Patient Safety/Quality Improvement

Otolaryngology–
Head and Neck Surgery

Endovascular Embolization in 1–7

Ó American Academy of
Otolaryngology–Head and Neck
the Treatment of Epistaxis Surgery Foundation 2019
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0194599819829743
http://otojournal.org
Phillip Huyett, MD1, Brian T. Jankowitz, MD2, Eric W. Wang, MD1,
and Carl H. Snyderman, MD, MBA1

Sponsorships or competing interests that may be relevant to content are dis- Received October 24, 2018; revised December 19, 2018; accepted
closed at the end of this article. January 18, 2019.

Abstract

O
ver the course of a lifetime, over half the population
Objective. The cost-effectiveness of endovascular embolization will experience an episode of epistaxis.1 The vast
(EE) for intractable epistaxis has been questioned, especially majority of these (~90%) resolve spontaneously or
as endoscopic surgical techniques have become standard of with modest measures such as gentle anterior nasal pressure,
care at many institutions. Our objectives were to review the light packing, and minimal cauterization. Severe, intract-
safety profile and effectiveness of EE for epistaxis. able, or posterior epistaxis occurs in approximately 10% of
Study Design. Retrospective case series. patients and, by definition, does not respond to minimal
intervention.
Setting. Tertiary care hospital. In these instances, complete or anterior-posterior packing
has historically been performed but may not provide suffi-
Subjects. There were 54 patients and 64 unique encounters. cient tamponade or continued hemostasis upon removal
Patients were 66.7% male, with a mean age of 64.5 years. of the packing in a significant number of patients.2
Bleeding disorders were present in 18.8%, hypertension was Hospitalization costs, patient discomfort and risks of postob-
present in 71.7%, and 61.1% were on anticoagulant/platelet structive sinusitis, arrhythmias related to vagal stimulation,
drugs. and posterior displacement of the pack into the airway make
Methods. Charts of patients undergoing EE for epistaxis this option less desirable.3,4 There is the additional cost of
between 2005 and 2015 were retrospectively reviewed. retreatment if epistaxis recurs following removal of packing.
Advances in endoscopic surgery and transarterial endo-
Results. The immediate bleeding control rate was 92.6%. vascular procedures have occurred in parallel since the late
Three patients died within 1 week of EE and were excluded 1980s. Accordingly, both endovascular embolization (EE)
from further analysis. Overall, 64.7% of the remaining of branches of the external carotid artery and endoscopic
patients had no further episodes of epistaxis. Thirteen sphenopalatine artery ligation (ESPAL) have become
patients (25.4%) rebled within 1 week, 11 of whom required increasingly popular treatment options for intractable or
repeat EE or operative control. Five patients (9.8%) rebled recurrent posterior epistaxis.5,6 Numerous studies demon-
more than 1 week following the procedure with 4 requiring strate similar efficacy with initial hemostasis rates over 90%
repeat EE or operative control. The major complication rate using either technique.3,6-27
was 7.4% and included transient stroke, diplopia, facial skin Beyond comparisons of efficacy, multiple studies demon-
necrosis, and extraperitoneal hemorrhage. strate the superior cost-effectiveness of early ESPAL over
Conclusion. While the immediate success rate of EE for epis- posterior nasal packing with admission,28-31 as well as
taxis was comparable to the literature, the overall short- ESPAL over EE.32,33 A recent report by Leung and
and long-term rebleed rate was high in this selected popula-
tion. The results suggest that patients who are referred for 1
Department of Otolaryngology, University of Pittsburgh Medical Center,
EE represent a high-risk group with increased risk of repeat Pittsburgh, Pennsylvania, USA
hemorrhage and morbidity. Patients who undergo EE for 2
Department of Neurological Surgery, University of Pittsburgh Medical
epistaxis should be carefully monitored for complications, Center, Pittsburgh, Pennsylvania, USA
including repeat hemorrhage. This article was presented at the AAO-HNSF 2016 Annual Meeting and
OTO EXPO; September 18-21, 2016; San Diego, California.

