Skip to main content

Cardiovascular Disease

Angiotensin and Coronavirus Infection: The Latest as of April 7

I wrote here the other day on the recent recommendation that people taking either ACE (angiotensin-converting enzyme) inhibitors or ARBs (angiotensin receptor blockers) should not alter their treatment regimens because of the coronavirus outbreak. Some background on the angiotensin system is here, and here’s an open-access review for those who know the basics and want to dive much deeper into the details. But now there’s even more to report on the subject, so I wanted to do an update post.

Recall that the coronavirus itself uses the ACE2 protein as an entry point into cells. One worry has been that the use of antihypertensive drugs (of either class mentioned above) might well cause ACE2 expression to increase, which seems as if it could be a bad idea, providing more targets for the virus to latch on to. But this survey of the literature found little evidence that these expression changes even happen. The animal data that show these effects, they report, tend to be via acute injury models or doses that are much higher than human patients encounter, and there seems to be no good evidence that it happens in humans. So that’s one thing to think about: a big part of the worry about antihypertension drugs may not be even be founded on a real problem.

We also have some clinical data: this preprint from a multicenter team in Wuhan retrospectively evaluates 43 patients with hypertension who were taking drugs in these two classes versus 83 hypertension patients who were not taking ACE inhibitors or ARBs, versus. 125 age- and gender-matched controls without hypertension at all. They also compared hospital admission statistics in general to patients’ medical histories. They first confirmed what others have found, that hypertension itself is a risk factor: the patients admitted for treatment had higher levels of hypertension than the general population, and once admitted those patients had higher death rates and longer hospital stays. But when they looked at the hypertension patients who were taking either ACE inhibitors or ARBs, their numbers were better. They had comparable blood pressure numbers to those taking other drugs, but they were a lower percent of critical patients (9.3% versus 22.9%, near miss on statistical significance) and had a lower death rate (4.7% versus 13.3%). The ACE/ARB cohort also had lower inflammation markers (c-reactive protein and calcitonin). So while the data are noisy, there may be a trend towards protection in those taking angiotensin-targeting drugs. All the more reason to heed the advice not to change therapies for people with hypertension.

This recent paper speculates that there could be a biphasic aspect to all this. Note that it assumes that ACE2 may be upregulated by these drugs, which (see above) may not even be the case – if that’s true, then we go from biphasic (bad before infection, good during infection) to more of a pure benefit:

At present, we cannot rule out that long‐term intake of ACEIs and/or ARBs may facilitate SARS‐CoV‐2 entry and virus replication. Conversely, it is yet unknown whether intake of ACEIs and/or ARBs, when infected, is beneficial with regard to pulmonary outcome. Possibly, we are dealing here with a double‐edged sword, depending on the phase of the disease: increased baseline ACE2 expression could potentially increase infectivity and ACEI/ARB use would be an addressable risk factor. Conversely, once infected, downregulation of ACE2 may be the hallmark of COVID‐19 progression. Consequently, upregulation by preferentially using renin‐angiotensin system blockade and ACE2 replacement in the acute respiratory syndrome phase may turn out to be beneficial.

What, though, if you deliberately give someone ACE2 protein? An email correspondent wrote to me a while ago wondering about doing that in a nebulizer, and my response was that it wasn’t a crazy idea, but that I didn’t know if anyone would go to the trouble of making and testing a whole new recombinant protein under the current conditions. I had forgotten about Apeiron! They have been developing an infused (not nebulized) recombinant version of the human ACE2 protein for some years now, and were in a deal with GSK to look at respiratory distress in general. The deal lapsed last year after a good deal of work in the clinic and a refocus at GSK away from respiratory disease in general. The clinical data were consistent with the mechanism: infusing ACE2 enzyme dropped the amount of angiotensin II protein, as it should, since it’s in there clipping it up, and raised the amounts of the shorter product peptides. So the mechanism was operating as it should – the problem was that it didn’t seem to help ARDS patients much (!)

But a viral infection that causes respiratory distress and enters cells through membrane-bound ACE2 proteins, well, that’s different. You could imagine that circulating ACE2 protein would be a decoy for the virus – it would bind as if it’s infecting a cell, but the protein in this case isn’t on a cell surface at all, so you could potentially soak up the virus and take it out of commission. They and their collaborators have just demonstrated this in cell culture, actually: adding the recombinant protein to a viral replication assay seems to have caused a large decrease in viral RNA, whereas adding the mouse version of ACE2 had no effect at all.

