Article In Brief
High quantities of inflammatory cytokines and markers of neurodegeneration were identified in the cerebrospinal fluid of cancer patients hospitalized for an extended period of time with moderate to severe neurological symptoms due to COVID-19. The paper shines light on why it has taken these patients so long to regain consciousness following sedation.
High levels of inflammatory cytokines and markers of neurodegeneration were found in the cerebrospinal fluid (CSF) of cancer patients hospitalized for over a month with moderate to severe neurological symptoms due to COVID-19.
The report, published February 13 in the journal Cancer Cell, offers possible insight into why hundreds of severely ill COVID-19 patients across the United States have taken weeks to months to regain consciousness after sedation is withdrawn.
The study was relatively small—including only 13 cancer patients with COVID-19 and three groups of controls who also had cancer from Memorial Sloan Kettering (MSK) Cancer Center— but it is the largest to date to identify neuroinflammation as the likely cause of severe neurological symptoms in COVID-19.
Strikingly, the MSK researchers found, the combination of symptoms and CSF findings strongly resembles the neurotoxicity sometimes seen in cancer patients treated with chimeric antigen receptor T cells (CAR-T), a condition known as immune effector cell-associated neurotoxicity syndrome (ICANS).
“That's why we measured the levels of all these inflammatory mediators in the spinal fluid and plasma,” said the senior author of the paper, Adrienne
Boire, MD, PhD, assistant attending neurologist and assistant member of the human oncology and pathogenesis program at MSK. “The deliriums were so reminiscent of what I sometimes see in CAR-T patients.”
Neurologists who specialize in critical care and neuro-infectious diseases said the findings align with what they have been seeing in their similarly ill patients with COVID-19.
“This is a really interesting and important paper,” said Serena S. Spudich, MD, the Gilbert H. Glaser Professor in the department of neurology and chief of neuroinfectious diseases and global neurology at Yale University's School of Medicine. “It contributes to the growing literature suggesting that neurologic presentations such as encephalopathy, delirium, and severe headaches in people with COVID-19 can be mediated by neuroinflammation.”
She and others emphasized, however, that the significance of the findings remains unknown for the large number of so-called COVID-19 long haulers: those with “brain fog” and other cognitive symptoms after developing COVID-19 that did not require hospitalization.
“This study only looked at patients who were moderately to severely ill and hospitalized with COVID-19 with significant neurologic symptoms, so it is unclear whether the findings can be extrapolated to other COVID-19 patients,” Dr. Spudich said. “But it is possible that this CNS inflammatory response is widespread at a subtle level, and that it manifests as prominent headaches and fatigue. The implication for those patients needs to be investigated.”
Seeking to build on prior published case reports of one to three patients, the MSK group began their study with 18 cancer patients hospitalized in the ICU for severe neurologic symptoms of COVID-19. They had either prolonged delirium (n=10), limbic encephalitis (n=four), refractory headaches (n=two), rhombencephalitis (n=one), or ischemic infarct (n=one).
CSF from 10 of the patients were analyzed at a median of 57 days from onset of respiratory symptoms associated with COVID-19, and 37 days from the onset of neurologic symptoms.
Because cancer is an inherently immunocompromised state, the COVID-19 group's results were compared to those from three cancer cohorts from the pre-pandemic era: one matched for age, cancer type, and presence of brain metastases; a second consisting of ICANS patients and a third with autoimmune encephalitis.
Cell count, protein, and glucose levels were all normal, and none of the patients had detectable levels of virus or viral proteins in their CSF.
Dr. Boire and colleagues found levels of 12 inflammatory mediators in the CSF of the COVID-19 patients to be nearly as high as those from patients with severe ICANS. The mediators included IL-6 and IL-8, interferon, and multiple types of chemokines. A protein involved in neurodegeneration, called matrix metalloproteinase-10 (MMP-10), was also seen at high levels, as was eukaryotic translation initiation factor 4E binding protein 1. As a group, the 12 mediators were significantly increased compared with the non-COVID-19 cancer controls (p=0.029).
The MSK group then compared levels of the 12 mediators to clinical scores of neurologic dysfunction at the time of lumbar punctures, as measured by the Karnofsky performance status (KPS) and disability rating scale (DRS). The levels of MMP-10, in both COVID-19- patients and controls, “correlated with the degree of neurologic insult by KPS and approached significance with the DRS,” the paper stated.
Additional proteomic analysis of CSF from the COVID-19 patients found an increase in the senescence marker 4E-BP1 and 10 other neuronal biomarkers, including neurofilament light chain protein (NfL).
“We find that cancer patients with neurologic sequelae of COVID-19 harbor leptomeningeal inflammatory cytokines in the absence of viral neuro-invasion,” the paper concluded. “The majority of these inflammatory mediators are driven by type II interferon and are known to induce neuronal injury in other disease states.”
Despite the findings, Dr. Boire told Neurology Today: “We can't determine causality. What we were doing was classic observational research, beginning with bedside examinations and then taking samples to the lab to see what we could learn. I can't say a particular cytokine is causative for the neurologic injury. But I can say we saw an accumulation of mediators of neurologic damage in patients with cognitive disturbances. And we didn't see those things in patients matched by age and cancer diagnosis in our control population.”