Keywords Corresponding Author:

Carl Snyderman, MD, MBA, UPMC Department of Otolaryngology, 200
epistaxis, endovascular, embolization, sphenopalatine artery Lothrop Street, EEI Suite 500, Pittsburgh, PA 15213, USA.
ligation Email: [email protected]
2 Otolaryngology–Head and Neck Surgery

coauthors34 analyzed the risks of ESPAL, embolization, and Indiana]; embosphere particles [Merit Medical, South
posterior packing and recommended ESPAL as first-line Jordan, Utah]; Onyx [ev3, Irvine, California]) and angio-
treatment for severe epistaxis. The objective of this study graphic findings (none, hypervascular blush, arteriovenous
was to review the past 10-year experience of EE for intract- malformation, pseudoaneurysm) were extracted from the
able epistaxis at a single institution, where both ESPAL and procedure transcriptions. Number of vessels treated was a
EE were adopted early and are frequently used. sum of the individual external carotid artery branches embo-
lized. Complications were categorized as major and minor
Methods as well as catheter related, local (facial, palatal), neurologic,
The institutional review board at the University of and ophthalmologic. Facial pain was included only if it
Pittsburgh approved this retrospective review of cases of remained present at the follow-up appointment so as to
epistaxis treated with EE from January 1, 2005, to January exclude pain related to nasal packing or manipulation.
1, 2015. Patients were identified within a database main- Follow-up was only interrogated within the UPMC hospital
tained by the interventional neuroradiology department at system, which has a large catchment area in the western
the University of Pittsburgh Medical Center (UPMC). Pennsylvania region.
Electronic medical records were reviewed to ensure a diag- SPSS version 23 (SPSS, Inc, an IBM Company,
nosis of epistaxis and treatment with EE. Exclusion criteria Chicago, Illinois) was used to perform statistical analyses,
included patients treated prophylactically or therapeutically and a P value less than .05 was considered significant.
for benign or malignant neoplasms of the head and neck, Categorical variables were compared using Pearson’s x2
those having previously undergone EE for epistaxis, and test, and numerical variables were compared using the
those with inadequate data in the medical record. All independent-samples t test. Fisher’s exact test was used to
patients were evaluated by the otolaryngology service. calculate the correlation between variables and local compli-
Selection of treatment was based on individual preference cations given the small sample size (n = 10).
of admitting/consulting surgeons but was influenced by a
departmental preference for ESPAL in suitable surgical Results
candidates. Over the 10-year study period, 54 patients were treated with
A patient was considered to have a bleeding disorder if EE across 64 unique encounters. Only the first encounter
he or she carried a diagnosis of an acquired or inherited, was included for analysis. The patients were an average of
nonpharmacologically induced proclivity toward nasal 64.5 years old (range, 23-97 years). Two-thirds of the
