The changing surgical management of juvenile nasopharyngeal angiofibroma

Eur Arch Otorhinolaryngol (2011) 268:599–607

HEAD AND NECK

The changing surgical management of juvenile nasopharyngeal angioWbroma

Suvi Renkonen · Jaana Hagström · Jyrki Vuola · Mika Niemelä · Matti Porras · Sanna-Maria Kivivuori · Ilmo Leivo · Antti A. Mäkitie

Received: 19 April 2010 / Accepted: 2 September 2010 / Published online: 17 September 2010© Springer-Verlag 2010

Abstract The management of juvenile nasopharyngealangioWbroma (JNA) has changed during the last decadesbut it still continues to be a challenge for the multidisciplin-ary head and neck surgical team. The aim of this study wasto review the used treatment approach and outcome of JNAin a single institution series of 27 patients diagnosed andtreated during the years 1970–2009. All patients were male,with the median age of 17 years (range 11–33 years).Surgery was used as the primary treatment in every case.Surgical approaches varied, transpalatal approach (N = 14)being the most common approach used in this series.During the last decade various other techniques were

applied, including endoscopic (N = 3) resection. Twopatients were additionally treated with antiangiogenicagents and one patient with stereotactic radiotherapy. Theprimary recurrence rate was 37% and it seemed to correlatewith vascular density of tumour and the surgical approachused. We suggest that the management of JNA should beplanned by an experienced head and neck surgeon, as partof a multidisciplinary team, preferably in a tertiary referralsetting, and the recent development of the available thera-pies should be taken into account to minimise the risk ofrecurrence.

Keywords Head and neck · Neoplasm · Benign · Surgery · Antiangiogenic · Vascular density · Radiotherapy

Introduction

Juvenile nasopharyngeal angioWbroma (JNA) is a rare, his-tologically benign, mesenchymal, vascular and highly cel-lular tumour. It originates from the superior margin of thesphenopalatine foramen, which is also a route for the sphe-nopalatine artery branching from the internal maxillaryartery [1, 2]. JNA accounts for 0.5% of all head and neckneoplasms and it is considered to be the most commonbenign neoplasm of the nasopharynx [3–6]. Despite itsbenign nature its growth pattern is locally destructive and itcan cause bone remodelling. Due to its propensity to boneerosion, the tumour has the potential of life-threateningcomplications, such as fatal epistaxis, intracranial extensionand intraoperative massive haemorrhage [7–9].

The extremely rich vascularisation together with thecomplex anatomical structures on the cranium (e.g. cranialnerves, carotid arteries, cavernous sinus and dura can beinWltrated) [10] and the aggressive growth pattern of the

S. Renkonen · A. A. Mäkitie (&)Department of Otorhinolaryngology, Head and Neck Surgery, Helsinki University Central Hospital and University of Helsinki, P.O. Box 220, 00029 Helsinki, Finlande-mail: [email protected]

J. Hagström · I. LeivoDepartment of Pathology, Helsinki University Central Hospital, Helsinki, Finland

J. VuolaDepartment of Plastic Surgery, Helsinki University Central Hospital, Helsinki, Finland

M. NiemeläDepartment of Neurosurgery, Helsinki University Central Hospital, Helsinki, Finland

M. PorrasDepartment of Neuroradiology, Helsinki University Central Hospital, Helsinki, Finland

S.-M. KivivuoriDivision of Hem/Onc and SCT, Hospital for Children and Adolescents, Helsinki University Central Hospital, Helsinki, Finland

123

600 Eur Arch Otorhinolaryngol (2011) 268:599–607

tumour makes its treatment a challenge. It is agreed thatsurgical excision is the treatment of choice for JNA [2, 5, 9,11]. Traditionally, surgical approaches used to resect JNA,have been transpalatal, transmaxillar (lateral rhinotomy ormidfacial degloving), Le Fort 1 osteotomy, and infratempo-ral fossa craniotomy approaches [2, 9, 12, 13]. Preoperativeselective embolisation of feeding arteries has decreasedintraoperative haemorrhage and made it possible to resecteven the largest tumours. However, recurrence rates as highas 50% (ranging from 6 to 50%) have been reported[14, 15]. There are numerous reports about endoscopicresections, started in 1993 in Barcelona and in 1994 in Graz[4, 11, 16–19]. Due to its minimal invasiveness, low morbidityrate, and typically low rates of recurrence this approach isgaining popularity. During the last decade, the increasedunderstanding of angiogenesis has allowed the develop-ment of new therapeutic substances. Since strong vascular-ity is a common feature of JNAs, it has been suggested thatantiangiogenic therapies ought to be considered in the man-agement of selected cases [20].

The purpose of this study was to investigate the usedtreatment approach and the outcome of JNA in a singleinstitution series of all patients diagnosed and treated at ourtertiary care academic teaching hospital during the past fourdecades (1970–2009).

