CNS Drugs (2023) 37:215–229 https://doi.org/10.1007/s40263-023-00993-x REVIEW ARTICLE Immunoregulatory and/or Anti‑inflammatory Agents for the Management of Core and Associated Symptoms in Individuals with Autism Spectrum Disorder: A Narrative Review of Randomized, Placebo‑Controlled Trials Gara Arteaga‑Henríquez1,2,3,4 · Laura Gisbert1,2,3,5 · Josep Antoni Ramos‑Quiroga1,2,3,5 Accepted: 13 February 2023 / Published online: 13 March 2023 © The Author(s) 2023 Abstract Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition with a so far poorly understood underlying pathogenesis, and few effective therapies for core symptoms. Accumulating evidence supports an association between ASD and immune/inflammatory processes, arising as a possible pathway for new drug intervention. However, current literature on the efficacy of immunoregulatory/anti-inflammatory interventions on ASD symptoms is still limited. The aim of this narrative review was to summarize and discuss the latest evidence on the use of immunoregulatory and/or anti-inflammatory agents for the management of this condition. During the last 10 years, several randomized, placebo-controlled trials on the effectiveness of (add-on) treatment with prednisolone, pregnenolone, celecoxib, minocycline, N-acetylcysteine (NAC), sul- foraphane (SFN), and/or omega-3 fatty acids have been performed. Overall, a beneficial effect of prednisolone, pregnenolone, celecoxib, and/or omega-3 fatty acids on several core symptoms, such as stereotyped behavior, was found. (Add-on) treatment with prednisolone, pregnenolone, celecoxib, minocycline, NAC, SFN, and/or omega-3 fatty acids was also associated with a significantly higher improvement in other symptoms, such as irritability, hyperactivity, and/or lethargy when compared with placebo. The mechanisms by which these agents exert their action and improve symptoms of ASD are not fully understood. Interestingly, studies have suggested that all these agents may suppress microglial/monocyte proinflammatory activation and also restore several immune cell imbalances (e.g., T regulatory/T helper-17 cell imbalances), decreasing the levels of proinflammatory cytokines, such as interleukin (IL)-6 and/or IL-17A, both in the blood and in the brain of individuals with ASD. Although encouraging, the performance of larger randomized placebo-controlled trials, including more homogeneous populations, dosages, and longer periods of follow-up, are urgently needed in order to confirm the findings and to provide stronger evidence. 1 Introduction * Josep Antoni Ramos‑Quiroga antoni.ramos@vallhebron.cat Autism spectrum disorder (ASD) is a neurodevelopmental 1 Department of Mental Health, Hospital Universitari condition characterized by early-appearing (and persistent) Vall d’Hebron, Passeig de la Vall d’Hebron, 119‑129, deficits in social interaction and communication, sensory- 08035 Barcelona, Catalonia, Spain motor problems, and by repetitive and/or stereotyped behav- 2 Group of Psychiatry, Mental Health and Addictions, Vall iors [1]. With a current global prevalence of about 1%, ASD d’Hebron Research Institute (VHIR), Barcelona, Catalonia, is associated with a high community, and individual burden Spain [2, 3]. However, and despite many efforts and significant 3 Biomedical Network Research Centre on Mental Health advances in our understanding of the neurobiology of this (CIBERSAM), Barcelona, Catalonia, Spain condition, its exact etiology remains elusive [4]. As a con- 4 NCRR‑The National Center for Register‑Based Research, sequence, pharmacological options for the management of Aahrus University, Aahrus, Denmark ASD are still limited, and almost exclusively aimed at target- 5 Department of Psychiatry and Forensic Medicine, Universitat ing associated (e.g., irritability, hyperactivity), but not core Autònoma de Barcelona, Barcelona, Catalonia, Spain symptoms [5, 6]. Vol.:(0123456789) 216 G. Arteaga‑Henríquez et al. cells may lead to an abnormal development of the brain, and Key Points thus, predispose to neurodevelopmental conditions, such as ASD (Fig. 1). In support of this idea, microglial dysfunction Individuals with autism spectrum disorder (ASD) has been related to core ASD symptoms, such as stereotyped show immune and inflammatory abnormali- behaviors, in a mouse model of obsessive-compulsive disor- ties, including increased blood levels of (activated) der [22]. Other reports have demonstrated the existence of a monocytes,and abnormal blood levels of lymphocyte positive association between the severity of ASD symptoms subpopulations,across the lifespan. and the levels of Th17 cells (and/or related cytokines, such Individuals with ASD are also characterized by a proin- as interleukin [IL]-17A) [23, 24]. flammatory activation of microglia in the brain, across Interestingly, an increasing body of evidence has sug- the lifespan. gested the existence of a systemically and chronically activa- tion of microglia and of the monocyte system in individuals Immune system abnormalities found in individuals with with ASD, across the lifespan [25]. Supporting this idea, a ASD are a possible pathway for new drug intervention. significantly higher microglial cell density, and/or a signifi- (Add-on) treatment with immunoregulatory/anti-inflam- cantly higher expression of microglial cell activation mark- matory agents, such as prednisolone, pregnenolone, ers, such as CD45, have been repeatedly found in the brain celecoxib, minocycline, N-acetylcysteine, sulforaphane, of individuals with ASD [26–28]. Moreover, elevated blood and/or omega-3 fatty acids may be beneficial for the levels of monocytes [29–35], and an increased expression management of core (e.g., stereotyped behavior) and of different monocyte activation markers (e.g., CD96, HLA- associated (e.g., irritability, hyperactivity, lethargy) DR) have been also demonstrated [36, 37]. Abnormal lev- symptoms in individuals with ASD. els of several lymphocyte subpopulations, including natural killer (NK), B, ­CD8+ T cytotoxic (Tc), and ­CD4+ T helper (Th) cells have been demonstrated, both in the brain and in the blood of individuals with ASD [25, 29, 38–41]. Among During the last decades, an increasing body of evidence the different Th cell subpopulations, an increase in Th17 cell has suggested the involvement of the immune system in levels, and a decrease in Treg levels, have been suggested the pathophysiology of this condition. Genome-wide asso- [42, 43]. These findings are consistent with the cytokine pro- ciation studies have demonstrated that variations in several file that has been described in individuals with ASD, with genes that encode proteins involved in the inflammatory subjects with ASD showing increased levels of proinflamma- response (e.g., human leukocyte antigen [HLA] genes), may tory cytokines and/or chemokines (e.g., interferon [IFN]-γ, increase the risk of developing ASD [7]. In addition, mater- IL-1β, IL-6, IL-12p40, IL-17A, IL-31, tumor necrosis fac- nal immune activation, and/or suffering from an infection tor [TNF]-α, chemokine C-C-motif ligand [CCL]-2), and during pregnancy also increase the risk of developing ASD decreased levels of anti-inflammatory cytokines (e.g., IL-10) in the offspring [8, 9]. Individuals with ASD often display in their blood, and/or cerebrospinal fluid [44–48]. abnormal immune responses [10, 11], and are also character- Current evidence on the role of the immune system in the ized by a higher incidence of comorbid immune-mediated pathophysiology of ASD has therefore increased the interest on conditions, such as diabetes mellitus [12], psoriasis [13], the potential use of immunoregulatory and/or anti-inflamma- autoimmune thyroiditis [14], and/or allergy [15]. However, tory agents for improving core and associated symptoms in, at no consistent and/or specific immunological mechanism has least, a subgroup of individuals with ASD. The aim of this emerged, so far. narrative review was to investigate the existing evidence on Contrary to previous beliefs, the central nervous system the effectiveness of these agents for the management of ASD . (CNS) is not immune-privileged, and immune mediators, such as microglial cells, and/or T lymphocytes are normally present in the brain [16]. These cells are crucial for a proper brain development and function. For example, microglial 2 Literature Search cells participate in programmed neuronal death and pro- motion of synaptogenesis, and are able to strip excess syn- 2.1 Search Strategy apses from developing neurons, allowing the integration of functional neuronal circuits [17–19] (Fig. 1). In addition, The PubMed, SCOPUS, and World of Knowledge electronic and while T-helper regulatory cells (Tregs) promote myeli- databases were independently