Do toxins trigger autistic regression?

Embedded Scribd iPaper - Requires Javascript and Flash Player
Do toxins trigger autistic regression?
McGinnis WR, Miller VM, Audya T and Edelson S. Neurotoxic brainstem impairment as proposed threshold event in autistic regression. CRC Press 2009
Intriguing phenomenon of autistic regression
• Widely recognized • Usually 18-24 months of age • Relatively rapid: days or weeks • Earlier problems in some children • Published incidence as high as 50%
ARI Database
Features of regression
• Vocalization loss: acquired words or babbling (29% / 9%. Lord 2004) • Loss of social function, in some cases unassociated with loss of vocalization (Goldberg 2003) • Gastrointestinal impairment (Madsen 2004; Goldberg 2004)
GI tract in regressed cohorts
• Radiographic fecal loading or megacolon (100%. Torrente 2002) • Reflux esophagitis (69%. Horvath 1999) • Enterocolitis (88%. Wakefield 1998, 2002)
Toxins are plausible triggers
• • Parallel increase in autism and environmental toxicants (Lathe 2008) Autism rates correlate with: 1. Presence of toxic landfills (Ming 2008) 2. Estimated environmental cadmium and mercury (Windham 2006) 3. Proximity to mercury point sources (Palmer 2009)
Toxins as triggers
• Elevated dental concentrations of mercury (Adams 2007) • Autistic symptoms correlate with mercuryconsistent porphyrins (Geier 2008) • Case reports: Geier: temporal association with mercury by injection Bradstreet, et al: on-off speech on DMSA
The case of R.K.
• Progressive loss of speech over weeks post one-time placement of 9 amalgams at age 4. • By age 6, had regained 200 words, but elevated blood [Hg] explicable only on basis of amalgams. • Subsequent loss of all speech immediately after one-time removal of amalgams without precautions.
Toxins as triggers
• Many neurotoxins are oxidative, and oxidative modification is increased in brain of autistic children (Evans 2008; LopezHurtado 2008; Sajdel-Sulkowska 2008) • Gliosis and neuronal loss in autism is consistent with toxicant effects (Kern 2006)
Circumventricular Organs (CVO)
Area postrema (AP) Posterior pituitary Median eminence (ME) Subfornical Organ Organum vasculosum Pineal Gland [Nucleus Tractus Solitarius (NTS)] • Portals for toxins: no blood-brain barrier (BBB) • In and around the brainstem
CVO are preferentially sensitive to a broad class of neurotoxins
• Cadmium • Monosodium glutamate (MSG) • Paraquat • Inorganic mercury (including inorganic mercury from metabolic conversion of organic and elemental forms)
Cadmium
• Injections accumulate only in brain outside BBB, including AP and pineal (Arvidson 1986) • Lipid peroxidation (Mendez-Armenta 2003) blocked by antioxidants (Kim 2008) • Inhibits complex II and III (Wang 2004)
MSG
• Injections accumulate only in brain outside BBB, including AP (Karcsu 1985) and ME (Meister 1989; Peruzzo 2000) • Lipid peroxidation, persistent for long periods (Bawari 1995; Singh 2003) • Excitotoxic. In autism, GAD much lower in brain (Fatemi 2002)
Paraquat
• Injections accumulate only in areas outside the BBB, including AP and pineal (Naylor 1995) • Increased TNF-α and superoxide production by microglia (Wu 2005) • Inhibits complexes II and IV (Palmeira 1995)
Inorganic mercury
• Worrisome levels in air, water, soil (McGinnis 2001) and food (Dufault 2009) • Pink Disease proves fractional systemic absorption in children (McGinnis 2001) • Injections accumulate in AP and brainstem motor nuclei (Arvidson 1992) • Persists in brain for years (Vahter 1994) • Immune stimulation (Havarinasab 2007) and increased microglia (Geier 2007)
Elemental mercury
Amalgam removal decreased plasma and red-cell inorganic mercury levels by 73% (Halbach 2008)
Daily oral organic (methyl) mercury in primates
• In 6 brain areas, inorganic mercury averaged x30 at 6 mos., x60 at 18 mos. • By far, highest inorganic mercury at pituitary, only CVO examined. • If stop organic dosing at 6 mos, inorganic mercury doubles in pituitary at 12 months, but not in regions with BBB. (Vahter 1994)
