Clin. Radiol. (1967) 18, 6 2 - 6 8
THE
NASOPHARYNGEAL
ANGIOFIBROMA*
P. J. FITZPATRICK
From the Princess Margaret Hospital, Toronto, Ontario
Twelve cases of angiofibromata of the nasopharynx are described. Ten patients were adolescent males, 2 were females. The aetiology is discussed but in this series endocrine dysfunction was not obvious. The technique Of irradiation used is described and the results presented. Irradiation appears to act by initiating thrombosis and fibrosis in the turnout which ultimately leads to regression. Sometimes there is a latent period of up to several months before subjective and objective improvement occurs. Early re-treatment after irradiation should be avoided since the chances of tumour regression occurring are high and the final results excellent.
NASOPHARYNGEAL angiofibroma is a rare lesion occurring predominately in young men. It has been described as being clinically malignant but histologically benign (Hubbard 1958). Hippocrates first described the bleeding nasal polyp in the fourth century B.C. and Chelius (1847) was the first to state that the fibrous nasal polyp 'commonly occurred in persons about the time of puberty'.
MATERIAL Since 1946, 13 patients have been seen and treated at The Prineess Margaret Hospital with a diagnosis of nasopharyngeal angiofibroma (Table 1). One patient, a woman of 74, had to be rejected from the
series as not being a true angiofibroma. O f the remaining 12 cases, 10 were male with an average age of 14.8 years. The 2 women were 19 and 36 years of age. None of these cases have been reported before and all were referred specifically for a course of irradiation. Most had severe symptoms with extensive disease and any surgical procudure would have been complicated and hazardous. Ciinieal Feat~res.--Nasal obstruction and epistaxis are the usual presenting symptoms and occurred in a severe degree together in most patients (Table 2). Nasal obstruction varied from partial blocking of one nostril to complete obstruction of the nasal airway. The tumour produced * Based on a paper read at the Canadian Association of Radiologists meeting in Toronto, February, 1965.
TABLE 1 NASOPHARYNGEALANGIOFIBROMA ALL CASESSEENAT ThE PRINCESSMARGARETHOSPITAL 1946-1963 No.
1
[
Sex
Age*
-----~
2 3
15 15
4
14
5 6 7~
Male
10
20 13 11 13 16 15
11
36
Female 2
8
I
9
i
1 if_
S = Surgery R= Radiation
H=Androgens
10
SR SR R R R R SSSSRS SSRS R SSSSSSR
H*
Case
H
1
Haemorrhage
SSR R
* Hypogonadal
11 12
14.8 years
Total * Note that following radiotherapy only two patients required subsequent surgery.
Nasal Obstruction
+44 444 +4 44 44 444 +4 444
+ +4 444 44 44 44 +4 44 44 44 44 +
9
12
44+
3
8 *AverageAge
/ /
TABLE 2 NASOPHARYNGEAL ANGIOFIBROMA PRESENTING SYMPTOMS
Treatment
* Hypogonadal 62
Facial Deformity + + 44
4++ +
4
THE NASOPHARYNGEAL
swelling of the cheek and facial deformity in 4 patients (Figs. 1, 2, 3) but the frog-face described in the literature (Martin, 1954; Rao, 1961, Gatumbi, 1963) as being associated with this disease has not
FIG. 1 Facial deformity with swelling of left cheek (Case 3).
ANGIOFIBROMA
63
been seen although one patient had broadening of the bridge of the nose. Another patient had marked proptosis and 5 more had radiological evidence of bone destruction. Ten of our patients had lesions demonstrable on routine skull x-rays (Figs. 4, 5, 6) and tomography and arteriography demonstrated the tumour more clearly (Figs. 7, 8). In all cases complete healing of bone destruction by turnout followed irradiation. The tumour erodes bone by pressure rather than by infiltration and it can push the soft tissues in front of it, Other symptoms include nasal speech, sneezing, rhinorrhea and deafness from blocking of the eustachian tubes. Cranial nerve palsies have been reported due to intracranial extension of the tumour, but none of our patients showed this feature (Weir, 1960; Hora, 1961). In young men the disease is said to regress with the arrival of sexual maturity. This has lead to the theory that excess oestrogen, or lack of androgen, is in Part responsible for the development of the tumour and to the use of androgens in therapy (Martin, 1948; Harma, 1958; Schiff, 1959). AETIOLOGY The tumour arises in the nasopharynx and its aetiology is in doubt. An attractive theory first put forward by Verneiul in 1877 suggested that the tumour arises from the perichondrium of the embryonic fibro-cartilage of the skull. This area
Fie. 2 FIG. 3 FIG. 2--Artist's diagram of Case 3. Nasopharyngeal Angiofibroma (Coronal Section). FIG. 3--Artist's diagram of Case 3. Nasopharyngeal Angiofibroma (Sagittal Section).
