Abstract
Vegrim HM, Dreier JW, Alvestad S, Gilhus NE, Gissler M, Igland J, Leinonen MK, Tomson T, Sun Y, Zoega H, Christensen J, Bjørk MH. JAMA Neurol. 2022;79(11):1-10. doi:10.1001/jamaneurol.2022.2977. Epub ahead of print. PMID: 36156660. Importance: Women with epilepsy are recommended high doses of folic acid before and during pregnancy owing to risk of congenital anomalies associated with antiseizure medications. Whether prenatal exposure to high-dose folic acid is associated with increases in the risk of childhood cancer is unknown. Objective: To assess whether high-dose folic acid supplementation in mothers with epilepsy is associated with childhood cancer. Design, Setting, and Participants: Observational cohort study conducted with nationwide registers in Denmark, Norway, and Sweden from 1997 to 2017. Analyses were performed during January 10, 2022, to January 31, 2022. Mother-child pairs were identified in medical birth registers and linked with information from patient, prescription, and cancer registers, as well as with sociodemographic information from statistical agencies, and were categorized by maternal diagnosis of epilepsy. The study population consisted of 3 379 171 children after exclusion of 126 711 children because of stillbirth or missing or erroneous values on important covariates. Exposures: Maternal prescription fills for high-dose folic acid tablets (≥1 mg daily) between 90 days before pregnancy start and birth. Main Outcomes and Measures: First onset of childhood cancer at younger than 20 years. Cox proportional hazards models were used to calculate adjusted hazard ratios with corresponding 95% CIs, adjusted for potential confounders. Cumulative incidence at aged 20 years was used as a measure of absolute risk. Results: The median age at the end of follow-up in the study population of 3 379 171 children was 7.3 years (IQR, 3.5-10.9 years). Among the 27 784 children (51.4% male) born to mothers with epilepsy, 5934 (21.4%) were exposed to high-dose folic acid (mean dose, 4.3 mg), with 18 exposed cancer cases compared with 29 unexposed, producing an adjusted hazard ratio of 2.7 (95% CI, 1.2-6.3), absolute risk if exposed of 1.4% (95% CI, 0.5%-3.6%), and absolute risk if unexposed of 0.6% (95% CI, 0.3%-1.1%). In children of mothers without epilepsy, 46 646 (1.4%) were exposed to high-dose folic acid (mean dose, 2.9 mg), with 69 exposed and 4927 unexposed cancer cases and an adjusted hazard ratio of 1.1 (95% CI, 0.9-1.4; absolute risk, 0.4% [95% CI, 0.3%-0.5%]). There was no association between children born to mothers with epilepsy who were prenatally exposed to antiseizure medications, but not high-dose folic acid, and an increased risk of cancer (absolute risk, 0.6%; 95% CI, 0.2%-1.3%). Conclusions and Relevance: Prenatal exposure to high-dose folic acid was associated with increased risk of cancer in children of mothers with epilepsy.
Commentary
The headline from the recent manuscript by Vegrim and colleagues 1 caused a collective gasp among providers who specialize in the treatment of persons with epilepsy and gestational capacity (PWEGC). In this study from the Nordic Register-Based Study of Antiepileptic Drugs in Pregnancy (SCAN-AED) collaboration, the authors report an association between “high-dose” folic prescriptions to pregnant patients with epilepsy and an increased risk of childhood cancer in their children. This is a registry-based study based on the birth, pharmacy, and cancer registries in Norway, Denmark, and Sweden. The health data from these countries can be linked because each citizen has a unique personal identification number. “High-dose” folic acid was defined as at least one prescription for 1 to 4 mg/day of folic acid dispensed to a pregnant patient between 90 days before conception and birth. Of note, in these countries a pharmacy can dispense enough pills for an entire pregnancy with one prescription. The cumulative incidence of childhood cancer by 20 years in children of mothers with epilepsy who received these high-dose prescriptions was 1.5% (95% CI, 0.5%-3.6%). The baseline cumulative incidence in the general population was 0.4% (95% CI, 0.4%-0.4%) and in children of mothers with epilepsy without high-dose folic acid prescriptions it was 0.6% (95% CI, 0.3%-1.1%). In a within group comparison of mothers with epilepsy, high-dose folic acid use had an adjusted hazards ratio (aHR) of 2.7 (95% CI, 1.2-6.3) for childhood cancer. There were too few affected children to allow for comparison of multiple dose categories. When analysis was restricted to exposure to maternal epilepsy and supplemental folic acid at doses <3 mg/day, however, the risk of pediatric cancer was not statistically increased compared with those exposed to maternal epilepsy without a prescription for high-dose folic acid aHR 2.6 (95% CI, 1.0-6.9).
