Long-term headache following cerebral venous thrombosis: a systematic review Authors: Diana Doukhi, Jerome Mawet, Caroline Roos, Mikael Mazighi, Isabelle Crassard Published: 2026 ABSTRACT AbstractBackgroundCerebral venous thrombosis (CVT) predominantly affects young adults and frequently presents with headache. Although functional outcomes are generally favorable, many patients report persistent or recurrent headache long after the acute event. This systematic review aimed to assess the prevalence, clinical characteristics, associated factors, and management of long-term headache following CVT.Main bodyA systematic search of PubMed and Embase (2000–2025) identified studies reporting original data on long-term headache after CVT in adult populations. Case series including fewer than 30 patients were excluded. Given substantial heterogeneity in study design, definitions, and outcome measures, a qualitative synthesis was performed. Fifteen studies including 2,136 patients were analyzed. The reported prevalence of post-CVT headache ranged from 14% to 59%. Headache most often resembled primary headache disorders, particularly tension-type and migraine-like phenotypes. Available data suggest that de novo headache may represent the most frequent presentation, although modification of pre-existing headache disorders also occurs. Pre-existing headache and depression were associated with post-CVT headache, whereas intracerebral hemorrhage was not. Associations with venous recanalization, thrombosis location, and anticoagulation delay were inconsistent. Secondary mechanisms, including intracranial hypertension, dural arteriovenous fistula, CVT recurrence, and medication-overuse headache, were identified in some patients. Data on management strategies were limited.ConclusionLong-term headache is a frequent but under-recognized complication of CVT that may significantly impact quality of life. The absence of a standardized definition contributes to heterogeneity across studies. After exclusion of secondary causes, post-CVT headache encompasses both de novo and modified primary headache phenotypes. A consensus term such as “new or modified recurrent headache attributed to past CVT” may help describe this spectrum and improve clinical characterization and future research.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s10194-026-02429-5.Keywords: Headache, Cerebral venous thrombosis, Migraine, Tension-type headache, Intracranial hypertension Background Cerebral venous thrombosis (CVT) is an important cause of stroke in young adults, with an estimated incidence of 12.1 per million person-years [1]. It is three times more frequent in women than in men [2,3]. The clinical presentation of CVT is highly variable, with headache being the most common symptom at onset, reported in up to 90% of patients [3] and occurring as an isolated symptom in about 14% [4]. Overall prognosis is generally considered favorable, with 80–90% of patients achieving functional independence (i.e., modified Rankin Scale score -mRS- 0–2) [3–5]. However, despite good functional outcomes measured by conventional disability scales, a substantial proportion of patients experience persistent long-term symptoms, particularly chronic or recurrent headache [6]. These symptoms remain insufficiently recognized and under-investigated, especially in working-age populations, in whom they may lead to functional impairment and reduced quality of life. In this systematic review, we aimed to evaluate the prevalence, clinical characteristics, associated factors, and management of long-term headache following CVT, in order to improve clinical recognition and optimize patient care. Methods Study design and reporting A systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines [7]. Search strategy A comprehensive literature search was performed in the Embase and PubMed databases. Two investigators (DD and IC) independently conducted the searches using predefined and database-specific search strategies. The full search strategies are provided in Supplementary Table 1. The search covered studies published between January 1, 2000, to August 1, 2025, and was restricted to human studies published in English. All identified records were imported into Zotero reference management software, and duplicates were removed prior to screening. Reference lists of relevant articles and literature known to be pertinent by the authors were also screened. Eligibility criteria Studies were eligible for inclusion if they met the following criteria: (1) included adult patients with CVT; studies also including some adolescents were eligible when adult patients were predominant or separate adult data were unavailable; (2) reported CVT confirmed by neuroimaging, including magnetic resonance