64
CLINICAL
RADIOLOGY
FIG. 4. Case 6. Lateral skull x-ray showing soft tissue mass in the nasopharynx' FI~. 5. Case 6. X-ray of base of skull to show tumour mass mainly in the right nasopharynx. F~c. 6. Case 12. A P skull x-ray to show tumour mass in the left side of the nose producing some bony destruction. ,, ~,~
FIG. 7
FIG. 8
FIG. 7. Case 12. Lateral tomogram o f skull showing large tumour completely filling the nasopharynx.' FIG. 8. Case 12. Carotid arteriogram showing extensive vascular supply to tumour in the nasopharynx.
THE NASOPHARYNGEAL
ANGIOFIBROMA
65
becomes ossified by age 25 and this may account undulating and may undergo hyaline or myxoid for the spontaneous regression of this tumour in degeneration. The stromal cells tend to be plump some cases. Ringertz (1938) postulated an origin and ovoid with steUate cytoplasmic processes and a from the anterior aspect of the first two vertebrae clear nucleus containing a fine chromatin mesh and several authors including Osborn (1959) have and a small, dark nucleolus (Fig. 12). The cytosuggested a sex-linked ectopic focus of nasal • plasmic processes may be hardly discernible and erectile tissue in the nasopharynx which responds to stromal cells resembling ganglion cells are a hormonal variations. Oestrogens stimulate t h e characteristic feature. The stromal cells are not production of vascular tissue and these turnouts particularly oriented to the intercellular substance consist initially of angiomatous tissue which and tend to radiate from the blood vessels, In other becomes more fibrous with time. All but one of areas the lesion bears a resemblance to embryonal our 12 patients were physically normal for their cartilage. Histologically, the lesion is to be differentiated age and sex. In one woman a buccal smear showed a normal chromosomal pattern. All new patients from nasal polypi, sclerosing haemangiomas, will be sexed in this manner and have endocrine haemangio-endotheliomas and fibrosarcomata. The studies including ketosteroid estimations. One presence of mucous glands and lymphoid male aged 20 (Case 5) had sparse pubic and axillary aggregates, as well as the eosinophilic infiltrate, hair, small testicles and a high pitched voice and characteristic of nasal polypi, serves to distinguish the latter. The lack of-mitotic activity in the stromal was described in the notes as being immature. cells and the relative uniformity of these structures distinguishes the lesion from sarcoma, although PATHOLOGY The tumour varies enormously in size. It is in some areas non-uniformity in the distribution usually nodular, dark red in colour and frequently of cells and the presence o f myxoid degeneration ulcerated. Evidence of recent haemorrhage is often in the stroma may make the destruction difficult. present. The tumour may be partly mobile or The endothelial cells lining the vascular structures completely adherent to the surrounding mucosa never show the atypieal features evident in an angiosarcoma. obscuring the site of origin. In this series, 9 cases had unequivocal histology of Histologically, the tumour consists of large dilated blood vessels similar to those seen in angiofibroma. Two of the remaining 3 were hemangiomata with varying amounts of fibrosis. reported as fibromata, but this only means that in The surface consists of the normal nasopharyngeal the tissue examined angiomatous elements were mucosa and, while no true capsule is present, not present, and it is well known that with age compression of its surface may produce a false nasopharyngeal angiofibromata Shrink and become more fibrous; in one case, the histological sections capsule. The criteria for the identification of juveniie could not be found. In all 3, the history, radioangiofibromata in this series were those described logical findings and response to irradiation were by Sternberg in 1954. The basic pattern is one in compatible with a diagnosis of angiofibroma. which endothelial-lined spaces are regularly D I F F E R E N T I A L DIAGNOSIS dispersed in a stroma consisting of fibroblasts and From the clinical standpoint the differential interwoven collagen fibres (Fig. 9). This lesion grows by both expansion and infiltration and it may diagnosis includes nasal polyps of varying types to contain nerves, bone and focal calcification, but fibromata, nasopharyngeal carcinoma, lymphonever mucous glands or lymphoid aggregates. In sarcoma, fibrosarcoma, haemangioendothelioma, the classic form the centre of the tumour contains carcinoma of the antrum, large adenoids, retrocavernous or aneurysmal vessels which alternatively pharyngeal lymphadenopathy and other rare become narrowed and slit-like (Fig. 10). Towards tumours such as chordoma, plasmacytoma and the periphery the vessels tend to become smaller craniopharyngioma. In three of our patients the and in the subepithelial area the lesion resembles referring diagnosis was carcinoma of the maxillary grartulation tissue. The walls of the vascular antrum. structures tend to develop an incomplete muscle The diagnosis is made on clinical grounds but eoat and they may undergo secondary changes routine skull x-rays are useful. A biopsy is desirincluding thrombotic occlusion, a diffuse vascuiitis able but because o f the :danger of severe or an obliterative hyaline thickening of the walls haemorrhage this must be done in the operating (Fig. 11). room under direct vision and with blood transfusion The stromal fibres are generally coarse and facilities ready.
