Authors: Casara Jean Ferretti, Benjamin Lê Cook, Aakash Mahant Mahant, Philip Chu, Yin Zhao, Bonnie P. Taylor, Betsy C. Herold, Eric Hollander
Categories: Research Articles, Autism spectrum disorder, Cytokines, Endophenotypes, Biomarkers, Cognition
Source: Neuroscience Applied
Authors: Casara Jean Ferretti, Benjamin Lê Cook, Aakash Mahant Mahant, Philip Chu, Yin Zhao, Bonnie P. Taylor, Betsy C. Herold, Eric Hollander
Cognitive inflexibility is a transdiagnostic endophenotype that presents across a range of disorders, including autism spectrum disorder (ASD), which is maintained through adulthood, and encompasses both cognitive and behavioral rigidity. There is evidence for immune dysfunction in a subgroup of ASD, including systemic inflammation, cytokine dysregulation and anti-brain autoantibodies. Although immunome pathways are involved in ASD pathophysiology, there is little known about how they relate to symptom domains or symptom severity, and whether such biomarkers may be useful in optimizing future clinical trials. We correlated baseline clinical measures of cognitive inflexibility and resultant irritability with immunome biomarker levels in children with ASD, aged 5–18 years with ABC-I scores ≥18, CGI-S scores ≥4 and SRS-2 scores ≥66T, at Albert Einstein College of Medicine (AECOM). Non-parametric Spearman correlations and estimated multivariable regression analyses adjusting for potential confounders, including other clinical variables, race, sex, and age were completed. Strong positive correlations (rs > .70) were found between the Montefiore Einstein Rigidity Scale – Revised (MERS-R) Total Score and the pro-inflammatory cytokines IL-6 (rs=.80), granulocyte-colony stimulating factor (GCSF; rs =.72), macrophage inflammatory protein-1 alpha (MIP-1a; rs =.71). The MERS-R subscales also had moderate and strong correlations with the immunome biomarkers. The MERS-R Total Rigidity Subscale Score had a strong positive relationship with IL-6 (rs = 0.7856). Using multivariable regression analyses significant relationships were found between the MERS-R Total Rigidity Subscale Score and proinflammatory cytokine IL-18 (p = 0.02), and a nonsignificant trend was found between it and IFN-alpha2 (β = −4.982, p = 0.058). The ABC-I was significantly correlated with pro-inflammatory cytokine IL-18 (p = .013), IFN-alpha2 (p = 0.039), the anti-inflammatory cytokine IL-10 (p = 0.02), and adaptive immunity cytokine IL-2 (p = 0.041). This preliminary data is the first to examine the relationship of clinical measures of cognitive inflexibility and immunome biomarkers in children with ASD, and may provide a framework for better understanding the relationship between immunome mechanisms, cognitive inflexibility, and ASD symptomatology. Clinicaltrials.gov: NCT03202303.
Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental disorder defined by social and communication deficits and restricted and repetitive behaviors (American Psychiatric Association, 2013) occurring in approximately 1 in 36 children, per the Centers for Disease Control (CDC) (Walensky et al., 2023). Individuals with ASD and their caregivers often present for treatment for symptoms that fall outside the core domains, including irritability, aggression, and self-injurious behaviors (Fig. 1). These behaviors are often related to the restricted and repetitive behavior domain (RRB) of ASD, with insistence on sameness behaviors, including cognitive and behavioral inflexibility, causing significant impairment and disruption to functioning across the lifespan. There are also high rates of medical and psychiatric comorbidities, with approximately 70% of individuals with ASD having at least one comorbid psychiatric diagnosis, including Attention-deficit/hyperactivity disorder (ADHD), Obsessive Compulsive Disorder (OCD), anxiety and mood disorders. At present there are no FDA approved treatments for the core symptoms of ASD, and approved treatments such as risperidone and aripiprazole only target irritability and aggression, and have significant metabolic and weight adverse effects. There is a need for a better understanding of the cognitive and behavioral flexibility present in those with ASD, how it relates to biomarkers, and how targeted treatments could improve quality of life and caregiver burden.Fig. 1ASD core and associated symptom domains (Breece et al., 2013, Hollander and Ferretti, 2023).Fig. 1
Cognitive inflexibility is a transdiagnostic endophenotype present across a range of disorders, including ASD (Uddin, 2021; Fineberg et al., 2021), and encompasses both cognitive and behavioral rigidity. Those with cognitive inflexibility cannot easily transition from one thought or behavior to another, and often become stuck in ritualistic or repetitive patterns of behavior. They are characterized as being cognitively rigid, with an intense focus on their own thoughts, beliefs, and behaviors to the exclusion of all others, which limits their ability to be flexible, adapt to changes in routine, and to manage unpredictability (Lecavalier et al., 2020). When expectations are violated, less flexible thinking can lead to behavioral rigidity, including challenges with new, unfamiliar, and unexpected situations, a need for sameness and order in the environment, and difficulty set-shifting (Yerys et al., 2009). Cognitive inflexibility, or rigidity, negatively effects adaptive skills, and is predictive of emotional and behavioral difficulties across the lifespan (Hollocks et al., 2022; Ozsivadjian et al., 2021) for both those with and without mental illness, and improved flexibility positively impacts both executive functioning and social skills (Strang et al., 2017). Within ASD, cognitive inflexibility is one component of the “insistence on sameness” construct within the RRB domain (Lecavalier et al., 2020; Strang et al., 2017; Scahill et al., 2015). Individuals with ASD present with greater levels of cognitive inflexibility than those without, and this specific executive functioning construct is related to both RRB severity and emotional and behavioral challenges across the lifespan, independent of other ASD symptoms (Hollocks et al., 2022). Cognitive and behavioral inflexibility is also predictive of adaptive behavior impairments, and contributes to social communication deficits and interpersonal challenges. It is also related to higher levels of anxiety, depression, aggression and opposition, and compounds psychiatric and medical comorbidities, including OCD, ADHD, and mood disorders (Ozsivadjian et al., 2021; Lei et al., 2022). The association between cognitive inflexibility and emotional symptoms in those with ASD continues to remain significant even when accounting for the covariance with RRBs (Hollocks et al., 2022). Disruptive and protest behaviors are a consequence of cognitive inflexibility, as when an individual struggles to adapt to changes in expectations or in the environment they experience increased levels of irritability, often resulting in externalizing outbursts (Ozsivadjian et al., 2021; Lawson et al., 2015). Irritability is well described in individuals with ASD, and the only FDA approved treatment for ASD targets this significant symptom which severely impacts quality of life and the ability to participate in educational, occupational, and therapeutic activities across the lifespan (Fung et al., 2016; Hirota and King, 2023). Understanding the mechanisms of, and directly targeting cognitive inflexibility, may thus reduce irritability, significantly improving quality of life for adults with ASD.
