Fenfluramine treatment in infantile autism. Neurochemical, electrophysiological, and behavioral effects

A Systematic Review on Autism and Hyperserotonemia: State-of-the-Art, Limitations, and Future Directions

Hyperserotonemia is one of the most studied endophenotypes in autism spectrum disorder (ASD), but there are still no unequivocal results about its causes or biological and behavioral outcomes. This systematic review summarizes the studies investigating the relationship between blood serotonin (5-HT) levels and ASD, comparing diagnostic tools, analytical methods, and clinical outcomes. A literature search on peripheral 5-HT levels and ASD was conducted. In total, 1104 publications were screened, of which 113 entered the present systematic review. Of these, 59 articles reported hyperserotonemia in subjects with ASD, and 26 presented correlations between 5-HT levels and ASD-core clinical outcomes. The 5-HT levels are increased in about half, and correlations between hyperserotonemia and clinical outcomes are detected in a quarter of the studies. The present research highlights a large amount of heterogeneity in this field, ranging from the characterization of ASD and control groups to diagnostic and clinical assessments, from blood sampling procedures to analytical methods, allowing us to delineate critical topics for future studies.

P300 amplitude and latency in autism spectrum disorder: a meta-analysis

Autism spectrum disorder (ASD) is an early onset neurodevelopmental disorder. Evidence suggests that ASD patients have abnormalities in information processing. Event-related potential (ERP) technique can directly record brain neural activity in real time. P300 is a positive ERP component which can measure the neuroelectrophysiological characteristics of human beings and has the potential to discover the pathological mechanism of ASD. However, P300 studies on ASD patients are incongruent and the disparities may be caused by several factors. By searching PubMed, Embase and Cochrane Library databases, a meta-analysis of P300 component difference between ASD group and typically developed (TD) control group was conducted. Results of amplitude and latency of P3b and P3a from included studies were synthesized. Random effect model was chosen and standardized mean difference (SMD) was calculated. Subgroup analysis was used to identify the source of heterogeneity and to test the effect of different experiment factors. A total of 407 ASD patients and 457 TD controls from 32 studies were included in this analysis. Reduced amplitude of P3b was found in ASD group (SMD = -0.505, 95 % CI -0.873, -0.138) compared with TD group, but no difference of P3b latency, P3a amplitude, or P3a latency was found between groups. Subgroup analysis showed that oddball paradigm elicited attenuated P3b amplitude in Pz electrode among ASD subjects. This meta-analysis suggests ASD patients have abnormalities in P300 component, which may represent for deficits in cognition, attention orientation and working memory processing, particularly in the decision-making processing condition.

Abnormal serotonergic development in a mouse model for the Smith-Lemli-Opitz syndrome: implications for autism

The Smith-Lemli-Opitz syndrome (SLOS) is a malformation/mental retardation syndrome resulting from an inborn error in 3beta-hydroxysteroid Delta7-reductase (DHCR7), the terminal enzyme required for cholesterol biosynthesis. Using a targeting strategy designed to virtually eliminate Dhcr7 activity, we have created a SLOS mouse model that exhibits commissural deficiencies, hippocampal abnormalities, and hypermorphic development of serotonin (5-HT) neurons. The latter is of particular interest with respect to current evidence that serotonin plays a significant role in autism spectrum disorders and the recent clinical observation that 50% of SLOS patients present with autistic behavior. Immunohistochemical analyses have revealed a 306% increase in the area of 5-HT immunoreactivity (5-HT IR) in the hindbrains of mutant (Dhcr7-/-) mice as compared to age-matched wild type animals. Amount of 5-HT IR was measured as total area of IR per histological section. Additionally, a regional increase as high as 15-fold was observed for the most lateral sagittal hindbrain sections. In Dhcr7-/- mice, an expansion of 5-HT IR into the ventricular zone and floor plate region was observed. In addition, the rostral and caudal raphe groups exhibited a radial expansion in Dhcr7-/- mice, with 5-HT IR cells present in locations not seen in wild type mice. This increase in 5-HT IR appears to represent an increase in total number of 5-HT neurons and fibers. These observations may help explain the behavioral phenotype seen in SLOS, and provide clues for future therapeutic interventions that utilize pharmacological modulation of the serotonergic system.

