New Drug Repurposing Approach Could Speed COVID Therapies To The Clinic
By Deborah Borfitz
September 7, 2021 | In under a year, researchers at the University of Michigan (U-M) have moved from discovery to phase 2 clinical studies for a pair of drug contenders in the fight against COVID-19. The credit goes to a program, newly launched when the pandemic struck, which “systemizes” drug repurposing by helping academic investigators look for existing drugs before trying to create new ones, according to Jonathan Sexton, Ph.D., assistant professor of internal medicine at U-M.
One of the clinical trials, a collaboration with Glanbia Nutritionals, will be conducted locally and look at the anti-viral and anti-inflammatory effects of the dietary supplement lactoferrin, Sexton says. The other planned study will test lactoferrin in combination with the experimental prostate cancer drug proxalutamide, which in a recent preprint claims to have reduced deaths in hospitalized COVID-19 patients in Brazil by 77%.
While the pharma-sponsored proxalutamide study sparked controversy, and Sexton agrees the findings need to be independently confirmed, he is intrigued. “It one of the most exciting small molecule therapeutic strategies that exists right now in terms of the magnitude of the effect, which is much greater than what we saw with remdesivir.”
Sexton’s interest in proxalutamide, which blocks the activity of androgens (male hormones such as testosterone), stems from a study that he and Arul Chinnaiyan, M.D., Ph.D., published late last year in PNAS (DOI: 10.1073/pnas.2021450118) pointing to the ability of androgen receptor inhibitors to reduce the SARS-CoV-2 host entry factors angiotensin-converting enzyme 2 (ACE2) and the cell surface transmembrane protease serine 2 (TMPRSS2), thereby halting infection.
Lactoferrin emerged as the star performer in another study, more recently published in PNAS (DOI: 10.1073/pnas.2105815118), which used U-M’s unique antiviral drug repurposing screening approach involving artificial intelligence (AI)-powered image analysis of human cell lines during infection with the novel coronavirus. More than 1,400 marketed drugs and compounds were tested in cells, either before or after viral infection, resulting in 17 hits with clinical translational potential.
Ten of those hits, including lactoferrin, were newly recognized. “Lactoferrin achieved the most complete efficacy of any of the drugs we tested because it works in several different ways,” says Sexton.
A glycoprotein found in secretory fluids, including breast milk, lactoferrin inhibits SARS-CoV-2 infection both by blocking viral entry into cells and modulating the host cell’s inflammatory response by suppressing the production of interleukin-6 (IL-6) that has been shown to trigger acute respiratory distress, he explains. Among the positives of lactoferrin are that it is widely commercially available and is a natural immune modulator, bridging innate and adaptive immunity in infants.
On the other hand, the pharmacokinetics (PK) of lactoferrin are “fairly poor,” Sexton adds. When ingested, lactoferrin exerts an antiviral effect in the upper gastrointestinal tract but how much of it survives the journey through the stomach depends on whether the host is fed or fasted.
PK studies in animals show that 70% of lactoferrin survives the stomach, if it’s empty, and is absorbed by the lymphatic system, he continues. The relevant players would by proteolytic products, the “clipped up pieces” of the protein that remain after digestion by the stomach, and it is difficult to anticipate the degree to which they would remain efficacious.
Agents that suppress IL-6, including the monoclonal antibody tocilizumab, are well studied in humans as a treatment for multiple diseases, says Sexton. With lactoferrin, the question is whether it sticks around long enough in human subjects to have the “pronounced efficacy” seen in cell cultures. “It’s really a complicated problem because lactoferrin is a protein and all the downstream products and their efficacy, half-lives and pharmacokinetics have not been established.”
The difference between people with asymptomatic SARS-CoV-2 infection and those who manifest symptoms up to the point of severe disease and death is not due to the virus but the host, which is another good reason to consider repurposing marketed medicines that target host factors, says Sexton. They can potentially modulate the physiology of people experiencing symptoms so they are less severe. “Our entire pharmacopoeia of human drugs that has been developed is a potential source of direct or indirect antivirals that can be very quickly deployed into clinical studies to see if they have efficacy.”
