HN 08-2005 Surgical Management of Parapharyngeal

7/27/2019 HN 08-2005 Surgical Management of Parapharyngeal

1/7

SURGICAL MANAGEMENT OF PARAPHARYNGEAL

SPACE MASSES

Seth M. Cohen, MD, MPH, Brian B. Burkey, MD, James L. Netterville, MD

Vanderbilt University Medical Center, Department of Otolaryngology Head & Neck Surgery,5025 Hillsboro Road, 7D, Nashville, TN 37215. E-mail: [email protected]

Accepted 2 February 2005

Published online 6 May 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/hed.20199

Abstract: Background. We sought to examine surgical tech-

niques used to remove parapharyngeal space (PPS) masses.

Methods. This retrospective search was conducted from

1980 to 2003. Age, sex, diagnosis, surgical approach, compli-

cations, and outcome were collected.

Results. One hundred sixty-six PPS masses were identi-

fied: 21 (12.7%) were malignant, 145 (87.3%) were benign, 76

(45.8%) were vascular, and 69 (41.6 %) involved the skull base.

Transcervical techniques were used in all cases. Removing the

styloid and its musculature and level II lymphadenectomies

enhanced exposure for vascular and skull base tumors. Thirty

transcervical transmastoid dissections (20.4%) facilitated

removal of vascular skull base tumors. To identify the facial

nerve, 20 transparotid transcervical approaches (13.6%) were

performed. Three mandibulotomies (2.0%) were required for

internal carotid artery involvement. Expected neurologic se-

quelae resulted from cranial nerve involvement by tumor.

Three patients (2.0%), all presenting with recurrent cancer,

had local recurrences.

Conclusion. Careful patient assessment and surgical tech-

niques allow the oncologically safe removal of benign, vascular,

and skull base PPS tumors. A2005 Wiley Periodicals, Inc. Head

Neck 27: 669675, 2005

Keywords: parapharyngeal; surgical treatment; pharyngeal neo-

plasms; paraganglioma; salivary gland neoplasms

Parapharyngeal space (PPS) masses are uncom-

mon lesions, accounting for 0.5% of all head andneck tumors.1 The complex anatomic relationships

and proximity of vital neurovascular structures

necessitates careful preoperative evaluation and

precise surgical techniques. Various surgical ap-

proaches, which may be combined with mandible-

splitting procedures to increase exposure, have

been described.2 The mandibulotomy approach

may cause scarring or injury to the inferior alve-

olar nerve or teeth and may require tracheotomy.

Because approximately 80% of PPS lesions are

benign, surgical removal should carefully consider

patient morbidity.35 This study further explores

the utility of transcervical approaches and spe-cific surgical techniques to enhance exposure for

resecting PPS masses.

MATERIALS AND METHODS

Approval for the study was obtained from the

Vanderbilt Institutional Review Board. A medical

records search for PPS masses was performed

from January 1980 to December 2003. Only

Correspondence to: S. M. Cohen

Reprints: J. L. Netterville, Department of Otolaryngology-Head & NeckSurgery, 2900 The Vanderbilt Clinic, Nashville, TN 37232.

Presented at the 6th International Head and Neck Cancer Conference,Washington, DC, August 7 11, 2004.

B 2005 Wiley Periodicals, Inc.

Surgical Management of Parapharyngeal Space Masses HEAD & NECK August 2005 669


2/7

tumors arising in and primarily involving the

PPS were included for analysis. Two carotid body

tumors with significant extension above the pos-

terior belly of the digastric were included. Data

regarding patient age, sex, diagnosis, location, and

characteristics of the mass, radiologic findings,

surgical approaches, complications, and outcome

were collected. Skull base involvement was de-fined as tumors eroding the skull base or extend-

ing into the jugular foramen or intracranially.

