1996 Autism Society National Conference Proceedings
What follows are proceedings from the Autism Society's national
conference held July 10-13, 1996 in Milwaukee, Wisconsin. The material
included here addresses some of the scientific and technical issues relevant
in organic acid testing for children with autism.
Organic Acid
Testing to Evaluate Abnormal Microbial
Metabolies in Urine of Children with Autism
Experience with Organic Acid Testing
to Evaluate Abnormal Microbial Metabolites in the Urine of Children With
Autism
For the past two years, we have evaluated by gas-chromatography mass-spectrometry
biochemical abnormalities that appear to be of microbial origin in urine
samples of children with autism and other developmental disorders. Our
interest in this phenomenon began when we found that certain putative
microbial metabolites appeared in higher than normal values in urine samples
of two brothers with autism (1).
These findings were of especial interest to us because of a report that
autistic children have a greater incidence of ear infections than age-matched
peers; that lower functioning autistic children had an earlier onset of
ear infections than their higher functioning autistic peers; and that
the ears of children with autism were anatomically positioned differently
than those of normal children, perhaps leading to greater ear infection
susceptibility (2). Intestinal overgrowth of yeast and anaerobic bacteria
are well documented sequelae of the common oral antibiotics used to treat
ear infections (3-10).
Therefore we considered the possibility that the
biochemical products of abnormal microorganisms may play a role in the
etiology of autism just as abnormal elevations of phenylalanine and its
metabolites cause PKU. Certain of the metabolites that had been
previously identified in urine samples include tartaric and citramalic
acids. Other compounds that we identified for the first time in urine
samples included arabinose(a carbohydrate), 3-methylmalic acid, 3-oxoglutaric
acid, phenylcarboxylic acid, and carboxycitric acid (1).
Since that time we have identified several additional compounds as commonly
increased in the urine samples of children with autism including dihydroxyphenylpropionic
acid, furandicarboxylic, hydroxymethylfuroic, and furancarbonylglycine.
We suspected that most of the above compounds were of microbial origin
based on reports that demonstrated the presence of these compounds or
closely related biochemicals in the culture media of yeast,fungi ,or bacteria
(1, 11-13).
During this period of time, we have gathered information about these
compounds and their possible role in autism:
By conducting a formal uncontrolled clinical trial of antifungal
drug therapy which involved 23 children with autism who were seen as
outpatients at the hospital;
By offering reference laboratory testing
to physicians who used this testing to evaluate possible standard inborn
errors of metabolism as well as putative abnormal microbial metabolites;
By testing culture media of different microorganisms to determine
which compounds were produced by different species.
if the urinary excretion of abnormal Krebs cycle
metabolites and/or abnormal carbohydrates are biochemical characteristics
in autistic children and
if antifungal treatment results in decrease or
elimination of abnormal Krebs cycle metabolites, metabolites such
as dihydroxyphenylpropionic acid,phenylcarboxylic acid, and/or carbohydrates
in autistic children and/r improvement in autistic symptoms.
The study was approved by the Investigational Review Board Of Children’s
Mercy Hospital. Funding for the studies was provided by grants from
the Katherine B. Richardson Foundation , from Pfizer Pharmaceutical
Corporation, and the parents of an autistic child.
Twenty-three autistic children were enrolled for the study. Each
child was classified as autistic according to the latest criteria
proposed by the American Psychiatric Association DSM-IV (1994).
After informed consent, a random urine sample was collected without
special preparation for organic acid analysis by gas chromatography/mass
spectrometry.
If the urine showed the presence of abnormal Krebs cycle metabolites
and/or elevated abnormal carbohydrates , the child was offered treatment
of suspected yeast infection with Mycostatin (NystatinS) 100,000
units q.i.d. orally for 10 days and another random urine sample
obtained and analyzed for organic acids. If the second urine sample
still showed the presence of abnormal metabolites, the child was
offered a second course of treatment with Mycostatin for 2 months
and another urine sample tested. If this still showed abnormal organic
acids, the child was offered to enter treatment with fluconazole
(Diflucan) 2 mg/kg/day as a single dose daily for 2 weeks. A second
two-week course of treatment was offered if abnormal metabolites
were still present in the urine and liver function was satisfactory
based on serum transaminase activities before and after fluconazole
therapy. A random urine sample was obtained two weeks later and
analyzed for organic acids. An additional urine sample was collected
four weeks after the end of therapy to determine the duration of
any biochemical normality associated with drug therapy.