Apeiron is jumping back into the clinic, and a trial has already started in China, with one getting underway in Europe as well to see if the treatment can decrease the number of patients going on to ventilators. This will be quite interesting; to the best of my knowledge this and perhaps camostat are the only agents directly targeting the ACE2 viral mechanism (and that one’s being studied as well). And we only have that by good luck, that Apeiron had already been looking at the mechanism for respiratory distress in general. Those earlier failed trials have now provided a quick entry into human studies!




36 comments on “Angiotensin and Coronavirus Infection: The Latest as of April 7”

  1. dennis says:

    On a tangentially related issue-are you seeing anything about the use of FXa inhibitors, given the role a few papers show for FXa in cleaving the spike protein and therefore, with viral entry? Given the number of folks on FXA inh’s, there is a population/natural experiment providing quick insights…

  2. Me says:

    Hi Derek I wonder if there’s anything in there about ethnicity? Black patients with hypertension are generally not prescribed ACE inhibitors due to reporting of worse outcomes for them on these drugs. Instead they are usually prescribed ARBs, diuretics or calcium-channel blockers.

    Wonder if this leads to different outcomes with COVID….

    1. Derek Lowe says:

      ARBs seem to be protective as well, though, if this latest work holds up.

    2. M says:

      It’s a study in Wuhan, China: I’d be surprised if there was even one black patient.

  3. Oudeis says:

    That’s really clever, and I’ve never heard of a drug working that way before.

    Are the other drugs that work by decoying pathogens?

    1. James says:

      As someone who is definitely not in medicine the only two things similar that I can come up with are:
      1) Chelation for heavy metals and radioactive species.
      2) Replacement(?) therapy for certain radioactive species. Like iodide tablets to flood the system with non-radioactive iodide to reduce how much radioactive iodide the thyroid takes up (and incorporates into proteins whose biological half-life is much longer than free iodide)

      1. loupgarous says:

        The analogy is pretty loose, but yeah, either way you’re arranging for a dangerous entity to be bound to something besides human tissue where it can cause disease. The ACE2 upregulators (and recombinant human ACE2) could do that (though this is largely handwaving, working to validate the hypothesis at this point’s not a waste of time).

      2. Cliff says:

        What you’re proposing is not similar at all. It’s New Age silliness.

    2. Derek Lowe says:

      It’s been proposed a number of times, such as using oligonucleotides as transcription factor decoys. It’s been hard to realize in practice, though. . .

      1. RonH says:

        There is a short roundup of some decoy receptor work on Wikipedia:

      2. Dan says:

        Enbrel is a decoy TNF receptor isn’t it?

        1. loupgarous says:

          Our good friend wikipedia thinks so:

          “It is a large molecule, with a molecular weight of 150 kDa, that binds to TNFα and decreases its role in disorders involving excess inflammation in humans and other animals, including autoimmune diseases such as ankylosing spondylitis,[6] juvenile rheumatoid arthritis, psoriasis, psoriatic arthritis, rheumatoid arthritis, and, potentially, in a variety of other disorders mediated by excess TNFα.”

    3. Chris Phoenix says:

      Back in the early 2000’s Chris Schafmeister developed a new way to link protein monomers, with two bonds instead of one. The resulting structure was quite rigid – he said you could predict the structure with a ball-and-stick model. At the time, there was talk of using this to hold virus-velcro at the right spacing and orientation to stick to several points on the virion at once, for very strong and selective binding. I don’t know if anything ever came of that.

  4. Toni says:

    Am I getting something wrong? I do not see any serious differences in blood pressure values between normal patients and hypertonic patients w/o drugs.

    1. Lester Kobzik says:

      Toni, I don’t see any mention of patients with hypertension that were not on any drugs, i.e. Table I divides the hypertensives into people getting ARBs/ACEinhibitors or those not (getting other types of anti-hypertension drugs “Non-ARBs/ACEis”)

      1. Toni says:

        Ah, right. Thanks to you.

  5. Daniel Barkalow says:

    The Apeiron thing seems like a good argument for having a list of drugs that are safe and have their designed effect, but aren’t effective at treating any known disease, and reporting safety on drugs whose trials find them ineffective. “This drug is a safe way to get a lot of extra-cellular human ACE2, with no significant helpful or harmful effects in patients with respiratory distress” could turn out to be a valuable thing to have proven, even if it is useless at the time. It’d probably be worth having the government fund getting that sort of failed trial written up as something that could be repurposed if it ever seems useful in the future.

