New-onset headache after transcatheter atrial septal defect closure: a systematic review

Abstract

Background

Migraine is a complex neurological disorder marked by severe headache and associated with various systemic symptoms. Atrial septal defect (ASD) closure, a common cardiac procedure, has been linked to the onset of new migraine episodes. This systematic review explores the incidence and management of migraine following transcatheter ASD closure.

Methods

A comprehensive literature search was conducted in PubMed, Scopus, and CINAHL Ultimate, covering studies from inception to August 2024. The review followed PRISMA guidelines and was registered on PROSPERO (CRD42024578609). Data extraction and risk of bias assessment were systematically performed by independent reviewers.

Results

A total of 831 studies were identified out of which 13 were included for the final analysis. The studies encompassed diverse populations globally, revealing varying onset times for migraine post-ASD closure, ranging from as early as 1 day to up to a year. Migraine incidence varied significantly, with several studies noting a higher prevalence in females and some cases showing a familial predisposition typically in the maternal side. The typical methods for ASD closure involved transcatheter approaches. Treatment modalities included non-steroidal anti-inflammatory drugs, calcium channel blockers, and anticoagulants, with some cases experiencing spontaneous resolution of symptoms. Modifications to anti-platelet therapy post-procedure, such as the introduction of clopidogrel along with aspirin, showed significant efficacy in treating and preventing migraine.

Conclusion

This review highlights a significant correlation between ASD closure and the onset of migraine, underlining the need for further investigation into preventive strategies and management. These findings suggest a complex interaction between structural heart alterations and migraine pathophysiology.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12055-025-01906-x.

Introduction

Migraine is characterized as a neurological disorder that involves episodes of intense headache accompanied by reversible neurological and systemic symptoms. Common symptoms associated with migraine include nausea, vomiting, sensitivity to light and sound, and visual disturbances, such as flashes of light [1]. A number of studies have detailed the pathophysiology linked to autonomic nervous system dysfunction and alterations in cerebral blood flow among migraine sufferers [2–6]. During the prodromal phase, individuals experiencing migraine exhibit cerebral hyperperfusion in hypothalamic regions that regulate the autonomic nervous system [3]. Additionally, patients with severe migraine attacks tend to show elevated diastolic blood pressure and diminished heart rate variability, indicating dysfunction within the autonomic nervous system [4]. These observations imply that alterations in cerebral hemodynamics, in conjunction with changes in the autonomic nervous system, play a significant role in the onset of migraine.

An atrial septal defect (ASD) occurs when there is an abnormality in the formation of the septum, leading to a connection between the right and left atria. If patients with ASD do not receive timely intervention, they could face complications later in life, including atrial arrhythmias, pulmonary hypertension, stroke, and right-sided heart failure. Quite a few individuals may experience new migraine headache following the procedure, typically emerging within days to weeks post-operation. This complication can result in a considerable increase in medical consultations, along with missed work or school days and decreased productivity [7].

Recently, multiple studies have suggested a possible link between the closure of ASD and the pathophysiology of migraine headache. Several theories have been put forward, such as micro-thromboembolism caused by irregular platelet activity, allergic responses, and the release of biochemical agents like serotonin. Nevertheless, the exact mechanism that explains this relationship is still not well understood. It has also been suggested that the headache occurring after ASD closure could be related to hemodynamic maladaptation due to a rapid increase in left ventricular preload [8]. The present systematic review aims to evaluate the present literature on the onset and management of migraine after closure of an ASD.

Methods

This review adheres to the Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA) guidelines [9] and has been registered with PROSPERO (ID: CRD42024578609).

Literature search

A comprehensive literature review was performed utilizing the following electronic databases: PubMed, Scopus, and CINAHL Ultimate. The search was performed from inception until 12 August 2024. The search string was as follows: (“ASD” OR “atrial septal defect”) AND (“headache” OR “migraine” OR “headaches” OR “migraines”) AND (“close” OR “closure” OR “closing” OR “closed” OR “closes” OR “postoperative” OR “post operative”).

Two authors (S.R.P. and H.S.A.) independently assessed the eligibility of each study. A total of 831 studies were identified based on the above search string. The articles were primarily screened by examining their titles and abstracts. Then, selected papers were subjected to a comprehensive full-text review. Finally, 13 studies which met the inclusion and exclusion criteria were included in the final analysis. Any disagreements that arose were resolved by discussion and consensus. The flowchart detailing the selection process is delineated in Fig. 1.

Fig. 1.

PRISMA flowchart

We included all studies evaluating new-onset headache or migraine in patients after transcatheter/percutaneous or surgical ASD closure. We included randomized controlled trial (RCT), prospective and retrospective interventional studies, observational studies, and case series. Studies exclusively evaluating patients with pre-existing migraine/headache; studies treating migraine/headache through ASD and/or patent foramen ovale closure; studies where ASD closure details are unavailable were excluded. There were no language restrictions on the included studies, and subsequently, English- and Chinese-language studies were identified. Case reports and case series with skewed or minimal information were excluded. A detailed PICOT (population, intervention, comparison, outcome, time) strategy used in the selection of relevant studies is found in Table 1.

