Cervical Artery Dissection in Young Women

Abstract

Objectives

To assess the risk of recurrence of cervical artery dissection (CAD) during pregnancy and puerperium in women with a history of CAD and then help physicians with providing medical information to women who wish to become pregnant.

Methods

Women aged 16–45 years who were admitted to our center for a CAD between 2005 and 2017 were identified from the hospital database, and those with spontaneous and symptomatic CAD were included. They were then contacted to answer a questionnaire that was specifically designed in regard to the recurrence of CAD and pregnancies after the primary CAD.

Results

Ninety-one patients satisfied our inclusion criteria, and 89 were included in the analysis. During a median follow-up of 7.0 years, 4 women (4.4%) had recurrent CAD, although none during pregnancy or puerperium. Eighteen women (20%) had a total of 20 full-term pregnancies, occurring at least 6 months after CAD. Of these 20 pregnancies, 13 (65%) were vaginal deliveries, and 7 (35%) were cesarean sections. The reason for the absence of pregnancies after the initial CAD was unrelated to the vascular event in 89% of cases, but 8% of the women had been advised by a physician to avoid any future pregnancy or they had been recommended to undergo abortion or sterilization.

Conclusion

In this study, there were no CAD recurrences during subsequent pregnancies or postpartum, irrespective of the type of delivery. Thus, pregnancy after a history of CAD appears to be safe.

Cervical artery dissection (CAD) is a leading cause of stroke in young adults.^1,2 Recurrences of CAD have a bimodal distribution, occurring mostly during the first month after the initial event, whereas the risk of a late recurrence is low (1% per year).^3,4

Many cases of CAD have been described in women in association with pregnancy or puerperium, thus suggesting that these conditions may be risk factors.^5-10 A recent analysis of a large population-based sample of women indicates that pregnancy is associated with a higher risk of CAD.¹¹ Puerperium, defined as the period of approximately 6 weeks after childbirth, appears to be the period with the highest risk, accounting for 6% of CAD.^8,11 Mechanical factors related to expulsive exertions, transient arterial wall abnormalities induced by hormonal changes, physiologic hemodynamic variations (such as an increase in cardiac flow or blood volume), or pathologic conditions with arterial hypertension (preeclampsia or hemolysis, elevated liver enzymes and low platelets [HELLP] syndrome) may play a role.^6,7,10,12

There have been few studies regarding the risk of recurrence during pregnancy and childbirth, and the lack of observational data makes it difficult to provide solid advice in this regard to women with a history of CAD who wish to become pregnant. Whether the delivery method has an effect also remains unclear. There has been only 1 study that specifically assessed this risk, and it indicated that this risk may, in fact, not be increased during pregnancy, childbirth, or puerperium.¹³

We investigated the risk of recurrence of CAD during pregnancy, childbirth, and puerperium in a large cohort of women with a history of CAD. We also recorded the obstetrical monitoring and the mode of delivery in women with further pregnancies or, if not, the reasons why.

Methods

Patients

Consecutive female patients who were aged from 16 to 45 years and who had been hospitalized between 2005 and 2017 in our stroke center for acute spontaneous CAD were screened. The diagnosis of CAD was confirmed by MRI and MR angiography or an angio-CT scan. Iatrogenic and traumatic CAD cases were excluded.

The patients were identified using a computerized extraction of the International Classification of Diseases, Tenth Revision (ICD-10) cervical dissection code (I67.0, I72.0, I72.5, and I72.8) from the French National Hospital Discharge Database (HDD), called PMSI (Programme de Médicalisation des Systèmes d'Information). In addition, using a database of our University Hospital, we extracted all patients with the keywords “cervical,” “carotid” or “vertebral,” and “dissection” in their medical reports. This method was used to improve completeness of the data, and it allowed 13 additional patients to be recruited for whom the ICD-10 coding was incorrect.

