Study type

Study topic

Disease /health condition
Human medicinal product

Study type

Non-interventional study

Scope of the study

Safety study (incl. comparative)

Data collection methods

Secondary use of data
Non-interventional study

Non-interventional study design

Cohort
Study drug and medical condition

Medicinal product name, other

Valproate and related substances: valproic acid, sodium valproate, valproate pivoxil, valproate semisodium, valpromide, valproate bismuth, calcium valproate, valproate magnesium.
Comparators: lamotrigine/levetiracetam.

Anatomical Therapeutic Chemical (ATC) code

(N03AG01) valproic acid
valproic acid
(N03AG02) valpromide
valpromide
(N03AX09) lamotrigine
lamotrigine
(N03AX14) levetiracetam
levetiracetam

Medical condition to be studied

Neurodevelopmental disorder
Congenital anomaly in offspring
Population studied

Short description of the study population

Inclusion criteria of the family units (i.e. family triads offspring-mother-father):
Pregnancies will be included if they meet all the following inclusion criteria: singleton pregnancies, with pregnancy outcome occurring during the study period with mothers’ continuous records in the database for ≥12 months prior to the date of their last menstrual period plus 2 weeks (LMP2) and with linkage to father within the study time period; fathers with continuous records in the database for ≥12 months prior to the LMP2 date of the linked mother, and treated with at least one anti-seizure medication (ASM) (valproate or lamotrigine or levetiracetam) from 12 months prior to LMP2 date, and up to 12 months after LMP2 date (this inclusion criteria will be modified for secondary outcomes and/or sensitivity analyses).

Exclusion criteria of the family units :
Family units will be excluded if any of the following criteria are met: pregnancies associated with in vitro fertilization (IVF), offspring being adopted, offspring with genetic disorder, offspring born to mother with a diagnosis of epilepsy or bipolar disorder (BD) prior to conception and/or during pregnancy, offspring born to mother treated with ASMs prior to conception and/or during pregnancy, offspring born to fathers who switch to, from or between ASMs during the window of interest, offspring born to mothers with infections resulting in major malformation and parents with CM related to chromosomal disorders.

Cohorts of offspring :
Two different cohorts will be derived:
- one cohort for NDD (primary outcome),
- one cohort for MCM (secondary outcome) excluding mothers exposed to teratogenic drugs during pregnancy and fathers exposed to teratogenic drugs during spermatogenesis.
The inclusion date (index date) will be the offspring birth date for the NDD cohort and the conception date for the MCM cohort. Offspring will be followed from index date up to 12 years of age for NDD, from index date to the first year of age for MCM.

Each individual offspring (live born or non-live birth for MCM) will be considered as the unit for analysis. If the father or mother has more than one offspring during the study period, the families can be included as many times as the number of considered offspring if the family unit is eligible for that pregnancy/child as long as inclusion/exclusion criteria are met. All variables collected for fathers and mothers will be re-calculated for each included pregnancy/offspring.

Age groups

  • In utero
  • Paediatric Population (< 18 years)
    • Neonate
      • Preterm newborn infants (0 – 27 days)
      • Term newborn infants (0 – 27 days)
    • Infants and toddlers (28 days – 23 months)
    • Children (2 to < 12 years)
    • Adolescents (12 to < 18 years)
  • Adult and elderly population (≥18 years)
    • Adults (18 to < 65 years)
      • Adults (18 to < 46 years)
      • Adults (46 to < 65 years)
    • Elderly (≥ 65 years)
      • Adults (65 to < 75 years)
      • Adults (75 to < 85 years)
      • Adults (85 years and over)

Special population of interest

Other

Special population of interest, other

Family unit (i.e. family triad offspring-mother-father)
Study design details

Study design

This will be a multi-country, real-world, population-based cohort study using routinely collected data from registries and databases in Denmark, Finland, Germany, Norway and Sweden.
The study will use all the data available to date in each country, allowing at least 1 year pre-inclusion period.

Main study objective

The primary objective is to estimate and compare the risk of NDD (as composite, all subtypes combined and by subtype) in offspring paternally exposed to valproate as monotherapy in the 3 months prior to conception (‘current users’) compared to offspring paternally exposed to lamotrigine/ levetiracetam as composite monotherapy in the same exposure window (‘current users’).
In the valproate exposed group, only monotherapy will be considered.
In the active comparator group, fathers exposed to lamotrigine monotherapy or levetiracetam monotherapy (‘composite monotherapy’) will be considered as the reference.