Compounds known as cationic amphiphilic drugs—phospholipidosis-causing substances that include the malaria drug hydroxychloroquine—are well known to achieve efficacy in cells in a dish, Sexton says. But they don’t generally work as antivirals in people because the concentration needed for efficacy would cause adverse events.
However, some drugs in this class are clearly worth exploring, he adds. He cites specifically clofazimine, one of the hits identified by the recent repurposing screening exercise, which is an old antibiotic used to treat leprosy. Clofazimine was also identified as a potential antiviral in a drug repurposing study by the Sanford Burnham Prebys Medical Discovery Institute (La Jolla, California) and a group in Hong Kong has demonstrated its efficacy in a hamster model.
Any compound that achieves in vivo efficacy warrants investigation, Sexton says, provided the safety profile is compatible with a COVID-19 patient. “We can pontificate about mechanisms of action but the only thing that really matters is if our discoveries in cells translates to a whole organism like a mouse or a hamster or a ferret.”
Leads generated by real-world evidence (e.g., analysis of efficacy data in electronic health records) are similarly worth pursuing, says Sexton. In the most recently published PNAS study, researchers observed that ipratropium bromide (Atrovent, commonly prescribed for asthma) had potent antiviral activity but albuterol (Ventolin, Proair, Proventil) did not and the two might be directly compared as potential COVID-19 treatments by looking at real-world signals of efficacy.
“As with all of our other drug repurposing hypotheses, what is most important is that we first stand up pilot clinical studies to demonstrate safety and tolerability in that patient population,” Sexton says. “I’m a big believer in targeting host factors, but we need direct antivirals as well. Ultimately, the winning strategy for viral diseases has been drug cocktails, to attack the infectious agent from many different angles… to achieve complete antiviral efficacy.”
The component drugs “really need to have an independent signal of efficacy before we start throwing them together in the clinic,” he adds. Many drug cocktails have been tested clinically with no solid evidence of efficacy as stand-alone products and, unsurprisingly, “a lot of those studies just don’t pan out and may cause harm.”
Taking a more “methodical approach” to identifying synergistic drug cocktails is the most promising path to eradicating SARS-CoV-2 and emergent strains, says Sexton. “Even with a highly vaccinated population, transmission is still possible, so we need prophylactic strategies as well as small molecule interventional treatments. There is a segment of the population that is not comfortable with vaccination; however, they will take small molecule drugs with no hesitation.”
Notably, the majority of the 17 hits from the recent antiviral drug repurposing exercise were generic drugs. Historically, it has been challenging to develop generics for new indications because of the perceived limited intellectual property that can limit commercialization, says Sexton but that is one of the areas where [taxpayer-supported] academic drug discovery can really pay off.”
Robust de novo drug discovery campaigns are still needed, he adds, because generics are not optimized for COVID-19. “They are a starting point and can serve as controls, and potentially mitigate some of the disease.”
Viral Life Cycle
Most of the antiviral drug repurposing screens that have been published to date have used a nonhuman cell line from African green monkeys, says Sexton, which is widely considered to be a good alternative host system for the cultivation of viruses but is not ideal for drug discovery purposes. By instead testing in several human cell systems, U-M researchers can look at clinically relevant effects beyond “protection against cell death” after viral inoculation.
For example, they can see which compounds inhibit the formation of syncytia—the large structures that result from the joining of cells. SARS-CoV-2 is an interstitial virus, spreading from cell to cell and eventually into a neighborhood of fused cells to create “one giant cell block that becomes the manufacturing site for new virus particles,” Sexton explains.
