Authors: Jane E. Zebrack (1Division of Allergy, Immunology, and Rheumatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; 2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California), Jaynelle Gao (1Division of Allergy, Immunology, and Rheumatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; 2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California), Britta Verhey (1Division of Allergy, Immunology, and Rheumatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; 2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California), Lu Tian (3Department of Biomedical Data Science, Stanford University School of Medicine, Stanford, California), Christopher Stave (4Lane Medical Library, Stanford University School of Medicine, Stanford, California), Bahare Farhadian (1Division of Allergy, Immunology, and Rheumatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; 2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California), Meiqian Ma (1Division of Allergy, Immunology, and Rheumatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; 2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California), Melissa Silverman (2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California; 5Division of Child and Adolescent Psychiatry and Child Development, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California), Yuhuan Xie (2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California; 5Division of Child and Adolescent Psychiatry and Child Development, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California), Paula Tran (2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California; 5Division of Child and Adolescent Psychiatry and Child Development, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California), Margo Thienemann (2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California; 5Division of Child and Adolescent Psychiatry and Child Development, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California), Jenny L. Wilson (6Division of Pediatric Neurology, Department of Pediatrics, Oregon Health & Science University, Portland), Jennifer Frankovich (1Division of Allergy, Immunology, and Rheumatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, California; 2Stanford PANS/Immune Behavioral Health Clinic and PANS Research Program at Lucile Packard Children’s Hospital, Stanford, California)
Categories: Original Investigation
Source: JAMA Network Open
Neurological Soft Signs at Presentation in Patients With Pediatric Acute-Onset Neuropsychiatric Syndrome
Authors: Jane E. Zebrack, Jaynelle Gao, Britta Verhey, Lu Tian, Christopher Stave, Bahare Farhadian, Meiqian Ma, Melissa Silverman, Yuhuan Xie, Paula Tran, Margo Thienemann, Jenny L. Wilson, Jennifer Frankovich
Abstract
Question
Do patients with pediatric acute-onset neuropsychiatric syndrome present with neurological soft signs associated with basal ganglia dysfunction?
Findings
In this cohort study of 119 patients with PANS, most patients presented with at least 1 neurological soft sign associated with the basal ganglia. The number of signs was associated with global impairment and the number of PANS symptoms, consistent with basal ganglia pathologic characteristics in PANS.
Meaning
This study suggests that targeted neurological examinations may help support the diagnosis of PANS.
Introduction
Pediatric acute-onset neuropsychiatric syndrome (PANS) and pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS) are abrupt-onset neuropsychiatric disorders thought to be triggered by infection.^1,2^ Basal ganglia inflammation may be an important mechanism in these disorders, based on imaging studies demonstrating basal ganglia swelling in the acute stage,^3^ microglia activation in the caudate and putamen,^4^ and microstructural changes that are most prominent in the basal ganglia.^5,6^ Sleep studies in this patient population indicating movements during rapid eye movement sleep^7,8,9^ implicate basal ganglia pathologic characteristics. Parkinson disease is a basal ganglia disorder in which movements in rapid eye movement sleep^10,11^ and the glabellar reflex^12^ can predate clinical onset of Parkinson disease.^13,14^ These findings have also been shown in PANS, suggesting PANS is a basal ganglia disorder.
Autoantibodies targeting cholinergic interneurons in the basal ganglia have been identified^15^ that could cause imbalance between excitatory and inhibitory synaptic transmission in the basal ganglia and may contribute to the cardinal symptoms of PANS. Animal models of PANDAS demonstrate an adaptive immune response involving autoantibodies and TH17 cells, leading to central nervous system pathologic characteristics, including neurovascular injury, blood-brain barrier disruption, activation of microglia, and loss of excitatory synaptic proteins.^16,17,18^
PANS classification criteria require abrupt onset or abrupt recurrence of obsessive-compulsive symptoms and/or eating restrictions with 2 or more additional new and abrupt-onset neuropsychiatric symptoms, which commonly include emotional lability, irritability, aggression, severely oppositional behaviors, behavioral regression and/or behavior outbursts, deterioration in school performance, sensory amplification, motor abnormalities, sleep disturbances, and urinary issues.^19,20,21^
Although “hard” neurological examination findings of basal ganglia dysfunction, such as chorea or dystonia, suggest a condition other than PANS, neurological soft signs (NSSs), such as voluntary movement overflow, have been described in attention-deficit/hyperactivity disorder (ADHD),^22,23,24,25,26,27,28,29,30,31,32^ obsessive-compulsive disorder (OCD),^22,23,33,34,35,36,37,38^ autism,^22,23,32,39,40,41,42,43^ and Sydenham chorea^44,45,46,47^ and may also indicate basal ganglia dysfunction. The prevalence of these findings in PANS is not known.