Cognitive inflexibility is also one mechanism through which co-occurring mental health comorbidities and behavioral and emotional difficulties are maintained through the lifespan (Hollocks et al., 2022). Over the lifespan, in combination with bullying in middle school, it leads to ruminative brooding, and higher rates of externalizing and disruptive behaviors in emerging adulthood (Fig. 2). Thus, identifying measures of cognitive inflexibility, and how this symptom domain is influenced by mechanisms of lipid and immune dysfunction identified by biomarker profiles, is critical in developing effective pharmacological and psychosocial interventions to improve quality of life for individuals with ASD and their families.Fig. 2Longitudinal effects of cognitive inflexibility (Hollander and Ferretti, 2023).Fig. 2
The etiology of ASD is complex and involves genetic, epigenetic, environmental, lipidomic and immune-inflammatory factors (Hsiao, 2013; Loke et al., 2015). There is significant evidence for immune dysfunction in ASD, including systemic inflammation, cytokine dysregulation and anti-brain autoantibodies (Meltzer and Van de Water, 2016; Lyall et al., 2014; Masi et al., 2017a). Both prenatal and postnatal exposures to immune triggers may lead to an altered immune system and thus altered neurodevelopment (Fox et al., 2012). Individuals with ASD have impaired immune system regulation and heightened inflammatory processes demonstrated by altered cytokine and chemokine profiles (Ashwood et al., 2011a) and increased microglial activation (Hsiao, 2013; Masi et al., 2017a; Pardo et al., 2005; Onore et al., 2012; Goines and Ashwood, 2013; Careaga et al., 2010). Changes in cytokines in ASD may be developmentally regulated as they differ when measured during the neonatal period compared to other developmental periods. Additionally, the use of immunomodulatory treatments in ASD have been experimentally shown to reduce RRBs, including cognitive and behavioral rigidity as measured by the Montefiore-Einstein Rigidity Scale – Revised (MERS-R) (Hollander et al., 2020). Lastly, individuals with intensely upregulated metabolome and immune genes in postmortem brain tissue may also have a neuroinflammatory condition in which atypical microglial activation derails normal neurogenesis and synaptogenesis across various conditions, including ASD (Chen et al., 2022). Most of the expression changes observed in brain tissue samples do not appear in gene expression patterns in blood samples from people with the same conditions. Thus, these changes may not reflect the specific disorders, but rather a transdiagnostic endophenotype that cuts across conditions, including ASD.
Immune-inflammatory dysfunction is found across the lifespan in individuals with ASD. Children with ASD have significant increases in plasma levels of Th1 cytokines, and elevated levels are associated with impaired communication and aberrant behaviors (Ashwood et al., 2011b). Increases in the pro-inflammatory cytokines, including the interleukins (IL), IL-1β, IL-6, IL-8, and IL-12p40, are associated with more regressive autism and more pronounced stereotypical behaviors (Ashwood et al., 2011a, 2011b). Reduced levels of TGF-β1 are associated with more severe symptoms of ASD, worsened stereotypy, irritability and hyperactivity, and reduced adaptive behavior, while increased levels of eotaxin, MCP-1, RANTES, IL-1β, IL-4, IL-6, IL-8 and IL-12p40 are associated with more severe irritability, hyperactivity, stereotypy, and lethargy; as well as greater impairments in visual reception, fine motor skills and receptive and expressive language, and increased impairments in adaptive behaviors (Masi et al., 2017a). PET ligand studies in ASD adults demonstrate microglia activation, with greater microglia activation associated with increased ASD severity (Suzuki et al., 2013). C-reactive protein (CRP) in peripheral blood is also higher in ASD vs controls (Yin et al., 2020), and is significantly higher in those with severe vs mild/moderate ASD symptoms. Macrophage migration inhibitory factor (MIF), an upstream regulator of innate immunity that promotes monocyte/macrophage-activation responses, has also been identified as a susceptibility gene for ASD and is also elevated in serum of depressed patients (Grigorenko et al., 2008; Musil et al., 2011). In examining the distribution of serum biomarkers (CRP, IL-6, MIF), severe cases of ASD had significantly elevated CRP, IL-6 and MIF vs mild-to-moderate ASD (Ning et al., 2019).
These cytokines have diverse and significant roles in the immune system. TNF-α is a proinflammatory cytokine involved in synapse maturation and stabilization, which at high levels has been shown to impair the development of the visual system by prematurely stabilizing developing synapses leading to increase local connectivity, and potentially, epilepsy (Masi et al., 2015; Lee et al., 2010). IFN-γ is responsible for coordinating the transition from innate to adaptive immunity by amplifying immune system sensitivity and promoting macrophage activation. By skewing the immune response toward the Th1 phenotype IFN-γ inhibits the proliferation of Th2 cells and subsequently anti-inflammatory cytokines. This is relevant to the hypotheses involving aberrant neuroconnectivity in ASD due to either excessive or insufficient synaptic pruning. IL-1β is known to affect the hypothalamic-pituitary-adrenal (HPA) axis and is involved in both acute and chronic inflammation. The proposed impaired chemokines in ASD, eotaxin, IL-8 and monocyte chemoattractant protein-1 (MCP-1), are involved in the recruiting neutrophils, eosinophils, monocytes and T-cells to areas of tissue inflammation (Masi et al., 2015, 2017a, 2017b).