bleeding, including end-stage liver and kidney disease. A patients were male (66.7%, n = 36). There were 33 nonsmo-
diagnosis of hypertension was only included if the patient kers (61.1%) and 21 smokers (38.9%). The etiology of
had a preexisting diagnosis documented in the medical bleeding was postsurgical in 4 patients, traumatic in 2
record. Blood pressure measurements in the emergency patients, and idiopathic in the remaining 48 patients.
department were considered unreliable given the sympa- Bleeding diatheses were present in 18.5% and were
thetic response to active severe bleeding and invasive proce- related to end-stage liver disease (n = 6) and end-stage renal
dures. Furthermore, medication was often given to control disease (n = 4). Hypertension was documented as a preexist-
blood pressure in the emergency room or as an inpatient. A ing diagnosis in 70.4% (n = 38), and 61.1% (n = 33) of
patient was considered to be on an anticoagulant/platelet patients were on anticoagulation medications. The mean
drug if he or she was on daily doses of aspirin, warfarin, hematocrit, INR, PTT, and platelet counts were 29.3%
clopidogrel, or similar agents. Laboratory measurements (range, 14.4%-41.3%), 1.3 (0.9-3.8), 34.7 seconds (22.8-
(hematocrit, international normalized ratio [INR], partial 107.2 seconds), and 202/mL (36-389/mL) respectively.
thromboplastin time [PTT], and platelet count) reflect test- In 10 patients (18.5%), the encounter represented their
ing immediately prior to EE. first episode of epistaxis. All but 7 patients (87.0%) were
The immediate success of EE was determined by the treated with packing initially, and in 9 patients (16.7%),
achievement of adequate hemostasis at the completion of the attempts at nasal cauterization were made. Five patients had
procedure. We considered a rebleed within 24 hours to be an undergone ESPAL and 2 patients had undergone ESPAL
immediate failure of EE. Short- and long-term failures were and anterior ethmoid artery (AEA) ligation prior to EE. In
defined as repeat episodes of epistaxis within 1week and all but 1 of these cases, ESPAL or AEA ligation was within
beyond 1 week, respectively. Minor failures were those that 1 week of the EE.
were treated conservatively with pressure, decongestants, All patients underwent successful angiography of the
minimal cauterization, or light packing. Major failures were bilateral common carotid arteries with 1 of 5 treating
those that required revision embolization or surgical treat- neurointerventionalists. Six patients (11.1%) were found to
ment. Patients who died within 3 days of the procedure were have a pseudoaneurysm or arteriovenous malformation, 12
excluded from analyses related to success rates. (22.2%) had a hypervascular blush, and the remaining
All patients underwent angiography of the bilateral 36 (66.7%) had no abnormal angiographic findings. All
common carotid arteries. The individual arteries embolized positive findings were related to branches of the external
(facial, internal maxillary), materials used (Tornado endo- carotid arteries only. Forty-eight patients (88.9%) were
vascular embolization coils [Cook Medical, Bloomington, treated with bilateral internal maxillary artery (IMA)
Huyett et al 3