Surgical approaches were categorized into sev-

eral different types. A transcervical approach ac-

cessed the PPS through the neck and did not use

a mandibulotomy. A transcervical transparotid

approach combined a neck and parotid dissection

to allow identification of the facial nerve. The

transoral transcervical approach allowed dual

access from the neck and oral cavity. An extended

approach was defined as cases in which the

digastric, styloid musculature, stylomandibular

ligament, and/or styloid process were removed.A transcervical mandibulotomy dissection ex-

tended the cervical exposure with a mandibu-

lotomy. Last, the transcervicaltransmastoid ap-

proach allowed exposure to proceed superiorly

from the neck by removing the mastoid and po-

tentially exposing the middle cranial fossa.

Data are presented in numerical and percent

form. Categorical data analysis was performed

with chi-square techniques or the Fisher exact

test. All statistical analysis used SigmaStat

software 2.03 (SPSS Inc., Chicago, IL).

RESULTS

Of 234 possible lesions, 177 PPS masses meeting

the inclusion criteria were identified. Because

they did not primarily involve the PPS, 57

tumors, primarily oropharyngeal squamous cell

carcinomas and tumors of the masticator space,

were excluded. Medical records could not be

obtained for 11 patients, leaving 166 patients in

the study, of which 70.0% were women. The me-

dian age was 47 years (range, 6 months79 years).

Among the PPS masses, 45 (27.1%) were of

salivary gland origin, 76 (45.8%) vascular, 88

(53.0%) neurogenic (including the vascular vagal

paragangliomas), and 22 (13.3%) of miscellaneous

origin (Table 1).

Twenty-one lesions (12.7%) were malignant,

most commonly malignant deep lobe parotid

tumors, and 145 (87.3%) were benign (Table 1).

Sixty-nine tumors (41.6%) either extended to or

involved the skull base, and 25 (15.1%) presented

as recurrent lesions. The mean maximum dimen-

sion of each lesion was 4.7 F 2.2 cm.

One fourth of the patients were asymptomatic

and had their tumor found incidentally (Table 2).

Symptoms were present for a mean duration of

14.3 months (range, 3108 months). Fullness in

the pharynx or neck was the most common

physical finding (Table 3).

Nineteen patients lesions (11.4%) were ob-

served for a mean of 35.2 months (range, 1

108 months). Eleven had asymptomatic neuro-

genic lesions, and eight had bilateral paragan-

Table 1. Diagnosis of parapharyngeal space masses.

Diagnosis No. masses (%)*

Salivary total 45 (27.1)

Benign salivary total 34 (75.6)

Deep lobe pleomorphic adenoma 22 (64.7)

Minor salivary pleomorphic adenoma 11 (32.4)

Myoepithelioma 1 (2.9)

Malignant salivary total 11 (24.4)

Malignant myoepithelial carcinoma 3 (27.3)

Low-grade adenocarcinoma 2 (18.2)

Carcinoma ex pleomorphic 1 (9.1)

Undifferentiated carcinoma 1 (9.1)

Mucoepidermoid carcinoma 1 (9.1)

Adenoid cystic carcinoma 1 (91)

High-grade adenocarcinoma 1 (9.1)

Acinic cell carcinoma 1 (9.1)

Vascular total 76 (45.8)

Paraganglioma 65 (85.5)

Carotid aneurysm 5 (6.6)

Vertebral aneurysm 1 (1.3)

Carotid body tumor 2 (2.6)

Hemangiopericytoma 1 (1.3)

Castlemans disease 1 (1.3)Hemangioma 1 (1.3)

Neurogenic total 88 (53.0)

Paragangliomay 65 (73.9)

Schwannomaz 16 (18.2)

Neurofibroma 7 (8.0)

Other malignant totalO 10 (6.0)

Metastatic thyroid carcinoma 4 (40.0)

Chondrosarcoma 1 (10.0)

Rhabdomyosarcoma 1 (10.0)

Metastatic esthesioneuroblastoma 1 (10.0)

Malignant carotid body 1 (10.0)

Other benign totalb 12 (7.2)

*Total percents calculated from all 166 tumors. Remainder calculated

from individual subclasses.

yParaganglioma includes 61 vagal and four sympathetic paragangliomas.zSchwannoma includes nine sympathetic, five vagal, one glossopharyn-

geal, and one facial nerve.Neurofibroma includes four vagal, two glossopharyngeal, and one sym-

pathetic neurofibroma.