screened by two authors nation [20], T-helper 17 cells (Th17) promote demyelination (GAH, LG) for relevant articles published between 1 Janu- [21]. It has been therefore hypothesized that abnormal levels, ary 1994, and 5 April 2022. The following search syntax or an abnormal function of microglia and of several immune was used for the PubMed database search, and adapted Immunoregulatory and/or Anti-inflammatory Agents in Autism Spectrum Disorder 217 Fig. 1  Immune system abnor- malities in ASD according to the different database index terms: (autism OR or French languages were included. Exclusion criteria were autistic OR autism spectrum disorder OR ASD OR perva- (1) individuals with otherwise diagnosed ASD; (2) other sive OR pervasive developmental disorder OR pervasive immunoregulatory/anti-inflammatory agents; (3) agents developmental disorders OR Asperger OR Asperger’s) AND requiring invasive techniques for their administration; (4) (celecoxib OR aspirin OR minocycline OR anti-inflamma- in vivo agents, such as probiotics; and (5) observational or tory OR NAC OR N-acetylcysteine OR NSAID OR ACTH quasi-experimental (non-randomized) studies. OR prednisone OR prednisolone OR corticosteroids OR hydrocortisone OR methylprednisolone OR dexamethasone 2.3 Quality Assessment OR cortisone OR sulforaphane OR omega-3 OR fatty acids OR immunomodulatory). The Jadad scale (also known as the Oxford quality scor- ing system) is considered as the standard method for evalu- 2.2 Eligibility Criteria ating randomized clinical trials [50] and consists of three items: (1) randomization; (2) blinding; and (3) description Inclusion and exclusion criteria were discussed and of patient’s withdrawals/dropouts. Scores range from 0 to 5 approved by all authors; inclusion criteria for human studies points; a Jadad score of 0–2 indicates that the randomized were established in the Population, Intervention, Compara- clinical trial is of low quality, whereas a score of 3–5 indi- tor, Outcomes, and Study (PICOS) format, according to the cates a high quality [50]. Two points relate to randomization Preferred Reporting Items for Systematic Reviews and Meta- and both are awarded if the answers to the following ques- Analyses (PRISMA) guidelines [49]: (1) participants: chil- tions are ‘yes’: ‘Was the study described as randomized?’; dren/adolescents/adults with a confirmed diagnosis of ASD and ‘Was the method of randomization appropriate?’. Two as assessed by Diagnostic and Statistical Manual, Fourth points relate to blinding and both are awarded if, again, Edition (DSM-IV), DSM-IV Text Revision (DSM-IV-TR), the answers to the following questions are ‘yes’: ‘Was the DSM, Fifth Edition (DSM-5) or International Classification study described as double-blinded?’; and ‘Was the blind- of Diseases, Tenth Revision (ICD-10) criteria; (2) inter- ing method appropriate?’. One point relates to a patient’s vention: at least one of the following immunoregulatory/ withdrawals/dropouts. anti-inflammatory agents as single or add-on therapy: cor- ticosteroids, neurosteroids, non-steroidal anti-inflammatory drugs (NSAID), minocycline, N-acetylcysteine (NAC), 3 Summary of Findings sulforaphane (SFN), omega-3 fatty acids; (3) comparator: placebo, or treatment as usual (TAU) plus placebo, corre- In total, 18 studies were included in our review. The char- sponding to the intervention set-up; (4) outcome: validated acteristics of the included studies are displayed in Table 1, rating scale; and (5) study design: randomized, placebo-con- and the total scores and quality ratings of the single studies trolled trials. Only studies in the English, German, Spanish, are presented in Table 2. 218 G. Arteaga‑Henríquez et al. 3.1 Pharmacological Interventions to either risperidone + placebo (n = 29), or to risperidone + pregnenolone (n = 30), and followed up over a 10-week 3.1.1 Corticosteroids (Prednisolone) period (Table 1). The primary outcome measure included the change, from baseline to week 10, in the ABC irritability Corticosteroids such as prednisolone are a class of steroid subscale score; secondary outcome measures included the hormones secreted from the adrenal gland in response to change in other ABC subscale scores (i.e., hyperactivity, stress. Since their discovery in the 1940s, corticosteroids lethargy, stereotypy, inappropriate speech). After 10 weeks have been used for the treatment of several immune and/or of continuous treatment, individuals allocated to risperidone inflammatory diseases based on their immunosuppressive + pregnenolone showed a statistically significantly higher and/or anti-inflammatory effects [51]. To date, two rand- improvement in irritability, stereotyped behavior, and omized, placebo-controlled clinical trials have investigated hyperactivity compared to those allocated to risperidone + the effectiveness of (add-on) prednisolone on improving core placebo. A placebo- or pregnenolone-specific effect on leth- and associated symptoms in individuals with ASD [52, 53]. argy and/or inappropriate speech was however, not found Overall, studies suggested a beneficial effect of prednisolone (Table 1). Pregnenolone had a good safety profile and was over placebo on irritability, lethargy, stereotyped behav- well tolerated; adverse events were mild to moderate, and ior, and/or hyperactivity (as assessed by the change, from included changes in appetite, dizziness, rash, diarrhea, head- baseline to week 12 in the Aberrant Behavior Checklist- ache, or abdominal pain. Again, both study groups did not Community Edition [ABC-C] respective subscale scores) statistically differ in relation to the frequency, severity, or in at least, a subgroup of children with a regressive form of type of adverse effects. The mechanism by which pregne- ASD (Table 1). In addition, a trend of prednisolone-specific nolone exerts its anti-inflammatory and/or immunoregu- improvement in the Language Development Assessment latory action is again, not fully understood. Interestingly, Tool (ADL) total score and in the Child Language Test in several reports have suggested that neurosteroids, such as Phonology, Vocabulary, Fluency, and Pragmatics (ABFW) pregnenolone, are also able to suppress microglial and Th17 total of communicative acts subscale score was also found cells proinflammatory activation in humans, and in murine in, in particular, children with autistic disorder and a history models of autoimmune conditions [60]. of developmental regression (Table 1). Adverse effects were mild to moderate and included hypertension, hyperglyce- 3.1.3 Non‑steroidal Anti‑inflammatory Drugs [Celecoxib] mia, and/or changes in appetite. Significant differences in relation to the frequency, type, or severity of adverse events Celecoxib acts as a NSAID that selectively inhibits the were not found between individuals under (add-on) treat- cyclooxygenase (COX)-2 enzyme. This agent is better tol- ment with placebo and individuals under (add-on) treatment erated than steroidal anti-inflammatory drugs, and is associ- with prednisolone. The mechanism by which prednisolone ated with a lower risk of gastrointestinal bleeding [61]. To exerts its anti-inflammatory and/or immunoregulatory action date, only one randomized, double-blind, placebo-controlled is not fully understood. Increasing evidence from animal and trial has investigated the effectiveness of celecoxib on ASD human studies has suggested that corticosteroids may modu- symptomatology [62]. In this study, a total of 40 medication- late microglial activation, and also restore Treg/Th17 cell naïve children with autistic disorder and moderate-to-high imbalances [54–56]. Corticosteroids are also capable of levels of irritability were randomly allocated to either risp- increasing the number of mature NK cells [57], and of reduc- eridone + placebo (n = 20), or to risperidone + celecoxib ing blood levels of pro-inflammatory cytokines, such as IL-6 (n = 20), and followed up over a 10-week period (Table 1). and/or IFNγ, within their first weeks of administration. The primary outcome measure was the change, from base- line to week 10, in the ABC irritability subscale score, while 3.1.2 Neurosteroids (Pregnenolone) secondary outcome measures included the change in other ABC subscale scores (i.e., lethargy/social withdrawal, stere- Pregnenolone is a steroid hormone precursor that is syn- otypic behavior, hyperactivity, inappropriate speech). After thesized in different steroidogenic tissues, the brain, and in 10 weeks of continuous treatment, individuals allocated to lymphocytes. It is known to act as an anti-inflammatory and/ risperidone + celecoxib showed a statistically significantly or immunoregulatory agent in several (neuro)inflammatory higher improvement in