Brainstem
• The “hard-drive” control panel for messages between brain and body • Rimland emphasized brainstem in 1964, but scant current neuropath interest • Many findings in autism consistent with brainstem dysfunction
Brainstem: medulla, pons, midbrain and [diencephalon]
Brain phylogeny
x2
Brainstem abnormalities
• Smaller medulla and midbrain on MRI (Courchesne 1997; Hashimoto 1993) • Reduced gray matter on MRI (Jou 2008) • Ectopic neurons and aberrant tracts (Bailey 1998) • Swollen medullary, thalamic, hypothalamic axon terminals (Weidenheim 2001) • Abnormalities of inferior and superior olives (Kemper 1993; Kulesza 2008)
Brainstem abnormalities
• Auditory brainstem response (Klin 1993; Kwon 2007) • Centrencephalic EEG (Gilberg 1983) • Heart rate, respiratory and vascular response (Bonvallet 1963; Althaus 2004) • Post-rotatory response (Ornitz 1983)
Suggest CVO impairment
• Pineal: abnormal melatonin production (Nir 1995; Kulman 2000; Tordjmann 2005) • Median eminence / posterior pituitary: abnormal oxytocin production (Modahl 1998; Green 2001)
Hypothalamus and pineal in relation to thalamus
Dorsal vagal complex (DVC)
1. Area postrema (AP) 2. Dorsal motor nucleus of vagus (DMV) 3. Nucleus tractus solitarius (NTS) The DVC mediates autonomic function of cervical, thoracic and abdominal viscera.
Portals for toxins
Dorsal vagal complex (DVC)
Dorsal vagal complex (DVC)
Area postrema
• No BBB • Highly vascular and very long residence time of blood in capillaries • So-called “emesis center” • Ablation increases consumption of water or concentrated salt water, food aversions, craving for carbohydrates and bland food • [Flavor aversion 2° Cd reversed by DMSA]
Nucleus tractus solitarius
• Lacks BBB on one side (Gross 1990) • Viscerosensory and visceromotor; parasympathetic and sympathetic efferent • Mediates social behavior, arousal, mood, emotion, anxiety, seizure activity and pain via limbic and cortical projections (Marvel 2004; Nemeroff 2006)
Secretin and NTS
• Highest binding of infused secretin at NTS; secretin activates NTS neurons (Yang 2004) • Several studies reported improvements in social behavior—may relate to NTS effect (Myers 2008) • Parent reports of sudden potty-training consistent with NTS effect (Beck, et al.)
Dorsal motor nucleus of the vagus (DMV)
• Visceromotor: peristalsis, esophageal sphincter tone, heart rate, pharyngeal and laryngeal musculature • Tensor palati to open eustachian tube • Viscerosecretory: digestion, floral balance
Phonation
• Significantly autonomic, “subconscious” • DMV visceromotor to all the intrinsic and extrinsic muscles of larynx and pharynx • Vagal alteration results in altered pitch (Shaffer 2005) • Vagal dysarthria should not be confused with classical “motor apraxia of speech”
DVC is anti-inflammatory
Suggest DVC impairment
• Excessive thirst (Terai 1999) • Salt-craving and flavor-aversion (ARI) • Otitis media (Konstantareas 1987; Rosenhall 1999) • Abnormal heart rate, respiratory and vascular (Bonvallet 1963; Althaus 2004) • Depressed cardiac parasympathetic and abnormal baroreflex (Ming 2005)
Suggest DVC impairment
• Esophageal reflux in 67% of regressed cohort (Horvath 1999) • Fecal loading or megacolon in 100% of regressed cohort (Torrente 2002) • Enterocolitis in 88% of regressed cohort (Wakefield 1998) • Paneth’s cells enlarged with granules (Horvath 1999)
Duodenal Paneth’s cells
(Courtesy of K. Horvath, Thomas Jefferson University)
Suggest DVC impairment
• 20% aged 3-5 identify better by pointing; of 23,685 children who regressed after 1 year, 4,141 had speech replaced by whisper for at least one week, and 679 whispered long-term (ARI parent survey) • Effective use of speech-generating device (Thunberg 2007) • Frank dysarthria reported in autistic subgroup (Weissman 2008)
Primary DVC impairment
• Sufficient explanation for core features of autistic regression: Loss of social skills Loss of vocalization Gastrointestinal disease • Impairment of other CVO might be expected to contribute to these core losses and other abnormalities
