Targets & Mechanisms
Long COVID’s big tent needs to get more precise
Breaking the disease into subtypes will be important for getting more companies into the space
Breaking the disease into subtypes will be important for getting more companies into the space.
The emerging consensus is that Long COVID should be addressed as a collection of distinct syndromes, even as advocates speak with one voice. Carving the disease into specific pathologies and biomarker-defined subsets will be important for bringing more drug developers into a disease space dominated by the public sector, patient groups and academia.
In the coming weeks, the Long COVID field will receive its first bolus of funding from NIH’s post-acute sequelae of SARS-CoV-2 infection (PASC) Initiative, supported by $1.15 billion in Congressional funding over four years. At an April 28 U.S. Congressional subcommittee hearing on Long COVID, NIH Director Francis Collins said the institute is evaluating 273 grant proposals, and will stand up core facilities and a large metacohort of patients for observational and interventional studies.
Some of the most centralized Long COVID initiatives have come from the World Health Organization; over the last eight months, it has established an ICD-10 code, clinical care guidelines and standardized case record form for the disease. A February meeting hosted by the organization and summarized in an April report was the first in a series aiming to define action priorities and present up-to-date information on the disease.
A key takeaway from both the Congressional hearing and the WHO meeting was that while inclusive case definitions will be critical for getting patients access to treatment, targeting the disease will require a divide-and-conquer approach.
“There are at least three or four pathways at play here, with different syndromes,” said Steven Deeks, a University of California San Francisco professor who testified at the Congressional hearing. “We’re going to have to slice and dice this, figure out the clinical phenotypes, figure out the mechanisms, and then figure out the therapies.”
The WHO ICD-10 code, which has been proposed for adoption by the U.S., reflects the move toward segmentation. Dubbing the disease “Post COVID-19 condition,” it instructs physicians to “code first the specific condition related to COVID-19 if known,” such as chronic respiratory failure, loss of smell or loss of taste, each of which have their own code.
Subdivisions within the disease are manifold. One major fault line is between patients who experience lingering symptoms after being hospitalized for COVID-19 versus those who experienced few or no acute symptoms; so far, most research has focused on the former.
Long COVID can also be split according to the organ systems affected, or alternatively the pathophysiology believed to drive disease; the handful of clinical trials testing therapies for Long COVID are divided along these axes.
Deeks believes it’s up to academics to de-risk the space by sorting out disease drivers. “We need the mechanisms, that’s what NIH will do. Once that happens, we will get industry involved,” he told the Congressional subcommittee.
Collins estimated researchers will have a better picture of the mechanisms driving Long COVID within six to nine months, and directions for targeting those mechanisms within a year.
But PureTech Health plc (LSE:PRTC; NASDAQ:PRTC) Head of Innovation Michael Chen thinks there are organ-specific aspects of Long COVID where the pathology is already clear enough for companies to bring their products to the table. PureTech is testing its idiopathic pulmonary fibrosis (IPF) candidate LYT-100 in a Phase II study of post-acute COVID-19 respiratory disease.
“It would be a little tragic to wait on the sidelines until all the biology is neatly worked out,” Chen told BioCentury.
PureTech portfolio company Akili Interactive Labs Inc., which develops digital therapeutics for neurological conditions, also is jumping in, having found alignment between its ADHD therapy and Long COVID brain fog, one of the most common and distressing symptoms reported by patients.
Akili CMO Anil Jina thinks drug developers have room to operate in emerging indications, despite shifting definitions. “I worked in the HIV space many years ago; you learn not to get worried about what things will be called,” he said.
Matching product to profile
Companies whose products have the potential to address obvious organ damage and functional deficits in Long COVID patients can start chipping away at those aspects of disease without dissecting their etiology.
For Akili, that means sidestepping the debate about whether the virus directly causes brain fog, or whether it emerges as a consequence of other stressors brought on by Long COVID. “We haven’t stressed too much about teasing apart whether it’s cause and effect,” Jina said.
“It would be a little tragic to wait on the sidelines until all the biology is neatly worked out.”
Akili is collaborating with Weill Cornell Medicine, New York-Presbyterian Hospital and Vanderbilt University to test AKL-T01, an adaptive video game designed to improve attention function, to treat patients with cognitive dysfunction following COVID-19.
The partners plans to conduct two randomized, controlled studies of AKL-T01, each with about 100 COVID-19 survivors who have exhibited a deficit in cognition, and measure changes in cognitive function and quality of life.
Jina said the company decided to move into the indication after learning about the attention deficits Weill Cornell physicians were seeing in Long COVID patients.
“When they showed us their work, we saw the aspects of cognition affected in these patients were the same aspects targeted by the mechanism of our main product, so we decided it makes sense for us to start something here.”
AKL-T01 is approved to treat children with ADHD, and has been studied in 12 different categories of disorders. The product has not been studied in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS), which shares many hallmarks with Long COVID, including neurological symptoms.
For PureTech, evidence from related coronaviruses supported its decision to enter the Long COVID space.