Patients and methods

Retrospective clinicopathological data of all patients whowere diagnosed for a histologically veriWed JNA during the40-year period between January 1, 1970 and December 31,2009 at the Helsinki University Central Hospital (HUCH),Helsinki, Finland were reviewed. The health-care districtincluded in this study includes approximately 1.5 millioninhabitants which represents almost one-third of the wholecountry. The majority of the cases were operated at theDepartment of Otolaryngology, Head and Neck Surgeryand a few cases at the Department of Neurosurgery. Hospi-tal surgery and discharge registries as well as the databaseof the Department of Pathology, HUCH, were used to iden-tify the patients. A total of 27 patients (all male, mean age17 years; range 11–33 years) were included. The hospitalrecords were reviewed and details were collected on thepatients’ age, sex, presenting signs and symptoms, theduration of symptoms, date of diagnosis, tumour site, imag-ing, preoperative embolisation, histopathology, stage of thetumour, feeding arteries, treatment modality and surgicalapproach, rate of recurrence, date of last follow-up visit,and status at last follow-up visit. No patients were lost tofollow-up. All patients had a minimum 7-month follow-up.

In addition, to investigate the importance of histopathol-ogical correlates of the tumour on clinical features and

tendency to recur, we examined the formalin Wxed and par-aYn embedded tumour samples of every patient (N = 27)for cell and vascular density. For vessel density, vesselsmarked by CD31 were counted and the mean counts permm2 were calculated.

For categorial, non-ordered variables, cross-tabulationswere analysed using the Chi square test or when the Chisquare test was not valid, Fisher’s exact test was used.We used SPSS version 15.0 software (SPSS, IL, USA) forall statistical analyses.

The Research Ethics Board of HUCH reviewed andapproved the study.

Results

The clinical data on 27 patients are presented in Table 1. Alltumours were located in the nasopharyngeal area and hadextension to the adjacent structures depending on tumoursize and growth pattern. The lesion was located on the rightside in 15 (55.6%) patients and on the left side in 12 (44.4%)patients. Seven (25.9%) cases showed extension to thepterygopalatine fossa, 13 (48.1%) cases to the sphenoidsinuses, eight (29.6%) cases to the ethmoid sinuses, Wve(18.5%) cases to the maxillary sinuses and two (7.4%) casesto the orbital, infratemporal and middle cranial fossae.Tumours were classiWed according to their location andgrowth pattern using the Andrews staging (Table 2) (27).Eight (29.6%) tumours were limited to the nasal cavity with-out bone destruction, and therefore represented Stage I.

For initial diagnosis, angiography was performed in 24(89%) patients. In 16 cases one primary feeding artery wasidentiWed and in all these cases, the blood supply originatedfrom the internal maxillary artery. Additional blood supplywas provided by ascending pharyngeal artery in Wve(31.3%) cases and by facial artery in one (6.3%) case.Selective preoperative embolisation was performed in 15(62.5%) cases out of the 24 patients with angiography.

Surgery was used as the primary treatment in all cases,as shown in Table 1. The various types of surgical approachincluded transpalatal, transmaxillary (either sublabialdegloving or lateral rhinotomy approach), endoscopic, andcraniotomy. The surgical approach used in each case wastypically chosen depending on the size and extension of thetumour. Three (11.1%) patients were treated with a combi-nation of two surgical approaches. The Wrst endoscopicresection was performed in 1998 and the three cases weretreated endoscopically (Stages I–II). The microscope forsurgery is routinely used and imaging guided navigationhas been used in all JNA operative procedures since 2005 atour institution.

There was no surgery-related mortality and no majorcomplications. The minor complications included oroantral

123

Eur Arch Otorhinolaryngol (2011) 268:599–607 601

Wstula (N = 1), secretory otitis media (N = 1), infraorbitalnerve damage (N = 1) or lacrimal duct dysfunction (N = 1).

JNA recurrence has to be considered, when clinical andradiological tumour suspicion correlates with clinicalsymptoms and growth during follow-up [21]. In the present

series 10 (37%) patients had persistent or recurrent diseasefollowing the primary operative procedure and theselesions were discovered on average 10.5 months (range 3–19 months) following the Wrst operative procedure. In everycase the recurrence was removed surgically. Except for twopatients, the re-surgery was done by combining two surgi-cal approaches. None of the recurrences was treated endo-scopically. In six patients (22.2%), despite a secondsurgical procedure, a recurrence was discovered after8 months (average) (range 5–12 months). Two patients(7.4%) underwent surgery four times. Following 108 and177 months of follow-up, respectively, both of thesepatients were free of disease. A correlation was foundbetween the number of recurrences and the surgicalapproach, as the tumours removed transpalatally had higherrisk of multiple recurrences when compared to the otherapproaches (Chi square, P = 0.007; Fisher’s exact test,P = 0.004). Of the total ten patients with a recurrence, at thelast follow-up visit, 50% (N = 5) had no evidence of diseaseand 50% (N = 5) were alive with disease and under follow-up. Following the three operative procedures, two (7.4%)patients in this series received antiangiogenic (celecoxib,thalidomide or etoposide) therapy; one patient receivedconcomitant celecoxib-therapy; the other patient was addi-tionally treated by stereotactic radiotherapy with a totaldose of 50.4 Gy.