imaging combined with magnetic resonance venography, CT/CT venography and/or digital substraction angiography; (3) provided original data on long-term headache following CVT, including prevalence, clinical characteristics, associated risk factors, or management strategies. Observational studies (including cohort, case–control, and cross-sectional studies) reporting original data were eligible. Studies were excluded if they involved exclusively pediatric populations; focused on highly selected CVT populations with specific etiologies (e.g., cancer- or infection-related CVT) that may have distinct clinical profiles and outcomes; were case series with fewer than 30 patients, to reduce the risk of small-sample and publication bias; were reviews, editorials, case reports, conference abstracts, or letters without original data. Study selection Titles and abstracts were independently screened for eligibility by two reviewers (DD and IC). Studies were excluded at this stage if they were out of scope. Full-text articles of potentially eligible studies were subsequently reviewed in detail for inclusion. Any discrepancies between reviewers were resolved through discussion, with adjudication by a third reviewer (JM) when necessary. Data extraction Data were independently extracted by two reviewers (DD and IC) using a standardized form. The following information was collected from each included study: first author, year of publication, country, study design, sample size, patient demographics (age and sex), prevalence of headache at admission, follow-up duration and timing of headache assessment, temporal definitions of long-term headache, prevalence estimates, headache clinical characteristics (phenotype, intensity, frequency, and duration), headache assessment methods, reported risk factors (including pre-existing headache, depression, intracranial hypertension, parenchymal lesions, thrombosis location, recanalization status, and anticoagulation timing, when available), and reported treatment strategies, including anticoagulant treatment, acute or preventive headache therapies, and intracranial pressure-lowering treatments. Data synthesis Given the heterogeneity of study designs, populations, definitions of long-term headache, and outcome measures, a qualitative narrative synthesis was performed. Study findings were summarized and compared descriptively, with particular attention to consistency across studies, sources of heterogeneity, and methodological limitations. Risk of bias assessment Risk of bias was assessed according to the study design. Observational cohort and case-control studies were evaluated using the Newcastle–Ottawa Scale (NOS). Cross-sectional studies were assessed using the Joanna Briggs Institute (JBI) critical appraisal checklist for analytical cross-sectional studies. The randomized controlled trial was evaluated using the Cochrane Risk of Bias 2 (RoB 2) tool. All assessments were performed independently by two reviewers (DD and IC), and disagreements were resolved by consensus or, when necessary, by consultation with a third reviewer (JM). Results The database search identified 6,987 records (PubMed: 1362; Embase: 5625). After duplicates were removed, the titles and abstracts of 5,824 studies were screened yielding 45 articles for full-text screening. Two additional studies related to the topic were included. After full-text screening, 15 studies were included in the final qualitative analysis [3,8–21] (Fig. 1). Flow chart of the systematic search Risk of bias assessment is summarized in Table 1. Table 1Risk of bias assessment of included studiesFirst Author, YearStudy DesignRisk of Bias ToolOverall Risk of BiasAguiar de Sousa, 2020 [8]Observational, prospective, multinationalNOSLowBossoni, 2022 [9]Cross-sectionalJBIModerateBreteau, 2003 [10]Observational, retrospectiveNOSModerateBuccino, 2003 [11]Observational, retrospectiveNOSModerateDias, 2024 [12]Observational, retrospectiveNOSLowFerro, 2002 [13]Observation, retrospective and prospectiveNOSLowFerro, 2004 [3]Observational, prospective, multinationalNOSLowField, 2023 [14]Randomized controlled trialRoB 2LowHiltunen, 2016 [15]Observational, retrospectiveNOSLowJi, 2020 [16]Observational, prospectiveNOSLowJi, 2021 [17]Observational, prospectiveNOSLowKoopman, 2009 [18]Case-controlNOSModeratePutaala, 2010 [19]Observational, retrospectiveNOSLowSaroja, 2019 [20]Cross-sectionalJBIModerateStrupp, 2002 [21]Observational, retrospectiveNOSModerateNOS: Newcastle–Ottawa Scale; JBI: Joanna Briggs Institute; RoB 2: Cochrane Risk of Bias 2 Risk of bias assessment of included studies NOS: Newcastle–Ottawa Scale; JBI: Joanna Briggs Institute; RoB 2: Cochrane Risk of Bias 2 Characteristics of the study populations Sample sizes ranged from 38 to 624 patients, yielding