OMalignant total also includes the hemangiopericytoma and a malignant

paraganglioma.

bOther includes one lipoma, one cystic hygroma, one parotid cutaneous

fistula, one rhabdomyoma, one encephalocele, one fibrous tumor, one on-

cocytic cyst, two branchial cleft cysts, one epidermoid cyst, one Thoro-

trast granuloma, and one meningioma.

HEAD & NECK August 2005670 Surgical Management of Parapharyngeal Space Masses


3/7

gliomas. Stereotactic radiation was used in two

patients with bilateral lesions to control the

enlarging mass.

Most PPS masses (69.1%) could be excised by

the transcervical technique alone. Tumors abut-

ting but not involving the skull base were three

times more likely to have a transcervical excisionalone vs other techniques (p = .004, chi-square;

Table 4). Vascular exposure and control was

sufficient to remove almost half of the vascular

masses. Most salivary and neurogenic masses

were also removed transcervically. Tumor loca-

tion in either the prestyloid or poststyloid space

did not influence the ability to perform a trans-

cervical resection (p > .05, chi-square).

Extending the transcervical approach by di-

viding the digastric, styloid musculature, stylo-

mandibular ligament, and/or styloid process

occurred in 89 cases (60.5%). Approximately twothirds of both prestyloid and poststyloid tu-

mors required these steps to facilitate exposure.

Tumors abutting or involving the skull base

more commonly required these maneuvers com-

pared with non skull base tumors (79.4% vs

55.9%; p = 0.008, chi-square). Tumor vascular-

ity, size, and malignancy did not correlate with

these techniques.

Excluding recurrent and malignant tumors, in

which a more extensive resection would be appro-

priate, a level II lymphadenectomy was used to

improve exposure for vascular control in 58.0%

of operations. Vascular lesions were statistically

significantly more likely than nonvascular masses

to have level II excised (93.8% vs 46.3%, respec-

tively; p V .001, chi-square). Tumor size and prox-

imity to the skull base were not associated with

a level II lymphadenectomy.

Thirty tumors (20.4%) required a transcervi-

cal transmastoid dissection to provide optimal

neurovascular control for excision. Vascular tu-

mors that eroded the skull base or had jugular

foramen extension with or without intracranial

spread were six times more likely to require a

transcervical transmastoid resection compared

with other approaches (p V .001, chi-square;Table 4). Five patients with paragangliomas with

skull base involvement also underwent preoper-

ative embolization. Tumor location, when control-

ling for vascularity and skull base involvement,

tumor size, and malignancy were not associated

with this approach.

A transcervical transparotid approach facili-

tated the removal of 20 tumors (13.6%) adhering

to the deep lobe of the parotid. Salivary lesions

were three times as likely to require this ap-

proach compared with other tumors (p V 0.001,

chi-square). Overall, one third of salivary masseswere treated in this manner (Table 4).

Three tumors (2.0%), two malignant salivary

gland tumors, and one vagal paraganglioma had

internal carotid artery involvement and required

a mandibulotomy to aid resection and internal

carotid artery replacement. Three masses (2.0%)

had prior transoral biopsies, prompting a com-

bined transcervical transoral technique.

Follow-up data were available for all but

10.9% of patients, with a mean of 27.9 months

(range, 1 168 months). Three patients (2.0%)

initially were seen at our institution with recur-

rent disease. These three patients, one with a

pleomorphic deep lobe parotid adenoma, one with

a deep lobe parotid high-grade adenocarcinoma,

and one with a deep lobe parotid acinic cell car-

cinoma, all had recurrent disease after excision.

One patient (0.7%) with chondrosarcoma had

persistent disease after a planned debulking

operation that was treated by stereotactic radio-

surgery. One patient (0.7%) with vagal para-

Table 3. Physical examination findings.

