Pregnancy Outcomes in Patients Exposed to OnabotulinumtoxinA Treatment

Abstract

Background and Objectives

A previous publication of pregnancy outcomes in onabotulinumtoxinA-exposed mothers demonstrated that the prevalence of major fetal defects (0.9%, 1/110) was comparable with background rates in the general population. There is continued interest to better understand the safety of onabotulinumtoxinA during pregnancy. This analysis evaluated pregnancy outcomes after onabotulinumtoxinA exposure to provide a cumulative 29-year update.

Methods

The Allergan Global Safety Database was searched from January 1, 1990, to December 31, 2018. Data from women (younger than 65 years or unknown) during pregnancy or ≤3 months before conception treated with onabotulinumtoxinA were assessed to estimate birth defect prevalence rates of live births only from prospective pregnancies.

Results

Of 913 pregnancies, 397 (43.5%) were eligible with known outcomes. Maternal age was known in 215 pregnancies: 45.6% were 35 years or older. Indication was known in 340 pregnancies: most frequent were aesthetic (35.3%) and migraine/headache (30.3%). The timing of exposure was known in 318 pregnancies: 94.6% were before conception or during the first trimester. OnabotulinumtoxinA dose information was known in 242 pregnancies; most (83.5%) were exposed to <200 U. Of 195 prospective pregnancies with 197 fetuses, there were 152 (77.2%) live births and 45 (22.8%) fetal losses (32 spontaneous, 13 elective). Of 152 live births, 148 (97.4%) had normal outcomes and 4 had abnormal outcomes. Among the 4 abnormal outcomes, there were 1 major birth defect, 2 minor fetal defects, and 1 birth complication. The prevalence rate for overall fetal defects was 2.6% (4/152, 95% CI 1.0%–6.6%) and 0.7% (1/152, 95% CI 0.1%–3.6%) for major fetal defects (3%–6% in the general population). Among cases of live births and known determinable exposure times, there was 1 birth defect with preconception exposure and 2 with first-trimester exposure.

Discussion

Although subject to reporting bias due to the nature of the postmarketing database review, this 29-year retrospective analysis of safety data in pregnant women exposed to onabotulinumtoxinA demonstrates that the prevalence rate of major fetal defects among live births is consistent with the rates reported in the general population. Although there are limited data available for second-trimester and third-trimester exposure, this updated and expanded safety analysis provides important real-world evidence to health care providers and their patients.

Classification of Evidence

This analysis provides Class III data that demonstrate that the prevalence rate of major fetal defects among live births subsequent to in utero onabotulinumtoxinA exposure is comparable with the reported background rates.

OnabotulinumtoxinA (BOTOX/BOTOX Cosmetic; Allergan, an AbbVie Company, North Chicago, IL) is a potent therapeutic neurotoxin that causes muscle relaxation and can provide clinical benefit for conditions characterized by the inappropriate contraction of skeletal and smooth muscles, pain (i.e., chronic migraine), and the overstimulation of cholinergically innervated glands (i.e., hyperhidrosis).¹ To date, onabotulinumtoxinA has been approved by the US Food and Drug Administration (FDA) and in the European Union for the treatment of 12 therapeutic indications and 3 aesthetic indications (Table 1).^2,3 Many of these indications are neurologic conditions; therefore, neurologists are likely to account for a large proportion of health care providers who treat with onabotulinumtoxinA in clinical practice.⁴ The high prevalence of migraine in the female population suggests that many exposed to onabotulinumtoxinA are expected to be women of childbearing age.⁵ Recent reviews have concluded that there are limited data on the safety of onabotulinumtoxinA during pregnancy.^6-8 Taking into consideration that approximately 45% of pregnancies are unintended,⁹ onabotulinumtoxinA exposure may inadvertently occur before conception or during the early stages of pregnancy in women undergoing routine treatment. Therefore, a greater understanding of the safety of onabotulinumtoxinA during pregnancy would be informative to health care providers and their patients.

Table 1.

