Systematic review of pharmacological treatments in fragile X syndrome

Fragile X syndrome, caused by a large expansion of a CGG trinucleotide repeat within the FMR1 gene located on chromosome X (q27.3) [1], is considered the most common cause of inherited mental retardation. It has been estimated that FXS syndrome affects approximately 1 in 4,000 males and 1 in 8,000 females [2]. However those estimates are based on population projections from cohorts of children with special education needs and thus they could underestimate the extent of clinical involvement as some individuals affected by the behavioural, emotional and/or learning disabilities of FXS could have IQs in the normal or borderline range [3, 4].

People with FXS have mental dysfunction that normally includes some degree of mental retardation, ranging from severe to mild. The dysfunction can also include Attention Deficit and/or Hyperactivity Disorder (ADHD) [5, 6], autism disorder or the autism spectrum disorder [7, 8] and speech and behavioural disorders, such as anxiety-related symptoms including Obsessive Compulsive Disorder-like and perseverative behaviours, emotional lability and aggressive or self-aggressive behaviours.

The rationale underlying pharmacological treatments in persons with FXS is diverse and it has been extensively covered elsewhere [9‐11]. After the discovery of FXS it was observed that chromosomes from people with fragile X needed to be cultured in solutions deficient in folic acid in order to reveal the defect and it was thought that maybe individuals with fragile X syndrome had a lack of folic acid in their bodies, or were unable to make optimal use of the folic acid they had. Therefore, it was argued that supplementing their dietary intake might remediate adverse developmental and behavioural effects of the condition and the first therapeutic trials focussed primarily on folic acid. More recently, pharmacological treatment strategies have been focussed mainly as supportive strategies designed to maximize social functioning. As behaviour in FXS can significantly impact functionality, symptom-based treatment of the most problematic behaviours in the individual with FXS can be quite helpful [12]. As many people with FXS had additional diagnosis of ADHD, stimulants like methylphenidate and dextroamphetamine-already efficacious in non FXS patients-have been assessed in people affected with FXS. In order to avoid some of the side effects observed in treatment with stimulants other researchers looked for a non-stimulant pharmacological treatment for ADHD symptoms in people with FXS and tested L-acetylcarnitine, a product that had been shown to inhibit in vitro the cytogenetic expression of the FXS-associated fragile site FRAXA [13]. CX 156, an AMPA receptor-positive modulator has been evaluated to study its effects on cognitive disability in people with FXS, given that recent understanding of defects in synaptic plasticity in the fmr1 knock-out KO mouse had led to the proposal of several pharmacological targets in FXS to attempt to normalize synaptic connectivity, including AMPA receptor activation [14].

To our knowledge, while several recently published articles include narrative reviews of this topic [10‐12, 15, 16], to date no comprehensive systematic review of the safety and efficacy of pharmacological interventions in people with FXS has been published.

In this systematic review we have not found reliable and solid evidence to support the recommendation of any specific medication for people with FXS. There are different reasons to justify this conclusion. Some reasons apply to all the treatments studied and others apply only to a specific intervention.

It is general the problem related with the small sample sizes of the studies, which can result in problems in two areas: increased probability of type 2 errors (positive effects found in some studies could have been statistically significant if the samples had been larger) and inadequate size to detect side effects with low incidence. It is also striking the fact that very few women have been included in the studies, which probably can be explained by the previously extended belief that the disease was a problem only in males. In relation to the small numbers of patients, and also to the scarcity of studies, it must be taken into account that FXS is a relatively new disease, the specific genetic mutation was identified in 1991, and that its prevalence is low, being included into the category of the so-called "rare diseases". It is understandable that few research groups, and with limited resources, have been willing and able to carry out clinical trials on people with FXS. It also takes time until research from animal models is translated to assessment in human beings [33].

Regarding to the design of the studies and the assessment of their potential risk of bias, only a few of them could be classified as being of acceptable or good methodological quality, although it must be taken into account that only two studies were published after the CONSORT statement on standards for reporting trials [34]. Indeed, many of the publications provide insufficient information to allow critical assessment of the methodological quality of the trials and proper evaluation of the risk of bias. So it is difficult to know whether a particular study is methodologically flawed in the design or whether the reporting of the methods is incomplete.