Each of the patient's baseline urine values served as controls.
Urine from 20 normal children of laboratory employees served as
additional normal controls. An additional 50 normal controls have
been collected and data will be available for the meeting. An assessment
of the severity of autistic behaviors was done by both a staff psychologist
working with a parent and by a teacher (if child was in school or
preschool) using the CARS scale (Childhood Autism Rating Scale)
This assessment was done twice: prior to starting therapy and at
the end of therapy. The use of two evaluators was employed to increase
the reliability of this test. Furthermore, parents and teachers
are the individuals who have usually observed the child longer than
any other people.
As indicated in Table 1, the mean values of all
of the above compounds with the exception of the phenylcarboxylic acid
compound decreased following 10 days of Nystatin therapy. The mean value
of the phenylcarboxylic compound actually increased by 89.3% following
Nystatin therapy (p=0.05 by paired t-test).
The percent decrease for dihydroxyphenylpropionic acid and furancarbonylglycine
is relatively small and is or marginal statistical significance. The percent
decrease for the additional compounds is larger (36.3-69.8%) with greater
degrees of statistical significance by the paired t-test. Please note
that four of the patient results are not included since these results
were late arriving but will be included in the final analysis.
Following 70 days of Nystatin therapy, mean values for
ten of the thirteen compounds decreased. The mean values for 3-oxoglutaric,
VMA analog, and phenylcarboxylic acid compound increased following 70
days of Nystatin therapy.
The percent decrease for dihydroxyphenylpropionic was
slight (11.8%) and was of marginal statistical significance. The percent
change of mean values compared to baseline ranged from 39.9-87.9 for
the remaining compounds with p values of the paired t-test ranging from
0.02-0.13. (The p value is the probability that the decreased values
are due to chance.)
The marginal decrease in dihydroxyphenylpropionic acid
led to the suspicion that this compound was not of fungal origin. Testing
was performed on several patients at the attending physician’s request
who had suspected or confirmed Clostridia infections and were treated
with metronidazole (Flagyl). We tested several of these patients before
and after metronidazole therapy and found a dramatic decrease in the
concentration of this compound from baseline in these patients after
drug therapy.
As shown in Table 3, there is a dramatic decrease in the
urinary concentration of dihydroxyphenylpropionic acid following the
administration of standard doses of the antibiotic Flagyl. In all four
patients, the concentrations of dihydroxyphenypropionic acid decreased
99% or more after two to three weeks on this drug. In the first patient
in the above series, dihydroxyphenylpropionic acid rapidly increased
following the cessation of Flagyl treatment.
Table
3 -Effect of Flagyl Therapy on Urinary
Excretion of Dihydroxyphenylprpionic Acid
Diagnosis
& Sex
Age
Length
of time (Days) from start of Flagyl Therapy
Urinary dihydroxyphenylpropionic acid*
Autism, male
4
0
435
6
184
16
1
21 (stopped Flagyl)
5
24
2
43
236
93
274
Previous C. difficile infection and
uncontrolled diarrhea, female
Thirteen of the parents and/or teachers of six of the
autistic children completed CARS evaluations both before and after therapy
with Nystatin.
The mean CARS score prior to therapy was 37.3 (sd= 4.2
) which is a rating of "severe autism" while the mean CARS
score after therapy was 32.6 (sd=5.1), a rating of "mild to moderate
autism."
This difference was rated as extremely significant by
the paired t-test (p< .001).
CARS Score
Diagnosis
15 - 30.0
Non-autistic
30.1 - 37.0
Mildly to moderately autistic
37.1 - 60
Severely autistic
Improvements cited by parents and teachers
include decreased hyperactivity, more eye contact, increased vocalization
, better sleep patterns, better concentration, increased imaginative
play, reduced stereotypical behaviors (such as spinning objects), and
better academic performance.
We were unaware of a possible role of metabolites of anaerobic
bacteria in autism until our formal research study was complete. The
overgrowth of anaerobic bacteria may be a complicating factor in the
use of antifungal therapy for treatment of autism.
In several cases , concentrations of dihydroxyphenylpropionic
acid increased following antifungal therapy. It is possible that anaerobic
bacteria may have proliferated when yeasts and/or fungi were reduced
in the microbial ecosystem.
If both yeast/fungal products and products of anaerobic
bacteria are involved in the mechanism of autism, more complex antimicrobial
therapies may be necessary to restore a balanced microbial ecology to
the gastrointestinal tract, perhaps by "reseeding" the gastrointestinal
tract with beneficial bacteria like Lactobacillus acidophilus. Such
therapy has proved effective in the treatment of individuals with recurrent
Clostridia difficile infections (14).
