Transnasal endoscopic resection of juvenile nasopharyngeal angiofibroma without preoperative embolization.

Abstract

Juvenile nasopharyngeal angiofibroma (JNA) is a benign, highly vascular, and locally invasive tumor. Because the location of these tumors makes conventional surgery difficult, interest in endoscopic resection is increasing, particularly for the treatment of lesions that do not extend laterally into the infratemporal fossa. We report the results of our series of 23 patients with JNA (stage IIB or lower) who underwent transnasal endoscopic resection under hypotensive general anesthesia without preoperative embolization of the tumor All tumors were successfully excised. The amount of intraoperative blood loss was acceptable. We observed only 1 recurrence, which was diagnosed 19 months postoperatively in a patient with a stage IIB primary tumor. We observed only 3 complications during follow-up-all synechia. We conclude that endoscopic resection of JNAs is safe and effective. The low incidence of recurrence and complications in this series indicates that preoperative embolization may not be necessary for lesions that have not undergone extensive spread; instead, intraoperative bleeding can be adequately controlled with good hypotensive general anesthesia.

Full text

Page 1

BORGHEI, BARADARANFAR, BORGHEI, SOKHANDON
ORIGINAL ARTICLE
740
ENT-Ear, Nose & Throat Journal  November 2006
Transnasal endoscopic resection of
juvenile nasopharyngeal angiofibroma
without preoperative embolization
From the Ear, Nose, Throat Research Center, Amir-Alam Hospital, Teh-
ran University of Medical Sciences, Tehran, Iran (Dr. P. Borghei,
Dr. Baradaranfar, and Dr. S.H. Borghei), and the Department of
Diagnostic Radiology, University of Michigan Medical Center,
Ann Arbor (Dr. Sokhandon).
Reprint requests: Peyman Borghei, MD, Ear, Nose, Throat Research
Center, Amir-Alam Hospital, North Sa’di Ave., Tehran, 11457
Iran. Phone: 98-912-115-2868; fax: 98-21-6676-0245; e-mail:
borgheii@sina.tums.ac.ir
Peyman Borghei, MD; Mohammad Hossein Baradaranfar, MD; Seyed Hebatodin Borghei, MD;
Farnoosh Sokhandon, MD
Abstract
Juvenile nasopharyngeal angiofibroma (JNA) is a benign,
highly vascular, and locally invasive tumor. Because the
location of these tumors makes conventional surgery
difficult, interest in endoscopic resection is increasing,
particularly for the treatment of lesions that do not extend
laterally into the infratemporal fossa. We report the results
of our series of 23 patients with JNA (stage IIB or lower)
who underwent transnasal endoscopic resection under
hypotensive general anesthesia without preoperative embo-
lization of the tumor. All tumors were successfully excised.
The amount of intraoperative blood loss was acceptable.
We observed only 1 recurrence, which was diagnosed
19 months postoperatively in a patient with a stage IIB
primary tumor. We observed only 3 complications during
follow-up—all synechia. We conclude that endoscopic
resection of JNAs is safe and effective. The low incidence
of recurrence and complications in this series indicates
that preoperative embolization may not be necessary for
lesions that have not undergone extensive spread; instead,
intraoperative bleeding can be adequately controlled with
good hypotensive general anesthesia.
Introduction
In 1906, Chaveau introduced the term juvenile nasopha-
ryngeal angiofibroma (JNA).1 This benign tumor is char-
acterized by aggressive local invasiveness and a tendency
toward local recurrence after incomplete resection; it occurs
primarily in adolescent males.2 JNA is a nonencapsulated,
submucosal, spreading tumor made up of fibrous con-
nective tissue and an abundance of endothelium-lined
vascular spaces.3
Many studies have indicated that JNA originates in the
pterygopalatine fossa at the aperture of the pterygoid canal.4
From there, it can extend to surrounding structures, includ-
ing the nasal cavity, sphenoid sinus and sella, infratemporal
fossa, inferior orbital fissure, and intracranial area.5
A diagnosis of JNA is suggested by the classic triad of
epistaxis, nasal obstruction, and a nasopharyngeal mass.
The presence of other symptoms depends on the direction
and extent of tumor spread.6,7 The treatment of choice for
localized primary tumors is surgical resection; for exten-
sive masses and for recurrent tumors, radiotherapy is also
recommended.8,9
Because the location of these tumors makes conven-
tional surgery difficult, interest in endoscopic resection
is increasing, particularly for lesions that do not extend
laterally to the infratemporal fossa.10,11 In this article, we
report the results of our series of transnasal endoscopic
resections without preoperative embolization of the tumor
in patients with JNA that did not extend laterally into the
infratemporal fossa.
Patients and methods
Our series included 23 patients with histopathologically
proven JNA (stage IIB or lower) who underwent transna-
sal endoscopic resection without tumor embolization at
our center from February 2000 through August 2004. All
patients were male. Their ages ranged from 10 to 26 years
(mean: 16.2). No patient underwent radiotherapy.
Staging. Tumor staging was based on the results of
computed tomography (CT) and/or magnetic resonance im-
aging (MRI). Tumors were staged according to the system
described by Radkowski et al (figure).12 Five tumors were
staged as IA, 9 as IB, 4 as IIA, and 5 as IIB (table 1).
Surgical technique. Hypotensive general anesthesia
was administered (target mean arterial pressure: 55 to 65
mm Hg) with the patient in the reverse Trendelenburg posi-
tion. Topical vasoconstriction was achieved by placing a

