Baseline characteristics of children in the International PANS Registry (IPR) Epidemiology Study

Introduction

Complex, neuropsychiatric conditions suspected of involving the immune system are not well understood, causing children to suffer from devastating, long-term psychiatric and neurological problems; academic, developmental and behavioural regression; and severe social and emotional disruption to patients and their families.1–5 Severe cases have led to in-patient hospitalisation and death.6 Paediatric acute-onset neuropsychiatric syndrome (PANS) is one such syndrome that is not well defined, but characterised by a sudden onset of obsessive–compulsive disorder (OCD) and/or severely restricted food intake and at least two other neuropsychiatric symptoms in previously healthy children.3–5 These may include anxiety, emotional lability, irritability/aggression, behavioural regression, deterioration in school performance, sensory and/or motor abnormalities, and somatic signs and symptoms.3 These symptoms tend to present in flare episodes. They are thought to follow a challenge to the immune system, triggered by exposure to infectious or other environmental factors and related to underlying immune dysregulation.7 Other names that have been used to classify patients with related conditions include: paediatric infection-triggered autoimmune neuropsychiatric disorder, paediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS),8–10 childhood acute neuropsychiatric symptoms, postinfectious or autoimmune basal ganglia encephalitis, autoimmune OCD and children’s postinfectious autoimmune encephalopathy.

PANS is a neuropsychiatric condition that may involve underlying neuroimmune pathophysiology.4 11 PANS is likely an ‘umbrella’ term for a broad, heterogenous syndrome comprised of definable subgroups that relate to causal factors, mechanistic pathways, disease course and treatment response.12 13 One known subgroup of PANS is related to a group strep A infection specifically: PANDAS.2 8 14 PANS-related conditions typically follow a relapsing-remitting clinical course with flares in psychiatric and physical symptoms typically lasting 3–4 months.15 16 Some patients develop a chronic illness course that requires ongoing treatment.17 Case management often involves long and multiple courses of treatment that target the psychiatric and physical symptoms, the source of the inflammation and the disturbance of the immune system.18 Although some patients’ symptoms go into remission, there is no established cure for PANS.

Causal factors for PANS are suspected to include bacterial and viral infections and other factors, such as genetics, environmental factors and metabolic disorders.11 The underlying biological mechanism linking infection and neuropsychiatric symptoms is hypothesised to be an autoimmune reaction, in which an abnormal immune response and/or dysregulation in the immune system causes damage to the brain tissue due to cross-reactive antibodies.19 PANS is observed in siblings and family clusters, suggesting genetic predisposition and environmental triggers may play important roles.20–23 Early genetic evidence further supports an association with PANS.24 25 Due to vague specification of the condition and a void of population-based data, the prevalence of these PANS-related neuroimmune conditions is not currently known.

A void of standardized, large-scale data on the broad spectrum of children and families affected by PANS and PANS-like features has hindered advancement of scientific knowledge relating to the underlying heterogeneity, pathophysiology, biomarker discovery, therapeutic targets for novel drug discovery, and ultimately, disease prevention.26 To date, one large-scale survey has been completed (n=698), but it did not include longitudinal follow-up and it restricted inclusion to those who had reportedly been diagnosed with PANS or PANDAS.15 Several small clinical study samples from specialty treatment centres in the USA and Sweden have also been described according to their phenotypic characteristics and clinical presentation, but these focus on patients who meet the current restrictive working criteria, requiring patients have an ‘abrupt, dramatic onset’ of symptoms.5 8 14 16 27 28

These gaps in knowledge systematically result in adverse repercussions for patients and their families, including undetected disease, mis or delayed diagnosis, inappropriate and/or delayed treatment and contribute to existing controversy around these conditions. Individuals from families with low socioeconomic position are likely particularly disadvantaged in this regard given care-seeking at specialty PANS clinics often requires families to cross state lines and pay medical costs out of pocket. The International PANS Registry (IPR) was created to facilitate and accelerate multidisciplinary and translational health research focused on this patient population, including but not limited to immunology, rheumatology, microbiology, psychiatry, neurology and genetics. The embedded IPR Epidemiology Study is an ongoing, registry-based, longitudinal, observational study that is, to our knowledge, the largest study sample of children with PANS and PANS-like features and their healthy siblings. In this paper, we describe the cross-sectional, baseline data of the first 1781 study participants.