irritability, lethargy/social with- diseases [58]. To date, only one randomized, placebo-con- drawal, and stereotyped behavior compared to individuals trolled study has investigated the effectiveness of pregne- allocated to risperidone + placebo. However, a celecoxib- nolone for the management of core and related symptoms specific improvement in other symptoms, such as hyperac- in subjects with ASD [59]. In this study, a total of 59 med- tivity and/or inappropriate speech was not found (Table 1). ication-naïve children and adolescents with ASD and mod- Celecoxib was well tolerated and both study groups did not erate-to-high levels of irritability were randomly allocated statistically differ in relation to the frequency, severity, or Table 1  Summary of randomized, placebo-controlled studies on immunomodulatory/anti-inflammatory therapies in autism spectrum disorder Agent Diagnosis (diagnostic Medication-free N Interven- Intervention (dose) Outcome Outcomes criteria applied) Age range tion period measures Structured/semi- (weeks) structured diagnostic interview Prednisolone Malek et al. (2020) [52] ASD (regressive) Only antipsychotic- Placebo (n = 13) 12 Risperidone: 0.5–2 mg/ ABC Prednisolone-specific [DSM-5] naïve Prednisolone (n = 13) day CARS improvement in: ADI-R Age range: 3–12 years Prednisolone: 1 mg/ ABC: irritability (p = 0.026), kg/day stereotypies (p = 0.026), hyperactivity (p = 0.039), lethargy (p < 0.001) CARS total score (p < 0.001) Rocha-Brito et al. Autistic disorder (DSM- No Placebo (n = 20) 24 Prednisolone: 1 mg/ ADL Trend of prednisolone-spe- * (2021) [53] IV) Pretreated with oral Prednisolone (n = 18) kg 8 weeks, 9–16 ABFW cific improvement in: NG albendazole (5 days) Age range: 3–7 years weeks 1 mg/kg/day CARS ADL (p = 0.057) every other day, final ABFW total of communica- 8 weeks: gradual dose tive acts (p = 0.054) reduction on alternate days Pregnenolone Immunoregulatory and/or Anti-inflammatory Agents in Autism Spectrum Disorder Ayatollahi et al. (2020) ASD (DSM-5) Yes Placebo (n = 29) 10 Risperidone: 0.5–3.5 ABC Pregnenolone-specific [59] ADI-R Pregnenolone mg/day improvement in: Moderate-to-high levels (n = 30) Age Pregnenolone: 200 mg/ ABC: irritability (p = 0.025), of irritability range:11-17 years day stereotypies (p = 0.029), hyperactivity (p = 0.004) Celecoxib Asadabadi et al. (2013) Autistic disorder (DSM- Yes Placebo (n = 20) 10 Risperidone: 0.5–2 mg/ ABC Celecoxib-specific improve- [62] IV) Celecoxib (n = 20) Age day ment in: ADI-R range: 4-12 years Celecoxib: 300 mg/day ABC: irritability (p < 0.001), Moderate-to-high levels lethargy/social withdrawl of irritability (p < 0.001), stereotype behavior (p < 0.001) Minocycline Ghaleiha et al. (2016) Autistic disorder (DSM- Yes Placebo (n = 23) 10 Risperidone: 0.5–2 mg/ ABC Minocycline-specific [67] IV-TR) Minocycline day improvement in: ADI-R (n = 23) Age range: Minocycline: 50 mg/day ABC: irritability (p = 0.003), Moderate-to-high levels 4-12 years hyperactivity/non-compli- of irritability ance (p = 0.002) N-Acetylcysteine 219 Table 1  (continued) 220 Agent Diagnosis (diagnostic Medication-free N Interven- Intervention (dose) Outcome Outcomes criteria applied) Age range tion period measures Structured/semi- (weeks) structured diagnostic interview Hardan et al. (2012) Autistic disorder (DSM- Yes Placebo (n = 15) 12 NAC: 900–2700 mg/day ABC NAC-specific improvement [75] IV-TR) NAC (n = 14) CGI-I in: ADI-R, ADOS Age range: 3–11 years CGI-S ABC: irritability (p < 0.001), CGI-S ≥4 SRS stereotyped/repeti- RBS-R tive behavior (p = 0.096) RBS: stereotypies (p = 0.014) SRS: social cognition (p = 0.037), mannerisms (p = 0.045) Ghanizadeh and Autistic disorder (DSM- Yes Placebo (n = 14) 8 NAC: 1200 mg/day ABC NAC-specific improvement Moghimi-Sarani IV-TR) ADI-R NAC (n = 17) in: (2013) [76] Age range: 3–16 years ABC: irritability (p = 0.035) Nikoo et al. (2015) [77] Autistic disorder (DSM- Yes Placebo (n = 20) 10 Risperidone: 0.5–2 mg/ ABC NAC-specific improvement IV-TR) NAC (n = 20) day in. ADI-R Age range: 4–12 years NAC: 600–900 mg/day ABC: irritability (p = 0.01), ABC-I ≥ 12 hyperactivity (p = 0.01) Wink et al. (2016) [78] ASD (autistic disorder, No Placebo (n = 12) 12 NAC: 60 mg/kg/day ABC No placebo- or NAC-specific Asperger’s disorder, NAC (n = 16) SRS improvement in any of PDD-NOS) Age range: 3–12 years VABS-II the primary or secondary (DSM-IV) CGI-I outcome measures ADI-R CGI-S (moderately ill) Dean et al. (2017) [79] Autistic disorder (DSM- No Placebo (n = 35) 24 NAC: 500 mg/day SRS No placebo- or NAC-specific IV-TR) NAC (n = 36) CCC​ improvement in any of Age range: 3–9 years RBS-R the primary or secondary DBC-P outcome measures PGI-I CGI-I CGI-S Sulforaphane G. Arteaga‑Henríquez et al. Table 1  (continued) Agent Diagnosis (diagnostic Medication-free N Interven- Intervention (dose) Outcome Outcomes criteria applied) Age range tion period measures Structured/semi- (weeks) structured diagnostic interview Singh et al. (2014) [84] Autistic disorder (DSM- No Placebo (n = 14) 18 SFN: 50–150 μM/day ABC SFN-specific improvement IV) SFN (n = 26) SRS in: ADOS Age range: 13–27 years CGI-I ABC: total score Moderate-to-severe (p < 0.001), irritability, lethargy, stereotypies, hyperactivity SRS: total score (p = 0.017), awareness, communication, motivation, mannerisms CGI-I: social interaction (p = 0.007), aberrant/ abnormal behavior (p = 0.014), verbal com- munication (p = 0.015) Momtazmanesh et al. ASD (DSM-V) Yes Placebo (n = 30) 10 Risperidone: 0.25–3.5 ABC SFN-specific improvement (2020) [85] ADI-R SFN (n = 30) mg/day in: Age range: 4–12 years SFN: 50–100 μM/day ABC: irritability (p < 0.001), hyperactivity (p = 0.015) Immunoregulatory and/or Anti-inflammatory Agents in Autism Spectrum Disorder Zimmerman et al. ASD (DSM-V) No Placebo (n = 23) 15 SFN: tablets contain- OASIS SFN-specific improvement (2021) [86] ADOS-2 SFN (n = 22) ing active myrosinase ABC in: VABS-II Age range: 3–12 years + broccoli seed and SRS-2 ABC: total score (p = 0.02), Moderate-to-severe sprout extracts (≅ 15 hyperactivity (p = 0.04), μmol/L SFN) inappropriate speech (p = 0.01) Omega-3 fatty acids Bent et al. (2011) [92] ASD (DSM-IV-TR) No Placebo (n = 12) 12 Omega-3: 350 mg EPA ABC No placebo- or intervention- ADOS Omega-3 (n = 13) + 230 mg DHA BASC specific effect in any of SCQ Age range: 5–17 years Placebo: safflower oil CGI-I the primary or secondary Moderate-severe Twice daily SRS outcome measures Peabody Pic- ture Vocabu- lary test Expressive Vocabulary test 221 Table 1  (continued) 222 Agent Diagnosis (diagnostic Medication-free N Interven- Intervention (dose) Outcome Outcomes criteria applied) Age range tion period measures Structured/semi- (weeks) structured diagnostic interview Bent et al. (2014) [93] ASD (NG) No Placebo (n = 28) 6 Omega-3: 350 mg EPA ABC Omega-3-specific improve- ADOS, ADI-R Omega-3 (n = 29) + 230 mg DHA SRS ment in: SCQ > 12 Age range: 5–8 years Placebo: safflower oil CGI-I ABC (parent rated): stereo- ABC-H ≥ 20 Twice daily typies (p = 0.05), lethargy (p = 0.01) Mankad et al. (2015) ASD (DSM-IV-TR) No Placebo (n = 19) 24 Omega-3: 0.75–1.5 g PDDBI Omega-3-specific worsening [94] ADOS Omega-3 (n = 18) EPA+DHA/day BASC-2 in externalizing behaviors ADI-R Age range: 2–5 years Placebo: refined olive VABS-II (p = 0.02) oil and medium chain PLS-4 TG/day CGI-I Mazahery et al. (2019) ASD (DSM-5) No Placebo (n = 16) 12 months Omega-3: 722 mg ABC Omega-3-specific improve- [95] Omega-3 (n = 23) DHA/day SRS-2 ment in: Age range: 2.5–8 years SPM ABC: irritability (p = 0.001), BPFAS lethargy (p = 0.02) Doaei et al. (2021) [96] Autistic disorder (DSM- Yes Placebo (n = 26) 8 Omega-3: 540 mg EPA GARS-2 Omega-3-specific improve- IV-TR) Omega-3 (n = 28) + 360 mg DHA/day ment in: ADOS Age range: 5–15 years Placebo: medium chain GARS-2: total score Moderate-to-severe TG/day (p = 0.001), stereotyped behaviors (p = 0.02), social communication (p = 0.02) ABC Aberrant Behavior Checklist, ABC-H ABC-Hyperactivity subscale, ABC-I ABC-Irritability subscale, ABFW Child Language Test in Phonology, Vocabulary, Fluency, and Pragmatics, ADI-R Autism Diagnostic Interview-Revised, ADL Language Development Assessment, ADOS Autism Diagnostic Observation Schedule, ASD autism spectrum disorder, BASC Behavior Assessment System for Children, BPFAS Behavioral Pediatrics Feeding Assessment Scale, CARS Childhood Autism Rating Scale, CCC​ Children’s Communication Checklist, CGI-I Clinical Global Impression-Improvement, CGI-S Clinical Global Impression-Severity, DBC-P Development Behavior Checklist, DHA docosahexaenoic acid, DSM Diagnostic and Statistical Manual of Mental Disorders, DSM-IV DSM Fourth Edition, DSM-IV-TR DSM Fourth Edition Text Revision, EPA eicosapentaenoic acid, GARS Gilliam Autism Rating