66
CLINICAL
RADIOLOGY
FIG. 9 FIG. 10 Fio. 9--Nasopharyngeal angiofibroma: Classic pattern with endothelial-lined spaces irregularly distributed in a cellular, fibrous stroma. H & E × 40. FiG. 10--Nasopharyngeal angiofibroma: Centrally the vessels are unevenly dilated, aneurysmal, and may develop an incomplete muscle wall. H & E x 125.
TREATMENT The treatment is debatable but from our own experience and survey of literature we would recommend the following broad policy. (1) If symptoms are minimal a 'wait and see' policy is adopted especially in adolescent males because of the possibility of tumour regression with sexual maturity (Harma, 1958; Martin, 1948). (2) If symptoms are becoming troublesome and the lesion is easily resectable surgery is the treatment of choice. (3) For more extensive lesions, where resection is difficult and hazardous, radiotherapy should be used. At The Princess Margaret Hospital the present policy is to treat the whole of the tumour, the nasopharynx and if necessary the surrounding air sinuses. Two opposing 6 × 6 cm. cobalt beam fields are used giving a mid-plane tumour dose of 3,000 rads in 3 weeks. In our hands this dose has been adequate with no patients requiring further
irradiation. Usually the haemorrhage was arrested and tamour regression occurred. The time for regression was very variable but over a period of months the tumours generally changed from being red and vascular to pale grey and firm and all bleeding ceased. In some patients tumour shrinkage gradually occurred over 2 years following radiation. Some authors have had to give large doses of irradiation to control haemorrhage, using local radium applicators to supplement external beam therapy. In repeated courses over many months these patients have received doses in the region of 16,000-18,000 rads. In such cases the high doses of irradiation together with repeated blood transfusions have been life-saving. Windeyer (1964) has recently reported a series of 15 patients none of whom received a tumour dose of less than 6,000 rads. His overall results were good and no necrosis was reported. It is our opinion however, that repeated radiation is potentially dangerous and
THE N A S O P H A R Y N G E A L
ANGIOFIBROMA
67
FIG. I1 Fla. 12 FI~. lla-~Nasopharyngeal angiofibroma: Secondary changes may include vasculitis (illustrated), thrombosis or hyaline sclerotic obliteration. H & E 125. FIG. 12--Nasopharyngeal angiofibroma: Stromal cells have plump ovoid nuclei and irregular, stellate cytoplasmic processes, not necessarily continuous with the hyaline bands. H & E 500.
doses in excess of 3,500 rads are best avoided in non-malignant disease, especially when long-term survival is expected. We believe that a tumour dose of 3,000 rads is just as likely to initiate tumour regression as higher doses and if symptoms cannot be controlled then radical surgery should be attempted. The morbidity from our radiation treatment policy of 3,000 rads in 3 weeks has been negligible, and there have been no complications. Lower doses of radiation have been used in some centres but the results with doses of 1,000 to 2,000 rads have not been good (Pimpinella, 1964), although a dose of 1,100 rads has been reported as producing a 5 0 ~ diminution of tumour size (Hemley, 1953). (4) Radical surgery when necessary means a face-splitting operation or palatal fenestration and this is usually preceeded by ligation of either the external carotid or internal maxillary artery. Haemorrhage is a major problem and radiation
should always be tried before embarking on a major operation. DISCUSSION The incidence of nasopharyngeal angiofibroma is unknown but it appears to be rare. Tumours of the nasopharynx are more common in some parts of Asia and Africa, but no reference to the frequency of angiofibroma could be found in reports from these areas. Several authors state categorically that the condition only occurs i n males; Martin et al. (1948) reported 29 such patients. They noted art average of two new patients a year in a head and neck clinic seeing 2,000 new cases annually. However a report by Handousa (1954) from Egypt reports 11 females among a total of 70 cases, and Figi (1950) reports 5 females among 53 cases seen in the same city over a 33-year period. A more recent series reported from Kenya (Gatumbi 1963)
68
CLINICAL RADIOLOGY
