Eight Cases of Pediatric Acute-Onset Neuropsychiatric Syndrome with Inflammatory Bowel Disease: Immunologic Intersections

Pediatric acute-onset neuropsychiatric syndrome (PANS) and pediatric autoimmune neuropsychiatric disorder associated with streptococcal infections (PANDAS) are immune-mediated diseases characterized by abrupt onset neurobehavioral changes. Inflammatory bowel disease (IBD) includes ulcerative colitis (UC) and Crohn’s disease (CD), chronic conditions characterized by gastrointestinal (GI) inflammation. Examining a series of children with both PANS and IBD may permit a better understanding of both illnesses and application of management strategies from one field to the other.

While understanding of PANS pathogenesis is developing, IBD is felt to arise from a dysregulated immune reaction to gut microbes in genetically predisposed individuals [1]. Inflammation may involve other organ systems; extraintestinal manifestations most commonly impact the musculoskeletal system (including arthritis and enthesitis), skin, and eye. IBD has known associations with other immune-mediated conditions, such as psoriasis (PsO) and sclerosing cholangitis. Neuropsychiatric disorders, such as fatigue, depression, and anxiety, are common comorbid conditions in IBD; rarer associations include bipolar disorder and schizophrenia [2, 3]. IBD and multiple sclerosis share a bidirectional risk association; peripheral neuropathy is also increased in IBD [4]. Based on a population-based Korean database study, IBD patients may also have an increased risk of Parkinson’s disease, a basal ganglia disorder [5, 6]. More than 1/3 of IBD patients are affected by an extraintestinal manifestation, which can predate IBD in 24% [7].

PANDAS and PANS are less well understood neuropsychiatric illnesses starting in childhood [8, 9]. Central features are abrupt onset of obsessive-compulsive disorder (OCD) and/or restricted eating, frequently occurring with severe anxiety, behavioral regression, school performance deterioration, mood disturbance, disordered sleep, urinary changes (polyuria or enuresis), and adventitious movements. Onset often follows an infection, most commonly group A streptococcus (GAS), and most cases have a relapsing-remitting course [10]. PANDAS was first defined in 1998 as pre-pubertal and GAS associated [8]. Recognition that many patients present after the window to evaluate for GAS and that other triggers are possible led to the broader diagnostic criteria for PANS in 2015 [9]. For the purposes of this article, the term PANS will be used to describe both.

Basal ganglia dysfunction appears to be central to PANS symptomatology. Evidence includes four imaging studies of PANS/PANDAS demonstrating changes in basal ganglia volume and diffusivity on MRI and microglial activation on PET, compared to controls [11‒14]. Additionally, patients with PANS have a high rate of movements and/or loss of atonia during REM sleep, an uncommon finding in pediatrics but a known predictor of Parkinson’s disease in adults [15, 16].

Immune dysregulation is implicated in basal ganglia abnormalities in PANS. Sera from children with PANS bind striatal cholinergic interneurons in mouse and human brain, altering their activity, a finding that improves in the recovered state [17‒19]. Repeated exposures to GAS in mice leads to migration of GAS-specific Th17 cells from nasal tissue into the brain, disruption of the blood-brain barrier, microglial activation, and brain deposition of IgG, providing further clues to the immune basis of PANS in humans [20]. Approximately one-third of children who undergo lumbar puncture for PANS have non-specific elevations of cerebrospinal fluid protein or albumin quotient, soft markers of neuroinflammation [21].

As in IBD, genetic predisposition for autoimmunity and inflammation appears to be important in PANS. Seventy-one percent of the first 47 PANS patients at the Stanford Immune Behavioral Health (IBH) Clinic had first-degree relatives with autoimmune or inflammatory disorders [22]. In a large survey of 490 children with parent-reported PANS, 85% had a family history of autoimmune conditions, including 29% of mothers and 65% of second-degree relatives. Hashimoto’s and PsO were the most reported autoimmune disorders in mothers (11% and 5.7%, respectively) [23].

