REVIEW ARTICLES

Cynthia K. Shortell, MD, SECTION EDITOR

A systematic review on internal jugular vein thrombosis

and pulmonary embolism
Panagiotis Drakos, MD, Benjamin C. Ford, MD, and Nicos Labropoulos, PhD, Stony Brook, NY

ABSTRACT
Objective: Whereas the internal jugular vein is the most common site of thrombosis in patients with deep venous
thrombosis (DVT) of the upper extremity, the association between internal jugular vein thrombus and pulmonary embolism
(PE) has not been clearly characterized. The objective of this paper was to determine the risk of embolization of an isolated
internal jugular vein thrombus causing a clinically overt PE, with the secondary objective of assessing the value of thera-
peutic anticoagulation in patients with isolated internal jugular vein thrombosis (IJVT) in improving clinical outcomes.
Methods: The National Center for Biotechnology Information, Cochrane Library, Web of Science, and Cumulative Index to
Nursing and Allied Health Literature were searched for articles. The relevant articles included were selected according to
the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Studies were included if they
specifically examined incidence of IJVT and incidence of PE and were excluded if they did not report on these rates
specifically or failed to specify the exact site of upper extremity DVT.
Results: Of the 274 articles screened, 25 were selected for full review following the Preferred Reporting Items for Systematic
Reviews and Meta-Analyses inclusion criteria. Seven of those provided adequate data and were included in the review.
There were only two studies demonstrating IJVT before PE that could probably establish causality, but this might be
confounded by the presence of concomitant upper extremity DVT in one of the cases and radiologic findings compatible
with resolving PE in another that might have preceded the presence of internal jugular vein thrombus. In the patients who
were found to have PE in the setting of IJVT, the overall observed mortality attributed to PE was low. In specific studies, the
use of anticoagulation did not reduce the mortality in those with isolated IJVT or affect the rate of thrombus resolution while
carrying the risk of bleeding complications in these patients, who often have severe comorbidities.
Conclusions: Despite the proximity of the jugular vein to the right side of the heart and the pulmonary vasculature, there
is little proof of propagation of the thrombus to cause a clinically overt PE. Whereas current practice is to treat the pa-
tients with IJVT in the same way as patients with lower extremity DVTs are treated, the lack of any survival benefit in those
with isolated IJVT and the risk of bleeding complications warrant further studies to characterize the need of medical
management in this population of patients. (J Vasc Surg: Venous and Lym Dis 2020;8:662-6.)
Keywords: Venous thrombosis; Upper extremity deep venous thrombosis; Jugular veins; Pulmonary embolism

Venous thrombosis, including deep venous thrombosis The prevalence of isolated IJVT varies in studies, with
(DVT) and pulmonary embolism (PE), is a common path- two studies showing a range of prevalence from 0.5%
ologic process with an annual incidence of approxi- to 3.47% in patients presenting with DVT.4,5 The most
mately 1 or 2 per 1000 adults.1-3 Whereas much is common clinical presentation is arm edema and pain,
known about the incidence, natural history, and treat- but it can also be completely asymptomatic.4
ment of lower extremity DVT (LEDVT), upper extremity IJVT has several recognized risk factors and causes. Central
DVT (UEDVT), specifically internal jugular vein thrombosis venouscatheter(CVC)use,whichisanimportantpartofman-
(IJVT), has not been analyzed as rigorously. agement in patients in both medical and surgical intensive
careunitsandduringsurgicalprocedures,isacommoninciting
From the Department of Surgery, Vascular and Endovascular Surgery Division, factor.DespitethatDVTisaknowncomplicationofCVCs,theinci-
Stony Brook University Hospital.
denceofCVC-relatedDVTisstillcontestedandvariesbysiteof
Author conflict of interest: none.
Correspondence: Nicos Labropoulos, PhD, Professor of Surgery and Radiology,
cannulationwitharangeof0%to67%basedononemeta-anal-
Director, Vascular Laboratory, Department of Surgery HSC T19-94, Stony Brook ysis.6 Other risk factors or causes of IJVT include radical neck
Medicine, 101 Nicolls Rd, Stony Brook, NY 11794-8191 (e-mail: nlabrop@yahoo. dissection,7,8ovarianhyperstimulationsyndrome,9malignant
com). neoplasms,andinfections,includingLemierresyndrome.Last,
The editors and reviewers of this article have no relevant financial relationships to
in cases with IJVT without central venous catheterization,
disclose per the Journal policy that requires reviewers to decline review of any
manuscript for which they may have a conflict of interest.
thrombusformationcouldbethefirstmanifestationofanoccult
2213-333X malignantneoplasm,whichmustberuledout.10-12
Copyright Ó 2020 by the Society for Vascular Surgery. Published by Elsevier Inc. Although the risk factors are recognized and there are
All rights reserved. data concerning UEDVT as a whole (encompassing
https://doi.org/10.1016/j.jvsv.2020.03.003

