What first led to your interest in PANS/PANDAS?

​When I opened my laboratory at the University of Oklahoma Health Sciences Center in Oklahoma City, I began by working on rheumatic fever because I thought it was actually a good example of autoimmunity caused by infection. I thought it would lead to a better understanding of how the immune system was affected by an infection attacking our bodies and causing disease. My PhD training was in the field of streptococcal immunology and infection.  I was working on a streptococcal vaccine in the 1970s with Dr Edwin Beachey and Dr Gene Stollerman who led a well-known streptococcal research group in Memphis, TN. After joining the faculty at the University of Oklahoma School of Medicine in the Department of Microbiology and Immunology, I initially decided to study autoimmunity and infection and the group A streptococcal sequelae. This led to the investigation of both heart and brain sequelae of streptococcal infections, which primarily affect children. Rheumatic fever peaks during the ages 5-15 years old which are the peak years of streptococcal infection in children.  When I opened my laboratory in 1980, medical researchers in immunology were looking for reasons for why autoimmune disease occurred and how it could be controlled. I wrote grants on rheumatic heart disease and discovered that cardiac myosin was the link between streptococci and heart disease and the NHLBI funded my laboratory with a career development award to study rheumatic heart disease caused by group A streptococci following a sore throat or pharyngitis. Rheumatic fever and heart disease is rare in children but occasionally children will get rheumatic fever which could lead to rheumatic heart disease. We began seeing patients who had Sydenham Chorea. Sydenham Chorea could be present with rheumatic heart disease. We collected blood from a hospitalized patient with Sydenham Chorea and began to produce monoclonal antibodies to study the disease pathogenesis. I was called by Dr Susan Swedo at the NIMH to study the streptococcal brain sequelae associated with infections. Dr. Swedo knew that I studied streptococcal sequelae and that she had identified a disease that she was interested in studying that was related to Sydenham Chorea.  Dr Swedo was investigating the effect of plasmapheresis on the outcome of Sydenham chorea. The monoclonal antibodies we produced from Sydenham chorea patients in Oklahoma were from B cells making the antibodies that attack the brain in Sydenham chorea. Because B cells produce the antibody in the Sydenham chorea, it actually directed us to the mechanism by which the antibodies work. Dr Swedo also found that plasmapheresis led to improvement of Sydenham chorea. Dr. Swedo wanted us to study her patients with pediatric autoimmune neuropsychiatric disorder associated with streptococci or PANDAS in addition to children with Sydenham Chorea. That is the story of how our research in PANDAS and PANS began. We published our first paper in Nature Medicine in 2003 on Sydenham Chorea and it was very well accepted because the mechanism had not been known previously which was that the antibodies signal neuronal cells to produce too much dopamine which affects movement and behaviors.

Is Sydenham Chorea a form of autoimmune encephalitis?

Yes.  Now we realize Sydenham chorea is a dopamine receptor and basal ganglia autoimmune encephalitis.  We actually put the human antibody gene derived from Sydenham chorea B cells into mice and the antibody expressed in mouse B cells targeted dopaminergic neurons in the basal ganglia in the brain of the mice.

Is dopamine always too high in kids who have PANS and PANDAS?

I don’t know that because we have not studied the dopamine neurotransmitter elevation in the brain in humans.  We do have an animal model where the antibodies were introduced directly into the brains through a catheter and the dopamine levels were elevated after injection of the antibodies into the brain.

So if Sydenham Chorea is a form of autoimmune encephalitis, then some autoimmune encephalitis can be treated just with antibiotics?