Given the study's finding of markers of neurologic inflammation, she said, “I do think that earlier and more liberal use of steroids in treating severely ill COVID-19 patients is probably going to lead to less frequent and perhaps less severe neurologic complications.”
A Fast-Growing Literature
The MSK paper was published amid a flurry of related research, some of it posted as preprints in medRxiv, examining CSF markers of inflammation and neurodegeneration in patients with severe neurologic sequelae of COVID-19.
Four days before the Cancer Cell paper was published online first on January 16, a Swedish group published results in Neurology of their tests on six such patients. They reported high CSF levels of neopterin (a marker of inflammation) and β2-microglobulin (sometimes associated with cancer). Median immunoglobulin G index, albumin ratio, and CSF white blood cell counts were normal in all six patients, but CSF NfL was elevated in two of them.
Another study of six COVID-19 patients with severe neurologic symptoms, posted as a preprint on December 9, also found increased levels of inflammatory markers in the CSF, including IL-12 and IL-12-associated innate and adaptive immune cell activation. The study, from a large group of researchers at Yale and other universities, also found what it called “increased proportions of B cells in the CSF relative to the periphery and evidence of clonal expansion of CSF B cells, suggesting a divergent intrathecal humoral response to SARS-CoV-2.”
Unique in the literature, the study also found that all six of the COVID-19 patients had anti-SARS-CoV-2 IgG antibodies in their CSF, as well as antibodies that recognized their own neural tissue.
Not all studies of hospitalized COVID-19 patients with severe neurologic symptoms have found high levels of cytokines or other signs of neuroinflammation. The title alone of a preprint posted on January 12 by the Hopkins Neuro-COVID-19 Group, for instance, indicates just how discordant its findings are from those by the MSK group: “Cerebrospinal fluid in COVID-19 neurological complications: no cytokine storm or neuroinflammation.”
That case-control study compared COVID-19 versus controls, not only to normal healthy patients but also to those with other viral infections (acute encephalitis and meningitis), autoimmune inflammation (neuromyelitis optica and autoimmune encephalitis), and also non-COVID-19 strokes. The design allowed the exclusion of pathology pathways that may also increase cytokines.
Carlos A. Pardo, MD, professor of neurology and pathology at Johns Hopkins University School of Medicine and senior author of that paper, offered a possible explanation for why the findings from his study differed from the study by Dr. Boire.
“Patients with cancer, previous metastatic lesions, and chemotherapy treatments have a different immunological setting,” he stated in an email to Neurology Today. “We are not able to make equivalent conclusions from that population to the general population with COVID-19.”
Dr. Boire, however, noted additional differences between their two studies. “They assayed six inflammatory cytokines, and we assayed 92 different inflammatory cytokines and 96 markers of neurotoxicity,” she said.
What's more, all the patients in her study were critically ill, compared to only eight of the 18 patients in Dr. Pardo's study, she said.
Even so, Dr. Pardo said he is not convinced that neuroinflammation or high levels of cytokines are the main driving factors causing the neurological complications in COVID-19. “It is extremely risky to recommend or state that such patients may get better with steroid use or even immunosuppressive treatments,” he told Neurology Today.
Michelle Monje, MD, PhD, an associate professor of neurology at Stanford University who is a neuro-oncologist, said, “The new paper is important and timely. In particular, the parallel between the COVID-19 patients' CNS cytokine profile and that of ICANS patients is a fascinating similarity.”
Ayush Batra, MD, assistant professor of neurology and pathology in the stroke and neurocritical care division at Northwestern University's Feinberg School of Medicine, said that despite the fact that Dr. Boire's study involved only cancer patients, it offered evidence that the neuroinflammatory markers seen in the COVID-19 patients were not due to cancer or its treatment.
“The findings will probably generalize to COVID-19 patients who do not have cancer but are otherwise critically ill,” said Dr. Batra, a co-founder of Northwestern's Neuro COVID-19 research group.
Dr. Batra said the paper “aligns well with some of the research we've done here. The patients we've tested haven't shown much evidence of direct viral invasion into the CNS. This paper is one of the first to show, in a granular manner, that inflammation is what's happening.”
Dr. Spudich, a coauthor of the Yale group's preprint study posted on December 9, said that while she agrees that Dr. Boire's three control groups make her findings likely generalizable, more research is necessary.
“We're talking about a small number of people,” Dr. Spudich said. “And there are many other factors in people who have a complicated disease like cancer that can cause brain inflammation.”
Allison Navis, MD, an assistant professor in the division of neuro-infectious diseases at the Icahn School of Medicine in New York, also expressed uncertainty as to whether the findings apply to COVID-19 patients who do not have cancer.
“These are already patients who are immunosuppressed,” she said. “So that limits the ability to which we can extrapolate the findings to those without cancer.”
Even so, Dr. Navis added, “It's an interesting and well-written article. It reinforces much of what we already understand about COVID-19.”
Drs. Boire, Navis, Spudich, and Batra had no disclosures. Dr. Monje receives a consulting fee for serving on the scientific advisory board of Cygnal Therapeutics.