a total of 2,136 patients. Studies were published between 2002 and 2024 and were predominantly conducted in Europe (n = 10), with additional studies from India, China, Brazil, and Canada. Female predominance was observed across studies, ranging from 40% to 92% of participants. Reported mean or median age ranged from 31 to 48 years. Follow-up duration varied widely, from 3 to 145 months [3,8–21] (Table 2). Table 2Patient characteristics of studies on headache following cerebral venous thrombosis (CVT)First Author, YearCountrySample SizeAge(Years)Female Sex (%)Time of Follow-up (months)Aguiar de Sousa, 2020 [8]Portugal6840 (14)56 (82)3Bossoni, 2022 [9]Brazil10038.3 (13.7)82 (82)12Breteau, 2003 [10]France5539 [16–68]42 (76)36Buccino, 2003 [11]Italy3838.5 (16.5)35 (92)42Dias, 2024 [12]Portugal13138.9 (12.6)104 (79)3.5Ferro, 2002 [13]Portugal14222Ferro, 2004 [3]Portugal62439.1 (16–86)465 (75)16Field, 2023 [14]Canada5348 [38.5–73.2]35 (66)12Hiltunen, 2016 [15]Finland16138 [27–54]106 (67)39Ji, 2020 [16]China16435 [25–49]87 (53)22–25Ji, 2021 [17]China32534 [24–46]191 (59)13Koopman, 2009 [18]Netherlands4431 [16–68]36 (82)63Putaala, 2010 [19]Finland9139.3 (18–75)64 (70)6Saroja, 2019 [20]India10034.97 (10.06)40 (40)26Strupp, 2002 [21]Germany4038.5 (19–66)30 (75)145Age is expressed as mean (SD or range) or median [IQR] Patient characteristics of studies on headache following cerebral venous thrombosis (CVT) Age is expressed as mean (SD or range) or median [IQR] Definition and prevalence Headache following CVT is not specifically defined in the third edition of the International Classification of Headache Disorders (ICHD-3) [22], unlike persistent headache attributed to ischemic stroke, intracranial hemorrhage, or arterial dissection. Consequently, heterogeneous terminology has been used in the literature, including persistent [9], residual [16,17], and post-CVT [12] headache. Definitions vary widely across studies, ranging from headache occurring at least three times with similar characteristics within the preceding three months [9] to headache occurring more than once per week [15]. One study focused exclusively on severe headache, defined as headache requiring bed rest or hospital admission [3]. None of the included studies reported the interval between the acute CVT event and the onset or recurrence of long-term headache symptoms. Accordingly, prevalence estimates show substantial heterogeneity. Fourteen of the fifteen included studies reported prevalence rates ranging from 14% to 59% of CVT patients [3,8–13,15–21] (Table 3). For consistency, the term post-CVT headache (PCH) will be used throughout this review [12]. Table 3Definition and prevalence of post CVT headacheFirst AuthorDefinition of post CVT headachePrevalence of post CVT headacheAguiar de Sousa [8]Reported head pain at the day of evaluation with intensity of at least 5 in the visual analogical scale(42)Bosson [9]At least three episodes of headache with the same characteristics within the 3 months prior59 (59)Breteau [10]Migraine, tension headache or other types according to the ICHD29 (52.7)Buccino [11]NS10 (29)Dias [12]Headache complain during the following visit60 (45.8)Ferro [13]NS59 (47)Ferro [3]Severe headache, requiring bed rest or hospital admission88 (14.1)Field [14]NSHiltunen [15]Headache more than once a week33 (20)Ji [16]NS43 (27.7)Ji [17]Remaining headacheattack (related to CVT or a history of chronic headache before CVT) within 1 month before the last follow-up visit.114 (35.1)Koopman [18]NS19 (43)Putaala [19]NS(23)Saroja [20]NS31 (31)Strupp [21]NS6 (15)NS: Not Specified; CVT: Cerebral Venous Thrombosis Definition and prevalence of post CVT headache Remaining headache attack (related to CVT or a history of chronic headache before CVT) within 1 month before the last follow-up visit. NS: Not Specified; CVT: Cerebral Venous Thrombosis Characteristics of PCH Phenotype Detailed semiological characteristics of PCH were reported in the study of Bossoni et al. [9], which included 100 patients. Headache was most commonly pressure-like (46%) or throbbing (37%). It was bilateral in 24% of patients and unilateral in 56%, with no specification as to whether laterality corresponded to the site of venous thrombosis. Pain most frequently involved the temporo-occipital region (22%), followed by the forehead (19%), and the temporal, vertex or nuchal regions (15% each) [9]. Associated symptoms were common, including nausea (42%), photophobia (39%), and phonophobia (34%) [9]. Headache worsening with physical exertion (56%) or during menses (58%) was also frequently reported [9]. Three studies characterized the headache phenotype associated with PCH [9,10,12]. The most frequent phenotype was tension-type headache (TTH)-like, accounting for 45–53% of