Two teens developed severe psychiatric symptoms such as paranoia, delusions and suicidal thoughts during mild COVID-19 infections. Now, scientists think they’ve identified a potential trigger: Rogue antibodies may have mistakenly attacked the teens’ brains, rather than the coronavirus.  

The researchers spotted these rogue antibodies in two teens who were examined at the University of California, San Francisco (UCSF) Benioff Children’s Hospital after catching COVID-19 in 2020, according to a new report on the cases published Monday (Oct. 25) in the journal JAMA Neurology. The antibodies appeared in the patients’ cerebrospinal fluid (CSF), which is a clear liquid that flows in and around the hollow spaces of the brain and spinal cord. 

But while such antibodies may attack brain tissue, it’s too early to say that these antibodies directly caused the troubling symptoms in the teens, the researchers wrote in the new study. That’s because many of the identified antibodies appear to target structures located on the inside of cells, rather than on the outside, co-author Dr. Samuel Pleasure, a physician-scientist and professor of neurology at UCSF, told Live Science in an email.

Related: 20 of the worst epidemics and pandemics in history 

“So, we suspect that either the COVID autoantibodies” — meaning antibodies that attack the body rather than the virus — “are indicative of an out of control autoimmune response that might be driving the symptoms, without the antibodies necessarily causing the symptoms directly,” he said. Future studies will be needed to test this hypothesis, and to see whether any other, undiscovered autoantibodies target structures on the surface of cells and thus cause direct damage, he added.

The study’s results demonstrate that COVID-19 may trigger the development of brain-targeting autoantibodies, said Dr. Grace Gombolay, a pediatric neurologist at Children’s Healthcare of Atlanta and an assistant professor at Emory University School of Medicine, who wasn’t involved in the new study. And they also hint that, in some cases, treatments that “calm down” the immune system may help resolve psychiatric symptoms of COVID-19, she told Live Science in an email. 

Both teens in the study received intravenous immunoglobulin, a therapy used to essentially reset the immune response in autoimmune and inflammatory disorders, after which the teens’ psychiatric symptoms either partially or completely remitted. But it’s possible the patients would have “improved on their own, even without treatment,” and this study is too small to rule this out, Gombolay noted.

Possible mechanism found, but many questions remain  

Other viruses, such as herpes simplex virus, can sometimes drive the development of antibodies that attack brain cells, trigger harmful inflammation and cause neurological symptoms, Gombolay said. “Thus, it is reasonable to suspect that an association could also be seen in COVID-19.”

Prior to their research in teens, the study authors published evidence of neural autoantibodies in adult COVID-19 patients. According to a report published May 18 in the journal Cell Reports Medicine, these adult patients experienced seizures, loss of smell and hard-to-treat headaches, and most of them had also been hospitalized due to the respiratory symptoms of COVID-19.

But “in the case of these teens, the patients had quite minimal respiratory symptoms,” Pleasure said. This suggests that there’s a chance of such symptoms arising during or after cases of mild respiratory COVID-19, Pleasure said.

Over the course of five months in 2020, 18 children and teens were hospitalized at UCSF Benioff Children’s Hospital with confirmed COVID-19; the patients tested positive for the virus with either a PCR or rapid antigen test. From this group of pediatric patients, the study authors recruited three teens who underwent neurological evaluations and became the focus for the new case study. 

One patient had a history of unspecified anxiety and depression, and after catching COVID-19 they developed signs of delusion and paranoia. The second patient had a history of unspecified anxiety and motor tics, and following infection they experienced rapid mood shifts, aggression and suicidal thoughts; they also experienced “foggy brain,” impaired concentration and difficulty completing homework. The third patient, who had no known psychiatric history, was admitted after exhibiting repetitive behaviors, disordered eating, agitation and insomnia for several days, when they hadn’t shown these behaviors previously.

As part of their neurological examinations, each teen underwent a spinal tap, where a sample of CSF is drawn from the lower back. All three patients had elevated antibody levels in their CSF, but only the CSF of patients 1 and 2 carried antibodies against SARS-CoV-2, the virus that causes COVID-19. In those two teens, it’s possible the virus itself infiltrated their brains and spinal cords, the study authors noted. “I would suspect that if there is direct viral invasion it is transient, but there is still a lot of uncertainty here,” Pleasure noted. 

These same patients also carried neural autoantibodies in their CSF: In mice, the team found that these antibodies latched onto several areas of the brain, including the brain stem; the cerebellum, located at the very back of the brain; the cortex; and the olfactory bulb, which is involved in smell perception. 

The team then used lab-dish experiments to identify the targets the neural antibodies grabbed onto. The researchers flagged a number of potential targets and zoomed in on one in particular: a protein called transcription factor 4 (TCF4). Mutations in the gene for TCF4 can cause a rare neurological disorder called Pitt-Hopkins syndrome, and some studies hint that dysfunctional TCF4 may be involved in schizophrenia, according to a 2021 report in the journal Translational Psychiatry.

These findings hint that the autoantibodies might contribute to a runaway immune response that causes psychiatric symptoms in some COVID-19 patients, but again, the small study cannot prove that the antibodies themselves directly cause disease. It may be that other immune-related factors, apart from the antibodies, drive the emergence of these symptoms. 

“These autoantibodies may be most clinically meaningful as markers of immune dysregulation, but we haven’t found evidence that they are actually causing the patients’ symptoms. There’s certainly more work to be done in this area,” co-first author Dr. Christopher Bartley, an adjunct instructor in psychiatry at the UCSF Weill Institute for Neurosciences, said in a statement.

In future studies, “it would … be helpful to examine CSF of children with COVID-19 who did not have neuropsychiatric symptoms,” as a point of comparison to those who did, Gombolay said. “However, obtaining CSF from those patients is challenging as CSF has to be obtained by a spinal tap, and a spinal tap is not typically done unless a patient has neurological symptoms.”

That said, the team is now collaborating with several groups studying long COVID, who are collecting CSF samples from patients with and without neuropsychiatric symptoms, Pleasure said. “In adults, it is not uncommon to have patients be willing to undergo a spinal tap for research purposes with appropriate informed consent and institutional review.” Using these samples, as well as some studies in animal models, the team will work to pinpoint the autoimmune mechanisms behind these troubling neuropsychiatric symptoms, and figure out how autoantibodies fit into that picture. 

Originally published on Live Science.