Approved Indications for OnabotulinumtoxinA Treatment^a

Given the additional pregnancy exposure since 2013 that was not included in the analysis conducted by Brin et al. in 2016² and the increased use of onabotulinumtoxinA in neurologic indications, this analysis was designed to replicate the methodology of that previous analysis including more recent data.

The objective of this analysis was to evaluate pregnancy outcomes after onabotulinumtoxinA exposure using an expanded data set to provide a cumulative 29-year safety update.

Methods

The Allergan Global Safety Database contains individual safety reports received from preapproval and postapproval sources worldwide, including Allergan-sponsored and partner-sponsored clinical trials, postauthorization studies, regulatory agencies, published literature, and postmarketing spontaneous reporting. The database includes prospective (fetal outcome unknown at initial reporting) and retrospective (fetal outcome known at initial reporting) pregnancies, in which the patient was exposed to onabotulinumtoxinA or botulinum toxin type A (BoNT/A) manufacturer unknown. When a pregnancy is reported, a minimum of 3 attempts are made to gather the following information: pregnancy exposure (including patient and treatment information), pregnancy outcome, consent to access medical information, and the physician's contact information (to gather additional pregnancy information).

For this analysis, the database was searched from January 1, 1990 (start of recorded cases), to December 31, 2018, for eligible reports. To be included in this analysis, pregnancies had to meet the following criteria: (1) eligible pregnancy, defined as a case in which onabotulinumtoxinA treatment occurred during pregnancy or ≤3 months before the estimated date of conception, and the patient's age was younger than 65 years (or unknown) and (2) known outcome, defined as a case in which the pregnancy had ended and birth type was known at the time of analysis. All pregnancies were reviewed by a medical safety physician to determine birth type, including live birth or fetal loss (resulting from spontaneous fetal loss or elective abortion and other causes of fetal loss [e.g., stillbirths]). Elective abortions may have been due to known prior exposure, and therefore, a conservative approach was used to evaluate fetal loss. In addition, the medical safety physician determined the presence or absence of fetal defect(s). Fetal defects were categorized as follows¹⁰: minor birth defect (defined as a birth defect with mostly cosmetic significance, rarely medically significant, or requiring surgical repair), major birth defect (defined as a birth defect with medical and/or social implications that often requires surgical repair), birth complication (defined as a complication during labor and/or delivery), genetic abnormality (defined as a pathogenic sequence variant or chromosome abnormality), and significant adverse event (defined as an event[s] considered to be significant by the medical reviewer). In addition to birth type and fetal defects, maternal characteristics were assessed, including age (grouped as younger than 35 years or 35 years or older [advanced maternal age]), indication, timing of exposure (grouped by stage of pregnancy), and dose used (grouped by 50 U increments). Owing to missing maternal characteristic data in some of the reports, the results are expressed as the percentage of individuals with known information. To determine whether maternal characteristics varied between prospective pregnancies with live births and those with spontaneous fetal loss, group differences were evaluated using χ² tests in SAS version 9.4 (SAS Institute, Cary, NC).

Owing to inherent bias associated with retrospective pregnancy reports,^11,12 only prospective pregnancies were used to estimate prevalence rates of fetal defects among live births, which is consistent with FDA guidance¹³ and the approach used in other reports.^11,14 Retrospective analyses of pregnancy outcomes have an inherent reporting bias caused by the tendency to report abnormal over normal outcomes, which can lead to a potential inflation of the true prevalence rate. The prevalence rate of overall and major birth defects for prospective pregnancies with live births were calculated using Poisson distributions in SAS and expressed as 95% CIs. This methodology is consistent with that used previously,² and the data reported in this manuscript are inclusive of, and expand on, those reported by Brin et al. in 2016.²

Standard Protocol Approvals, Registrations, and Patient Consents

Similar to other published analyses^2,15 using deidentified and anonymous data, informed consent was not required.^2,15

Data Availability

AbbVie is committed to responsible data sharing regarding the clinical trials we sponsor. This includes access to anonymized, individual, and trial-level data (analysis data sets), as well as other information (e.g., protocols, clinical study reports, and analysis plans), as long as the trials are not part of an ongoing or planned regulatory submission. This includes requests for clinical trial data for unlicensed products and indications.