IL-6 is an important neuroimmune factor that is shown to be involved in brain development and several neurological disorders. It can act to both promote neural growth and cause neuronal death, depending on its concentration, brain region and the cell type affected. IL-6 facilitates the communication between the CNS and immune system, and has been linked to the impairment of neural cell adhesion and migration, synapse formation and synaptic plasticity. Elevated levels have been linked to detrimental effects on synaptic plasticity in addition to physiological and pathological effects on learning and memory. Its role in promoting neuronal differentiation in the adult hippocampus may be comprised if it is chronically elevated, as seen with altered levels of synaptic proteins in the hippocampus during chronic IL-6 administration. Alterations in neural cell adhesion and migration may lead to an imbalance of inhibitory and excitatory synaptic transmissions, including elevated excitatory synapse formation and impaired inhibitory synapses, resulting in reduced paired-pulse inhibition and short-term synaptic plasticity. The impaired excitatory/inhibitory balance theory is one of the leading hypotheses in ASD etiology, and the role of IL-6 demonstrates that it may be linked to the immune-inflammatory hypotheses (Wei et al., 2013).
As mentioned above, individuals with ASD have been observed to have altered cytokine profiles compared to healthy controls, and to be in a chronic state of cytokine induction. Many studies have shown those with ASD have a dampened Th2 anti-inflammatory cytokine response and an increased Th1 pro-inflammatory cytokine response, including an increased innate and adaptive immune response through the Th1 pathway, suggesting that localized brain inflammation and autoimmune dysfunction may be involved. More specifically, increases in IL-6, IL-8 and IL-23 have been observed, with an increased Th1 to Th2 ratio when compared to controls. These aberrations have been shown in amniotic fluid (elevated IL-4, IL-10, tumor necrosis factor – α (TNF-α), transforming growth factor – beta (TGF-β)), plasma and peripheral blood mononuclear cells. Further studies have shown increased proinflammatory cytokines in the brain and cerebrospinal fluid (CSF) of individuals with ASD, including higher levels of proinflammatory IL-6, IL-8, IL-12 and interferon-gamma (IFN-γ) and TNF-α; in addition to elevated proinflammatory Th17 cytokines (Masi et al., 2017a; Pardo et al., 2005; Masi et al., 2015; Masi et al., 2017b; Chez et al., 2007; Garbett et al., 2009; Abdallah et al., 2013; AL-Ayadhi and Mostafa, 2012; Croonenberghs et al., 2002; Li et al., 2009; Ricci et al., 2013; Chez and Guido-Estrada, 2010). This upregulation of cytokines has been linked to the activation of microglia and astroglia in those with ASD. Meta-analysis on studies that compare the plasma and serum concentrations of cytokines in unmedicated ASD and healthy participants found significantly altered concentrations of cytokines, further strengthening the hypothesis of an abnormal cytokine profile and immune system dysregulation in ASD that results in increased proinflammatory signals and CNS impairment. The largest effect sizes showed an elevation of proinflammatory IFN-γ and a reduction in anti-inflammatory TGF-β1, while increased levels of proinflammatory IL-1β, IL-6, eotaxin, IL-8 and MCP-1 were also significant. IL-8 is also shown to be associated with the level of paternal and maternal cytokines in families of children with ASD in addition to the individual with ASD themselves, suggesting its role in ASD pathogenesis (Shen et al., 2021). As mentioned above, elevated levels of IL-8 have been found in multiple studies of those with ASD, and has been associated with ASD severity and cognitive and adaptive ability, in addition to social communication (Ashwood et al., 2011b; Masi et al., 2015; Li et al., 2009; Korzeniewski et al., 2018). Shen et al. (2021) also found that in addition to IL-8, IL-7, interferon-alpha (IFN-alpha), and interferon gamma-induced protein-10 (IP-10) were also associated with an ASD diagnosis when controlling for parental cytokine levels.
An altered inflammatory response, observed as elevated cytokines, has been linked to ASD symptoms. Those who have fluctuating behavioral symptoms following immune insults have additional innate immune abnormalities and significantly altered transcript profiles of peripheral blood monocytes (Jyonouchi et al., 2011). Constipation, the most prominent GI symptom in those with ASD, has also been associated with increased levels of blood myeloid dendritic cells (Breece et al., 2013). These cells act as messengers between the innate and adaptive immune system, and present antigen material to the T-cells of the immune system. Higher levels of these cells are also associated with increased repetitive behaviors in those with ASD, larger amygdala volumes, developmental regression and a later onset of ASD symptoms. Typically developing children with altered dendritic cell frequencies are shown to have poorer adaptive behavior skills. Associations with the amygdala are relevant to ASD etiology, as it is implicated in alterations of core social abilities in ASD and the modulation of anxiety, in addition to having a role in modulating GI activity (Breece et al., 2013). Reduced adaptive behavior skills were also associated with lower plasma levels of TGF-β1, a crucial regulator of the immune system and mediator in the development of autoimmune and systemic inflammation. Children with ASD and low plasma levels of TGF-β1 have worse behavioral scores on the Aberrant Behavior Checklist (ABC) including higher levels of irritability, social withdrawal, hyperactivity and stereotypies. These results were seen in both children with early onset ASD and those who had regressed, but not for typically developing children or those with developmental disabilities (Ashwood et al., 2009). In contrast, other studies have showed higher cytokine profiles in those with ASD and regression compared to those with ASD without regression (Ashwood et al., 2011b). Both populations had higher levels than typically developing children, and higher levels of Th1 proinflammatory cytokines IL-1β, IL-6, IL-8 and IL-12p40 were associated with severity of core behaviors on the ADI-R and more severe aberrant behaviors on the ABC, in particular increased stereotypical behaviors. However, there are also studies showing no differences in immune abnormalities between those with early onset ASD and those with developmental regression, and more clarification is needed (Ashwood et al., 2011c). Increases in Th2 cytokine production are associated with better behavior scores on the ABC (IL-13), improved adaptive behaviors on the Vineland Adaptive Behavior Scale (VABS), and better developmental and cognition scores on the Mullen Scales of Early Learning (MSEL) (IL-10 and IL-5). In contrast, increases in Th1 proinflammatory cytokines are associated with severity of ASD behaviors on gold standard measures of ASD, the Autism Diagnostic Interview-Revised (ADI-R), the Autism Diagnostic Observation Schedule (ADOS) and aberrant behaviors on the ABC. Levels of chemokines MCP-1, eotaxin and RANTES observed to be increased in the plasma of ASD children compared to typically developing and developmentally disabled controls were also associated with severity of behaviors, cognition and adaptive function observed on the MSEL, ABC and VABS (Ashwood et al., 2011b).