Page 2

Volume 85, Number 11
741
TRANSNASAL ENDOSCOPIC RESECTION OF JUVENILE NASOPHARYNGEAL ANGIOFIBROMA WITHOUT PREOPERATIVE EMBOLIZATION
cottonoid pledget soaked in 1:1,000 epinephrine solution
for 10 minutes.
The surgical technique varied slightly from case to case,
depending on the tumor’s size and degree of extension, but
all dissections followed a series of the same basic steps. The
tumor was slowly pulled down with a retractor to expose
the sphenopalatine foramen. If the sphenopalatine artery
was evident through the sphenopalatine foramen, it was
cauterized. We did not clip the artery, and we did not use
cautery to shrink the tumor.
The tumor was detached from the nasal mucosa and sphe-
noid sinus and pushed medially and inferiorly. If extension
to the lower part of the pterygopalatine fossa was present,
the posterior portion of the inferior turbinate was resected.
These maneuvers exposed the superior, inferior, and medial
borders of the tumor. The posterior wall of the maxillary
sinus was resected, and the retromaxillary periosteum
was elevated, which provided access to the infratemporal
fossa. The tumor and surrounding soft tissue were pulled
inferomedially, and the pterygopalatine fossa was cleaned.
The entire tumor was removed via the mouth. The nasal
cavity was packed with Merocel for 48 hours.
The first follow-up endoscopic evaluation was performed
3 weeks postoperatively and repeated 2 months later.
Subsequent follow-ups with CT or MRI were scheduled
for the third month and then every 6 months up to 2 years
and once yearly thereafter. Overall, the length of individual
follow-ups ranged from 13 to 57 months (mean: 33).
Outcome variables included estimated perioperative
blood loss, recurrences, and complications. Any patient
whose symptoms returned or who exhibited an expanding
tumor on imaging was considered to have experienced a re-
currence and was therefore a candidate for reoperation.
Results
The estimated average amount of intraoperative blood loss
per patient and the average number of blood transfusion
packs required per patient were positively correlated with
tumor stage (table 1):

• stage IA: 666 ml blood loss; 0 packs for transfusion
• stage IB: 680 ml; 0.2 packs
• stage IIA: 1,068 ml; 0.75 packs
• stage IIB: 1,310 ml; 1.8 packs

During follow-up, 3 patients developed asymptomatic
synechia between the nasal septum and one of the turbi-
nates.
Only 1 patient (4.3%) experienced a recurrence. This
patient’s primary tumor had been staged as IIB. The recur-
rence was diagnosed 19 months postoperatively after the
patient had complained of repeated episodes of epistaxis
and occasional nasal obstruction. The patient underwent
reoperation and remained disease-free at 28 months.
Discussion
A diagnosis of JNA requires close attention to symptoms.
In most cases, the aforementioned triad of epistaxis, nasal
obstruction, and a nasopharyngeal mass in a young male is
suggestive of JNA. The presence of anterior bowing of the
posterior wall of the maxillary sinus on imaging (Holman-
Miller sign) is a known pathognomonic finding.13
Many surgical approaches to JNA have been used, in-
cluding the transpalatal approach, medial maxillectomy,
midfacial degloving, LeFort I surgery, lateral rhinotomy,
and the infratemporal fossa approach. However, transnasal
endoscopic resection is becoming an attractive alternative
to these external approaches as our surgical expertise and
our knowledge of the intranasal anatomy improve. Still,
the proper role of endoscopic resection for the management
of angiofibromas remains a subject of debate.4
Hypotensive general anesthesia. We did not perform
preoperative embolization because we believe that intraop-
erative bleeding of lesions that have not spread extensively
can be controlled by good hypotensive general anesthesia.
Hypotensive general anesthesia may be appropriate for sev-
eral types of operations, including head and neck surgery,
neurosurgery, large orthopedic procedures, and a variety
of plastic surgical procedures. The possible complications
of hypotensive general anesthesia primarily involve the
Stage Description
IA Involvement limited to the nose and/or naso-
pharynx
IB Extension into one or more sinuses
IIA Minimal extension into the pterygopalatine
fossa
IIB Occupation of the entire pterygopalatine
fossa with or without erosion of the orbital
apex
IIC Involvement of the infratemporal fossa with
or without extension to the cheek or posterior
to the pterygoid plates
IIIA Erosion of the skull base (the middle cranial
fossa/base of the pterygoids); minimal intra-
cranial extension
IIIB Erosion of the skull base; extensive intra-
cranial extension with or without cavernous
sinus invasion
Figure. The Radkowski staging system for juvenile nasopharyn-
geal angiofibroma.12