Increasingly, PANS appears to occur in the setting of systemic inflammatory dysregulation, with immune-mediated comorbidities outside the neurologic system. In a Swedish cohort of 45 patients, 24% had a preexisting autoimmune or inflammatory disorder [24]. Of 193 consecutive children with PANS followed longitudinally at the Stanford clinic, by the age of 14, 28.3% developed arthritis and 7.5% developed another autoimmune disease, most commonly thyroiditis (4.1%) and PsO (2.6%) [25, 26]. Arthritis was typically “dry,” with novel features on musculoskeletal imaging including joint capsular thickening in 55% and distal interphalangeal joint tenderness in 81.8%, findings described in adult psoriatic arthritis [25, 26]. Intriguingly, onset of guttate PsO, like PANS, has also been associated with GAS infections [27], and a recent population study of over 200,000 American adults suggests PsO is associated with OCD, when compared to matched controls (2.11-fold increased odds, 1.6–2.7 95% CI, p < 0.001) [28].

In PANS, a child’s psychiatric symptoms arise so severely and suddenly that they may overshadow somatic complaints. However, closer interrogation often hints at systemic inflammation. For example, at Stanford IBH clinic, 58% of PANS cases have associated fatigue and 47% have unrefreshing sleep [22]. GI symptoms occur in 42% [22].

IBD has established treatment protocols that focus on immune modulation [29‒31]. In recent years, there has been an explosion of advanced therapies for IBD, but only infliximab and adalimumab, both anti-TNFα agents, are approved for use in children by the US Food and Drug Administration (FDA) currently. Children tend to have more extensive intestinal involvement and aggressive IBD courses than adults, with potential impact on growth [32]. In refractory cases, off-label use of other biologic therapies, such as anti-integrin or anti IL-12/23 agents, may be required to achieve remission.

For PANS, a three-pronged treatment guideline was published in 2017 [33‒36]. According to this protocol, suspected infections should be addressed, paying special attention to eradication of GAS. Second, inflammation should be managed with medication such as nonsteroidal anti-inflammatory drugs (NSAIDs), corticosteroids, or in severe or refractory cases, intravenous immunoglobulin (IVIg). Rarely, more aggressive immune modulators are used for persistent disease when there are clear markers of autoimmunity/inflammation and response to corticosteroids; however, the PANS guideline does not address many treatments commonly used in IBD. Some antibiotics have recognized anti-inflammatory properties [37]; azithromycin may owe its effectiveness in reducing PANS-related OCD to this effect [38]. Third, rehabilitation and support of the healing brain are key to management and include cognitive behavioral therapy and careful titration of medications like selective serotonin reuptake inhibitors (SSRIs), second generation antipsychotics, and alpha-2 agonists.

Since 2012, Stanford’s Immune Behavioral Health (IBH) Clinic has followed hundreds of PANS patients into young adulthood and is uniquely positioned to recognize long term outcomes. Eight patients at IBH have now received IBD diagnoses. Separately, the pediatric IBD Clinic at the F. Widjaja IBD Institute at Cedars-Sinai Medical Center noted that 3 of their patients also have PANS, and Dartmouth’s Neuroimmune Psychiatric Disorders (NIPD) Clinic has 1 patient with both diagnoses. Children who have both PANS and IBD provide an opportunity to examine the intersection of these illnesses.

To our knowledge, only two published cases examine this pairing of diagnoses [39, 40]. By examining a larger group through a multidisciplinary lens, we hope to shed light on features that put a child at risk for both diseases and identify potential management strategies.

The Stanford Immune Behavioral Health (IBH)/PANS Clinic (Palo Alto, CA) is a tertiary academic program that receives referrals of children and young adults with abrupt neuropsychiatric deterioration. Patients are screened for clinic entry by review of medical records, discussion with referring physicians, and patient/parent questionnaires. Longitudinal multidisciplinary care includes psychiatry, psychology, rheumatology, immunology, and general pediatrics. IBD evaluation and care is facilitated by board-certified pediatric gastroenterologists affiliated with Stanford University.

Cedars-Sinai Pediatric IBD Clinic (Los Angeles, CA) is a tertiary academic center that receives referrals of children with suspected IBD from the southern California region and beyond. Children in this clinic with PANS received their neurobehavioral care outside the Cedars-Sinai system.

The Dartmouth Neuroimmune Psychiatric Disorders (NIPD) Clinic is a tertiary academic clinic that provides longitudinal care and consultations for patients with neuropsychiatric symptoms after infection or in association with autoimmune diseases. Children with IBD are evaluated and treated within the broader Children’s Hospital at Dartmouth/Dartmouth Health system in northern New England.

To be included in this series, patients were required to have IBD diagnosed by a gastroenterologist, as well as PANS diagnosed by a treating physician from an appropriate specialty (for instance, pediatrics, neurology, rheumatology, allergy/immunology, and psychiatry).

