Factors Associated with Symptom Persistence in PANS: Part II—Presenting Features,Medical Comorbidities,and IVIG Treatment History

Introduction

Pediatric Acute-Onset Neuropsychiatric Syndrome (PANS) is characterized by obsessive-compulsive symptoms and/or restricted food intake, as well as neuropsychiatric and somatic symptoms including anxiety, emotional lability, depression, irritability, aggression, oppositional behaviors, behavioral regression, sleep disturbances, urinary frequency, sensory or motor abnormalities, and deterioration in school performance (Swedo et al., 2012). Onset and exacerbations often follow infections, frequently with Streptococcus pyogenes (termed Pediatric Autoimmune Neuropsychiatric Disorder Associated with Streptococcus: PANDAS), but can also be triggered by environmental factors. Consensus guidelines on evaluation (Chang et al., 2015) and intervention (Swedo et al., 2017) provide a standard approach to this condition, including behavioral support (Thienemann et al., 2017), immunomodulatory therapies (Frankovich et al., 2017), and management of infections (Cooperstock et al., 2017).

The frequency and duration of PANS exacerbations vary widely across affected patients. Although generally considered a relapsing-remitting illness (Cooperstock et al., 2017), complete relief of symptoms between exacerbations is not always achieved (Gromark et al., 2022). In a 34-patient cohort followed by clinician assessment over 2-5 years, 2 patients were classified as remitted, 20 as relapsing–remitting, and 12 as having a chronic-static/progressive course (Gromark et al., 2022). Compared to those with relapsing-remitting or remitted disease, levels of IL-1-β, TNF-α, and IgM were significantly elevated in patients who experienced chronic symptoms, suggesting a more inflammatory state. Similarly, in a large-scale survey sample of 698 cases, 19% had experienced PANS symptoms every day since onset, while 19% had been asymptomatic more than 75% of days (Calaprice et al., 2017). However, no large-scale study to date has attempted to identify what clinical or treatment-related features differentiate individuals with persistent versus intermittent symptoms. Understanding potential risk factors for symptom persistence might empower providers to take preventive measures to prevent chronic disease. Moreover, in other autoimmune neurological diseases, for example, multiple sclerosis and neuropsychiatric lupus (Bates, 2011; Schwartz et al., 2019), patients experiencing different disease courses respond differently to therapeutic interventions; thus, understanding the range of courses experienced by PANS patients may aid our understanding of variability in therapeutic response.

In 2017 and 2018, Calaprice and colleagues described the symptoms, course, medical and family histories, and treatment experiences of 698 patients with PANS, the largest survey cohort to date (Calaprice et al., 2018; Calaprice et al., 2017). Although the authors reported briefly that receiving sufficient antibiotic treatment to resolve the infection associated with PANS onset was associated with lower probability of recurrence, a thorough investigation of factors associated with symptom persistence was beyond the scope of the original analyses.

The current study further analyzes presentation, treatment, and medical history traits associated with PANS symptom persistence, as measured by the respondent-reported % days symptom-free. We hypothesized that more persistent PANS symptoms could reflect a state of chronic, systemic immune activation and that this would be reflected in associations between symptom persistence and (1) higher rates of comorbid inflammatory conditions (specifically allergies, chronic sinusitis, musculoskeletal pains, rashes, recurrent high fevers, and/or asthma) and immune deficiencies, (2) a stronger tendency for PANS symptoms to flare in response to a range of pathogens and other immune-activating environmental triggers, and (3) lack of effective immunomodulatory treatment. We further hypothesized that more persistent disease would be associated with poorer outcomes, including more numerous symptoms and functional difficulties.

Methods

Results

The demographic and disease-history characteristics of the sample for this analysis are described in the companion to this article (Calaprice-Whitty et al., 2023).

Discussion

Our findings reveal compelling associations between the persistence of PANS symptoms as measured by respondent-reported % days symptom-free and subjects' medical comorbidities, inciting infections for recurrences, responses to environmental triggers, and IVIG treatment experiences. Increased symptom persistence in our study was also associated with more pervasive debility in a wide range of realms, including more problematic neuropsychiatric and behavioral symptoms, cognitive and perceptual symptoms, sleep symptoms, urinary issues, fatigue, pain, and developmental issues, as well as an inability to attend school even with accommodation. Although it is not possible to assert causality for the associations revealed in this observational study, the life-altering consequences of persistent symptoms make it important to explore all potential opportunities to shift patients toward more intermittent symptoms and more durable remissions. Encouragingly from this perspective, symptom persistence appears to be impacted by potentially modifiable variables, including infectious and environmental exposures, although intrinsic factors such as a genetic vulnerability to immune dysfunction also likely play a role.

