An old drug is getting new attention, but is it for the right reasons?
In regular press briefings on COVID-19, President Trump and members of his administration have repeatedly touted a potential treatment: hydroxychloroquine. On Twitter the president described it as potentially ?one of the biggest game changers in the history of medicine.? However Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID) and prominent member of the coronavirus task force, has tried to temper enthusiasm for the therapeutic regimen, citing a lack of evidence. So what?s driving the hype around this old drug, and is it deserved?
What is (hydroxy)chloroquine?
Centuries ago, indigenous Peruvians used Cinchona tree bark to treat chills and fever. Later, Europeans realized this herbal medicine could treat malaria and in 1820 the active ingredient, quinine, was extracted. Chloroquine, a related chemical (called an analog) was discovered in 1934 by Bayer Laboratories and approved in the US to prevent malaria in 1947. Hydroxychloroquine (HCQ) is another analog with fewer side effects and since it?s the most widely used version in the US, I will refer to it for the rest of this post.
What is HCQ usually used for?
HCQ is approved for the prevention and treatment of malaria, which is a disease caused by single-celled microorganisms of the Plasmodium genus. The parasite is not a bacterium so it can?t be treated with traditional antibiotics, and it has plagued the human race for thousands of years. The reason malaria is so deadly is because the parasite lives in red blood cells and digests hemoglobin, a critical protein that transports oxygen throughout the body. HCQ is thought to kill Plasmodium by preventing it from breaking down heme, a toxic byproduct of hemoglobin digestion. The parasite detoxifies heme within an acidic organelle (specialized compartment in a cell) called a food vacuole. HCQ is a weak base so it accumulates in the food vacuole and disrupts the parasite?s metabolism.
HCQ also seems to have broad immune suppressive effects that benefit patients with auto-immune disease. The drug is also used to treat some auto-immune diseases including Systemic Lupus Erythematosus (lupus) and Rheumatoid Arthritis (RA). Lupus affects many different organ systems including the skin, joints, brain, heart, kidneys and lungs, while RA primarily involves the joints. Patients with lupus or RA experience chronic inflammation (often in the form of swelling or rashes) because their immune system is always activated and on high alert, attacking normal tissues during flares. It?s not entirely clear why HCQ is effective against these diseases, but it?s been hypothesized to act in a couple of different ways. Firstly, it?s thought to act by neutralizing an acidic organelle called the lysosome. Specialized immune cells called antigen-presenting cells use the lysosome to break down pathogens and present pieces of them to other immune cells and activate a response. In an auto-immune disease context, antigen-presenting cells display pieces of normal cells instead of bacteria and viruses, so HCQ prevents that from happening. A second way the drug affects the immune system is by disrupting signaling via the Toll-like receptor (TLR) pathway which requires another acidic organelle, the endosome. TLRs recognize foreign DNA (or self DNA in the case of auto-immune disease) and use endosomes to transmit the information to the nucleus and activate an immune response. So HCQ might help auto-immune patients by preventing the immune system from attacking the body.
What is the evidence that HCQ might work for treating COVID-19?
Many viruses have pH-dependent steps in their life cycle that HCQ could potentially perturb. In fact, other known coronaviruses utilize acidic endosomes to infect target cells and chloroquine has been shown to inhibit this process in a petri dish (in vitro). Moreover, sometimes the immune system?s overreaction to a viral infection causes more damage than the virus itself, and HCQ has a proven track record of immunosuppression. It has the added advantage of being an old, widely used drug that is generally safe (more on that later). With these reasons in mind, many clinical trials are now underway to test whether HCQ is an effective treatment against COVID-19. A few of these trials are completed but the results have been a mixed bag, so let?s take a closer look at what they found.
Researchers from Zhejiang University in China tested whether HCQ could reduce the presence of SARS-CoV-2 RNA in throat swabs (which would suggest it helps clear the infection). Thirty patients were split into two groups and half received HCQ. After seven days of treatment, they analyzed viral RNA in throat swabs from the two groups and saw no difference (13/15 in the HCQ arm and 14/15 in the control arm were negative). This trial could be interpreted two ways: HCQ does nothing or the researchers waited too long to look for the virus. For example, maybe HCQ was effective and reduced viral load at after three days while the controls cleared it at six days, but we can?t know because of the way the study was designed.
The first trial to suggest that HCQ is effective against COVID-19 came out of France and received a lot of attention. Researchers at the M?diterran?e Infection University Hospital Institute in Marseille recruited 26 patients for the HCQ arm of the study, while 16 control patients were treated in four different centers in South France. To be included, patients had to be over twelve years of age and positive for SARS-CoV-2. All patients received treatment for symptoms and some were given antibiotics to prevent secondary infections, but the HCQ patients also received the drug for ten days. On day six, 70% of the HCQ patients had reduced viral levels compared to only 12.5% of control patients. The effect was even more dramatic when taking the antibiotic, azithromycin (Z), into account: 100% of patients who got HCQ + Z were negative for the virus on day six, while 57.1% of patients who got HCQ alone had less viral RNA.