Standardized clinical information on disease presentation, diagnosis, and treatment, in addition to family and past medical history, was gathered from review of the electronic medical record, which included information on patient/parent-reported symptoms, clinician impression, and standardized survey instruments. Date of PANS onset was based on the first episode of abrupt neuropsychiatric deterioration meeting PANS criteria [9]. Date of IBD diagnosis was used, rather than date of first IBD symptoms, due to the high rate of chronic GI symptoms and obscuring psychiatric symptoms in the cohort. This study was approved by the Institutional Review Boards at Stanford University, Dartmouth Health Children’s, and Cedars-Sinai Medical Centers.

A total of 12 patients met inclusion criteria (8 from Stanford, 3 from Cedars-Sinai, and 1 from Dartmouth). However, three Stanford cases were excluded from analysis because gastroenterology records were not available to confirm IBD diagnoses. An additional Stanford case who responded to biologic therapy for a combination of PANS, arthritis, and suspected early perianal CD was excluded due to diagnostic uncertainty regarding IBD. Of the 8 remaining cases, 3 were diagnosed with UC and 5 with CD.

Demographic information is shown in Table 1. Five of 8 cases were male (63%); the majority were White. There were 3 cases of UC, numbered UC-1 through UC-3, and 5 cases of CD, similarly numbered. Mean age at PANS onset was 9.3 years (range 3–16) and mean age at IBD diagnosis was 15.6 (range 8–23). PANS preceded IBD in 7 cases, by a mean of 8.4 years (range 1–14 years), and UC preceded PANS in 1 case by 8 years. The mean age of the patients at the time of this report is 19.8 years (range 12–27 years). Patients were followed by the authors for an average of 11.3 years since PANS onset or IBD diagnosis, whichever came first.

Medical comorbidities in each subject and their first or second-degree relatives are summarized in Table 2. Comorbid inflammatory conditions were common in this patient cohort: eczema (n = 6), asthma (n = 5), arthralgia or inflammatory arthritis (n = 5), low immunoglobulin levels (n = 3), and chronic rhinosinusitis (n = 2). Pre-PANS comorbid neurodevelopmental disorders were also common: attention deficit/hyperactivity disorder (ADHD) (n = 2) and autism spectrum disorder (ASD) (n = 1). The patient diagnosed with ASD at age 8 (CD-5) was no longer felt to have ASD once his neuropsychiatric symptoms improved following treatment of IBD. 88% (7/8) of patients had a first-degree relative with an immune-mediated disease: PsO or psoriatic arthritis (n = 5, 71%), idiopathic thrombocytopenia purpura (n = 1), celiac disease (n = 1), rheumatoid arthritis (n = 1), and Hashimoto’s thyroiditis (n = 1). Several had a history of IBD in more distant relatives, including a grandfather and an aunt with UC and a great-grandfather with CD.

Characteristics of PANS symptoms, onset, and course are detailed in Table 3. Antecedent infections to PANS (and therefore possible triggers) included GAS in 2 cases and a mixture of other predominantly upper respiratory illnesses in the remainder. In CD-4, onset of neurobehavioral symptoms occurred after a 2 day hospital stay for a sepsis-like illness with sore throat in which throat culture recovered Streptococcus dysgalactiae. Throat cultures or PCR were performed in 6 cases, but most occurred weeks to years after symptom onset and returned negative. Only UC-2 had a positive throat culture for GAS. GAS rapid antigen testing was unrevealing, often performed beyond the window for detecting GAS. UC-1 had a mild elevation of anti-streptolysin O antibodies (ASO) drawn 2 weeks after onset (212 IU/mL, with normal range = 16–200), peaking at 252 a month later and then normalizing, consistent with clinical suspicion of a GAS trigger in this case. UC-2 had an elevated anti-DNase B titer of 189 U/mL (normal 0–120) drawn at PANS diagnosis during a relapse, 2 years after initial symptom onset. CD-3 and CD-4 did not have GAS serology testing done. In the remaining cases, anti-streptococcal antibodies were within normal ranges. ESR and CRP levels were within normal limits for all subjects except CD-4, who had elevations to 45 mm/h and 256 mg/L, respectively, during infection with S. dysgalactiae, with normalized CRP and improved ESR to 25 mm/h afterward.