Table 1.

PICOT table

Element	Description
Population	Patients of any age who have undergone transcatheter/percutaneous or surgical closure of an ASD with no prior history of migraines or chronic headaches
Intervention	Atrial septal defect closure performed either by transcatheter/percutaneous techniques or by open surgical methods
Comparison	Comparison may include patients without new-onset headaches or migraines after ASD closure or patients evaluated before ASD closure. Studies without a comparison group are also included if they report on post-closure headaches
Outcomes	Primary outcome: incidence of new-onset headaches or migraines following ASD closure Secondary outcomes: time to onset, frequency, duration, severity of the headaches, treatment modalities used, effectiveness of management, and any long-term neurological or quality-of-life outcomes related to post-closure headaches
Time	Time frame for the development of headaches after ASD closure, with both short-term (within 30 days of closure) and long-term (months to years after closure) follow-up considered. The time frame for headache onset, progression, and resolution will also be evaluated when reported in studies
Study types	The review includes randomized controlled trials, prospective and retrospective interventional studies, observational studies (cohort, case–control, cross-sectional), and case series

Data was extracted independently by two authors (A.F.D. and M.Z.) and entered into a spreadsheet (Microsoft Excel 2021). Extracted data included study details such as demographics, migraine particulars, treatment strategies, and outcomes. Any discrepancies were resolved after discussion or after consulting a third author (H.S.A.).

Risk of bias was also assessed independently by two authors (S.R.P. and A.F.D.) using scales specific to different studies. RCTs were evaluated during the Cochrane RoB 2 tool (Risk of Bias 2.0), non-randomized interventional studies were evaluated using ROBINS-I tool (Risk Of Bias In Non-randomized Studies of Interventions), and case series were evaluated using the JBI tool for case series [10–12].

Results

Demographics

The demographic details of included studies are summarized in Table 2. The studies under review were conducted in various countries, including Japan, the UK, Spain, Canada, the USA, India, and China. The year of publication ranged from 2005 to 2023 [13–25]. Sample sizes varied, with Castellini et al.’s [17] case series including only 3 participants, while a study conducted by Smerdziński et al. [25] in Poland involved as many as 840 participants. Some studies included only adults or pediatric population, while others included both. Six studies [13, 14, 17, 18, 20, 23] discussed only adults, five studies [16, 19, 21, 22, 24] reported the inclusion of both adult and pediatric population, and two studies [15, 25] evaluated only the pediatric age group. Matsuoka et al. [21], although included both adult and pediatric age groups, predominantly focused on the pediatric age group. Two studies [17, 24] reported the presence of a positive family history of headache or migraine typically in the maternal side. The study wise risk of bias can be found in Supplementary File 1.

Table 2.

Demographics

Author	Year	Country	Sample size	Adult/pediatric/both	Age at ASD closure	Age at presentation	Sex (M:F)	Study design
Wilmshurst et al. [13]	2005	UK	Aspirin alone group = 71 Aspirin + clopidogeral group = 89	Adult	Aspirin alone group: 37.0 ± 10.5 years Asipirn + clopidogeral group: 41.7 ± 10.1 years	-	Aspirin alone group: 40:31 Aspirin + clopidogeral group: 61:29	Clinical audit
Mortelmans et al. [14]	2005	Belgium	75	Adult	51 ± 19 years	-	16:59	Cross-sectional study
Fernández-Mayoralas et al. [15]	2007	Spain	97 (4 patients from this sample who experienced migraine were selected for detailed report)	Pediatric	Case 1: 16-year-old male Case 2: 13-year-old male Case 3: 14-year-old male Case 4: 4-year-old female	Case 1: 48 h after ASD closure, at the age of 16 Case 2: 24 h after ASD closure, at the age of 13 Case 3: 3 days after ASD closure, at the age of 14 Case 4: 7 days after ASD closure, at the age of 4	3:1	Case series
Rodés-Cabau et al. [16]	2008	Canada	260	Both	Mean, 39 ± 19 years	-	1:1.47 (75:110)	Cross-sectional study
Castellini et al. [17]	2009	Italy	3	Adult	Case 1: 42 Case 2: 23 Case 3: 21	Case 1: 44 Case 2: 23 Case 3: 25	1:2	Case series
Wei et al. [18]	2012	Taiwan	40	Adult	New-Onset MHAs, 23 ± 3.3 years Without new-onset MHAs, 29 ± 3.3 years	-	New-onset MHAs, 1:7 Without new-onset MHAs, 1:3	Observational study (prospective)
Kato et al. [19]	2013	Japan	207	Both	Range, 4 to 79 years	-	0.42:1 (61:146)	Cross-sectional
Sharma et al. [20]	2020	USA	53	Adult	42 years (range, 24 to 71 years)	-	4:5.5 (40:55)	Ambispective interventional study
Matsuoka et al. [21]	2020	Japan	167	Both (primarily pediatric)	9 to 17	-	0.62:1 (64 males, 103 females)	Observational study (retrospective)
Huang et al. [22]	2020	China	567	Both	34.4 ± 19.4	-	1:2.8 (148 were male)	Observational study (retrospective)
Wintzer-Wehekind et al. [23]	2021	Canada	171	Adult	Mean age, 38	-	0.62:1 (65: 106)	Randomized clinical trial
Gangopadhyay et al. [24]	2022	India	325	Both	3 years to 65 years	Range, 3.5–35 years	0:3 (all 5 patients presented with headaches were female)	Observational study (retrospective)
Smerdziński et al. [25]	2023	Poland	840	Pediatric	Median 5.5 (3 to 11 years)	13 (7.35–16) years	1:2 (269 male patients (32%) and 571 female patients (68%))	Observational study (retrospective)