These clinical trial data can be requested by any qualified researchers who engage in rigorous, independent scientific research, and will be provided following review and approval of a research proposal, Statistical Analysis Plan and execution of a Data Sharing Agreement. Data requests can be submitted at any time after approval in the United States and Europe and after acceptance of this manuscript for publication. The data will be accessible for 12 months, with possible extensions considered. For more information on the process, or to submit a request, visit the following link: abbvieclinicaltrials.com/hcp/data-sharing.

Results

Distribution of Pregnancies

In total, 913 pregnancies were retrieved (January 1, 1990–December 31, 2018) from the Allergan Global Safety Database (Figure). After excluding pregnancies that failed to meet the inclusion criteria (461 pregnancies did not have a known outcome [labeled “unknown”] and 55 were not eligible [labeled “ineligible” because the injection date was >3 months before the estimated date of conception]), 397 (397/913, 43.5%) eligible pregnancies with a known outcome were included in this analysis (Figure). Most eligible pregnancies reported exposure to onabotulinumtoxinA (357/397, 89.9%), with the remainder BoNT/A manufacturer unknown (40/397, 10.1%). Of the 397 pregnancies, 195 (195/397, 49.1%) were prospective and 202 (202/397, 50.9%) were retrospective. All BoNT/A manufacturer unknown were retrospective pregnancies.

Figure. Distribution of Pregnancies.

^aPregnancies in which onabotulinumtoxinA injection occurred >3 months before the estimated date of conception were considered ineligible.

Maternal Characteristics

Characteristics of pregnancies with a known outcome are presented in Table 2. Advanced maternal age (35 years or older) was observed in 45.6% of total pregnancies (98/215). Neurologic conditions were reported in 180/340 (52.9%) pregnancies, including migraine/headache (103/340, 30.3%) and movement disorders (41/340, 12.1%). The other most frequently reported indication was aesthetic (120/340, 35.3%). The most common dose categories of onabotulinumtoxinA were 100 to <200 U (40.1%, 97/242) and <50 U (32.2%, 78/242). Approximately 95% of onabotulinumtoxinA exposure occurred during the first trimester or before the estimated date of conception.

Table 2.

Characteristics of Prospective and Retrospective Pregnancies With a Known Outcome^a

Fetal Outcomes

Of the 195 prospective pregnancies (197 fetuses, including 2 sets of twins), there were 152 live births (152/197, 77.2%) (Figure). Most live births were normal (148/152, 97.4%); 4 were characterized as abnormal (4/152, 2.6%). Fetal loss was largely due to spontaneous fetal loss (32/45, 71.1%), and the remainder due to elective abortion (13/45, 28.9%). There were 4 cases of intrauterine death and 3 premature births. Of those live births with known fetal defects and known timing of exposure, 1 pregnancy was observed with onabotulinumtoxinA exposure during the preconception period (ventricular septal defect) and 2 pregnancies were observed with exposure during the first trimester (talipes equinovarus and cardiac murmur).

Of the 202 retrospective pregnancies (207 fetuses, including 5 sets of twins), there were 160 live births (160/207, 77.3%) (Figure). Similar to prospective pregnancies, most live births were normal (155/160, 96.9%); 5 had abnormal outcomes (5/160, 3.1%). Of the 202 retrospective pregnancies (207 fetuses), 22.7% had fetal loss (47/207). Fetal loss was largely due to spontaneous fetal loss (40/47, 85.1%), and the remainder due to elective abortion (7/47, 14.9%). There was 1 premature birth and no intrauterine deaths. Of those with known fetal defects and known timing of exposure, 1 was exposed to onabotulinumtoxinA during the preconception period (laryngomalacia) and 2 were exposed during the first trimester (tracheoesophageal fistula/esophageal atresia and brain neoplasm).