Decrease in 6R-5,6,7,8-tetrahydrobiopterin content in cerebrospinal fluid of autistic patients

The levels of pterin compounds in the lumbar cerebrospinal fluid (CSF) of ten neurologically normal and twenty autistic children were measured by two different liquid chromatographic systems, a conventional and a newly developed direct method. Among pterin compounds 7,8-dihydroneopterin (NH2) and 6R-5,6,7,8-tetrahydrobiopterin (R-BH4) levels in autistic children were significantly reduced to 66.1 and 41.5%, respectively, of those found in the controls. The autistic children were divided into three subgroups according to their CSF homovanillic acid (HVA) level and age: the group with an elevated HVA level, the younger (< 7 year old) group with a normal HVA level, and the older (> 7 year old) group with a normal HVA level. In comparison, the further reduction was observed in NH2 and R-BH4 levels in the younger group and in the 7,8-dihydropterin (PH2) level in the group with an elevated HVA level. Thus, these results suggest that the endogenous biosynthesis of R-BH4 in the brain may be reduced in autistic children as compared with that in neurologically normal children.

Functional consequences of fenfluramine neurotoxicity

Male Sprague-Dawley rats were trained to discriminate the anorectic drug d,l-fenfluramine (2.0 mg/kg intraperitoneally administered) from its vehicle using a food-motivated (fixed-ratio 10 schedule) two-lever operant task. Once trained, doses of 0.5, 1.0 and 1.5 mg/kg fenfluramine tested 20 min after IP administration produced dose-responsive discrimination performance. Subsequently, noncontingent twice-a-day administrations of 1 ml/kg saline were made for 4 days and the dose-effect relationship redetermined on the 13th to 15th day after initiation of the chronic saline regimen. Results of these dose-response experiments indicated that there was no significant effect upon fenfluramine discrimination after multiple saline injections or after 10 days without training. Following four days of retraining, 6.25 mg/kg fenfluramine twice-a-day for four days was followed 10 days later by another dose-response determination. This purportedly neurotoxic regimen of fenfluramine significantly increased the rats' ability to discriminate fenfluramine. These results suggest the possibility that chronic release of serotonin or selective damage to serotonin-containing neurons produced by fenfluramine may lead to postsynaptic supersensitivity as manifested by the functionally increased discriminative performance observed.

A controlled crossover trial of fenfluramine in autism

We report a 12 month double-blind randomized crossover trial of fenfluramine in 20 children with the syndrome of autism. On active drug most of the children lost weight and blood serotonin levels fell by an average of 60%. There was a fall in urinary dopamine (DA) and noradrenaline (NA) levels and increased excretion of homovanillic acid (HVA). Some of the children showed improvement in tests of cognitive and language function, although the results did not achieve overall statistical significance. Event-related brain potentials (ERPs) were obtained in seven subjects on an auditory choice reaction time task. Side effects of the drug included irritability and lethargy. Fenfluramine may have a limited place in the management of some patients with autistic disorder.

Event-related potentials and monoamines in autistic children on a clinical trial of fenfluramine