Page 3

BORGHEI, BARADARANFAR, BORGHEI, SOKHANDON
742
ENT-Ear, Nose & Throat Journal  November 2006
nervous system. The most common are dizziness and
cerebral thrombosis. Less common complications include
hemiplegia, retinal thrombosis, and even death.
During hypotensive general anesthesia, a target blood
pressure of 50 to 65 mm Hg is safe for young, otherwise
healthy patients. This type of anesthesia is probably more
risky in the elderly and in those who have underlying
organ dysfunction.14 In our study, no patient experienced
any anesthesia-related complication.
Preoperative embolization. Another reason to avoid
embolization is that it has potential complications of its
own.15 They include nerve injury (e.g., facial nerve palsy)
and devitalization of tissues such as the overlying skin and
nasal mucosa. Intentional embolization of the branches of
the internal carotid artery may result in accidental embo-
lization of the brain and ophthalmic artery.
Even so, embolization does have its proponents, as
several authors have described their favorable experiences
with embolization during transnasal endoscopic excision
of JNAs (table 2). For example, Li et al reported that
patients with stage IIC or lower lesions who underwent
preoperative embolization lost significantly less blood
intraoperatively than did those who were not embolized
(mean: 637 and 1,136 ml, respectively; p < 0.05).15 Also,
among patients who required transfusion, embolized
patients required significantly less blood (mean: 400 and
836 ml; p < 0.01).
Likewise, Liu et al reported that preoperative emboliza-
tion was associated with significantly less intraoperative
bleeding in patients with stage IA tumors (mean: 275 and
840 ml, respectively).16 They also compared the amount of
blood loss in 17 patients with stage IB and higher tumors—8
of these patients had undergone embolization of feeding
arteries and 9 had undergone ligation of an external carotid
artery—but they found no significant difference.
Nicolai et al reported that the amount of intraoperative
blood loss in 15 patients, all of whom underwent preop-
erative embolization, ranged from 80 to 600 ml (mean:
372).17 Moulin et al also reported a significant decrease
in intraoperative blood loss.18
Both Siniluoto et al19 and Elasfour et al20 reported that
embolization not only reduced intraoperative blood loss,
but it also contributed to improved surgical results.
Tranbahuy et al recommended intratumoral emboliza-
tion.21 Hazarika et al reported that preoperative emboliza-
tion followed by KTP laser-assisted endoscopic excision
of JNA is superior to radical approaches.22
Staging. As all of our patients were staged as IIB or
lower, our findings support the effectiveness of the trans-
nasal endoscopic approach for tumors that have extended
to the nasal cavity, nasopharynx, paranasal sinuses, and
pterygopalatine fossa. Other authors have recommended
this procedure for tumors up to stage IIC, citing the benefits
of fewer postoperative complications and a more rapid
Table 1. Results of treatment with endoscopic transnasal technique

Pt.
No. units
transfused
Blood
loss (ml)
Follow-up
(mo) Stage Recurrence Complication
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
IA
IA
IA
IA
IA
IB
IB
IB
IB
IB
IB
IB
IB
IB
IIA
IIA
IIA
IIA
IIB
IIB
IIB
IIB
IIB
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
1
2
2
2
1
2
2
550
850
570
660
700
550
620
800
450
1,150
650
750
550
600
820
1,100
1,050
1,300
1,200
1,300
1,600
1,350
1,100
46
42
36
30
20
57
54
48
48
46
35
22
16
14
44
17
47
13
47
24
22
18
16
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
+
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Synechia
–
Synechia
–
–
Synechia
–