Age at presentation was reported in four studies, most of them being case series [15, 17, 24, 25]. In the study by Gangopadhyay et al. [24], which described five cases, the age at presentation ranged from 3.5 to 35 years, with a median age of 26 years. The pediatric population included in Smerdziński et al.’s [25] study presented at ages ranging from 7.35 to 16 years. Similarly, the four cases reported by Fernández-Mayoralas et al. [15] involved patients aged between 4 and 16 years, with the presentation occurring between 24 h and 7 days after ASD closure. On the other hand, the age at which ASD closure was performed varied across the studies. Gangopadhyay et al. [24] reported an age range of 3 to 65 years for ASD closure. Similarly, Sharma et al. [20] and Kato et al. [19] noted age ranges of 24 to 71 years and 4 to 79 years, respectively. In contrast, Smerdziński et al. [25] reported a median age of 5.5 years for ASD closure in their study.

Gender distribution varied significantly across the studies. Ten studies [14, 16–23, 25] showed a higher female population compared to male. A study conducted by Smerdziński et al. [25] reported a male-to-female ratio of 1:2, while Matsuoka et al.’s [21] study showed a male-to-female ratio of 0.62:1. On the other hand, some studies indicated a male predominance [13, 15].

ASD measurements

Table 3 discusses the details regarding the ASD measurements and residual small defect. Five studies provided details on ASD dimensions [16, 19, 21, 23, 25]. Rodés-Cabau et al. [16] reported that patients with new-onset migraine headache (MHAs) had a median ASD size of 16 mm, compared to 15 mm in those without MHAs. In the study by Gangopadhyay et al. [24], among the five cases described, the ASD size ranged from 16 × 18 mm to 32 × 30 mm. Smerdziński et al. [25] divided their study population into two groups: those with transient headache episodes (THE), who had an ASD size ranging from 8 to 14.5 mm, and those without THE, whose ASD size ranged from 8 to 13 mm. In another study, Kato et al. [19] noted that the de novo MHA group (n = 23) had a mean ASD size of 16.8 ± 4.7 mm.

Table 3.