The goal of this study was to characterize the prevalence of NSSs that may be associated with basal ganglia dysfunction and examine the association between the number of NSSs and the impairment levels at clinical presentation of PANS. We hypothesized a high prevalence of NSSs among patients with PANS and that patients with more NSSs would exhibit more severe disease states.
Methods
Study Setting
This was a retrospective cohort study conducted at the Stanford Children’s Immune Behavioral Health (IBH) Clinic, where children and young adults with PANS and PANDAS are treated by multiple subspecialties, including child psychiatry, rheumatology, immunology, pediatrics, and psychology. Informed written consent from parents and adult participants and written assent from competent minor participants were obtained after the nature of the research was fully described and before any data were collected. Data were stored in a secure database. Approval was given by the Stanford University institutional review board. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies was followed.
Study Population
The IBH Clinic saw 135 new patients between November 1, 2014, and March 1, 2020, who consented to the research and received a diagnosis of PANS (eAppendix 1 in Supplement 1) from a child psychiatrist (M.S., Y.X., P.T., or M.T.). Neurological examinations were completed as part of the physical examination by clinicians in the IBH Clinic (B.F., M.M., and J.F.). We included neurological examination findings from patients’ initial presentation to the IBH Clinic. Occasionally, at the first visit, there was no comprehensive neurological examination conducted that included NSSs that may be associated with basal ganglia dysfunction; these abridged examinations occurred for various reasons, including time restraints or patient distress. In these cases, the neurological examination from the second or third visit was used for data collection. Patients were excluded if they had no comprehensive neurological examination that included these NSSs within the first 3 visits and 3 months after the first visit (n = 16). The final study sample included 119 patients. Most examinations included (93 of 119 [78.2%]) were from the patient’s first visit to the IBH Clinic.
Data Sources
Demographic characteristics, neurological examination findings, PANS symptoms, and impairment scale scores were abstracted from electronic medical records and prospective parent or caregiver questionnaires between December 13, 2020, and September 25, 2023. These data were maintained within our institutional review board–approved REDCap database. Race and ethnicity were self-reported. Categories included American Indian or Alaska Native, Asian, Black or African American, Jewish, Native Hawaiian or Other Pacific Islander, White, or unknown. Race and ethnicity were assessed to discern whether results are generalizable to a racial and/or ethnic population.
Measures and Variables
Our clinical protocol was established through training with a pediatric movement disorder specialist (Terence Sanger, MD) and trainings by Susan Swedo, MD, based on her observations of the National Institutes of Health (NIH) PANDAS cohort.^48,49^ The eTable in Supplement 1 indicates the procedures and findings for each neurological test. The NSSs evaluated included (1) glabellar tap reflex, (2) abnormal tongue movements, (3) milkmaid’s grip, (4) choreiform movements, (5) spooning, and (6) overflow movements. A patient could present with a minimum of zero or a maximum of 6 NSSs that may be associated with basal ganglia dysfunction.
Prior to each clinic visit, parents and patients completed an electronic questionnaire to report interim medical or psychiatric symptoms. Impairment scales included the Global Impairment Score (GIS), a validated measure of disease severity developed for PANS,^50^ and the Caregiver Burden Inventory (CBI), a measure of caregiver burden validated for patients with PANS.^51,52,53^ The GIS ranges from 1 to 100, with higher scores indicating greater impairment, and the CBI ranges from 0 to 96, with higher scores indicating greater caregiver burden.
Statistical Analysis
Statistical analysis was conducted between September 26, 2023, and November 22, 2024. Summary statistics were used to describe our cohort’s demographic and clinical characteristics. Continuous variables were summarized using mean (SD) values. Categorical variables were summarized using counts and proportions. Selected demographic and clinical characteristics were compared between patients with different prevalences of NSSs using 2-sample t tests. The following regression models were conducted, with models 1, 2, and 3 being Poisson regression and model 4 being linear Model controlling for sex and race and ethnicity, is age associated with the number of NSSs?Model controlling for sex, race and ethnicity, and age, is PANS duration associated with the number of NSSs?Model controlling for sex, race and ethnicity, and age, is the number of NSSs associated with the number of PANS symptoms?Model controlling for sex, race and ethnicity, and age, is the number of NSSs associated with the GIS? (The analysis was based on 107 patients after excluding 12 patients with missing GISs.)We conducted a sensitivity analysis by fitting the 4 models using the 71 patients who were examined for all 6 NSSs that may be associated with basal ganglia dysfunction. Model 4 sensitivity analysis was based on 67 patients after excluding 4 patients with missing GISs.
