Can you share in layman's terms the purpose of PET scans in children with PANS/PANDAS and what the PET scans show?
The clinical type of PET scan that people order is a PET scan of glucose metabolism and those are not useful in PANDAS and PANS. They are useful in epilepsy because they help us find the epileptic focus but it is not helpful at all in PANS or neuropsychiatric disorders. What I do is a different kind of PET scan, a research scan. PET allows us to look at many different chemicals in brain so there are some research chemicals that we use to image different functions of the brain.
We have a compound that images inflammation in the brain, which a clinical PET of glucose metabolism doesn't. So if a patient’s doctor ordered a PET scan, they would probably get a glucose PET scan and that will not yield useful info. In Detroit, where I spent over twenty years, we had a different compound called PK 11195 and I published a paper on the PANS, PANDAS, and Tourette’s population, but that compound was not ideal because of the background noise of that scan, which we call nonspecific binding. Therefore, you needed a strong signal in order to see results with that scan. We saw some abnormalities in basal ganglia in PK11195 PET scans from PANS, but we also saw some abnormalities in Tourette’s and they were not overlapping. The pattern was different between Tourette’s and PANS patients.
We now have a better compound called GE180. I’ve scanned a number of patients so far and am not finding very much, but those patients were stable and not highly symptomatic at the time. They were not in the acute phases. What I'm interested in is to get fresh new cases that just presented and scan them before any treatment. That hasn't been the case so far.
Is there a reason that you have chosen PET scans over functional MRI, or any other imaging technique? Are volumetric changes typically present on imaging in children with PANDAS or only evidence of microglial activation?
The volumetric MRIs can occasionally show you swollen basal ganglia. There is one paper on that. I think that is kind of nonspecific because there is variation in volume between individuals. The normal variation in basal ganglia volume is quite wide. To take one patient and measure the basal ganglia and compare it to a population—you’re probably not going to find anything unless it is a very, very dramatic case like the one that was published. Most kids will not be so dramatic. I think volumetric MRI is useful if you study a whole population of patients with PANDAS and compare them to a normal population. That might be helpful in understanding the disorder, but to take one person and try to get clinical information would not be easy. The same thing for fMRI. It just tells you how the brain is working and which areas of the brain are participating in certain tasks. It doesn’t get at the heart of the matter which is, is there inflammation or not? The biology of it is what we are targeting. We are asking if there inflammation or not, and neither fMRI nor volumetric MRI will give that information, so I think that PET of inflammation is by far the best.
There is a great deal of heterogeneity in PANS. What percentage of children with PANS who have had imaging have evidence of inflammation on PET scan? Are your findings in PANS identical to your findings in PANDAS?
Well we haven't come that far yet. We've just done a handful of cases. With more research, we should be able to answer those questions. I can tell you clinically that eight of ten kids who come to see me—and the parents are sometimes convinced their child has PANS—actually don't have it. But some clearly have PANS, and we have treated those with steroids, IVIG, plasmapheresis and they’ve gotten better. A significant number of the children I have seen already have some other underlying disorder and then they get an infection, and show a worsening and then the parents believe that it is PANS and that is totally wrong. I always give the example of epilepsy. When patients with epilepsy get an infection, the seizures typically get worse because infections lower the seizure threshold. That doesn’t make epilepsy PANDAS or PANS, even if the infection is strep. Epilepsy is common and strep is common, so if a child with epilepsy gets strep, the seizures get worse, but that does not make epilepsy a PANS disorder.
In children with worsening of neuropsychiatric symptoms following infection, are you saying that it is not always due to neuroinflammation?
Yes. In many it is not due to neuroinflammation. We have to carefully go through the story and evolution of symptoms. I spend some time in going through the history. Until we have a biomarker for PANS, we have to rely on a good clinical history. If I see a child who was normal and then suddenly they’re a different kid with OCD, anxiety, urinary incontinence, and other neurobehavioral features and maybe some movement abnormalities and that happens overnight, that is probably autoimmune. That is how autoimmune disorders present—very abruptly. But if a child was autistic, and I’ve had a number of those, and if that child has an infection and the autism gets worse, then some parents come in and say “this is PANDAS” and that just isn’t true. He’s autistic. PANS doesn’t give you autism. You’d be amazed how many people come in with that kind of a story and they get infuriated when I tell them I don’t believe their child has PANS. I’ve had a number of such encounters and it is kind of sad because I really want to help their child. This also happens with OCD, where often there is a strong family history of anxiety and OCD and the child who had some anxiety and/or OCD gets worse after an infection and they say, “Well, this is PANS or PANDAS.”
Have you imaged any children who have autism as well as PANS and if so, what were your findings?
No I have not imaged those kids because I didn't diagnose them as having PANS. I wouldn’t make that diagnosis so I wouldn't image those kids. PANS does not cause autism.