ASD details

Author	ASD size	Qp/Qs	Atrial septal aneurysm	Residual small ASD after closure
Wilmshurst et al. [13]	-	-	-	Yes
Mortelmans et al. [14]	Mean, 23 ± 7 mm	1.9 ± 0.4	-	-
Fernández-Mayoralas et al. [15]	Range, 10 to 25 mm	Case 1: approximately 1:9 Case 2: 2:1 Case 3: 2:1 Case 4: approximately 2:1	-	-
Rodés-Cabau et al. [16]	ASD median size for patients with new-onset migraine headaches, 16 mm ASD median size for patients without new-onset migraine headaches, 15 mm	Patients with new-onset migraine headaches median, 2.1 Patients without new-onset migraine headaches median, 2.1	Patients without new-onset migraine headaches: 13% of these patients had an atrial septal aneurysm. Patients with new-onset migraine headaches: None	-
Castellini et al. [17]	Case 1: 25 mm Case 2: size not mentioned Case 3: size not mentioned	-	-	Case 1: no residual shunt Case 2: minimal residual shunt Case 3: no residual shunt
Wei et al. [18]	New-onset MHA, 20 ± 0.9 mm Without new-onset MHA, 16 ± 1 mm	New-onset MHAs, 2.9 ± 0.22 Without new-onset, 2.3 ± 0.18	-	With new-onset migraine: 1. No residual small ASD, 14 out of 24 patients (58%) 2. Transient residual small ASD, 7 out of 24 patients (29%) 3. Persistent residual small ASD, 3 out of 24 patients (13%) Without new-onset MHAs: 1. No residual small ASD, 11 out of 16 patients (69%) 2. Transient residual small ASD, 5 out of 16 patients (31%) 3. Persistent residual small ASD, 0 out of 16 patients (0%)
Kato et al. [19]	No MHA (n = 155): The mean ASD size was 16.5 ± 5.4 mm De novo MHA (n = 23): The mean ASD size was 16.8 ± 4.7 mm MHA exacerbated (n = 11): The mean ASD size was 17.0 ± 5.9 mm MHA without change (n = 11): The mean ASD size was 19.4 ± 5.3 mm MHA improved (n = 7): The mean ASD size was 18.3 ± 5.7 mm	No MHA (n = 155): Qp/Qs = 2.3 ± 1.0 De novo MHA (n = 23): Qp/Qs = 2.4 ± 0.9 MHA exacerbated (n = 11): Qp/Qs = 2.8 ± 1.3 MHA without change (n = 11): Qp/Qs = 2.5 ± 0.8 MHA improved (n = 7): Qp/Qs = 2.4 ± 0.9	-	-
Sharma et al. [20]	-	-	-	Yes (after the removal of the device, 53 patients (91%) had primary direct repair of the residual defect while 5 patients (9%) required patch closure)
Matsuoka et al. [21]	9.2 to 16.0 mm	1.6 to 2.3	-	-
Huang et al. [22]	Direct ASD size is not mentioned (diameter of occluder for symptomatic group = 19.80 ± 6.67 mm, for asymptomatic group = 17.40 ± 7.28 mm)	-	-	-
Wintzer-Wehekind et al. [23]	Measured by transesophageal echocardiography (TEE): Aspirin + clopidogrel group: Mean size was 15.7 mm (SD 5.7 mm) Aspirin + placebo group: Mean size was 16.9 mm (SD 5.7 mm) Measured by balloon (for balloon sizing): Aspirin + clopidogrel group: Mean size was 20.3 mm (SD 6.3 mm) Aspirin + placebo group: Mean size was 21.7 mm (SD 5.3 mm)	Aspirin + clopidogrel group: ratio was 1.93 (SD 0.84) Aspirin + placebo group: ratio was 1.89 (SD 0.81)	Aspirin + clopidogrel group: 15.5% (11 patients) Aspirin + placebo group: 16.4% (11 patients)	At hospital discharge: Aspirin + clopidogrel group: 27 patients (32.1%) Aspirin + placebo group: 26 patients (29.8%) At 3-month follow-up: Aspirin + clopidogrel group: 8 patients (12.7%) Aspirin + placebo group: 9 patients (10.3%)
Gangopadhyay et al. [24]	Patient 1: 24 mm Patient 2: 16 mm Patient 3: 23 × 18 mm Patient 4: 32 × 30 mm Patient 5: 16 × 18 mm	-	-	Patient 1: small intradevice residual shunt Patient 2: small intradevice residual shunt Patient 3: no residual shunt Patient 4: no residual shunt Patient 5: no residual shunt
Smerdziński et al. [25]	Patients with transient headache episodes (THE): 12 mm (range 8 to 14.5 mm) Patients without THE: 10 mm (range of 8 to 13 mm)	-	Yes (in the case of 63 patients (7.5%), the atrial septum was qualified as aneurysmatic)	Yes (residual shunt after 24 h: 12.5% of patients with transient headache episodes (THE) had a residual shunt compared to 4.9% of patients without THE (P = 0.03)) Residual shunt after 1 year: 7.5% of patients with THE had a residual shunt compared to 1.0% of patients without THE (P < 0.001)

Residual small defect

The presence of a residual defect was observed in seven studies [13, 17, 18, 20, 23–25]. Among the five patients described by Gangopadhyay et al. [24], two had a small intradevice residual shunt, while the other three patients had no residual shunt. In Wintzer-Wehekind et al.’s [23] study, at hospital discharge, 27 patients (32.1%) in the aspirin + clopidogrel group and 26 patients (29.8%) in the aspirin + placebo group had a small residual defect after the procedure for ASD closure. At the 3-month follow-up, 8 patients (12.7%) in the aspirin + clopidogrel group and 9 patients (10.3%) in the aspirin + placebo group still had a residual defect. In Sharma et al.’s [20] study, after the removal of the device, 53 patients (91%) underwent primary direct repair of the residual defect, while 5 patients (9%) required patch closure. In Wei et al.’s [18] study, 58% of patients had no residual small ASD, 29% had a transient residual small ASD, and 13% had a persistent residual small ASD.

ASD closure

Among the various methods for ASD closure (Table 4), the studies reviewed primarily focused on the transcatheter approach. All studies performed transcatheter closure. However, Sharma et al. [20] evaluated the surgical removal of the offending device in patients with refractory nickel allergy and subsequent surgical closure. The Amplatzer ASD occluder device was the most frequently used, being utilized in 12 studies [13–21, 23–25]. In the study by Smerdziński et al. [25], a variety of septal occlusion devices were employed, including the Amplatzer Cribriformis occluder, Hyperion, Cocoon, Cardi-O-Fix, Figulla, HeartR, and Cera. Additionally, the studies by Sharma et al. [20] and Matsuoka et al. [21] utilized the Helex septal occluder and Occlutech Figulla Flex II, respectively.

Table 4.

