Involuntary Adventures in Autism

A Stranger in a Strange Land
Involuntary Adventures in Autism
James R. Laidler, MD

Prologue:

My interest in autism began with the diagnosis of my oldest son in 1997 and my youngest son in 1998. Prior to this, I had encountered autism only as an intellectual challenge, a diagnosis to make and pass on to the unfortunate parents. When my interest became more personal, I began to look more closely at what was known about autism and its treatment. The conventional medical literature had little to offer in the way of treatment other than education and speech therapy, so that was what we did.

My wife (who is also a doctor) was not as easily satisfied, however, and she explored "alternative medi-cine" and an organization known as "DAN!" (Defeat Autism Now!). She went so far as to go to a confer-ence held by DAN! and came home very excited about some therapies that I found rather crazy. Not being the confrontational sort, I held my tongue (more or less) and let her do these crazy things to our kids. My thought was that I would be able to show her that these treatments did not work and then we would go back to the therapies I was more comfortable with. Unfortunately, the treatment (gluten-free, casein-free diet) worked spectacularly. This left me to figure out why it worked and thus began my exploration of the scien-tific basis of autism treatments.

Introduction:

The popular press makes a great deal about the dichotomy between "mainstream medicine" and "alterna-tive medicine". If one takes the long view, however, this division blurs considerably. Many of the treat-ments that were initially considered "alternative" have become "mainstream" as their efficacy was proven (e.g. antibiotic treatment for ulcers) and many "mainstream" therapies have slipped into the "alternative" camp as they fell out of favor (e.g. herbal medicine). The real division in medicine is between the therapies that work and those that don't. Unfortunately, this division can only be seen in hindsight.

Although double blind, placebo-controlled studies are the "gold standard" for evaluating medical therapies, most treatments in current use were validated by the "silver standard" of clinical trial. In other words, they worked. Once a treatment is shown to work, it becomes very difficult, if not frankly unethical, to do a pla-cebo-controlled study. This does not mean that the treatment cannot be scientifically evaluated in humans, since there are many other valid means of testing.

Most, if not all, of the biomedical treatments for autism have not been formally evaluated for efficacy; they are still in the phase of development that I have come to call "clinical tinkering". This is the stage where clinicians put an idea from basic science into practical use and is the breeding ground for most of the prac-tical therapies in current use. The next step, and one being done right now by a few centers, is to critically evaluate the results of these therapies.

In the absence of quantitative information supporting biomedical treatments for autism, how can we sepa-rate the wheat from the chaff? One way is to evaluate the scientific basis of the treatment to see if it holds water. A large number of biochemical and biological abnormalities have been identified in autistic children (and adults) and treatments directed at these abnormalities are likely to be helpful to some extent. Treat-ments for diseases that share some commonality with autism may also be helpful, but the commonality may be spurious.

The greatest difficulty in critically evaluating biomedical treatments for autism is separating spontaneous improvement from improvement due to the treatment. The second greatest difficulty is discriminating be-tween real improvement and wishful thinking (placebo effect). Since many of the features of autism are difficult or impossible to objectively evaluate, the effect of the treatment on the evaluator (parent, re-searcher, doctor or therapist) cannot be ignored. These difficulties were demonstrated in dramatic fashion when a recent double-blind, placebo-controlled study of secretin was completed. One child, who had improved markedly during the trial, was found to have received the placebo (saline); his parents were anxious to find the composition of the placebo so they could continue giving it.

The Gluten- and Casein-free (GF/CF) Diet:

We are indebted to Dr. Kalle Reichelt for discovering the biological basis for the success of the GF/CF diet when he found high levels of casomorphins in the urine of autistic people . Dr. Jon Pangborn has discussed how an intestinal enzyme (DPP IV) defect that may be responsible for this high level. These peptides are the result of incomplete digestion of the milk protein, casein (casomorphins) and the wheat (and other grains) protein, gluten (gliadomorphins). Unfortunately, none of these scientists has yet been able to prove that the high level of urinary casomorphins and gliadomorphins contributes to autism.

People without overt autistic disorders also have casomorphins and gliadomorphins detectable in their urine after eating casein , and gluten , indicating that the mere presence of these peptides is not abnormal. How-ever, studies of b-casomorphin and its derivatives have shown that it has powerful effects on sleep , pain , neuronal development , immune function and hormone function , , . There are even studies connecting post-partum psychosis , to circulating casomorphins resulting from the production of breast milk . This would suggest that there is a considerable range in the sensitivity to casomorphins.