Antibiotics like penicillin kill bacteria in infections but they can have other effects—such as on the microbiome in the colon or the brain beyond the elimination of the infection. We have one animal model where we used antibiotics and the animals improved. Animals were immunized (not infected) with streptococcal antigens and the antibiotics had an effect either on the microbiome or they were anti-inflammatory or affected dopamine output. In that animal model the aberrant behaviors subsided and the dopamine levels went down when the antibiotics were given.  I do not know the mechanisms for this response to antibiotics in the absence of infection.  Sydenham Chorea is the brain manifestation of Rheumatic Fever.  If you were diagnosed with Sydenham Chorea, you would be treated with penicillin or azithromycin or other antibiotics until you were 21.  That is what the guidelines are for treatment of rheumatic fever, and rheumatic fever can be just a single symptom of Sydenham Chorea. It doesn’t have to be the valvular heart disease (heart murmur) or arthritis.

Do we know if Sydenham Chorea is a B cell or T cell mediated disease?

We would say it is B cell mediated disease because it’s the antibodies that get into the brain but it doesn’t mean that T cells don’t help those B cells produce the antibody or that T cells do not get into the brain and affect behaviors. Antibody is produced with T cell help usually. We see lots of IgG in serum with the same specificity of human monoclonal antibodies we produced from the patient.

Is PANDAS autoimmune encephalitis?

Both Sydenham chorea and PANDAS are considered dopamine receptor and /or basal ganglia autoimmune encephalitis which occurs after streptococcal infection. This conclusion is based on data collected on patients, animal models and specifically a transgenic mouse model where the antibodies in SC target the basal ganglia of the brain. The autoantibodies in SC and PANDAS are targeted against the D1 and D2 dopamine receptors. The autoantibodies in SC and PANDAS cause the neuronal cells to release too much dopamine which lead to the symptoms of PANDAS. Studies are in progress examining the mechanism of the autoantibody impact on the D1 dopamine receptor.  Thus, PANDAS would be a type of encephalitis, which means inflammation of the brain. It might not seem as severe as Sydenham Chorea because the chorea is an involuntary movement disorder which can be quite dramatic. In PANDAS, neuropsychiatric symptoms can be overwhelming for children and their families.  OCD and tics, characteristic of PANDAS, may be seen in Sydenham Chorea prior to the movement disorder onset.

Is the same true for PANS and PANDAS?

PANDAS is specifically caused by strep but in PANS, strep could be just one of many triggers. There are people who are resistant to treatment and may not have autoantibodies. Years of relapsing and remitting disease may or may not be associated with autoantibodies after a long time since the infections, and the neuronal circuits may be set in the brain and those may respond better to neuropsychiatric drugs. Dr Tanya Murphy has explained that the drugs should be initiated at very low doses and titrated up based on a child’s response. Regular doses of psychotropic drugs at the beginning of treatment may make the symptoms worse according to Dr Murphy. She has also had several successful antibiotic trials.

How many Cunningham panels have you completed?

Over 7000 at my laboratory and Moleculera Labs combined.  Moleculera Labs has been open about five years and has received more samples over time as the disease and the autoantibody panel have become more recognized.

Is the Cunningham Panel helpful for adults or is it only meant to be used with children?

We only have evidence of its use in children but that doesn’t mean it couldn’t be useful for adults. In fact we have many physicians that utilize the panel in adults.  We are in the process of establishing adult controls so all we can do is report what we find compared to pediatric controls. We have studies in progress of adult chorea and PANDAS (children) from a sampling at Mayo Clinic in Rochester, MN and also two studies of Lyme disease samples but those results are yet to be published.

Can you comment on the study funded by Autism Speaks in which you looked for the presence of anti-neuronal antibodies in children with autism?

Yes. First we are so grateful for the funding from Autism Speaks. The research into autism would not have been possible without the Autism Speaks Trailblazer awards twice, and we are in the process of finishing the paper to report what we found in the study. It clearly shows that ASD can be comorbid with PANDAS symptoms. If PANDAS is discovered in an ASD patient, it might potentially be treated the same way you normally treat PANDAS, with IVIG and/or antibiotics.  Clearly there are children with autism and overlapping PANDAS symptoms and the anti-neuronal autoantibodies are evident in certain patients with the PANDAS symptoms.

Can you comment on the Swedish study?