662
Journal of Vascular Surgery: Venous and Lymphatic Disorders Drakos et al 663
Volume 8, Number 4

subclavian, axillary, and internal jugular vein thromboses anticoagulation was given at the discretion of the
as one entity), there is not significant literature address- attending physician. Four studies provided data on PE
ing the incidence of isolated IJVT, the incidence of PE rate in patients with isolated IJVT, showing rates of
in patients with isolated IJVT, and the best management. 0.26%,17 0.5%,5 10.3%,4 and 14.3%.15 CTA or V/Q scan was
The aim of this review was to assess the incidence of PE performed only for symptomatic patients. None of the
as a complication of IJVT. studies provided complete data for PE sources from
other veins.
METHODS
Five databases were searched using search terms and DISCUSSION
keywords to identify abstracts and titles related to IJVT Concurrent diagnosis. The aim of this study was to
and PE. The National Center for Biotechnology Informa- evaluate the incidence of PE attributable to IJVT. Several
tion, Cochrane Library, Web of Science, and Cumulative In- of the selected studies described patients who had
dex to Nursing and Allied Health Literature were searched radiographically diagnosed PE and were then found to
using the main terms “internal jugular vein,” “thrombosis,” have an IJVT while a source was being pursued. Studies
and “pulmonary embolism.” Each term was expanded to dependent on these cases for their data introduce bias
include synonyms and subsets. The Boolean terms “and” that is likely to increase the demonstrated risk of PE
and “or” were used to combine terms. from IJVT by selecting for patients who already have
the desired end point.
Inclusion and exclusion criteria. Studies were included In our review, several of the studies that ultimately were
if they examined the specific incidence of IJVT and the selected included groups of patients in whom a diag-
incidence of PE. Articles were excluded if they did not nosis of PE was found concurrently with the diagnosis
specifically include the incidence of PE associated with of IJVT. Only the two prospective studies directly speci-
IJVT or failed to specify site of UEDVT. fied that IJVT was found before PE,14,17 thus demon-
strating a possible temporal relationship. One of these
RESULTS
studies also included patients who were found to have
There were 274 articles identified in the initial search. These
PE before diagnosis of IJVT but separated those patients
articles were screened by title and abstract for inclusion and
from the rest of the population in the results,17 although
exclusion criteria. Of the initial 274, there were 25 articles
a study of the Computerized Registry of Patients with
selected for full review. An additional five articles were found
Venous Thromboembolism (RIETE) did show that pa-
by cross-referencing and were included. Of these 30 articles
tients with DVT in the upper extremity were less likely
selected for full-text review, 7 ultimately provided adequate
to present with PE at diagnosis (10%) compared with
data and were included in the review (Fig).
those with LEDVT.18
Of the seven studies selected, five were retrospective
cohort studies and two were prospective cohort studies, Concomitant thrombosis. Another aim of this study
all providing low levels of evidence (Table). There was was to assess the incidence of PE specifically associated
heterogeneity in the study groups, with one paper study- with isolated IJVT. Many previous studies on UEDVT
ing only patients with pre-existing malignant disease, have tended to group axillary, subclavian, and internal ju-
one studying patients in intensive cardiac rehabilitation gular vein thromboses into the umbrella term UEDVT
after cardiac surgery, one studying patients admitted to and based statistics on this larger category. Concomitant
a surgical intensive care unit, and four studying all thrombosis of the axillary vein or the subclavian vein in-
patients sent for DVT studies at their respective institu- troduces a second possible source of embolization.
tions. DVT was detected with duplex ultrasound in all Therefore, determining the theoretical risk of PE from
studies, and PE was detected by either ventilation- an isolated IJVT is confounded by the presence of
perfusion (V/Q) scan or computed tomography angiog- concomitant thrombosis in other veins.
raphy (CTA; Table). Three of the studies were designed Of the selected studies, six studies specified whether
to study UEDVT as a whole entity but gave information the IJVT was isolated or occurred concomitantly with
on isolated IJVT as a covariate,13-15 whereas the other subclavian or axillary vein thrombosis (Table).4,5,13,15,16
four studies were directly interested in isolated IJVT. The observed relationship of isolated IJVT to PE in com-
Only five of the studies, including one of the prospective parison to concomitant thrombosis and PE is mixed;
studies, were designed with PE as one of the primary two studies demonstrated an increased incidence of
outcomes studied.4,5,13,16,17 Neither of the prospective PE in patients with isolated IJVT,13,15 whereas two others
studies randomized which patients were given anticoa- did not demonstrate this increased risk.4,5
gulation. Frizzelli et al17 gave patients anticoagulation if
the IJVT met certain imaging criteria (thrombus Anticoagulation therapy. There is currently no defini-
>2 cm long or occupying >50% of lumen), whereas tive evidence-based treatment guideline for UEDVT. In
Malinoski et al14 gave anticoagulation at the discretion the absence of specific studies, there is a 2C recom-
of the attending physician. In all retrospective studies, mendation to consider anticoagulation.19
664 Drakos et al Journal of Vascular Surgery: Venous and Lymphatic Disorders
July 2020