ASD device

Author	Device used for closure	Device manufacturer
Wilmshurst et al. [13]	Amplatzer septal occluder	-
Mortelmans et al. [14]	Amplatzer septal occluder	-
Fernández-Mayoralas et al. [15]	Amplatzer septal occluder	-
Rodés-Cabau et al. [16]	Amplatzer septal occluder	AGA Medical, Golden Valley, MN, USA
Castellini et al. [17]	Amplatzer septal occluder	Ab-bott Laboratories
Wei et al. [18]	Amplatzer septal occluder	AGA Medical, Golden Valley, MN, USA
Kato et al. [19]	Amplatzer septal occluder	AGA Medical, Golden Valley, MN, USA
Sharma et al. [20]	Amplatzer septal occluder; Helex septal occluder	St Jude Medical Corporation, St Paul, Minn; W. L. Gore & Associates, Inc, Medical Products Division, Flagstaff, Ariz
Matsuoka et al. [21]	Amplatzer septal occluder; Occlutech Figulla Flex II	Ab-bott Laboratories; Occlutech International AB
Huang et al. [22]	Amplatzer septal occluder	-
Wintzer-Wehekind et al. [23]	Amplatzer septal occluder	St. Jude
Gangopadhyay et al. [24]	Amplatzer septal occluder	AGA Medical Corporation, Plymouth, MN, USA
Smerdziński et al. [25]	Amplatzer septal occluder in 685, Amplatzer Cribriformis occluder in 9, Hyperion in 52, Cocoon in 29, Cardi-o-Fix in 25, Figulla in 21, HeartR in 17, Cera in 11	-

Anticoagulant use

Anticoagulant use was prevalent across all studies (Table 5) before ASD closure, given the high risk of device thrombosis associated with the septal occluder device. In the studies reviewed, heparin was administered before closure in eight studies [15, 16, 18, 19, 21, 23–25]. In these studies, all patients, including controls, received intravenous heparin at dosages of either 50 U/kg or 100 U/kg. Clopidogrel was also frequently used as an anticoagulant prior to ASD closure [15, 16, 18, 19, 21, 24]. Across various studies, a total of 14 patients received clopidogrel alone, without other anticoagulants. In these studies [15], dosages for adults were either 300 mg/day or 325 mg/day, and 37.5 mg/day for children. Five studies [16, 18, 19, 21, 24] also utilized a dosage of 75 mg/day for patients. The study by Wintzer-Wehekind et al. [23] implemented a randomized, blinded treatment assignment before ASD closure, dividing participants into two groups: one receiving aspirin and a placebo, and the other receiving clopidogrel and aspirin, with daily doses of 325 mg and 75 mg, respectively.

Table 5.

Anticoagulation details

Author	Clopidogrel	Aspirin	Heparin	Warfarin	Ticlopidine	Change in drugs	Combination drugs
Wilmshurst et al. [13]	-	Aspirin 150–300 mg/day	-	-	-	Aspirin for 6 months plus clopidogrel 75 mg daily for the first month	Aspirin with clopidogrel in 3 patients
Mortelmans et al. [14]	-	-	-	-	-	-	-
Fernández-Mayoralas et al. [15]	Clopidogrel 300 mg	Aspirin 50–500 mg/day	All patients 4000 U IV	-	-	-	-
Rodés-Cabau et al. [16]	Clopidogrel 75 mg/kg	Aspirin (325 mg/day in adults, 80–125 mg/day in children)	All patients 100 U/kg	Warfarin administered	-	-	Aspirin and clopidogrel in 29 patients; aspirin and warfarin in 3 patients
Castellini et al. [17]	-	Aspirin 200 mg/day; 100 mg/day; 100 mg/day	-	-	-	-	-
Wei et al. [18]	Clopidogrel 75 mg/day adults; 37.5 mg/day children	Aspirin (200 mg/day adults; 3–5 mg/kg/day children)	All patients 50 U/kg	-	-	Aspirin changed to clopidogrel in 4 patients	-
Kato et al. [19]	Clopidogrel 75 mg/day	Aspirin (75–330 mg/day)	All patients — heparin	Warfarin administered	Ticlopidine (200 mg/day)	-	Aspirin and ticlopidine (100–200 mg/day) in 42 patients; aspirin and warfarin in 4 patients; ticlopidine and warfarin in 2 patients
Sharma et al. [20]	-	-	-	-	-	-	-
Matsuoka et al. [21]	Clopidogrel 75 mg/kg	Aspirin 100–200 mg/day for adults; 5 mg/kg/day for pediatric population	All patients 100 U/kg	Warfarin administered	-	-	-
Huang et al. [22]	-	-	-	-	-	-	-
Wintzer-Wehekind et al. [23]	-	-	Aspirin (80 mg/day) and placebo in 1 group; aspirin and clopidogrel (75 mg/day) in other group	-	-	-	Aspirin (80 mg/day) and placebo in 1 group; aspirin and clopidogrel (75 mg/day) in other group
Gangopadhyay et al. [24]	Clopidogrel 3 mg/kg	All patients received aspirin 5 mg/kg	All patients 100 U/day	-	-	-	-
Smerdziński et al. [25]	-	-	All patients 100 U/day	-	-	-	-

Aspirin was the second most frequently used anticoagulant in the studies reviewed. Four studies [15–17, 21] administered aspirin prior to ASD closure, with dosages ranging from 150 to 325 mg/day. Warfarin was administered in three studies [16, 19, 21] and ticlopidine in one study [19]. In the study by Kato et al. [19], 5 patients who developed adverse reactions to clopidogrel were switched to ticlopidine at a dose of 200 mg/day.

Among the studies reviewed, combination therapies were also utilized, most commonly aspirin with clopidogrel in three studies [15, 17, 24]. These combination therapies provided more effective anticoagulant management for the prevention of thrombosis during ASD closure. Notably, the study by Kato et al. [19] also included combinations of aspirin and ticlopidine, ticlopidine and warfarin, as well as aspirin and warfarin.