They found that the sensitivity of the Cunningham panel was approximately the same as our sensitivity if used as a panel all together.  The Swedish study identified 100% of their PANDAS/PANS children or adults with symptoms. At my laboratory, we found that the sensitivity was 91%. These sensitivity numbers are very close. However, they focused on the tests singly rather than as the panel, and they also found low specificity due to the control population that they compared with the PANDAS patients they enrolled in their study. Controls in their study were not assayed for streptococcal infections or other infections. My laboratory and Dr. Harvey Singer had already published 4 sets of controls demonstrating that some control sets could be elevated if not screened for pharyngitis or tonsillitis or other infections. Children or adults with pharyngitis or tonsillitis will demonstrate the anti-neuronal autoantibodies and you can cannot use these subjects as a healthy control for comparison. If you don’t screen for that in your control group then your results might display low sensitivity. Also, if adults are in the control group, then the control group may also be high and give a lower specificity. The interesting thing is the autoantibodies in pharyngitis go away quickly in a couple of weeks, but the ones in PANDAS and Sydenham Chorea do not go away and remain elevated. Dr Hilla Ben Pazi at Shaare Zedek Medical Center in Jerusalem and I are ready to publish this analysis in a new article. Our specificity for the test panel is approximately 70%.

A major issue with the Swedish study is that invalid blood collection tubes were unknowingly used for sampling, as was acknowledged in the Corrigendum to their article. We find that collection tubes with clot activators and other additives sporadically interfere within the assay and the results. This would lead to unreliable data and could explain the difference between all our previous work and their results.

Even if the specificity is low, you’re not ever going to treat someone who doesn’t have symptoms. The antibodies are not present in other diseases such as multiple sclerosis and the specificity of the autoantibodies for the ganglioside antigen is a different specificity than seen in the Guillain Barre Syndrome. It is important that you can identify nearly all children with the disease using the autoantibody test panel. So far we do not see false negatives with the autoantibody panel. If you are positive and have symptoms then it is conclusive that you have a diagnosis of PANDAS or Sydenham Chorea, or if you want to call it something else, you could say that it confirms the diagnosis of dopamine receptor or basal ganglia autoimmune encephalitis. In addition, autoantibodies present in OCD and tics are not surprising since inflammation has been reported in Tourette’s syndrome. The antibodies will remain high over time during the symptoms over months and then the titers decrease when the symptoms improve based on our studies of samples collected at Yale University and the NIMH.

Are you planning on adding any other anti-neuronal autoantibodies to the Cunningham panel?

Not at this time. Although we are researching additional markers and will be screening the samples we have to see if we can identify additional biomarkers that may assist in the future.  Possibly other ways of testing that might be genetic or related to genetics to determine other risk factors.

What percentage of children with positive Cunningham panels have had an abrupt onset vs. subacute onset and in tracking them over time, do both groups respond to similarly to immunomodulatory treatments?

Those PANDAS children who are positive for the autoantibodies against the dopamine receptor(s) and have a positive CaMKII or other antibody in the panel are likely to have a good response to immunomodulatory treatment. The autoantibodies against the dopamine receptors indicate an autoimmune dopamine receptor encephalitis while autoantibodies against lysoganglioside or tubulin suggest a basal ganglia encephalitis if they are not positive when tested against the dopamine receptors.

Do you have a sense for how many children with PANS might also have mitochondrial disorders?

No, it is probably more in ASD.

It seems IVIG is often failing or provides relief for a time but not lasting relief. Do you know why this is or what the next steps should be?

I do not really know the answer to that question. However, we find that the autoantibody panel may likely be an excellent predictor of a positive response to the IVIG. This is because autoantibodies provide a rational basis for immunotherapy and may suggest that immunotherapies would work. IVIG also protects children with borderline immunodeficiency against infections and is also thought to be anti-inflammatory.


~Thank you to Dr. Cunningham for taking the time to be interviewed by FCND President Anna Conkey.