Concerning specifically folic acid, it must be taken into account that apart from one study [26] the trials that assessed folic acid versus placebo were crossover studies and that the possibility of a 'carry-over' of treatment effect from one period to the next cannot be discarded. A carry-over effect means that the observed difference between the treatments depends upon the order in which they were received; hence the estimated overall treatment effect will be affected (usually underestimated, leading to a bias towards the null) [35]. One study showed positive statistically significant results on IQ measurements in a small group of 8 prepuberal children, but not in older patients [25]. In younger children it cannot be discarded that normal neuro-developmental changes could explain impressive improvements in some children in periods of time as short as six months; and in that study most of the children with the largest positive effects also had significant improvements in the placebo phase.

Regarding stimulants, the only RCT that assessed their effect in 15 children with FXS also diagnosed with ADHD [5] found some positive effects of metylphenidate, but it can only be considered as an exploratory study due to the very short length of the treatment -one week-, the small sample size and the crossover design with lack of wash-out period between treatments.

In relation to L-acetylcarnitine (LAC) a recently published study in children with FXS also diagnosed with ADHD was classified by the authors as a phase II study [13]. They found statistically significant stronger reductions of hyperactivity and improvement of social behaviour in patients treated with LAC, compared with the placebo group, as assessed by Conners' Global Index-Parents (CGI-P) and Vineland Adaptive Behavior Scales - Survey Form (VABS). LAC doses in that study ranged from 20 to 50 mg/kg/day. A previous smaller study carried out by the same research group assessed LAC at a much higher dose, 100 mg/kg/day and found significant statistical differences in favour of LAC in Conners' Abbreviated questionnaire for parents, but not for teachers [32]. Even though it is difficult to put value on the impact in the quality of life and social or intellectual performance of affected children of changes in scales assessed in a subjective way [31, 36], the positive results observed with LAC merit replication, with appropriate samples sizes in children with FXS and ADHD, since no significant side-effects in the LAC group were reported.

However the mechanisms underlying the pathogenesis of FXS still remain unclear and the wide variety of treatments under evaluation probably reflects a lack of a single underlying pharmacological mechanism for the several impairments and symptoms associated with FXS. Several other medications (e.g., fenobam, lithium, aricept, memantine) that have been evaluated in open-label non comparative studies are now possible candidates for clinical controlled trials [11, 37‐40]. Other products (e.g. minocycline, alpha-tocopherol, melatonin) that have been being tested recently in animal models could open new prospects for future treatments [41‐43].

Nevertheless given that intellectual, behavioural, emotional and/or learning performance in people with FXS is strongly influenced by different social factors, pharmacological treatments must be understood and assessed in the context of other interventions being in place. Often proposed areas of non pharmacological interventions include modifications in the home environment, more-tailored behavioural interventions and classroom environments, language and occupational therapy, and attention to social factors. Unfortunately, as a recent review points out, few studies have been published on the effectiveness of behavioural or social interventions among patients with FXS [11, 44‐46]. So it is advisable to promote studies that assess the effect of combined pharmacological and non-pharmacological interventions. In particular, it should be tested whether combined treatments could be more beneficial if administered in the early years in children with FXS.

Compared to previous reviews on pharmacological interventions on FXS, this review is the first one done following standardized methodology for systematic reviews. It also provides readers with complete data on the main characteristics of the studies, including aspects relating with their methodological quality, and their results.

There is no robust evidence to support recommendations on pharmacological treatments in people with FXS in general or in those with an additional diagnosis of ADHD or autism. Available data are insufficient to draw firm conclusions or to make recommendations. Some positive results found in various studies should be confirmed by further parallel randomised trials, properly designed and with adequate statistical power, before these treatments could be translated to standard medical practice. New trials should include affected women and test whether pharmacological treatments could be more beneficial if administered in the early years in affected children, either alone or combined with non-pharmacological interventions.