How sensitive and specific is PET as a biomarker for neuroinflammation?
We've done a bunch of cases with multiple sclerosis and other clearly document autoimmune disorders and we’ve seen very strong signals. We submitted an abstract on our preliminary experience based on seven or eight patients and the MS patients are striking. You see a very strong signal in the brain. I’m seeing subtle findings in some of the neurobehavioral cases but as I explained, those were random cases as we were testing our new compound (GE180). We haven’t begun a systematic study because we don’t have the funding for it yet.
At some point we will write a grant and we may scan some of Susan Swedo's patients from NIH. I would love to scan patients who have gone to Susan, and Susan is convinced they have PANS. She would have done the evaluation, and we could scan the patients here. We’re only two hours away from the NIH. After the scan, the kids could then go back to Susan for treatment of some sort, maybe plasmapheresis or IVIG, and then we could scan them again and compare the two images to see if there is a change. I would love to do that study. That would be a nice, tightly controlled study. That is what I am waiting for.
Is PET in PANS/PANDAS purely a research tool at the present time or is it being used as a diagnostic tool or tool to monitor disease activity right now?
The PET of inflammation is purely research now. PET is a clinical tool for glucose metabolism for epilepsy, tumors, and dementia, but that is a different kind of PET measuring glucose metabolism. There are over 300 chemicals you can image with PET, so when you say PET, you have to explain what kind of PET. Which chemical? Which metabolism are you looking at? In PANS, we want to look at inflammation in the brain.
Do you have thoughts on why some children respond so quickly to antibiotics and anti-inflammatories while others have a much more severe disease process that requires immunosuppression?
I don't really know. I think the antibiotics should be used to treat acute infection. I don't believe in using chronic antibiotics. If you have an infection, you treat it for two weeks and then you stop the antibiotics. The immunosuppression--some kids respond beautifully. I had a five-year-old girl and I was convinced she had PANS. I admitted her and gave five days of intravenous steroids and then discharged her. I saw them shortly thereafter and the parents were very grateful that I had given them back their daughter. This was now a normal child. She looked so good compared to the first time I saw her, I didn’t think I even needed to see her again. So that is rewarding. That child absolutely had an autoimmune condition that responded beautifully to steroids and we have had a number of similar cases like that as well. Others seem to have a more resistant course, but I cannot explain why.
If neuroinflammation is present in PANS as your PET studies show, what are your thoughts about glutamate excitotoxicity as a therapeutic target?
Well glutamate excitotoxicity is damaging to the brain. That is how areas of the brain are damaged, through glutamate mechanisms of excitotoxicity. I don’t think we have proven that the symptoms of PANS are due to glutamate excitotoxicity. I haven’t seen any paper like that so if you say I'm going to give a medication that is counteracting glutamate, that would be pure speculation.
There is nothing really good out there that counteracts glutamate. There is an anti-convulsant that counteracts one subset of glutamate receptors. Glutamate is complicated. There are four different receptors. There is no overall strong glutamate antagonist. People have been searching for that for a long, long, long time because of stroke. When you have a stroke, it brings in all this glutamate toxicity so people have been seeking drugs that can counteract that and they’ve not been very successful.
Long-term antibiotics are used in Sydenham Chorea. Long term antibiotics in PANS and PANDAS seems to control flares in some children. Do you have thoughts on why this might be effective for some?
I don't think long-term antibiotics are used in Sydenham Chorea. I think they’re treated acutely and then they’re stopped and if the Sydenham’s patient has choreiform movements, they use other medications. They use Zoloft for OCD. They use various dopaminergic drugs for chorea but they don’t use chronic antibiotics.
If there is an element of rheumatic heart disease, then yes, those kids would take chronic antibiotics because of the danger to the heart but not for pure Sydenham’s Chorea. They aren’t treated with antibiotics long term. Long-term antibiotics have a serious effect on your microbiome. It kills the bacteria and allows a fungal overgrowth. There is more and more data now that shows that the microbiome interacts with your total body including your brain and if you’re artificially slanted towards fungus and get rid of bacteria, I think that changes things. I don’t know how good it is for you. Chronic antibiotics also give rise to resistant strains that can be harmful.
Can you comment on any additional research you have planned or underway?
My next step is to really stop doing random cases and do a serious study with appropriate design. I am waiting for something to happen, hopefully through Susan [Swedo] because I trust her to make the diagnosis accurately and she has NIH resources. We could design it such that patients go through her to get their screening for PANS or PANDAS and then we scan the patient and they go back for treatment and then we scan them again and compare the two scans. If the binding pattern of GE180 is different between the two scans, then that is very valuable information. We can actually quantify the signal. We can measure the binding potential and compare those numbers before and after treatment.
FCND Chair Anna Conkey sincerely thanks Dr. Chugani, Division Chief of Neurology within the Nemours Neuroscience Center, for this interview.