The marked decrease in dihydroxyphenylpropionic acid following
treatment with Flagyl, an antibacterial agent with specificity toward
anaerobic bacteria and no antifungal properties (15,16) is consistent
with the production of this compound by one or more species of anaerobic
bacteria.
Phenylpropionic acid and monohydroxyphenylpropionic acid
which are very closely related biochemically to this compound are produced
by several species of Clostridia (17 ). However we failed to identify
this compound in multiple culture media samples in which multiple species
of Clostridia were cultured.
Our failure to isolate dihydroxyphenylpropionic acid*
from pure cultures of Clostridia could be due to multiple reasons:
A precursor of the compound such as monohydroxyphenylpropionic
acid may be produced by the anaerobic bacteria and then be converted
to dihydroxyphenylpropionic acid by another microorganism and/or by
human metabolism.
The anaerobic bacteria producing this compound may
be difficult to grow in vitro.
The culture media for this organism may not provide
the nutrients needed for the biosynthesis of this compound by this
organism. We were very interested in a possible role in the mechanism
for autism for this compound because it is related biochemically to
the neurotransmitters dopamine and norepinephrine and because it is
an inhibitor of dopamine decarboxylase, the enzyme responsible for
the conversion of DOPA to dopamine (18).
* This compound has recently been definitiely identified as 3-(3-hydroxyphenyl)-3-hydroxy
propionic acid.
Shaw W, Kassen E,and Chaves E. "Increased excretion of analogs
of Krebs cycle metabolites and arabinose in two brothers with autistic
features." Clin Chem 41:1094-1104, 1995.
Kontstantareas M and Homatidis S. "Ear infections in autistic
and normal children." J Autism and Dev Dis 17:585, 1987.
Kennedy M and Volz P. "Dissemination of yeasts after gastrointestinal
inoculation in antibiotic-treated mice." Sabouradia 21:27-33,
1983.
Danna P et al. "Role of Candida in pathogenesis of antibiotic
associated diarrhea in elderly patients." Lancet 337: 511-14,
1991.
Ostfeld E et al. "Effect of systemic antibiotics on the microbial
flora of the external ear canal in hospitalized children." Pediat
60: 364-66, 1977.
Kinsman OS et al. "Candida albicans gastrointestinal colonization
and invasion in the mouse:effect of antibacterial dosing, antifungal
therapy, and immunosuppression." Mycoses 32:664-74, 1989.
Van der Waaij D. "Colonization resistance of the digestive
tract--mechanism and clinical consequences." Nahrung 31:507-17,
1987.
Samonis G et al. "Antibiotics affecting gastrointestinal colonization
of mice by yeasts." Chemotherapy 6: 50-2, 1994.
Samonis G et al. "Prospective evaluation of the impact of
borad-spectrum antibiotics on the yeast flora of the human gut."
European Journal of Clinical Microbiology& Infectious Diseases 13:665-7,
1994.
Samonis G et al. "Prospective evaluation of the impact of
broad-spectrum antibiotics on gastrointestinal yeast colonization
of humans." Antimicrobial Agents and Chemotherapy 37: 51-53,
1993.
Sumiki Y. "Fermentation products of mold fungi. IV.Aspergillus
glaucus.I." J Agr Chem Soc Jap 5 : 10, 1929.
Sumuki Y. "Fermentation products of molds." J Agr Chem
Soc Jap 7 :819,1931.
Kawarda A et al. "Biochemical studies on bakanae fungus."
Bull Agr Soc Jap 19: 84, 1955.
Gorbach S et al. "Successful treatment of relapsing Clostridium
difficile colitis with Lactobacillus" GG. Lancet 1987 ii: 1519
,1987.
Rollo I Miscellaneous drugs used in the treatment of protozoal infections,
pg .1151. In Goodman and Gilman’s The Pharmacological Basis of
Therapeutics. Fourth Edition.The MacMillan Co., NY, NY, 1970.
The Physician’s Desk Reference, pgs 2224-6, 1993.
Elsden S et al. "The end products of the metabolism of aromatic
amino acids by Clostridia." Arch Microbial 107: 283-8, 1976.
Hartman W et al. "Competitive inhibition of 3,4-dihydroxyphenylalanine
(DOPA) decarboxylase in vitro." J Biol Chem 216 : 507-529, 1955.