“A lot is known about the inflammatory and fibrotic component of Long COVID, specifically in the lung, which goes back to experience with SARS and MERS,” Chen said. “In our minds, that component of it is well-understood, and provides a rationale for us to intervene with LYT-100.”
Even within the organ-specific framework, the company is limiting its scope to a narrow subset of patients. Its Phase II study is recruiting patients who have been hospitalized for COVID-19, required oxygen support, and had pneumonia detected by imaging, with a minimum of two lung lobes involved.
“We’re focusing on patients we expect will have respiratory complications,” said Chen. He said the company might consider including other patient populations once it has data from its ongoing trial, and once more is known about other patient subtypes.
Divide and treat
Research to uncover and treat underlying drivers of Long COVID will be especially dependent on identifying which mechanisms are at play in which patients.
Even for patients whose disease is caused by the same overarching mechanism, for example cytokine-driven inflammation, “variable degrees of damage might be based on individual factors, such as genetics, prior exposures, and management of the acute infection,” said Hana Akselrod, co-director of the COVID-19 Recovery Clinic at The George Washington University. “If we solve that conundrum, we can advance more personalized therapies.”
Deeks thinks the demographics of Long COVID support the autoimmune hypothesis.
“Women in every cohort are more likely to get this than men,” he said. “This to me is the strongest hint we have in terms of the biology, because women in general are more susceptible to many autoimmune diseases.”
Potential therapeutic interventions for autoimmune-driven Long COVID include steroids and IVIG, said Collins.
Immune profiling methods, such as TCR sequencing technology from Adaptive Biotechnologies Corp. (NASDAQ:ADPT) and an autoantibody discovery platform from the lab of Yale University professor Aaron Ring, could identify the patients most likely to benefit. Ring and fellow Yale professor Akiko Iwasaki have launched a Long COVID immune profiling study together with David Putrino, a professor at Icahn School of Medicine at Mount Sinai.
Another prominent theory is that Long COVID is driven by persistent viral reservoirs in immune privileged sites, which could harm tissues both directly and indirectly via inflammation.
Viral reservoirs are not typical of coronaviruses, and persistent SARS-CoV-2 has been challenging to detect. But some evidence supporting the hypothesis is emerging; in a May 3 Science Translational Medicine study, an Institut Pasteur team showed patients with long-term loss of smell had viral transcripts, SARS-CoV-2-infected cells and inflammation in their olfactory mucosa.
A medRxiv preprint, along with anecdotal reports, indicate COVID vaccination improves symptoms in a portion of Long COVID patients. That could be because those patients are finally clearing lingering virus, but larger studies are needed to confirm the effect, determine which patients are most likely to benefit, and account for confounding factors.
“It seems to be happening more with people who haven’t been sick as long, which makes me think that symptom resolution would have happened anyway,” said Hannah Davis, co-founder of the Patient-Led Research Collaborative, which is conducting and publishing research on the Long COVID patient community.
Iwasaki and Ring are studying the effects of COVID vaccines in Long COVID patients, using immune profiling before and after vaccination to look for correlations between immune changes to symptom changes, and measuring mucosal immunity in saliva. The study is open to any adults who had COVID-19 and have experienced symptoms that interfere with normal activities for more than two months.
Untangling the complexity of what’s driving disease will require minimizing noise wherever possible.
One critical task is harmonizing the way patient data are reported, using the WHO case record form or another standard.
“Everyone who has a cohort or a clinic measures stuff differently, and they report stuff differently. As a consequence, the epidemiology is a mess,” Deeks said.
Another is addressing the stress faced by Long COVID patients as a result of their condition.
“You want to offer people standardized social work support and stress reduction methods, something to reduce the background noise," said Akselrod. “It’s not a traditional area for drug development, but the risk of not addressing the stress and psychosocial dimensions is a bias toward the null hypothesis.”
It will also be critical to account for biases created by unequal access to the Long COVID clinics springing up around the world, and avoid drawing broad conclusions from data focused on specific subtypes of post-COVID patients.
For example, Davis said, a TCR profiling dataset disclosed by Adaptive disproportionately featured hospitalized patients in their 60s and 70s, while the Long COVID patient populations her organization has surveyed are primarily younger people who were never hospitalized.
“One worry I have is people accidentally creating biased datasets, and making easy conclusions using machine learning that don’t apply to Long COVID,” she said.
Prevention for all
A form of intervention that’s less dependent on dissecting disease subtypes is preventing Long COVID by preventing or treating acute infections. Collins said part of NIH’s meta-cohort will include patients in trials for antiviral mAbs.
By clearing the virus more effectively and preventing “chaotic cytokine inflammatory storms,” antivirals and prophylactic vaccines could avoid long-term tissue damage, said Akselrod. “I think we’ll prevent a lot of Long COVID, just like we’re preventing a lot of critical acute COVID.”
NIH’s first foray into Long COVID interventions falls under the prevention umbrella.
The ACTIV-4 trial includes an arm testing the anticoagulant Eliquis apixaban in patients discharged from the hospital or convalescing; the endpoint is reduction in thrombotic complications within 45 days of being hospitalized for moderate and severe COVID-19.