All subjects had normal blood counts and metabolic panels at PANS diagnosis, with the following exceptions: CD-3 and CD-5 were mildly anemic; CD-4 had 20% monocytes on WBC differential, UC-1 and UC-2 had 9% eosinophilia and borderline hyperglycemia of 127 and 101 mg/dL, respectively. Of the five subjects who had anti-nuclear antibody (ANA) drawn, two (UC-1 and CD-3) had titers of 1:80; the remainder were negative. UC-1 had IgA levels below normal, UC-3 had low IgG and IgM, and CD-3 had low IgG/IgG1. Other testing is outlined in Table 3.

Subjects received a variety of treatments (Table 3), including treatments to address infections, anti-inflammatories, and immunomodulators for post-infectious inflammation (NSAIDS, corticosteroids, and IVIg), disease modifying anti-rheumatic drugs (DMARDS) for arthritis (methotrexate and sulfasalazine), and neuropsychiatric support (i.e., SSRIs, etc.). At the time of this report, most patients are in PANS recovery, with functioning at or near pre-morbid baseline. UC-2 is still experiencing mild symptoms, and CD-3 continues to improve dramatically, with current mild to moderate symptoms.

IBD clinical presentations are outlined in Table 4, including Montreal classifications showing a range of disease location, severity, and behavior. All three UC cases were prescribed biologics except for UC-3, who is managed with oral and topical 5-ASA due to reluctance to start biologics. Of the 5 CD cases, three are maintained on biologics. CD-3 improved on methotrexate, sulfasalazine, and IVIg for PANS and is in the process of obtaining insurance approval for upadacitinib to address chronic arthritis and IBD. None of the cases have had Clostridium difficile or cytomegalovirus infection or have been hospitalized for severe IBD. Joint pain and arthritis occurred in 5 cases (63%) and 2 patients (CD-1 and CD-2) have had orthopedic hip surgeries.

Since PANS diagnosis preceded IBD in 7 of 8 cases, we explored whether PANS improved after IBD was successfully treated. It was difficult to determine if there was a relationship between IBD treatment and PANS improvement in CD-3 because treatment for both largely overlapped in time. In UC-2, IBD is still active, so the impact of IBD control on PANS cannot yet be determined. This leaves 5 cases for which the effect of IBD treatment on PANS can be examined. In all 5 cases, once IBD treatment was started, there were minor fluctuations of PANS symptoms but no major relapses. For CD-5, there was a very clear improvement of PANS symptoms after each dose of infliximab, given every 6–8 weeks for IBD. His OCD and sleep normalized with each treatment and worsened again 2–3 weeks prior to his next scheduled dose. UC-1 is the only subject who had just one PANS flare; as this flare was resolving, he was diagnosed with IBD, raising the possibility that the successful management of colitis contributed to his more benign neuropsychiatric course. In the single case where IBD preceded PANS (CD-4), GI symptoms remained stable as PANS was treated.

We provide full narrative case descriptions in online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000543969) for 1 case each of UC and CD (UC-1 and CD-3). We also describe an additional case with PANS, enthesitis-related arthritis, and signs of early perianal CD that resolved with early treatment with sulfasalazine and adalimumab but was excluded from our series due to lesser diagnostic certainty for IBD. Together, these cases illustrate the complexity of clinical practice, where PANS and IBD presentations evolve over time, impacting treatment decisions.

We describe 8 individuals diagnosed with both PANS and IBD. We found (1) an extremely high rate of immune-mediated illness, especially PsO, in first-degree relatives, (2) frequent comorbidity with arthralgias or arthritis, and (3) a tendency for PANS to predate IBD development and to improve with treatment of IBD.

The occurrence of UC and CD in the setting of PANS lends further support to an immune mechanism for PANS, which unfortunately still lacks an easily available biomarker and relies on clinical diagnosis. We feel that the co-existence of IBD and PANS diagnoses in our cohort cannot be attributed to coincidence alone. In 2016, the prevalence of IBD in the pediatric population was 77.0/100,000 [41]; thus, we would expect 0 or 1 child in the Stanford IBH clinic to have concomitant IBD to date, rather than 5.

