01From immune activation to neurodevelopmental change
The autism cascade model is a systems-biology framework that maps the sequential biological events connecting upstream immune activation to downstream neurodevelopmental change. It is not a single-gene or single-pathway model. It is a convergence model — one that explains how multiple upstream inputs (infections, environmental exposures, gut dysbiosis, genetic susceptibility) produce a common downstream phenotype through a shared biological mechanism.
The cascade proceeds through six levels. Each level is supported by published, peer-reviewed research. The full integrated model has not yet been validated in a single large clinical trial — which is precisely what this initiative exists to address.
The Autism Cascade — Six Levels from Trigger to Regression
▶ IMIG / IVIG Entry Point
Level 1 — Upstream Triggers & Immune Activation
Chronic or recurrent immune activation from infections, gut dysbiosis, environmental exposures, or genetic susceptibility drives sustained elevation of pro-inflammatory cytokines — principally IFN-γ, TNF-α, and IL-6. This is where immunoglobulin therapy acts: suppressing the cytokine environment before the downstream cascade loads.
IMIG intervenes here — upstream of everything that follows
Level 2
IDO1 Activation & Kynurenine Pathway Diversion
IFN-γ upregulates indoleamine 2,3-dioxygenase 1 (IDO1), the rate-limiting enzyme in the kynurenine pathway. Tryptophan — the sole precursor to serotonin and a key input to NAD⁺ synthesis — is diverted away from normal metabolism into the kynurenine pathway. Quinolinic acid (QUIN) accumulates as a downstream kynurenine metabolite. QUIN is a potent NMDA receptor agonist.
Level 3
NAD⁺ Depletion & SIRT1 Collapse
Kynurenine pathway diversion reduces tryptophan availability for NAD⁺ synthesis via the salvage pathway. Simultaneously, the inflammatory environment increases PARP activation — a major consumer of NAD⁺. The combined effect is NAD⁺ depletion. SIRT1, a NAD⁺-dependent deacetylase, loses its cofactor. SIRT1 regulates four critical downstream programs: mitochondrial biogenesis (PGC-1α), neuroinflammation (NF-κB), synaptic plasticity (CREB/BDNF), and the somatostatin system. All four fail simultaneously.
Level 4
Somatostatin Elevation & cAMP/CREB Suppression
Somatostatin (SST) rises as a compensatory response to neuroinflammation. SST suppresses adenylyl cyclase (AC), collapsing the cAMP → PKA → CREB signalling axis. CREB (cAMP response element-binding protein) is the transcription factor that drives BDNF expression and synaptic remodelling. Without CREB activity, the molecular machinery for learning, plasticity, and language acquisition is suppressed. The SST brake, once elevated, requires weeks of consistently reduced inflammatory tone to recalibrate — it cannot do so during the troughs between IVIG infusions.
Level 5–6
NMDA Overactivation, Synaptic Regression & Reduced Plasticity
QUIN-driven NMDA overactivation produces excitotoxic stress at developing synapses. Simultaneously, the loss of CREB-driven BDNF expression means synaptic architecture is not maintained or built. The system enters a low-plasticity state characterised by reduced language acquisition, reduced social responsiveness, and in the regression phenotype, active loss of previously acquired skills. In children with profound ASD (Level 2–3), this state has typically been established over years of cumulative inflammatory loading.
Section 02 — The SST Normalisation Argument
02Why recovery requires sustained suppression — not periodic intensity
The most clinically significant implication of the cascade model for immunoglobulin therapy is the somatostatin normalisation timeline. SST elevation is a compensatory response — it rises because neuroinflammation is chronically active, and it drives AC/cAMP suppression as a downstream consequence. Reducing it requires not just a period of reduced inflammation, but a sufficiently sustained period that the SST system genuinely recalibrates.
The timescales involved are measured in weeks, not days. SIRT1 recovery requires sustained NAD⁺ availability. CREB-driven BDNF expression and the synaptic remodelling it enables operate on timescales of weeks to months. This creates a specific prediction about IVIG's oscillating pharmacokinetic profile.
IVIG peak — acute cytokine suppression
IFN-γ, TNF-α, and IL-6 fall. IDO1 activity decreases. Kynurenine pathway flux reduces. Parents and clinicians observe behavioural improvements. SIRT1 begins to recover as NAD⁺ availability improves.
IVIG trough — cytokine rebound
Serum IgG returns toward pre-infusion levels. Inflammatory cytokines begin rising again. IDO1 reactivates. The partial SST normalisation achieved in weeks 1–2 is interrupted before completion. SIRT1 recovery stalls. Gains plateau or partially reverse — a pattern documented in at least four independent IVIG studies.
IMIG steady-state — the sustained window the biology requires
IMIG's slower absorption profile produces a sustained IgG elevation without the sharp trough. Inflammatory suppression is maintained continuously rather than periodically. The SST system is given the uninterrupted window it needs to genuinely recalibrate. SIRT1 recovery can proceed. CREB/BDNF-dependent plasticity can re-engage. The six-month minimum trial duration is derived from this timeline — it is the threshold at which cascade recovery becomes measurable.
The Pharmacokinetic Argument
IMIG's lower peak is not a limitation — it is the feature
IVIG's high serum IgG peak produces dramatic early responses, which is why the clinical signal in the literature is so clear. But the trough that follows may be working against the very biology IVIG is trying to restore. IMIG never reaches IVIG's peak — but it never drops as far. For cascade biology, where the therapeutic target is weeks of consistently reduced neuroinflammation rather than periodic peaks of immune suppression, IMIG's steady-state profile is mechanistically preferable, not merely more convenient.
Section 03 — PANS as Mechanism Validation
03Why the PANS result is critical for the ASD argument
Pediatric Acute-onset Neuropsychiatric Syndrome (PANS) is a condition in which autoantibodies — often triggered by streptococcal or other infections — target basal ganglia structures, producing acute behavioural regression, OCD, and neuropsychiatric symptoms. Its mechanism is well-characterised: autoantibody-mediated neuroinflammation producing acute-onset CNS dysfunction. It is, in a meaningful sense, a clean experimental model for the immune-neurological interface.
Why PANS Validates the ASD Mechanism
A well-characterised autoimmune mechanism producing a known CNS phenotype — and responding strongly to IMIG
In the Fourie & Armstrong 2024 case report, five PANS children treated with monthly IMIG showed a mean improvement score of 4.4 out of 5. All five scored above zero. This is a near-ceiling response for a case series of this type.
The significance for the ASD argument is direct: if IgG-mediated immunomodulation via the intramuscular route produces this level of clinical effect in a condition with a well-characterised autoimmune mechanism, it validates that the route and formulation work. The question for ASD is not whether IMIG can modulate the immune system — the PANS result establishes that it can, via this route, at this dose. The question is whether the ASD subpopulation with measurable immune dysregulation shares enough mechanistic overlap with PANS to respond similarly.
The cascade model predicts that they do — and that the more modest ASD response in the case report (+2.9/5) reflects the timing effect of intervening after years of established neuroinflammatory loading, rather than a fundamental mechanistic difference. The testable prediction: earlier intervention, in younger children with recently elevated biomarkers, should produce a response closer to the PANS result.
This is not a speculative extrapolation. It is a specific, falsifiable hypothesis that the Proposed IMIG for Autism Controlled Pilot Study with biomarker-stratified patient selection can directly test. The trial design — including the patient selection criteria, the biomarker panel, and the age envelope — is built around this prediction. See the Trial page for the full design rationale.
For the full evidence base — Rossignol & Frye 2021, Fourie & Armstrong 2024, and what the data collectively points toward — see the Evidence page.