Patients with a recurrence were between 10 and 17 yearsold at the time of diagnosis, except for one patient who was24 years old. All patients with more than one recurrencewere below 15 at the time of diagnosis. Of the nine patientsabove 17 years at the time of diagnosis, only one had arecurrence.

The median follow-up time was 106 months (range 7–360 months). At the last follow-up visit all patientsremained either free of disease (N = 20, 74.1%) or werealive with minimal signs of persistent disease (N = 6,22.2%), but without evidence of signiWcant further growthor malignant transformation.

The histological vascular architecture of the samplesshowed vessels of diVerent numbers, size and shape in aWbrous stromal tissue as detected with the endothelialmarker CD31 (Fig. 1). The numbers of blood vesselsranged between 160 and 6,200 vessels per mm2 (median2,340). We found a signiWcant correlation (Fisher’s exacttest, P = 0.040) between the recurrence rate and the vascu-lar density of the tumour.

Discussion

This study was undertaken to evaluate the used treatmentmodalities and the obtained outcome of JNA at a single ter-tiary care academic institution. A retrospective review was

Table 1 Clinical characteristics of 27 JNA patients

Number of Patients 27

Sex

Male 27 (100%)

Age (years) 17 (11–33)

Symptoms

Nasal obstruction 13 (48%)

Epistaxis 8 (30%)

Nasal discharge 3 (11%)

Duration of symptoms (months) 7 (2–24)

Stage of the tumour

I 8 (30%)

II 14 (52%)

IIIa 2 (7%)

IIIb/IV 3 (11%)

Imaging

TT 17 (63%)

MRI 15 (56%)

Embolisation 15 (56%)

Treatment

Surgery

Transpalatal 14 (52%)

Transmaxillar 5 (19%)

Endoscopic 3 (11%)

Craniotomy 2 (7%)

Combination 3 (11%)

Radiotherapy (50.4 Gy) 1 (3%)

Antiangiogenic therapy 2 (7%)

Recurrence 10 (37%)

Follow-up (months) 106 (7–360)

Status at the last follow-up visit

No evidence of disease 21 (78%)

Alive with disease 6 (22%)

Table 2 Andrews staging system for juvenile nasopharyngeal angio-Wbroma (1989)

I Tumour limited to the nasal cavity and nasopharynx

II Tumour invading the pterygopalatine fossa or maxillary, ethmoidal and sphenoid sinuses; with bone destruction

III Tumour invading the infratemporal fossa or orbital region: (a) without intracranial involvement and (b) with extradural intracranial involvement

IV Tumour with intradural intracranial involvement: (a) without or (b) with inWltration of cavernous sinus, pituitary fossa, or optic chiasma

123

602 Eur Arch Otorhinolaryngol (2011) 268:599–607

done on all patients with a JNA diagnosed and treated at theHelsinki University Central Hospital in Finland from 1970to 2009. A change was observed in the surgical approachpreferred in the management of these tumours during thefour decades. New and seemingly promising treatmentmodalities were used.

JNA is a rare diagnosis and the patient series in the liter-ature are relatively small (median 30, mean 33, range 12–95) [2, 5–7, 9, 15, 18, 19, 22]. In the present retrospective40-year single institution (currently covering an area of 1.5million people) series there were only 27 patients, althoughthe treatment of these tumours in Finland is centralised totertiary care centres.

The diagnosis of JNA is based on a precise clinical his-tory and examination of the patient, and imaging (CT orMRI). Tissue biopsies should be avoided due to the highlyvascular nature of the tumour. The most common present-ing symptom is persistent nasal obstruction with repetitiveepistaxis. Angiography is used to deWne the feeding arteriesof the tumour and embolisation of the vessels can be per-formed at the same time. In the present series, embolisa-tions were used consistently since the Wrst procedures wereperformed in the early 1990s.

Various staging systems according to the location andextent of JNA have been proposed. In this study, we usedAndrews staging (Table 2), according to which type Itumours are limited to the nasopharynx and nasal cavity,and bone destruction is negligible or limited to the spheno-palatine foramen. Other classiWcation systems are theChandler and Radowski classiWcations, which have nocommits on bone destruction. It is noteworthy that in ourseries, most recurrences occured in the Stage II tumours, as6 out of 14 (43%) Stage II tumours recurred and were sub-jected to re-surgeries. Understaging is one possible expla-nation for this result and obviously these lesions should bemanaged with more radical surgery.

It is generally agreed that surgery is the treatment ofchoice for JNA [5, 9]. Other methods that have been usedas primary treatment include hormone therapy, chemother-apy and radiotherapy [3, 23, 24]. In the present series, nopatients received these as Wrst line treatment. Surgery aimsat a complete and safe resection of tumour, with minimalmorbidity and loss of blood. Transpalatal, transmaxillar(lateral rhinotomy or midfacial degloving), Le Fort 1 oste-otomy and infratemporal fossa approaches [2, 9, 12, 13]used to be the traditional surgical methods commonly usedto remove JNA. Recent reports recommend endoscopic sur-gery [4, 11, 16–19]. However, the most suitable choice foreach patient remains under discussion.