Fetal Loss

Among all prospective and retrospective pregnancies, 69 (72 fetuses, including 3 sets of twins) resulted in fetal loss due to spontaneous fetal loss (cases with known data shown in eTable 1, links.lww.com/WNL/C790; see Figure for flowchart). Of the total pregnancies, a higher proportion of women who experienced spontaneous fetal loss were of advanced maternal age (35 years or older; 32/53, 60.4%). The most frequently treated indications were aesthetic (35/66, 53.0%), migraine/headache (16/66, 24.2%), and movement disorders (9/66, 13.6%). Most onabotulinumtoxinA exposure occurred during the first trimester (41/52, 78.8%). The most frequently reported dose was <50 U (21/44, 47.7%). Approximately half of pregnancies reported a gestational age of 1 to <2 months at the time of fetal loss (26/49, 53.1%).

A comparison of maternal characteristics in prospective pregnancies with live births (n = 152) vs prospective pregnancies with spontaneous fetal loss (n = 32) (Table 3) revealed a nonsignificant statistical trend for patients who experienced spontaneous fetal loss to be of advanced maternal age (35 years or older; 14/24, 58.3%) compared with those with live births (41/108, 38.0%; comparison, p = 0.067). No statistical differences were found between prospective pregnancies of live births and spontaneous fetal loss for timing of onabotulinumtoxinA exposure (p = 0.990) or dose (p = 0.205).

Table 3.

Comparison of Characteristics Between Prospective Pregnancies With Live Births and Spontaneous Fetal Loss^a

For completeness, the characteristics of eligible prospective (n = 434) and retrospective (n = 22) pregnancies with known and unknown fetal outcomes are presented in eTables 2 and 3 (links.lww.com/WNL/C790). In brief, the pregnancies with and without known outcomes had similar maternal characteristics, aside from a trend toward lower onabotulinumtoxinA doses in the latter group.

In all prospective and retrospective pregnancies, there were 20 elective abortions (pregnancies with known data presented in eTable 3, links.lww.com/WNL/C790; see Figure for flowchart). In contrast to the higher maternal age observed in pregnancies of spontaneous fetal loss, elective abortions were more common in women younger than 35 years (9/15, 60.0%). The following indications were observed most often among pregnancies with elective abortion: aesthetic (5/16, 31.3%), migraine/headache (4/16, 25.0%), and urological disorders (3/16, 18.8%). Similar to spontaneous fetal loss, most onabotulinumtoxinA exposure occurred during the first trimester (14/17, 82.4%). The most frequently reported dose was ≥200 U (5/12, 41.7%). The most common gestational age at the time of elective abortion was 1 to <2 months (6/15, 40.0%). The most common reasons reported for elective abortion were personal/social (8/15, 53.3%) and fetal disorder (4/15, 26.7%). There were 4 fetal defects reported in elective abortion cases (2 prospective cases, 2 retrospective cases). Both prospective cases reported timing of exposure as first trimester (agenesis of corpus callosum and Down syndrome); one of the retrospective cases reported timing of exposure as first trimester (Down syndrome), and the other retrospective case had unknown timing of exposure (neural tube defect).

Fetal Defects

Of all prospective and retrospective pregnancies (n = 397), 13 fetal defects were reported (Figure; Table 4). There were 6 major fetal defects: 2 prospective pregnancies (1 live birth, 1 elective abortion) and 4 retrospective pregnancies (3 live births, 1 fetal loss). Of the prospective pregnancies, ventricular septal defect was reported in a case of live birth, which was asymptomatic and did not require intervention, and a case of agenesis of the corpus callosum, in which the fetus was electively aborted at 7-month gestation. Of the retrospective pregnancies with live births, there was a report of tracheoesophageal fistula/esophageal atresia, which was successfully repaired with surgery, a case of cleft lip and cleft palate, and a case of diaphragmatic hernia, which underwent surgical repair and required prolonged intubation. Finally, a retrospective case of neural tube defect resulted in an elective abortion at week 19 of gestation.

Table 4.