  6. mayfin says:

    I’m no doctor, but Aperion APN-01 sounds like it could be a case of “so near, yet so far” …

    Given it failed to show efficacy in trials against ARDS, it may well not be that useful once the infection is established. As a decoy, it would seem most useful at the start of an infection, or better yet as a prophylactic – but how in practical terms do you deploy that for a drug you’ve described as an infusion ? Hook healthcare workers up to an IV for 30 mins before the start of each shift ?

  7. Adam Zweifach says:

    As far as I know, ACE inhibitors don’t touch ACE2, so whatever is going on- if anything- it must be complicated. One could certainly start speculating, but IMO that’s not warranted until we’re much more certain that there’s something real there.

    Good thing no one has suggested taking ACE inhibitors or ARBs just because there’s nothing to lose and they’re safe, or this post’s comments would be as polluted as the last hydroxychlorquine post’s.

    1. loupgarous says:

      I take lisinopril now, have done for years. I’m driving into New Orleans for a PET scan (can’t get one anyplace closer, and that’s where my oncologists practice). I’m a test subject for that hypothesis AND following medical advice from several vectors regardless of the hypothesis.

      1. Chris says:

        Would love to hear how it goes. I’m on Lisinopril now.

  8. Elliot says:

    One wrinkle not mentioned here in the ACE/RAS/Coronavirus story is that Angiotensin(1-7), the product of ACE2 cleavage appears to be as potent down regulator of IL-6 and other inflammatory cytokines thought to be involved in the disease pathology. Turns out there is a widely available, trivial to synthesize, and FDA approved(!) compound which is an “off-target” agonist of ACE2 and has been shown to potently reduce IL-6 and inflammation. So I wonder why noone looking into Diminazene (DIZE) as a therapuetic? Sure, its widely available status is as a veterinary antitrypanosomal (think African sleeping sickness) although it has been used successfully for that in humans as well. It fell out of favor due to side effects in some animals, but people are desperate and nothing like Actemra or remdesivir is going to scale well. Doesn’t sound any less crazy than that other antiparasitic with known severe side effects being given out like candy.

  9. bks says:

    On This Week in Virology podcast #598 a doctor in the NYC area describes using steroids and Tocilizumab (I think that’s what he says) for severe cases with good clinical effect, at about the 10:20 mark:

  10. KazooChemist says:

    Never fear, help is on the way. Martin Shkreli is seeking a furlough from prison so he can discover a cure for COVID-19.

    1. Pedwards says:

      I’m setting up a clinical trial to test if repeated high-velocity application of a clenched fist to a person’s cranium will help reduce their viral load. I think he is the perfect candidate for the one spot in the treatment arm.

    2. loupgarous says:

      Shkreli sure has chutzpah. He seems to have forgotten that Daraprim, his moneymaker, was discovered by Burroughs Wellcome, who weren’t looking for a way to gouge patients and their insurance companies. That company ploughed its earnings, back then, into more research.

      I know, “once upon a time”. But anyone who takes Shkreli’s word that he can do anything but separate other people from their money is deluded.

    3. x says:

      If we ever discover that price gouging cures disease, he’ll have a solid case.

  11. Some idiot says:

    Two questions (I am a process chemist, and not a clinician/pharmacist…)

    1: If it were to be used IV as a decoy, then I would guess that the protein concentrations used IV would have to be pretty high to be competitive, but I would guess that someone has done the maths on that one…? And would depend on the stage of the progression, of course…
    2: Given that the virus enters through airways (nose as well?), then an inhaled formulation (or nose spray?) would presumably give much more protection as a prophylactic (or bang for your buck), since the decoy would be on the “outside”. The doses required would also be significantly lower (as an asthmatic who has been on inhaled formulations for decades now, but is old enough to remember the side effects from when taken orally, I can vouch for that!!! (-: ).

    Could be useful for health workers…?

    Any thoughts/comments?

    1. Ilya says:

      Good questions. I wonder how much of the protein is even able to penetrate from the bloodstream through the endothelium, the basal membrane and reach airway epithelium given the aggressive inflammatory environment full of proteases

  12. Greg says:

    Perhaps a naïve question but if ACE2 is acting (mainly) as decoy for COVID19 in terms of delaying/preventing infection then this is unlikely to be dependent on its enzymatic activity i.e. is a mutant (catalytically dead) version of ACE2 equally as efficacious? The Monteil et al., 2020 publication that has just appeared talks about this but doesn’t test this hypothesis by creating such a mutant. Would be really important to perform this experiment pre-clinically in my opinion.