It is critical to identify and refine this potential transdiagnostic endophenotype linked to immunome alterations and cognitive inflexibility, as it could both identify a potential subgroup within ASD, and also provide more targeted and personalized interventions. Although immunome pathways are involved in ASD pathophysiology, there is little known about how they relate to symptom severity, symptom domains, and which biomarkers may be useful in optimizing future clinical trials. We have completed a first step in identifying these relationships, by comparing constructs of cognitive inflexibility with immunome biomarker measures in the child ASD population.
Human subjects included 20 children aged 5 to 14 (8.450 ± 3.170) diagnosed with ASD by a medical or mental healthcare provider, confirmed by DSM-5 criteria and supported by the Autism Diagnostic Observation Schedule-2 (ADOS-2). All participants had an engaged caregiver who was consented, and completed rating scales, providing feedback to the clinician throughout the study. All participants had an Aberrant Behavior Checklist – Irritability subscale (ABC-I) score of 18 or higher, a Social Responsiveness Scale −2 (SRS-2) score of 66T, and a Clinical Global Impression – Severity score of 4 or greater at baseline. Written informed consent and assent were obtained from all study participants and their caregivers. The study was IRB approved by the Albert Einstein College of Medicine Institutional Review Board (IRB # 2017–8538). A consort diagram has been provided (Fig. 3).Fig. 3Consort diagram.Fig. 3
The preliminary data presented in this publication was collected at baseline as a part of a larger DOD funded clinical trial that is still ongoing. The study objective of this 12-week, double-blind, placebo-controlled trial is to determine the efficacy and safety of cannabidivarin (CBDV) versus placebo in children and adolescents diagnosed with ASD with high levels of irritability. For the purposes of this publication, we will only focus on the data collected at the baseline visit in those subjects with biomarker samples collected, as the study is ongoing and data are still blinded. The twenty participants visited the Autism and Obsessive Compulsive Spectrum Program at the Psychiatry Research Institute of Montefiore-Einstein (PRIME) to complete their baseline visits. During this visit, they completed a battery of clinician and caregiver-rated questionnaires, in addition to clinical and research laboratory measures, physical exams and vitals. Baseline bloodwork was collected by the Clinical Research Center (CRC), a member of the Institute for Clinical and Translational Research (ICTR) at Einstein-Montefiore. Immunome, lipidomic, and metabolomic biomarker measures were stored in the ICTR Biomarker and Biorepository Core (BBC) biorepository until panel analyses were completed. Blood sampling was not completed under fasting conditions, and the majority of samples were collected in the late afternoon. Immunome analyses were completed by the lab of Betsy Herold, MD at Albert Einstein College of Medicine. Lipidomic, metabolomic and endocannabinoid biomarker measures are still being analyzed.
Montefiore-Einstein Rigidity Scale – Revised (MERS-R): Individuals with ASD often have restricted interests, repetitive behaviors, insistence on sameness and difficulty transitioning. These characteristics can be defined as impairments in cognitive flexibility, a construct of executive functioning which is often defined as the ability to shift between discrepant tasks and demands. A lack of flexibility is observed clinically as cognitive and behavioral rigidity, which when present in ASD significantly impacts functioning. Cognitive and behavioral rigidity, in addition to the severity of the individual's response to changes (protest), and rigidity in general, are not well captured in available outcome measures (Children's Yale-Brown Obsessive Compulsive Scale for ASD (CYBOCS-ASD), Repetitive Behavior Scale-Revised (RBS-R)) as they are often either non-specific, and do not focus on rigidity, or overly specific, using very specific examples of rigid behaviors but omitting many others. The MERS-R is a clinician-administered assessment that measures three domains of rigid behavior in children/adolescents and adults, (1) Behavioral Rigidity, (2) Cognitive Rigidity, and (3) Protest (Taylor et al., 2022). All available information is used to rate items, including clinical observations, interviews with the subject and caregiver/significant other, and clinical judgment. The assessment focuses on the prior 2-week interval. Questions are ranked on a scale of 0 (none) to 4 (extreme), and assess time spent/frequency, interference, distress associated with disruption and control (difficulty getting unstuck) and severity. The Behavioral Rigidity Domain measures difficulty managing behavior in new, unfamiliar or unexpected situations, including discomfort when a deviation from expectation occurs, or when a change interrupts plans/activities. It also measures difficulties with transition and insistence on sameness. The Cognitive Rigidity Domain measures lack of flexible thinking, including repetitive/perseverative questioning, checking for reassurance, difficulty switching conversation topics, and black and white, concrete, or literal thinking. Lastly, the Protest Domain measures behavioral reactions in response to an interruption to a subject's rigidity, including arguing, screaming, verbal/physical aggression, irritability and non-responsiveness. The MERS-R demonstrates good internal consistency for the total score (Cronbach's α = 0.89), behavioral rigidity domain subscale (Cronbach's α = 0.83, protest domain (Cronbach's α = 0.88) and total rigidity score subscale (Cronbach's α = 0.82); and adequate internal consistency for the cognitive rigidity domain subscale (Cronbach's α = 0.72)(Taylor et al., 2022). It has significant convergent validity with the Children's Yale-Brown Obsessive Compulsive Scale for ASD (CY-BOCS-ASD), good to excellent test-retest reliability over 2 weeks (ICC = 0.66–0.79), and fair to good test-retest reliability over 12 weeks (ICC = 0.56–0.66) (Taylor et al., 2022). The MERS-R was also shown to be a sensitive outcome measure of change in a treatment study examining the effect of an immunomodulatory treatment, Trichuris Suis Ova, on rigidity in ASD (Hollander et al., 2020). The MERS-R has been used in several federal, industry and foundation sponsored studies of 10.13039/100007726ASD with good reliability and validity.