Finding No. patients (%)*

Neck mass 51 (30.7)

Pharyngeal fullness 42 (25.3)

Vocal cord paresis/paralysis 18 (10.8)

Normal examination 16 (9.6)

Palatal weakness 14 (8.4)

Tongue weakness 13 (7.8)

Shoulder weakness 7 (4.2)

Facial nerve weakness 7 (4.2)

Serous otitis media 6 (3.6)

Horners syndrome 3 (1.8)

Cutaneous fistula 1 (0.6)

None listed 11 (6.6)

*Percent is calculated from all 166 patients.

Table 2. Symptoms.

Symptom No. patients (%)*

Asymptomatic 42 (25.3)

Dysphagia/globus pharyngeus 12 (7.2)

Hoarseness 12 (7.2)

Pain 10 (6.0)

Tinnitus 8 (4.8)

Hearing loss 6 (3.6)

Othery 6 (3.6)

*Percent is calculated from all 166 patients.

yOther includes one with malfitting dentures, one snoring, one chok-

ing, one cough, one dyspnea with supine positioning, and one na-

sal regurgitation.



4/7

ganglioma had distant metastasis to the spine

8 months after primary excision, which was re-

sected; the patient was disease-free at the 3-year

follow-up. Radiation therapy was additionally

given for the carcinoma ex pleomorphic, muco-

epidermoid carcinoma, and malignant myoepithe-

lial carcinomas and was refused by the patients

with low-grade adenocarcinoma and metastatic

thyroid carcinoma.

Surgical outcomes were divided into sequelae

that were expected from the operation (ie, from

planned cranial nerve sacrifice) and unexpected

outcomes (ie, true complications). Cranial nerve

sacrifice secondary to tumor involvement resulted

in numerous neurogenic sequelae (Table 5).Patients with paragangliomas and malignant

lesions were particularly at risk. Silastic medial-

ization with or without arytenoid adduction for

vocal fold paralysis, palatal adhesion for glosso-

pharyngeal paralysis, and facial nerve grafts

and eyelid procedures were used to assist pa-

tient rehabilitation.

Unexpected complications also resulted from

PPS surgery (Table 6). Despite cranial nerve

preservation, palatal weakness occurred postop-

eratively in four patients, three of whom were

asymptomatic. Two patients had iatrogenic vo-

cal cord weakness but recovered function at

2 months. Six patients had facial nerve paresis

despite having a transcervical transparotid ap-

proach with identification of the facial nerve, and

all had resolution by 9 months. First bite

syndrome, pain associated with the first bite of

food, and trismus were the most common non-

neurogenic complications.

DISCUSSION

Masses of diverse origins present in the PPS and

are treated by head and neck surgeons. Obtaining

adequate exposure amid the complex anatomy

can be challenging. However, careful preoperative

Table 4. Surgical approaches to the parapharyngeal space based on tumor characteristics.

Tumor type

% cases by cervical approach

Transcervical

Transcervical

transmastoid

Transcervical

transparotid

Transcervical

transoral Mandibulotomy

Salivary 62.5 0 28.1 9.4 3.1

Neurogenic* 94.4 5.6 0 0 0

Vasculary 41.7 56.3 2.1 0 0

Malignant 55.6 0 33.3 0 5.6Abutting skull basez 55.3 17.0 19.1 4.3 2.1

Involving skull base 12.2 73.0 2.7 0 0

OtherO 53.8 7.7 38.5 0 0

Note. Percentages may not equal 100% because of missing data.

*Neurogenic includes all neural lesions except vagal paragangliomas.

yVascular includes vagal paragangliomas and other vascular lesions.

zAbutting skull base includes lesions adjacent to but not involving the skull base.Involving skull base includes lesions eroding the skull base or with jugular foramen or intracranial extension.

OOther includes one lipoma, one cystic hygroma, one parotid cutaneous fistula, one rhabdomyoma, one encephalocele, one fibrous tumor, one oncocytic

cyst, two branchial cleft cysts, one epidermoid cyst, one Thorotrast granuloma, and one meningioma.