Fig. Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram for the systematic review
of internal jugular vein thrombosis (IJVT) and pulmonary embolism (PE).

The risk of PE from UEDVT differs in various studies. This studies were patients randomized to anticoagulation or
undetermined risk makes a cost-benefit analysis of treat- not; one study treated all patients with a “high-risk”
ment murky; but given the known benefits of anticoagu- IJVT,17 and the other treated patients at the discretion
lation in preventing PE in patients with LEDVT, the of the attending physician. Last, one of the reviewed
estimated risk of PE in patients with UEDVT, and evi- studies showed that administration of anticoagulation
dence that the risk of recurrent venous thromboembo- did not achieve statistical significance to reduce overall
lism and PE in patients with UEDVT is not significantly mortality in a group of patients with isolated IJVT,
different compared with patients with LEDVT,18 most cli- although this can be confounded by the fact that pa-
nicians empirically treat with anticoagulation. Current tients who are given anticoagulation are generally
guidelines suggest the same initial and long-term anti- healthier, and improved survival cannot necessarily be
coagulation therapy as would be used in patients with attributed to protection from PE.
LEDVT.18,20 Furthermore, although patients with UEDVT The data from the reviewed articles is insufficient to
less frequently have signs of clinically overt PE at time make a recommendation supporting or refuting the
of presentation, their 3-month clinical outcomes are use of anticoagulation in patients with isolated IJVT.
similar, and the rate of recurrence in patients with The risks of anticoagulation should be weighed because
catheter-related UEDVT is higher, thus suggesting that most patients with UEDVT would recover without
appropriate therapy should not differ.18,20 sequelae, although a small percentage could have
In all reviewed articles, therapeutic anticoagulation was disabling long-term symptoms,21 and PE was not listed
achieved with heparin, low-molecular-weight heparin, or as the primary cause of death in these patients.16 In addi-
heparin bridge to warfarin. In studies focusing on CVC- tion, anticoagulation treatment carries the risk of intra-
related IJVT, CVCs were also frequently removed at the cranial and gastrointestinal bleed with documented
discretion of the attending physician, with only removal deaths in the treatment arm of studies.22 However, Man-
of the catheter significantly affecting complete resolu- sour et al13 suggested that patients with non-catheter-
tion of IJVT in one study.5 In neither of the prospective related DVTs diagnosed with cancer are more prone to
Journal of Vascular Surgery: Venous and Lymphatic Disorders Drakos et al 665
Volume 8, Number 4

Table. Study type, pulmonary embolism (PE) rate, and method of detection
Study Study type Isolated IJVT Concomitant IJVT PE rate in isolated IJVT, % (No.) PE detection method
Mansour et al13 Retrospective 22 38 22.7 (9) Not specified
17
Frizzelli et al Prospective 386 e 1.6 (6) CTA
Gbaguidi et al4 Retrospective 29 e 10.4 (3) CTA
Malinoski et al14 Prospective e 64 e CTA
Sheikh et al16 Retrospective 74 142 2.7 (2) V/Q scan
Ascher et al5 Retrospective 21 61 Confounded by LEDVT V/Q scan
15
Major et al Retrospective 28 e 14.3 (4) CTA and V/Q scan
CTA, Computed tomography angiography; IJVT, internal jugular vein thrombosis; LEDVT, lower extremity deep venous thrombosis; V/Q, ventilation-
perfusion.

complications and therefore may benefit from more be significant clot burden to cause a clinically overt PE,
aggressive treatment compared with those with throm- which may explain the lack of documented mortality
bosis in other sites of UEDVT. from internal jugular vein thrombi.