Data Collection

The medical records were reviewed to obtain the characteristics of the initial event. The women who satisfied the inclusion criteria were then contacted by phone or mail to document any recurrences or subsequent pregnancies after the primary CAD. A structured questionnaire was devised to obtain information about the recurrence of CAD (signs or symptoms of dissection confirmed by imaging and whether it led to hospitalization), the number of pregnancies before and after CAD, as well as the place of obstetrical monitoring, the method of delivery, and the management of antithrombotics in case of further pregnancies. For cases with no further pregnancy, the reasons were explored: medical advice, fear of a recurrence, or any other reason not related to CAD. The interview was conducted over the phone by an experienced neurologist. If a woman was not able to answer a question and provide useful information regarding subsequent pregnancies, obstetrical records were consulted. For patients who reported any further significant event, the medical records were checked, in particular to confirm the diagnosis of CAD recurrence.

Standard Protocol Approvals, Registrations, and Patient Consents

An information form was sent by mail, and all participants provided their informed consent. The study was approved by the regional Ethics Committee and registered in ClinicalTrials.gov (NCT04253535).

Statistical Analysis

A χ² test (or Fisher test) and the Student t test were used to explore group differences, and a p value of <0.05 was considered statistically significant. The recurrence of CAD was assessed by a Kaplan-Meier analysis and a log-rank test to compare the survival curves. The event rate was calculated in person-years using a Poisson distribution 2-sided test with the exact method and 95% CI. The statistical analysis was carried out with R 3.3.1 statistical software.

Data Availability

Anonymized data not published within the article could be shared by request from any qualified investigator.

Results

A total of 91 patients met our inclusion criteria. One patient died during the acute phase of stroke because of CAD, and another one was lost to follow-up. All 89 patients who were contacted agreed to participate in the study and answered the questionnaire, and there was no need to check the obstetrical records.

The characteristics of the population are shown in the Table. Of the 89 initial CAD, 8 (9%) occurred during pregnancy or puerperium (1 during the first quarter and 7 in puerperium, mostly during the first week after delivery). None of the women had signs or symptoms of a heritable connective tissue disorder (e.g., Marfan or Ehlers-Danlos syndrome), although 6 patients (7%) had signs of potential underlying arteriopathy such as cervical or renal fibromuscular dysplasia or intracranial aneurysms.

Table.

Baseline Characteristics of the Included Patients

During a median follow-up of 7.0 ± 3.7 years (from 2 to 14 years), 4 women (4.4%) had recurrent CAD, 1 of whom had 2 recurrences. The event risk in person-years was 5.7 × 10⁻³ (95% CI, 0.002–0.015). All these recurrences affected women who did not become pregnant after the primary CAD (Figure). Two of 5 recurrences occurred within 3 months of the primary CAD, and the 3 others (60%) occurred after 12 months (at 1, 5, and 10 years). The recurrences involved an artery other than the first CAD in 4/5 cases. None of the recurrences led to a stroke. Only 1 patient had an underlying arteriopathy (renal fibromuscular dysplasia). The cumulative risk of recurrence was 2.2% after 2 months, 3.3% over the first year, and 6.4% over 10 years.

Figure. Kaplan-Meier Curve of CAD Recurrence.

The Kaplan-Meier curves show no recurrence of CAD in the 18 women with pregnancies after the initial CAD, whereas there were 4 women with CAD recurrence among the 71 without pregnancy. CAD = cervical artery dissection.

During the follow-up, 18 women (20%) had a total of 20 full-term pregnancies, starting at least 6 months after the first CAD. These women were younger at the time that CAD occurred than the women without further pregnancies (31 ± 3 year old vs 39 ± 4.8), and they had half as many previous pregnancies (Table). There were no recurrences of CAD during the pregnancies and puerperium. Half of the pregnancies were monitored in a University Hospital, each time at the request of a neurologist or obstetrician, in light of the history of CAD, except for 1 woman who had preeclampsia during a previous pregnancy. Of the 20 pregnancies, 13 (65%) were vaginal deliveries and 7 (35%) were cesarean deliveries. Of the 7 deliveries by cesarean section, 5 were scheduled because of a history of CAD, mainly to avoid expulsive efforts, and 2 were emergency deliveries for other reasons (preeclampsia and intrauterine growth restriction). Fourteen women (78%) were still being treated with an antithrombotic drug at the time of pregnancy (aspirin in 13/14 cases), with no change during pregnancy. Most of the women taking aspirin (85%) stopped their treatment several days before childbirth (between 1 and 4 weeks). For 2 patients, treatment with aspirin was substituted with anticoagulation medication (therapeutic doses of low-molecular-weight heparin) after childbirth, for 3 months, based on the advice of a neurologist. Of the 22% of women who had not been treated with an antithrombotic for several months or years, no treatment was readministered because of pregnancy.