Sounds promising right? But there were some flaws in how the trial was carried out. Firstly, the trial was not randomized or blinded so both the researchers and the patients knew if they were getting the experimental treatment. This can introduce bias on the side of the researchers and placebo effect for the patients. Second, the control patients and HCQ patients were treated at different hospitals, which is not ideal. A patient?s outcome can differ dramatically depending on where they get treated. Third, patients who were excluded from the HCQ arm due to pre-existing conditions like long QT syndrome (a heart issue) were included in the control arm. Perhaps these people were already sicker to begin with? It?s important for both arms to have very similar patients in them so that the only variable is treatment. Fourth, the researchers ended up excluding six of the 26 HCQ patients for various reasons, including death (1), transfer to the ICU (3), quit due to side effects (1), and left the hospital (1). Patients drop out of trials all the time for reasons like this, but the results might have looked a lot less exciting if, for example, the very sick HCQ patients who ended up in the ICU were included. Fifth, it was unclear how many of the control patients received the antibiotic. Perhaps they did worse because they didn?t receive Z and developed secondary infections, not because HCQ is working. And finally, sixth, the endpoint the researchers chose (viral RNA levels on day 6) don?t tell us much about how sick the patients were. A high frequency of false negatives has been reported for the PCR test, with patients testing negative one day and positive the next, so it?s possible that a person negative for the virus on day 6 is still quite ill. The ultimate goal of treatment is to heal patients, not just reduce the amount of virus in their system.
Unfortunately, there were additional problems beyond the design and execution of the trial that raise questions about the credibility of the research group. First, based on the timeline reported in the paper, it seems the researchers started the trial before they got ethical approval to do so (a big no-no). Second, the peer-review process is questionable, because one of the authors is the editor-in-chief of the journal it was published in and it only took 24 hours to be accepted. Let me provide some insight into how peer review normally works: the editor reads your manuscript, finds 2?4 experts to review it (which takes about three weeks, if they?re timely) and then you spend at least three months improving the study before the reviewers take a second look at it. This is an emergency, but 24 hours is not a lot of time to carefully evaluate such an important contribution to medicine. The bottom line is, while this trial provides some evidence HCQ + Z is effective against SARS-CoV-2, it?s certainly not definitive.
A second French group out of Paris attempted to replicate the same HCQ + Z treatment regimen in the Marseille study to see if it was effective in eleven COVID-19 patients. They found, after six days of HCQ treatment, 8/10 patients were still positive for SARS-CoV-2 RNA on day 6 (one patient died). This is in contrast to the Marseille study which found that none of the HCQ + Z patients were positive on day 6. The Paris group reported no improvement in clinical outcomes for their patients.
The Marseille research group conducted a second trial with 80 patients receiving HCQ and found a reduction of viral RNA over 14 days. In the control arm they found? nothing because there was no control arm. The patients enrolled for this study had mild disease: few had fevers, none needed to be hospitalized and most had already been through five days of symptoms. These are people who were probably going to get better on their own, so again, not great evidence that HCQ is effective.
The city of Wuhan was the epicenter of the coronavirus outbreak and therefore presents a unique opportunity to test potential treatments. Researchers there were able to conduct a small randomized controlled trial (yay!) during the month of February to test the effectiveness of HCQ. They enrolled 62 COVID-19 patients, which were split into two groups with similar characteristics in terms of age, sex, and severity of illness (all had mild pneumonia confirmed by chest CT scan). After five days of standard treatment, with or without HCQ, the patients were assessed for fever, cough and pneumonia (much more informative than viral RNA levels). Recovery from fever and cough occurred on average one day earlier in the HCQ group. Pneumonia improved on day 6 for 80% of HCQ patients compared to 55% of control patients. Four patients got sicker over the course of the trial and they were all in the control group, and two patients in the HCQ group had bad side effects. So what does this mean? Well, it seems like HCQ sped up recovery by one day for these patients. That?s great but it?s not the miracle cure that some in the government and media have been howling about. But at least we can believe the results of this trial: it was well controlled, patients were randomized, and they measured clinical outcomes instead of viral RNA. The only ways to improve this study would be to have it double-blinded (neither the patients nor the researchers know who?s getting HCQ) and enroll thousands of patients, which is exactly what new clinical trials around the world are trying to do.
HCQ clinical trials in progress
As of 4/7/2020, there are 239 clinical trials for ?coronavirus? listed at clinicaltrials.gov. Twelve of them are testing the effectiveness of chloroquine or HCQ in treating COVID-19 and it looks like the only one that?s been completed is the Wuhan study. Trials are in the recruiting phase in Norway, France, and Brazil and more trials have been registered (but not started yet) in China, Canada, Australia, France, Thailand and the US. The two American trials taking place at the University of Minnesota and Washington University School of Medicine in St. Louis, Missouri will test whether HCQ can prevent symptomatic COVID-19 in healthcare workers who have been exposed to the virus.
What?s the danger in moving ahead with hydroxychloroquine while we wait for trial results?
For most patients there is little danger, but for some it?s not without serious risks. HCQ has been used for decades so we know it can be safe under the right conditions, but it can cause serious side effects, interact adversely with other drugs, and be fatal for people with certain pre-existing conditions. The most common side effects during short-term use are nausea, diarrhea, vomiting, headache, and reduced appetite. One of HCQ?s less common but more serious side effects is heart arrhythmia, which can be fatal for people with cardiac issues. Other medications, including the azithromycin (Z), can also cause heart arrhythmia so these drug interactions must be taken into consideration before HCQ is prescribed. Another side effect of HCQ is hypoglycemia (low blood sugar), so using this drug can be risky for diabetics. Heart disease and diabetes are two of the most prevalent chronic conditions in the US so these adverse effects should not be taken lightly. Another reason doctors should be cautious when prescribing HCQ is because the demand is already starting to exceed the supply. Lupus and RA patients take HCQ everyday to prevent flares, and they?re already being forced to ration doses because they can?t get their prescriptions filled. People with a proven need for HCQ are suffering because of unmerited hysteria over the drug.
Doctors are legally allowed to prescribe FDA-approved drugs like hydroxychloroquine for ?off-label? purposes (beyond what they were originally approved for), and maybe it will work against COVID-19. It would be fantastic if it works. But we won?t really know for sure until the clinical trials are done. In the meantime, let?s hope that doctors are being thoughtful when prescribing HCQ and that it really does help these patients get better.