Jump to comment:

• Risk of developmental delay in children born to epileptic mothers treated with antiepileptic agents
  bertrand de Toffol
  Published on: 13 April 2016
• Delays caused by increased testosterone
  James M Howard
  Published on: 13 April 2016

• Published on: 13 April 2016

  Risk of developmental delay in children born to epileptic mothers treated with antiepileptic agents

  • bertrand de Toffol, Prof Neurology
  • Other Contributors:
    • catherine Billard, bertrand de Toffol and Elisabeth Autret-Leca

  Dear Editor,

  We read the retrospective study by Adab et al [1] with interest; this paper raises the issue of risk of developmental delay in children aged 6 years and over born to epileptic mothers treated with antiepileptic agents during pregnancy. More specifically, foetal exposure to valproate could involve significant risk of impairment of verbal IQ in such children, despite no change in full scale IQ.

  ...

  Show More

  Dear Editor,

  We read the retrospective study by Adab et al [1] with interest; this paper raises the issue of risk of developmental delay in children aged 6 years and over born to epileptic mothers treated with antiepileptic agents during pregnancy. More specifically, foetal exposure to valproate could involve significant risk of impairment of verbal IQ in such children, despite no change in full scale IQ.

  These data are of considerable importance for clinicians whose task it is to inform and treat women of childbearing age: we therefore wish to obtain as much detailed information as possible in addition to the data published in this article so as to better assess the type and degree of cognitive risk involved.

  The practical dilemma is perfectly set out by the authors: the risk of death of pregnant mothers associated with epileptic seizures during pregnancy is 10-fold higher than among the general population and prophylactic treatment against seizures is consequently indicated. The extent of foetal malformation can at present be accurately estimated overall (global risk 2 to 3 times greater than among the general population) with specific damage depending on individual drugs (neural tube defects with valproate and, to a lesser extent, with carbamazepine), which does not in itself rule out prescription of an antiepileptic provided adequate foetal monitoring is performed. However, if the risk of developmental delay were objectively demonstrated in children exposed to certain antiepileptic drugs, this could lead to reassessment of the benefit/risk ratio of the incriminated treatments during pregnancy, or even to avoidance of such treatment.

  The authors provide a particularly valuable discussion of the potential bias associated with the retrospective nature of their study. The efficacy of valproate in the treatment of both partial seizures and generalised seizures is well established [2]. The published study data support this efficacy: of the 63 women treated with valproate alone for all syndromes, 47 presented no seizures (or experienced only non- convulsive seizures), i.e. 75% success rate for satisfactory control of epilepsy during pregnancy.

  We feel it is essential that details of cognitive evaluation for children born to mothers treated with valproate based be provided, centred on the following questions:

  1. The IQ results are expressed as a mean value and confidence interval, making it awkward to assess the degree of inter-individual variation. It would be useful to provide the range of values recorded for each group

  2. VIQ is a composite measurement of verbal intelligence based on five subtests designed to investigate different areas: Information, Comprehension, Similarities, Vocabulary and Arithmetic. It would therefore be preferable to provide the individual scores in each of these subtests, at least for the group of children exposed to valproate exhibiting the greatest impairment. Specific impairment concerning one or more subtest results would constitute a strong argument supporting a direct effect of valproate. In contrast, a heterogeneous impairment profile would tend rather to point to multifactorial origin. Demonstration of a specific deficiency would be particularly revealing: a deficit in verbal reasoning (Similarities) would tend to suggest a direct effect on verbal intelligence of exposure to valproate, while such a conclusion would be much less clear-cut in the event of deficits in subtests more sensitive to socio-economic conditions (Information, Comprehension).

  3. The results shown in figure 1 are puzzling: 42% of children exposed to valproate had a VIQ of less than 70 (last two bars to the right of the figure). It would be of value to provide details of the precise cognitive profile of this subgroup together with the PIQ values. If the PIQ of these children (or certain of these children), already impaired in terms of verbal IQ, is also deficient, then we are dealing with mentally impaired children. As things stand, it is impossible to evaluate this point adequately, since the FSIQ and PIQ values are given indiscriminately for all children exposed to valproate. Precise data concerning this subgroup are required in order to determine the reality of Specific Language Impairment (if the PIQ of these children is unchanged), or to illustrate the heterogeneous nature of the different situations in the absence of specificity of cognitive profiles. Such marked impairment of verbal intelligence in a child aged over 6 years in effect generally has a negative impact on learning at school.
  Under these conditions, it is normally possible to establish a precise diagnosis. However, the precise diagnoses available for certain children exposed to valproate such as Asperger syndrome or dyspraxia are surprising since they normally correspond rather to impairment of non-verbal functions and the link with impairment of VIQ is far from clear.