Table 5. Expected sequelae.

Sequelae No. patients (%)*

Vocal cord paralysis 43 (29.3)


Shoulder weakness 18 (12.2)

Tongue weakness 15 (9.0)

Facial nerve weakness 13 (7.8)

Horners syndrome 12 (7.2)

Gastrointestinaly 3 (2.0)

*Percent is calculated from all 147 surgical patients.

yGastrointestinal is persistent nausea/vomiting, gastroesophageal reflux

requiring Nissen operation.

Table 6. Unexpected complications.

Complication No. patients (%)*

First bite syndromey 18 (12.2)

Facial nerve weakness 6 (4.1)Trismus/temporomandibular joint pain 6 (4.1)


Cerebrospinal fluid lead 3 (2.0)

Pneumonia 2 (1.2)

Vocal cord paralysis 2 (1.2)

Seroma/hematoma 2 (1.2)

Orocutaneous fistula 1 (0.7)

Endocarditis 1 (0.7)

*Percent is calculated from all 147 surgical patients.

yFirst bite syndrome is pain associated with first bite of food.



5/7

assessment and application of meticulous surgical

techniques allow the oncologically sound removal

of PPS tumors. The transcervical approach, which

can be extended and/or combined with other tech-

niques, serves as the main method.

Benign tumors predominate among PPS le-

sions. Compared with other reports of 80% benign

and 20% malignant tumors, we found 87.3% and12.7%, respectively.3,4 Patients may harbor these

masses for long periods before symptoms arise.

Tumors often grow to at least 2.5 to 3 cm before

they are detected.6 Almost one fourth of our

patients reported no symptoms (Table 2). Carrau

et al4 found that 20% of masses were also found

incidentally. By comparison, Hughes et al3 found

that 10.4% of benign lesions had cranial neuro-

pathies at presentation. Similarly, 22 (15.2%) of

our 145 patients with benign masses had preop-

erative cranial nerve deficits.

Given the benign nature of most PPS masses

and the potential for preoperative cranial nervedeficits, the surgeon must carefully consider the

treatment. Because of their asymptomatic tumors

and age, 11 patients (6.6%) had their vagal

paragangliomas observed. Because paraganglio-

mas usually grow very slowly, observation may be

the first line of treatment in elderly patients,

particularly those with minimal symptoms and

comorbidities that increase their anesthetic risk.

Furthermore, the characteristic salt-and-pepper

appearance is sufficient for diagnosis, and obser-

vation is recommended in these patients.7 De-

pending on the overall health picture, radiationtherapy may also be considered in elderly pa-

tients with symptomatic tumors. Furthermore,

delaying resection for patients with partial

cranial nerve paralysis, allowing complete paral-

ysis to occur, is an option. Patients often com-

pensate better to a gradual loss of cranial nerve

function compared with an acute loss at sur-

gery, facilitating the recovery from cranial nerve

deficits.8 Patients with preexisting contralateral

palsies of cranial nerves 10 or 12 also require

careful consideration. Eight patients (4.8%) had

preexisting cranial nerve palsies from prior para-

gangliomas and underwent observation of their

current vagal paraganglioma. Bilateral deficits

could severely impair the airway and swallowing

function. To avoid bilateral cranial nerve defi-

cits, these patients may be observed until tumor

growth is determined. External beam radiation

or, for lesions smaller than 3 cm, stereotactic

radiation, can then be used for treatment.7 Pa-

tients must be adequately counseled about all

treatment options and the potential expected

outcomes, especially those pertaining to cranial

nerve sacrifice.

