Seasonal influenza vaccine is specifically recommended for women who are or who might become pregnant in the flu season. This special population is targeted for vaccination because pregnant women are at increased risks of serious complications if infected with influenza virus. Despite this recommendation, recent evidence indicates that still fewer than 50% of women in the United States are vaccinated during pregnancy ( MMWR Morb Mortal Wkly Rep. 2016 Dec 9;65[48]:1370-3 ).
Potential reasons for this lack of uptake are concerns about safety of the vaccine for mothers and fetuses ( Vaccine. 2012 Dec 17;31[1]:213-8 ). This has highlighted the need for systematic safety surveillance for influenza vaccination with each subsequent seasonal formulation. To that end, season-specific studies of birth and infant outcomes since the 2009 season have been conducted; findings have been generally reassuring ( Vaccine. 2016 Aug 17;34[37]:4443-9 ; Vaccine. 2016 Aug 17;34[37]:4450-9 ).
However, a recently published analysis of data from the Vaccine Safety Datalink (VSD) raised concern about the potential risk of spontaneous abortion following seasonal influenza vaccination ( Vaccine. 2017 Sep 25;35[40]:5314-22 ). The VSD is a collaborative project between the Centers for Disease Control and Prevention’s Immunization Safety Office and several U.S.-based health care organizations. The VSD monitors safety of vaccines and conducts studies about rare and/or serious adverse events following immunization. In their recent analysis, the authors used a case-control design to evaluate risks for spontaneous abortion in the 2010-2011 and 2011-2012 influenza seasons. A total of 485 cases of spontaneous abortion and 485 individually matched controls were selected from six geographically diverse health care plans across the United States. The a priori exposure window of interest was vaccination within 28 days before the event of spontaneous abortion.
The authors found a doubling of risk for spontaneous abortion within that 28-day exposure window, but no association if the vaccination took place outside that period. This was in contrast to null findings for a similar analysis that the same group had conducted for vaccination in the 2005-2006 and 2006-2007 seasons. Of further interest, the authors noted even higher risks among women who had also been vaccinated for influenza in the previous season (adjusted odds ratio, 7.7; 95% confidence interval, 2.2-27.3). The highest odds ratios were among women who had been vaccinated in the 2010-2011 season and had also been vaccinated with monovalent pandemic H1N1 vaccine in the 2009-2010 season (aOR, 32.5; 95% CI, 2.9-359.0).
The VSD findings raise interesting questions about the biologic plausibility of strain-specific risks for spontaneous abortion, and risks of receiving a second vaccine containing the same strain in a subsequent season. However, this study should be interpreted with caution. With respect to the overall finding of a doubling of risk for spontaneous abortion, this is inconsistent with previous studies. A systematic review of 19 observational studies, 14 of which included exposure to the 2009 monovalent pandemic H1N1 strain, noted hazard ratios or odds ratios for spontaneous abortion ranging from 0.45 to 1.23 and 95% confidence intervals that crossed or were below the null ( Vaccine. 2015 Apr 27;33[18]:2108-17 ). More recently, the Vaccines and Medications in Pregnancy Surveillance System investigators evaluated spontaneous abortion in pregnancies exposed to influenza vaccine over four seasons from 2010 to 2014 and found an overall hazard ratio of 1.09 (95% CI, 0.49-2.40).
However, there are a number of limitations that must be considered. Many previous studies, including the VSD analysis, could have had misclassification of exposure, especially in recent years when vaccines are often received in nontraditional settings. The VSD study findings could have been influenced by unmeasured confounding. For example, there could be differential vaccine uptake in women with comorbidities that are also associated with spontaneous abortion, such as subfertility and psychiatric disorders.
In summary, at present the data viewed as a whole do not support a change to the current recommendation that pregnant women be vaccinated for influenza regardless of trimester. However, these data do call for continued surveillance for the safety of each seasonal formulation of influenza vaccine, and for further exploration of the association between repeat vaccination and spontaneous abortion in other datasets.
Dr. Chambers is professor of pediatrics and director of clinical research at Rady Children’s Hospital and associate director of the Clinical and Translational Research Institute at the University of California, San Diego. She is also director of MotherToBaby California, a past president of the Organization of Teratology Information Specialists, and past president of the Teratology Society. She has no direct conflicts of interest to disclose, but has received grant funding to study influenza vaccine from the Biomedical Advanced Research and Development Authority (BARDA) in the Department of Health and Human Services, and from Seqirus Corporation.