Migraine characteristics

The incidence of migraine following ASD or patent foramen ovale (PFO) closure varied significantly across studies, ranging from 5 to 87%. Some studies reported high incidence rates of new-onset migraine or headache, with rates between 73 and 87.5% [20, 22]. Other studies found intermediate incidence rates, where migraine occurrence ranged from 12 to 17.5% [16, 19, 21]. In contrast, Smerdziński et al. [25] reported a lower incidence, with only 4.8% of patients experiencing transient headache. The detailed characteristics of these migraine are summarized in Table 6.

Table 6.

Migraine profile

Author	Types of migraine	With/without aura	IHS criteria followed	Time between closure and migraine onset
Wilmshurst et al. [13]	5/30 patients had severe attacks of migraine; 2/5 had very severe attack	-	Yes	1 month
Mortelmans et al. [14]	22/75 developed new-onset migraine	8 had symptoms of aura; 14 without aura	ICHD-2	48 months
Fernández-Mayoralas et al. [15]	Intense pulsating headache with photophobia; Intense pulsating headache along right half of head; severe pulsating headache with photophobia; intense headache with photophobia	-	Yes	Range, 1 to 7 days
Rodés-Cabau et al. [16]	Migraine headache attack seen in 13 patients	With aura in 9 patients	Yes (according to the International Headache Society criteria 1988 and its revised version for pediatric migraine)	Median: 10 days
Castellini et al. [17]	Migraine with aura; visual disturbances, severe non-throbbing headache; non-migraine headache, mild non-pressure type headache; migraine with aura	With aura; with aura; with aura	ICHD-2	Case 1: 24 h Case 2: about 2 months Case 3: about 3 months
Wei et al. [18]	Migraine headache attack; mild to moderate in 4; severe in 20	Aura was present in 13 patients; photophobia in 10 patients; phenophobia in 15 patients	Yes	Median: 1 month
Kato et al. [19]	De novo migraine headache in 23; migraine headache exacerbated in 11; migraine headache without change in 11; migraine headache improved in 7	Aura seen in 13/23 patients with de novo migraine	ICHD-2	Within 30 days
Sharma et al. [20]	-	-	-	1 day to 1 year
Matsuoka et al. [21]	New-onset migraine headache in 47/167 patients	Aura symptoms seen in 11/47 patients	ICHD-3	7 (3–10) days
Huang et al. [22]	New-onset migraine in 51/567 patients	-	-	-
Wintzer-Wehekind et al. [23]	New-onset migraine in 27/171; continued migraine in 9/171	Aura seen in 7 patients	ICHD-2	3 months
Gangopadhyay et al. [24]	Severe headache reported by 5/325 patients	Aura symptoms seen in 4/5 patients	Yes	2 weeks
Smerdziński et al. [25]	Transient headache episodes noticed in 40 out of 840 patients	-	The authors defined THE as headaches occurring up to 24 h after the procedure of more than mild intensity not reported in the patient’s history and lasting no more than 24–48 h	24 h

Aura symptoms

Aura was a common entity estimated along with migraine symptoms. A total of 10 studies [14–19, 21–24] reported the presence of an aura associated with the migraine attacks. In the study by Gangopadhyay et al. [24], a total of four out of five patients presented with migraine attacks, in contrast to the study by Fernández-Mayoralas et al. [15], which reported no aura symptoms among all three patients that were included.

Latency period

The period between the closure of ASD and the appearance of MHA varied from 1 day to 1 year among the studies under review. Among the four cases described by Fernández-Mayoralas et al. [15], the latency period ranged from 1 to 7 days. Wei et al. [18], in their study, noted a median latency period of 1 month for new-onset MHA. Sharma et al. [20] depicted wide variability in the latency period, which ranged from 1 day to 1 year. Kato et al. [19] noted the appearance of symptoms within 30 days of ASD closure. In Castellini et al.’s [17] case series, one patient developed MHA within 1 day, while two other patients developed MHA after 2–3 months.

Atrial septal aneurysm

Three studies [16, 24, 25] reported the presence of an atrial septal aneurysm after ASD closure. Rodés-Cabau et al. [16] noted that 13% of patients without MHA had an atrial septal aneurysm while none of the patients with MHA developed the same. Wintzer-Wehekind et al. [23] in their study noted that in the aspirin + clopidogrel group, 15.5% (n = 11) of the patients and in the aspirin + placebo group, 16.4% (n = 11) of the patients developed an atrial septal aneurysm. In Smerdziński et al.’s [25] study, out of 840 patients, the atrial septum was found to be aneurysmal in 63 patients (7.5%).

Diagnostic criteria

Nine studies [13–19, 21, 23, 24] used the International Classification of Headache Disorders (ICHD) criteria by the International Headache Society (IHS) for the diagnosis of the headache disorders. Two studies [20, 22] did not specify any criteria. All studies used the latest edition at the time of publication whereas three studies [13, 15, 24] did not specify the iteration used. Rodés-Cabau et al. [16] utilized the IHS-1988 criteria and its revised version for pediatric migraine. Although Smerdziński et al. [25] did not follow the IHS/ICHD criteria, they defined “THE” as headache occurring up to 24 h after the procedure, of more than mild intensity, not reported in the patient’s history, and lasting no more than 24–48 h.