Cohort description

Study design and protocol

The IPR Epidemiology Study is an ongoing, registry-based, longitudinal, observational study by which we will investigate the epidemiology of the broad spectrum of PANS-related conditions. The IPR rationale and study protocol is described elsewhere (manuscript under review). To date, the IPR Epidemiology Study includes survey data based on parent report for suspected PANS cases and their siblings. We designed the IPR Epidemiology Study to enable ongoing contact with participants and to facilitate ancillary studies and integration of new data elements in the future. Study participants will be invited to participate in periodic follow-up surveys to track longitudinal observations on all cases and siblings; siblings will be asked to report development of any new PANS-like features (online supplemental figure 1).

Supplemental material

Participants and eligibility criteria

Children, adolescents and adults in the USA and Canada suspected to have/had PANS or PANS-like features and their siblings who are not currently affected by these conditions are welcome to enrol in the IPR. We recruited patients by convenient sampling directing interested parties to the publicly accessible link on the IPR’s website, using Facebook, mailing lists (including emails and newsletters), partnerships with other non-profit organisations, clinical referral networks, local community events, and international conferences.

The IPR Epidemiology Study participants are a subsample of the IPR Recruitment Pool. Inclusion criteria for the IPR Epidemiology Study is completion of the following baseline surveys: household introduction and demographic surveys and the individual eligibility, demographics, maternal health, prenatal/perinatal/postnatal periods and comorbidities surveys; cases are additionally required to complete the diagnosis, symptoms and onset surveys. Individuals (cases and siblings) may be included in the IPR Epidemiology Study even if their household members are not. Households are designated as ‘complete household inclusion’ when all individual members enrolled in the IPR from one household have met the Epidemiology Study inclusion criteria.

The data described here were collected using the Dacima Software platform.

Patient and public involvement

The IPR and embedded Epidemiology Study are based in a collaborative community-based research partnership between the two authors, an epidemiologist (EM) and the director of the IPR sponsor (JG), the Pediatric Research and Advocacy Initiative (PRAI). PRAI, a non-profit organisation supported by the parent/patient community raised funds from families affected by PANS and related conditions to fund the establishment of the first registry for families affected by PANS and related conditions. Many of the survey question responses we describe in this manuscript were originally developed via parent discussions and interactions on social media and inquired about via informal polling in online parent support groups. Through these platforms, thousands of families from around the world were able to compare notes on common clinical reports, histories and presentations they observed in their children who they suspect have PANS. JG led the initial compilation of these questions and topics and EM guided their conversion into a formal set of surveys. This is the first study that addresses their collective desire for these topics to be further investigated and more deeply considered in the scientific and clinical communities.

Statistical analyses

In this paper, we describe the IPR Epidemiology Study subsample, including their individual and household demographics; neurologic, psychiatric and behavioural symptoms; symptom onset; phenotypic and other characteristics; phenotypic traits and other characteristics; comorbid conditions; treatment and medication experience; family health history; disease impact and experience; and interest in sharing existing data and making additional data contributions. We use descriptive statistics, including means and SD to describe continuous measures and frequencies and proportions to describe categorical measures. We stratify descriptions by cases and siblings. Analyses were conducted in SAS V.9.4 (SAS Institute, Cary, North Carolina, USA).

Findings to date

IPR Epidemiology Study participant profile

Data collection cut-off for this report was December 2021. At this point in time, 1781 participants met the inclusion criteria for the IPR Epidemiology Study. The IPR Epidemiology Study includes 1010 households, 1179 cases and 602 siblings. Nearly all adult respondents were the primary caregiver of the child(ren) they enrolled in the study (99% of cases, 98% of siblings) and primarily biological mothers (94% of cases and 93% of siblings). At the time of enrolment, cases were older than the enrolled siblings. Most participants were under 18 years of age (94%), though adults whose parent has power of attorney are also included (table 1, online supplemental figure 2). Most (81%) of the cases report that a physician has told them their child has PANS and/or PANDAS.