Aberrant Behavior Checklist -Irritability Subscale (ABC-I): The ABC is an informant rating instrument that was empirically derived by principal component analysis to measure behavior in those with developmental disability and ASD (Aman, 1985; Aman M, 2012). It contains 58 items that resolve into 5 subscales, including the Irritability Subscale used in this analysis, which is composed of 15 items that address the presence of aggression, tantrums and/or self-injury. The ABC was designed to be completed by any adult who knows the individual well, such as a parent/caregiver or teacher. This instrument measures behavior on a four-point severity scale where 0 = no problem at all, 1 = behavior is a problem but in a slight degree, 2 = problem is moderately serious, and 3 = problem is severe in degree. The ABC has been used in multiple ASD trials to measure changes in irritability and social withdrawal, with success. It demonstrates good to excellent internal consistency for every subscale (Cronbach's α = 0.85–0.94), except the inappropriate speech subscale which as acceptable levels of internal consistency (Cronbach's α = 0.77)(Aman, 1985; Kaat et al., 2014). The Aberrant Behavior Checklist – Irritability (ABC-I) subscale assesses an individual's level of irritability, which can be exacerbated due to cognitive inflexibility, and challenges adapting to changes in the environment. The scores on the subscale range from 0 to 45. The ABC-I subscale has an excellent internal consistency level (Cronbach's α = 0.92)(Kaat et al., 2014). Subjects scored an 18 or higher on ABC-I at screening and baseline to be included in this study.
Blood was collected from participants at their in-person baseline visit at the Clinical Research Center (CRC) of the ICTR, and plasma isolated, divided into aliquotes and stored at −80 °C in the BARC biorepository until the time of analysis. Cytokine concentrations in plasma were measured using an 11-plex Milliplex MAP Human Cytokine/Chemokine Magnetic Bead Panel (Millipore). Samples were prepared per the manufacturer's instructions. Data were acquired on a Luminex Magpix (Luminex Corporation) and analyzed in the Belysa Milliplex Analyst program (Millipore). All the standard curves had an R^2^ between 0.99 and 1.0. Analytes measure were granulocyte-colony stimulating factor (G-CSF; CSF-3), granulocyte-macrophage colony-stimulating factor (GM-CSF), IFN-α2, IFN-γ, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, IL-13, IL-17A, IL-18, IP-10, MCP-1, macrophage inflammatory protein-1 alpha (MIP-1α), macrophage inflammatory protein-1 beta (MIP-1β), TNF-α and TNF-β. Values below the lower limit of detection (LLOD) were set at half the LLOD. GM-CSF was undetectable in all subject's samples and was not included in the final analysis. Pro-inflammatory cytokines generally include IL-6, IL-1, TNF-alpha, IL-12, IL-18, IFN-gamma, while anti-inflammatory cytokines include IL-4, IL-10, IL-13, TGF-beta although functions are complex and may vary.
Data were visually inspected using histograms and boxplots. Stata 18 software was used to complete the analyses (StataCorp, 2023). Categorical variables were summarized using univariate descriptive statistics. We assessed Spearman correlations between behavioral and biomarker measures, choosing a nonparametric statistical analysis because our biomarker measures are not expected to follow a specific distribution. Multivariable regression analyses were estimated to examine the association between clinical/behavioral measures (MERS-R Total and Subscale scores, ABC-Irritability subscale) and immunome biomarkers, while covarying/adjusting for confounders (age, sex, race, and ethnicity).
Participant demographics, baseline clinical characteristics and immunome biomarker values are summarized in Table 1. Participant ages ranged from 5 to 14 years (M = 8.450 ± 3.17) (Table 1). Eighteen participants were male and two were female, 50% of the sample identified as Black race and 40% of the sample identified a Hispanic or Latino. Participants clinical and behavioral characteristics were captured using measures of adaptive behavior (Vineland-3 ABC score, M = 49.70 ± 16.5), irritability (ABC-I, M = 24.25 ± 9.96), social communication (SRS-2 Total T Score, M = 85.05 ± 9.72), global clinical impression (CGI-S, M = 4.35 ± 0.48), and cognitive and behavioral inflexibility (MERS-R Total Score, M = 33.0 ± 6.0; MERS-R Total Rigidity Subscale Score, M = 22.3 ± 4.27). Further details are presented in Table 1. Immunome biomarker data is summarized in Table 2.Table 1Baseline demographic and clinical characteristics.Table 1Total (N = 20)CharacteristicN%Sex Male1890% Female210%Race Asian15% Black1050% White420% Multiracial315% Other210%Ethnicity Hispanic/Latino840% Not Hispanic/Latino1050% Not Reported210%NMeanSDRangeAge (Years)208.4503.1705–14Behavioral MeasuresNMeanSDRangeVineland-3 Adaptive Behavior Composite Score2049.75016.54930.0–79.0Clinical Global Impression Scale – Severity (CGI-S)204.3500.4894–5Aberrant Behavior Checklist – Irritability Subscale (ABC-I)2024.2509.9684.0–42.0Social Responsiveness Scale – 2 (SRS-2) Total T-Score1985.0539.72766–102Montefiore Einstein Rigidity Scale (MERS-R) Total Score1033.06.01820–38MERS-R Total Rigidity Subscale Score1022.34.27013–26Table 2Baseline immunome biomarker values in children with ASD.Table 2Immunome BiomarkersNMeanSDRangeGranulocyte-colony stimulating factor (GCSF)2063.56249.97518.010–199.070Interferon alpha 2 (IFN-α2)2044.53950.39111.075–242.430Interferon gamma (IFN-γ)204.9602.3734.250-14.20Interleukin – 1 Beta (IL-1β)209.01913.7280.965-64.470Interleukin 2 (IL-2)201.1882.6560.245-12.180Interleukin 4 (IL-4)203.5312.4410.520-9.650Interleukin 5 (IL-5)204.9303.4561.22-14.94Interleukin 6 (IL-6)201.1671.3310.5–5.14Interleukin 8 (IL-8)202.2951.1760.49-5.37Interleukin 10 (IL-10)2021.57342.9954.860-200.180Interleukin 12p70 (IL-12p70)206.1147.4920.190-31.010Interleukin 13 (IL-13)2021.16327.6920.215-98.470Interleukin 17A (IL-17A)207.25811.3761.795-49.280Interleukin 18 (IL-18)2076.05656.27017.02–221.530Interferon gamma-induced protein 10 (IP-10)20216.64999.31360.79–420.94Monocyte Chemoattractant Protein 1 (MCP-1)20330.709175.448173.060–963.20Macrophage inflammatory protein-1 alpha (MIP-1α)2020.54717.1347.390-76.240Macrophage inflammatory protein −1 beta (MIP-1β)2020.3346.7179.140-34.70Tumor Necrosis Factor alpha (TNF-α)2028.42517.8523.69–70.48Tumor Necrosis Factor beta (TNF-β)207.1087.2734.705-35.290
Strong positive correlations (rs > .70) were found between the MERS-R Total Score and the pro-inflammatory cytokines IL-6 (rs=.80), granulocyte-colony stimulating factor (GCSF; rs =.72), and MIP-1a (rs =.71). A moderate negative correlation was found between the MERS-R Total Score and the pro-inflammatory cytokine MCP-1 (rs = -.53) and a moderate positive correlation was found between it and the anti-inflammatory cytokine IL-10 (rs =.65). The MERS-R subscales also had moderate and strong correlations with the immunome biomarkers. The MERS-R Behavioral Rigidity subscale had moderate positive correlations with IL-6 (rs = 0.68), GCSF (rs = 0.60), MIP-a (rs = 0.61), IL = 10 (rs = 0.62), IFN-gamma (rs = 0.42) and TNFβ (rs = 0.47); the MERS-R Cognitive Rigidity subscale had a moderate negative correlation with IL-5 (rs = −0.61), and the MERS-R Total Rigidity Subscale Score had a strong positive relationship with IL-6 (rs = 0.7856), moderate positive correlations with MIP1-a (rs = 0.69), GCSF (rs = 0.53), IL-10 (rs = 0.52), IFN-gamma (rs = 0.46), TNFβ (rs = 0.46), and IL-17a (rs = 0.41), and a moderate negative correlation with MCP-1 (rs = −0.51). (Table 3). No other pairwise comparisons reached a level of significant correlation. Scatterplots of significant relationships can be seen in Fig. 4.Table 3Spearmen correlations (ρ): MERS-R and immunome biomarkers.Table 3MERS-R Behavioral Rigidity SubscaleMERS-R Cognitive Rigidity SubscaleMERS-R Total Rigidity Subscale ScoreMERS-R Total ScorerprprprpIL-6 (pro-inflammatory)0.68760.00660.18790.55870.78560.00710.80720.0048GCSF (pro-inflammatory)0.60770.0212−0.02880.92920.53380.11200.72170.0185MIP-1a (pro-inflammatory)0.61790.01850.07820.80910.69330.02620.71640.0198
IL-10 (anti-inflammatory)0.62430.01700.16130.61640.52600.11830.65250.0409MCP-1 (pro-inflammatory)−0.39600.1610−0.33340.2895−0.51380.1287−0.53050.1147IFN-gamma (pro-inflammatory)0.42030.1346−0.14060.66300.46850.17200.46700.1735TNFβ (pro-inflammatory)0.47360.0871−0.14060.66300.46850.17200.46700.1735IL-17a (pro-inflammatory)0.37170.1907−0.01200.97050.41000.23930.43970.2036IL-5 (adaptive immunity)0.00340.9909−0.61600.0330−0.31600.3738−0.27830.4362IFN-alpha2 (pro-inflammatory)0.35220.2168−0.13020.68670.36030.30650.40250.2488
IL-1β0.24410.4003−0.22630.47950.16870.64130.26910.4521IL-20.30030.29680.00001.00000.39130.26350.39010.2651IL-40.07060.8104−0.11310.72630.15340.67230.16820.6423IL-8−0.21700.4561−0.25460.4246−0.22630.5295−0.18290.6130IL-12p700.11010.7078−0.37800.22570.03720.91880.13580.7083IL-130.29450.3068−0.23510.46200.29500.40790.31580.3741IL-180.07830.79020.23660.45900.15290.67320.35980.3072IP-100.14100.63080.04300.89440.22020.54100.26830.4536MIP-1b0.31770.2683−0.11110.73090.23850.50690.32320.3624TNF-α0.26180.3659−0.26170.41120.24470.49570.29880.40170.00–0.19 = very weak correlation; 0.20–0.39 = weak correlation.0.40–0.69 = moderate correlation; 0.70–0.89 = strong correlation.Fig. 4Scatterplots of immunome biomarkers with cognitive inflexibility measures.Fig. 4
Using multivariable regression analyses significant relationships were found between the MERS-R Total Rigidity Subscale Score and proinflammatory cytokine IL-18 (p = 0.02), and a nonsignificant trend was found between it and IFN-alpha2 (β = −4.982, p = 0.058). Nonsignificant trends were also found between the MERS-R Behavioral Rigidity Subscale score and pro-inflammatory cytokines IL-8 (β = −0.83, p = 0.060) and IFN-alpha2 (β = 14.57, p = 0.099), and anti-inflammatory cytokine IL-10 (β = −48.91, p = 0.08), while the MERS-R Cognitive Rigidity Subscale had a nonsignificant relationship with the pro-inflammatory cytokine MCP-1 (β = 33.91, p = 0.065) after adjustment for age, sex, race and ethnicity. The ABC-I was significantly associated, after adjustment for age, sex, race and ethnicity with pro-inflammatory cytokine IL-18 (p = .013), IFN-alpha2 (p = 0.039), the anti-inflammatory cytokine IL-10 (p = 0.02), and adaptive immunity cytokine IL-2 (p = 0.041) (Table 4).Table 4Multiple Regression Analyses Covarying for sex, age and Immunome Biomarkers and Cognitive Inflexibility Measures.Table 4Immunome BiomarkerCoefficientStandard ErrorsR-squaredBetatP > |t|MERS-R Behavioral Rigidity SubscaleIFN-alpha214.5762.2780.99711.266.400.099IL-8−0.83490.79100.9989−1.26−10.560.060IL-10−48.9126.6040.9977−1.29−7.410.085MERS-R Cognitive Rigidity SubscaleMCP-133.9133.4991.00.2159.690.065MERS-R Total Rigidity ScoreIL-18−16.5730.51151.000-.790−32.500.02IFN-alpha2−4.9820.45160.9994−1.05−11.030.058MERS-R Total ScoreIL-5−2.6680.17830.9976−5.63−14.960.043ABC-Irritability subscaleIL-1812.8000.25511.0002.1450.170.013IL-107.161.190.98342.145.990.027IFN-alpha23.6960.22520.99942.7416.410.039IL-20.03890.00251.000.56015.340.041IL-51.0980.8650.99874.1412.690.050IL-80.33820.07910.99893.7812.590.050