We evaluated confounders of some NSSs that could reveal an alternative explanation for the finding, such as age or comorbid conditions. We conducted 2 subset χ^2^ (1) whether or not overflow movements were associated with younger age and (2) whether or not spooning was associated with hypermobility.
All statistical tests were considered statistically significant if the 2-sided P < .05. We used Bonferroni correction to account for multiple testing (P < .025). SAS, University Edition (SAS Institute Inc), was used for analysis.
Results
The study cohort included 119 patients with PANS (Table 1). At presentation, patients’ ages ranged from 3.9 to 22.4 years (mean [SD] age, 10.4 [3.6] years). At PANS onset, the mean (SD) age was 8.2 (3.6) years (range, 1.6-18.6 years). Patients presented to the IBH Clinic a mean (SD) of 2.2 (2.7) years (range, 0.01-11.3 years) after initial PANS onset. Most patients (105 [88.2%]) were in a disease flare at the time of evaluation. The cohort included 66 boys (55.5%) and 53 girls (44.5%), 3 Asian patients (2.5%), 14 multiracial patients (11.8%), 88 non-Hispanic White patients (73.9%), and 13 patients of other race or ethnicity (10.9%). Obsessions, compulsions, anxiety, sleep disturbance, and irritability were the most common presenting PANS symptoms. Of 29 symptoms assessed via questionnaire, patients presented with a mean (SD) of 13.2 (4.3) symptoms.
Neurological soft signs that may be associated with basal ganglia dysfunction were common in our 71 patients (59.7%) presented with overflow movements, 62 (52.1%) with choreiform movements, 37 (31.1%) with abnormal tongue movements, 36 (30.2%) with milkmaid’s grip, 26 (21.8%) with glabellar tap reflex, and 21 (17.6%) with spooning (Table 2). These proportions were notably higher among the subset of patients who had a complete neurological examination, meaning they were examined for all 6 NSSs.
All 119 patients were examined for at least 1 NSS; 95 (79.8%) presented with at least 1 NSS (Table 3), and the mean (SD) number of NSSs was 2.1 (1.6). Among those examined for all 6 NSSs (71 [59.7%]), 91.5% (65 of 71) presented with at least 1 NSS, and the mean (SD) number of NSSs was 2.6 (1.4) (Figure).
Compared with the 24 patients with no NSSs, the 95 patients with 1 or more NSSs and the 26 patients with 4 or more NSSs had more PANS symptoms (0 vs ≥1 NSS: mean [SD], 11.5 [4.2] vs 13.6 [4.3] symptoms [P = .04]; 0 vs ≥4 NSSs: mean [SD], 11.5 [4.2] vs 15.1 [4.9] symptoms [P = .008]) and shorter PANS duration (0 vs ≥1 NSS: mean [SD], 3.2 [3.3] vs 1.9 [2.5] years [P = .03]; 0 vs ≥4 NSSs: mean [SD], 3.2 [3.3] vs 1.5 [2.5] years [P = .04]) (Table 3). Compared with patients with no NSSs who provided GISs (13 of 24), patients with 4 or more NSSs who provided GISs (22 of 26) reported worse global functioning of borderline significance (mean [SD] GIS, 56.0 [22.6] vs 40.6 [26.7]; P = .05). These results did not survive the Bonferroni correction (P = .003). There was no significant difference in age or CBI score.
On Poisson regression, neither age at first visit nor PANS duration was associated with the number of NSSs (models 1 and 2). Model 3 showed that the number of NSSs was significantly associated with the number of PANS symptoms, with 1 more sign increasing the number of symptoms by 5% on average (1.05; 95% CI, 1.02-1.08; P = .002; Bonferroni-adjusted P = .04). On linear regression, model 4 showed that the number of NSSs was significantly associated with GIS, with 1 more sign increasing the GIS by 2.86 points on average (2.86; 95% CI, 0.09-5.62; P = .04). This finding was not significant after adjusting for multiple testing (Bonferroni-adjusted P = .08).
