2026 Volume 13 Pages 21-26
To investigate whether oral contraceptives induce acute ischemic stroke in patients with moyamoya disease. We retrospectively reviewed female patients with moyamoya disease who were taking oral contraceptives during their reproductive years (12-49 years of age) at the time of diagnosis or during follow-up. The patients' detailed clinical history, including the time course of oral contraceptive usage and acute ischemic stroke occurrence, was evaluated. For patients with radiological evidence of acute ischemic stroke, the imaging pattern of stroke was assessed to investigate the underlying mechanisms. Among 589 female patients, 12 patients took oral contraceptives at some point during the period from 2012 to 2023; 5 patients were taking oral contraceptives at the time of diagnosis, and 4 of them were diagnosed with moyamoya disease after acute ischemic stroke. Among the 7 patients who started oral contraceptives during follow-up, 2 patients developed acute ischemic stroke several months after starting oral contraceptives. Notably, none of the hemispheres that developed acute ischemic stroke previously underwent revascularization, and none of the 5 hemispheres with prior revascularization developed acute ischemic stroke. The imaging pattern of acute ischemic stroke in 7 patients suggested the contribution of a thromboembolic mechanism. Oral contraceptives may increase the risk of acute ischemic stroke via thromboembolic mechanisms in patients with moyamoya disease, but its risk might be manageable in hemispheres with prior revascularization and stable disease conditions for years. Active education of patients regarding the possible risk of oral contraceptives and close communication with gynecologists seem necessary to provide the best treatment strategies for female patients with moyamoya disease who require contraception or treatment of menstrual disorders.
Oral contraceptives (OCs) are well-established as a means of effective contraception and can be used to treat menstrual disorders such as dysmenorrhea, hypermenorrhea, and premenstrual syndrome. In Japan, OCs were introduced in 1999 as medications not covered by insurance, and their use for menstrual disorders has been covered by insurance since 2008. While conflicting results exist1) and the risk of developing thromboembolism seems much lower than the odds during pregnancy,2) studies have shown that OCs increase the risk of acute ischemic stroke (AIS) and venous thromboembolism.3-6) The risk of developing thromboembolism increases during the first 3-4 months after starting OC, in patients over 40 years of age4) and in patients with comorbidities, such as hypertension, diabetes mellitus,5) and migraine.1) In addition to the well-known risk of thrombosis, postmortem histopathological analysis of patients who died from ischemic stroke while taking OC confirmed that OC use induces intimal hyperplasia with and without associated thrombosis in intracranial arteries.7) In Japanese guidelines for gynecological practice,8) "cerebrovascular disease" is listed as a contraindication of OCs together with venous thromboembolism and coronary artery disease as conditions caused by thrombosis. However, whether the risk of AIS is increased by OC use in each specific cerebrovascular disease, such as moyamoya disease (MMD), remains unclear.
MMD is a rare steno-occlusive cerebrovascular disease of unknown origin.9) It is common in female patients of reproductive age because the onset age of adult patients peaks at 30-40 years, and many female children treated during recent decades have reached reproductive age.10,11) While females with MMD are potential candidates for OCs,12,13) and some reports have suggested OCs as possible triggers of AIS in patients with MMD,14-16) these reports included only patients taking OCs before the diagnosis of MMD, and the risk of OCs in patients who have already been diagnosed with MMD or who have undergone revascularization surgery remains unclear. Moreover, details of OCs, such as estrogen dosage and usage period, have rarely been reported; thus, it is difficult to assume that OCs induce AIS in reported patients with MMD. Imaging findings of AIS that could provide underlying mechanisms are also rare. Because the number of Japanese women taking OCs is increasing, opportunities for patients with MMD to request prescriptions for OCs are expected to increase in the future.
Here, we present our case series of female patients with MMD and OC usage with a detailed history of usage to assess the risk of OCs in this disease population. We also evaluated the radiological findings of magnetic resonance imaging (MRI) to estimate the underlying mechanism of AIS during OC usage.
We used a prospectively collected database of patients with MMD in our department with the approval of the ethical committee of the local institution (M2000-1331), and informed consent was obtained from all participants. We extracted data from female patients who visited us during their reproductive years (12-49 years of age) from 1968 to 2024. Medical charts were checked for OC use and detailed information (indications, types of OCs, and duration) and comorbidities (migraine, hypertension, dyslipidemia, type 2 diabetes mellitus, obesity [body mass index >2517)], and smoking history) while patients were taking OCs.