Treatment and prognosis

Migraine attacks were treated with a variety of modalities, including non-steroidal anti-inflammatory drugs (NSAIDs) and calcium channel blockers (Table 7). Spontaneous resolution was reported in a total of eight studies [13–19, 21, 23–25]. In the study by Castellini et al. [17], three cases were included, with two patients experiencing spontaneous resolution. Notably, one of these patients was treated with lamotrigine to manage migraine attacks. Additionally, Wintzer-Wehekind et al. [23] found that migraine attacks spontaneously resolved within 6 to 12 months.

Table 7.

Management

Author	Spontaneous resolution	NSAIDS (except aspirin)	CCB	Removal of device	Resolution after anti-platelet change	Recurrence
Wilmshurst et al. [13]	Yes	-	-	-	Yes	Yes
Mortelmans et al. [14]	Yes	-	-	-	-	-
Fernández-Mayoralas et al. [15]	Yes	Yes (ibuprofen)	Flunarizine	-	-	-
Rodés-Cabau et al. [16]	Yes (4 patients experienced transient MHAs that spontaneously disappeared)	-	-	-	-	-
Castellini et al. [17]	Yes (for cases 2 and 3 but in case 1 recurrent attacks seen)	Yes (for 1 case)	Flunarizine used for cases 1 and 3	-	-	Yes (for case 1 only)
Wei et al. [18]	Yes (7 out of 24 with new-onset migraine)	-	-	-	Yes	-
Kato et al. [19]	Yes (de novo MHA group: spontaneous resolution in 8 patients; exacerbation of existing MHA group: in 1 patient)	Yes (ibuprofen)	Lomerizine	-	Yes	-
Sharma et al. [20]	The study does report some observations about the resolution of symptoms following the removal of the nickel alloy device	-	-	Yes	-	-
Matsuoka et al. [21]	Yes	-	-	-	Yes	-
Huang et al. [22]	-	-	-	-	-	-
Wintzer-Wehekind et al. [23]	Yes (most migraine attacks that occurred within the first 3 months after transcatheter ASD closure resolved or improved spontaneously within 6 to 12 months)	Yes (ibuprofen in 1 case)	-	-	Yes	Yes
Gangopadhyay et al. [24]	Yes (3 out of 5 patients (60%))	Yes (ibuprofen was used in 1 patient after paracetamol was reported to be ineffective)	-	-	-	-
Smerdziński et al. [25]	Yes	-	-	-	-	-

The most common approach for managing migraine included the use of aspirin and anti-platelet therapy, which were tested in numerous studies. Aspirin was commonly administered in doses ranging from 50 to 500 mg/day, both pre- and post-procedurally, and combined with clopidogrel in some cases [13, 15–19, 21, 24]. The combination of clopidogrel and aspirin was found to be more effective than aspirin alone in preventing migraine attacks, suggesting the involvement of platelets in migraine pathogenesis, particularly through serotonin mechanisms [13, 23, 24]. Furthermore, anti-platelet therapy was reported to reduce the frequency and severity of migraine, with clopidogrel being the most frequently prescribed agent [13, 23, 24]. Clopidogrel was the most frequently prescribed anti-platelet agent among the studies.

Another frequently employed treatment strategy involved the use of calcium channel blockers (CCBs). Medications such as flunarizine and lomerizine were used to manage migraine symptoms, particularly in the post-closure period following ASD procedures. These treatments were utilized in several studies, with flunarizine [15, 17] being reported in a few cases and lomerizine [19] noted in another. Other treatments included ibuprofen [15, 19, 23, 24], which was used in some studies to alleviate acute migraine symptoms, and lamotrigine [17], which was reported to result in the resolution of migraine attacks in at least one patient.

Less common but notable strategies included device removal and residual defect closure, which showed significant improvement in symptoms post-procedure, with some patients reporting complete resolution of migraine after device explantation [20]. Additionally, the potential impact of larger ASD sizes and low CGRP (calcitonin gene-related peptide) levels prior to ASD closure in predicting new-onset migraine was noted, suggesting a more individualized approach to treatment might be beneficial [18, 21].

While no single strategy emerges as a definitive treatment for migraine post-ASD closure, the most commonly employed approaches were aspirin combined with anti-platelet therapy, CCBs, and ibuprofen. These strategies, along with other targeted therapies such as lamotrigine and device explantation, offer a variety of options for managing migraine attacks post-ASD closure.

Discussion

The present systematic review is to the best of our knowledge the first of its kind to evaluate the association between new-set headache and ASD closure. While previous reviews have looked into the association between migraine and PFO closure [26–33], there has been no comprehensive review on the present topic. The role of migraine after ASD closure has been attributed to various pathophysiological mechanisms, but in the present review we have noted a significant association between anti-platelet use (aspirin and clopidogrel) and their association with the incidence and management of headache. We also noted a focus in the adult population only. This could be associated with the fact that children and adolescents may not be able to articulate or explain headache as a significant complication [34, 35] or because that earlier closure allows for spontaneous resolution of headache in children as well, as seen in many of the presented cases.