In large surveys, the mean age at PANS onset is 7.5 years [42]. IBD, on the other hand, has a bi-modal distribution, with the major peak at 15–29 years [43], and a second group of 10–15% diagnosed at 60 years or older [44]. Since the typical onset of IBD is older than PANS, subjects would be expected to develop PANS prior to IBD. In fact, PANS preceded IBD in our series by a mean of 6.4 years (7 of 8 cases had PANS a mean of 8.4 years before IBD, and 1 case had IBD precede PANS by 8 years).

Most of our patients (88%) had a history of asthma and/or eczema. This is strikingly higher than previously described in a survey of 69 patients with PANS, where 19% had atopic dermatitis, 16% asthma, 48% allergic rhinitis, 16% IgE-mediated food allergy, and 4.3% celiac disease, with only allergic rhinitis more prevalent than in the general population [45].

Patients in our case series also had a high rate of arthralgias or clinically apparent arthritis during their course (5 of 8 patients, 63%). This rate is even higher than the 28% prevalence of radiographically confirmed arthritis in the general PANS population at Stanford IBH clinic [26]. Ma et al. [26] found that most patients with PANS and arthritis had the forms of arthritis connected with IBD (enthesitis-related arthritis and/or spondyloarthritis). In the general IBD population, arthralgia without arthritis occurs in 5–16%, while symptomatic arthritis/enthesitis occurs in 10–20% of CD and 4–14% of UC [7]. Up to 50% of CD patients may have asymptomatic arthritis [7], and arthritis in patients with IBD has been associated with increased concomitant autoimmune and inflammatory conditions [46]. Together, these findings suggest that arthritis could be a phenotypic marker of IBD risk in PANS.

Additionally, 7 of 8 patients had a first-degree relative with an immune-mediated disorder, including 5 with PsO or psoriatic arthritis, much higher than previously described in the general PANS population. A previous study of the Stanford PANS cohort found immune-mediated comorbidities in 14% of siblings and 36% of parents, with PsO in 1.2% and 7.4%, respectively [47]. Given the striking rate of PsO in close family members in our series (63%), it is possible that this subset of PANS shares immune mechanisms with PsO and IBD.

Many children with PANS have GI symptoms and weight loss (42% and 63%, respectively) [22], especially during PANS flares. While it is not practical to perform endoscopy on every child with PANS and GI symptoms, our data suggest those with arthritis plus a family history of PsO may be at higher risk of IBD and require long term monitoring. In these cases, clinicians should have a lower threshold to check inflammatory markers such as fecal calprotectin and to refer to gastroenterology for further evaluation of GI inflammation when indicated.

Early identification of bowel inflammation is worthwhile as our experience suggests that PANS course may improve when comorbid IBD is treated. In all 5 cases where PANS predated IBD sufficiently to permit an examination of clinical course, neuropsychiatric symptoms improved or stabilized after IBD treatment was established. Identifying IBD opens the door to immunological treatments, including biologics, that are inaccessible to most children with PANS and that may also be helpful in improving neuropsychiatric outcomes. Research is needed to determine if these immunologic therapies, which target TNFα, IL-12/23, and other pathways, might improve PANS symptoms even in the absence of IBD.

The increased prevalence of IBD in the PANS population suggests shared mechanisms of disease. We, therefore, examine the IBD knowledge base for information that may shed light on PANS.

PANS and IBD disease similarities include an association with antecedent infection. Infections implicated as triggers in PANS include group A streptococcus (most common), influenza, SARS-CoV-2, and mycoplasma [9, 48, 49]. IBD onset has likewise been associated with infections, especially GI infections with Campylobacter spp., Salmonella spp., C. difficile, and norovirus [50, 51]. Less well understood is the role of non-GI infections as potential triggers for IBD. One recent trial investigated a potential role for group A streptococcus as a trigger in 91 patients with CD divided into new onset, in remission, active disease, and controls. While no patients had positive throat cultures or recognized GAS infection in the past year, anti-streptolysin O (ASO) titers were elevated in half of new onset CD cases, a significant difference compared to controls (OR 4.41 in multivariate analysis, 95% CI, 1.35–14.37, p = 0.014) [52].