In the present series, transpalatal approach was mostcommonly (51.6%) used for primary surgery. It was espe-cially common (54.5%) in the treatment of Stage I–IItumours. Furthermore, all patients during the decades1970–1989 were treated using either transpalatal (62.5%),combination of transpalatal and transmaxillar (25.0%) orsublabial approach (12.5%). During the last decade (2000–2009) only 41.7% had exclusively transpalatal surgery,while the rest were treated using other approaches. Trans-palatal approach is advantageous when removing tumourslimited to nasopharynx, nasal cavity and sphenoid sinuseswhile no facial incisions are needed. However, it has obvi-ous limitations. If lateral tumour growth to pterygopalatineand infratemporal fossae occurs, [25] it leads to higher riskof recurrence. In accordance with earlier reports, in ourseries the transpalatal approach was associated with a sig-niWcant risk of persistent disease and with more than onere-surgery. The most common complication among patientstreated with transpalatal approach is palatal Wstula, whichusually tends to heal spontaneously [2, 25]. Poor exposurefor large tumours and the reported elevated rates of recur-rence by us and others [8], make this approach less favour-able in our armamentarium for JNA surgery. Nevertheless,it can be combined with other approaches in selected cases(Table 3).

For the extensive tumours, transmaxillar approaches,including midfacial degloving and lateral rhinotomy, aresuitable. These allow surgical exposure also for tumoursexpanding into pterygopalatine and infratemporal fossae,paranasal sinuses, orbit, ethmoid sinus and medial part ofthe cavernous sinus. The lateral rhinotomy approach oVersexposure for removal of an extensive JNA. However, theuse of midfacial degloving has gained popularity, as itavoids facial scars while maintaining excellent surgicalexposure [8, 13, 14]. In addition, the transoral Le Fort Iosteotomy approach provides a wide exposure and goodvisibility of the tumour margins without visible skin scars.This technique is thought to be suitable for tumours invad-ing the skull base and orbit. It is noteworthy that non-erupted teeth may be damaged and rupture of the palatal

Fig. 1 CD31 staining showing slit vessels (£20)

123

Eur Arch Otorhinolaryngol (2011) 268:599–607 603

vessels during down-fracturing of the maxilla can causeexcessive blood loss [26]. Thus, the preoperative embolisa-tion is essential [8, 27–29]. In the management of the larg-est tumours with intracranial extension, uncontrolledhaemorrhage, neurologic deWcits, subtotal resections andrecurrences have been reported [3, 24, 29, 30]. To resectthese tumours, various craniotomies, both infratemporalapproach and the combination of anterior and lateralapproaches have been used [31]. The infratemporalapproach may result in conductive hearing loss on the oper-ated side, and in the sacriWce of at least the mandibulardivision of the trigeminal nerve [31, 32]. Furthermore, allexternal approaches are invasive, and thus a slight risk offacial growth retardation is present in this adolescentpatient population.

Over the past 10 years, management of JNA has under-gone a signiWcant evolution. Advances in technology andimprovements in endonasal techniques have enabled theuse of endoscopic surgery. Initially, endoscopic approachwas used exclusively for tumours limited to the nasophar-ynx, the nasal cavity, the ethmoid and sphenoid sinuses andslight extensions to the pterygopalatine fossa (Fig. 2).Recently, selected cases of tumours expanding to thepterygopalatal fossa, infratemporal fossa and even withlimited intracranial involvement have been managed endo-scopically [19, 33, 34]. However, the consensus is thatendoscopic approach has limitations and is not recom-mended for tumours with extensive intracranial extensions,or extensions laterally to the cavernous sinus [11, 18]. Theadvantages of endoscopic approach are enhanced surgicalvisualisation with magniWed and multiangled views, whichis thought to prevent overlooking of tumour remnants,along with the avoidance of facial scars and bone disruption[11]. Endoscopic surgery is also associated with lower mor-bidity, low rate of complications, and shorter duration ofhospitalisation [16, 17]. Many authors have described theimportance of exploring the basis of the sphenoid to

remove all tumours invading the cancellous bone at theskull base [14, 15, 28], which might be better achieved withendoscopic resection [14, 15, 28]. When compared with tra-ditional surgical approaches, the risk of recurrence isreported to be lower [35]. In addition, it is possible to com-bine the endoscopic technique with open approach, whichallows delicate inspection of the surgical cavity and enables

Table 3 Operative approach and outcome according to disease stage

Combined = TP + TM or TP + ES

TP transpalatal approach

ES endoscopic approach

TM transmaxillary approach

CR craniotomy

Stage Number of patients

Surgery Recurrence

TP TM ES CR Combined

I 8 5 1 2 0 0 2 (1 £ TP, 1 £ TM)