Several surgical approaches have been de-

signed to facilitate removing PPS tumors. Con-

sistent with other reports, the transcervical

approach was our most common technique.4,5,9

However, concern about the exposure to themedial and superior aspects of the PPS and

control of vascular lesions has been raised.10

A midline mandibulotomy has been recom-

mended to increase exposure for lesions in the

superior PPS, tumors larger than 8 cm, those

encasing the internal carotid artery, and ma-

lignant tumors invading the skull base and

vertebral bodies.2,10 Because of the morbidity

associated with midline mandibulotomy, alterna-

tive mandibulotomy techniques have been de-

signed to enhance superior PPS exposure.11,12

Although we did use a mandibulotomy for two

malignant salivary tumors and one vagal para-ganglioma with internal carotid involvement, a

mandibulotomy was not required for the 76

vascular lesions or the 69 tumors extending

superiorly to the skull base. Overall, our man-

dibulotomy rate of 2.0% is lower than other rates

of 6.4% to 20.5%.2,3,5,13 Although differences in

tumor characteristics could explain our low

mandibulotomy rate, the transcervical approach

with its extensions is a versatile technique.

Through a cervical crease or facelift/hairline

incision, many PPS masses, regardless of location

relative to the styloid process, can be excisedtranscervically. Not only were most salivary and

neurogenic tumors removed, but lesions abutting

the skull base were three times more commonly

extirpated transcervically alone than by any

other method (p = .004, chi-square; Table 4).

Specific maneuvers were used to increase the

exposure provided by the transcervical approach.

Dividing the stylomandibular ligament and ante-

rior dislocation of the mandible can increase

exposure by 50%.4 Subsequently, the styloid

process, styloid musculature, and posterior belly

of the digastric can be removed, further enhanc-

ing exposure. Tumors adjacent to the skull base

more commonly required these extensions than

non skull base tumors to move the mandible

away from the operative field (p = .008, chi-

square). As the PPS is further exposed, the 9th

to 12th cranial nerves, internal jugular vein,

and internal carotid artery are more easily

identified and traced to the skull base, thereby

avoiding a mandibulotomy.



6/7

Almost twice as many patients with vascular

tumors, primarily vagal paragangliomas, had a

level II lymphadenectomy compared with those

with nonvascular masses (p V .001, chi-square).

Because metastatic disease is the primary de-

terminant of a malignant paraganglioma, lymph

nodes should be sampled for metastatic spread

when managing paragangliomas.

14

In addition,clearing level II isolates the great vessels and

its branches, which can then be traced to the

skull base, facilitating vascular control.

However, situations arise that necessitate

more advanced techniques. Through a postau-

ricular C incision, the transcervicaltransmas-

toid technique allows increased access to the skull

base. This lateral approach to the skull base was

more commonly used for vascular tumors eroding

the skull base and for tumors extending intra-

cranially through the jugular foramen, mainly

vagal paragangliomas (p V 0.001, chi-square;

Table 4). Proximal and distal control of thejugular bulb and internal carotid artery are

obtained by systematically approaching the skull

base from the neck and mastoid. A mastoidectomy

exposes the labyrinth, facial nerve, sigmoid sinus,

jugular bulb, internal carotid artery, and dural

surfaces involved by the tumor.7 The internal

jugular vein is ligated in the neck and mobilized

while the sigmoid sinus is controlled at the skull

base. A bloodless field reduces the chances of

traumatizing the cranial nerves. Removing the

jugular bulb and potentially mobilizing the facial

nerve further exposes the jugular foramen. Moreintracranial extension may require identifying

the internal carotid artery deep in the glenoid

fossa. The artery may be traced through the

carotid canal and even into the middle cranial

fossa.7 Hence, neurovascular control is obtained

inferiorly in the neck and progresses superiorly as

dictated by the tumors anatomy.

Potential injury to the facial nerve from

tumors that are adherent to the deep lobe of the

parotid may require the use of a combined

transcervical transparotid approach. Although

only one third of salivary tumors required this

combined dissection, they were three times more

likely to require this type of excision compared

with other tumors (pV 0.001, chi-square; Table 4).

In other series, the transcervical transparotid

approach was the primary method and has been

suggested for all deep lobe and many extraparotid

salivary tumors.2,10 Yet, in our study, no patients

undergoing a transcervical excision alone for

salivary lesions had facial nerve sequelae. Sim-

ilarly, Malone et al5 found no facial nerve injuries

in their transcervical resections. However, if

resection may injure the facial nerve, the surgeon

should be prepared to convert to a transparotid

dissection to identify and preserve this structure.