Table 1

Child participant demographics (N=1781)

Case and sibling participants are mostly male (57% and 53%, respectively), from singleton births (94% and 93%, respectively), white (95% of cases and siblings) and privately insured (80% of cases) (table 1). The median level of education was fifth grade for cases and fourth grade for siblings, and median income for the year prior to the child’s birth was US$85 000–US$100 000 (for cases and siblings).

To date, the IPR Epidemiology Study includes 1010 participating households. For most of these households (83%), all of their children enrolled in the IPR met inclusion criteria for the Epidemiology Study subsample. Most households (83%) include 1 or 2 children and some (17%) include at least 3 children (online supplemental table 1). Some of the households in the IPR Epidemiology Study included a sibling (39%) and some included more than one case (20%). Most household survey respondents were partnered (married, remarried or living as married; 88%), employed (60%) and owned or are buying their homes (87%) (online supplemental table 2). The median reported annual household income for the year prior to IPR enrolment was US$125 000–US$150 000. For most households, the highest level of maternal and paternal educations was bachelor’s degrees (37% and 35%, respectively) or master’s degrees or professional degrees beyond a bachelor’s (35% and 26%, respectively).

Neurologic, psychiatric and behavioural symptoms

Overall, the most common symptoms reported among IPR Epidemiology Study participant cases were anxiety (94%), emotional lability (92%) and obsessions (90%), followed by somatic signs and symptoms (84%), irritability, aggression, severe oppositional behaviour (83%) and sensory abnormalities (81%) (table 2). For nearly all symptoms, 70%–80% of participants reportedly experienced an onset of that particular symptom that felt ‘abrupt and dramatic’. Nearly all participants (90%) were affected by these symptoms before puberty. When asked if these symptoms can be explained by a known neurologic or medical disorder (such as Sydenham chorea, systemic lupus erythematosus, Tourette disorder or others), very few (n=39, 3%) replied affirmatively.

Table 2

Neurologic, psychiatric and behavioural symptoms by parent report, N=1179

Symptom onset

Overall, slightly over half of cases reportedly had a ‘sudden and dramatic’ nature of onset (58%) and just less than half (42%) report ‘a gradual progression’ nature of their symptom onset. Among those who reported a ‘sudden and dramatic’ onset of symptoms, 54% also reported they noticed early signs or symptoms in hindsight (table 3). Among those who generally described their child’s symptom onset as ‘a gradual progression’, 88% reported a dramatic increase in symptoms occurred at some point in time. ‘A gradual progression’ nature of onset or early signs/symptoms noted in hindsight begin at 4 years of age on average; a ‘sudden and dramatic’ nature of onset or a subsequent dramatic increase in symptoms occurs around 7 years of age on average. Most of those who reported a ‘sudden and dramatic’ nature of onset quantify the onset timing as within 1 week (62%), whereas only 26% of those who report a dramatic increase in symptoms following ‘a gradual progression’ nature of onset estimate the time over which the dramatic increase occurs as within 1 week. Most of those who report the nature of onset as ‘sudden and dramatic’ or ‘a gradual progression’ recalled a time when their child was symptom free (69% and 59%, respectively).

Table 3

Nature of the onset of PANS feature by parent report, n=1179

We observed three main patterns in the parent reports of initial symptom onset: (1) 27% reported sudden onset with no report of any early signs and symptoms in hindsight, (2) 68% reported a (a) gradual onset with an eventual dramatic increase in symptoms (37%) or (b) sudden onset with reported early signs and symptoms in hindsight (31%) or (3) 5% reported gradual onset without ever experiencing a dramatic increase in symptoms. These patterns will be the focus of subsequent analyses of extant IPR Epidemiology Study data, including evaluating their potential association with phenotypic traits, maternal health and other disease characteristics.

Suspected triggers

The leading suspected triggers for the initial onset of PANS symptoms and subsequent flare episodes were similar (online supplemental table 3). The most common suspected trigger of flare episodes is infection or illness (84%); approximately half of parents also point to anxiety or stress (50%), environmental allergies (40%) and being around someone who is strep positive (40%). Among those who suspected the initial onset of PANS symptoms was triggered by infection or illness (71%), 61% suspected the infection was strep throat.