  4. Certain results require further explanation:
  a) Why do children exposed to phenytoin have a higher VIQ than children in the other comparative groups? Is the difference significant?
  b) Why is it that in contrast with those exposed to valproate in monotherapy, the VIQ of children exposed to valproate in polytherapy does not differ significantly from that of unexposed children (Table 4). Does combined administration of certain antiepileptic drugs play a protective role? Are plasma concentrations of valproate lower in polytherapy than in monotherapy?

  5. As regards genetic factors, it is essential to take account not only of the intelligence level of mothers but also of familial history of learning disorders both in the group of children considered as normal (VIQ>80) and in children with impairments (VIQ<80) for each group exposed to valproate. The distribution of environmental factors (socio-cultural level of mothers and fathers, alcoholism among mothers) should be defined for each study group in view of their extremely harmful effects on cognition among children subject to such factors.

  6. The incidence of tonic-clonic seizures during pregnancy should also be specified for the specific subgroup of severely impaired children.

  Irrespective of the methodological problems posed by this study, it nevertheless highlights the importance of systematic monitoring of drugs following their initial market launch. Doubts concerning foetal exposure to medicines centre on risk of malformations, which may be detected rapidly at birth. However, systematic and centralised recording is necessary to demonstrate any increase over time in initially low risks. For instance, 15 years elapsed before any description and evaluation was offered concerning the incidence of neural tube defects associated with valproate [3], and there was a corresponding period of 27 years for carbamazepine following its market launch [4].

  Obviously, systematic study of any psychomotor and/or behavioural consequences of exposure of the foetal brain during its development to medicines displaying cerebral tropism is more difficult, since the observation period required is much longer (3 to 7 years). It was thus several decades after the initial market launch before evidence could be obtained regarding any retrospective suspicions surrounding older antiepileptic drugs (phenytoin, phenobarbital, valproate).

  This period of uncertainty could have been shortened by the introduction of systematic prospective monitoring of psychomotor and behavioural effects in children exposed in utero immediately after commercialisation in order to quantify the associated risks and compare them for different antiepileptic drugs. The absence of systematic prospective monitoring creates an unacceptable time lapse in risk evaluation. The data provided by Adab et al, although retrospective, cannot be concealed from existing mothers or future mothers treated with valproate. In all likelihood, the drug information centres (French regional pharmacovigilance centres) will now add to the risk of malformation for valproate, first reported, that of impairment of verbal intelligence quotient, even in the absence of change in FSIQ.

  In this context, none of the drugs proposed as an alternative to valproate, such as lamotrigine for instance, can be considered as presenting lower risk of retarded development in principle since there are no available studies. However, the absence of studies cannot be equated with absence of risk. While the wish expressed by the authors ("It is essential that adequately controlled prospective studies are established now to identify the level of risk for cognitive impairment in children of women taking both new and established AEDs during pregnancy") may well be utopian, the health authorities should nevertheless insist that all foetal exposure is accompanied by prospective monitoring of exposed children (psychomotor and behavioural testing between the ages of 5 and 7 years) immediately following the commercial launch of a medicine in order to evaluate the associated risks of developmental delay. Today, this risk is unknown for all new antiepileptic drugs.

  References

  (1) Adab, N, Kini U, Vinten J et al. J Neurol Neurosurg Psychiatry 2004;75:1575-1583

  (2) The Scottish Intercollegiate Guidelines Network (SIGN). Guideline 70, http:/www.sign.ac.uk/guidelines/published/index.html (accessed 9 August 2004).

  (3) Robert E, Guilbaud P. Maternal valproic acid and congenital neural tube defects. Lancet 1982;2(8304):937

  (4) Rosa FW. Spina bifida in infants of women treated with carbamazepine during pregnancy. N Engl J Med 1991;324(10):674-7

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 13 April 2016

  Delays caused by increased testosterone

  • James M Howard

  Dear Editor,

  I suggest the developmental delays and intellectual problems, exhibited by children of mothers who take valproate for epilepsy, are caused by increased testosterone.

  Learning disabilities have been connected with increased testosterone levels in children (Physiol Behav. 1993 Mar;53(3):583-6). Adab et al, determined that valproate increased the need for "additional educational nee...

  Show More

  Dear Editor,

  I suggest the developmental delays and intellectual problems, exhibited by children of mothers who take valproate for epilepsy, are caused by increased testosterone.

  Learning disabilities have been connected with increased testosterone levels in children (Physiol Behav. 1993 Mar;53(3):583-6). Adab et al, determined that valproate increased the need for "additional educational needs" but carbamazepine does not in children of epileptic mothers (J Neurol Neurosurg Psychiatry. 2001 Jan;70(1):15-21). Valproate increases testosterone levels in women but carbamazepine does not (Seizure. 2003 Sep;12(6):323-9).

  I suggest developmental delays and lower IQs in children of women taking valproate results from increases in inappropriate levels of testosterone during critical times of brain growth and development in utero.

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top