One thing that is curious about the study’s findings is that the children of mothers without epilepsy who received high-dose folic acid did not have an increased risk of cancer compared to the general population. This was true both in the group of mothers who took anti-seizure medications (ASMs) for other indications and in those who did not take ASMs. The discrepancy challenges the assumption that prescription folic acid by itself is the only issue here. There are a few between-group differences that may be important. Among those who received supplemental folic acid, the average daily dose was greater in pregnant patients with epilepsy compared to those without epilepsy (4.3 vs 2.9 mg/day). The authors did do an analysis of mothers in the general population prescribed >4 mg/day and did not find an increased cancer risk in their children. However, a limitation of the study is that it cannot account for compliance or the cumulative dose of folic acid consumed over a pregnancy, which might be greater in the epilepsy+ASM group. Another important difference within the maternal epilepsy group is that the folic acid + ASM group were on higher doses of ASMs compared to those who were not prescribed folic acid. Thus, supplemental folic acid prescriptions may be a marker of patients with more difficult to control epilepsy and it is possible that the higher doses of ASMs or something about the epilepsy plays a role in the pediatric cancer risk. Finally, it is also possible that there is some unmeasured factor in the pregnant patients with epilepsy (e.g., undiagnosed genetic predisposition) that contributed to the impact of folic acid. Future research should likely include genetic polymorphism affecting folic acid metabolism and folic acid receptor antibodies as well as B12 and folate levels.
In order to decide how to integrate this new information into clinical practice, we should examine the origins of the common practice of folic acid supplementation and specifically high-dose folic acid supplementation for patients with epilepsy planning to conceive. In 1991, a randomized study demonstrated that, compared to no periconceptual supplementation, 4 mg of folic acid/day reduced the risk of neural tube defects (NTD) in pregnant patients with a history of a NTD in a prior pregnancy. 2 The use of 4 mg in this study was based on convenience (there was a pill this size available) and there is some data that lower doses may be sufficient in this population. 3
Based on this benefits to families with a prior NTD, several organizations have recommend supplementation with 4 or 5 mg to prior to conception for groups considered high-risk, including patients taking ASMs. 3 The rationale for including patients taking ASMs as a high-risk group is that some ASMs (valproate and carbamazepine) have been associated with NTD. Furthermore, several ASMs including valproate as well as carbamazepine, lamotrigine, phenobartbital, and phenytoin interfere with folate metabolism. 4 Valproate in particular has been shown to decrease brain and placenta uptake of folic acid metabolites. 4 Never-the-less, no study has been able to demonstrate a clear benefit of folic acid supplementation in the prevention of NTD or other malformations in pregnant patients taking ASMs. Recently, the Australian Epilepsy Pregnancy Register was unable to demonstrate a significant dose-dependent effect of folic acid supplementation on malformation risk. 5
Although the practice of periconceptual folic acid supplementation for PWEGC originates in trying to prevent NTD, there are several lines of evidence that it is beneficial for other reasons. Periconceptual folic acid supplementation has been associated with a decreased risk of miscarriage 6 and preterm birth 7 in pregnant patients with epilepsy. Additionally, it has been associated with better cognitive outcomes in exposed children including higher full-scale and verbal IQ as well as fewer autistic traits. 8 –10 It is not known if these benefits are specific to pregnant patients with epilepsy or those taking ASMs, as similar associations have been seen in the general population. Again, however, when it comes to these additional benefits of periconceptual folic acid supplementation, no study has demonstrated a dose-related benefit. In fact, in the general population, daily folic acid supplementation >5 mg during pregnancy was associated with lower psychomotor development scores in exposed children at 1 year of age. 11
To summarize what we know about folic acid supplementation in pregnancy: The wrong dose of folic acid supplementation is 0 mg, but the right dose is unknown. In the absence of clear evidence, it may be best to take an approach somewhere in the middle. In the United States, a standard prenatal vitamin has 0.8 mg of folic acid. Folic acid is also available by prescription in 1 mg pills, which many clinicians prefer because of a presumption that patients are more likely to comply with a prescription than a vitamin. A reasonable strategy would be to recommend 1 mg of folic acid for PWEGC, adding a prenatal vitamin prior to conception. Some authors suggest patients taking valproate or carbamazepine should be prescribed more. It may also be reasonable to still recommend up to 4 mg to patients with a family history of NTD, especially in a prior child. The study by Vegrim et al should make us pause and re-evaluate our practices and where they come from. Not because one study should completely change practice (Indeed, this study will need to be replicated), but because it helps us question if more of a good thing is always better.
Footnotes
Acknowledgments
I would like to thank Drs Kimford Meador, Håkon Magne Vegrim, and P. Emmanuela Voinescu for discussing this article with me.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.