Summary of Fetal Defects in Prospective and Retrospective Pregnancies

Overall Prevalence Rates of Fetal Defects

Consistent with FDA guidance¹³ and other reports,^11,14 only prospective pregnancies were used to estimate prevalence rates of fetal defects in live births (N = 152). In this analysis, the prevalence rate for overall fetal defects among live births (including all defects regardless of severity) was 2.6% (4/152, 95% CI 1.0%–6.6%) and for major fetal defects was 0.7% (1/152, 95% CI 0.1%–3.6%) (see eTable 4, links.lww.com/WNL/C790). In those with known timing of exposure, the prevalence rate of fetal defects in those exposed during the preconception period was 3.1% (1/32, 95% CI 0.6–15.7) and in those exposed during the first trimester was 2.0% (2/98, 95% CI 0.6–7.1).

Neurologic Conditions

Approximately 65% (220/340) of pregnancies had a therapeutic indication, of which 81.8% (180/220) were neurologic (see eTable 5, links.lww.com/WNL/C790). A total of 57.2% (103/180) and 22.8% (41/180) of those with neurologic conditions were being treated for migraine/headache and movement disorders, respectively. Of prospective pregnancies with any neurologic indication, 2 were characterized as abnormal (2/74, 2.7%). Similarly, of retrospective pregnancies with any neurologic indication, 2 were characterized as abnormal (2/106, 1.9%).

This analysis provides Class III data that demonstrate that the prevalence rate of major fetal defects among live births subsequent to in utero onabotulinumtoxinA exposure is comparable with reported background rates.

Discussion

This manuscript summarizes pregnancy outcomes after onabotulinumtoxinA exposure using the Allergan Global Safety Database to provide a cumulative 29-year safety update. In total, 913 pregnancies were retrieved from the database, with 397 pregnancies eligible with known outcome, which were nearly evenly distributed between prospective (n = 195) and retrospective (n = 202) pregnancies. Analysis of maternal characteristics revealed that age was slightly skewed toward women younger than 35 years compared with those of advanced maternal age (35 years or older). The most common indications were aesthetic and migraine or headache, with onabotulinumtoxinA dosing consistent with typical use.^3,16 Most onabotulinumtoxinA exposure occurred during the first trimester, followed by preconception exposure. There were 3 fetal defect events among those with preconception exposure and 5 among those with first-trimester exposure.

This retrospective safety analysis expands on a previous publication² that summarized safety data from the Allergan Global Safety Database over a 24-year period and included 232 eligible pregnancies with known outcomes. This 29-year update, encompassing the previous data, includes 397 eligible pregnancies with known outcomes. This updated analysis added 165 eligible pregnancies (58 prospective and 107 retrospective), an approximate 71% increase. In the original report, aesthetic was the most common reason for using onabotulinumtoxinA (50.5% of total pregnancies), followed by movement and pain disorders (16.8% and 14.2%, respectively). In the current manuscript, therapeutic treatment with onabotulinumtoxinA (64.7%) was more common than aesthetic treatment (35.3%). Neurologic conditions (52.9%), such as migraine/headache (30.3%) and movement disorders (12.1%), were the most frequently reported therapeutic indications. In this updated analysis, there was a large increase in the number of migraine or headache cases (n = 103) compared with the previous study (n = 24).² OnabotulinumtoxinA was approved for the treatment of chronic migraine in 2010, allowing for 8 years of postapproval data in the current manuscript compared with only 3 years in the original. Taking into consideration that the prevalence of migraine is the highest in women of reproductive age, women report migraines negatively affecting their plans for pregnancy, up to 80% of women who had migraines before conception will continue to have manifestations during pregnancy requiring intervention, and women with chronic migraine who discontinued onabotulinumtoxinA treatment during pregnancy showed a relapse in their condition,^17,18 this expanded analysis can provide important real-world evidence to health care providers and patients.

Most fetal outcomes from prospective (77.2%) and retrospective (77.3%) pregnancies resulted in live births. The incidence of spontaneous fetal loss was 16.2% (32/197) for prospective pregnancies and 19.3% (40/207) for retrospective pregnancies. In the general population, spontaneous fetal loss is expected to occur in approximately 10%–28% of known pregnancies,¹⁹ with increased risk with advanced maternal age.²⁰ In this analysis, the incidence of fetal loss due to elective abortion was 6.6% (13/197) for prospective pregnancies and 3.4% (7/207) for retrospective pregnancies. According to surveillance data from the Centers for Disease Control and Prevention (CDC), the rate of US legal-induced abortion is 11.6 abortions per 1,000 women (aged 15–44 years).²¹ Direct comparisons between the abortion rate found in this analysis and the US rate are not possible due to differences in how the denominators were defined. In this analysis, the denominator included only pregnant women, while the CDC denominator included all women of childbearing potential (aged 15–44 years).