cases, while migraine-like headache was reported in 18–48% of patients [9,10,12]. In one study, 12% of patients reported more than one headache phenotype [9]. Less common headache presentations were described in the study by Dias et al. [12], which included 131 patients. These comprised other headache types (23%), including non-elsewhere-classified (10%), stabbing headache (3%), facial pain (2%), and unspecified headache (8%). Secondary headache was identified in 10% of patients, including dural arteriovenous fistula (5%), intracranial hypertension (3%), and worsening or recurrent CVT (2%). Intracranial hypertension was further reported in other studies. Aguiar de Sousa et al. (n = 68). described headache with typical features of intracranial hypertension in 28% of patients with PCH [8], while Ji et al. confirmed intracranial hypertension by lumbar puncture in 33 of 43 patients presenting with severe PCH [17], supporting the involvement of specific pathophysiological mechanisms in PCH, which will be discussed later in this review. Finally, semiological characteristics of PCH differed significantly from those of the headache experienced during the acute CVT phase (p < 0.001) [9]. Intensity and frequency Only three studies reported data on headache intensity and frequency (n = 567) [9,13,17]. PCH intensity was reported as mild or moderate in 62% of patients [17]. Specifically, mild headache was reported in 29–78% of patients [9,13], moderate headache in 5–48% [9,13], and severe headache in 17–38% [9,13,17]. Mean headache frequency reached 13.6 ± 9.9 headache days per month in one study [9]. Burden In the study by Dias et al. [12] (n = 100), no significant association was observed between the presence of PCH and functional outcome at 6 months, as measured by the mRS (mRS < 2: PCH 46% vs. no PCH 54%, p > 0.999). Similarly, Hiltunen et al. reported no association between headache and unemployment in univariate analysis (OR 1.01, 95%CI 0.23–4.44; p = 0.98) [15]. In contrast, Koopman et al. found significantly higher mean Headache Impact Test (HIT-6) scores in PCH patients compared with controls (51.0 ± 10.3 vs. 43.3 ± 7.5; p < 0.001), although the average score remained below the threshold generally considered clinically disabling (HIT-6 < 56) [18]. In the same study, higher psychosocial impact scores on the Sickness Impact Profile (SIP-68) were significantly associated with most CVT sequelae scores, including HIT-6 scores (Spearman ρ = 0.55; p < 0.001) [18]. Functional outcomes were further reported in the study by Bossoni et al. [9], which documented a mean of 2.3 ± 5.1 days of work absence per month in PCH patients. In addition, the number of days with inability to perform instrumental activities of daily living varied according to headache phenotype, being higher in patients with migraine-like PCH (14.2 ± 9.1 days per month) compared with those with TTH-like PCH (0.6 ± 1.6 days). De novo headache In the study by Dias et al. [12] (n = 100), among the 60 patients with PCH, 44 (73.3%) developed a new-onset headache, including 21/60 (35%) with a de novo tension-type headache and 6/60 (10%) with a de novo migraine-like headache. Similarly, in the study by Bossoni et al. [9], 62.7% of the 59 patients with PCH had no prior history of headache. In a smaller cohort, Buccino et al. [11] reported that none of the 10 patients with PCH had complained of headache prior to the cerebrovascular event. Associated factors Pre-existing headache The prevalence and characteristics of headache disorders preceding CVT were reported in several studies, with heterogeneous findings [9,11,12,17]. In the studies by Bossoni et al. [9], Dias et al. [12], and Ji et al. [17], a pre-existing headache disorder was identified in 18–25% of patients. These headaches were mainly primary, including tension-type headache in 44–48% and migraine in 42–48% of patients with a headache history [9,12]. Among patients with a prior headache disorder, two studies specifically examined the relationship between the pre-existing headache phenotype and PCH (n = 231), with inconsistent findings [9,12]. One study found no significant difference between PCH and the pre-existing headache type (p = 0.9) [9], whereas another study reported that PCH was similar to the pre-existing headache in only 39% of patients [12]. Several studies specifically examined the association between pre-existing headache and the occurrence of PCH. In the study by Dias et al. [12], a prior headache disorder was more frequently observed in patients with PCH than in those without PCH (n = 131; 38% vs. 14%; p = 0.001). Similarly, Bossoni et al. [9] reported a higher prevalence of prior headache in the PCH group compared with controls (n = 100; 34% vs. 7%; OR 6.4, 95%CI 1.7–36.3). In the study