Anatomical consistency
• Observation: social regression often precedes loss of vocalization (Goldberg 2003) and may be unaccompanied by loss of vocalization (Goldberg 2003; Lord 2004) • Explanation: no BBB at superior aspect of NTS to impede toxin entry, but entry to DMV requires diffusion from AP or NTS
Parkinsonian parallels to autism
• Environmental factors strongly suspected • Inflammation may be “causative factor” (Whitton 2007) • Digestive symptoms frequent or dominant (Spellman 1977), with disordered motility (Cersosimo 2008), lax GE sphincter and reflux in 61% (Bassotti 1998) • DMV is consistent first site of pathology
Ramifying pathology of PD
(Braak H., et al. Cell Tissue Res 2004;31:121-134)
Ramifying pathology of PD
Possible mechanisms for ramifying brain pathology in autism
• De-afferentiation may disturb the development of higher brain structures (Tanguay 1982; Gessaga 1986; Geva 2008) OR • Cumulative effects of toxins / oxidative stress OR • Diffusion of inflammatory cytokines produced by CVO in response to toxicants
Microglial activation
• Brainstem has highest microglial density • Many toxic exposures are associated with release of excitatory cytokines associated with neuronal cell loss (Mangano 2009) • TNF-α is an cytokine suspected to play a pathogenic role in PD (McCoy 2003), and may be significant in autism
TNF-α
• Elevated in CSF of regressed cohort (Chez 2007) and cohort with 10/12 regressed (Zimmerman 2005) • High CSF/blood ratios suggest elevation due to increased brain production (Chez 2007) • AP and ME lack blood-CSF barrier as well as BBB (Broadwell 1983)
Endotoxin (LPS) model
• LPS poorly transits BBB • Systemic LPS induces immediate robust TNF-α only in CVO and adjacent structures, most intensely in AP and ME • TNF-α expression in NTS at 1.5 hours, marked by 18 hours • TNF-α absent in DMV initially, but present at 18 hours (Breder 1994)
Inflammatory toxins
• Cadmium potently stimulates inflammatory cytokines, including TNF-α (Souza 2004) • MSG increases TNF-α in brain which is unprotected by BBB, with resulting neuronal death (Chaparro-Huerta 2002) • Paraquat increases LPS-stimulated TNF-α from monocytes x18 (Erroi 1992) • Inorganic mercury accumulation in CVO associates with increased glia (Vahter 1994)
Cytokine transmission via CSF
• A pattern of inflammatory cytokine diffusion along nerve bundles suggests a diffusion pathway along small channels outside myelinated axons (Agnati 1995) • Experimentally, cytokines circulate from lateral ventricle via white matter nerve bundles of the corpus callosum, external capsule and striatum all the way to the amygdala (Vitkovic 2000)
Evans TA, et al., Am J Biochemistry and Biotechnology 2008;4(2):61-72.
Predicted threshold effects
• Oxidative stress—regardless of cause— would lower neurophysiological threshold for regression resulting from toxic effects on CVO: 1. Additive to oxidative neurotoxicity of CVO-preferential toxins 2. Cholinergic—especially muscarinic — systems are particularly sensitive to oxidative stress
Implications of site-specificity
• Regressive threshold may be reached by isolated or cumulative exposure to one compound, or additive/cumulative exposures to distinct compounds. • Impaired development or function of CVO by gestational or first-year factors not modulated directly by BBB would lower the threshold for regression triggered by CVO-preferential toxins after BBB maturation.
Possible gestational influences
• Vulnerable period for thalidomide as risk factor for autism (20-24 days) corresponds to the timing of formation of the medullary motor nuclei (Rodier 1997) • Cord-blood levels of mercury correlate with decreased autonomic activation of heart rate and brainstem auditory evoked potentials (Grandjean 2004)
CVO studies
Morphology Receptor density Oxidative modification Cytokine levels Toxin levels Vagus