Journal reference: Pasternak B, Hviid A. Use of acyclovir, valacyclovir, and famciclovir in the first trimester of pregnancy and the risk of birth defects. JAMA 2010;304(8):859-866 ^[1]

Link: http://dx.doi.org/10.1001/jama.2010.1206

Published: 25 August 2010

Evidence cookie says...

No association was found between the use of aciclovir (Zovirax), or valaciclovir (Valtrex), in early pregnancy and major birth defects. Too few participants were exposed to famciclovir (Famvir) for meaningful analysis. Aciclovir is the most widely studied antiviral. It should be antiherpetic antiviral of choice in early pregnancy. Note: there are some limitations in this study that may underestimate birth defects. These drugs have not been definitively determined safe in pregnancy.

More details:

Article details

Study design:

retrospective population based cohort study

Study aim:

to investigate associations between aciclovir, valaciclovir and famciclovir use in the first trimester and risk of major birth defects

Methods summary:

• data obtained from national registries in Denmark between 1 January 1996 to 30 September 2008
• large nationwide historical cohort
• cohort of all live births established from Medical Birth Register, and used to calculate date of conception
• Prescription Drug Register used to obtain data of all community prescriptions for oral aciclovir, valaciclovir and famciclovir
• birth defects identified from National Patient Register
• individual patient level data available from all three registers and were linked by unique personal identification numbers (available in Denmark)
• statistics:
  • logistic regression used to calculate prevalence odds ratios (POR) with 95% confidence intervals
  • regression models for potential confounding factors
• study had 80% power to detect a 47% relative increase (POR 1.47) in the risk of birth defects in those exposed to any antiherpetic antiviral drug

Results summary:

• 837 795 live births included in cohort
• 2.4% overall were diagnosed with major birth defect in first year of life
• 2.2% (40/1804) of babies exposed to antiviral had a major birth defect
• 2.4% (19 920/835 991) of unexposed babies had a major birth defect
• prevalence odds ratio (POR) adjusted for confounders and other variables:
  • POR = 0.89 (95% CI 0.65-1.22)
  • there was no association with major birth defects in the children of women exposured to aciclovir, valaciclovir or famciclovir, compared to the children of women not exposed
• there was similarly no association detected in supplemental analyses of dermatological aciclovir
• early exposure of antiherpetic antivirals within 4 weeks pre-conception demonstrated a non-statistically significant trend toward increased risk of heart, eye and central nervous system defects

Study conclusion:

In this large nationwide cohort, exposure to aciclovir or valaciclovir in the first trimester of pregnancy was not associated with an increased risk of major birth defects.

Participants:

• mothers and their live born infants in Denmark between 1996-2008 included in national registry
• 837 795 live births included in the study
• majority of women were born in Denmark:
  • those exposed to antiherpetic drugs: ~ 91%
  • those not exposed: ~ 85%
• extensive demographic details and clinical characteristics of the mothers given in the study:
  • birth year
  • age at conception
  • level of education
  • socioeconomic class
  • birth place
  • place of residence at time of conception
  • parity
  • smoking status in pregnancy
  • history of birth defects in siblings
  • maternal diseases including diabetes, infectious disease in first trimester, history of sexually transmitted infection, history of anogenital herpes, immunodeficiency, high risk pregnancy
  • maternal use of antibiotic in first trimester
  • maternal use of antineoplastic or immunomodulating drug, 3 months before conception through the the first trimester
  • maternal use of oral corticosteroid in first trimester

Methodological weaknesses

• historical study extracting data from three separate registries
  • subject to systematic biases relating to data entry and use of the registries:
    • birth defect register only captured hospital level diagnosis
    • defects diagnosed in primary care after 1 year of age were not captured
    • drug prescription registries measuring medication use are likely to over-represent actual use of medication as there is no measure of compliance
    • it is probable that there is a systematic bias towards the null hypothesis
• majority of the participants were locally born Danish women
  • homogeneous population limited to one country limit generalisability of results to other ethnicities
• dataset does not include inpatient antiviral drug treatment
• absolute number of babies exposed to famciclovir and valaciclovir were small
  • in the case of famciclovir, the numbers were too small to make a meaningful comparison
• study did not include abortions
  • if there is an association between exposure and spontaneous abortion or planned abortion, this would not have been captured and thus bias towards the null

Methodological strengths

• potential confounding factors (maternal risk factors, and diseases during pregnancy) were examined
• alternative analyses conducted to test robustness of study findings
• this is the largest study to assess valaciclovir exposure

Biases and conflicts of interests

• funding source includes a grant provided by the Lundbeck Foundation (note: Lundbeck does not appear to be a manufacturer antiherpetic antiviral drugs)

Clinical relevance to primary health care

Prior to this study, there was little published research literature on the safety of antiherpetic antiviral drugs, valaciclovir (Valtrex) and famciclovir (Famvir) in particular. This study demonstrates the absence of an association between antiherpetic antiviral drugs as a group, aciclovir (Zovirax) and valaciclovir as individual agents, and major birth defects in a large historic population group with more than 800,000 live births.

There are some limitations to the evidence. Firstly, the absolute numbers exposed to valaciclovir is low, and to famciclovir very low (26 exposed pregnancies). Meaningful analysis of the safety of famciclovir cannot be made.  Secondly, it is probable that the use of registry data introduces a bias towards the null hypothesis; i.e., if there was a real association, it would be underestimated in the results. Lastly, the study had 80% power to detect a 47% relative increase in risk of major defects; it might not have detected a real increased risk of major defects of lesser magnitude.

Aciclovir remains category B3 ^[2] but is the most widely studied antiviral. It should be antiherpetic antiviral of choice in early pregnancy.

References

1. Pasternak B, Hviid A. Use of acyclovir, valacyclovir, and famciclovir in the first trimester of pregnancy and the risk of birth defects. JAMA 2010;304(8):859-866
2. eTherapeutic guidelines: Antibiotic version 14. November 2010

Editor: Michael Tam