The MMR/Autism Controversy:

"The lady doth protest too much, methinks." Hamlet, Act III, scene 2, line 242

Much poor science has been done in recent years in order to save the MMR vaccine from careful scrutiny. Most of this has been in the form of poorly designed and executed retrospective studies on the epidemiol-ogy of autism with respect to the introduction of the MMR vaccine in the US and UK. A case-by-case, line-by-line discussion of these studies is beyond the scope of this talk, but I will attempt to address the most egregious (and laughable) errors seen in most of them.

Admittedly, these flaws do not prove that the MMR vaccine causes autism. What they do show is that the studies done so far do not exonerate the MMR vaccine.

While there is not yet a "smoking gun" pointing to the MMR vaccine, there are a large number of studies that indicate how it might potentially cause some of the problems seen in autism. The first of these are stud-ies showing that the passive reporting of vaccine adverse effects grossly underestimates their incidence , . These studies undermine the comforting statistics proffered by the CDC and other public health organiza-tions, since their vaccine safety statistics are based solely on such passive reporting.

When vaccine recipients are followed actively, the incidence of adverse effects rises by a factor of five or ten, even for serious complications such as meningitis and febrile convulsions. If these reactions, which occur soon after the vaccination and are not subtle, are under-reported, how much more so for something subtle in nature and delayed in onset, like autism?

Measles:

The measles virus, even the vaccine strains, cause a significant and prolonged (up to 3 months) immuno-suppression. Certain vaccine strains, specifically the Edmonston-Zagreb strain, cause more damage to the immune system than others . This immune system damage is not just a hypothetical problem; vaccine tri-als in Africa resulted in higher than normal mortality in vaccinated children because of their inability to resist bacterial and parasitic infections , . Generally, the vaccine strains most effective at inducing immu-nity cause more immunosuppression.

Mumps:

Like the wild-type virus, the mumps vaccine is capable of causing meningitis and does so in about one in five thousand recipients. One particular vaccine strain, Urabe AM-9, has been documented to cause menin-gitis in one recipient in three thousand , . Again, the rate of meningitis varies directly with the effective-ness of the vaccine.

Rubella:

Rubella, both wild-type and vaccine strains, can cause an autoimmune arthritis which affects females more often than males .

The Autism Epidemic: Truth or Rumor?:

In discussions with parents and educators, there is no question that the incidence of autism is increasing, and rapidly. Epidemiologic studies attempting to analyze the question are contradictory. Part of the prob-lem is the way the diagnosis of autism has evolved over the past decades. Another problem is that there is one definition for educators and a different one for the medical community. Add in such ambiguities as PDD (Pervasive Developmental Disorder) and MSDD (Multi-System Developmental Delay), "autistic spectrum disorder" (ASD), and whether some children with ADD/ADHD are autistic, and it is amazing that any useful information is available.

Ironically, the educational definition of autism is probably the most useful for evaluating the epidemiology of autism, since it is very formalized, fairly consistent and, most importantly, has not changed significantly in the past several years. The medical definition of autism has changed so often and is so subjective that the statistics from one year cannot be compared to the next.

The US Department of Education presents an annual report to Congress covering the Individuals with Dis-abilities Education Act (IDEA). This report contains over 300 pages of tables describing the population of students who fall under the IDEA. Buried in these tables are the numbers of children with autism by age, sex, race and location.

Figure 2
Figure 2 shows a "snapshot" of two different disabilities (autism and mental retardation) in a single school year (1998-1999). The curve of the line for mental retardation is similar to that of most other disabilities. The initial sharp rise represents the time to make the diagnosis and the decline after age 16 represents those children leaving the school system. The curve for autism is strikingly different. Since the majority of chil-dren with autism are diagnosed by age 3 and the graph starts at age 6, the initial rise in the curve is not visi-ble. More importantly, the curve shows a steady decline with increasing age. This decline could be due to children "recovering", but autism is a "life-long" diagnosis and its definition does not allow for recovery. The only other explanation is that the incidence of autism is increasing. Those who argue that the rise in autism is due to improved or broadened diagnosis will have to explain why the diagnosis apparently cannot be made in older children.

Mercury and Autism: Cause or Association?

A number of people have argued that there is an association between exposure to mercury, specifically in the form of thimerosal (thiomersal in the UK) in vaccines, and autism. On the face of it, this would be a difficult hypothesis to test, since thimerosal-containing vaccines are nearly universally administered and, in many instances, are legally mandated. With 99+% of all children in the US, UK and Europe receiving roughly equivalent doses of thimerosal, the association is not very statistically significant. One possible means of finding an association is to examine the rise in autism in concert with the increasing number of vaccinations; the two numbers rise in rough synchrony, implying an association between them.