    1. Derek Lowe says:

      Good question. The enzymatic activity didn’t seem too useful in the ARDS trial, but we don’t know about the lung pathology in this case.

  13. Birone says:

    “But this survey of the literature* found little evidence that these expression changes even happen. The animal data that show these effects, they report, tend to be via acute injury models or doses that are much higher than human patients encounter, and there seems to be no good evidence that it happens in humans.”

    In reaching this conclusion, the authors somehow ignore one of their own comments:

    “In one study [Furuhashi et al.] the investigators observed that subjects using the ARB Olmesartan had increased ACE2 levels, but several other ARBs and ACEIs had no effect.”

    In the Furuhashi study urinary ACE2 protein concentration was assayed in 617 subjects, including 101 untreated subjects and 100 hypertensives treated with various drugs.with the finding that “In contrast with other antihypertensive drugs, olmesartan may uniquely increase urinary ACE2 level”.

    Taken alone, this is not strong evidence. Notably n(olmesartan) was only 13 and the olmesartan group had higher HbA1c; so a preference towards prescribing olmesartan to diabetics – eg for reno-protective effects – could have had a confounding effect (iirc diabetics tend to already have raised urinary ACE2?). And for the five ARBs consideredm p<0.05 would be observed ~25% of the time even if all mean urinary ACE2 followed the null hypothesis distribution .

    BUT "Whether the effects of several ARBs on ACE2 expression are same remains uncertain…. The present finding that the plasma Ang II level was decreased after replacement of candesartan with olmesartan without any changes in PRC suggests that [**the effect of olmesartan on ACE2 expression is more stimulating**] than that of candesartan" – so as long as candesartan doesn't decrease ACE2 expression, olmesartan increases it relative to baseline.

    AGAIN this is not perfect evidence: Furushahi et al.'s candesartan group did in fact have slightly ('insignificantly') lower urinary ACE2. So we can't be totally sure of the baseline for olmesartan.

    BUT there is some animal evidence pointing to olmesartan raising ACE2 expression:

    One might also posit an evolutionary basis for the existence of a homeostasis mechanism governing ACE2 shedding, conserving it on the membrane when scarce, and eliminating it from the cytoplasm when that's gets 'overcrowded' with ACE2. So modulation of ACE2 expression might also modulate the soluble-to-membrane bound ACE2 molecule ratio (a better candidate for interfering with SARS-CoV2 binding than soluble ACE2 levels on a standalone basis?).

    And Olmesartan, an inexpensive, FDA-approved, well-studied drug might just raise ACE2 expression in a way that raises that ratio to interfere with SARS-CoV2 binding, and meanwhile also has other helpful effects vs SARS-CoV2 (it moderates blood glucose, has lung- and other organ protective effects, and may stimulate innate immunity in sufficiently high doses.)

    I'm NOT proposing anyone takes olmesartan on the strength of this. But I do think that amongst the ARBs*** it's worth checking out as a therapeutic. And at the very least the evidence points to differential affects amongst the ARBs, justifying their individual treatment in the statistical analyses (where possible).

    ***which were fairly effective against Ebola, whose lethality also stems largely from epithelial dysregulation:

  14. Jasmina says:

    Please, read these and think about it.
    Immunology 1998 94 376–379
    The effect of renin–angiotensin system inhibitors on pro- and anti-inflammatory cytokine production
    A. C. T. M. PEETERS, M. G. NETEA, B. J. KULLBERG, T. THIEN & J. W. M. VAN DER MEER Division of General Internal Medicine, Department of Medicine, University Hospital Nijmegen, Nijmegen, The Netherlands

    Cell Mol Neurobiol. Author manuscript; available in PMC 2010 Sep 1.
    Published in final edited form as:
    Cell Mol Neurobiol. 2009 Sep; 29(6-7): 781–792.
    Published online 2009 Mar 4. doi: 10.1007/s10571-009-9368-4
    PMCID: PMC2718067
    NIHMSID: NIHMS118627
    PMID: 19259805
    BENICKY Julius, SÁNCHEZ-LEMUS Enrique, PAVEL Jaroslav, and SAAVEDRA Juan M*