In this analysis, the prevalence rate of overall fetal defects in prospective live births, which includes all defects regardless of severity, was 2.6% (4/152, 95% CI 1.0%–6.6%) and for major fetal defects was 0.7% (1/152, 95% CI 0.1%–3.6%). Cases of live births and known determinable exposure times were 1 birth defect (1/32, 3.1%, 95% CI 0.6–15.7) with preconception exposure and 2 birth defects (2/98, 2.0%, 95% CI 0.6–7.1) with first-trimester exposure. In addition, an exploratory sensitivity analysis was conducted that included pregnancies with unknown birth outcomes. The prevalence rate was 3.9% (6/154, 95% CI 1.8%–8.2%), suggesting that including these data, which assumed that a fetal defect occurred in unknown cases, had no effect on the prevalence of fetal defects.

Published fetal birth defect rates vary between countries and reporting centers. The CDC reported that major structural and genetic birth defects occur in approximately 3% of US births.²² The Texas Birth Defects Registry reported structural or chromosomal birth defects in 4.3% of live births.²³ Registry data from the United Kingdom found a slightly lower incidence, with major and system-specific congenital anomalies occurring in approximately 1.7%–2.0% of children younger than 1 year.²⁴ The March of Dimes Foundation estimates that 6% of total births worldwide have a serious birth defect of genetic or partially genetic origin.²⁵ Given these data, the prevalence rate of major birth defects in live births found in this analysis (0.7%) is within/below that observed in the general population. Compared with the previous analysis² (overall defects: 2.7% [3/110, 95% CI 0.6%–8.0%]; major defects: 0.9% [1/110, 95% CI 0.02%–5.1%]), the prevalence rates in this expanded analysis were slightly lower than those reported previously.

Of 397 pregnancies with known outcomes, 13 fetal defects were reported. No discernible patterns in characteristics were observed, as fetal defects were reported for prospective and retrospective pregnancies, in cases of live birth and fetal loss, and in patients treated for a variety of indications. In addition, no consistent types of malformation by organ were observed. Compared with the previous analysis,² 4 additional fetal defects (major) were identified: corpus callosum agenesis (prospective elective abortion), cleft lip and cleft palate (retrospective live birth), diaphragmatic hernia (retrospective live birth), and neural tube defect (retrospective elective abortion).

Published reports of fetal exposure to botulinum toxin from botulism (17 publications describing 19 pregnant women, which have been summarized previously^2,26) found no evidence of infantile botulism nor any direct adverse effects on the pregnancy or fetus. A literature search of BoNT/A treatment during pregnancy (or up to 3 months before conception) for therapeutic or aesthetic indications, which may be inclusive of the pregnancies reported in this analysis, revealed 21 publications describing 109 women and 118 pregnancies.^18,27-46 Some reports described other BoNT/A products or did not specify the manufacturer, and we acknowledge the noninterchangeability of these findings. Of the published pregnancies, there were 1 report of a therapeutic abortion²⁹ and 5 reports of spontaneous fetal loss.^{18,28,29,36,46} Of the spontaneous fetal loss, 2 were in women with a history of previous miscarriage without BoNT/A treatment^28,29; 1 was “a result of an inherent anembryonic pregnancy, most likely without any relation to the BoNT/A injection”³⁶; and the last 2 were medically unremarkable.^18,46 There were 2 reports of premature birth, with one “not thought to be related to the drug”³⁹ and the other was “within background average statistics.”⁴⁶ In addition to the pregnancies above, a recent retrospective study evaluated the safety of onabotulinumtoxinA for the treatment of chronic migraine during pregnancy and concluded that all 10 pregnancies assessed resulted in “live full-term births to healthy babies with no organ malformations.”⁴⁷