Currently, the strongest argument connecting mercury and autism is the similarity between some effects of mercury toxicity and some of the non-behavioral features of autism. Many of the more verbal people with autism have reported that their visual fields are severely constricted, resulting in "tunnel vision". Visual field restriction is also a prominent feature of methylmercury poisoning . Another common feature of au-tism is difficulty with visual perception and recognition as well as a "fragmenting" of the visual image. Similar problems have been reported in patients suffering from mercury poisoning .

Abnormalities in immune function are frequently reported in autism, ranging from autoimmunity to poor cellular immunity and reduced ability to clear bacterial infections. Mercury, even in "non-toxic" amounts, has been shown to cause T-cell death , autoimmunity (especially anti-brain antibodies) , poor monocyte function and prolonged/exaggerated humoral immunity , . Many of these effects result from alteration of apoptosis (programmed cell death) . Mercury causes premature apoptosis in many cell lines (including neuronal cell lines and T-cells) and delays apoptosis in B-cells and plasma cells. Premature cell death has obvious implications, but delaying programmed cell death can be just a problematic since this is the body's way of terminating a number of functions, such as antibody production.

Mercury is so toxic that it can cause dysfunction in almost any organ system. This tends to support the hy-pothesis that mercury could be responsible for at least a portion of the patients with autism. Mercury's ef-fects are so wide-ranging that they can mimic a broad spectrum of diseases. Unfortunately, the converse is also true: many diseases may have features that look like mercury toxicity.

DMSA and Autism:

Working on the hypothesis that autism may be caused by mercury, several clinicians treated autistic chil-dren with mercury chelating agents (primarily DMSA). Although no outcome studies have yet been done, the anecdotal reports are encouraging. This was initially thought to be final proof that mercury causes au-tism, but some of the children who showed improvement did not have significant excretion of mercury (or other toxic metals). The only conclusion that can be drawn is that some children with autism benefit from treatment with DMSA.

DMSA is used primarily to chelate heavy metals (such as lead, mercury and cadmium), but it does have other actions. It is a powerful antioxidant and can help with the antioxidant depletion seen in autistic chil-dren. It also can act as a sulfur donor, although it is usually given in doses too small to be a major source of sulfur. DMSA can also bind to disulfide-containing compounds, such as the gliotoxin produced by patho-genic yeast. Finally, DMSA may have actions not yet known.

All of this led Dr. Rimland of the Autism Research Institute (ARI) to sponsor a "think-tank" on mercury and DMSA. The results of this conference were published in a position paper, which is available from the Autism Research Institute. The basic principles for using DMSA in autism are simple, however, and they are outlined below.

Pre-treatment:

DMSA has been used extensively in children to treat lead poisoning; that experience has shown that DMSA is generally safe. DMSA does cause some degree of temporary bone marrow suppression in everyone who receives it; this suppression invariably resolves when the drug is stopped. However, some people have ex-perienced potentially dangerous drops in their white blood cell and platelet numbers while taking DMSA. For this reason, a complete blood count (CBC) with platelet count is essential to give a starting measure before beginning DMSA therapy.

A few people will show a rise in certain liver enzymes (ALT and AST) during DMSA therapy, suggesting liver injury, although no clinical symptoms of liver damage have been reported. Checking the blood level of the major liver enzymes (ALT, AST, and alkaline phosphatase) prior to starting therapy will give a base-line measure.

Many practitioners are testing urine, hair and stool for mercury and other heavy metals prior to starting DMSA (or after a "provocation" dose of DMSA or DMPS). The value of these tests lies primarily in their ability to encourage the parents and practitioners to start and continue DMSA therapy. It is not at all clear that DMSA exerts its effect in autism through removal of mercury or other metals

In order to derive the maximum benefit from DMSA therapy, nutritional status and gastrointestinal condi-tion should be optimized. A number of people have interpreted this to mean that all abnormalities must be completely corrected before therapy can begin; however, the key word is "optimized". Many autistic indi-viduals have dietary idiosyncrasies and gastrointestinal problems that defy complete correction; these prob-lems should be addressed as completely as possible.