54 paents

4 immediate re- 50 immediate

bleeds success

3 endoscopically
managed 10 short-term re- 2 died within 3 days 38 short-term
bleeds (unrelated to EE) success
(incl. 3 AEA lig)

9 endoscopically
1 died within 3 days 5 long-term
managed 33 long-term success
(unrelated to EE) re-bleeds
(incl. 3 AEA lig)

1 minor re-bleed
1 repeat EE (treated
conservavely)

1 with mulple
connued re-bleeds 3 repeat EE
despite EE and ESPAL

2 minor re-bleeds
1 endoscopically
(treated
managed
conservavely)

Figure 1. Outcomes of patients who underwent endovascular embolization. AEA, anterior ethmoid artery; EE, endovascular embolization.

embolization; the remaining patients had unilateral IMA In the 7 patients with a history of ESPAL or AEA liga-
embolization. Twenty-five patients also underwent facial tion for epistaxis, EE had long-term success in 3 patients.
artery embolization: 6 bilateral and 18 unilateral. Only 6 Of the remaining 4 patients, 1 died within 3 days and 3 had
patients had exclusively unilateral treatment. Tornado endo- major short-term failures that ultimately responded to AEA
vascular embolization coils were used in 44 patients, embo- ligation. EE was therefore efficacious in 50% (3/6) of surgi-
sphere particles in 41 patients, and Onyx in 12 patients. cal failures with the remaining 3 requiring treatment of ves-
The immediate control rate of epistaxis was 92.6%. sels not accessed by EE.
Patient outcomes are depicted in Figure 1. Three patients In the 15 patients who failed EE and required repeat EE
died within 3 days of the procedure due to withdrawal of or surgery, 10 subsequently achieved surgical control, 4
care in a patient with terminal esophageal cancer, pulmonary underwent successful repeat EE, and 1 patient continued to
embolism, and cardiac arrest, all of which occurred following have recurrent severe epistaxis despite both EE and endo-
completion of EE. Overall, 64.7% of the remaining patients scopic management. Of the 10 successful surgical cases fol-
had no known further episodes of epistaxis. Thirteen (24.1%) lowing failed EE, 9 were performed within 1 week and 6
patients rebled within 1 week, 11 of whom required repeat involved AEA ligation. Three of the 4 long-term failures
embolization or operative control. Five had recurrent epis- were successfully treated with repeat EE.
taxis greater than 1 week following the procedure, with 4 The median patient follow-up was 601 days (range, 0-3705
requiring repeat embolization or operative control. days). The overall complication rate of EE was 20.4%
As shown in Table 1, no demographic or treatment vari- (14 complications in 11 patients). As shown in Table 2,
ables predicted treatment failures (short or long term) with the complications included transient stroke, diplopia, palatal
the exception of patients who were on anticoagulation medi- ulcers, facial pain, facial skin necrosis, and extraperitoneal
cations (odds ratio [OR], 5.31; 95% confidence interval hemorrhage. The major complication rate was 7.4% (facial skin
[CI], 1.29-21.9; P = .015). Patients who had 3 or more ves- necrosis, vision change, stroke, and extraperitoneal hemorrhage).
sels embolized had similar failure (39.1% vs 32.1%, P = No patient or treatment variable correlated with increased risk
.603) and total complication rates (16.0% vs 24.1%, P = of local complications (facial/palatal ulcers, pain, numbness,
.517) compared to patients who had 1 to 2 vessels embo- necrosis) except age greater than 60 years, which was found to
lized. No laboratory value corresponded with increased risk be protective (Table 3).
of failure. When comparing the first 27 cases performed
with the most recent 27 cases, there was no improvement in Discussion
long-term hemostasis (56.0% vs 73.1%, P = .202) or major This study finds that EE is effective at controlling severe
failures (40.0% vs 19.2%, P = .102). epistaxis within the first 24 hours following treatment. The
4 Otolaryngology–Head and Neck Surgery

Table 1. Relationship of Demographic and Treatment Characteristics to Treatment Failures (Short and Long Term).a
Characteristic n OR 95% CI P Value

Age .60 y 30 1.15 0.36-3.74 .806

Male 33 1.14 0.34-3.83 .829
Smoker 21 0.47 0.14-1.64 .233
Hypertension 37 1.37 0.35-5.30 .648
Anticoagulant/platelet drug 31 5.31 1.29-21.9 .015
Bleeding disorder 10 0.39 0.07-2.08 .259
Positive angiography finding 17 1.00 0.30-3.38 1.000
IMA bilateral 45 3.04 0.33-28.2 .309
Facial artery 22 0.76 0.24-2.46 .651
Coils 41 0.32 0.11-1.89 .278
Particles 38 0.54 0.15-1.95 .343
Onyx 12 1.43 0.38-5.38 .597
Abbreviations: CI, confidence interval; IMA, internal maxillary artery; OR, odds ratio.
a
Patients who died within 3 days are excluded.