In a sensitivity analysis using the 71 patients examined for all 6 signs, neither age at first visit nor PANS duration was associated with the number of NSSs that may be associated with basal ganglia dysfunction. The number of NSSs was significantly associated with the number of PANS symptoms, with 1 more sign increasing the number of symptoms by 6% on average (1.06; 95% CI, 1.02-1.10; P = .001; Bonferroni-adjusted P = .02). The number of NSSs was significantly associated with the GIS, with 1 more sign increasing the GIS by 5.19 points on average (5.19; 95% CI, 1.96-8.43; P = .002; Bonferroni-adjusted P = .04). Because patients can have up to 6 NSSs, the estimated association with symptoms and GIS per NSS is significant.
There was no significant association found between overflow movements and being 9 years of age or younger (patients aged ≤9 years, 31 of 42 [73.8%]; patients aged >9 years, 40 of 69 [58.0%]) (Table 2). There was also no significant association found between spooning and joint hypermobility.
Discussion
To our knowledge, this is the first study to describe the prevalence of NSSs which may be associated with basal ganglia dysfunction among patients with PANS at clinical presentation. We observed a high frequency of NSSs among patients with PANS who presented to the Stanford Children’s IBH Clinic, with more NSSs associated with more symptoms and greater impairment.
Previous studies have reported neurological findings among patients with PANS at the time of magnetic resonance imaging^6^ as well as during evaluation at the NIH.^49^ These studies are fairly consistent with our findings (Table 4).^6,27,28,29,30,31,32,36,37,38,40,41,42,43,44,45,46,47,49,54,55,56,57,58,59,60,61^ The slightly higher rates of NSSs in previous studies could be explained by selection bias; those cohorts are representative of more severe cases (only a small fraction of patients with PANS receive imaging or specialized care from the NIH program).
Neurological soft signs are subtle examination findings thought to indicate abnormalities in neurological circuit development.^62^ These are contrasted with “hard neurological signs,” such as chorea and dystonia, which are often associated with structural and/or genetic basal ganglia abnormalities.^63^ However, distinction on examination is not always clear; choreiform movements may be a mild form of chorea, and motor overflow may be a mild form of dystonia. This has clinical implications. For example, in a child with acquired chorea without genetic findings, Sydenham chorea is the probable diagnosis. However, if the chorea is more subtle (“choreiform”), the child is more likely to have PANS, assuming PANS criteria are met. Moreover, research classification and enrollment in trials are challenging in borderline cases.
Based on a PubMed search (eAppendix 2 and eReference in Supplement 1), NSSs that may be associated with basal ganglia dysfunction are variably seen in developmental and psychiatric conditions, including ADHD, OCD, autism, schizophrenia, and Sydenham chorea (Table 4).^6,27,28,29,30,31,32,36,37,38,40,41,42,43,44,45,46,47,49,54,55,56,57,58,59,60,61^ Neurological soft signs that may be associated with basal ganglia dysfunction are typically not observed among school-aged children, with the exception of overflow movements, which are normal among children younger than 6 years^64^ and gradually disappear at 7 to 14 years of age,^65^ with a significant decrease at 9 years of age.^55,59^ Younger age was not associated with more NSSs in our study, confirming that the high prevalence of NSSs is not associated with early developmental stage.
Similar to what is reported in the literature about typically developing youths,^55,59^ in our cohort, patients 9 years of age or younger had a higher frequency of overflow movements compared with patients older than 9 years (74% vs 58%). However, this difference was not statistically significant. Moreover, the rates of overflow movements in both the younger and older subset of patients are notably high compared with the expected rates among typically developing youths (Table 4), confirming the validity of evaluating this finding among patients with PANS, especially those older than 9 years.
Although there are more than 70 NSSs and more than 160 examination maneuvers,^60^ the maneuvers in this study were chosen because PANS is a suspected basal ganglia inflammatory disorder,^3,4,5,6,7,8,9^ and these maneuvers are more likely to be associated with basal ganglia dysfunction. However, the pathophysiology is likely more complex, and other circuits may be involved. For example, in the glabellar tap sign, the basal ganglia and interconnected cortical structures have a defined role in the timing and reciprocity of blinking. Thus, persistent blinking may reflect basal ganglia dysfunction and/or frontal lobe dysfunction.^54,61^ Inability to maintain tongue protrusion and milkmaid’s grip are both manifestations of the motor impersistence observed in chorea. Tongue deviation could be due to dystonia, but it can also be observed in tongue weakness or structural abnormalities.^56,57^ However, in our patients with tongue deviation (n = 10), there were no structural changes or tongue atrophy, and tongue deviation improved as the disease flare resolved. Overflow movements are involuntary posturing in body regions outside of the area of intentional movement and may be secondary to pathologic findings in the basal ganglia, cerebellum, and sensorimotor cortex.^58^ Given the preponderance of data implicating basal ganglia involvement in PANS,^3,4,5,6,7,8,9^ our proposed set of NSSs likely reflect abnormalities in circuits involving the basal ganglia. It is possible that preexisting basal ganglia dysfunction predated and predisposed to PANS. Thus, the association between NSSs and the clinical state should be explored. Neurological soft signs that may be associated with basal ganglia dysfunction resolving with patient recovery would contradict basal ganglia involvement being a static predisposing factor.