    BMJ Case Rep. 2014; 2014: bcr2014205462.
    Published online 2014 Jul 17. doi: 10.1136/bcr-2014-205462
    PMCID: PMC4112303
    PMID: 25035451
    Case Report
    ACE inhibitors: upper respiratory symptoms
    Paulette Pinargote, Denisse Guillen, and Juan C Guarderas

    Conference Highlights » AAN 2019 Conference
    Publish Date May 8, 2019
    ACE Inhibitors, ARBs May Be Effective Migraine Prophylactic Treatments
    Florence Chaverneff, Ph.D.
    Current Drug Therapy
    ACE inhibitors and ARBs: Managing potassium and renal function
    Tasnim Momoniat, MBCHB, MRCP (UK), Duha Ilyas, MBBS, MRCP (UK) and Sunil Bhandari, MBCHB, FRCP, PhD, M CLIN EDU, FHEA
    Cleveland Clinic Journal of Medicine September 2019, 86 (9) 601-607; DOI:

    Rapid response to:
    China coronavirus: mild but infectious cases may make it hard to control outbreak, report warns
    BMJ 2020; 368 doi: (Published 28 January 2020) Cite this as: BMJ 2020;368:m325

  15. An Old Chemist says:

    I have found an interesting correlation between statin usage and covid-19 mortality. Here is my hypothesis:

    Statin-Covid19 Hypothesis (SCH): People who are taking statins are at a significantly greater risk for covid-19 complications and mortality.

    I will use covid death rate per million of population in each country as the covid metric to support the SCH, especially since the reliability and availability of testing and testing accuracy vary so widely. But I believe that this hypothesis would remain basically unchanged even if we used the number of confirmed infections (instead of deaths) as the covid metric.

    I have picked a bunch of representative countries from the northern hemisphere (in case covid is seasonal). See details below.

    I don’t have statin usage data for all the countries listed here. So far I have gathered some data by searching for each country, one at a time. I am still looking for a single paper that compares the statin usage in different countries. Here is the outline of my evidence so far.

    Statin usage by country: is much higher in developed/richer countries like US, UK etc where the death rates per million of population are the highest. It is very low in countries like Ethiopia where the covid death rate is very low. Cuba, our neighbor, is an interesting example. Because of US trade embargo, US drugs are not easily available there, so their statin usage is quite low (haven’t found the exact %). Their death irate is 7 per million compared to about 100 for nearby Florida.
    Statin usage by age: is highest in the same age groups where the covid death rate is highest. The patterns match rather well.
    Statin usage by gender: More men use statins (the % varies by country). More men also die of covid.

    Physiological Evidence:
    It is well-known that statins suppress the immune system.

    “These different approaches show that statins affect the immune response on multiple layers including signalling, gene transcription, epigenetic modifications and immune metabolism.”
    “The potency of statins with respect to immunosuppression is lower than that of many of the established immunosuppressive drugs…
    “They found that levels of antibodies directed against the influenza virus were lower in patients taking statins, compared with patients who were not taking statins.”

    Note that covid mortality is a function of many variables, so a near-perfect correlation with any single variable (like statin usage) should not be expected.
    Source: Johns Hopkins, May 19
    Country #deaths Popul. #deaths/Million
    Vietnam 0 96.5 Mil. 0
    Ethiopia 5 112 0.04
    Sri Lanka 9 21.3 0.4
    B.Desh 349 163 2.1
    India 3,163 1,370 2.3
    Japan 749 127 5.9
    Egypt 645 100 6.5
    Cuba 79 11.3 7.0
    Phillippines 837 108 7.8
    Russia 2,837 146 19
    Israel 277 8.5 33
    Mexico 5,332 128 42
    Iran 7,057 82.9 85
    Germany 8,007 83.5 96
    Canada 5,842 37.4 156
    US 89,354 330 271
    Sweden* 3,698 10.0 370
    UK 34,796 67.5 515
    Italy 32,007 60.6 528
    *Sweden data is without an economic shutdown
    *** ** *
    Here is a paper with data on statin use in Europe.
    Interestingly, statin use correlates with severe irritability

  16. Lydia says:

    I have a question. My mom caught Covid a while back and drank rooibos tea when she was developing a slight urge to cough and tightness in her chest. She drank a cup and it eased the tightness, then she drank another cupful before she went to bed. When she woke up in the morning, her lung symptoms were gone. She thinks drinking the tea kept her from getting worse and going to the hospital. Is there any scientific explanation for why rooibos could have that sort of effect?

Comments are closed.