Both IBD and PANS patients have an altered gut microbiome [53]. In a study of 30 children with PANS, ASO titers elevations >500 IU/mL within 5 months of microbiome testing were significantly negatively correlated with Dehalobacterium, Corynebacterium, Gemella, and Lactobacillus and positively correlated with Odoribacter [54]. However, untangling cause and effect in this relationship is challenging. Management of PANS frequently includes antibiotics to treat and prophylax against GAS; therefore, disruption of the microbiome would be expected. The common observation that PANS improves transiently with antibiotics, even in the absence of demonstrable infection, suggests a direction for future research in these antibiotic-responsive cases. Perhaps “anti-inflammatory antibiotics” such as azithromycin [37] exert short-term beneficial shifts in gut microbiome and epithelium that alleviate systemic and brain inflammation. NSAIDS, often used as first-line PANS immune modulation, can disrupt bowel epithelial barrier function [55], potentially promoting the development of IBD in a predisposed child. Predisposition to IBD is felt to be complex and influenced by genetic factors, environment, microbiota, and immune response.

The concept of gut-brain axis, the connection between GI and neurologic function via autonomic nerves, hormones, neurotransmitters, microbes, and immune pathways [56], may explain why GI complaints are common in the PANS population and may track flares. It may also provide a mechanistic link between PANS and IBD.

PANS and IBD populations share an increased rate of immunoglobulin deficiencies, which were found in 3 of our 8 subjects (38%). In a large survey of 656 PANS cases, 62 (9.5%) were reported to have low total IgG or IgG subclass or inadequate response to pneumococcal vaccine, and 9 (1.4%) had low IgA [57]. Another review of 114 PANS cases found low total IgG in 26%, IgG subclasses in 27%, IgM in 16%, and IgA in 11%. Children with low IgA, IgG, or IgG subclass levels were more likely to achieve remission with IVIg at 12 month follow-up [58]. IBD patients also frequently have low immunoglobulin levels; a survey of 324 patients found low IgG, IgG1, IgA, and IgM levels in 22%, 23%, 8%, and 11%, respectively [59]. Low IgG, IgG G1, and IgG G4 have been associated with poorer clinical outcomes, such as need for small bowel resection [60, 61].

Iron deficiency and low ferritin levels are common in both PANS and IBD [62, 63]. Prevalence of iron deficiency and hypoferritinemia in PANS is 3–8% and 27% respectively, with hypoferritinemia rising to three quarters during PANS flare [62]. In IBD, 12% of patients have iron deficiency (defined as low ferritin, or low transferrin saturation if ferritin is not available) and 43% have iron deficiency with concurrent anemia [63]. Although data on ferritin were not available for all our subjects, two had documented hypoferritinemia.

Despite these similarities in presentation, there are striking differences in the current management strategies of IBD and PANS. There may be lessons from decades of IBD experience that may be applied to the more recently described PANS.

For instance, a core principle of IBD care is risk stratification to guide therapy. Over the last 2 decades, IBD management has shifted to a precision based approach, with individuals at high risk for complications receiving early aggressive therapy, often with biologics and immune suppressants [64]. Reducing the time to effective treatment reduces IBD complications [65]. Because younger age at onset is associated with poorer IBD outcomes [66, 67], earlier advanced treatments are often indicated for pediatric IBD. However, concern for lifetime medication effects, lack of pediatric efficacy and safety data, and slower FDA approvals conspire to delay access to newer, more effective IBD treatments for children. Regardless, from 2010 to 2019, biologic use in pediatric CD increased from 53% to 79% in a large 90 site survey of 4,035 patients [68], with initiation of therapy decreasing from 978 to 52 days after diagnosis. At the same time, stricturing, penetrating, and perianal disease dropped from 24% to 10%, suggesting again that early therapy improves outcomes.

In contrast, in PANS, there is no current guidance on how to risk stratify. It would be ideal to identify which children will have more benign courses (single episode, or relapsing-remitting with full return to baseline) and who will suffer more chronic, progressive disease. Clinicians currently lack markers to identify which patients respond to routine PANS interventions (clearance of GAS, NSAIDS, oral corticosteroid bursts) and who will require more intensive treatments (IVIg and aggressive immunomodulation). Furthermore, access to intensive treatments is obstructed by insurer denials. Therefore, treatment is usually approached in a “step-up” fashion, where widely available, inexpensive, oral therapies are tried first, and more costly and invasive options are saved for refractory cases. A recent survey demonstrated that delays in treatment are associated with more persistent PANS symptoms [69], suggesting that an early, aggressive treatment approach may be a better strategy for some patients, as in IBD. While additional research is sorely needed in this area, many rheumatological illnesses have shown better outcomes with early aggressive care, such as the reduction in joint damage in rheumatoid arthritis through early use of DMARDS [70]. Likewise, a child’s developing brain would be expected to benefit from rapid diagnosis and treatment of inflammation in PANS, as is recognized to be the case in neuropsychiatric systemic lupus erythematosus, autoimmune encephalitis, and multiple sclerosis [71‒74]. Our series suggests that the presence of arthritis and family history of PsO in a child with PANS may identify candidates for aggressive, early immune modulation due to a high risk for future autoimmunity.