In a double blind, crossover study of the response of autistic subjects to fenfluramine, event-related potentials (ERPs) were recorded from 7 subjects on an attention-demanding auditory choice reaction time task (ACRT). ACRT, IQ and biochemical measures were taken after 5 months placebo and 5 months fenfluramine treatment. After fenfluramine treatment blood serotonin levels fell, urinary catecholamine levels fell and the HVA/DA ratio rose. IQ and ACRT performance improved. On the ACRT subjects were asked to press a button to a rare target (500 Hz, P = 0.14) and to ignore higher pitched rare (2,000 Hz, P = 0.14) and frequent non-targets (1,000 Hz). After fenfluramine treatment N1 latencies increased. The scalp distribution of ERP maxima changes slightly with treatment. P3 maxima elicited by rare non-targets were recorded more rostrally after fenfluramine treatment. After rare non-targets N1 amplitudes at Fz decreased but P3 amplitudes at Pz increased. Early negativity after the rare non-target (particularly on the right side) was negatively correlated with the HVA/DA ratio. Subtraction of the P3 component elicited in a passive condition where no response was required from the active condition showed that P3 positivity to targets was halved with treatment. (In contrast Nd increased on fenfluramine treatment). Overall, N1 and P3 components showed greatest responsiveness to rare non-targets on fenfluramine. N1 but not P3 changes may represent slight improvement of attention-related function with treatment. Small changes in ERP latency and distribution, associated with the neuroleptic action of fenfluramine may be partly responsible for a mild improvement of IQ and ACRT performance on medication.

Effects of high-dose fenfluramine treatment on monoamine uptake sites in rat brain: assessment using quantitative autoradiography

Fenfluramine is an amphetamine derivative that in humans is used primarily as an anorectic agent in the treatment of obesity. In rats, subchronic high-dose d,l-fenfluramine treatment (24 mg/kg subcutaneously, twice daily for 4 days) causes long-lasting decreases in brain serotonin (5HT), its metabolite 5-hydroxyindoleacetic acid, and high-affinity 5HT uptake sites. Moreover, this high-dose treatment regimen causes both selective long-lasting decreases in fine-caliber 5HT-immunoreactive axons and appearance of other 5HT-immunoreactive axons with morphology characteristic of degenerating axons. Determination of the potential neurotoxic effects of fenfluramine treatment using immunohistochemistry is limited from the perspectives that staining is difficult to quantify and that it relies on presence of the antigen (in this case 5HT), and the 5HT-depleting effects of fenfluramine are well known. In the present study, we used quantitative in vitro autoradiography to assess, in detail, the density and regional distribution of [3H]paroxetine-labeled 5HT and [3H]mazindol-labeled catecholamine uptake sites in response to the high-dose fenfluramine treatment described above. Because monoamine uptake sites are concentrated on monoamine-containing nerve terminals, decreases in uptake site density would provide a quantitative assessment of potential neurotoxicity resulting from this fenfluramine treatment regimen. Marked decreases in densities of [3H]paroxetine-labeled 5HT uptake sites occurred in brain regions in which fenfluramine treatment decreased the density of 5HT-like immunostaining when compared to saline-treated control rats. These included cerebral cortex, caudate putamen, hippocampus, thalamus, and medial hypothalamus. Smaller, but nonetheless significant, decreases in density of [3H]paroxetine-labeled 5HT uptake sites were noted in brain regions in which partial sparing of 5HT-like immunoreactive fibers had been reported following fenfluramine treatment, specifically septum, lateral hypothalamus, and amygdala. In contrast, [3H]mazindol autoradiography revealed that total catecholamine (i.e., dopamine and norepinephrine) uptake sites in cerebral cortex, caudate putamen, and locus coeruleus, areas in which [3H]paroxetine-labeled 5HT uptake sites were significantly decreased, were unaffected by this fenfluramine treatment. These data support the hypothesis that subchronic, high-dose fenfluramine treatment causes selective degeneration of 5HT axons in rat brain. Since pharmacokinetic studies show that the dosing regimen used in this study exposes rat brain to concentrations of fenfluramine that are approximately 600 times greater than those resulting from the therapeutic oral dose, caution must be exercised in extrapolating these data to humans.

A two-year follow-up of autistic children treated with fenfluramine

Six children with autism, who were treated with fenfluramine as part of a multi-center double-blind placebo-controlled crossover design study, were continued on an open trial of this medication. Follow-up evaluation of these children after 27 months indicated that numerous problems arose in the management of these children, most often resulting in discontinuation of the medication. Particularly noted were development of tolerance, appetite and weight problems, and requirements for other kinds of interventions including introduction of other psychotropic medications and change in custodial circumstances.