Scale, NAC N-acetylcysteine, NG not given, OASIS Overall Anxiety Severity and Impairment Scale, PDDBI Pervasive Developmental Disorder Behavior Inventory, PDD-NOS Pervasive Developmental Disorder Not Otherwise Specified, PGI-I Patient Global Impression of Improvement Scale, PLS-4 Preschool Language Scale, RBS-R Repetitive Behavior Scale-Revised, SCQ Social Communication Questionnaire, SFN sulforaphane, SPM Sensory Processing Measure questionnaire, SRS Social Responsiveness Scale, TG triglycerides, VABS-II Vineland Adaptive Behavior Scale, Second Edition. *In the study performed by Rocha-Brito and colleagues (2021) [53], patients were treated, before the starting of the trial, with oral albendazole for five days. G. Arteaga‑Henríquez et al. Immunoregulatory and/or Anti-inflammatory Agents in Autism Spectrum Disorder 223 Table 2  Quality assessment of included studies Study agent Quality score Interpretation Malek et al. (2020) [52] Prednisolone **** High quality Rocha-Brito et al. (2021) [53] ***** High quality Ayatollahi et al. (2020) [59] Pregnenolone ***** High quality Asadabadi et al. (2013) [62] Celecoxib ***** High quality Ghaleiha et al. (2016) [67] Minocycline ***** High quality Hardan et al. (2012) [75] N-Acetylcysteine ***** High quality Ghanizadeh and Moghimi-Sarani (2013) [76] ***** High quality Nikoo et al. (2015) [77] ***** High quality Wink et al. (2016) [78] ***** High quality Dean et al. (2017) [79] ***** High quality Singh et al. (2014) [84] Sulforaphane **** High quality Momtazmanesh et al., (2020) [85] ***** High quality Zimmerman et al. (2021) [86] **** High quality Bent et al. (2011) [92] Omega-3 ***** High quality Bent et al. (2014) [93] ***** High quality Mankad et al. (2015) [94] ***** High quality Mazahery et al. (2019) [95] ***** High quality Doaei et al. (2021) [96] ***** High quality type of adverse effects, which included changes in appetite, of the intervention phase at week 10, a significantly higher abdominal pain, dizziness, insomnia, nausea, and/or seda- improvement in irritability and hyperactivity was found in tion. A possible mechanism, by which celecoxib improves the subgroup of patients allocated to (add-on) minocycline these symptoms in individuals with autistic disorder is, when compared to those allocated to (add-on) placebo. An again, the inhibition of microglial/monocyte [63] and/or intervention-specific effect on the other symptoms assessed Th17 cells activation [64]. Interestingly, several reports have was however, not found (Table 1). Minocycline had a good suggested that different proinflammatory cytokines, such as safety profile and was well tolerated, adverse events were IL-1β, IL-6, and/or IL-17A, could serve as indicators for mild to moderate and included diarrhea, headache, increase predicting clinical response to celecoxib in individuals with in appetite, dizziness, insomnia, nausea, and/or sedation. immune-mediated conditions, such as ankylosing spondylitis Both study groups did not statistically differ in relation to [65]. the frequency, severity, or type of adverse effects. Again, studies have found that minocycline is able to inhibit micro- 3.1.4 Minocycline glial/monocyte proinflammatory activation [68–70] and to regulate the Th17/Treg cells axis, decreasing the levels of Minocycline is a second-generation tetracycline antibiotic several proinflammatory cytokines such as IL-1β, IL-6, with well-known antioxidant, immunoregulatory, and/or TNFα, IFNγ, and/or IL-17A, both in the brain and in the anti-inflammatory properties [66]. To date, only one rand- periphery [71]. omized, double-blind, placebo-controlled trial has assessed the effectiveness of minocycline on improving core and 3.1.5  N‑Acetylcysteine associated symptoms in individuals with ASD [67]. In this study, a total of 46 medication-naïve children diagnosed with NAC is a synthetic N-acetyl derivative of the endogenous autistic disorder and with moderate-to-high levels of irrita- amino acid L-cysteine [72], which acts as a precursor of bility were randomly allocated to either risperidone + pla- the antioxidant enzyme glutathione (i.e., the most abundant cebo (n = 23), or to risperidone + minocycline (n = 23), and antioxidant in the brain) [73]. An increasing body of evi- followed up over a 10-week period (Table 1). The primary dence suggests that NAC may also act as an immunoregula- outcome measure was the change, from baseline to week tory and/or anti-inflammatory agent [74]. Therefore, several 10, in the ABC irritability subscale score, while second- randomized, placebo-controlled clinical trials have inves- ary outcome measures included the change in other ABC tigated the effectiveness of (add-on) treatment with NAC subscale scores (i.e., lethargy/social withdrawal, stereotypic on improving core and associated symptoms in individuals behavior, hyperactivity, inappropriate speech). At the end with ASD (Table 1). In general, studies suggest that (add-on) 224 G. Arteaga‑Henríquez et al. treatment with NAC may be beneficial for the management immunoregulatory, and/or anti-inflammatory properties of irritability [75–77] in,at least, a subgroup of children [83]. To date, three randomized, placebo-controlled trials and adolescents with autistic disorder (Table 1). In all these have assessed the effectiveness of this agent on improv- studies, children and adolescents were medication-naïve, ing core and/or related symptoms in individuals with ASD and the diagnosis of autistic disorder was previously con- (Table 1). Overall, studies suggest that (add-on) treatment firmed by a (semi)-structured interview (i.e., the Autism with SFN may improve hyperactivity in children, adoles- Diagnostic Interview-Revised [ADI-R], and/or the Autism cents, and adults with autistic disorder and/or ASD [84, Diagnostic Observation Schedule [ADOS]). The improve- 85]. In all these studies, the improvement in hyperactivity ment in irritability was assessed by the change, from base- was assessed by the change from baseline until week 10 line until week 8 [76], 10 [77], or 12 [75] in the respective [84], 1586 , or 18 [85] in the respective ABC subscale score ABC subscale score. NAC was administered at a dose range (Table 1). The dose of SFN ranged between 50 and 150 μM/ of 600–2700 mg/day (Table 1). In only one of four stud- day; one studyused glucoraphanin-rich broccoli seed extract ies assessing irritability as an outcome measure, (add-on) tablets containing myrosinase, instead of SFN [86] (Table 1). treatment with NAC was not associated with a significantly Mixed findings were found for other symptoms assessed, higher improvement in this symptom, compared with pla- such as irritability, lethargy, stereotyped/repetitive behavior cebo (Table 1) [78]. In this study, children were not med- (i.e., ABC subscale scores), mannerisms, awareness, motiva- ication-naïve, and were diagnosed with ASD (i.e., autistic tion, social communication (i.e., SRS subscale scores), and disorder, Asperger’s disorder, and/or pervasive developmen- social interaction, aberrant/abnormal behavior, and/or verbal tal disorder not otherwise specified [PDD-NOS]) (Table 1), communication (CGI-I subscale scores) (Table 1). Differ- something which could have influenced the findings. Mixed ences in the duration of the treatment period, in the question- findings were found for other symptoms assessed, such as naires/scales used for assessing symptoms, and in the dosage hyperactivity (i.e., ABC subscale score), stereotyped/repeti- and/or composition of the study agent could have influenced tive behavior (i.e., ABC and/or Repetitive Behavior Scale the findings. SFN had a good safety profile, and was well [RBS] subscale scores), mannerisms (i.e., Social Respon- tolerated; adverse effects included abdominal pain, increased siveness Scale [SRS] subscale score), and/or social cogni- flatulence, constipation, diarrhea, vomiting, increased appe- tion (SRS subscale score) (Table 1) [75–79]. Differences in tite, weight gain, headache, irritability, increased aggres- the duration of the treatment period, in the questionnaires/ sion, dizziness, sedation, insomnia, rashes, exacerbation of scales used for assessing symptoms, and/or in the dosage seasonal allergies, and/or fever. Significant differences in of the study agent could have also influenced results. In all relation to the frequency and/or severity of adverse effects studies, NAC had a good safety profile and was well toler- were not found between individuals allocated to (add-on) ated; significant differences in relation to the frequency and/ placebo, and those allocated to (add-on) SFN. The mecha- or severity of adverse effects were not found between indi- nism of action by which SFN improves these symptoms in viduals treated with (add-on) placebo, and those treated with individuals with ASD is not known at all. SFN may exert (add-on) NAC. Adverse effects were mild to moderate and an anti-inflammatory effect on microglia [87]. Interestingly, included gastrointestinal symptoms (e.g., abdominal pain, in a study performed on Black and Tan Brachyury (BTBR) diarrhea and/or constipation, changes in appetite, nausea), T+ Itpr3tf/J mice (i.e., a strain of mouse model that is most headache, rash, insomnia, and/or fatigue. The mechanism noted for its phenotypic similarities to humans on the ASD of action by which NAC improves irritability in individu- scale), SFN was able to ameliorate autism-like behaviors als with autistic disorder is not fully understood. NAC can (i.e., reduced self-grooming/marble burying behavior, reverse microglial proinflammatory activation [80], and increased social interaction) through suppression of Th17- several studies have suggested that NAC could also regu- related signaling both in the periphery, and in the brain (i.e., late the Treg/Th17 axis in individuals with inflammatory SFN-treated BTBR mice were characterized by a reduced conditions, such as chronic obstructive pulmonary disease expression of STAT3, RORC, IL-17A, and/or IL-23R in (COPD) [81]. In addition, animal models of experimental ­CD4+ Th cells) [88]. In another study performed on children autoimmune encephalomyelitis have also demonstrated a with ASD, SFN was associated with nuclear factor erythroid suppressive action of NAC on Th17 cells [82]. 2-related factor 2 (Nrf2) stimulation, resulting in an inhibi- tory effect on nitrative stress markers and pro-inflammatory 3.2 Dietary Interventions cytokines [89]. 3.2.1 Sulforaphane 3.2.2 Omega‑3 Fatty Acids SFN is an isothiocyanate derived from Brassica veg- Omega-3 polyunsaturated fatty acids (PUFAs) include etables (in particular, from broccoli) with antioxidant, α-linolenic acid (ALA), stearidonic acid (SDA), Immunoregulatory and/or Anti-inflammatory Agents in Autism Spectrum Disorder 225 eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA), (add-on) placebo, and those allocated to (add-on) omega-3 and docosahexaenoic acid (DHA). Due to their well-known fatty acids. The mechanism of action by which omega-3 fatty anti-inflammatory actions [90, 91], most trials assessing acids exert their action is not fully understood. Again, evi- dietary interventions in ASD have used omega-3 PUFAs as dence suggests that omega-3 fatty acids exert their action the study agent, but with mixed findings (Table 1). (Add- by (at least in part) increasing the expression of FoxP3 and on) treatment with omega-3 fatty acids resulted beneficial the differentiation of Tregs, while inhibiting Th17 promo- for the management of lethargy in two of three studies that tion, and reducing IL-17A production [97]. In addition, these used the change in the ABC respective subscale score as the compounds have also been found to reverse microglial pro- outcome measure [93, 95]. In both studies, study participants inflammatory activation [98]. were children (age range 2.5–8 years) diagnosed with an ASD, and omega-3 supplements consisted of DHA (with or without EPA). The daily dose of DHA was ≅ 460–722 mg/ 4 Discussion day, and the daily dose of EPA was ≅ 700 mg/day (Table 1). In the remaining study [92], study participants were chil- Overall, our findings suggest a beneficial effect of (add-on) dren and adolescents diagnosed with a moderate-to-severe treatment with prednisolone, pregnenolone, celecoxib, mino- ASD, and the sample size was lower, something which could cycline, and/or NAC on irritability; a beneficial effect of have influenced the findings (Table 1). (Add-on) treatment (add-on) treatment with prednisolone, pregnenolone, mino- with omega-3 was also beneficial for the management of cycline, and/or SFN on hyperactivity; a beneficial effect stereotyped behaviors in two of four studies assessing this of (add-on) treatment with prednisolone, celecoxib, and/ symptom (by the change in the ABC and/or in the GARS-2 or omega-3 fatty acids on lethargy, and a beneficial effect respective subscale scores) (Table 1) [93, 96]. Study partici- of (add-on) treatment with prednisolone, pregnenolone, pants were children and adolescents (age range 5–15 years) celecoxib, and/or omega-3 fatty acids on stereotyped behav- diagnosed with ASD or with autistic disorder. In both stud- ior, over placebo. All agents had a good safety profile and ies, the intervention product was composed of DHA (dose were well tolerated; significant differences in relation to the range 360–460 mg/day) + EPA (dose range 540–700 mg/ frequency and/or severity of adverse effects were not found day) (Table 1). Mixed findings were found for other symp- between individuals treated with placebo and those treated toms assessed, such as irritability (i.e., ABC subscale score), with the different above-mentioned compounds. The mech- and/or social communication (i.e., GARS-2 subscale score) anism, by which all these agents exert their action is not (Table 1). Differences in the duration of the treatment fully understood. Interestingly, all agents are able to inhibit period, in the scales used for assessing symptoms, and/or microglial proinflammatory activation, and to restore Treg/ in the dosage or composition of the study agent could have Th17 imbalances in at least,a subgroup of individuals with influenced results. Interestingly, in a study performed on ASD, decreasing the levels of proinflammatory cytokines children with ASD [94], (add-on) treatment with 750 mg such as IL-6 and/or IL-17A in both the brain, and the periph- EPA + 1500 mg DHA/day was associated with a worsen- ery. Although encouraging, findings should be considered in ing in externalizing behaviors (BASC-2) at week 24, when light of several limitations. The majority of studies included compared with placebo. In this study, all participants were children and adolescents as participants and therefore, find- younger than 5 years of age, and omega-3 was adminis- ings may not be applicable to adults with ASD. Moreo- tered at higher doses in comparison with the other studies, ver, the sample size was low in the majority of studies, and makingpossible that omega-3 fatty acidsworsens external- most studies excluded non-verbal participants with severe izing behaviors in this age group, when given at high doses. intellectual disabilities, and included study participants Moreover, the majority of study participants were minimally with different diagnoses (i.e., ASD vs. autistic disorder vs. verbal and therefore, potential gastrointestinal distress may PDD-NOS vs. Asperger’s syndrome). Therefore, findings have been captured as reports of externalizing behaviors. may only be applicable to a small subgroup of individu- Omega-3 fatty acids had a good safety profile and were well als living with this condition. The severity of baseline ASD tolerated, no serious adverse events were reported during the symptoms was inconsistent among trials, and the use of study. Most adverse events reported were mild to moderate concomitant psychiatric medications and/or of behavioral and included neuropsychiatric symptoms (e.g., decreased treatments was allowed in several studies, something which energy, headache), sleep disturbances (e.g., insomnia, early could have influenced the findings. Differences in the active awakening), nutritional or gastrointestinal symptoms (e.g., treatment composition and/or in the dosage used, as well as changes in appetite, abdominal pain), dermatological, or in the duration of the treatment period, could also have influ- others, such as eye swelling. No statistically significant dif- enced the results. Unfortunately, baseline levels of biologi- ferences in relation to the frequency and characteristics of cal (immune/inflammatory) markers were not assessed in adverse events were found between patients allocated to the majority of studies. This would have been interesting in 226 G. Arteaga‑Henríquez et al. order to identify (immune/inflammatory) predictors of treat- Code availability Not applicable. ment response and thus, to also identify those individuals Author contributions GAH: conceptualization, data curation, investi- who would benefit from a particular treatment regimen, at gation, methodology, validation, visualization, writing—original draft, baseline (personalized medicine). review, and editing. LG: data curation, investigation, methodology, validation, visualization: review, and editing. JARQ: writing—review, and editing. All authors read and approved the final submitted manu- script and agree to be accountable for this work. 5 Conclusions and Future Perspectives Open Access This article is licensed under a Creative Commons