Most of the 71 women (63, 89%) with no pregnancies after the dissection reported that the reason for not becoming pregnant was not directly related to their history of CAD. They most often reported no subsequent pregnancies because of their age (the mean age at the time of the primary CAD was 39 ± 4.8 years). Six women (8%) stated that their physician (a neurologist or obstetrician-gynecologist) advised them against any future pregnancies because of their history of CAD, and 1 of them had a voluntary termination of their pregnancy based on the advice from their obstetrician-gynecologist. Another woman stated that her gynecologist encouraged her to undergo tubal ligation. Finally, 2 patients (3%) abstained from further pregnancy because of fear of recurrence of a CAD.

Of the 8 women who had their initial CAD during pregnancy or puerperium, 1 had a subsequent pregnancy, with a cesarean delivery because of preeclampsia, and the 7 other women did not have any subsequent pregnancies because of personal reasons (age and no longer desire for pregnancy).

Discussion

In our observational study, we noted an overall risk of recurrence of CAD of 4% in >7 years of follow-up, with a bimodal distribution as previously described,^3,4 which was unrelated to pregnancy or puerperium. Indeed, there were 20 full-term pregnancies, representing 20% of our cohort of women, without any recurrences of a cerebrovascular event or CAD. This suggests that the risk of recurrence in women with a history of CAD may not be increased during pregnancy, childbirth, or puerperium.

CAD is a leading cause of ischemic stroke in young adults.¹ In practice, neurologists and gynecologists are regularly asked by women who have a history of CAD about the risks related to a further pregnancy. Several authors have studied the risk of stroke recurrence during pregnancy and puerperium, irrespective of its cause.^14-16 All these studies suggest that this risk is very low. One study focused on the risk of recurrence after CAD.¹³ This study identified 92 women with a history of CAD, but only 60% were included (11 declined to participate, and 28 were lost to follow-up). During a mean follow-up of 6 years, 13 pregnancies occurred without a CAD recurrence. In a French cohort of 373 young women with an arterial ischemic stroke, the stroke was related to a CAD in 79 of them.¹⁵ During a mean follow-up of 5 years, 11 women had a full-term pregnancy, with no recurrence. Finally, when we considered all childbearing women with a history of CAD reported in the largest series to date (the 2 referred to above and our own), a total of 221 women had a long follow-up after CAD (>5 years), with a particular focus on their reproductive history, and none had a recurrence during pregnancy.^13,15 However, pregnancies were reported for only 44 of the women (including unpublished data from Lamy et al.).¹⁵

Half of the pregnancies in our study were monitored in high-level university centers as a result of medical advice based on a history of CAD. However, such monitoring is not usually recommended in women with a history of stroke.^17,18 This should remain true for CAD, except in the context of heritable connective tissue disorders or uncontrolled hypertension, because CAD is not a chronic disease, and it has a very low recurrence rate that is probably not influenced by further pregnancy, and it carries no theoretical fetal risk.^13,15

Cesarean delivery was performed in one-third of our subjects, whereas in France, during the same period, only 20% of childbirths were by cesarean sections.¹⁹ Cesarean was recommended by the practitioners to avoid expulsive efforts, whereas the other subjects had vaginal deliveries without any complications. The current guidelines do not support systematic use of cesarean delivery when there is no obstetrical indication,^17,18 and our results, like those from previous studies,^13,15 are in keeping with this assertion.