II 14 7 3 1 0 3 6 (2 £ TP, 2 £ Combined, 1 £ TM, 1 £ ES)

IIIa 2 0 1 0 1 0 1 (1 £ CR)

IIIb/IV 3 2 0 0 1 0 1 (1 £ TP)

Fig. 2 Photograph of an endoscopically removed Stage I JNA (a) andthe preoperative sagittal CT image (b)

123

604 Eur Arch Otorhinolaryngol (2011) 268:599–607

further resection, if necessary. Herman et al. [36] haveshowed that combining endoscopic techniques with tradi-tional surgery decreased the likelihood of tumour recur-rence. The conWnes of endoscopic approach are the loss oftwo-handed dissection, and the weak visibility in case ofbleeding, which is why most of the authors agree, that pre-operative embolisation of the feeding vessels is essentialfor hemostasis and visualisation of the surgical Weld [9, 18].We currently tend to use imaging guided navigation in allJNA operative procedures for documenting and safety rea-sons as has been also suggested by others [4]. Another use-ful dissection system with obvious advantages allowingquick and eVective endoscopic removal of the tumour is theCavitron ultrasonic surgical aspirator. Preoperative use ofintratumoural injections of cyanoacrylate glue has beensuggested to help dissection of JNA [37].

Although surgery is considered to be the treatment ofchoice in most cases of JNA, controversies still existregarding the appropriate management of the mostadvanced tumours, especially those with intracranial exten-sion. In addition to the fact that diYcult tumour locationincreases the risk of postoperative recurrence, also thepotential morbidity related to the intracranial surgery needsto be addressed. Consequently, it is presented that thosetumours with radiographically evidenced intracranialextension, or, respectively, lesions deemed incompletely

resectable, without notable risk of morbidity, could beeVectively cured with external beam radiotherapy (RT)[23, 24, 38]. It has been shown, that radiotherapy and surgeryare evenly eVective in the treatment of JNAs [24]. How-ever, the well known long-term complications of RT, suchas secondary malignancies, endocrine hypofunction andcataract, must be taken into account and compared with therisks of intracranial surgery.

Recurrences are known to be common in the manage-ment of JNA and the majority of them occur in 1 year aftersurgery [7]. Because JNA is a unifocal benign disease,recurrence is often thought to represent residual tumour dueto incomplete resection. In 2007, Tyagi et al. [15] showedthat evident residual tumour was an independent risk factorfor recurrence. However, despite the fact, that tumourrecurrence almost certainly represents tumour persistence,there are reports of incompletely resected tumours that havenot progressed after a notable time of follow-up [15], andspontaneous involution of JNA has been suggested to hap-pen infrequently [39, 40]. Thus, residual disease should notalways be considered as a recurrence, because tumour rem-nants can be found by postoperative imagining in patientswho remain symptom-free [30], and therefore one mustdiVerentiate true new tumour growth from symptomless,non-growing tumour residual (Fig. 3) [10]. Detection ofresidual growth as early as possible is essential for

Fig. 3 Representative images of JNA recurrences. a Coronal CT image of a recurrent JNA with erosion of the skull base. The tumour was removed through sublabial approach. b, c MRI images of recurrent JNAs with no further growth of the tumour under 3–4-year surveillance

123

Eur Arch Otorhinolaryngol (2011) 268:599–607 605

treatment planning. However, follow-up imaging shouldnot be performed earlier than 6 weeks, as an early CT orMRI scan can give a false impression of a residual tumourdue to various postoperative artefacts [15].

The age of the patient seems to predict whether therecurrences will develop or regress spontaneously, astumour remnants may be expected to continue their growthin young patients and to disappear in the elderly [36]. Thecorrelation between the young age and the recurrence ratehas been shown in numerous studies [15, 36, 41], andtumour growth has been considered to be related with thehormonal environment during male puberty [40, 42–44].In the present series it was also noted that only one patientout of nine patients over 17 years old had a recurrence. Wesuppose, that this may reXect the spontaneous decrease ofgrowth potential of these tumours after puberty.

Recurrence rates of on average 32% have been described[14, 15, 36]. These seem to correlate with the extent oftumour growth especially to lateral and superior directionsmaking surgical excision more diYcult. In case of tumourextension to the skull base, when total removal is especiallydiYcult, recurrence rates of even 40–50% have beenreported [36]. In addition to evident residual tumour, otherrisk factors associated with recurrence are e.g. extensionsinto the pterygoid fossa and base of sphenoid bone, intra-cranial extensions medial to the cavernous sinus, feedingvessels from the internal carotid artery and young age [15].Although it is well established that preoperative tumourembolisation reduces intraoperative bleeding and decreasesthe need of blood transfusion, its beneWts still remain con-troversial, as some authors describe that preoperativeembolisation may disrupt the tumour border and hence leadto incomplete resection [11, 28]. In the present series, only2 out of the 10 patients with recurrence were embolised.