Attri- The evidence that immune dysfunction may play a role in bution-NonCommercial 4.0 International License, which permits any the pathophysiology of at least a subgroup of individuals non-commercial use, sharing, adaptation, distribution and reproduction with ASD is now considerable. An increasing body of evi- in any medium or format, as long as you give appropriate credit to the dence has supported the existence of a chronic activation of original author(s) and the source, provide a link to the Creative Com- mons licence, and indicate if changes were made. The images or other not only microglia, but also of the monocyte/macrophage third party material in this article are included in the article's Creative system in individuals with ASD, across the lifespan. In addi- Commons licence, unless indicated otherwise in a credit line to the tion, abnormal levels of several lymphocyte subpopulations material. If material is not included in the article's Creative Commons (e.g., Tregs, Th17 cells) have been repeatedly demonstrated. licence and your intended use is not permitted by statutory regula- tion or exceeds the permitted use, you will need to obtain permission Therefore, immunoregulatory and/or anti-inflammatory directly from the copyright holder. To view a copy of this licence, visit agents may represent a possibility for more personalized http://​creat​iveco​mmons.​org/​licen​ses/​by-​nc/4.​0/. treatment regimens among individuals diagnosed with this condition. Several randomized, placebo-controlled trials on the effectiveness of (add-on) treatment with prednisolone, References pregnenolone, celecoxib, minocycline, NAC, SFN, and/or omega-3 fatty acids on core and related symptoms of ASD 1. American Psychiatric Association. Diagnostic and statistical man- have been performed, with promising findings. However, ual of mental disorders (5th ed). American Psychiatric Associa- tion; 2013. https://​doi.​org/​10.​1176/​appi.​books.​97808​90425​596. larger randomized, placebo-controlled trials including, for 2. Baxter AJ, Brugha TS, Erskine HE, et al. The epidemiology example, more homogeneous study populations, and longer and global burden of autism spectrum disorders. Pscyhol Med. periods of follow-up are urgently needed to confirm the 2015;45(3):601–13. https://​doi.​org/​10.​1017/​S0033​29171​40017​ findings. 2X. 3. Baio J, Wiggins L, Christensen DL, et al. Prevalence of autism spectrum disorder among children aged 8 years-autism and devel- Declarations opmental disabilities monitoring network, 11 sites, United States, 2014. MMWR Surveill Summ. 2018;67(6):1–23. https://​doi.​org/​ Funding No funding was secured for this study. Open access publica- 10.​15585/​mmwr.​ss670​6a1. tion was supported by the Group of Psychiatry, Mental Health and 4. Strathearn L. The elusive etiology of autism: nature and nurture? Addiction of Vall d'Hebron Research Institute, Barcelona, Catalonia, Front Behav Neurosci. 2009;3:11. https://​doi.​org/​10.​3389/​neuro.​ Spain. 08.​011.​2009. 5. DeFilippis M, Wagner KD. Treatment of autism spectrum dis- Conflict of interest All authors have no financial relationships to dis- order in children and adolescents. Psychopharmacol Bull. close relevant to this article. Josep Antoni Ramos-Quiroga was on 2016;46(2):18–41. the speakers’ bureau and/or acted as a consultant for Janssen-Cilag, 6. Pandina G, Ring RH, Bangerter A, et al. Current approaches to Novartis, Shire, Takeda, Bial, Shionogi, Sincrolab, Novartis, BMS, the pharmacologic treatment of core symptoms across the lifespan Medice, Rubió, Uriach, Technofarma and Raffo in the last 3 years, and of autism spectrum disorder. Child Adolesc Psychiatr Clin N Am. also received travel awards (air tickets + hotel) from Janssen-Cilag, 2020;29(2):301–17. https://​doi.​org/​10.​1016/j.​chc.​2019.​12.​004. Rubió, Shire, Takeda, Shionogi, Bial and Medice for taking part in 7. Harville T, Rhodes-Clark B, Bennuri SC, et al. Inheritance of psychiatric meetings. The Department of Psychiatry chaired by Josep HLA-Cw7 associated with autism spectrum disorder (ASD). Antoni-Ramos-Quiroga received unrestricted educational and research Front Psychiatry. 2019;10:612. https://d​ oi.o​ rg/1​ 0.3​ 389/f​ psyt.2​ 019.​ support from Janssen-Cilag, Shire, Oryzon, Roche, Psious, and Rubió 00612. in the last 3 years. 8. Lyall K, Ashwood P, van de Water J, et al. Maternal immune- mediated conditions, autism spectrum disorders, and developmen- Availability of data and material Data sharing was not applicable to tal delay. J Autism Dev Disord. 2014;44(7):1546–55. https://​doi.​ this article as no datasets were generated and/or analyzed during the org/​10.​1007/​s10803-​013-​2017-2. current study. 9. Lombardo MV, Moon HM, Su J, et al. Maternal immune activa- tion dysregulation of the fetal brain transcriptome and relevance to Consent to participate Not applicable. the pathophysiology of autism spectrum disorder. Mol Psychiatry. 2018;23(4):1001–10013. https://​doi.​org/​10.​1038/​mp.​2017.​15. Ethics approval Not applicable. 10. Ashwood P, Wills S, van de Water J. The immune response in autism: a new frontier for autism research. J Leukoc Biol. Consent for publication Not applicable. 2006;80(1):1–15https://​doi.​org/​10.​1189/​jlb.​12057​07. Immunoregulatory and/or Anti-inflammatory Agents in Autism Spectrum Disorder 227 11. Hughes HK, Ko EM, Rose D, et al. Immune dysfunction and 27. Rodriguez JI, Kern JK. Evidence of microglial activation in autism autoimmunity as pathological mechanisms in autism spectrum and its possible role in brain underconnectivity. Neuron Glia Biol. disorders. Front Cell Neurosci. 2018;12:405. https://​doi.​org/​10.​ 2011;7(2–4):205–13. https://​doi.​org/​10.​1017/​S1740​925X1​20001​ 3389/​fncel.​2018.​00405. 42. 12. Tromans S, Yao G, Alexander R, et al. The prevalence of diabe- 28. Ahmad SF, Ansari MA, Nadeem A, et al. Involvement of CD45 tes in autistic persons: a systematic review. Clin Pract Epidemiol cells in the development of autism spectrum disorder through dys- Ment Health. 2020;16:212–25. https://​doi.​org/​10.​2174/​17450​ regulation of granulocyte-macrophage colony-stimulating factor, 17902​01601​0212. key inflammatory cytokines, and transcription factors. Int Immu- 13. Zerbo O, Leong A, Barcellos L, et al. Immune mediated con- nopharmacol. 2020;83:106466. https://​doi.​org/​10.​1016/j.​intimp.​ ditions in autism spectrum disorders. Brain Behav Immun. 2020.​106466. 2015;46:232–6. https://​doi.​org/​10.​1016/j.​bbi.​2015.​02.​001. 29. Ashwood P, Corbett BA, Kantor A, et al. In search of cellular 14. Muskens JB, Verlders FP, Staal WG. Medical comorbidities in immunophenotypes in the blood of children with autism. PLoS children and adolescents with autism spectrum disorders and ONE. 2011;6:e19299. https://d​ oi.o​ rg/1​ 0.1​ 371/j​ ourna​ l.p​ one.0​ 0192​ attention deficit hyperactivity disorders: a systematic review. Eur 99. Child Adolesc Psychiatry. 2017;26(9):1093–103. https://​doi.​org/​ 30. Tonhajzerova I, Ondrejka I, Mestanik M, et al. Inflamma- 10.​1007/​s00787-​017-​1020-0. tory activity in autism spectrum disorder. Adv Exp Med Biol. 15. Kotey S, Ertel K, Whitcomb B. Co-occurrence of autism and 2015;861:93–8. asthma in a nationally-representative sample of children in the 31. Zhao HX, Yin SS, Fan YG. High plasma neopterin levels in Chi- United States. J Autism Dev Disord. 2014;44(12):3083–8. https://​ nese children with autism spectrum disorders. Int J Dev Neurosci. doi.​org/​10.​1007/​s10803-​014-​2174-y. 2015;41:92–7. 16. Estes ML, McAllister AK. Immune mediators in the brain and 32. Pardo CA, Farmer CA, Thurm A, et al. Serum and cerebrospinal peripheral tissues in autism spectrum disorder. Nat Rev Neursoci. fluid immune mediators in children with autistic disorder: a lon- 2015;16(8):469–86. https://​doi.​org/​10.​1038/​nrn39​78. gitudinal study. Mol Autism. 2017;8:1. 17. Paolicelli RC, Bolasco G, Pagani F, et al. Synaptic pruning by 33. Kutlu A, Cevher BN. Does increased neutrophil-lymphocyte microglia is necessary for normal brain development. Science. ratio predict autism spectrum disorder? Anadolu Psikiyatri Derg. 2011;333(6048):1456–8. https://​doi.​org/​10.​1126/​scien​ce.​12025​ 2018;19:607–14. 29. 34. Hesapcioglu ST, Kasak M, Kurt ANC, Ceylan MF. High mono- 18. Schafer DP, Lehrman EK, Kautzman AG, et al. Microglia sculpt cyte level and low lymphocyte to monocyte ratio in autism spec- postnatal neural circuits in an activity and complement-dependent trum disorders. Int J Dev Disabil. 2019;65:73–81. manner. Neuron. 2012;74:691–705. https://d​ oi.o​ rg/1​ 0.1​ 016/j.n​ eu- 35. Ceylan MF, Hesapcioglu ST, Yavas CP, et al. Serum ischemia- ron.​2012.​03.