The antithrombotic regimen should be reassessed before all pregnancies in women with a history of stroke, including CAD.¹⁷ According to current recommendations, there is no need to withdraw low-dose aspirin for delivery or breastfeeding.²⁰ Furthermore, as the risk of recurrence of CAD does not appear to increase during pregnancy, restarting a prophylactic antithrombotic in women who have stopped taking the treatment for several months or years does not appear to be warranted in our opinion. Moreover, there is no evidence or guideline for starting anticoagulation therapy or replacing antiplatelet with anticoagulant medication at the time of pregnancy or puerperium because there is no clear difference between these 2 types of drugs in terms of preventing ischemic recurrence²¹ and the use of anticoagulants is usually limited to a few months after CAD, followed by antiplatelet therapy.²²

All pregnancies started more than 6 months after the first CAD, as reported in previous studies,^13,15 whereas recurrences have been reported to be more frequent within the first months after the initial dissection.⁴ We cannot exclude that the recurrence rate could be significant when pregnancies start during the first 6 months after the initial dissection. No recommendations specify the minimum delay before starting a pregnancy after a history of CAD.

The physicians involved (i.e., the neurologist and obstetrician) should be well aware of the very low risk of recurrence and the favorable pregnancy and childbirth outcomes after a CAD, before making recommendations regarding avoidance of pregnancies, inducing abortion or encouraging sterilization, as was reported by 8% of our patients. Our results, and those of previous studies,^13,15 suggest that a history of CAD does not appear to warrant limiting further pregnancy or modification of its course, unless there is another medical reason to do so.

In this study, we found that 9% of the instances of CAD occurred during pregnancy and most often during puerperium. The proportion of CAD occurring during these periods has rarely been mentioned to date in large series of CAD,^3,23-25 although pregnancy is considered to be a risk factor.¹¹ It was approximately 6% in 1 monocenter prospective registry⁸ and 4% in a case-control study.¹¹

The data collection was retrospective for recurrence or further pregnancy. This study design is justified by the rarity of CAD among all strokes, as well as the low incidence of recurrence, and it allows for a long follow-up time. A recall bias, particularly for cerebrovascular events, is possible, but it seems unlikely as the medical professionals had informed the patients of the symptoms of recurrence of CAD or stroke. However, we cannot exclude that asymptomatic CAD could have occurred or that some transient or minor events were not mentioned by the patients. The monocentric design of this study implies that the medical practices were relatively uniform, especially the recommendations in regard to further pregnancies and the management of the antithrombotic therapies. The small number of patients and, thus, the small number of pregnancies, are another limitation of this study. The small number of pregnancies after CAD may be partly explained by the mean age of the patients at the time of the CAD event. Indeed, the women who did not have a pregnancy after CAD were significantly older, and their mean number of previous pregnancies was higher.

Because of the low number of subsequent pregnancies in women with a history of pregnancy-associated CAD, we cannot provide an estimate of the risk of recurrence in this subset of women if they become pregnant again. Furthermore, none of the 4 women with recurrence became pregnant. We, therefore, cannot exclude that these women could have peripartum recurrences in case of pregnancy.

In conclusion, this study showed that no recurrences of CAD during subsequent pregnancies occurred for at least 6 months after the initial CAD event, and it suggests that pregnancy and childbirth follow a normal course in women with a history of CAD. Thus, a history of CAD does not appear to warrant medical contraindication to pregnancy, and there is no evidence in favor of cesarean delivery.

TAKE-HOME POINTS

• → Pregnancy is a risk factor for CAD, and it has been identified in 4–9% of female patients.

• → There are limited data regarding the risk of recurrence during subsequent pregnancies, irrespective of the putative precipitating factor of the initial CAD. Thus, it may be difficult for a physician to provide medical information to women who wish to become pregnant after a CAD.

• → In this cohort of 91 women with a history of CAD and a median follow-up of more than 7 years, 20% eventually became pregnant. No recurrence of CAD occurred during pregnancy or puerperium, irrespective of the type of delivery.

• → This study suggests that pregnancy and childbirth after a history of CAD follow a normal course and are safe.

Appendix. Authors

Study Funding

No targeted funding reported.

Disclosure

The authors report no disclosures relevant to the manuscript. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.