The present study noted a gender-related predisposition, with more women experiencing headache compared to men. However, attributing this sex predilection specifically to ASD closure may be premature, as evidence suggests that women generally have a higher burden of headache, both in frequency and intensity, compared to men [36–38]. However, this could also be attributed to hemodynamic changes, as the most striking sex-related difference in cardiovascular adjustments is observed in vascular tone. Women tend to exhibit reduced vasoconstriction and lower vascular resistance compared to men [39].

It seems plausible that both intracranial and extracranial arteries play a nociceptive role in migraine [40–42]. Intracranial vessel dilation, particularly in meningeal arteries, has been implicated in migraine. In one study, during the experimental induction of migraine, dilation of both the middle meningeal and middle cerebral arteries were observed [40]. These experimental and clinical findings suggest that vasodilation is neither necessary nor sufficient to cause migraine pain [43]. However, it should be noted that they do not rule out the possibility of vasoconstriction acting as an initial trigger for subsequent migraine pain. While the vascular theory of migraine [44] is important and critical, its role in the context of the present study may be more related to high exacerbation of hemodynamic variables in women than men in ASD patients, which could be implicated but is purely hypothetical as no studies on this have been performed [45].

We noted significant improvement in headache in those patients without spontaneous resolution with changes and modifications to anti-platelet therapy. A common consensus reached is that addition of clopidogrel along with aspirin to the traditional anti-platelet therapy is effective in managing and reducing incidence of migraine. This could be attributed to reduced inflammation or improved vascular integrity in migraine. However, studies by Spencer et al. [46] and Sommer et al. [47] show that anti-platelet regimens including a P2Y12 inhibitor, compared to those excluding it, resulted in a lower rate of new-onset migraine. While this evidence is associated predominantly with PFO, they may be extrapolated to ASD as well.

The findings related to ventricular efficiency, ventricular-arterial coupling, and the possible role of migraine-related peptides like CGRP have all been discussed in varying detail in the included papers. CGRP is a well-known peptide implicated in migraine along with pituitary adenylate cyclase-activating peptide (PACAP) [48, 49]. However, there is limited evidence at present to see if there is any relevant association between these peptides and headache in ASD.

Future studies could look into the role of age and gender-related associations between ASD closure and new-onset headache. In particular, the evaluation of hemodynamic changes and their association with headache in men and women would be of significant interest. Additionally, the role of residual defects and nickel allergy [50] are significant aspects that would merit consideration. In particular, these cases predominantly appear to affect children and are burdened with re-operation and surgical intervention [51–53]. Therefore, it may be prudent to evaluate for nickel allergy in these patients prior to device implantation [54]. Further long-term studies to assess the persistence of headache and controlled trials to evaluate the efficacy of different therapeutic approaches post-ASD closure are other important topics. In particular, with the rise in use of monoclonal antibodies [55] and gepants [56] for headache disorders, their use for those with non-resolving or persistent headache after ASD closure warrants consideration.

Limitations

This review adheres to PRISMA guidelines and is registered with PROSPERO, featuring studies in two different languages from eleven different countries, making it the most comprehensive contemporary review on this topic to date. However, it is important to acknowledge certain limitations. The included studies vary significantly in size and type, ranging from large-scale studies to small case series. This diversity might introduce bias, particularly if some patients, attributing their headache to neurological issues rather than a consequence of ASD closure, sought advice from neurologists. Additionally, the onset of headache months or even years post-closure could potentially be linked to different pathologies, as it is challenging to definitively ascertain the headache’s origin in these cases. Furthermore, the variability in follow-up duration, management strategies, and ASD closure techniques—dependent on physician preference—could lead to inconsistent outcomes without a clear, unified explanation for all findings reported.

Conclusion

The present systematic review highlights the incidence of headache and migraine post-transcatheter ASD closure, identifying varying rates across studies, with high incidence rates (up to 87.5%) in some cases, and lower rates (as low as 4.8%) in others. The choice and dosage of anticoagulants, typically aspirin, clopidogrel, and heparin, were critical in minimizing thrombotic risks associated with the procedure. The combination of aspirin and clopidogrel was found to be especially effective, further pointing to platelet involvement in migraine pathogenesis. The review reiterates that proactive management with anti-platelet therapy can mitigate migraine incidence post-ASD closure. Effective treatment options include both preventive (aspirin and clopidogrel) and symptomatic relief (NSAIDs and CCBs), with device removal and residual defect closure showing promise for refractory cases.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

H.S.A. contributed to the conception and design of the study. S.R.P. and H.S.A. performed the study selection, while M.Z. and A.F.D. were responsible for data extraction. H.S.A., V.V., A.F.D., and M.Z. contributed to drafting the manuscript. H.S.A. and S.R.P. critically revised the manuscript for important intellectual content. All authors provided final approval of the version to be published and agree to be accountable for all aspects of the work.

Funding

None.

Data availability

No new data was generated in the present study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Yes.

Conflict of interest

None.