Surveillance method. The “gold standard” for the diag- CONCLUSIONS

nosis of PE is CTA. Despite this, many centers and older In patients with UEDVT, IJVT is the most common,
studies use V/Q scans as the main tool for diagnosis of especially in those with CVCs and in the presence of
PE. Of the selected articles, two used V/Q scanning active malignant disease. The risk of PE is low, and
only, one used a mix of V/Q and CTA scans, and one studies have failed to establish a clear-cut temporal rela-
did not specify (Table). Use of a V/Q scan in the diag- tionship between IJVT and subsequent development of
nosis of PE in these studies can again allow bias because PE. Current practice is to treat patients with isolated
of the difference in sensitivity and specificity between IJVT with systemic anticoagulation. Our systematic re-
CTA and V/Q scans. view shows that the data are inadequate to form an
In the diagnosis of UEDVT, Doppler ultrasound has evidence-based recommendation on the use of anticoa-
been reported to range from 78% to 100% in sensitivity gulation in patients with isolated IJVT and suggests that
and from 82% to 100% in specificity. The internal jugular further studies are required to establish the role of anti-
vein has been the most observed site.23 coagulation in this cohort.

PE-associated mortality. With venous thrombosis, the

AUTHOR CONTRIBUTIONS
greatest fear is of PE, but the severity of PE is not the
Conception and design: PD, BF, NL
same in all patients. Whereas large, saddle emboli can
Analysis and interpretation: PD, BF, NL
cause rapid hemodynamic collapse, smaller subsegmen-
Data collection: PD, BF
tal emboli can cause only minor clinical symptoms. In
Writing the article: PD, BF, NL
the reviewed studies, only one death due to PE was re-
Critical revision of the article: PD, BF, NL
ported, in a patient with lung cancer, of the total 24 cases
Final approval of the article: PD, BF, NL
of PE described. Despite the variation in incidence of PE
Statistical analysis: Not applicable
associated with IJVT in these studies, the overall
Obtained funding: Not applicable
observed mortality was low. Specifically, in non-
Overall responsibility: NL
catheter-associated UEDVTs as shown from the analysis
of RIETE,24 although PE was a rare occasion, it was fatal in
22% of cases. REFERENCES
On the other hand, regarding the mortality associ- 1. Naess IA, Christiansen SC, Romundstad P, Cannegieter SC,
ated with UEDVT including IJVT, many authors are Rosendaal FR, Hammerstrom J. Incidence and mortality of
considering it a marker of severity of illness rather venous thrombosis: a population-based study. J Thromb
Haemost 2007;5:692-9.
than a possible source of thromboembolism that 2. Delluc A, Tromeur C, Le Ven F, Gouillou M, Paleiron N,
could lead to a fatal PE with documented increased Bressollette L, et al. Current incidence of venous throm-
Acute Physiology and Chronic Health Evaluation III boembolism and comparison with 1998: a community-
scores of patients with UEDVT compared with based study in Western France. Thromb Haemost
LEDVT.25 2016;116:967-74.
3. Tagalakis V, Patenaude V, Kahn SR, Suissa S. Incidence of
It is our hypothesis that in isolated IJVT, the cross flow and mortality from venous thromboembolism in a real-
from the ipsilateral subclavian vein will prevent signifi- world population: the Q-VTE Study Cohort. Am J Med
cant growth of an isolated IJVT. Therefore, there will not 2013;126:832.e13-21.
666 Drakos et al Journal of Vascular Surgery: Venous and Lymphatic Disorders
July 2020