consisted of 5 patients, all males, ranging in age from 10 to 35 years. These patients were seen during one year among 5,500 patients attending the ear, nose and throat department of one Hospital. The history usually seems to commence about puberty, but there is some age scatter. It is often stated that most cases show sexual underdevelopment (Martin, 1948; Redoglia, 1950; Sternberg, 1954), but this was not so in our series where only one patient could be described as slightly hypogonadal. Spontaneous regression .is said to occur, but it is difficult to find a case reported in the literature in which this has occurred in the absence of surgical intervention or radiotherapy. None of our patients died as a result of their disease or treatment, but fatalities from spontaneous haemorrhage or surgical complications have been reported (Martin, 1948; Gatumbi, 1963; Pimpinella, 1964). RESULTS A N D C O N C L U S I O N S Following irradiation there was a latent period during which time haemorrhage became less frequent and less severe whilst nasal obstruction gradually cleared. There was both objective and subjective improvement within 2 months of completing irradiation in 7 patients and within 4 months in 3 patients. In these 10, further tumour regression continued for periods up to 2 years. The other cases took 7 and 12 months to show objective change following irradiation. It would seem that radiotherapy plays its part by initiating changes leading to thrombosis and fibrosis which subsequently result in regression. Two of our male patients were treated with androgens with no improvement. Prior to radiotherapy 16 operations had been carried out on 6 patients (Table 1). In only 2 patients was surgery necessary after irradiation. F r o m this experience and with the known gradual and continued response to irradiation for many months afterwards, it would seem advisable to avoid re-treatment by irradiation or surgery, should regression of the lesion initially appear to be limited. ADDENDUM Since this paper was submitted a further patient, a male aged 18, with a normal physical develop. ment and a year's history of epistaxis and nasal obstruction has been treated by the described irradiation technique (Case 13). Prior to being referred a severe nasal haemorrhage occurred which necessitated hospitalization and blood transfusions. He had a nasal voice, bilateral impairment
of heating and considerable swelling of both cheeks with broadening of the nose. A vast haemorrhagie and ulcerated tumour completely obliterated the nasopharynx causing palatal depression and it was easily visible through the open mouth. X-rays demonstrated the tumour mass, but showed no evidence of bone destruction. Exhaustive endocrine investigations and chromosomal studies were quite normal. By the end of the third week of treatment all bleeding had ceased and the tumour itself was cleaner and smoother. One month later the tumour had shrunk, the facial swelling had subsided and hearing had returned to normal. Since the patient lived over 1,000 miles from this hospital he was discharged home with a six months return appointment. He was then completely asymptomatic, the only abnormality being a very small nubbin of smooth yellowish tumour on the r o o f of the nasopharynx. At one year no abnormality is detectable. Aeknowledgements.--These patients have been treated by various members of the Staff of The Princess Margaret Hospital and I am grateful to them for allowing me to publish this series of cases. Dr. T. C. Brown reviewed the microscopic slides and clarified the histological basis on which the diagnosis of angiofibromata are made. My sincere thanks to to the Department of Medical Photography for preparation of the Figures and Tables and to my secretary Mrs. Margaret McIntyre, for typing the manuscript. REFERENCES CHELIUS,J. M. (1847). System of Surgery, vol. 2 (translated by John F. South). London: Henry Renshaw. FIGI, F. A. & DAVIS,R. E. (1950). Laryngoscope, 60, 794. F1NERMAN,W. B. (1951). Arch. Otolaryng., 54, 620. GATUMBI,I. & LINSELL,C. A. (1963). Brit. J. Cancer, 18, 1. HANDOUSA,A., FAraD,H. & ELWI,A. M. (1954). J. LaryngoL, 68, 647. HARMA,R. A. (1958). Acta oto-laryng. (Stockh.), SuppL, 146. HEVmEV,S. D., et al. (1953). N.Y. St. J. Med., 53, 1861. HIPPOCRATES. Oeuvres Compldtes, Traduction Nouvelle par E. Littre. Paris: J. B. BaiUiere, 10v., 1839-1861. HORA,J. F. & W~LLER,W. A. (1961). Ann. OtoL (St. Louis), 70, 164. HUBBARD,E. M. (1958). Arch. Path., 65, 192. MARTIN, H., EHRLICrI,H. E. & ABELS,J. C. (1948). Ann. Surg., 127, 513. MARTIN,J. S. (1954). J. Laryng., 68, 39. OSBORN,D. A. (1959). J. Laryng., 73, 295. PI~PIN~LLA,R. J. (1964). or. Pediat., 64, 260. RAt, A. B. N. (1961). J. Laryng., 75, 1086. REDOGLIA,E. (1950). Arch. ital. OtoL, 61, 77. RINGERTZ,N. (1938). Acta oto-laryng. (Stockh.), Suppl. 27, 1-405. SCHIFF,M. (1959). Laryngoscope, 69, 981. STImNBERG,S. S. (1954). Cancer, 7, t5. VERNEUIL(1877). Quoted by H. Benseh in Be#rage zur Beurtheilung der chirurgischen Behandlung der Nasenrachenpolypen, Berlin: G. Schade. WEIR, C. D. & DOlG, J. A. (1960). J. Laryng., 74, 685.
WINDEVER,SIR B. (1964). Unpublished data.