Phenotypic traits and other characteristics

Table 4 describes various phenotypic traits and other characteristics across cases and siblings. Many of these, to our knowledge, have not been previously collected for this patient population. Among cases, some of these are reportedly more intense during a symptom flare.

Table 4

Phenotypic traits and other characteristics of cases and siblings by parent report, n=1781

Health conditions

Overall, 18% of cases and 7% of siblings reportedly have an autoimmune or autoinflammatory condition other than PANS/PANDAS (online supplemental table 4). Approximately one-quarter (27%) of cases and 12% of siblings have at least one food allergy diagnosed through formal testing. Among them, the most common food allergies were lactose intolerance/cow’s milk allergy (57% cases, 46% siblings), gluten intolerance/wheat allergy (55% cases, 44% siblings) and eggs (32% cases, 29% siblings). Most participants had received the most common vaccines (online supplemental table 5). Approximately one-third of cases (36%) and siblings (30%) have been vaccinated according to a modified vaccine schedule.

Family health history

Among those with known and reported health of the biological mother, 39% of participants’ biological mothers have an autoimmune, autoinflammatory or other immune-mediated condition. Most of the case participants (90%, n=1060) reported a known health history for their nuclear, biological family, including that of the participant’s mother, father and siblings (online supplemental table 6). The most common reported conditions in the nuclear, biological family were allergies (66%), mental illness (56%) and autoimmune conditions (48%). Several participants had a brother (29%) and/or sister (25%) with PANS and some had a mother (15%) or father (9%) with a suspected PANS diagnosis. On both the maternal and paternal sides of the family, PANS/PANDAS were believed to be observed in cousins (8%–10% and 4%–7%, respectively), aunts/uncles (7%–8% and 3%–5%, respectively) and grandparents (4%–5% and 2%–3%, respectively). Reports may be more common on the maternal side.

Treatment and medication experience

1094 of the IPR Epidemiology Study sample had completed the ‘treatment’ survey at the time of data analysis. The treatments that have most commonly ever been tried in the study sample include antibiotics (88%), supplements (81%) and oral non-steroidal anti-inflammatory drugs (NSAIDs) (79%) (table 5). The treatments that were most often reported as ever having at least partially alleviated symptoms in this study sample were the same: antibiotics (74%), NSAIDs (62%) and supplements (52%). ‘Tincture of time’/waiting it out was the approach most reported as having made symptoms worse (42%) and the only approach where worsening of symptoms was more commonly reported than alleviation.

Table 5

Treatment experience overview, n=1094 cases

Disease impact, experience and healthcare utilisation

Parents consistently reported a high degree of disease-related disruption in their child’s lives (9.0±1.5 on a 10-point scale) and a high degree of severity (7.9±1.9 on a 10-point scale) (online supplemental table 7). Approximately a quarter of cases had been hospitalised for disease-related reasons and 60% had their ability to attend school compromised in some way due to PANS (39% had stopped attending school for at least a month at some point). Most cases were receiving special accommodations in school, mostly individualised education programmes, 504 plans, home-bound home schooling and help from a resource teacher.

Most participants reported having difficulty locating (84%) and accessing (87%) appropriate care and treatment (online supplemental table 8). The primary barriers to access they reported were a shortage of specialty providers and treatment services where they live as well as cost and or a lack of insurance coverage. Most participants (64%) waited more than a month for a new patient appointment with the first specialist they saw with their child. Overall, patients had seen an average of five providers (SD: 5.2) seeking treatment and management of PANS symptoms.

Participants reported that, on average, less than half (46%, SD: 37%) of their PANS-related care and treatment costs in the past year were covered by insurance and/or other healthcare coverage. Overall, participants spent an estimated US$5447 (SD: US$9698) out of pocket on PANS-related care in the prior year and report an estimated US$11 177 (SD: US$22 665) as the most spent out of pocket on PANS-related care in any 1 year.

