Therapeutic Radiology Joseph N. Fields, PhD, MD2 * Karen J. Halverson, MD * Venkata R. Devineni, MD . Joseph R. Simpson, MD, PhD * Carlos A. Perez, MD
Juvenile Nasopharyngeal Angiofibroma: Efficacy of Radiation TherapyL From 1962 to 1984, 13 patients with juvenile nasopharyngeal angiofibroma (JNA) were treated with megavoltage radiation therapy. Followup ranged from 40 to 255 months (median, 136 months). Two patients received radiation therapy as the initial treatment; the other 11 patients had undergone unsuccessful previous surgical treatment (median, three resections). Gross tumor was evident at the start of radiation therapy in seven patients, and orbital, sphenoid sinus, or intracranial extension was noted in eight of 13 (62%). Doses ranged from 3,600 to 5,200 cGy (median, 4,800 cGy in daily fractions of 180-200 cGy). Tumor was controlled in 11 patients (85%) after irradiation. Two patients were treated with embolization for residual mass; both remained asymptomatic and without evidence of tumor 134 and 83 months after embolization, respectively. With the exception of xerostomia and caries, no significant chronic morbidity was seen. This review and other studies demonstrate that megavoltage radiation therapy is an effective and appropriate treatment for advanced and recurrent JNA; its routine use for early tumors remains controversial. Index terms: Angiofibroma, 10.314, 22.314, 23.314, 263.314 * Nasopharynx, neoplasms, 263.314 * Nasopharynx, therapeutic radiology, 263.1299 * Orbit, neoplasms, 22.314 * Orbit, therapeutic radiology, 22.1299 * Paranasal sinuses, neoplasms, 23.314 * Paranasal sinuses, therapeutic radiology, 23.1299 * Therapeutic radiology, in infants and children * Therapeutic radiology, postoperative
Radiology 1990; 176:263-265
JUVENILE nasopharyngeal
angiofibroma (JNA) is a relatively uncommon and histologically benign neoplasm that primarily affects male adolescents. This locally aggressive tumor characteristically expands into the paranasal sinuses, pterygomaxillary and infratemporal areas, and the orbit and intracranial fossae. Serious morbidity and even death may occur from hemorrhage or intracranial extension. Both surgery and radiation therapy are reported to be highly effective in the treatment of these tumors, but the relative merits of the procedures remain controversial. Some authors (1,2) espouse primary radiation therapy for essentially all lesions, including the very early ones, whereas others (3-5) prefer a primary surgical approach even for the most extensive tumors. Each approach has its strong proponents who have produced analyses of efficacy, toxicity, and cost that support their respective positions. The high rates of tumor control obtained with radiation therapy at some institutions (2,6-9) have not been the universal experience, and some authors (3,10) question the efficacy of this modality in the treatment of JNA. The effectiveness and toxicity of radiation therapy are addressed in this report, which presents the Mallinckrodt Institute experience with modern megavoltage radiation therapy as primary or adjuvant treatment for histologically proved JNA. PATIENTS AND METHODS Twenty patients with JNA were referred to the Mallinckrodt Institute of Radiology, Washington University Medical
I From the Radiation Oncology Center, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St Louis. Received December 26, 1989; revision requested February 19, 1990; revision received March 12; accepted March 19. Address reprint requests to C.A.P., Suite 5500, 4939 Audubon Ave, St Louis, MO 63110. 2 Current address: Department of Radiation Oncology, Memorial Medical Center, Springfield,
Ill.