In the present series the overall recurrence rate was 37%which is in good accordance with the literature [21]. Recur-rences were discovered on average 10.5 months after theWrst seemingly complete tumour removal. Most recurrences(43%) occurred in the group of Stage II tumours. Other fac-tors that seemed to aVect the possibility of recurrence wereyoung age, as only one out of nine patients over 17 yearshad recurrence, high vascular density of the tumour and thetranspalatal approach used. Patients operated transpalatallyhad the highest possibility (29%) for multiple recurrences.

The growth of JNA is thought to be inXuenced by hor-monal factors, as these tumours almost exclusive occur inadolescent males. Many studies have aimed to demonstratethe role of androgen, estrogen and progesterone and partic-ularly their nuclear receptors, for JNA behaviour, but so farthe concept of hormone receptors remains controversial[43, 45, 46]. Current studies on the proliferation of stromaland vascular components of JNA have supported thehypothesis of the angiogenic growth factors playing a role

in the pathogenesis [20, 46–50]. Angiogenesis is known tobe an essential process during the growth and developmentof metastasis in solid tumours, as a prerequisite of tumourcell proliferation [51]. More importantly, during the lastdecade, increased understanding of angiogenesis hasallowed the development of new therapeutic substances.Particularly, studies revealing the signiWcant role of vascu-lar endothelial growth factor (VEGF) in the regulation ofnormal and pathological angiogenesis have made the block-ade of VEGF signalling a major target for the therapy ofmalignant tumours [51]. A common feature of JNAs is theirstrong vascularity. Brieger et al. [20] showed that VEGF issecreted by JNA, suggesting that it contributes to the strongvascularity of this tumour and more importantly that antian-giogenic therapy ought to be considered. Our Wnding ofvascular density of the tumour correlating with recurrencerate strongly supports the idea of using antiangiogenic ther-apy for JNA. In the present series, two patients with inoper-able tumour remnants were successfully experimentallyadministered antioangiogenic drugs (selecoxib, thalido-mide, etoposide) [52]. Therefore, on a guarded note we rec-ommend that in selective cases as with patients withhistologically veriWed highly vascular tumours in surgicallychallenging locations, antiangiogenic adjuvant therapymight be beneWcial. None of our patients has receivedinhibitors of VEGF, which have been widely used in cancertherapy [53]. In the future, this could possibly be contem-plated also in this patient population, knowing that VEGFis shown to be expressed in JNA.

Conclusions

Despite the technical and surgical development during thepast decades management of JNA is still challenging.Treatment should be planned by an experienced head andneck surgeon, as part of a multidisciplinary team, prefera-bly in a tertiary referral setting. Tumour recurrence is com-mon, but the risk could be minimised by taking intoaccount the recent evolution in surgical techniques and alsoby considering the antiangiogenic therapy in selected cases.

Acknowledgment This study was supported by The HelsinkiUniversity Central Hospital Research Fund, Helsinki, Finland.

ConXict of interest None.

References

1. Schuon R, Brieger J, Heinrich UR, Roth Y, Szyfter W, Mann WJ(2007) Immunohistochemical analysis of growth mechanisms injuvenile nasopharyngeal angioWbroma. Eur Arch Otorhinolaryn-gol 264:389–394

123

606 Eur Arch Otorhinolaryngol (2011) 268:599–607

2. Tosun S, Ozer C, Gerek M, Yetiser S (2006) Surgical approachesfor nasopharyngeal angioWbroma: comparative analysis and cur-rent trends. J Craniofac Surg 17:15–20

3. Antonelli AR, Cappiello J, Di Lorenzo D, Donajo CA, Nicolai P,Orlandini A (1987) Diagnosis, staging, and treatment of juvenilenasopharyngeal angioWbroma (JNA). Laryngoscope 97:1319–1325

4. Enepekides DJ (2004) Recent advances in the treatment of juve-nile angioWbroma. Curr Opin Otolaryngol Head Neck Surg12:495–499

5. Mann WJ, Jecker P, Amedee RG (2004) Juvenile angioWbromas:changing surgical concept over the last 20 years. Laryngoscope114:291–293

6. Paris J, Guelfucci B, Moulin G, Zanaret M, Triglia JM (2001)Diagnosis and treatment of juvenile nasopharyngeal angioWbroma.Eur Arch Otorhinolaryngol 258:120–124

7. Glad H, Vainer B, Buchwald C, Petersen BL, Theilgaard SA,Bonvin P, Lajer C, Jakobsen J (2007) Juvenile nasopharyngealangioWbromas in Denmark 1981–2003: diagnosis, incidence, andtreatment. Acta Otolaryngol 127:292–299

8. Marshall AH, Bradley PJ (2006) Management dilemmas in thetreatment and follow-up of advanced juvenile nasopharyngealangioWbroma. ORL 68:211–216

9. Scholtz AW, Appenroth E, Kammen-Jolly K, Scholtz LU, Thumf-art WF (2001) Juvenile nasopharyngeal angioWbroma: manage-ment and therapy. Laryngoscope 111:681–687

10. Hofmann T, Bernal-Sprekelsen M, Koele W, Reittner P, Klein E,Stammberger H (2005) Endoscopic resection of juvenile angioW-bromas—long term results. Rhinology 43:282–289