Another concept central to modern IBD care is “treat-to-target.” That is, medication management should be escalated until a set goal is achieved, such as clinical, biochemical, or endoscopic remission [75]. IBD patients who achieve normalization of fecal calprotectin and endoscopy suffer fewer complications [76, 77]. Unfortunately, it is more challenging to identify biomarkers for central nervous system diseases like PANS. Routine bloodwork is generally unrevealing or non-specific. Abnormalities on brain imaging in PANS may be apparent only when groups of patients are compared to controls [11‒14] or if individuals are followed serially over time and imaging is compared to their healthy state [78, 79]. Treat-to-target efforts currently rely on parent-reported symptoms and subtle exam findings. We do not yet know what the treatment target should be for PANS, nor whether children should be treated until completely back to baseline or merely on an improving trajectory. In contrast, IBD treatment is divided into induction of remission and maintenance care [80]. Currently, the concept of maintenance therapy is variably applied in PANS, with some clinicians focusing just on treating flares, and others treating beyond flare resolution, aiming for deeper disease quiescence.

Treatment of IBD differs from PANS in other notable ways. In IBD, antibiotics are only occasionally part of management. Prior antibiotics are associated with higher likelihood of incident IBD, especially when given in the first year of life or repeatedly [81‒85]. It remains to be seen whether antibiotics frequently given in the setting of PANS might contribute to the later development of IBD.

The use of NSAIDs in PANS and IBD also differs. NSAIDs are employed in PANS to reduce symptoms and duration of flares [86]. However, NSAIDS are traditionally avoided as possible precipitants of IBD per American College of Gastroenterology guidelines [81, 87, 88]. Current European consensus guidelines state there is no evidence for a relationship between NSAID use and UC flare but a potential association with CD [89]. Thus, it is unclear whether NSAIDS used for PANS could contribute to eventual development of IBD (CD subtype) in these children.

Th17 lymphocytes and their signature cytokine IL-17A have been implicated in PsO, psoriatic arthritis, spondyloarthritis, multiple sclerosis, and PANS [90‒92]. Recently, IL-17A inhibition in the mouse model of PANDAS partly restored blood-brain barrier function and microglial expression of chemokine genes [91]. Several IL-17 inhibitors (i.e., secukinumab, ixekizumab) are FDA-approved to treat PsO and psoriatic arthritis. Since PsO is so prevalent in first-degree relatives in this case series and in the general PANS population [24], a possibility exists that these medications may help a subgroup of individuals with PANS (manuscript in process). However, IL-17A inhibition is felt to trigger or exacerbate IBD and is generally avoided in this population [92, 93]. Prior to starting a IL-17A inhibitor for PANS, it would be ideal to identify and exclude those patients at risk for incipient IBD.

Immune-mediated inflammatory diseases have been traditionally categorized by the primary organ system affected, with care siloed into respective medical specialties. There is growing recognition that these illnesses are in fact systemic and require multidisciplinary care. Immune diseases are increasingly categorized by the signature “cytokine hub” that is dysregulated [94]. Our observation that half of the children in our series have first-degree relatives with PsO and personal histories of enthesitis-related arthritis and spondyloarthritis hints that our subjects are predisposed to the family of IL-23 driven diseases. In IBD, selection of a biologic agent is often influenced by comorbidities; for example, patients with PsO and IBD may be good candidates for certain advanced treatments indicated for both conditions. Similarly, it is possible that neuropsychiatric symptoms in patients with the triad of PANS, arthritis, and family history of PsO could respond to treatments that work well for both PsO and IBD. Insurance barriers for most immune modulators for PANS persist currently, limiting accrual of clinical experience [69]. Therefore, funding of research on newer immune-based therapies for PANS is needed. In our series, most patients’ PANS improved once IBD was treated, suggesting that IBD should be treated with a standard-of-care approach, if it develops in the setting of PANS.