Cognitive inflexibility is a transdiagnostic endophenotype that negatively affects functioning and quality of life across the lifespan, but has yet to be explored fully in the literature. Identifying how this symptom domain is influenced by mechanisms of immune dysfunction, a well-characterized feature of ASD, is critical to both better understanding the pathophysiology of ASD, and developing effective pharmacological and psychosocial interventions to improve quality of life for individuals with ASD and their families. The aim of this preliminary data analysis was to dimensionally explore the transdiagnostic endophenotype of cognitive inflexibility, its resultant irritability, and immunome biomarkers in children with ASD. Using nonparametric Spearman correlations we found a strong correlation between the MERS-R total and subscale scores and multiple immunome biomarkers, including IL-6, GCSF, MIP-1a, and IL-10. We further analyzed this relationship using multivariable regression analyses covarying for sex, race and age, and found significant relationships between the MERS-R Total Rigidity Score and IL-18, and between the ABC-I subscale and IL-18, IFN-alpha2, IL-10 and IL-2. It is possible, particularly due to the small sample size, that adjustment of age, sex, race and ethnicity in the multivariable regression analyses could have influenced the significance of the correlations, and notably different immunome biomarkers were significant after adjusting for covariates. Additionally, the ABC-I did not show strong correlations with immunome biomarkers, but did show significance in the multiple regression analyses. Due to the small sample size we discuss the immunome biomarkers significant in both the spearman correlations and in the multiple regression analyses.
For individuals with ASD, irritability is often exacerbated in response to changes in routines and the environment, and can be representative of an individual's level of cognitive and behavioral inflexibility, and future analyses will examine correlations between these behavioral measures. Elevations in IL-6 have consistently been found in individuals with ASD, and are noted to cause changes in excitatory and inhibitory synaptic formations that disrupt the excitatory/inhibitory balance. Previous work has shown that elevations in IL-6 are associated with more severe repetitive behaviors as measured by the Repetitive Behavior Scale-Revised (RBS-R)(Hughes et al., 2022). Thus, it is not surprising that IL-6 levels are strongly correlated with symptoms of cognitive and behavioral inflexibility, which are a part of the restricted and repetitive behavior domain.
The chemokine, MCP-1, recruits circulating monocytes and neutrophils (as well as other immune cells) to sites of inflammation and has also been shown to be elevated in those with ASD compared to healthy controls (Masi et al., 2015). However, MCP-1 also promotes tissue repair and remodeling. Notably, we found a moderate negative correlation with cognitive and behavioral inflexibility using the MERS-R Total Score suggesting that it may be protective, and this relationship will require further study.
Levels of IL-18, a cytokine that influences both cellular and humoral immunity, have also been found to be dysregulated in those with ASD, with decreases observed in serum and increases in the brain of both individuals with ASD and mouse models (Businaro et al., 2016). Our regression analyses found that IL-18 was significantly associated with both cognitive and behavioral rigidity and resultant irritability, via the MERS-R Total Rigidity and ABC-I subscales further confirming its role in ASD pathophysiology.
The regulatory cytokine IL-10 is often found to have decreased levels in those with ASD, with significantly lower levels in those with ASD and gastrointestinal symptoms compared to those without (Ashwood, 2023). We found that IL-10 levels had a positive strong correlation with cognitive and behavioral inflexibility symptoms in children with ASD, and was significantly associated with irritability. Further analyses will be completed examining comorbidities to see if similar subgroups are present related to gastrointestinal symptoms.
Increased levels of IFN-alpha2 have also been associated with gastrointestinal symptoms in children with ASD, and in our population was significantly related to irritability. Gastrointestinal symptoms may cause increased irritability, and further examination of this relationship will be important in our next analyses.
IL-5 is produced by Th2 and mast cells, and affects IgE and eosinophil mast cell mediated responses, stimulates B Cells to secrete immunoglobins, and is a mediator of eosinophil differentiation and activation (Li et al., 2009; Molloy et al., 2006; Suzuki et al., 2011). Within individuals with autism there are conflicting results on IL-5 levels, with studies showing no significant difference compared to controls (Li et al., 2009; Nour-Eldine et al., 2022), lower levels of IL-5 compared to controls (Shim et al., 2024), and slightly elevated levels of IL-5 compared to controls (Suzuki et al., 2011; Saghazadeh et al., 2019). More research is needed to examine the role of this cytokine in ASD symptoms.
There are limited results on the role GCSF, MIP-1a, IL-2, and IFN-alpha2 play in the pathophysiology and symptomatology of ASD, and our work indicates there may be specific inflammatory profiles linked to cognitive and behavioral inflexibility that need additional exploration.