11. Andrade NA, Pinto JA, Nobrega Mde O, Aguiar JE, Aguiar TF,Vinhaes ES (2007) Exclusively endoscopic surgery for juvenilenasopharyngeal angioWbroma. Otolaryngol Head Neck Surg137:492–496

12. Midilli R, Karci B, Akyildiz S (2009) Juvenile nasopharyngealangioWbroma: analysis of 42 cases and important aspects of endo-scopic approach. Int J Pediatr Otorhinolaryngol 73:401–408

13. Yiotakis I, Eleftheriadou A, Davilis D, Giotakis E, Ferekidou E,Korres S, Kandiloros D (2008) Juvenile nasopharyngeal angioW-broma stages I and II: a comparative study of surgical approaches.Int J Pediatr Otorhinolaryngol 72:793–800

14. Howard DJ, Lloyd G, Lund V (2001) Recurrence and its avoidancein juvenile angioWbroma. Laryngoscope 111:1509–1511

15. Tyagi I, Syal R, Goyal A (2007) Recurrent and residual juvenileangioWbromas. J Laryngol Otol 121:460–467

16. Borghei P, Baradaranfar MH, Borghei SH and Sokhandon F(2006) Transnasal endoscopic resection of juvenile nasopharyn-geal angioWbroma without preoperative embolization. Ear NoseThroat J 85:740–743, 746

17. Douglas R, Wormald PJ (2006) Endoscopic surgery for juvenilenasopharyngeal angioWbroma: where are the limits? Curr Opin Ot-olaryngol Head Neck Surg 14:1–5

18. Nicolai P, Berlucchi M, Tomenzoli D, Cappiello J, Trimarchi M,Maroldi R, Battaglia G, Antonelli AR (2003) Endoscopic surgeryfor juvenile angioWbroma: when and how. Laryngoscope113:775–782

19. Onerci TM, Yucel OT, Ogretmenoglu O (2003) Endoscopic sur-gery in treatment of juvenile nasopharyngeal angioWbroma. Int JPediatr Otorhinolaryngol 67:1219–1225

20. Brieger J, Wierzbicka M, Sokolov M, Roth Y, Szyfter W, MannWJ (2004) Vessel density, proliferation, and immunolocalizationof vascular endothelial growth factor in juvenile nasopharyngealangioWbromas. Arch Otolaryngol Head Neck Surg 130:727–731

21. Carrillo JF, Maldonado F, Albores O, Ramirez-Ortega MC, Onate-Ocana LF (2008) Juvenile nasopharyngeal angioWbroma: clinicalfactors associated with recurrence, and proposal of a staging sys-tem. J Surg Oncol 98:75–80

22. Mistry RC, Qureshi SS, Gupta S (2005) Juvenile nasopharyngealangioWbroma: a single institution study. Indian J Cancer 42:35–39

23. Lee JT, Chen P, Safa A, Juillard G, Calcaterra TC (2002) The roleof radiation in the treatment of advanced juvenile angioWbroma.Laryngoscope 112:1213–1220

24. Reddy KA, Mendenhall WM, Amdur RJ, Stringer SP, Cassisi NJ(2001) Long-term results of radiation therapy for juvenile naso-pharyngeal angioWbroma. Am J Otolaryngol 22:172–175

25. Dubey SP, Molumi CP (2007) Critical look at the surgical ap-proaches of nasopharyngeal angioWbroma excision and “totalmaxillary swing” as a possible alternative. Ann Otol RhinolLaryngol 116:723–730

26. Belmont JR (1988) The Le Fort I osteotomy approach for nasopha-ryngeal and nasal fossa tumors. Arch Otolaryngol Head Neck Surg114:751–754

27. Davis KR (1987) Embolization of epistaxis and juvenile nasopha-ryngeal angioWbromas. AJR Am J Roentgenol 148:209–218

28. Lloyd G, Howard D, Phelps P, Cheesman A (1999) Juvenile angi-oWbroma: the lessons of 20 years of modern imaging. J LaryngolOtol 113:127–134

29. Browne JD, Jacob SL (2000) Temporal approach for resection ofjuvenile nasopharyngeal angioWbromas. Laryngoscope 110:1287–1293

30. Jones GC, DeSanto LW, Bremer JW, Neel HB III (1986) JuvenileangioWbromas. Behavior and treatment of extensive and residualtumors. Arch Otolaryngol Head Neck Surg 112:1191–1193

31. Andrews JC, Fisch U, Valavanis A, Aeppli U, Makek MS (1989)The surgical management of extensive nasopharyngeal angioWbro-mas with the infratemporal fossa approach. Laryngoscope 99:429–437

32. Zhang M, Garvis W, Linder T, Fisch U (1998) Update on the infra-temporal fossa approaches to nasopharyngeal angioWbroma.Laryngoscope 108:1717–1723