DMSA Dosing:

The optimal dosing for DMSA in autism is not entirely clear because it is not clear what the DMSA is do-ing (if anything). The vast experience using DMSA for lead poisoning has led to some very clear guidelines for maximum doses. As always, the minimum dose that produces the desired effect should be used. No single dose of DMSA should exceed 10 mg/kg body weight, with a maximum single dose of 500 mg re-gardless of weight. The maximum daily intake should not exceed 30 mg/kg body weight, with a maximum daily intake of 1500 mg regardless of weight. Many people have received doses much greater than these without significant ill effect, but these maximum doses are generally considered safe and effective (for heavy metal poisoning). Likewise, many people have reported good results with doses much less than these maximum doses. As a rule, start with a low dose and gradually increase until the desired effect is attained or the maximum dose is reached.

Dosing interval is an area of some controversy; good results have been reported with dosing intervals of anywhere from three to twelve hours. It is likely that there is no ideal dosing interval and that any perceived difference is spurious. If the effect of DMSA is assumed to be due to removal of heavy metals, there is no need to maintain a steady blood level and any dosing interval can be used. Dosing every eight hours allows for uninterrupted sleep (for both patient and parents) and is reported to be effective at least as often as any other interval.

When DMSA is used for lead poisoning, it is given for 14 - 19 consecutive days, followed by a "rest" pe-riod of at least 14 days. The long treatment period is needed because lead is incorporated into bone where it acts as a reservoir. Mercury and other heavy metals do not have a bone reservoir and so a shorter treatment period is acceptable. Three days of DMSA therapy is sufficient to reduce the blood mercury, cadmium or arsenic level to below detectable levels in all but the most severe poisoning. A "rest" period of eleven days or longer allows time for the metal inside cells to approach equilibrium with the blood and extracellular fluid. This is the rationale for recommending a regimen of three days of DMSA followed by eleven days of "rest". Other regimens may be more effective if DMSA is working by other means.

Concurrent testing and therapies:

Because of the possibility of bone marrow suppression and liver enzyme elevation, a complete blood count, platelet count, and serum liver enzyme level should be repeated after a month of DMSA therapy. If these results are normal (or at least unchanged from the pre-therapy levels), they can be repeated at greater inter-vals during therapy. New abnormalities should be closely followed and the DMSA dosing should be re-duced or stopped.

a-Lipoic acid (ALA) is a nutritional supplement that may be of special use during DMSA therapy. It is a dithiol fatty acid that is present in the cells of the body as an antioxidant and natural heavy metal chelating agent. Its inclusion into most regimens using DMSA to treat autism is based on a single oft-quoted but rarely translated Russian study. This study showed that ALA given with mercury increased the absorption of mercury and its entry into cells of the brain and liver. Presumably, ALA should also accelerate removal of mercury from the cells, although this hypothesis has not been tested. Some practitioners who use ALA routinely claim that it causes significant fecal excretion of mercury and this is consistent with the known method of action.

In addition to its potential action as a heavy metal chelator, ALA also is a powerful antioxidant and is sec-ond only to glutathione in this respect. Numerous studies have shown that ALA is helpful in diseases such as diabetic neuropathy where excessive oxidation is a significant feature. Since antioxidant depletion is often reported in autistic individuals, ALA may be helpful in autism as well.

Dosing ALA in autism is a matter of considerable debate. ALA is a natural product of mammalian cells and no reports of overdose in humans have been published. Studies using ALA for diabetic neuropathy have used up to 25 mg/kg/day for up to five years without complication; a starting dose of 5 mg/kg/day should be well tolerated by most people. Some practitioners delay starting ALA until after a few months of DMSA therapy for fear of moving mercury into sensitive tissues, but this is extremely unlikely unless the recipient is ingesting a significant amount of mercury.

General caveats:

The chain of evidence connecting mercury to autism is extremely weak; it relies on a known exposure (vaccines containing thimerosal), similarity of symptoms between mercury poisoning and autism, and im-provement after treatment with DMSA. The improvement seen after DMSA is the weakest link of all; it is based on anecdotal reports and is not a consistent finding. Even if DMSA is eventually shown to have effi-cacy in treating autism, that would still not prove that mercury causes autism.

Much of the biomedical treatment of autism is based on supposition and hypothesis; there is little or no objective evidence to support it. Any or all of the hypotheses supporting autism treatment could be wrong; the improvements seen may be placebo effect or spontaneous improvement. The placebo effect is ex-tremely pervasive, especially when the measured outcome is not measurable or quantifiable. Until placebo-controlled studies are done the possibility remains that any or all of these therapies may be ineffective.

Autism is most likely not a single disease. Rhett's disease, Landau-Kleffner disease and Fragile-X disease were all once included in "autism" and are now known as separate diseases. It is only sensible that more will follow. It should therefore come as no surprise that no therapy works for all people who have autism.