Fenfluramine treatment of twenty children with autism

The effects of fenfluramine were examined on 20 children with autism over a 48-week period utilizing a double-blind placebo-controlled crossover design. Blood and urine samples and psychological tests (Griffith's Developmental Scales and Real Life Rating Scale) were obtained at each crossover period. The only significant improvement was a decrease in abnormal motor behavior. We did not find any significant improvement in intellectual functioning or any correlation between good clinical response and low baseline serotonin levels or high baseline IQ. Serotonin decreased 53% after fenfluramine treatment and rebounded to a level 35% higher than baseline following a placebo period. Fenfluramine and the active metabolite norfenfluramine were determined in plasma samples.

Electrophysiological effects of fenfluramine or combined vitamin B6 and magnesium on children with autistic behaviour

The authors compared the effects of fenfluramine or combined vitamin B6 and magnesium treatment on the evoked potential conditioning of 12 children whose autistic behaviour had improved clinically following treatment. The children who were clinically sensitive to combined vitamin B6 and magnesium developed a conditioning phenomenon and the fenfluramine-sensitive children showed an enhancement of the Cz evoked response amplitude. Results are discussed with reference to behaviour modifications observed during treatment.

Review of fenfluramine in the treatment of the developmental disabilities

Fenfluramine, a serotonin reducing agent, has been the subject of intense research effort in recent years. A variety of biochemical studies summarized suggest that some autistic children and many nonautistic severely retarded individuals have elevated blood serotonin concentrations. The research on fenfluramine's clinical efficacy is thoroughly reviewed from a methodological perspective. All studies assessing the drug's effects on blood serotonin have observed reductions in whole blood serotonin to about 50% of baseline concentrations. Although there were early reports of drug enhancement of IQ, there is no good evidence that this is the case. However, there are data to suggest that fenfluramine may enhance social relatedness, reduce stereotypic behavior, lessen overactivity, and improve attention span in some autistic children, although these results do not appear consistently across studies. The animal literature on the neurotoxicity of fenfluramine is reviewed, and a number of limitations in this research are identified that raise questions about its relevance to the pharmacotherapy of children.

Urinary dopamine metabolites as indicators of the responsiveness to fenfluramine treatment in children with autistic behavior

Modifications in serotonin and dopamine metabolism were evaluated in 13 children with autistic behavior and related to their responsiveness to fenfluramine treatment. A double-blind medication-placebo crossover design was used. Each patient received 1.5 mg/kg fenfluramine daily for 3 months followed and preceded by placebo for 1 month. Clinical improvement was observed in 6 children (responders). It included reduction of behavioral symptoms such as motor activity, anxiety, mood disturbances, and distractibility. Modifications of serotonin (5-HT), dopamine (DA), and DA metabolites [homovanillic acid (HVA) and dihydroxyphenylacetic acid (DOPAC)] were assessed at urinary levels. Responders and nonresponders showed a significant decrease of urinary 5-HT levels on fenfluramine. The main differences between the two groups of subjects were found with HVA, the major metabolite of DA. Fenfluramine significantly increased HVA levels in responders whereas no significant modification was found in nonresponders. Moreover the initial level of HVA (lower in responders) significantly differentiated the two groups. These results suggest that the clinical response to fenfluramine could be related to the dopaminergic action of this drug and that urinary DA metabolite levels could be considered as indicators of the responsiveness to fenfluramine treatment in children with autistic behavior.

Diagnosis of autism and other pervasive developmental disorders

Problems in the diagnosis of autism and other pervasive developmental disorders are reviewed. Modification and application of recent developments in neuroimaging, molecular genetics and population studies are presented. A prioritized list of needed investigations is discussed.

Clinical effects of fenfluramine on children with autism: a review of the research

A review of research studies published to date on the effects of fenfluramine on children with autism is presented. The current status of the fenfluramine research on children with autism is assessed. The review analyzed the methodological aspects of the research, the toxicity of fenfluramine, and the relationship between fenfluramine, neurotransmitter activity, cognitive ability, and subsequent behavioral change. The review of published data indicated that fenfluramine had positive effects on the reduction of hyperactivity and stereotypic behaviors in 33% of the subjects. The best responders were children with the highest baseline IQs. The conclusions address the need for appropriate subgrouping of autistic syndromes, which may lead to identification of responders to pharmacological treatments. The need for further study of the possible long-term adverse side effects of flenfluramine is noted. Further experimental research on the effects of fenfluramine on children with autism is endorsed.