4. Gbaguidi X, Janvresse A, Benichou J, Cailleux N, Levesque H, 17. Frizzelli R, Tortelli O, Di Comite V, Ghirardi R, Pinzi C,
Marie I. Internal jugular vein thrombosis: outcome and risk Scarduelli C. Deep venous thrombosis of the neck and pul-
factors. QJM 2011;104:209-19. monary embolism in patients with a central venous catheter
5. Ascher E, Salles-Cunha S, Hingorani A. Morbidity and mor- admitted to cardiac rehabilitation after cardiac surgery: a
tality associated with internal jugular vein thromboses. Vasc prospective study of 815 patients. Intern Emerg Med 2008;3:
Endovascular Surg 2005;39:335-9. 325-30.
6. Rooden CJ, Tesselaar ME, Osanto S, Rosendaal FR, 18. Munoz FJ, Mismetti P, Poggio R, Valle R, Barron M, Guil M,
Huisman MV. Deep vein thrombosis associated with central et al. Clinical outcome of patients with upper-extremity
venous cathetersda review. J Thromb Haemost 2005;3: deep vein thrombosis: results from the RIETE Registry.
2409-19. Chest 2008;133:143-8.
7. Quraishi HA, Wax MK, Granke K, Rodman SM. Internal ju- 19. Tait C, Baglin T, Watson H, Laffan M, Makris M, Perry D,
gular vein thrombosis after functional and selective neck et al. British Committee for Standards in Haematology.
dissection. Arch Otolaryngol Head Neck Surg 1997;123: Guidelines on the investigation and management of
969-73. venous thrombosis at unusual sites. Br J Haematol
8. Harada H, Omura K, Takeuchi Y. Patency and caliber of the 2012;159:28-38.
internal jugular vein after neck dissection. Auris Nasus Lar- 20. Cote LP, Greenberg S, Caprini JA, Tafur A, Choi C, Munoz FJ,
ynx 2003;30:269-72. et al. Comparisons between upper and lower extremity
9. Lee SH, Park JW, Han M, Bae JW, Cho YJ. Internal jugular vein deep vein thrombosis: a review of the RIETE registry. Clin
thrombosis with OHSS. J Clin Ultrasound 2017;45:450-2. Appl Thromb Hemost 2017;23:748-54.
10. Erkoc R, Uzun K, Yuca K, Etlik O, Dogan A. Internal jugular 21. Joffe HV, Kucher N, Tapson VF, Goldhaber SZ; Deep Vein
vein thrombosis two different etiologies. Eur J Gen Med Thrombosis (DVT) FREE Steering Committee. Upper-ex-
2005;2:123-8. tremity deep vein thrombosis: a prospective registry of 592
11. Carrington BM, Adams JE. Jugular vein thrombosis associ- patients. Circulation 2004;110:1605-11.
ated with distant malignancy. Postgrad Med J 1988;64:455-8. 22. Levy MM, Albuquerque F, Pfeifer JD. Low incidence of
12. Boedeker CC, Ridder GJ, Weerda N, Maier W, Klenzner T, pulmonary embolism associated with upper-extremity
Schipper J. [Etiology and therapy of the internal jugular vein deep venous thrombosis. Ann Vasc Surg 2012;26:
thrombosis]. Laryngorhinootologie 2004;83:743-9. 964-72.
13. Mansour A, Saadeh SS, Abdel-Razeq N, Khozouz O, 23. Blaivas M, Stefanidis K, Nanas S, Poularas J, Wachtel M,
Abunasser M, Taqash A. Clinical course and complications of Cohen R, et al. Sonographic and clinical features of upper
catheter and non-catheter-related upper extremity deep extremity deep venous thrombosis in critical care patients.
vein thrombosis in patients with cancer. Clin Appl Thromb Crit Care Res Pract 2012;2012:489135.
Hemost 2018;24:1234-40. 24. Newton DH, Monreal Bosch M, Amendola M, Wolfe L, Perez
14. Malinoski DJ, Ewing T, Patel MS, Nguyen D, Le T, Cui E, et al. Ductor C, Lecumberri R, et al. Analysis of noncatheter-
The natural history of upper extremity deep venous throm- associated upper extremity deep venous thrombosis from
boses in critically ill surgical and trauma patients: what is the the RIETE registry. J Vasc Surg Venous Lymphat Disord
role of anticoagulation? J Trauma 2011;71:316-21; discussion: 2017;5:18-24.e1.
321-2. 25. Hingorani A, Ascher E, Markevich N, Yorkovich W, Schutzer R,
15. Major KM, Bulic S, Rowe VL, Patel K, Weaver FA. Internal ju- Mutyala M, et al. Risk factors for mortality in patients with
gular, subclavian, and axillary deep venous thrombosis and upper extremity and internal jugular deep venous throm-
the risk of pulmonary embolism. Vascular 2008;16:73-9. bosis. J Vasc Surg 2005;41:476-8.
16. Sheikh MA, Topoulos AP, Deitcher SR. Isolated internal ju-
gular vein thrombosis: risk factors and natural history. Vasc
Med 2002;7:177-9. Submitted Oct 30, 2019; accepted Mar 9, 2020.