Secretin: A treatment for autism?
by Ronald J. Kallen, M.D.
What is secretin?
What is secretin used for?
Can secretin affect the brain?
The complexity of the brain: A hypothetical scenario
Does secretin have a psychophysiologic effect?
A critical review of the Horvath article
Comment on Horvath article
The final verdict on secretin awaits more research
What is secretin?
Secretin is a hormone consisting of a single chain of 27 amino acids and is therefore referred to as a peptide hormone. (There are 20 different amino acids that can be "strung together" in the peptide chains of proteins. Secretin is a mixture of 12 of the 20 amino acids. The amino acid, leucine, accounts for 6 of the 27 in secretin.) In terms of molecular size, it is about one-twentieth the size of the common plasma protein, albumin. A hormone is a substance released into the blood stream, often in response to a particular stimulus, so that it can deliver a "message" to a more distant site or "target tissue." In the instance of secretin, specialized cells lining the upper small intestine (the duodenum) detect the arrival of the acidic contents of the stomach. The partially digested food has been acidified by the hydrochloric acid produced by cells in the lining of the stomach. Further digestion of food requires action by pancreatic enzymes that work best after the acidic stomach contents are neutralized by the alkaline digestive fluids produced by the pancreas. The release of secretin by the "S" cells of the duodenum is triggered by the acidic pH (usually a pH less than 4) of the stomach contents. The secretin enters the general circulation. As the blood containing secretin flows through the pancreas, secretin molecules bind to receptors on pancreatic acinar cells (the target tissue) causing increased production and secretion of alkaline digestive fluids. The net effect is that secretin travels a round-about route by way of the blood stream to exert an effect on pancreatic cells that are only several centimeters away from the duodenum, where it originated.
Although the acid pH of stomach contents is the main stimulus for secretin release, other stimuli for release include fatty acids, bile, and certain components of spicy foods (such as 1-phenylpentanol).
Everyone produces secretin, even persons with autism, with every meal (especially meals with a high content of protein). After each meal, as the stomach empties its contents into the upper small intestine, there is a pulse of secretin released into the blood. So, even persons with autism have daily exposure to secretin.
Return to Overview
What is secretin used for?
The main use of secretin has been as a provocative maneuver to test the ability of the pancreas to increase production of alkaline digestive fluid. This test is rarely done now by gastroenterologists. It is also used as a diagnostic test for a rare tumor (gastrinoma) in adults. The source of secretin for medical use is pork intestine. Human secretin has been synthesized and differs from porcine secretin by two amino acids (the other 25 amino acids in the peptide chain are the same as those in the porcine preparation). Both forms of secretin have equivalent potency in terms of their effects on the pancreas. Human secretin is not commercially available.
Commercially-available porcine secretin (Ferring Laboratories) has the potential for an allergic or acute hypersensitivity reaction. Ferring advises that a small test dose be injected first, especially if there is a history of allergy or asthma.
Secretin has no known therapeutic use for gastrointestinal disease. There is no known disease attributable to a deficiency of secretin. It is not clear that secretin has a physiological role outside of the gastrointestinal tract although there is preliminary evidence in rats that there may be secretin receptors in the brain. Studies in rats have also shown that brain cells have genetic information for directing the synthesis of a precursor of secretin. If secretin is actually produced by brain cells it is not yet clear that it plays an active role as a neuropeptide neurotransmitter in the central nervous system.
Return to Overview
Can secretin affect the brain?
For secretin produced by the small intestine to have an effect on the brain, the following sequence of events would have to occur: Secretin released into the circulation would have to cross the blood-brain barrier to reach brain cells. Although it does not seem likely that this would occur, since there is no known transporter for secretin across the blood-brain barrier, it would then have to "home in" on cell membrane receptors with a specific affinity for secretin. Secretin, like most other hormones, has a relatively brief duration of action. Presumably its putative neurobiological effect would be relatively short-lived, as is the case for the action of secretin on the pancreas where it is rapidly degraded within minutes by a membrane peptidase.
This typical behavior of a hormone is difficult to reconcile with the testimonials as to remarkable long-term benefits on behavior in children with autism after a single injection. This raises the unlikely consideration that, despite its brief biological half-life, it somehow brings about a sustained effect on behavior. This implies a "trophic" or growth-factor-type effect on cells of the central nervous system. In essence, brain cells would have to "turn on" production of a substance that, in some way, affects social cognition. Whether the substance is a protein (such as a membrane receptor) or a neurotransmitter, such a mechanism would require that the single injection of secretin leverage a sustained effect at some unknown point in the sequence from transcription through post-translational processing. The bottom line is that secretin, during its relatively brief bioactive existence, must somehow "switch on" sustained production of "something" that sustains the behavioral change.
Return to Overview
The complexity of the brain: A hypothetical scenario
The above is a superficial account of complex events in cell biology, extending from "switching on" a gene, to synthesis of gene products crucial to the neurobiological basis of social cognition. In an attempt to convey this complexity, the following is a purely hypothetical scenario: Suppose that a certain subclass of serotonin receptors was not optimally active in autism. (It is known that there are at least a dozen such receptors. (Link to a recent review of the serotonin system in autism). This unique subclass might be confined to a unique topology in the brain, perhaps the hippocampus/amygdaloid complex and certain social-cognition-specific extensions to other brain regions (cerebellum, frontal cortex, visual cortex, auditory processing area, etc.). A specific treatment for autism might interact with the transcriptional-translational process mentioned above to ultimately upregulate a particular subclass of serotonin receptors, perhaps by inserting receptors in the cell membrane in specific brain regions. Alternatively, the effect of secretin on the "second messenger" that transduces the signal actuated by serotonin-receptor interaction at the cell membrane might translate into changes in cell membrane stability; or calcium ion flux across cell membranes; or certain metabolic sequences in the cell interior. It is known that secretin acts on the pancreas via the second messenger, cyclic AMP. There are myriad possibilities and it remains for basic research to establish which one (or more than one) of these mechanisms are susceptible to therapeutic intervention.