Even a drug that did nothing more than hinder this cell-to-cell fusion process may be a clinical success because syncytia formation is associated with lung tissue damage that is diagnosable with imaging in the clinic. “A compound protecting against syncytia formation it is not going to rescue all of the infected cells, but it is going to arrest the viral life cycle at a particular stage… [and] you just couldn’t discover that with another technology,” says Sexton.
The drug screening approach of the U-M research team uses machine vision and AI methods to observe individual infected cells in their natural progression to adjacent infected cells and fusion into syncytia, he says. When the cell population gets perturbed with a drug, they can see if the compound is arresting the viral life cycle at entry, at replication, or at syncytia formation. This screenings approach produces information “orders of magnitude” greater than conventional drug discovery approaches.
The work is one of the first major discoveries to come out of the new U-M Center for Drug Repurposing, “coincidentally” established in November 2019 under the leadership of the Michigan Institute for Clinical & Health Research to find potential therapeutics for the thousands of human diseases lacking any treatment—especially rare and orphan diseases being ignored by big pharma. Its founder is George A. Mashour, M.D., Ph.D., one of the co-authors on the PNAS study.
“We decided to be disease-agnostic and develop the infrastructure [including AI-based image analysis] so we could support investigators who are interested in drug discovery to first look for existing drugs,” says Sexton. Then COVID-19 came along, confronting scientists with the “absolute moonshot” of finding treatments for an unknown virus and the daunting task of setting up a biosafety level 3 biocontainment facility.
Sexton notes that their screen of the 1,425 marketed drugs and compounds had an initial 132 hits that got “triaged” to the 17 that were sufficiently potent and compatible with human dosing. Many of the others had efficacy but only at higher concentrations—including, most notably, hydroxychloroquine.
The overall hit rate was about 1%, which would be surprisingly high if de novo compounds were being screened, he says. The hit rate on compounds without any known bioactivity would typically be between 10 and 100 times lower.
U-M researchers will be initiating one of the planned phase 2 clinical studies with philanthropic support and the other with funding from Glanbia Nutritionals. “The compounds we are going forward with have been tested against numerous strains of the SARS-CoV-2 virus and shown to be quite effective,” says Sexton. “This is a strategy that is going to be particularly useful for emergent strains as well.”
In a newly submitted manuscript, the research team presents preliminary data supporting the effectiveness of lactoferrin against every variant of concern—including “a nearly 10-fold increase in efficacy in hyper-virulent strains like B117 and Delta” by blocking infection at the entry mechanism, according to Sexton. “This points to an interesting difference between the strains and their entry mechanism, which we are working on elucidating.”
Separately, the research team will be further investigating one of the surprise findings from their recent antiviral drug screen—MEK-inhibitors (typically prescribed to treat cancer) appear to exacerbate SARS-CoV-2 infection. “All drugs in the same class resulted in tripling of infection in this cell system,” says Sexton.
“When we’re putting two drugs together, we ideally want to see synergy… but very often what happens is that they actually worsen the disease state,” he continues. “We need to get that information out there so we can look for real-world evidence of that exacerbation before avoiding such drugs,” he continues.
“This may not translate to worsening in the human condition, but it is certainly something that needs to be investigated,” he says. “This may translate to other viral diseases.”
To come up with a cohort of patients large enough to be conclusive, the U-M team is tapping data derived from electronic health records of different institutions and harmonized by the National COVID Cohort Collaborative (N3C), shares Sexton. The clinical data lake contains information on millions of patients who have tested positive for COVID-19 and an even greater number of negative controls, which have proven invaluable in drug research against the virus.
“Our number one concern is the risk of putting a drug into a medically unstable patient, and we can address that situation simply by looking at the outcomes of patients who are coincidentally on the drug… relative to a matched cohort,” he says. “That can arm us with enough knowledge to de-risk the clinical studies we run and… perhaps detect signals of efficacy or exacerbation that can be examined further” or—perhaps more realistically—prompt the use or search for an alternative therapy.