Setting

Large longitudinal patient-level registries with family linkage (father-mother-child) were selected for this study, representing five European countries: Denmark, Finland, Germany, Norway and Sweden.
Main characteristics of the chosen databases:
- Father infant linkage available for all databases
- High rate of linkage for live birth outcomes in most databases (>90%)
- High rate of father infant linkage with non-live birth outcomes expected
- Number of live births captured per year: between 40,000 and 100,000 in Danish, Finnish, Norwegian and Swedish National registries, and between 115,000 and 210,000 in the German database (GePaRD).
- Valuable information on key variables including on pregnancy characteristics, identification of siblings and birth sex; data on parents’ epilepsy and bipolar disorder diagnoses, study medication and indication;
- ASM exposure window classification possible;
- Valuable information on NDD (all data sources are exhaustive) and MCM.

Comparators

-The risk of NDD in offspring paternally exposed to valproate as monotherapy will be estimated and compared to offspring paternally exposed to lamotrigine/ levetiracetam as composite monotherapy in the same exposure window.
In addition to 'current users', two other exposure windows will be considered: ‘former users’ defined as fathers exposed up to 3 months prior to conception, but not during the 3 months prior conception (i.e. prior to spermatogenesis but not during), and ‘later users’ defined as fathers not exposed before conception, but exposed after.
-The risk of NDD in offspring paternally exposed to valproate as monotherapy will be estimated and compared to offspring paternally exposed to valproate as monotherapy between the different exposure windows.
-The risk of MCM in live born offspring paternally exposed to valproate as monotherapy in current users will be estimated and compared in live born offspring paternally exposed to lamotrigine/levetiracetam in current users.
-Only in the group of offspring paternally exposed to valproate: the risk of MCM in live and non-live born i) in ‘former users’ vs ‘later users’, and ii) in ‘current users’ vs ‘later users’ will be estimated and compared.

Outcomes

NDD outcomes:
- at least one code (ICD-10) of NDD (composite outcome),

NDD by subtype outcomes:
- at least one code per NDD subtype (possible diagnosis), or
- at least 2 codes of any subtype (co-occurence), or
- at least 2 codes within the same NDD subtype (confirmation), or
- at least 2 codes within the same subtype with ADHD specificity, 1 ADHD ICD code and at least 1 ADHD medication or no ADHD code but at least 2 ADHD medications,
from offspring birth date to the occurrence of the outcome

MCM outcome: at least one ICD-10 code of MCM recorded from live birth (or from conception) to 1 year of age (excluding MCM associated with maternal infections)

Data analysis plan

The following analyses will be performed:
Descriptive analysis: Description of the study population, offspring outcomes and follow-up will be conducted, by data source, for each exposure group and each exposure window.
Crude analyses of primary (NDD) and secondary (MCM) outcomes will be conducted.
In addition, adjusted comparative analyses and risk estimate will be conducted using propensity score methods:
-Selection of variables for propensity score (PS): each covariate will be assessed regarding its association with the outcome, and with the exposure. PS will be estimated using two different approaches: 1) Common PS models (same PS in all the countries for the main analysis) and 2) Database-driven PS models (PS specific to each country in a sensitivity analysis, to rule out residual confounding due to covariates not initially anticipated a priori).
-Risk of NDD estimation
-Risk of MCM estimation
Controlling for confounding: Models listed above will be performed on the whole population in each data source. To further assess confounding and robustness, results will also be stratified if numbers allow it, by:
· Indication for treatment,
· Sub-type of epilepsy,
· Calendar year of birth in suitable categories (e.g. 2 or 3 years).
· The risk factors parental NDD/MCM.
Landmark methods will be used to analyze such a long follow-up period, by choosing specific time points based on clinical relevance for diagnosis of NDD.
Sensitivity analyses: Multiple sensitivity analyses will be conducted to assess residual confounding and/or the robustness of the main analyses results
Meta-analysis: All comparative analyses will be considered for pooling to achieve a more precise estimate of the observed effect size and identify any potential country-specific patterns in the data