Rarely, the transoraltranscervical technique

is necessary. Transoral excisions alone have been

criticized for their increased risk of bleeding,tumor spillage, and increased recurrence.15,16

Yet, prior transoral biopsy may require a com-

bined technique, secondary to the tumor scarring

to the oral mucosa.10 Because of prior biopsies

and transoral attempts at removal, three patients

had this combined approach to increase exposure

and allow excision. No recurrences from the

combined transcervical transoral approach oc-

curred in our study. In each case, vascular control

was obtained from the neck. Blind dissection

transorally places the internal carotid artery

at risk.

With these procedures, only five patients(3.4%) had a recurrence or persistence of disease,

three presenting to our institution with a recur-

rent tumor. Eleven other malignant tumors and

12 benign lesions that presented as recurrent

masses were managed without further recur-

rence. Although follow-up was missing in 10.9%,

the transcervical, transcervical transmastoid,

and transcervical transparotid approaches serve

as major techniques for the oncologic removal of

PPS masses.

Besides proper surgical technique, timely

rehabilitation can further maximize patient out-come and should begin with preoperative coun-

seling. One third of our patients had expected

neurologic sequelae, some more than one, because

of the necessity of cranial nerve sacrifice for

tumor extirpation. Hence, a significant portion

of patients will require counseling about and po-

tentially rehabilitation for potential postopera-

tive deficits. In other reports, these occurrences

with other unexpected complications have been

discussed.2,35 An important distinction must be

drawn between sequelae expected to occur from

cranial nerve deficits that result from planned

surgical sacrifice and outcomes not expected to

occur (ie, true complications). By recognizing ex-

pected functional deficits, patients can be pre-

pared for them preoperatively.

Patients with neurogenic tumors, particularly

vagal paragangliomas, are particularly at risk for

having postoperative sequelae.2,4 In addition, four

of our patients with malignancies required cra-

nial nerve sacrifice. Thus, these patients should



7/7

be prepared for expected functional deficits.

Perioperative speech and swallowing therapy

can be instituted to facilitate the recovery of

these patients. Injection laryngoplasty or silastic

medialization can improve the glottal incompe-

tence, speech, and swallowing dysfunction in

patients with vocal cord paralysis.1719 Unilat-

eral palatal adhesion also improves the velopha-ryngeal incompetence and nasal regurgitation in

patients with glossopharyngeal nerve sacrifice.20

When determining appropriate treatment, the

surgeon must consider not only the preoperative

and operative assessment but also the postoper-

ative care required.

Some points regarding the study design are rel-

evant. With our institution being a referral cen-

ter, some patients may have elected to be followed

by local physicians, limiting our length of follow-

up. Hence, our ability to discuss tumor recurrence

and long-term outcome is reduced. Because of its

retrospective nature, certain aspects of the datamay have been missing. If operative dictation did

not include each step, misclassification bias could

have occurred when determining which proce-

dures were used. Nevertheless, our study provides

insights about surgical approaches to PPS masses.

CONCLUSIONS

Because of their complexity, PPS masses are

challenging lesions to treat. Patient comorbidities

and preoperative functional deficits, as well as

the tumor type, pathology, histology, and anat-

omy, influence the management. Tumor location,prestyloid or poststyloid, by itself does not seem

to influence the type of resection. Tumor vascu-

larity, skull base involvement, and potential risk

to the facial nerve are key factors. The trans-

cervical approach alone allows the removal of

tumors abutting the skull base, most vascular,

salivary, and neurogenic PPS tumors. In addition,

the transcervicaltransmastoid technique enhan-

ces exposure for vascular tumors with jugular

foramen or intracranial involvement. Internal

carotid artery involvement may necessitate a

mandibulotomy. Last, PPS masses intimatelyinvolved with the facial nerve should be removed

in a transcervicaltransparotid fashion. The

surgeon must counsel the patient preoperatively

and balance the techniques required for resec-

tion with the potential adverse results from the

surgical procedure. With its modifications, the

transcervical approach allows the graduated,

safe oncologic removal of most PPS lesions, in-

cluding vascular tumors and those with skull

base extension.