Page 4

Volume 85, Number 11
743
TRANSNASAL ENDOSCOPIC RESECTION OF JUVENILE NASOPHARYNGEAL ANGIOFIBROMA WITHOUT PREOPERATIVE EMBOLIZATION
recovery and earlier discharge in addition to less bleed-
ing.12,23-25 However, involvement of the infratemporal fossa,
anterior skull base, and orbit still requires a more invasive
surgical approach,26 although Onerci et al recommended
endoscopic resection for lesions with minimal extension
into the cranium.27
Recurrence. Scholtz et al performed transnasal endo-
scopic excision on 7 patients with JNAs that had extended
into the nasal cavity, the nasopharynx, the pterygopalatine
fossa, and the ethmoid, sphenoid, and maxillary sinuses.10
They reported no recurrences.
Likewise, Wormald and Van Hasselt reported no recur-
rences in 7 patients with lesions that had extended into
the nasal cavity, paranasal sinuses, and pterygopalatine
fossa, including those with minimal invasion into the
infratemporal fossa.28
Finally, Nicolai et al reported only 1 recurrence in 15
patients.17 The recurrent lesion was small and did not
extend laterally. That patient had undergone conservative
surgery that had not included dissection of the pterygo-
palatine fossa; also, the middle turbinate was spared. The
amount of intraoperative blood loss in these 15 patients,
all of whom underwent preoperative embolization, ranged
from 80 to 600 ml (mean: 372).
Complications. In our study, we observed only 3 com-
plications—all synechia. Others have also reported a low
incidence of complications.10,24,25
Conclusions
In summary, we conclude that the transnasal endoscopic
technique is an acceptable approach to the resection of
JNAs without extensive lateral extension. Our study
demonstrated a low degree of bleeding, a low recurrence
rate, and minimal complications.
Although we did not perform preoperative emboliza-
tion of the feeding arteries, we did not encounter massive
intraoperative bleeding. Therefore, we believe that intra-
operative bleeding can be adequately controlled with good
hypotensive general anesthesia in patients whose tumor
does not extend into the infratemporal fossa or cranium.
References
1. Chaveau C. Histoire des Maladies du Pharynx. Paris: Balliere;
1906.
2. Iannetti G, Belli E, De Ponte F, et al. The surgical approaches to
nasopharyngeal angiofibroma. J Craniomaxillofac Surg 1994;22:
311-16.
3. Ungkanont K, Byers RM, Weber RS, et al. Juvenile nasopharyngeal
angiofibroma: An update of therapeutic management. Head Neck
1996;18:60-6.
4. Lloyd G, Howard D, Phelps P, Cheesman A. Juvenile angiofibroma:
The lessons of 20 years of modern imaging. J Laryngol Otol 1999;
113:127-34.
5. Batsakis JG. Tumors of the Head and Neck: Clinical and Pathologi-
cal Considerations. 2nd ed. Baltimore: Williams & Wilkins; 1979:
296-300.
6. Kennedy JD, Haines SJ. Review of skull base surgery approaches:
With special reference to pediatric patients. J Neurooncol 1994;20:
291-312.
7. Gullane PJ, Davidson J, O’Dwyer T, Forte V. Juvenile angiofibroma:
A review of the literature and a case series report. Laryngoscope
1992;102:928-33.
8. Wiatrak BJ, Koopmann CF, Turrisi AT. Radiation therapy as an
alternative to surgery in the management of intracranial juvenile
nasopharyngeal angiofibroma. Int J Pediatr Otorhinolaryngol 1993;
28:51-61.
9. Gudea F, Vega M, Canals E, et al. Role of radiation therapy for
“juvenile” angiofibroma. J Laryngol Otol 1990;104:725-6.
Table 2. Review of the results of other studies

Author
Type of
surgery
Highest
stage

Recurrences
Mean blood
loss
Mean blood
transfused
Li et al,15 1998


Conventional


IIC


1 of 21


With preoperative
embolization: 637 ml;
without: 1,136 ml
With preoperative
embolization: 400 ml;
without: 836 ml
Liu et al,16 2002



Scholtz et al,10 2001


Wormald and
Van Hasselt,28 2003
Conventional


I (see text)


0 of 10


With preoperative
embolization: 275 ml;
without: 840 ml
Not reported
Transnasal
endoscopic

Transnasal
endoscopic
IIB 0 of 7 Not reported Not reported

IIC 0 of 7 Not reported Not reported
Nicolai et al,17 2003

Transnasal
endoscopic
IIC 1 of 15 372 ml Not reported
Continued on page 746