed the text of their article to distinguish 3 rather than 2 groups. This changes nothing substantive in my critique.

  2. Frank Savoretti, JD, MDMarch 30, 2020 at 2:03 AM

    EXCELLENT review article! Thank you. Send it to the NEJM & NYTIMES & WSJournal for wider exposure. Have not seen anything similar yet in the medical or general media.

  3. I really appreciate this post Susan, thank you ❤

  4. Susan, thanks and great synthesis article. You should be writing for UptoDate, forget NY Times!
    Steve Novek

  5. I am a 73-year old male smoker with only two coronary arteries left, one of those has 3 stents in it. I am reasonably convinced that I had a mild infection of the virus in January, which took 10 days to clear up.
    I have been taking Telmisartan daily for 14 years. I read that the drug changes the ACE2 producing cells which are apparently the preferred target of the virus when it reaches the lower respiratory tract and lung.
    Any view?
    Thanks, Newell White.

    1. Good question. The ACE connection with coronavirus is well known, but experts are not at all clear on how it works out. Some think drugs like telmisartan could worsen covid-19, others think they could help cure covid-19. In the end all the important organizations have come out with official guidelines saying that people who are already taking ACE inhibitors and sartans should not stop or change these drugs if they get sick with covid-19. In case you are interested, you can go to

  6. really excellent. thanks for posting!!!!!

  7. You're adoptive country's stats are garbage:
    I thought you could read Italian.

    1. Yes, there have been problems with Italian death statistics, with underreporting due to the inability to continue the widespread testing and contact tracing the NHS was achieving at the onset of the outbreak, as supplies ran out and hospitals near the epicenter were overwhelmed. That inability is in turn due to chronic and acute underfunding, related in large part to EU-imposed austerity. Fortunately despite the reporting problems there is good reason to believe that the situation is improving. I might suggest you consider taking a more polite tone in your comments.

    2. Bravo on a wonderful, clear review paper amid much noise in the media. Your article definitely deserves much broader attention. And super bravo for taking the high ground in your responses, especially to this poster.

    3. Craken, By all means make your points, but perhaps try to keep the tone a bit more civil?

    4. I had read about nastiness on the web, but the person who hides behind the handle of Craken has given me the first opportunity to see it in person. Thanks for the support.

  8. Thank you for the information - it’s comforting to have a clear scientific explanation.

  9. So glad to have your learned take on present medications and results, as we go through this. You are breaking trail here, with reason and information to help build our knowledge bank. Mega-thanks! Stay safe.

  10. Great article! Have you heard of this other possible treatment, leronlimab? It seems to be showing great results with COVID-19.
    Thanks again for the article. Be well.

    1. No I didn't, actually, thank you for drawing it to my attention. I'll include it in my next update. But I gather it's only been tried in two patients so far, so it's far too early to make predictions.