​026. modified albumin levels, myeloperoxidase activity and peripheral 19. Coomey R, Stowell R, Majewska A, et al. The role of microglia blood mononuclear cells in autism spectrum disorder (ASD). J in neurodevelopmental disorders and their therapeutics. Curr Top Autism Dev Disord. 2021;51(7):2511–7. Med Chem. 2020;20(4):272–6. https://​doi.​org/​10.​2174/​15680​ 36. Enstrom AM, Onore CE, Van de Water JA, et al. Differential 26620​66620​02211​72619. monocyte responses to TLR ligands in children with autism spec- 20. McIntyre LL, Greilach SA, Othy S, et al. Regulatory T cells trum disorders. Brain Behav Immun. 2010;24(1):64–71. https://​ promote remyelination in the murine experimental autoimmune doi.​org/​10.​1016/j.​bbi.​2009.​08.​001. encephalomyelitis model of multiple sclerosis following human 37. Torres AR, Westover JB, Rosenspire AJ. HLA immune function neural stem cell transplant. Neurobiol Dis. 2020;140:104868. genes in autism. Autism Res Treat. 2012;2012:959073. https://​ https://​doi.​org/​10.​1016/j.​nbd.​2020.​14868. doi.​org/​10.​1155/​2012/​959073. 21. Larochelle C, Wasser B, Jamann H, et al. Pro-inflammatory T 38. DiStasio MM, Nagakura I, Nadler M, et al. T lymphocytes and helper 17 directly harms oligodendrocytes in neuroinflammation. cytotoxic astrocyte blebs correlate acorss autism brains. Ann Neu- Proc Natl Acad Sci USA. 2021;118(34):e2025813118. https://d​ oi.​ rol. 2019;86(6):885–989. https://​doi.​org/​10.​1002/​ana.​25610. org/​10.​1071/​pnas.​20258​13118. 39. Wasilewska J, Kaczmarski M, Stasiak-Barmuta A, et al. Low 22. Luo Y, Chen C, Wei C, et al. BDNF alleviates microglial inhibi- serum IgA and increased expression of CD23 on B lymphocytes tion and stereotypic behaviors in a mouse model of obsessive- in peripheral blood in children with regressive autism aged 3–6 compulsive disorder. Front Mol Neurosci. 2022;15:926572. years old. Arch Med Sci. 2012;8:324–31. https://​doi.​org/​10.​3389/​fnmol.​2022.​926572. 40. Siniscalco D, Mijatovic T, Bosmans E, et al. Decreased num- 23. Al-Ayadhi LA, Mostafa GA. Elevated serum levels of interleukin- bers of CD57+CD3− cells identify potential innate immune 17A in children with autism. J Neuroinflamm. 2012;9:158. https://​ differences in patients with autism spectrum disorder. In Vivo. doi.​org/​10.​1186/​1742-​2094-9-​158. 2016;30:83–9. 24. Moaaz M, Youssry S, Elfatatry A, et al. Th17/Treg cells imbalance 41. De Giacomo A, Gargano CD, Simone M, et al. B and T immu- and their related cytokines (IL-17, IL-10 and TGF-β) in children noregulation: a new insight of B regulatory lymphocytes in autism with autism spectrum disorder. J Neuroimmnol. 2019;337:577071. spectrum disorder. Front Neurosci. 2021;15: 732611. https://​doi.​org/​10.​1016/j.​jneur​oim20​19.​577071. 42. Ellul P, Rosenzwajg M, Peyre H, et al. Regulatory T lymphocytes/ 25. Arteaga-Henríquez G, Lugo-Marín J, Gisbert L, et al. Abnormal Th17 lymphocytes imbalance in autism spectrum disorders: evi- blood levels of lymphocyte subpopulations in individuals with dence from a meta-analysis. Mol Autism. 2021;12(1):68. https://​ autism spectrum disorder: a systematic review and meta-analysis. doi.​org/​10.​1186/​s13229-​021-​00472-4. Int J Mol Sci. 2022;23(22):14329. https://​doi.​org/​10.​3390/​ijms2​ 43. Ahmad SF, Zoheir KMA, Ansari MA, et al. Dysregulation of Th1, 32214​329. Th2, Th17, and T regulatory cell-related transcription factor sign- 26. Morgan JT, Chana G, Pardo CA, et al. Microglial activation and aling in children with autism. Mol Neurobiol. 2017;54(6):4390– increased microglial density observed in the dorsolateral prefron- 400. https://​doi.​org/​10.​1007/​s12035-​016-​9977-0. tal cortex in autism. Biol Psychiatry. 2010;68(4):368–76. https://​ 44. Masi A, Quintana DS, Glozier N, et al. Cytokine aberrations in doi.​org/​10.​1016/j.​biops​ych.​2010.​05.​024. autism spectrum disorder: a systematic review and meta-analysis. 228 G. Arteaga‑Henríquez et al. Mol Psychiatry. 2015;20(4):440–6. https://​doi.​org/​10.​10138/​mp.​ autoimmunity by a neurosteroid. J Immunol. 2016;197(7):2598– 2014.​59. 609. https://​doi.​org/​10.​4049/​jimmu​nol.​16010​38. 45. Saghazadeh A, Ataeinia B, Keynejad K, et al. A meta-analysis of 61. Chan FKL, Ching JYL, Tse YK, et al. Gastrointestinal safety pro-inflammatory cytokines in autism spectrum disorders: effects of celecoxib versus naproxen in patients with cardiothrombotic of age, gender and latitude. J Psychiatr Res. 2019;115:90–102. diseases and arthritis after upper gastrointestinal bleeding 46. Saghazadeh A, Ataeinia B, Keynejad K, et al. Anti-inflammatory (CONCERN): an industry-independent, double-blind, double- cytokines in autism spectrum disorders: a systematic review and dummy, randomized trial. Lancet. 2017;389(10087):2375–82. meta-analysis. Cytokine. 2019;123: 154740. https://​doi.​org/​10.​1016/​S0140-​6736(17)​30981-9. 47. Ahmad SF, Ansari MA, Nadeem A, et al. Upregulation of inter- 62. Asadabadi M, Mohammadi MR, Ghanizadeh A, et al. Celecoxib leukin (IL)-31, a cytokine producing CXCR1 peripheral immune as adjunctive treatment to risperidone in children with autistic cels, contributes to the immune abnormalities of autism spectrum disorder: a randomized, double-blind, placebo-controlled trial. disorder. J Neuroimmunol. 2020;349:577430. https://​doi.​org/​10.​ Psychopharmacology. 2013;225(1):51–9. https://​d oi.​o rg/​1 0.​ 1016/j.​jneur​oim.​2020.​577430. 1007/​s00213-​012-​2796-8. 48. Nadeem A, Ahmad SF, Attia SM, et al. Oxidative and inflamma- 63. Villa V, Thellung S, Bajetto A, et al. novel celecoxib analogues tory mediators are upregulated in neutrophils of autistic children: inhibit glial production of prostaglandin E2, nitric oxide, and role of IL-17A receptor signaling. Prog Neuuropsychopharmacol oxygen radicals reverting the neutoinflammatory responses Biol Psychiatry. 2019;90:204–11. https://d​ oi.o​ rg/1​ 0.1​ 016/j.p​ npbp.​ induced by misfolded prion protein fragment 90–231 or lipopol- 2018.​12.​002. ysaccharide. Pharmacol Res. 2016;113(PtA):500–14. https://​ 49. Page MJ, McKenzie JE, Bossyt PM, et al. The PRISMA 2020 doi.​org/​10.​1016/j.​phrs.​2016.​09.​010. statement: an updated guideline for reporting systematic reviews. 64. Paulissen SMJ, van Hamburg JP, Davelaar N, et al. Synovial BMJ. 2021;372:n71. https://​doi.​org/​10.​1136/​bmj.​n71. fibroblasts induce Th17 pathogenicity via the cyclooxygenase/ 50. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of prostaglandin E2 pathway, independent of IL-23. J Immunol. reports of randomized clinical trial: is blinding necessary? Control 2013;91(3):1364–72. https://​doi.​org/​10.​4049/​jimmu​nol.​13002​ Clin Trials. 1996;17:1–12. 74. 51. Williams DM. Clinical pharmacology of corticosteroids. Respir 65. Zhang Y, Ning C, Zhou H, et al. Interleukin-1b, interleukin-6, Care. 2018;63(6):655–70. https://d​ oi.o​ rg/1​ 0.4​ 187/r​ espca​ re.0​ 6314. and interleukin-17A as indicators reflecting clinical response 52. Malek M, Ashraf-Ganjouei A, Moradi K, et al. Prednisolone as to celecoxib in ankylosing spondylitis patients. Ir J Med Sci. adjunctive treatment to risperidone in children with regressive 2021;190(2):631–8. https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 11845-0​ 20-0​ 2366-5. type of autism spectrum disorder: a randomized, placebo-con- 66. Singh S, Khanna D, Kalra S. Minocycline and doxycycline: more trolled trial. Clin Neurophamacol. 2020;43(2):39–45. than antibiotics. Curr Mol Pharmacol. 2021;14(6):1046–65. 53. Rocha-Brito A, Teixeira Vairo GP, Henriques Dias AP, et al. https://​doi.​org/​10.​2174/​18744​67214​66621​02101​22628. Effect of prednisolone on language fucntion in children with autis- 67. Ghaleiha A, Rasa SM, Nikoo M, et al. Minocycline as adjunctive tic spectrum disorder: a randomized clinical trial. J Pediatr (Rio treatment to risperidone in children with autistic disorder: a ran- J). 2021;97(1):22–9. https://​doi.​org/​10.​1016/j.​jped.​2019.​10.​012. domized, double-blind placebo-controlled trial. J Child Adolesc 54. Sugama S, Takenouchi T, Fujita M, et al. Corticosteroids limit Psychopharmacol. 2016;26(9):784–91. https://​doi.​org/​10.​1089/​ microglial activation occurring during acute stress. Neuroscience. cap.​2015.​0175. 2013;232:13–20. https://​doi.​org/​10.​1016/j.​neuro​scien​ce.​2012.​12.​ 68. Campbell JH, Burdo TH, Autissier P, et al. Minocycline inhibi- 012. tion of monocyte activation correlates with neuronal protection in 55. Mathian A, Jouenne R, Chader D, et al. Regulatory T cell SIV neuroAIDS. PLoS ONE. 2011;6(4):e18688. https://​doi.​org/​ responses to high-dose methylprednisolone in active systemic 10.​1371/​journ​al.​pone.​00186​88. lupus erythematosus. PLoS ONE. 2015;10(12):30143689. https://​ 69. Kobayashi K, Imagama S, Ohgomori T, et al. Minocycline selec- doi.​org/​10.​1371/​journ​al.​pone.