For patients who presented radiological evidence of AIS while OCs were taken, MRI findings were also assessed to investigate possible AIS mechanisms. The location of stroke was categorized as follows:18,19) internal (deep) watershed white matter (junctions between penetrating branches and the major cerebral vessels), external watershed white matter (a strip of paramedian subcortical areas near the vertex), cortex, and the basal ganglia.
In our institution,20) we surgically treat patients by indirect revascularization if hemodynamic disturbances exist on perfusion studies (elongated transit time on dynamic susceptibility contrast and/or decreased cerebral blood flow on arterial spin labeling MRI in pediatric patients; and increased oxygen extraction fraction on 15O-gas positron emission tomography in adult patients). When patients are asymptomatic, or even when patients present with ischemic symptoms, if the hemodynamic disturbance of viable brain is mild, patients are conservatively followed up without surgery; these patients are usually prescribed antiplatelet agents (mainly aspirin, but clopidogrel or cilostazol are also used if they were prescribed by the referring hospital during the study period). Pediatric patients are routinely prescribed aspirin for 3-6 months after the surgery, and then it is usually discontinued.
Among the 717 female patients who were part of the 1,053 patients visiting us during the study period, 589 patients visited us during their reproductive years, and 19/589 (3.2%) had documentation of OCs in their medical charts at some point during the period from 2012 to 2023. One patient took OCs for a while but quit 1 year before the diagnosis, so this patient was excluded from evaluation. Five patients used OCs when they were diagnosed with MMD, and 7 patients started OCs during follow-up (Table 1). Three patients started OCs from 2012 to 2014, while 9 patients started OCs after 2019. Six patients consulted neurosurgeons whether they could use OCs or not, sometimes at the request of their gynecologists, but no one started OCs afterward.
Details of Patients Who Took Oral Contraceptives (OCs) Before or After the Diagnosis of Moyamoya Disease (MMD)
| Case | OCs before/after diagnosis of MMD |
Onset age | Onset type | Previous Surgery | RNF213 p.R4810K | Comorbidities | Age at OC usage | Antiplatelets at OC usage | Ischemic events during OC usage | Reasons for OC use | Type of OC | Estrogen dosage (mg/day) | Usage period |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AIS: acute ischemic stroke; ASA: aspirin; Bil: bilateral; CLP: clopidogrel; DLP: dyslipidemia; DRSP: drospirenone; DSG: desogestrel; EDAS: encephalo-duro-arterio-synangiosis; EE: ethinylestradiol; hetero: heterozygous variant; HTN: hypertension; IDA: iron deficiency anemia; LNG: levonorgestrel; mos: months; N/A: not available; NET: norethisterone; NF1: neurofibromatosis type 1; NG: norgestrel; PMS: premenstrual syndrome; PPD: pack per day; Rt: right; T2DM: type 2 diabetes mellitus; TIA: transient ischemic attack; UF: uterine fibroids; wild: wild type; wks: weeks; yrs: years | |||||||||||||
| #1 | Before | 38 | AIS | hetero | T2DM, DLP, obesity | 38 | AIS | PMS, birth control | LNG+EE | 0.02 | 2 mos | ||
| #2 | Before | 44 | AIS | wild | 44 | AIS | Metrorrhagia, menoxenia | NG+EE | 0.05 | 2 wks | |||
| #3 | Before | 34 | IS, epilepsy | wild | Depression, current smoker (1PPD) | N/A | AIS | PMS | DRSP+EE | 0.02 | N/A | ||
| #4 | Before | 25 | AIS | hetero | IDA | 25 | AIS | Hypermenorrhea | LNG+EE→ DSG+EE | 0.03 | 7 mos(6 mos→ 1 mos) | ||
| #5 | Before | 48 | Headache | N/A | 47 | ? IS | PMS | DRSP+EE | 0.02 | 17 mos | |||
| #6 | After | 20 | TIA | Rt EDAS | hetero | 24 | AIS, TIA, MRA progression | Dysmenorrhea | LNG+EE | 0.02 | 2 mos (intermittent usage for 2 yrs) | ||
| #7 | After | 20 | Headache | hetero | HTN | 21 | AIS, TIA, MRA progression | Menoxenia | LNG+EE | 0.02 | 6 mos | ||
| #8 | After | 41 | TIA | hetero | 48 | Asymptomatic IS, MRA progression | Hypermenorrhea | NET+EE | 0.02 | 12 mos | |||
| #9 | After | 21 | Headache | N/A | Ex-smoker (till 20 yrs old) | 27 | ASA | TIA, MRA progression | Dysmenorrhea | DRSP+EE | 0.02 | 18 mos | |
| #10 | After | 2 | TIA | Bil EDAS | hetero | IDA | 23 | TIA | PMS | DRSP+EE | 0.02 | 60 mos (continuing) | |
| #11 | After | 3 | Epilepsy | Bil EDAS | N/A | NF1 | 19 | TIA | Menoxenia | N/A | N/A | 12 mos | |
| #12 | After | 42 | Headache | hetero | HTN, migraine, obesity | 43 | CLP | None | Multiple UF, hypermenorrhea | N/A | N/A | 1 mo | |
Among patients who used OCs at the time of diagnosis, 4/5 (80%, #1-4 of Table 1) patients were diagnosed with AIS, and 2/4 (50%) developed symptoms within 2 months after starting OCs. One patient (#4) also had iron deficiency anemia due to hypermenorrhea at the time of AIS. Although 1 patient (#5) was diagnosed with headache and did not have radiological evidence of AIS, she developed left arm paresthesia 3 months after starting OCs, and MRI obtained 1 year after this symptom revealed old ischemic stroke in the right parietal lobe that might be the cause of her previous symptoms. Two patients (#1 and #2) revealed misery perfusion on 15O-gas positron emission tomography and received indirect revascularization. The remaining patients lacked misery perfusion and were conservatively treated. All patients discontinued OCs after the diagnosis of MMD.