by Ji et al. [17], a history of chronic headache was more frequently observed in patients with severe residual headache in univariate analysis (n = 325; OR 2.683, 95%CI 1.312–5.489; p = 0.007). Taken together, these findings suggest a consistent association between pre-existing headache and the occurrence of PCH. By contrast, the presence of headache at CVT onset did not differ significantly between patients with and without PCH (85% vs. 75%; p = 0.144). Depression Depression has been evaluated in 6 of 15 included studies [9,11,12,15,18,20]. In the study by Dias et al. [12] (n = 131), depression prior to CVT was reported in 44% of patients with PCH, a prevalence significantly higher than that observed in patients without PCH (19%; p = 0.005) [12]. After adjustment for age, female sex, and recanalization status, previous depression remained associated with an increased likehood of PCH (OR 7.9, 95%CI 1.6–31.1; p = 0.009). At long-term follow-up after CVT, when assessed concomitantly with PCH, depression was reported in 18–32% of patients with PCH across studies [11,15,18,20]. In addition, depression severity (evaluated by the Patient Health Questionnaire -PHQ-) was higher in patients with PCH compared with those without PCH (PHQ score 10.5 ± 7.2 versus 6.1 ± 5.6; p = 0.002) [9]. Intracerebral hemorrhage Data on the association between intracerebral hemorrhage and PCH are reported in three studies (n = 556), consistently showing no significant relationship [9,12,17]. In the study by Bossoni et al. [9], the prevalence of intraparenchymal hemorrhage did not differ between patients with PCH and controls (20.3% vs. 24.4%, p = 0.630), and no association was observed with other lesions, including venous infarction or subarachnoid hemorrhage. Similarly, Dias et al. [12] reported no significant difference in the frequency of intraparenchymal hemorrhage between patients with PCH and those without PCH (64.3% vs. 48.0%, p = 0.232). In the study by Ji et al. [17], cerebral hemorrhage was not associated with headache severity, with similar rates observed in patients with severe residual headache and those with no or mild-to-moderate headache (23.3% vs. 22.3%; OR 1.05, 95%CI 0.49–2.26; p = 0.893). Recanalization The association between PCH and recanalization status was evaluated in six studies, using heterogeneous recanalization classifications - ranging from binary (yes/no) [9] to ordinal (no/partial/complete) [17,19,21] and study-specific grading systems [8,12]- as well as variable assessment time points, resulting in inconsistent results across studies [8,9,12,17,19,21]. Several studies reported an association between absent recanalization and headache persistence or severity. In the study by Putaala et al. [19], headache was the most frequent residual symptom among patients without recanalization (44%), compared with those with partial (36%) or complete recanalization (10%) (p = 0.009). Similarly, Strupp et al. [21] reported focal neurological deficits or headache in 85% of patients without recanalization, compared with 33% of patients with partial or complete recanalization (p < 0.005). Along the same lines, in the study by Ji et al. [17], absence of recanalization was more frequent in patients with severe residual headache. In multivariable logistic regression analysis, absence of recanalization remained independently associated with severe headache (OR 10.158, 95%CI 4.194–24.600; p < 0.001). In contrast, other studies did not identify a significant association between recanalization status and PCH. In the study by Dias et al. [12], recanalization status did not differ significantly between patients with and without PCH (p = 0.598). Although most patients achieved at least partial recanalization, the proportion of PCH tended to decrease with higher recanalization grades (absent: 50%, incomplete: 52%, complete: 36%; p = 0.389). Similarly, Bossoni et al. [9] reported no significant difference in recanalization rates between patients with PCH and controls (75% vs. 66%; p = 0.690). Aguiar de Sousa et al. [8] also found no significant difference in headache frequency, including headache with features of intracranial hypertension, between patients with complete recanalization and those with poorer recanalization grades, although absent recanalization was observed in only 5% of patients. Thrombosis location Data on the association between thrombosis location and PCH are limited. No study assessed the relationship between PCH laterality and thrombus location. In the study by Bossoni et al. [9] (n = 100), no significant association was found between thrombosis location or the number of thrombosed sinuses and the presence or characteristics of PCH. In the study by Dias et al. [12] (n = 131), thrombosis of the superior