Our small sample of children and adolescents with ASD had high rates of both irritability (ABC-I M = 24.25, out of max score of 45) and cognitive and behavioral rigidity (MERS-R Total Score M = 33, out of max score of 48), in addition to severe social impairment as measured by the Social Responsiveness Scale (SRS-2, M = 85.05). Per typical ratios in ASD, we had more males with ASD than females with ASD participate in this study. Many of the clinical trials completed in the ASD community struggle with recruiting diverse samples due to roadblocks and struggles that impact research in other populations as well. Our location at Albert Einstein College of Medicine in the Bronx, NY has helped us in reaching a more diverse population, providing data that could be relevant in examining immune and inflammatory biomarkers between individuals with ASD of different racial and ethnic backgrounds. Our preliminary dimensional data correlating measures of cognitive inflexibility and irritability with immune-inflammatory markers is a first step in better understanding the role inflammation has in ASD pathophysiology, the transdiagnostic endophenotype of cognitive inflexibility, and how they are related. This transdiagnostic endophenotype may be measured not only with scales of cognitive inflexibility, but also of related symptoms and behaviors.
Disruptive and protest behaviors, including aggression and irritability, are often a consequence of cognitive inflexibility, as individuals struggle with changes to their expectations or in their environment. The relatedness of these symptoms, cognitive inflexibility and irritability, is observed in our results, as correlations were seen between inflammatory biomarkers and elevated scores on scales measuring both constructs (the MERS-R and ABC-I). The inflammatory biomarkers that correlated with each scale and its subscales varied, with IL-10, MCP-1 and IFN-alpha having correlations with at least two subscales. Further factor analyses will need to be completed in order to better understand the relationship of these inflammatory biomarkers, and if there is a particular subset of them related to the cognitive inflexibility endophenotype. In future analyses, we will also examine additional metabolomic biomarkers and clinical and behavioral measures, and their relationship to immunome biomarkers, cognitive inflexibility, and irritability.
Immune dysfunction and inflammation are well-recognized as contributors to ASD symptomatology, and anti-inflammatory treatments have shown reductions in RRBs and rigidity. Cognitive inflexibility, and its resultant disruptive behaviors, are underrecognized in ASD, but contribute over the lifespan to the greatest caregiver burden and functional disability for those with ASD and related disorders, including challenges with work performance, relationships, and the ability to live in a mainstream setting. Little is known about the relationship of cognitive and behavioral inflexibility to the immune dysfunction present in those with ASD, and the preliminary data provided by this proposal is key to the development of better treatment targets and outcome measures. Although the majority of research in this area show that those with ASD have increased levels of Th1 cytokines, and decreased levels of Th2 cytokines, with an altered Th1:Th2 ratio, the results regarding specific cytokines vary across studies. This makes it difficult to discern the specific effects altered cytokine levels may have on behaviors, and the mechanism of immune-inflammatory treatments.
It is critical to identify and refine this endophenotype which may be linked to immunome alterations, and could identify a potential subgroup within ASD, in order to provide more targeted interventions. Although immunome pathways are involved in ASD pathophysiology, there is little known about how they relate to symptom severity, symptom domains, and which biomarkers may be useful in optimizing future clinical trials. Irritability is well described in individuals with ASD, with the only Food and Drug Administration (FDA) approved treatments targeting this significant symptom domain which severely impacts quality of life and the ability to participate in educational, occupational, and therapeutic activities across the lifespan. Understanding the mechanisms of, and directly targeting cognitive inflexibility, may thus reduce irritability, significantly improving quality of life for those with ASD.
The strengths of our study include the use of a unique clinician-rated measure of cognitive and behavioral inflexibility, the MERS-R. Additionally, although other symptom domains have been examined in correlation with immune-inflammatory biomarkers, this is the first study to examine cognitive and behavioral inflexibility, which has significant effects on quality of life and functioning in those with ASD and related disorders across the lifespan.
The limitations of our study include our sample size, particularly of females with ASD. Furthermore, our age range is limited to children between the ages of 5 and 14, and therefore we do not know if these correlations would continue to present in children under 5, or in young and older adults with ASD. As we expect that cognitive inflexibility is transdiagnostic and present in both other developmental disorders and healthy controls, it would be beneficial to collect these biomarkers in these populations to complete between group differences. Lastly, due to our sample size we have limited our correlation data to the total score on our measure of cognitive inflexibility, the MERS-R and its subscales, and our regression data to irritability, per the ABC-I, and the MERS-R and its subscales. Additionally, we note that correlation at one point in time is not evidence of causation. Future analyses will include multiple correlation analyses with other subscales, and a factor analysis of the inflammatory biomarkers.
This preliminary data analysis is the first to examine the relationship of clinical measures of cognitive and behavioral inflexibility and immunome biomarkers in children with ASD, and serves as a first step toward better understanding the connection between immunome mechanisms and ASD symptomatology. We have also provided additional support for the MERS-R, our clinician-informed measurement of cognitive and behavioral inflexibility. Our preliminary data informs the development of an endophenotype of cognitive and behavioral rigidity in the 10.13039/100007726ASD population, and provides further support regarding the dysfunction of immune-inflammatory pathways which will allow for the development of more specific and personalized treatments that target this symptom domain. In targeting the symptom of inflexibility using both psychological and psychiatric treatments, irritability and aggression may indirectly be decreased, allowing for improved quality of life at both the individual and family level. These results also set the stage for similar work in larger populations of those with ASD and those with related conditions who also present with severe levels of inflexibility, including Prader-Willi syndrome (PWS), and OCD.
Written informed consent and assent was obtained from all participants and their caregivers. The Albert Einstein College of Medicine Institutional Review Board approved the protocol.
The study is currently ongoing, and the data remains blinded. At study completion data will be provided to the National Institute of Mental Health Data Archive (NDA).
This work was supported by the 10.13039/100000005Department of Defense (DOD) [grant number AR160104, W81XWH-17-1-025301]. Cannabidivarin (CBDV), matching placebo and additional support was provided by Jazz (formerly GW) Pharmaceuticals.
The authors declare the following financial interests/personal relationships which may be considered as potential competing Eric Hollander, MD reports financial support was provided by 10.13039/100000005U.S. Department of Defense. Eric Hollander, MD reports financial support and equipment, drugs, or supplies were provided by 10.13039/100011096Jazz Pharmaceuticals Inc. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.