Interest in sharing existing data and making additional data contributions

Most participating households (83%) provided permission for their deidentified data to be shared with any third party that the PRAI, the IPR sponsor, recommends. Another 12% were more selective, but still provided permission for their deidentified data to be shared with specific third parties. Nearly all of them (93%) gave permission for their data to be shared specifically with physicians and scientists/researchers. Overall, 94% of the IPR Epidemiology Study households are interested in sharing their data with physicians and scientists/researchers.

Additionally, 84% of all IPR Epidemiology participants are interested in sharing biological samples and/or specimens from their child. Most would be willing to consider sharing any sample or specimen type, but participants seemed most receptive to sharing saliva, hair, fingernails and/or urine. Similarly, overall, 75% of participants reported a willingness to share a complete set of their child(ren)’s deidentified medical records, though only about 17% had already obtained the records and were ready to share them at the time of the survey. Case and sibling participants were similar in their willingness to share specimens and medical records.

In summary, this paper provides an overview of the baseline data we collected from participants in the IPR Epidemiology Study between July 2020 and December 2021. Enrolment is open and ongoing and, to this end, the IPR invites families affected by these conditions to enrol and dramatically expand our sample size (www.pansregistry.org). The IPR provides researchers with detailed phenotypic information from thousands of individuals and families affected by PANS and related conditions to accelerate clinical research and identify the causes of these conditions. We invite external research collaborators to use the IPR database for analyses and ancillary study recruitment. This cohort provides a previously absent opportunity to conduct analyses that address unresolved research questions, including further investigation of the possible involvement of immune dysregulation in the disease processes affecting their enrolled child(ren).

Strengths and limitations

This study is unique because the participant sample is selected from the IPR, the first large-scale, centralised epidemiologic registry in the world designed to accelerate research on the broad spectrum of PANS. The IPR Epidemiology Study was designed to be dynamic and flexible, to longitudinally follow participants, facilitate ancillary study recruitment based on detailed selection criteria and expand the scope of the database (eg, integrate medical record information, biospecimens). The Study has also been set up to enrol and prospectively track healthy siblings who may develop PANS symptoms in the future. It is important to note that the study investigators recognise that siblings of self-reported PANS cases should not be considered study controls. We will keep detailed documentation of the data collected over time and monitor our success at tracking participants over time, comparing their baseline characteristics to those we lose to follow-up.

Participation in the IPR Epidemiology Study is not restricted to those who meet the current working criteria for PANS.11 While current recommendations for evaluating suspected PANS cases include a comprehensive medical and psychiatric history, a thorough physical exam, and an overall evaluation by a treating physician,3 this approach limits the potential for conducting large-scale studies as it is challenging to pool large sample sizes and to standardise and centralise this type of data. While many study participants report PANS-like features but likely do not meet the full set of current working criteria for the condition, the strength of the IPR Epidemiology Study is that it includes thousands of participants with related conditions. Furthermore, the IPR Epidemiology Study enables study of the broad spectrum of PANS-like conditions and application of more restrictive selection criteria within the study sample to meet specific research selection criteria. It is important to note that the survey data we have presented here is based on parental self-report and subject to recall bias, particularly when reporting clinical information that has not been recorded in writing and/or recalling events that occurred in the distant past. To date, the IPR Epidemiology Study includes a convenient sample of participants that is not generalisable nor representative of any region, demographic, nor specific patient population. Furthermore, the IPR Epidemiology study sample is subject to selection bias related to our recruitment approach and the status of PANS within the US medical system. It is possible that those included represent, on average, parents who are the most motivated to further solidify PANS as a diagnosis and to accelerate research that will lead to better treatment for their children.

Although we asked participants to report any codiagnosis with neurological, developmental and other conditions for cases and siblings, the response field was open text and we do not believe we collected consistent and complete replies. In the first follow-up survey, we have included a multiple-choice list of codiagnoses as well as a list of known infections that preceded a flare episode (for cases only) to better ascertain these characteristics and experiences.

Future plans include increasing sample size, collecting longitudinal survey data, recruiting appropriate study controls and expanding the scope of the database, prioritising medical record data integration and creating a linked biobank. Secondary analyses of extant data will prioritise identifying subgroups by phenotypic traits, maternal health and characteristics of disease onset.

This post was originally published on https://bmjopen.bmj.com