sagittal sinus (SSS), straight sinus, and cortical veins was more frequent in patients without PCH than in those with PCH (SSS: 34% vs. 60%, p = 0.003; straight sinus: 12% vs. 29%, p = 0.002; cortical vein thrombosis: 8% vs. 22%, p = 0.040). In univariable logistic regression analysis, thrombosis of the SSS (OR 0.33, 95%CI 0.15–0.68; p = 0.003), cortical veins (OR 0.33, 95%CI 0.11–0.98; p = 0.047), and straight sinus (OR 0.32, 95%CI 0.12–0.84; p = 0.021) were associated with lower odds of PCH. After adjustment for age, sex, and recanalization status, only SSS thrombosis remained independently associated with reduced odds of PCH (OR 0.15, 95%CI 0.03–0.56; p = 0.005). In the study by Ji et al. [17], in univariate analysis, thrombosis of the right lateral sinus (n = 325; OR 3.35, 95%CI 1.62–6.91; p = 0.001) and occlusion of two or more sinuses (OR 2.34, 95%CI 1.08–5.07; p = 0.031) were more frequent in patients with severe residual headache. Anticoagulation delay In the only study assessing anticoagulation delay – defined according to the interval between symptom onset and initiation of anticoagulant therapy, later initiation of treatment was associated with PCH, with higher rates in the late-treatment group compared with the early-treatment group (n = 164; 35% vs. 20%; p = 0.029) [16]. This association was also observed in the subgroup of patients with isolated intracranial hypertension (52% vs. 16%; p = 0.007). By contrast, anticoagulation duration was not associated with PCH in another study (n = 100; p = 0.647) [9]. Secondary headache Secondary headache mechanisms associated with PCH were reported in several studies. In the study by Ji et al., (2021) [17], isolated intracranial hypertension was independently associated with severe residual headache in multivariable logistic regression analysis (OR 3.309, 95%CI 1.43–7.63; p = 0.005), together with CVT recurrence (OR 4.72, 95%CI 1.63–13.60; p = 0.004). Dural arteriovenous fistula, a recognized cause and potential complication of CVT, was more frequently observed in patients with PCH than in those without PCH in the study by Dias et al. [12] (12% vs. 1%; p = 0.023). In addition, Bossoni et al. [9] reported a case of dural fistula characterized by red-flag features, including pulsatile tinnitus, aural fullness, progressive worsening of headache, and lack of response to analgesic treatment. Finally, medication-overuse headache was reported more frequently in the PCH group than in controls (34% vs. 0%; p < 0.001) [9]. Management Management of headache after CVT can be challenging, particularly in patients receiving long-term anticoagulation, in whom non-steroidal anti-inflammatory drugs (NSAIDs) are generally avoided. Data on the management of PCH are limited. In the study by Aguiar de Sousa et al. [8], approximately one quarter of patients were treated with analgesics or acetazolamide. In the study by Breteau et al. [10], 49% of patients reported using analgesics at least once per week. Discussion Despite generally reported favorable functional outcomes after CVT, post-CVT headache emerges as a frequent and clinically relevant condition, with potential impact on quality of life in a predominantly young and working-age population. Available data suggest that PCH most often presents as a de novo headache, although modification of pre-existing headache disorders also occurs, and typically resembles primary headache phenotypes such as tension-type headache or migraine. Pre-existing headache and depression appear to be associated with the presence of PCH. However, this review highlights heterogeneity in the available data, which may contribute to an underestimation of the true burden of PCH. Headache is not systematically assessed during follow-up, and commonly used disability scales, such as the mRS, lack sensitivity for capturing headache-related disability. Even in studies specifically investigating headache after CVT, evaluation remains challenging. A major source of heterogeneity across studies is the absence of a consensus definition of PCH, which is not formally recognized in the ICHD-3 [22],. The use of heterogeneous terminology – persistent, residual, or post-CVT headache – reflects this lack of standardization and may obscure clinically meaningful distinctions. In particular, the term persistent may be misleading, as it can suggest a continuous headache, whereas PCH covers a wide range of clinical situations, including de novo headache, modification or exacerbation of a pre-existing headache disorder, and secondary headache mechanisms, with either continuous or discontinuous patterns. In this context, once secondary causes have been excluded, a consensus term such as « new or modified recurrent headache attributed to past CVT », could be proposed. Additional