33. Roger G, Tran Ba Huy P, Froehlich P, Van Den Abbeele T, KlossekJM, Serrano E, Garabedian EN, Herman P (2002) Exclusively endo-scopic removal of juvenile nasopharyngeal angioWbroma: trends andlimits. Arch Otolaryngol Head Neck Surg 128:928–935

34. Wormald PJ, Van Hasselt A (2003) Endoscopic removal of juve-nile angioWbromas. Otolaryngol Head Neck Surg 129:684–691

35. Pryor SG, Moore EJ, Kasperbauer JL (2005) Endoscopic versustraditional approaches for excision of juvenile nasopharyngealangioWbroma. Laryngoscope 115:1201–1207

36. Herman P, Lot G, Chapot R, Salvan D, Huy PT (1999) Long-termfollow-up of juvenile nasopharyngeal angioWbromas: analysis ofrecurrences. Laryngoscope 109:140–147

37. Budzynowska K, Pietniczka M, Dowzenko A, Borowska K, Czep-iel W (2008) Safe extirpating of AFJ after preoperative tumorobliteration with tissue adhesive glue. Otolaryngol Pol 62:408–411

38. McAfee WJ, Morris CG, Amdur RJ, Werning JW, MendenhallWM (2006) DeWnitive radiotherapy for juvenile nasopharyngealangioWbroma. Am J Clin Oncol 29:168–170

39. Jacobsson M, Petruson B, Ruth M, Svendsen P (1989) Involutionof juvenile nasopharyngeal angioWbroma with intracranial exten-sion. A case report with computed tomographic assessment. ArchOtolaryngol Head Neck Surg 115:238–239

40. Weprin LS, Siemers PT (1991) Spontaneous regression of juvenilenasopharyngeal angioWbroma. Arch Otolaryngol Head Neck Surg117:796–799

41. Petruson K, Rodriguez-Catarino M, Petruson B, Finizia C (2002)Juvenile nasopharyngeal angioWbroma: long-term results in preop-erative embolized and non-embolized patients. Acta Otolaryngol122:96–100

42. Farag MM, Ghanimah SE, Ragaie A, Saleem TH (1987)Hormonal receptors in juvenile nasopharyngeal angioWbroma.Laryngoscope 97:208–211

123

Eur Arch Otorhinolaryngol (2011) 268:599–607 607

43. Lee DA, Rao BR, Meyer JS, Prioleau PG, Bauer WC (1980) Hor-monal receptor determination in juvenile nasopharyngeal angioW-bromas. Cancer 46:547–551

44. Liang J, Yi Z, Lianq P (2000) The nature of juvenile nasopharyn-geal angioWbroma. Otolaryngol Head Neck Surg 123:475–481

45. Gatalica Z (1998) Immunohistochemical analysis of steroid hor-mone receptors in nasopharyngeal angioWbromas. Cancer Lett127:89–93

46. Saylam G, Yucel OT, Sungur A, Onerci M (2006) Proliferation,angiogenesis and hormonal markers in juvenile nasopharyngealangioWbroma. Int J Pediatr Otorhinolaryngol 70:227–234

47. Nagai MA, Butugan O, Logullo A, Brentani MM (1996) Expres-sion of growth factors, proto-oncogenes, and p53 in nasopharyn-geal angioWbromas. Laryngoscope 106:190–195

48. Ngan BY, Forte V, Campisi P (2008) Molecular angiogenicsignaling in angioWbromas after embolization: implications fortherapy. Arch Otolaryngol Head Neck Surg 134:1170–1176

49. Ponti G, Losi L, Pellacani G, Rossi GB, Presutti L, Mattioli F,Villari D, Wannesson L, Alicandri Ciufelli M, Izzo P, De Rosa M,Marone P, Seidenari S (2008) Wnt pathway, angiogenetic and

hormonal markers in sporadic and familial adenomatous polypo-sis-associated juvenile nasopharyngeal angioWbromas (JNA).Appl Immunohistochem Mol Morphol 16:173–178

50. Zhang PJ, Weber R, Liang HH, Pasha TL, LiVolsi VA (2003)Growth factors and receptors in juvenile nasopharyngeal angioW-broma and nasal polyps: an immunohistochemical study. ArchPathol Lab Med 127:1480–1484

51. Grimm D, Bauer J, Schoenberger J (2009) Blockade of neoangio-genesis, a new and promising technique to control the growth ofmalignant tumors and their metastases. Curr Vasc Pharmacol7:347–357

52. Kivivuori SM, Riikonen P, Valanne L, Lönnqvist T and Saarinen-Pihkala UM (2010) Antiangiogenic combination therapy afterlocal radiotherapy with topotecan radiosensitizer: improved qual-ity of life for children with inoperable brainstem gliomas. ActaPaediatr (in press)

53. Hashizume H, Falcon BL, Kuroda T, Baluk P, Coxon A, Yu D,Bready JV, Oliner JD, McDonald DM (2010) Complementaryactions of inhibitors of Angiopoietin-2 and VEGF on tumor angi-ogenesis and growth. Cancer Res 70:2213–2223

123