major treatment for recurrent severe epistaxis following EE

Table 2. Complications following Endovascular Embolization for
Epistaxis.a (29.4%). For comparison, reported failure rates (major and
minor) beyond 24 hours following ESPAL range from 13%
Complication Number Percentage of Patients to 22% and 7% to 9% at our institution.25,31,35,36 In this
study, we found no factors that correlated with an increased
Neurologic
risk of failure besides the use of anticoagulation medications
Transient strokeb 1 1.85
(Table 1). Without a randomized trial comparing EE to
Ophthalmologic
ESPAL, it is not possible to determine if the lower rate of
Vision change (diplopia)b 1 1.85
failure with ESPAL is due to selection bias or greater
Catheter site
efficacy.
Retroperitoneal hemorrhageb 1 1.85
Potential explanations for recurrent epistaxis following
Groin hematoma 1 1.85
EE include failure to completely embolize the targeted ves-
Failed catheterization 0 0.00
sels, failure to embolize the correct arterial territory, or
Local
bleeding from a new site. Subsequent angiograms often
Palatal ulcers 3 5.56
reveal sustained occlusion of embolized vessels, arguing
Nasal/facial pain 4 7.40
that recanalization of the embolized vessel is uncommon.
Nasal/facial/palatal numbness 2 3.70
The identification of the correct vascular territory during EE
Facial skin necrosisb 1 1.85
is challenging given the rarity of positive findings on angio-
a
There were 14 complications reported in 11 patients. graphy (66.7% had a normal angiogram), most often result-
b
Denotes major complications. ing in unnecessarily wide-field treatment with embolization
of vessels bilaterally (88.8% were treated bilaterally).
Interestingly, treating an increased number of vascular terri-
immediate success rate of 92.6% found within this cohort is tories increased neither the success rate nor the complication
comparable to numerous prior studies. In patients with rate. The high prevalence of anticoagulant/platelet medica-
severe or life-threatening epistaxis, the immediate control tions and association with failure in the EE group might
rate is what is reported most often in the literature. indicate bleeding from multiple different sites.
However, given reports in the literature of an equivalent EE is not without significant potential risk (Table 2).
immediate success rate using ESPAL, we were interested in Prior studies have noted lower complication rates (2%-17%)
determining the longer-term rebleed and complication rate compared to our overall complication rate (20.4%), which
with EE. may be a reflection of the lack of long-term follow-up, dif-
The overall rebleed rate beyond 24 hours was 35.3% (18 ferential definitions of complications, and variations in tech-
patients), a high figure that is not readily reported in the lit- nique. The major complication rate of 7.4% is acceptable,
erature. Most of these patients (n = 13) experienced a failure but the severity of these potential complications should
within 1 week, and all but 2 required repeat EE or ESPAL. prompt caution in decision making. As shown in Table 3,
Similarly, 4 of 5 patients who experienced recurrent epis- no treatment variables, such as bilateral treatment or embo-
taxis more than 1 week after EE required repeat EE or lization of the IMA and facial arteries, resulted in higher
ESPAL. This equates to a nearly 1 in 3 chance of requiring rates of local complications. As mentioned, the experience
Huyett et al 5