However, the above scenario implicating an effect on serotonin receptors is just a "for instance." There is no known direct relationship between the serotonergic system and secretin in the central nervous system. It is of some interest that histochemical studies in some mammalian species have shown that secretin and serotonin co-localize in "entero-endocrine" cells of the small intestine.
Return to Overview
Does secretin have a psychophysiologic effect?
If the primary site of action of secretin is in the pancreas how does that translate to "improved language and social skills" claimed by some? We are all familiar with one principle psychophysiological effect of digestion: the sensation of relief from the tension of hunger, or outright pleasure, of a satisfying meal. This is referred to as satiety. Although other "gut" hormones have been referred to as "satiety hormones" it is not clear that secretin is one of them. Even if it played a role in satiety, this does not explain a more complex and sustained effect on language and social interaction.
Return to Overview
A critical review of the Horvath article
The only published report of secretin administration to persons with autism is:
Horvath K et al, Improved social and language skills after secretin administration in patients with autistic spectrum disorders. J Assoc Acad Minority Physicians 1998;9:9-15.
The authors describe the effects of secretin in 3 boys with autism and chronic diarrhea. All were subjected to a test of pancreatic function involving endoscopic placement of a cannula through the stomach into the duodenum for withdrawl of pancreatobiliary fluid after intravenous administration of secretin. The usual response to secretin is an increased rate of flow of fluid (up to four fold) and an increased alkaline (sodium bicarbonate) content of the aspirated fluid. The three subjects were found to have an apparently exaggerated increase in pancreatobiliary fluid flow rate and the expected increase in alkalinity. Biopsies showed mild inflammation of the esophagus (1 of 3), stomach (2 of 3), and the duodenum (3 of 3, but described as "minimal" in one). Pancreatic enzyme activities were normal. One had decreased mucosal lactase activity and probably had lactose intolerance as a basis for diarrhea (although the authors do not comment on this finding). In all three boys "chronic diarrhea" resolved within 5 weeks of secretin administration and all are said to have had "dramatic improvements in their behavior" including "eye contact and expressive language." All three received a second dose of secretin some months later and showed further improvement. The authors attribute these behavioral effects to secretin.
Return to Overview
Comment on Horvath article
All three subjects were said to meet DSM-IV criteria for autism and the diagnosis was "confirmed by pediatric neurologists experienced in evaluating pervasive developmental disorders." However, there is no mention of specific details of the diagnostic assessment. Each child had a "developmental/psychological evaluation prior to secretin administration." Again, there is no mention of the specific techniques used. Apparently two of the three had autistic regression. Following intravenous administration of secretin, evaluation was based on "notes of therapists and teachers," parent interviews, and videotape recordings. Two had a "more structured evaluation." Again, it is not clear if this was done both before and after secretin administration. The same comment applies to mention of the use of the Childhood Autism Rating Scale.
Case 1 is initially said to have been given the diagnosis of PDD-NOS after a multidisciplinary evaluation. Elsewhere he is described as having "changed from a low-functioning autistic child to a social, nonautistic, speech-delayed child." These inconsistencies cast some doubt as to whether or not this child was truly autistic in the first place.
No explanation is provided for the apparently sudden cessation of "chronic diarrhea" in at least two of the three boys following a single injection of secretin.
Return to Overview
The final verdict on secretin awaits more research
As of this writing (February 1999) there are no published controlled trials of the administration of secretin to persons with autism. The NIH has invited proposals for possible funding of studies of secretin in autism. (See NIH statement on secretin.)
In the absence of objective biological markers for social cognition, observations of behavior must be especially rigorous. The diagnosis of autism must be clearly established beforehand using autism-specific behavior rating scales and, perhaps, videotaped observations. Possible effects on behavior would then be assessed subsequent to secretin administration, at predefined intervals of followup. Again, autism-specific rating scales and videotapes should be used by experienced professionals who are "blind" as to which treatment the child received. Videotapes recorded before and after secretin must be done under similar circumstances, with the child engaged in the same tasks, if they are to be directly comparable.
The study design, as with any newly-proposed treatment, must include proper controls. It should be double-blind; that is, neither the patients (and their families) or the clinical investigators know if the subject received secretin or a placebo. A proper placebo-controlled trial means that each child is "blindly" allocated to receive either secretin or placebo. Given the difficulty of assessing behavioral effects objectively, it will not be sufficient to avoid a placebo-control group by giving secretin to all of the subjects and having each child "serve as his own control."
These comments assume that the circumstances of the study of Horvath et al, namely, general anesthesia, endoscopy, biopsies, and cannula placement, are not confounding variables to be taken into account. It is unlikely that an Institutional Review Board would approve a research protocol that carried the potential risks inherent in general anesthesia.
The study design should also assure that other factors are held constant. For example, if one child is also receiving intensive one-to-one behavioral intervention concurrent with secretin administration and another is not, they are not comparable. Another variable is the "subtype" of autism. It is possible that children who appeared "normal" during the first one to two years of life and then experienced "autistic regression" (which occurs in about 30 per cent) are a different autism subtype than those who had obvious developmental problems early in infancy.
It may take a few years to resolve the secretin matter. There are no "quick and dirty" shortcuts to the research process. If the research is not able to meet the high standards of peer-review of methodology and data analysis, then the data are meaningless.
Return to Overview on this Page
NIH statement on secretin
|Return to list of commentaries| |Links to other autism resources|