heterogeneity arises from variable follow-up durations (3-145 months), inconsistent headache assessment methods, and the absence of structured headache interviews across studies, which complicates estimation of prevalence and comparison of results. The absence of reporting regarding the timing of headache onset after CVT further limits characterization of the natural history of PCH. Nevertheless, by analogy with persistent headache disorders in the ICHD-3 (e.g., persistent post-traumatic headache) [22], for which chronicity is conventionally defined after 3 months, a temporal threshold of 3 months after the acute CVT event appears to be the most appropriate time point for the first standardized evalaution of long-term headache following CVT. In clinical practice, the evaluation of headache after CVT should first aim to exclude secondary causes. Intracranial hypertension, dural arteriovenous fistula, and CVT recurrence represent relevant and potentially treatable conditions. This step is particularly crucial in patients presenting with de novo headache or atypical features. Clinical warning signs such as pulsatile tinnitus, worsening headache, or poor response to analgesics should prompt specific evaluation for dural arteriovenous fistula, including dedicated vascular imaging, as standard imaging protocols may fail to detect this diagnosis. Systematic fundus examination may also represent a pragmatic and accessible approach to detect intracranial hypertension during follow-up. By contrast, a stable headache phenotype closely resembling migraine or tension-type headache in the absence of red flags may argue against an ongoing secondary headache mechanism. However, once these secondary causes have been excluded, it may remain difficult to determine, particularly in younger patients, whether the headache primarily reflects CVT sequelae or the emergence of a primary headache disorder. Once secondary causes have been excluded, PCH may reflect either the emergence of a new headache disorder or the modification of a pre-existing primary headache. Although de novo headache appears to be the most common presentation, the relationship between prior headache and PCH remains insufficiently characterized. Some studies report limited similarity with previous headache patterns, while others suggest partial overlap. These discrepancies should be interpreted with caution. The reported prevalence of tension-type headache and migraine among patients with PCH appears comparable to that observed in the general population [23], and CVT predominantly affects young women, a population in which primary headache disorders – especially migraine – are already common [23]. In this context, the relatively modest proportion of patients with a documented pre-existing headache disorder may in fact be lower than expected, suggesting possible under-ascertainment. Indeed, pre-existing headache history was not systematically or prospectively assessed in most studies, raising the possibility of misclassification. The underlying mechanisms of PCH remain incompletely understood and are likely multifactorial. Several hypotheses have been proposed, including destabilization of a pre-existing primary headache disorder, unmasking of a previously subclinical headache susceptibility, and persistent nociceptive activation following the acute event [12]. In some patients, headache may be attributed to intracranial hypertension, potentially influenced by incomplete venous recanalization. The dural venous sinuses are innervated by trigeminovascular afferents, which constitute a plausible nociceptive substrate for headache after CVT [24]. Venous congestion and elevated intraluminal pressure may activate these meningeal afferents and promote central sensitization [25,26]. Intracranial pressure-related mechanisms may further contribute to migraine-like features through activation of the trigeminovascular system and release of calcitonin gene-related peptide (CGRP) from trigeminovascular terminals – a mechanism also implicated in migraine and idiopathic intracranial hypertension [24]. Such mechanisms may partly explain the clinical overlap between PCH and migraine phenotypes and could have implications for future targeted therapeutic strategies, including anti-CGRP therapies. Additionally, in dural arteriovenous fistula, altered venous flow and secondary intracranial pressure disturbances may contribute to headache persistence. The lateralization of PCH in relation to the site of venous thrombosis has not been investigated. Whether headache localization reflects local pathophysiological mechanisms, such as regional venous congestion or trigeminovascular activation, or is influenced by cognitive or emotional factors – such as patient awareness of thrombosis