Event-related potentials in autistic and healthy children on an auditory choice reaction time task

Event-related potentials (ERPs) were recorded from midline (Fz, Cz, Pz) and lateral sites (F3, F4, P3, P4) in autistic children (n = 7) and age-matched controls (n = 9) on an auditory choice reaction time task. Subjects were asked to press a button to an infrequent target (500 Hz, P = 0.14) and to ignore higher pitched infrequent (2000 Hz, P = 0.14) and frequent (1000 Hz) non-targets. Autistic subjects made twice as many errors of omission as controls and showed a higher criterion (beta) for targets. Maximum ERP peak amplitudes showed a more varied scalp distribution in the autistic group. N1 latencies were consistently shorter in the autistic group and in 3 subjects the target P3 latencies were markedly longer than for the controls. Compared to controls, the N1 amplitude of the autistic response was larger to the rare stimuli (particularly to non-targets). The amplitude of the P3 component was smaller in the autistic group (particularly to the target). The stimuli were also presented in a passive condition requiring no response. After subtraction of the waveform obtained in the passive condition from that obtained in the active condition or subtraction of the waveform elicited by the rare non-target from that elicited by the target, N1 target amplitude was larger in control than in autistic children. Autistic subjects showed more early negativity to the rare non-target at left frontal and a larger P3 to the target at right parietal sites. ERPs of autistic children are more responsive to stimulus features (e.g. high/rare non-target tone) and less responsive to their associations or meaningfulness (e.g. target P3). Attention-related ERPs of autistic children show signs of precocious (right dominance for P3) and delayed development (P3 not maximal at parietal sites).

Fenfluramine treatment of autism: relationship of treatment response to blood levels of fenfluramine and norfenfluramine

Nine children meeting DSM-III criteria for infantile autism were treated with fenfluramine hydrochloride or inactive placebo in a double-blind crossover trial that followed the protocol for the fenfluramine multicenter study. Parents of the two children who had had the highest fenfluramine blood levels wished to have their children continue on fenfluramine, although the improvement they saw could not be demonstrated on the various rating instruments employed. The results of the study, while providing minimal support in themselves for the effectiveness of fenfluramine, do raise the possibility that fenfluramine blood levels might be related to treatment response.

Blood serotonin levels in adults, autistic and non-autistic children--with a comparison of different methodologies

Recent interest in conditions associated with increased blood serotonin level has highlighted the need for consistency between assay methods to allow for more accurate delineation of serotonin variables. To this end, comparison was made between a spectrofluorimetric technique frequently used in the past and two potentially more specific high performance liquid chromatographic procedures. Normal ranges and diurnal variations for blood serotonin in adults, normal, autistic children and children with developmental dysphasia were also determined. No significant difference was found between serotonin level in blood drawn by simultaneous venepuncture and capillary (fingerprick) collection. Whilst there was no evidence of circadian rhythm, seasonal variation with mean blood serotonin levels significantly lower in summer than in two successive winters was suggested. Blood serotonin values in normal children tended to decline with increasing age. No similar maturational effect was apparent in autistic children. The mean level for autistic children in winter was significantly higher than that for normal children in the same season; despite this there was considerable overlap of blood serotonin levels between normal and autistic groups. Serotonin levels determined by the three different methodologies showed a high correlation but differed significantly: caution should be exercised when comparing blood serotonin results where different methods are employed.