Patients who started OCs during follow-upSeven patients started OCs after the diagnosis of MMD without consulting their neurosurgeons; nevertheless, all patients reported their MMD history to their gynecologists. Three of the seven (43%, #6-8) patients developed AIS several months after starting OCs; notably, all strokes occurred in the nonsurgically-treated hemispheres. A patient who previously received right-sided revascularization (#6) discontinued and restarted OCs intermittently for 2 years because she frequently experienced transient ischemic attacks (TIAs) of the right extremities while taking OCs, but at the same time, her dysmenorrhea clearly improved. She developed AIS in the left hemisphere but not in the surgically-treated hemisphere.
Additionally, TIA occurred in 3/7 (#9-11) patients (1 conservatively treated patient and 2 postoperative patients) several months after starting OC. One patient (#9) developed involuntary movements of the right hand while taking OCs, but symptoms disappeared afterward. Two postoperative patients (#10, 11) who were stable for a decade developed transient leg weakness several months after starting OC. Their transient symptoms eventually disappeared during the continuation of OCs.
The patient with prior right-sided revascularization (#6) subsequently received left-sided revascularization because 15O-gas positron emission tomography revealed misery perfusion in the left hemisphere. The remaining patients were conservatively treated because misery perfusion was not detected. All patients discontinued OCs within 18 months, except for 1 patient with prior bilateral revascularization during childhood (#10) who strongly wanted to continue OCs because they relieved her severe premenstrual symptoms.
Radiological findings of AISMRI of the AIS was available for 7 patients. Six patients had multiple strokes involving the cortex or subcortical white matter in middle cerebral artery territories (Fig. 1). One patient experienced one large parietal stroke involving both the deep white matter and cortex. Overall, the AIS of all patients involved external watershed white matter (Fig. 2). Two patients who developed AIS also exhibited progression of arterial stenosis after taking OCs during follow-up (#6 and #7 in Table 1). No hemispheres with prior revascularization (ipsilateral hemisphere of 1 patient after unilateral revascularization and bilateral hemispheres of 2 patients after prior bilateral revascularization) revealed radiological evidence of AIS.
Representative magnetic resonance images of patients who developed AIS while taking oral contraceptives before (#1, 2, 4) or after (#6-8) the diagnosis of moyamoya disease.
All patients presented radiological evidence of multiple strokes (the white arrows indicate ischemic stroke lesions in internal and/or external watershed white matter, and white arrowheads indicate ischemic lesions in the cortex and the basal ganglia). Two of the three patients who developed acute AIS also experienced progression of arterial stenosis (#6, left anterior cerebral artery; #7, left middle cerebral artery). A patient with prior right-sided revascularization did not develop AIS in the treated hemisphere (#6). Note that the number of each patient is identical to the number in Table 1.
AIS: acute ischemic stroke
Locations of acute ischemic stroke lesions in seven patients.
While some previous studies reported patients diagnosed with MMD and AIS while taking OCs,14-16) our series is unique because we also evaluated patients who used OCs after diagnosis and during serial MRI follow-up at the outpatient clinic. A substantial number of patients developed AIS, and radiological findings of AIS exhibited involvement of external watershed white matter in all patients.