Table 3. Relationship of Patient and Treatment Variables to All with traumatic facial injuries and epistaxis, angiography has
Local Complications (10 Complications, 7 Patients). both diagnostic and therapeutic benefits. EE is also pre-
Characteristic n OR 95% CI P Value ferred for bleeding secondary to head and neck and skull
base tumors since the contributing vessels may not be surgi-
Age .60 y 33 0.71 0.19-2.71 .617 cally accessible.
Male 36 0.32 0.08-1.26 .094 Endoscopic surgery and EE have complementary roles in
Smoker 21 0.48 0.11-2.12 .478 the management of epistaxis. All but 1 patient had lasting
Hypertension 38 0.62 0.15-2.54 .708 hemostasis with either EE or EE and ESPAL. Although lim-
Anticoagulant/platelet drug 33 1.92 0.45-8.26 .497 ited by relatively small numbers, EE was successful in treat-
Bleeding disorder 11 0.33 0.04-2.90 .426 ing half of surgical failures (3/6). The remaining 3 patients
Positive angiography finding 18 0.70 0.16-3.04 .733 who failed EE after surgical failure required treatment of
Facial artery 24 0.66 0.17-2.58 .736 vessels not accessible by EE (AEA). Similarly, AEA liga-
.2 vessels embolized 25 0.60 0.15-2.35 .517 tion was needed to achieve hemostasis in 6 of the 13 pri-
Coils 44 1.03 0.19-5.71 1.000 mary short-term EE failures. In the long-term primary EE
Particles 41 0.81 0.18-3.64 1.000 failures, all 3 patients who underwent repeat EE achieved
Onyx 12 1.42 0.31-6.46 .693 long-term hemostasis. Ultimately, the choice of surgery vs
EE depends on multiple factors, including the availability
Abbreviations: CI, confidence interval; OR, odds ratio.
and experience of rhinologic surgeons and interventional
radiologists.
of the interventional radiologist may be an important factor. This study is limited by its retrospective design and, in
Although there was no statistical difference in success or particular, the heterogeneous indications for EE. There was
complication rate between the first and second half of the an inherent selection bias in the choice of therapy due to
cases, the study was not designed to power such an analysis. choices of individual surgeons and an evolving clinical care
EE is often chosen in patients with medical comorbidities pathway. Presumably, patients who were selected for EE
that are considered high risk for surgery. This may introduce had more significant comorbidities and risks for bleeding.
a treatment bias that predisposes to complications. Although the experience reported in this study suggests that
Our institution recently instituted an epistaxis clinical endoscopic surgery is preferred due to better short-term and
care pathway that emphasizes the preferential use of endo- long-term efficacy and favorable safety profile, randomized
scopic surgery (ESPAL 6 AEA ligation) in all cases of trials are necessary to truly compare long-term efficacy
severe epistaxis.33 This was based on a number of factors. and identify subpopulations of patients who are better
First, we found early ESPAL to be cost-effective when com- treated with EE. Until these populations are identified, we
pared to traditional packing with increased duration of hos- advocate that endoscopic surgery remain the treatment of
pitalization, cardiac monitoring, and higher failure rate.28 choice with EE reserved for refractory cases or unique
ESPAL has similarly been found to be more cost-effective scenarios.33,34
than EE.32 Second, ESPAL affords the otolaryngologist the
opportunity to perform a thorough examination of the nasal Conclusions
cavity, greatly increasing the chance of identifying the
EE provided a high immediate success rate in patients with
source of bleeding, which is uncommon during angiography.
severe epistaxis, although the overall short- and long-term
This allows for targeted ligation of the sphenopalatine, ante-
rebleed rates were high. In light of the significant potential
rior, or posterior ethmoid arteries (which are not accessible
risks and higher cost, patients with epistaxis should be con-
endovascularly) or local treatment of severe mucosal bleed-
sidered for endoscopic surgery first. The role of EE in the
ing. At our institution, such targeted treatment has resulted
treatment of epistaxis needs to be better defined but serves
in excellent sustained efficacy (91%-93%) and a more
an important role in refractory cases and unique scenarios.
benign side effect profile compared to EE, typically nasal
crusting and palatal numbness. Following the implementa- Acknowledgments
tion of the clinical care pathway, no cases of epistaxis Special thanks to Li Wang and Dan Winger.
required EE for over a year.
Even so, EE plays an extremely important role in the
Author Contributions
treatment of epistaxis. EE is an excellent treatment alterna-
tive to nasal packing for patients in whom general anesthe- Phillip Huyett, acquisition/analysis/interpretation of data, drafted
sia is contraindicated. For example, in our study cohort, EE and wrote the manuscript, final approval, accountability for the
work; Brian T. Jankowitz, interpretation of data, critical review
was performed in patients with end-stage liver disease, who
(revised and edited the manuscript), final approval, accountability
notoriously have diffuse mucosal bleeding related to pro- for the work; Eric W. Wang, interpretation of data, critical review
found clotting factor deficiencies and are extremely poor (revised and edited the manuscript), final approval, accountability
surgical candidates. In this small sample size (n = 6), all for the work; Carl H. Snyderman, interpretation of data, critical
were successfully treated with EE with no immediate-, review (drafted, wrote and edited the manuscript), final approval,
short-, or long-term failures or complications. In patients accountability for the work.
6 Otolaryngology–Head and Neck Surgery