In general, the specific comorbidities and symptoms that were more often reported for subjects with more persistent PANS symptoms are common in many rheumatologic conditions, that is, pain (headaches, musculoskeletal), immune deficiencies, dermatologic and sinus issues, digestive concerns, fatigue, irritability, and brain fog (Dickerson et al., 2007; Groven et al., 2018; Hansen et al., 2020; Kogan et al., 2020). Subjects with more persistent symptoms did not, however, suffer higher rates of allergies or asthma. It is conceivable that these common conditions are more likely to be treated on a consistent basis with systemic anti-inflammatory medications and that such medications may in turn mitigate PANS symptoms; the relationships between comorbid conditions, medications, and PANS symptoms deserve further exploration.

Greater symptom persistence was also associated in our sample with symptom reactivity to a wider range of immune stimuli, including infections in cohabitants, vaccinations, and environmental irritants, as well as with infections that are often persistent, for example, chronic sinus infections and EBV, and with flares in comorbid conditions. These observations suggest that, in contrast to subjects with more intermittent disease, persistently symptomatic PANS subjects may suffer a more consistently active inflammatory response. Peripheral inflammation has known negative effects on cognitive function and emotional processing, with potential mechanisms including disruption of the blood–brain barrier and activation of pro-inflammatory cytokines, T/B cells, and monocytes (Bauer and Teixeira, 2019; Bollen et al., 2017; Huang et al., 2021; Miller, 2020; Platt et al., 2017; Van Erp et al., 2022; Yarlagadda et al., 2009). The greater levels of immune compromise, lower responsiveness to the initial antibiotic course, and broader range of infectious triggers associated with symptom persistence in our sample also suggest that the ratio of immune challenge to immune competence may play a role in symptom persistence. Whether some patients may benefit from increased antibiotic dosing or duration requires further study.

The muted response to IVIG treatment in subjects with more persistent PANS symptoms is also of interest in this context. In our sample, subjects with greater symptom persistence were more likely to have received IVIG, possibly because of their more problematic PANS symptoms and/or their higher rates of relatively pervasive immune deficiency. In other chronic inflammatory disorders, including multiple sclerosis, lupus, and rheumatoid arthritis, IVIG treatment is reserved for clearly relapsing-remitting, refractory, or special cases (e.g., pregnancy), as the benefits for chronic cases generally do not outweigh the liabilities (Dalakas, 2004; Hoffman and Enk, 2019; Katz-Agranov et al., 2015; Mulhearn and Bruce, 2015; Shoenfeld and Katz, 2005; Sorensen, 2003). Instead, standard treatment for chronic cases relies on corticosteroids, DMARDS (Disease Modifying Anti-Rheumatic Drugs), B-cell suppressors, and/or cytokine inhibitors to provide more consistent suppression of the inflammatory state than periodic IVIG can achieve.

Children with chronic/persistent PANS possess many clinical features found in other chronic rheumatic conditions such as lupus, psoriatic arthritis, and ankylosing spondylitis, all of which are also seen at elevated frequencies in PANS patients and/or first-degree relatives (Calaprice et al., 2017; Chan et al., 2020; Fernell et al., 2021; Frankovich et al., 2015; Murphy et al., 2010). In fact, 28% of patients meeting PANS criteria at Stanford's clinic developed chronic arthritis confirmed by MRI or ultrasound, often years later (Ma et al., 2023). Interestingly, it is not uncommon for psychiatric symptomatology, including OCD, to antedate development of the physical manifestations of autoimmune illnesses such as Sjogrens and inflammatory bowel disease (Ananthakrishnan et al., 2013; Chang et al., 2021; Walker et al., 2008). These observations suggest the need for long term follow-up of PANS patients to determine risk of developing additional rheumatologic conditions and for research into the role of DMARDS, B-cell suppressors, and cytokine inhibitors in treating persistent PANS.

In our sample, developmental concerns were present even in significant minorities of those with low-symptom persistence, while the majority of those with highly persistent symptoms carried one or more developmental diagnoses and suffered high rates of respondent-perceived developmental lags relative to peers in every functional realm probed. Possible explanations for these associations are that PANS and developmental issues share a common pathophysiologic mechanism; that they share common risk factors; or that either condition predisposes a child for the other (e.g., social skills lag due to loss of opportunities to develop such skills during PANS exacerbations, Tona et al., 2017). Our survey did not inquire whether developmental diagnoses were rendered before or after the PANS diagnosis; thus, it is also possible that, because of diagnostic overshadowing (i.e., an individual's new symptoms are attributed to the previously-diagnosed condition rather than being investigated independently, thought to occur often in mental health conditions; The Joint Commission, 2022), individuals with preexisting developmental disorders may not have received prompt workups at PANS onset, resulting in delayed diagnosis and treatment and leading to a more persistent course (Calaprice-Whitty et al., 2023). In any case, our findings point to the importance of inquiring about developmental concerns during PANS evaluations.

Our findings also suggest the potential utility of interventions to improve occupational participation. Occupational therapists may identify opportunities to accommodate deficits during exacerbations and assist with “catch up” development as exacerbations subside (Tona and Posner, 2011). Ultimately, further research exploring the root cause of the association between PANS and developmental issues, as well as research examining the effectiveness of occupational therapies, could help professionals develop appropriate interventions to promote skill development.