We have found that MRI revealed multiple AISs involving external watershed white matter, internal watershed white matter, and cortex in our patients. According to "washout theory", thromboembolism may induce AIS in watershed areas, either directly or combined with hypoperfusion and reduced clearance of emboli.21) A previous study found that microembolic signals on transcranial Doppler were significantly associated with ischemic events during follow-up in MMD.22) Therefore, in addition to AIS involving cortical lesions, lesions of watershed white matter may also be triggered by thromboembolisms. Even in patients with mild hemodynamic disturbances, the elongated transit time and slow blood flow23) in affected arteries of MMD may trigger AIS. Because the risk of developing thromboembolism increases especially during the first 3 months after starting OCs, at least in the 3 patients who developed AIS within 3-4 months after starting OCs (#1, #2 and #6 in Table 1), we can suspect the OCs as a probable trigger of thrombosis and a possible trigger of arterial stenosis or transient ischemic symptoms.
In addition to the well-known possible risk of developing thromboembolism, previous histopathological studies have suggested that OC usage induces intimal thickening composed of collagen and smooth muscle cells in intracranial arteries.7) In 2 patients who started OCs during follow-up, we observed the progression of arterial stenosis as well as AIS. As MMD is progressive in nature, we cannot prove that OCs induce arterial stenosis in these patients without histopathological evaluation of stenosed intracranial arteries.
All patients who received OCs after MMD diagnosis started OCs without consulting their neurosurgeons, but gynecologists prescribed OCs while being aware of the history of MMD. While specific risk of AIS in MMD under OCs usage remains unclear, the risk of ischemic stroke might be higher in some specific patients with MMD, such as those with recent ischemic symptoms. Perhaps gynecologists might not be aware of the risk of ischemic stroke in MMD because hemorrhagic stroke is more common in adult patients with MMD in Japan. While OCs may increase the risk of AIS, anemia induced by hypermenorrhea may also increase the risk of AIS in patients with intracranial arterial stenosis,24,25) so active treatment of hypermenorrhea is important to reduce stroke risk in patients with MMD. Active education of patients regarding the possible risk of OCs and close communication with gynecologists seem necessary to provide the best treatment strategies for patients. Medications with reduced risks of thromboembolism, such as progesterone-only pills26) and novel estetrol-containing OCs,27) would be reasonable options for patients with MMD. Also, because the risk of developing thromboembolism increases primarily during the first 3-4 months after starting OCs and gradually decreases afterward, if patients are to use OCs, it is better to adhere to continuous standard usage rather than starting and quitting intermittently, as some of our patients did.
Notably, AIS occurred only in hemispheres of adult-onset patients without prior revascularization, and no hemisphere with prior revascularizations and stable disease conditions for years developed AIS. Numerous studies have shown that revascularization surgery decreases the risk of AIS in patients with MMD.28) The reduced transit time and fast blood flow guaranteed by prior revascularization may have a protective effect against AIS by improved clearance of emboli.29,30) It is also possible that patients with prior revascularization have a lower risk of AIS just because they no longer have an active disease status and are in stable conditions after successful treatments. Compared to patients with prior revascularization and stable disease conditions for years, recently diagnosed adult-onset patients may have a higher risk of disease progression and AIS, even though their hemodynamic disturbance is relatively mild. Although we could not prove that revascularization decreases the risk of AIS while OCs are taken, patients with prior revascularization with stable hemodynamic status for years may have a relatively small risk of AIS from OCs.
The limitations of this case series include its retrospective nature with probable selection bias; we did not calculate and compare the stroke occurrence rates of patients with and without OC usage because of the small sample size, various disease statuses, ages, comorbidities, and potential bias. Some patients might have used OCs but did not disclose the information to us. Nevertheless, this is the largest case series of patients with MMD who use OCs, with a detailed clinical history of each patient. Epidemiological studies with large sample sizes are warranted to clarify the OC-related risk in patients with MMD.
In conclusion, OCs may increase the risk of AIS via thromboembolic mechanisms in patients with MMD, and traditional OCs increasing the risk of thromboembolism may be better avoided. Nevertheless, the risk might be manageable in hemispheres with prior revascularization and stable disease conditions for years. Active education of patients regarding the possible risk of OCs and close communication with gynecologists seem necessary to provide the best treatment strategies for female patients with MMD who require contraception or treatment of menstrual disorders.
All authors have no conflict of interest.
This work was partially supported by Japan Society for the Promotion of Science KAKENHI (grant number. 23K08514 given to Shoko Hara). Otherwise, the authors have no relevant financial or nonfinancial interests to disclose.