16. Oguni T, Korogi Y, Yasunaga T, et al. Superselective emboli-

Disclosures
sation for intractable idiopathic epistaxis. Br J Radiol. 2000;
Competing interests: None. 73:1148-1153.
Sponsorships: None. 17. Kumar S, Shetty A, Rockey J, Nilssen E. Contemporary surgi-
Funding source: The statistical analysis performed in this project cal treatment of epistaxis: what is the evidence for sphenopala-
was supported by the National Institutes of Health through grant tine artery ligation? Clin Otolaryngol Allied Sci. 2003;28:
UL1TR000005. 360-363.
18. Cullen MM, Tami TA. Comparison of internal maxillary
References artery ligation versus embolization for refractory posterior
1. Small M, Murray JA, Maran AG. A study of patients with epistaxis. Otolaryngol Head Neck Surg. 1998;118:636-642.
epistaxis requiring admission to hospital. Health Bull (Edinb). 19. Strong EB, Bell DA, Johnson LP, Jacobs JM.Intractable epis-
1982;40:20-29. taxis: transantral ligation vs. embolization: efficacy review
2. Schaitkin B, Strauss M, Houck JR. Epistaxis: medical versus and cost analysis. Otolaryngol Head Neck Surg. 1995;113:
surgical therapy: a comparison of efficacy, complications, and 674-678.
economic considerations. Laryngoscope. 1987;97:1392-1396. 20. Seno S, Arikata M, Sakurai H, et al. Endoscopic ligation of
3. Klotz DA, Winkle MR, Richmon J, Hengerer AS. Surgical the sphenopalatine artery and the maxillary artery for the treat-
management of posterior epistaxis: a changing paradigm. ment of intractable posterior epistaxis. Am J Rhinol Allergy.
Laryngoscope. 2002;112:1577-1582. 2009;23:197-199.
4. Zayyan E, Bajin MD, Aytemir K, Yilmaz T. The effects on 21. Soyka MB, Nikolaou G, Rufibach K, Holzmann D. On the
cardiac functions and arterial blood gases of totally occluding effectiveness of treatment options in epistaxis: an analysis of
nasal packs and nasal packs with airway. Laryngoscope. 2010; 678 interventions. Rhinology. 2011;49:474-478.
120:2325-2330. 22. Howe DJ, Wazir U, Skinner DW. Outcomes of endoscopic
5. Brinjikji W, Kallmes DF, Cloft HJ. Trends in epistaxis emboli- sphenopalatine artery ligation for epistaxis: a five-year series
zation in the United States: a study of the Nationwide from a single institution. Ear Nose Throat J. 2012;91:70-72.
Inpatient Sample 2003-2010. J Vasc Interv Radiol. 2013;24: 23. Voegels RL, Thome DC, Iturralde PP, Butugan O. Endoscopic
969-973. ligature of the sphenopalatine artery for severe posterior epis-
6. Christensen NP, Smith DS, Barnwell SL, Wax MK. Arterial taxis. Otolaryngol Head Neck Surg. 2001;124:464-467.
embolization in the management of posterior epistaxis. 24. O’Flynn PE, Shadaba A. Management of posterior epistaxis
Otolaryngol Head Neck Surg. 2005;133:748-753. by endoscopic clipping of the sphenopalatine artery. Clin
7. Dubel GJ, Ahn SH, Soares GM. Transcatheter embolization in Otolaryngol Allied Sci. 2000;25:374-377.
the management of epistaxis. Semin Intervent Radiol. 2013;30: 25. Snyderman CH, Goldman SA, Carrau RL, Ferguson BJ,
249-262. Grandis JR. Endoscopic sphenopalatine artery ligation is an
8. Willems PW, Farb RI, Agid R. Endovascular treatment of effective method of treatment for posterior epistaxis. Am J
epistaxis. AJNR Am J Neuroradiol. 2009;30:1637-1645. Rhinol. 1999;13:137-140.