REFERENCES

1.

Stell PM, Mansfield AO, Stoney PJ. Surgical approaches totumors of the parapharyngeal space. Am J Otolaryngol1985;6:9297.

2. Olsen KD. Tumors and surgery of the parapharyngealspace. Laryngoscope 1994;104:128.

3. Hughes KV, Olsen KD, McCaffrey TV. Parapharyngealspace neoplasms. Head Neck 1995;17:124130.

4. Carrau RL, Meyers EN, Johnson JT. Management oftumors arising in the parapharyngeal space. Laryngoscope1990;100:583589.

5. Malone JP, Agrawal A, Schuller DE. Safety and efficacyof transcervical resection of parapharyngeal space neo-plasms. Ann Otol Rhinol Laryngol 2001;110:10931098.

6. Som PM, Biller HF, Lawson W, Sacher M, Lanzieri ZF.Parapharyngeal space masses: an updated protocol basedupon 104 cases. Radiology 1984;153:149156.

7. Netterville JL, Jackson CG, Miller FR, Wanamaker JR,

Glasscock ME. Vagal paraganglioma: a review of 46patients treated during a 20 year period. Arch OtolaryngolHead Neck Surg 1998;124:1133 1140.

8. Eisele DE, Netterville JL, Hoffman HT, Gantz BJ. Para-pharyngeal space masses. Head Neck 1999;21:154159.

9. Pensak ML, Gluckman JL, Shumrick KA. Parapharyngealspace tumors: an algorithm for evaluation and manage-ment. Laryngoscope 1994;104:11701173.

10. Som PM, Biller HF, Lawson W. Tumors of the para-pharyngeal space: preoperative evaluation, diagnosis andsurgical approaches. Ann Otol Rhinol Laryngol 1981;Suppl 90:315.

11. Teng MS, Genden EM, Buchbinder D, Urken ML. Sub-cutaneous mandibulotomy: a new surgical access for largetumors of the parapharyngeal space. Laryngoscope 2003;113:18931897.

12.

Smith GI, Brennan PA, Webb AA, Ilankovan V. Verticalramus osteotomy combined with a parasymphyseal man-dibulotomy for improved access to the parapharyngealspace. Head Neck 2003;25:1000 1003.

13. Miller FR, Wanamaker JR, Lavertu P, Wood BG. Mag-netic resonance imaging and the management of para-pharyngeal space tumors. Head Neck 1996;18:67 77.

14. Jackson CG. Neurotologic skull base surgery for glomustumors: diagnosis for treatment planning and treatmentoptions. Laryngoscope 1993;103(suppl 6):1722.

15. Ward PH, Jenkins HA, Hanafee WN. Diagnosis andtreatment of carotid body tumors. Ann Otol RhinolLaryngol 1978;87:614620.

16. Heeneman H, Maran A. Parapharyngeal space tumors.Clin Otolaryngol 1979;4:57 66.

17. Netterville JL, Jackson CG, Civantos F. Thyroplasty inthe functional rehabilitation of neurotologic skull base

surgery patients. Am J Otol 1993;14:460464.18. Netterville JL, Civantos FJ, Jackson CG. Rehabilitation of

cranial nerve deficits after skull base surgery. Laryngo-scope 1993;102:45 54.

19. Netterville JL, Stone RE, Luken EJ, Civantos FJ, OssoffRH. Silastic medialization and arytenoids adduction: areview of 116 procedures: the Vanderbilt experience. AnnOtol Rhinol Laryngol 1993;102:41344.

20. Netterville JL, Vrabec JT. Palatal adhesions. ArchOtolaryngol Head Neck Surg 1994;120:218221.


Documents

HN 08-2005 Surgical Management of Parapharyngeal