A significant strength of this analysis, which increases the external validity of these results, is the large number of pregnancies within the Allergan Global Safety Database. Nearly 400 eligible pregnancies across both aesthetic and therapeutic indications from the United States and international sources are described in this manuscript. The safety of onabotulinumtoxinA during pregnancy remains an important topic for health care providers and their patients, and this update provides additional real-world information. Owing to differences in manufacturing, formulation, and potency evaluation among BoNT/A products, which can affect clinical dosing, duration of action, and adverse events, these onabotulinumtoxinA safety data are not interchangeable with those of other products.^48,49

As is common to retrospective database analyses, this analysis has several limitations. First, to ensure potentially important safety data were not overlooked, reports of BoNT/A manufacturer unknown were included in this analysis; therefore, data in this manuscript may not be exclusive to pregnancies associated with onabotulinumtoxinA exposure. Nevertheless, none of these cases were in the prospective cohort. Second, this analysis is restricted to data within the Allergan Global Safety Database, which includes only reported pregnancies and does not represent all onabotulinumtoxinA-exposed pregnancies. In addition, the database may not contain full details, limiting in-depth analyses on risk factors in this population, including concomitant medication exposure, comorbidities, and other potentially relevant covariates. Third, most onabotulinumtoxinA exposure data are from the first trimester, which is important, as it is the period of most organogenesis; however, consequently, less is known about onabotulinumtoxinA exposure throughout the later stages of pregnancy. Based on preclinical data, it is unclear whether a level of potential systemic onabotulinumtoxinA exposure would be significant enough to affect the fetus of treated mothers.⁵⁰ Fourth, exposure was primarily (83.5%) <200 U (per labeling, the maximum dose is 400 U), so the effects of higher doses could not be evaluated. Most exposed mothers were treated for neurologic and aesthetic conditions, which may limit generalizability. Fifth, abnormal outcomes were assessed at the time of birth, but data for later time points during postnatal development were often unavailable. This was not a comparative or controlled study, and although the analysis adds to the evidence base, conclusions about the safety of onabotulinumtoxinA during pregnancy should be made with caution and continue to be monitored.

This 29-year retrospective analysis of safety data in onabotulinumtoxinA-exposed mothers demonstrates that the prevalence rate of major fetal defects among live births is consistent with rates reported in the general population. Most of the exposures were in the preconception period and first trimester of pregnancy, with limited data available for second-trimester and third-trimester exposure. This updated and expanded safety analysis provides important real-world evidence to health care providers and their patients.

Glossary

BoNT/A

botulinum toxin type A

CDC

Centers for Disease Control and Prevention

FDA

US Food and Drug Administration

Appendix. Authors

Study Funding

This analysis was sponsored by Allergan (prior to its acquisition by AbbVie); they contributed to the concept, acquisition of data, interpretation of data, statistical analysis, drafting, revision, and approval of the final version.

Disclosure

Neither honoraria nor any other form of compensation was provided for authorship. Financial arrangements, within the past year, of the authors with companies whose products may be related to the present manuscript are listed, as declared by the authors. M.F. Brin: employee of AbbVie and receives salary and stock/stock options as compensation. R.S. Kirby: received grant funding from the Centers for Disease Control and Prevention, Florida Department of Health, and Maternal and Child Health Bureau/Health Resources and Services Administration; employee of University of South Florida. A. Slavotinek: none reported as relevant to this publication. A.M. Adams: employee of AbbVie and receives salary and stock/stock options as compensation. L. Parker: employee of AbbVie and receives salary as compensation. A. Ukah: employee of AbbVie and receives salary and stock/stock options as compensation. L. Radulian: employee of AbbVie and receives salary and stock/stock options as compensation and is a PhD student at the University of Medicine and Pharmacy. M.R.P. Elmore: employee of AbbVie and receives salary and stock/stock options as compensation. L. Yedigarova: former employee of AbbVie. I. Yushmanova: employee of AbbVie and receives salary and stock/stock options as compensation. M.F. Brin and L. Parker had full access to all the data in the analysis and take responsibility for the integrity of the data and the accuracy of the data analysis. Go to Neurology.org/N for full disclosures.