9. Elden L, Montanera W, Terbrugge K, Willinsky R, Lasjaunias 26. Sharp HR, Rowe-Jones JM, Biring GS, Mackay IS. Endoscopic
P, Charles D. Angiographic embolization for the treatment of ligation or diathermy of the sphenopalatine artery in persistent
epistaxis: a review of 108 cases. Otolaryngol Head Neck Surg. epistaxis. J Laryngol Otol. 1997;111:1047-1050.
1994;111:44-50. 27. Gottumukkala R, Kadkhodayan Y, Moran CJ, Cross de WT
10. Tseng EY, Narducci CA, Willing SJ, Sillers MJ. Angiographic 3rd, Derdeyn CP. Impact of vessel choice on outcomes of
embolization for epistaxis: a review of 114 cases. polyvinyl alcohol embolization for intractable idiopathic epis-
Laryngoscope. 1998;108:615-619. taxis. J Vasc Interv Radiol. 2013;24:234-239.
11. Fukutsuji K, Nishiike S, Aihara T, et al. Superselective angio- 28. Dedhia RC, Desai SS, Smith KJ, et al. Cost-effectiveness of
graphic embolization for intractable epistaxis. Acta Otolaryngol. endoscopic sphenopalatine artery ligation versus nasal packing
2008;128:556-560. as first-line treatment for posterior epistaxis. Int Forum
12. Vitek J. Idiopathic intractable epistaxis: endovascular therapy. Allergy Rhinol. 2013;3:563-566.
Radiology. 1991;181:113-116. 29. Moshaver A, Harris JR, Liu R, Diamond C, Seikaly H. Early
13. Gurney TA, Dowd CF, Murr AH. Embolization for the treat- operative intervention versus conventional treatment in epis-
ment of idiopathic posterior epistaxis. Am J Rhinol. 2004;18: taxis: randomized prospective trial. J Otolaryngol. 2004;33:
335-339. 185-188.
14. Duncan IC, Fourie PA, le Grange CE, van der Walt HA. 30. Nikolaou G, Holzmann D, Soyka MB. Discomfort and costs in
Endovascular treatment of intractable epistaxis—results of a 4- epistaxis treatment. Eur Arch Otorhinolaryngol. 2013;270:
year local audit. S Afr Med J. 2004;94:373-378. 2239-2244.
15. Sadri M, Midwinter K, Ahmed A, Parker A. Assessment of 31. McDermott AM, O’Cathain E, Carey BW, O’Sullivan P,
safety and efficacy of arterial embolisation in the management Sheahan P. Sphenopalatine artery ligation for epistaxis:
of intractable epistaxis. Eur Arch Otorhinolaryngol. 2006;263: factors influencing outcome and impact of timing of surgery.
560-566. Otolaryngol Head Neck Surg. 2016;154:547-552.
Huyett et al 7

32. Rudmik L, Leung R. Cost-effectiveness analysis of endoscopic 35. Abdelkader M, Leong SC, White PS. Endoscopic control of
sphenopalatine artery ligation vs arterial embolization for the sphenopalatine artery for epistaxis: long-term results. J
intractable epistaxis. JAMA Otolaryngol Head Neck Surg. Laryngol Otol. 2007;121:759-762.
2014;140:802-808. 36. Gede LL, Aanaes K, Collatz H, Larsen PL, von Buchwald C.
33. Vosler PS, Kass JI, Wang EW, Snyderman CH. Successful National long-lasting effect of endonasal endoscopic spheno-
implementation of a clinical care pathway for management of palatine artery clipping for epistaxis. Acta Otolaryngol. 2013;
epistaxis at a tertiary care center. Otolaryngol. Head Neck 133:744-748.
Surg. 2016;155:879-885.
34. Leung RM, Smith TL, Rudmik L. Developing a laddered algo-
rithm for the management of intractable epistaxis: a risk analy-
sis. JAMA Otolaryngol Head Neck Surg. 2015;141:405-409.