The results of this study underscore the recommendations for clinical evaluation from the 2013 PANS Consensus Conference (Chang et al., 2015), which called for inquiring about a history of rashes over the past 6 months and for serologic testing relevant to potential rheumatologic diagnoses if a rash, joint pain, fatigue, or elevated inflammatory markers are seen. In our sample, such symptoms correlate with higher PANS symptom persistence and may support the provider not only in differential diagnosis but also in risk-stratifying patients to consider for more aggressive treatment. Our findings also support Chang et al.'s recommendation to maintain a low index of suspicion to screen for immune deficiencies. Immune abnormalities are so commonplace in the PANS population that universal screening may even be warranted, including not only measurement of total immunoglobulins but also of IgG subclasses and specific antibody deficiencies that may be detected by inadequate response to multivalent pneumococcal vaccination; any of these may be clinically relevant to the ability to clear a specific infection and to the development of autoimmunity (Jefferis and Kumararatne, 1990; Morell, 1994).

This study has the weaknesses associated with any retrospective observational survey, including that participants are self-selected and diagnoses are self-reported. With respect to self-selection, respondents who represent more serious cases may be more engaged with PANS patient communities and hence more likely to respond to solicitations for research participation; thus it is possible that our sample skews in this direction and is not representative of the PANS population as a whole. On the flipside, however, it is also possible that respondents who represent more serious cases may find it more burdensome to participate in research activities with no expected personal benefit; thus our sample may also skew the other way. The fact that no benefit whatsoever was offered to subjects to complete the lengthy survey would be expected to minimize participation by those not meeting diagnostic criteria. Although each of these factors may play some role in shaping our sample, the fact that our sample characteristics are consistent with those seen in practice-based studies of PANS (e.g., gender and age distribution, time to diagnosis, range of acuity of onset, symptom frequencies, recurrence rates) suggests our sample to be representative of the general PANS population that participates in research (Frankovich et al., 2015; Johnson et al., 2019; Lepri et al., 2019; Murphy et al., 2015).

It is also possible that recall bias plays a role in a few of our findings. For example, caregivers of subjects with more persistent symptoms may be more likely to track or recall events that led to symptom exacerbations or may be more focused on tracking individual symptoms or functional deficits as they occur. To mitigate this effect, we offered, in the survey, “don't recall” or “don't know” options for every question so that respondents who could not recall clearly did not feel compelled to guess, and we then excluded such responses from the analysis. Indeed, if recall biases such as these were in play, we would expect their effect to be most apparent for more subtle events or symptoms that might escape the notice of less-focused caregivers. Instead, the most pronounced differences were observed in symptoms that are likely to be memorable to all, for example, aggression, urinary incontinence, pain, sleep problems such as night terrors and nightmares, and diagnoses of developmental disorders.

In contrast, a major strength of our survey format is that it achieves a large broad sample otherwise not possible for an uncommon disease. This naturalistic study represents “real world” community medicine and not simply the experience of PANS specialty clinic populations, thus serving to confirm the generalizability of studies of those populations. Our naturalistic data may also be hypothesis-generating for prospective studies. For example, we urge future research on the intersection between PANS and other rheumatologic illnesses, as well as prospective studies of disease modifying anti-rheumatic drugs (DMARDS) in PANS patients at high risk of further autoimmunity. Such research should include long-term follow-up of both medically-supervised and community-based cohorts and consideration of the hypothesis that PANS may in fact represent an early and/or subsyndromal presentation of a rheumatologic condition.

Future research may also consider the relationship between % symptom-free days and disease course according to classical rheumatologic classifications, that is, relapsing-remitting vs. chronic. Although the measure % symptom-free days shows, importantly, what proportion of time a person feels and functions normally (and does in fact correlate in our sample with function, for example, the ability to attend school), it does not entirely distinguish between individuals with frequent short flares and those with few or single long bouts. The effectiveness of programs focusing on occupational engagement is another area deserving of further investigation. Finally, research relevant to the identification and avoidance or management of a variety of infections and inflammatory triggers could help in the development of strategies to prevent more persistent and pervasive disease manifestations.

Conclusion and Clinical Significance

The present study sheds light on the wide-ranging experiences of PANS subjects in a large community sample and identifies features associated with symptom persistence. Specifically, subjects experiencing greater PANS symptom persistence also reported higher rates of: medical comorbidities, including rashes, pain (head, muscle, joint), digestive concerns, chronic sinusitis, and immune deficiencies; symptom exacerbations in association with infections in close contacts, vaccinations, a variety of environmental triggers, flares of comorbid medical conditions, and infections with EBV, mycoplasma, or “sinus infections”; developmental diagnoses and respondent-perceived developmental lags; and difficulties attending school. They were also more likely to receive IVIG but less likely to find it effective. Together with our companion article, the results presented here suggest that both intrinsic and extrinsic factors, including exposures and access to interventions, may influence the symptom persistence experienced by individual patients. Adherence to published assessment and treatment recommendations may improve disease course, with pervasive benefits on health, function, and quality of life.