To our knowledge, this is the first case series to document the co-occurrence of two uncommon conditions, IBD and PANS. Firm conclusions are difficult to draw from small, observational series. Long follow-up of prospective cohorts of PANS and IBD patients will be needed to confirm our observations, ideally in a population with broader racial, socioeconomic, and geographic diversity. Nevertheless, our series is hypothesis generating for future research. Further research to characterize fecal microbiome in PANS could shed light on any similarities with IBD, hopefully advancing stool microbiota testing to a point of clinical utility. Additional immunological and genetic evaluation of the patients in our series could elucidate common mechanisms for IBD and PANS. Examination of a large cohort of PANS patients for other GI conditions (e.g., food sensitivities, irritable bowel syndrome [IBS], and celiac disease) and other similar syndromes (e.g., long COVID and ME/CFS, which also have prominent fatigue and GI symptoms) could spotlight shared disease mechanisms.

In conclusion, both PANS and IBD are complex, immune-mediated disorders requiring multidisciplinary care, due to manifestations outside the primary organ system. Our examination of 8 individuals with both PANS and IBD suggests (1) an extremely high rate of immune-mediated disorders, especially PsO, in first-degree relatives, (2) frequent comorbid arthralgias or arthritis, and (3) a tendency for PANS to predate IBD development and to improve with standard IBD treatment. These findings lend support to an immunologic basis of PANS and suggest that cytokines important in IBD and arthritis, such as IL-23, IL-17, and TNFα, may also mediate neuropsychiatric disease for some children. The triad of PANS, arthritis, and family history of PsO may represent a subset of PANS at heightened risk for additional immune disorders that may benefit from increased surveillance, including evaluation of GI symptoms with tests appropriate for bowel inflammation. Research is needed to determine if treatment with medications effective for IBD, PsO, or arthritis may alleviate neuropsychiatric symptoms of PANS in this group. Because they may be at heightened risk for IBD, patients with the PANS and arthritis or family history of PsO should be monitored closely if receiving IL-17 inhibitors.

Both PANS and IBD have severe, lasting consequences on the lives of affected children. Future research is needed to identify potential cytokine targets for PANS, to understand which children would benefit from these advanced therapies, and to clarify treatment strategies and goals. Until then, our series sheds some light on aspects of IBD care that might be applied to improving PANS management.

We would like to thank the Neuroimmune Foundation and Sam Pleasure, MD, PhD, for organizing/facilitating this special edition following the 2024 Inflammatory Brain Disorders Conference, and Loren G. Miller, MD, MPH, for assistance with manuscript preparation. We are grateful for our patients and families who understand treatment limitations and continue to lend their time and cooperation to research participation.

Written informed consent was obtained from all adult patients and from the parent/legal guardians/next of kin of all patients younger than 18 years to participate in the study and for publication of this case report and any accompanying images. This study protocol was reviewed and approved by the Stanford Panel on Human Subjects Institutional Review Board (IRB), Approval #26922; Cedars-Sinai IRB, Study #2918; and Dartmouth Hitchcock Medical Center IRB, Approval #23171.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder. However, J.C.M. receives funding from the Alex Manfull Fund Foundation, the PACE Foundation, the Arizona Biomedical Research Center, The CF Foundation, NIEHS R01 ES033988-09; NIH UG3 0D023275-09. S.R. serves on the advisory for Janssen Pharmaceuticals, Prometheus Laboratories, and Pfizer Pharmaceuticals.

Angela W. Tang, Jennifer Frankovich, and Shervin Rabizadeh conceptualized the project. Angela Tang, Paula M. Prieto Jimenez, Ian K.T. Miller, Juliette C. Madan, and Jaden Nguyen performed primary data collection. Meiqian Ma, Melissa Silverman, Bahare Farhadian, Cindy Manko, Jennifer Wilson, and Yinka Davies assisted with patient data collection. Angela W. Tang, Paula M. Prieto Jimenez, and Jennifer Frankovich validated data. Angela W. Tang performed data analysis and wrote original draft. Angela W. Tang, Paula M. Prieto Jimenez, Ian K.T. Miller, Meiqian Ma, Juliette C. Madan, Melissa Silverman, Jennifer Wilson, Jaden Nguyen, Bahare Farhadian, Alka Goyal, Cindy Manko, Yinka Davies, Shervin Rabizadeh, and Jennifer Frankovich reviewed and edited the manuscript.

Drs. Rabizadeh and Frankovich are co-senior authors.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.