​01436​89. tively inhibits M1 polarization of microglia. Cell Death Dis. 56. Fu XQ, Cai JY, Li MJ. Prednisone may rebuild the immunologic 2013;4(3):e525. https://​doi.​org/​10.​1038/​cddis.​2013.​54. homeostasis: alteration of Th17 and Treg cells in the lympho- 70. Han Y, Zhang L, Wang Q, et al. Minocycline inhibits microglial cytes from rats’ spleens after treated with prednisone-containing activation and alleviates depressive-like behaviors in male adoles- serum. Mol Genet Genom Med. 2019;7(7):e00800. https://​doi.​ cent mice subjected to maternal separation. Psychoneuroendocri- org/​10.​1002/​mgg3.​800. nology. 2019;107:37–45. https://d​ oi.o​ rg/1​ 0.1​ 016/j.p​ syneu​ en.2​ 019.​ 57. Eddy JL, Krukowski K, Janusek L, et al. Glucocorticoids regu- 04.​021. late natural killer cell function epigenetically. Cell Immunol. 71. Florou DT, Mavropoulos A, Dardiotis E, et al. Tetracyclines 2014;290(1):120–30. https://​doi.​org/​10.​1016/j.​celli​mm.​2014.​05.​ Dminish in vitro IFN-y and IL-17-producing adaptive and 013. innate immune cells in multiple sclerosis. Front Immunol. 58. Murugan S, Jakka P, Namani S, et al. The neurosteroid preg- 2021;12:739186. https://​doi.​org/​10.​3389/​fimmu.​2021.​739186. nenolone promotes degradation of key proteins in the innate 72. Aldini G, Altomare A, Baron G, et al. N-Acetylcysteine as an immune signaling to suppress inflammation. J Biol Chem. antioxidant and disulphide breaking agent: the reasons why. Free 2019;294(12):4596–607. https://​doi.​org/​10.​1074/​jbc.​RA118.​ Radic Res. 2018;52(7):751–62. https://d​ oi.o​ rg/1​ 0.1​ 080/1​ 07157​ 62.​ 005543. 2018.​14685​64. 59. Ayatollahi A, Bagheri S, Ashraf-Ganjouei A, et al. Does preg- 73. Wu G, Fang YZ, Yang S, et al. Glutathione metabolism and its nenolone adjunct to risperidone ameliorate irritable behavior implications for health. J Nutr. 2004;134(3):489–92. https://​doi.​ in adolescents with autism spectrum disorder: a randomized, org/​10.​1093/​jn/​134.3.​489. double-blind, placebo-controlled clinical trial? Clin Neurophar- 74. Zhou N, Yang X, Huang A, et al. The potential mechanism of macol. 2020;43(5):139–45. https://​doi.​org/​10.​1097/​WNF.​00000​ N-acetylcysteine in treating COVID-19. Curr Pharm Biotechnol. 00000​00405. 2021;22(12):1584–90. https://d​ oi.o​ rg/1​ 0.2​ 174/1​ 38920​ 10219​ 9920​ 60. Aggelakopoulou M, Kourepini E, Paschalidis N, et al. ERβ- 12282​12043. dependent direct suppression of human and murine Th17 75. Hardan AY, Fung LK, Libove RA, et al. A randomized controlled cells and treatment of established central nervous system pilot trial of oral N-acetylcysteine in children with autism. Biol Immunoregulatory and/or Anti-inflammatory Agents in Autism Spectrum Disorder 229 Psychiatry. 2012;71(11):956–61. https://​doi.​org/​10.​1016/j.​biops​ of Th17 related signaling and rectification of oxidant-antioxidant ych.​2012.​01.​014. imbalance in periphery and brain of BTBR T+tf/J mice. Behav 76. Ghanizadeh A, Moghimi-Sarani E. A randomized double blind Brain Res. 2019;364:231–224. https://d​ oi.o​ rg/1​ 0.1​ 016/j.b​ br.2​ 019.​ placebo controlled clinical trial of N-acetylcysteine added to 02.​031. risperidone for treating autistic disorders. BMC Psychiatry. 89. Nadeem A, Ahmad SF, Al-Ayadhi LY, et al. Differential regula- 2013;13:196. https://​doi.​org/​10.​1186/​1471-​244X-​13-​196. tion of Nrf2 is linked to elevated inflammation and nitrative stress 77. Nikoo M, Radnia H, Farokhnia M, et al. N-Acetylcysteine as an in monocytes of children with autism. Psychoneuroendocrinol- adjunctive therapy to risperidone for treatment of irritability in ogy. 2020;113:104554. https://​doi.​org/​10.​1016/j.​psyne​uen.​2019.​ autism: a randomized, double-blind, placebo-controlled clinical 104554. trial of efficacy and safety. Clin Neuropharmacol. 2015;38(1):11– 90. Simonetto M, Infante M, Sacco RL, et al. A novel anti-inflam- 7. https://​doi.​org/​10.​1097/​WNF.​00000​00000​63. matory role of omega-3 PUFAs in prevention and treatment of 78. Wink LK, Adams R, Wang Z, et al. A randomized placebo-con- atherosclerosis and vascular cognitive impairment and demen- trolled pilot study of N-acetylcysteine in youth with autism spec- tia. Nutrients. 2019;11(10):2279. https://​doi.​org/​10.​3390/​nu111​ trum disorder. Mol Autism. 2016;7:26. https://​doi.​org/​10.​1186/​ 02279. s13229-​016-​0088-6. 91. Giacobbe J, Benoiton B, Zunszain P, et al. The anti-inflammatory 79. Dean OM, Gray KM, Villagonzalo KA, et al. A randomized, role of omega-3 polyunsaturated fatty acids metabolites in pre- double-blind, placebo-controlled trial of a fixed dose of N-acetyl clinical models of psychiatric, neurodegenerative, and neurologi- cysteine in children with autistic disorder. Aust N Z J Psychiatry. cal disorders. Front Psychiatry. 2020;11:122. https://​doi.​org/​10.​ 2017;51(3):241–9. https://​doi.​org/​10.​1177/​00048​67416​652735. 3389/​fpsyt.​2020.​00122. 80. Tripathi A, Thangaraj A, Chivero ET, et al. N-Acetylcysteine 92. Bent S, Bertoglio K, Ashwood P, et al. A pilot randomized con- reverses antiretroviral-mediated microglial activation by attenuat- trolled trial of omega-3 fatty acids for autism spectrum disorder. J ing autophagy-lysosomal dysfunction. Font Neurol. 2020;11:840. Autism Dev Disord. 2011;41(5):545–54. https://​doi.​org/​10.​1007/​ https://​doi.​org/​10.​3389/​fneur.​2020.​00840. s10803-​010-​1078-8. 81. Liu X, Hu Z, Zhou H. N-Acetylcysteine improves inflammatory 93. Bent S, Hendren RL, Zandi T, et al. Internet-based, randomized, response in COPD patients by regulating Th17/Treg balance controlled trial of omega-3 fatty acids for hyperactivity in autism. through hypoxia inducible factor-1α pathway. Biomed Res Int. J Am Child Adolesc Psychiatry. 2014;53(6):658–66. https://​doi.​ 2021;2021:6372128. https://​doi.​org/​10.​1155/​2021/​63721​28. org/​10.​1016/j.​jaac.​2014.​01.​018. 82. Fu G, Xu Q, Qiu Y, et al. Suppression of Th17 cell differen- 94. Mankad D, Dupuis A, Smile S, et al. A randomized, placebo- tiation by misshapen/NIK-related kinase MINK1. J Exp Med. controlled trial of omega-3 fatty acids in the treatment of young 2017;214(5):1453–69. https://​doi.​org/​10.​1084/​jem.​20161​120. children with autism. Mol Autism. 2015;6:18. https://​doi.​org/​10.​ 83. Ruhee RT, Suzuki K. The integrative role of sulforaphane in 1186/​s13229-​015-​0010-7. preventing inflammation, oxidative stress and fatigue: a review 95. Mazahery H, Conlon CA, Beck KL, et al. A randomized con- of a potential protective phytochemical. Antioxidants (Basel). trolled trial of vitamin D and omega-3 long chain polyunsaturated 2020;9(6):521. https://​doi.​org/​10.​3390/​antio​x9060​521. fatty acids in the treatment of irritability and hyperactivity among 84. Singh K, Connors SL, Macklin EA, et al. Sulforaphane treatment children with autism spectrum disorder. J Steroid Biochem Mol of autism spectrum disorder (ASD). Proc Natl Acad Sci USA. Biol. 2019;187:9–16. https://​doi.​org/​10.​1016/j.​jsbmb.​2018.​10.​ 2014;111(43):15550–5. https://d​ oi.o​ rg/1​ 0.1​ 073/p​ nas.1​ 41694​ 0111. 017. 85. Momtazmanesh S, Amirimoghaddam-Yazdi Z, Moghaddam HS, 96. Doaei S, Bourbour F, Teymoori Z, et al. The effect of omega-3 et al. Sulforaphane as an adjunctive treatment for irritability in fatty acids supplementation on social and behavioral disorders of children with autism spectrum disorder: a randomized, double- children with autism: a randomized clinical trial. Pediatr Endo- blind, placebo-controlled clinical trial. Psychiatry Clin Neurosci. crinol Diabetes Metab. 2021;27(1):12–8. https://​doi.​org/​10.​5114/​ 2020;74(7):398–405. https://​doi.​org/​10.​1111/​pcn.​13016. pedm.​2020.​101806. 86. Zimmerman AW, Singh K, Connors SL, et al. Randomized con- 97. Kim JY, Lim K, Kim KH, et al. N-3 polyunsaturated fatty acids trolled trial of sulforaphane and metabolite discovery in children restore Th17 and Treg balance in collagen antibody-induced with autism spectrum disorder. Mol Autism. 2021;12(1):38. arthritis. PLoS ONE. 2018;13(3):e0194331. https://​doi.​org/​10.​ https://​doi.​org/​10.​1186/​s13229-​021-​00447-5. 1371/​journ​al.​pone.​01943​31. 87. Subedi L, Lee JH, Yumnam S, et al. Anti-inflammatory effect 98. Chehimi M, Ward R, Pestel J, et al. Omega-3 polyunsaturated of sulforaphane on LPS-activated microglia potentially through fatty acids inhibit IL-17A secretion through decreased ICAM-1 JNK/AP-1/NF-kB inhibition and Nrf2/H0-1 activation. Cells. expression in T cells co-cultured with adipose-derived Stemm 2019;8(2):194. https://​doi.​org/​10.​3390/​cells​80201​94. cells harvested from adipose tissues of obses subjects. Mol Nutr 88. Nadeem A, Ahmad S, Al-Harbi NO, et al. Nrf2 activator, sul- Food Res. 2019;63(11):e1801148. https://​doi.​org/​10.​1002/​mnfr.​ foraphane ameliorates autism-like symptoms through suppression 20180​1148.