| Domain | Best-supported finding | Evidence type/sample | Confidence/limitation |
|---|---|---|---|
| Definition/diagnosis | PANDAS is best understood as a **clinical syndrome** within the broader PANS construct: abrupt onset OCD and/or tics, childhood onset (3 years to puberty), episodic dramatic exacerbations, temporal association with GAS infection, and neurologic abnormalities such as motor hyperactivity or choreiform movements. No definitive confirmatory test exists. Established practice: clinical diagnosis after excluding alternative neurologic/psychiatric disease. (pqac-00000008, pqac-00000001, pqac-00000006) | 2024 Delphi consensus; prior review of working definitions | **Moderate** for use as current clinical framework; **low-to-moderate** for nosologic certainty because autoimmune causality remains debated. |
| Epidemiology | Male predominance is repeatedly reported: ~59.9% male in a 2022 systematic review and ~65% male in survey/cross-sectional data; male:female ratio reported as 3.33:1. Typical onset is prepubertal, around 6.3 years for tics and 7.4 years for OCD; children 5-12 years appear most susceptible. True incidence/prevalence remain uncertain and likely undermeasured. (pqac-00000010, pqac-00000007) | Consensus synthesis of cross-sectional/survey/systematic-review data | **Moderate** for age/sex pattern; **low** for population incidence/prevalence due to lack of robust population-based studies. |
| Pathophysiology | Leading hypothesis: GAS triggers molecular mimicry and post-infectious neuroimmune responses. Human/in vitro data show PANDAS-associated sera/monoclonal antibodies can enhance **dopamine D1 receptor** signaling; D1R autoantibodies discriminated PANDAS/PANS cohorts from controls with ~72%, 93%, and 79.5% accuracy in separate cohorts, whereas D2R more strongly characterized Sydenham chorea. Proposed chain: recurrent GAS -> anti-GAS/anti-neuronal immunity -> dopaminergic signaling changes and basal-ganglia inflammation. (pqac-00000012, pqac-00000013, pqac-00000016) | Human serum/monoclonal antibody study; consensus review | **Moderate** for biologic plausibility; **low-to-moderate** for direct causality in routine patients because biomarkers are not universally validated and conflict-of-interest concerns exist in parts of the biomarker literature. |
| Biomarkers | No circulating biomarker or autoantibody panel is currently pathognomonic. The Cunningham/Moleculera panel may provide auxiliary information, but sensitivity/specificity and clinical utility remain unclear. MRI/EEG abnormalities can occur but are not diagnostic. (pqac-00000015, pqac-00000007, pqac-00000004) | Consensus statement; review evidence | **High** that no validated diagnostic biomarker exists; major limitation is poor standardization and lack of prospective validation. |
| Prognosis | Course is usually **relapsing-remitting**, but a substantial minority have chronic/progressive illness. In a Swedish PANS follow-up (n=34, median 3.3 years), 2 remitted, 20 were relapsing-remitting, and 12 had chronic-static/progressive course. Consensus synthesis states roughly one-third chronic-progressive and two-thirds relapsing-remitting; OCD (62%) and tics (50%) were common follow-up symptoms. In a prior PANDAS longitudinal cohort, 72% had at least one exacerbation, 12% had clinically significant OCD at follow-up, and 9% had chronic-progressive course. (pqac-00000018, pqac-00000014) | Longitudinal cohort studies; consensus synthesis | **Moderate** because follow-up data exist but sample sizes are small and combine PANS/PANDAS constructs. |
| Acute GAS treatment | When **active GAS infection is documented**, standard anti-streptococcal treatment is recommended; this is the most established infectious-disease intervention. Small prospective data cited in consensus report suggest OCD symptoms may rapidly improve after eradication therapy in new-onset PANDAS, but evidence base is small. Children with psychiatric symptoms but no evidence of GAS should not routinely receive antibiotics. (pqac-00000014, pqac-00000024) | Consensus/guideline-style recommendations; small prospective studies | **Moderate** for treating documented GAS infection; **low** for expecting consistent neuropsychiatric remission from antibiotics alone. |