location – remains unknown. PCH also poses practical challenges, as it may lead to repeated emergency department visits, hospital admissions, and repeated investigations driven by concern for secondary causes. Careful and detailed clinical history-taking is therefore essential, especially as headache at CVT onset is often reported to differ from patients’ usual headache patterns [10]. Identification of associated or aggravating factors is also important, as these may contribute to headache persistence or severity. Several factors have been explored in relation to PCH, including pre-existing headache disorders, depression, recanalization status, thrombosis location, and anticoagulation delay. Recognition of such factors may help identify patients at higher risk of developing PCH and allow earlier intervention. Among these, depression appears particularly relevant, as it is common and potentially reversible. Systematic screening for comorbid depression and anxiety should therefore be considered an integral component of PCH management. Management strategies for PCH remain poorly defined, and evidence-based recommendations are lacking. The use of non-steroidal anti-inflammatory drugs is often limited due to long-term anticoagulation. In this context, greater consideration of preventive treatments aimed at reducing attack frequency may be appropriate. However, no study specifically evaluated preventive therapies in PCH, and available evidence remains extremely limited. In clinical practice, management may therefore reasonably be guided by the predominant headache phenotype; for example, headaches with migraine-like features could be treated using conventional anti-migraine therapies. Given the potential involvement of trigeminovascular and CGRP-mediated mechanisms in PCH pathophysiology, botulinum toxin or CGRP-targeted therapies could represent therapeutic options in some cases, although dedicated evidence is currently lacking. Screening for and managing comorbid depression and medication overuse should also be part of routine care, although their direct impact on PCH outcomes remains uncertain. Beyond symptomatic treatment, the question of optimizing CVT management to reduce the risk or severity of PCH may also be raised. Strategies aimed at improving venous recanalization, such as endovascular thrombolysis or thrombectomy, could theoretically influence headache outcomes, although available data are insufficient to support such an approach specifically for PCH prevention. Similarly, early initiation of anticoagulation may affect long-term headache persistence. Indeed, the association between anticoagulation delay and PCH may reflect the impact of anticoagulation on venous recanalization, which could represent a more direct determinant of headache persistence. The potential role of prolonged anticoagulant treatment may also be questioned; however, venous recanalization mainly occurs during the acute and subacute phases of CVT, and the impact of late or extended anticoagulation on long-term headache outcomes remains uncertain. Several limitations should be acknowledged. By restricting inclusion to studies with more than 30 patients, small case series (particularly those involving patients undergoing decompressive craniectomy or experiencing severe intracranial hypertension) may have been excluded. These populations may present distinct mechanisms, such as post-craniectomy headache. In addition, variability in headache assessment tools, follow-up duration, and outcome measures further complicates comparisons across studies. In conclusion, PCH represents a frequent yet under-recognized clinical issue that may significantly impact quality of life after CVT. Improved definitions, systematic and structured headache assessment, and standardized outcome measures are needed. Prospective studies are required to better delineate headache phenotypes, clarify underlying mechanisms, identify risk factors, support the development of diagnostic criteria, and develop targeted management strategies for patients with PCH. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements Not applicable. Author contributions IC and JM directed the project. DD and IC performed the systematic search. DD wrote the initial manuscript. IC, JM, CR and MM reviewed the manuscript. All authors read and approved the final manuscript. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Data availability All data used in this study are derived from previously published studies and are available in the cited references. No new datasets were generated. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All data used in this study are derived from previously published studies and are available in the cited references. No new datasets were generated.