Some clinicians have theorized that spooning is a manifestation of joint hypermobility rather than dystonic posturing, given the high prevalence of hypermobility among patients with PANS.^66,67^ In our cohort, there was no significant association between spooning and joint hypermobility, suggesting it may be a relevant dystonic feature in PANS. Nonetheless, future studies should define whether spooning is associated with other NSSs associated with basal ganglia dysfunction. More broadly, the ability to distinguish joint or muscle symptoms from subtle neurological findings should be further explored in this patient population.
In the imaging study,^6^ patients with choreiform movements at the time of magnetic resonance imaging had more diffuse brain abnormalities. In synthesis with our study, patients with PANS who present with NSSs may have not only greater inflammation but also greater disease severity. Thus, NSSs may aid diagnosis and clinical monitoring of PANS, which currently has no reliable biomarker. Because 84% of youths with PANS undergo relapses and remissions,^68^ determining the clinical state is crucial to managing symptoms and treatment. Most patients in our study (105 [88.2%]) were in a disease flare during examination at presentation, so the study may not have been powered for assessing the association between NSSs and disease state. Further study is needed to understand the evolution of NSSs associated with basal ganglia dysfunction during the course of disease.
Some NSSs may help differentiate between developmental and psychiatric conditions. For example, choreiform movements are typically absent in youths with autism^42^ but are commonly present in youths with PANS,^6,49^ ADHD,^29^ OCD,^38^ and Sydenham chorea.^45^ Thus, if a patient with a diagnosis of autism exhibits choreiform movements, further diagnostic workup for other developmental and psychiatric conditions may be warranted. Because many NSSs observed in patients with PANS are also observed in patients with ADHD, OCD, autism, and Sydenham chorea (Table 4), the presence of a specific NSS may not distinguish PANS from these other conditions. However, these conditions could have unique NSS profiles that could aid in diagnosis. This possibility underscores the significance of completing and analyzing neurological examinations in these pediatric populations. Moreover, because NSSs are often general markers of psychopathologic findings, they may help determine patients’ neuropsychiatric status, regardless of diagnosis.
Limitations
There are some limitations to this study. Although the neurological examination was standardized, variability in methods and findings may have occurred between clinicians, and some data were missing. We were unable to conduct a comprehensive neurological examination including NSSs at every patient’s first visit, and examinations were sometimes performed during remission or resolution of a disease flare. Fourteen patients (11.8%) had their NSSs evaluated for the first time when they were not in a flare state. A challenge with this patient population is that patients who are doing poorly from a behavioral or mental health standpoint may not be able to cooperate with neurological examinations but may exhibit high levels of basal ganglia dysfunction. Thus, we may have underestimated the true prevalence of NSSs associated with basal ganglia dysfunction among patients with PANS. Last, the prevalence of NSSs among healthy youths and youths with OCD, ADHD, autism, and Sydenham chorea is not well established, so it is unclear if NSSs can distinguish PANS from other pediatric conditions. Future studies should use blinded reviews of examination videos to define the rates of NSSs associated with basal ganglia dysfunction among patients with PANS and other developmental and psychiatric conditions, as well as healthy youths.
Conclusions
In this cohort study, children and young adults presenting with PANS commonly had multiple NSSs associated with basal ganglia dysfunction on physical examination, with more NSSs assocated with greater impairment. Although diagnosis of PANS is based primarily on history, these findings suggest that targeted neurological examinations may support diagnosis. Furthermore, the high frequency of NSSs in PANS provides further evidence that pathogenic mechanisms of PANS may involve the basal ganglia, which could be the target of psychotherapeutic medications and rehabilitation. As these examinations are seldom routinely performed by neurologists or pediatricians, we advocate for their use in evaluating youths for immune-mediated neuropsychiatric deterioration.