Prevalence of Pelvic Vein Pathology in Patients with Cryptogenic Stroke and Patent Foramen Ovale Undergoing MRV Pelvis

UMMS Affiliation

Department of Neurology

Publication Date


Document Type



Adult; Aged; Embolism, Paradoxical; Female; Foramen Ovale, Patent; Humans; Male; Middle Aged; Prevalence; Retrospective Studies; Risk Factors; Stroke; Thrombophilia; Venous Thrombosis


Nervous System Diseases | Neurology


BACKGROUND: A substantial proportion of ischemic strokes has no any identified underlying cause. Notably, the prevalence of a patent foramen ovale (PFO) is increased in cryptogenic stroke (CS) populations, which may serve as a conduit for paradoxical emboli originating from deep vein thrombosis (DVT) including the pelvic veins. Yet, there are no published guidelines for the assessment of pelvic veins as part of the stroke workup and few studies have systematically investigated pelvic veins as a potential source for paradoxical emboli in CS patients. Further, there is a relative paucity of data regarding pelvic DVT in CS and results have been conflicting. Hence, we sought to determine the prevalence of pelvic DVT in select CS patients with PFO who underwent magnetic resonance venography (MRV).

METHODS: We retrospectively identified patients (n = 50) who underwent contrast-enhanced pelvic MRV at the discretion of the treating physician for the evaluation of CS in the presence of a PFO during hospitalization at a single academic stroke center between January 2011 through December 2013. Multivariable logistic regression analyses were used to assess for factors independently associated with the presence of an abnormal MRV pelvis.

RESULTS: Patients (47 +/- 13 years of age) had MRV performed 4 +/- 3 days after their incident stroke. Nine patients had an abnormal MRV (18%). Of these, four (8%) had pelvic vein thrombosis and 5 (10%) a May-Thurner anatomic variant. All patients with pelvic DVT were subsequently anticoagulated with warfarin (none had abnormal hypercoagulability testing). Clinical clues suggesting paradoxical embolism were present in as many as 40% of patients. On multivariable logistic regression, a history of any risk factors predisposing to DVT (OR 6.7; coefficient 1.9; BCa 95% CI 0.08-20.2; p = 0.014) as well as the number of predisposing risk factors (OR 3.9; coefficient 1.4; BCa 95% CI 0.25-4.2; p = 0.005) predicted the presence of pelvic vein pathology on MRV.

CONCLUSION: We demonstrate a relatively high prevalence of pelvic DVT among select CS patients emphasizing the importance of considering the pelvic veins as a potential source for emboli particularly in the presence of risk factors known to predispose DVT. Because patients were included at the treating physician's discretion, our results reflect 'real-life' practice. Our results may be of clinical importance as inclusion of pelvic vein imaging in CS patients with PFO had impactful therapeutic and nosologic implications. Further study is needed to define patients most likely to benefit from pelvic vein imaging.

DOI of Published Version



Cerebrovasc Dis. 2015;39(3-4):216-23. doi: 10.1159/000376613. Epub 2015 Mar 14. Link to article on publisher's site

Journal/Book/Conference Title

Cerebrovascular diseases (Basel, Switzerland)


Co-author Nils Henninger is a doctoral student in the Millennium PhD Program (MPP) in the Graduate School of Biomedical Sciences (GSBS) at UMass Medical School.

Related Resources

Link to Article in PubMed

PubMed ID