| CBT/SSRIs | Best-supported symptomatic therapy is standard OCD care: CBT/ERP and SSRIs. Response rates appear similar to non-PANDAS pediatric OCD. Small PANDAS CBT data: n=7 with symptom improvement; 3/6 in remission at 3 months. Survey/retrospective evidence and consensus place CBT/SSRIs as first-line symptomatic treatment. (pqac-00000022, pqac-00000024, pqac-00000007) | Small interventional study; reviews/consensus | **Moderate** for use in practice because evidence is extrapolated from pediatric OCD and supported by small PANDAS studies. |
| Prophylactic antibiotics | Evidence is inconsistent and insufficient for routine prophylaxis. A randomized placebo-controlled crossover trial (n=37) found penicillin prophylaxis did **not** reduce exacerbation rates or tic/OCD severity; a noncontrolled 12-month study (n=23) reported 96% fewer GAS infections and 61% fewer neuropsychiatric exacerbations. Reviews and clinical guidance do **not** recommend routine prophylactic antibiotics. (pqac-00000022, pqac-00000024, pqac-00000021) | One randomized trial; one uncontrolled study; reviews | **Low-to-moderate** because results conflict and better-controlled data are lacking. |
| Steroids/NSAIDs | Anti-inflammatory therapy is used selectively, not as universal standard care. Retrospective evidence suggests corticosteroids shortened flares by ~3.5 weeks; consensus notes evidence for NSAIDs/steroids is scarce and they should be individualized, generally under specialist oversight. (pqac-00000022, pqac-00000007, pqac-00000021) | Retrospective study; expert consensus | **Low** due to nonrandomized evidence and uncertain patient selection. |
| IVIG/PLEX | Immunomodulation remains **investigational/specialist care**. Earlier controlled trial data suggested improvement with IVIG or plasma exchange in severe cases, but a later randomized double-blind IVIG trial in 35 children found no significant difference vs placebo at 6 weeks; longer-term improvements occurred across groups and may reflect natural history. Open-label studies in PANS show improvement (e.g., 10-child 2022 trial; 2024 study of 10 boys with improved psychometric scores and reduced pro-inflammatory monocytes), but these do not resolve efficacy. PLEX retrospective severe-PANS series reported ~65% improvement at 6 months and 78% at longer follow-up. (pqac-00000023, pqac-00000025, pqac-00000017, pqac-00000021) | Randomized trials plus open-label/retrospective studies | **Low-to-moderate** for efficacy; strongest limitation is placebo response, spontaneous improvement, small samples, and heterogeneity. |
| Tonsillectomy | Not supported as effective treatment. Larger retrospective/prospective studies found no meaningful differences in OCD/tic severity, GAS titers, or remission between surgery and no surgery; consensus and reviews do not recommend routine tonsillectomy/adenoidectomy for PANDAS. (pqac-00000023, pqac-00000004, pqac-00000007) | Retrospective and prospective observational studies | **Moderate** for concluding lack of supportive evidence; absence of randomized trials remains a limitation. |
| Genetics/familial predisposition | No monogenic cause is established. Family autoimmunity is enriched: ~20% of first-degree relatives had at least one serious autoimmune diagnosis; rheumatic fever history was reported in 3% of mothers, 1% of fathers, and 14% of grandparents in one survey synthesis. Exploratory WES in severe PANS identified variants in 11 genes (including PLCG2, NLRC4, CACNA1B, SHANK3, GRIN2A, RAG1, SYNGAP1), but these are susceptibility candidates, not diagnostic PANDAS genes. (pqac-00000010, pqac-00000016, pqac-00000008) | Family-history surveys; exploratory WES in severe PANS | **Low-to-moderate** for susceptibility signal; **low** for clinical genetic testing utility in PANDAS today. |
| Animal models | Recurrent intranasal GAS mouse models support a mechanistic neuroimmune pathway: GAS-specific Th17 cells from NALT/tonsil-associated immunity can enter the brain, with BBB breakdown, IgG deposition, microglial activation, and loss of excitatory synaptic proteins in the absence of viable CNS bacteria. Important limitation: pathology localized strongly to olfactory-connected regions, with **no extensive basal ganglia BBB damage** observed, so model recapitulation is incomplete. (pqac-00000020, pqac-00000019) | Human tonsil immunology plus recurrent-GAS mouse model | **Moderate** for biologic plausibility; **low-to-moderate** for full disease fidelity to human PANDAS. |


*Table: This table summarizes the strongest currently available evidence across diagnosis, mechanisms, treatment, prognosis, genetics, and models for PANDAS. It separates routine clinical practice from investigational or hypothesis-driven areas and highlights the main limitations of the evidence base.*