Compulsive respiratory stereotypies in children with autistic features: polygraphic recording and treatment with fenfluramine

Syncopes due to compulsive respiratory stereotypies were studied in eight patients with autistic features. Most had been referred for episodes thought to be intractable epileptic seizures. Polygraphic recording showed two types of syncope, one induced by prolonged apnea and the other by a prolonged Valsalva maneuver. Fenfluramine, 1.5-3 mg/kg per day, was given in an open trial. In four of five cases with frequent Valsalva maneuvers, respiratory stereotypies and syncopes were suppressed for 2-18 months. Patients with periodic apneas were more severely retarded and had less clear benefit. Side effects consisted of dose-dependent sedation and mild weight loss which stabilized without interrupting treatment. We suggest that these syncopes are volitional and may be associated with pleasant sensations. A double-blind placebo-controlled trial of fenfluramine seems warranted in such patients.

Effects of fenfluramine on autistic individuals residing in a state developmental center

The effects of fenfluramine on 21 maladaptive behaviors in 20 autistic individuals were examined over a 9-month period utilizing a double-blind, cross-over, placebo-controlled design. Raters carried out time-sampled observations in the school and residence. In addition, videotaped data were collected in controlled settings and assessed by the raters at the conclusion of the study. Some individuals displayed negative side effects such as tension, agitation, insomnia, and sweating during the 16-week period they received fenfluramine. The results demonstrated that fenfluramine caused no significant reductions in maladaptive behaviors. The lack of any significant positive results from this medication and the side effects observed strongly indicate the need for caution in the use of fenfluramine with autistic persons.

No effect of chronic fenfluramine on the P300 component of the event-related potential

Five autistic children were administered a visual version of the oddball paradigm under both drug (chronic fenfluramine) and placebo conditions. The P300 component of the event-related potential (ERP) was not affected by fenfluramine, although the drug significantly lowered blood concentrations of serotonin. It was concluded that serotonin as a neurotransmitter does not play a major role in production of P300. This conclusion fits with the facts that: (1) the general functional role of serotonin in the nervous system is to globally inhibit the overt expression of behavior, and (2) P300 is generally insensitive to response selection processes.

Autism and pervasive developmental disorders: concepts and diagnostic issues

The purpose of this report is to bring up to date available information on the defining features and diagnostic issues relating to autism and related disorders. We review the validity of the syndrome based on our last review (Rutter, 1978; Schopler, 1978). Subsequent data have produced refinement in our understanding of both diagnostic criteria and the nature of the basic deficit. Controversies over both the boundaries and the heterogeneity within the autism syndrome are evaluated according to available evidence. Diagnostic rating instruments for expediting systematic sample selection are critiqued, and leads for new research directions are suggested.

Visual augmenting/reducing and P300 in autistic children

Visual event-related potentials were recorded from five male autistics and five matched controls. Sensory effects were investigated by having subjects passively view flashes of three different but equiprobable intensities (augmenting/reducing paradigm). Cognitive effects were examined by having subjects count infrequent, target, flashes of one intensity embedded within a series of frequent, nontarget, flashes of a different intensity (oddball paradigm). In the augmenting/reducing paradigm, the sensory N100 wave of autistic but not controls showed a significant increase in amplitude (augmenting) as flash intensity increased. The cognitive P300 wave of autistics did not differ from controls in the oddball paradigm. Unlike controls, autistics had an equally large P300 in the no-task augmenting/reducing paradigm. It is concluded that autistics may experience a degree of stimulus overload in the visual modality.

The organic etiology of infantile autism: a critical review of the literature

Various theories have been provided in the literature regarding the etiology of infantile autism. It seems that the biological causation dominates the thinking of mental health practitioners and researchers irrespective of discipline. Areas of research include the following: neuropathological studies of the brain, autopsies, electroencephalograms, epileptic seizures, brain lateralization, studies in asymmetry; neurochemistry; genetics; and pre-, peri- and postnatal factors. A critical analysis of these studies has indicated that the evidence of organic factors tends to be rather weak and furthermore has been found to be contradictory. It is felt that the major problem lies in the fact that a large number of investigators include in their sampling children with various mental and physical disabilities and label them autistic. In general, there seems to be little information regarding the selection and little background information is offered on these children. It is suggested that a standardized and world wide diagnostic system be constructed which will provide objective etiological results.