¢ RSNA, 1990
Center, St Louis, between 1962 and 1984. Five patients were treated surgically and received no radiation therapy. Two patients were irradiated at outside institutions. The remaining 13 patients, who were treated with radiation therapy at our institution, are the subject of this report. Follow-up is current in all patients and ranges from 40 to 255 months, with a median of 136 months. All patients were boys. Age at diagnosis ranged from 11 to 17 years, with a median of 14 years. The most common symptoms at presentation nasal obstruction, reported by 11 pa-
were
tients (85%), and epistaxis, also reported by 11 patients (85%); however, exophthalmos/periorbital edema and auditory changes were also seen in two patients (15%) each. Diagnosis was histologically confirmed in all patients, on the basis of previous excision in 11, and on the basis of biopsy alone in two. The majority of patients (10 of 13) in this series were evaluated by means of angiography to define the blood supply to the tumor and also its intracranial extent. Skull radiographs (five patients) and tomograms (four patients) were used to assess tumor extent during the 1960s, with computed tomography (CT) (four patients) and magnetic resonance imaging (one patient also underwent CT) used more recently. The extent of tumor, determined radiographically before initial treatment, is shown in the Table. Tumors were staged retrospectively according to the following system (3): stage I, confined to the nasopharynx and/or nasal fossae; stage II, extending into the sphenoid sinus and/or pterygomaxillary fossae; stage III, extending beyond stage II limits into the maxillary sinus, ethmoid sinuses, orbits, infratemporal fossae, cheeks, or palate; and stage IV, extending into the intracranial fossae. Of the 13 patients treated with radiation therapy, 11 had undergone unsuccessful surgical treatment, with a mean of three resections each (range, one to five). Gross recurrent tumor was evident at the start of radiation therapy in seven of
Abbreviation: JNA geal angiofibroma.
=
juvenile nasopharyn-
263
Data from 13 JNA Patients Treated with Radiation Therapy No. of Prior
Extension Beyond Nasopharynx and
Patient
Age (y)*
Nasal Cavity
Stage'
Surgical Procedures
1 2 3 4 5
14
17 14 13 14
S None IC PMF None
II I IV II I
2 4 4 5 1
6 7
13 11
PMF M, PMF, orbit
II
III
0 3t
8
14
S, M, PMF,
III
9 10 11 12
13 13 12 15
II
13
14
S, PMF None E,S, PMF, S, M, E, PMF, orbit S, M, PMF, orbit, IC
Tumor Status at XRT G G G
Dose
(cGy/fx/d) 3,660/21/29 4,140/22/37
Statust NED (173) NED (119)
Remarks Caries
Microscopic G G
NED (255) NED (127) NED (204)
Caries Jehovah's
G
5,018/26/37 4,962/24/31 5,000/25/38 5,000/29/45 5,000/25/36
NED (179) Recurrence (41), NED (107 [after salvage
Estrogen therapy
2
Microscopic
5,200/26/38
Poor healing, reconstructive surgery
51 1 1t
Microscopic
I III
III
3
Microscopic
IV
01
G
5,000/25/35 3,500/17/22 4,816/24/49 4,500/25/34 4,000/20/28
Recurrence (29), NED (137 after embolization) NED (157) NED (75) NED (50) NED (44) NED (40)
Estrogen therapy
G G
embolization])
Witness
Note.-E ethmoid sinuses, fx = no. of treatment fractions, G gross residual tumor (at initiation of radiation therapy), IC = intracranial, M = maxillary sinus, NED - alive without evidence of disease, PMF = pterygomaxillary fossa, S - sphenoid sinus, XRT = radiation therapy. * At initial presentation (reference 3). t Months after termination of radiation therapy in parentheses. t Initial treatment included embolization.
these patients, as well as in the two patients who received irradiation as initial treatment after biopsy alone. Four patients received elective irradiation after gross excision of recurrent tumor; in all these patients, the radiation therapy was
delivered after two or more recurrences of the tumor. The interval from initial diagnosis to irradiation in the 11 patients with unsuccessful surgery ranged from 9 to 72 months, with a median of 20 months. Prior to radiation therapy, four patients had also been treated with embolization at the time of initial arteriography, and one of these patients underwent a second embolization at the time of recurrence. Two patients in this series received estrogen treatment. All patients were treated with megavoltage radiation (cobalt-60, 25 MV) through opposed lateral portals. Simulation and custom blocking were used to ensure minimal irradiation of normal tissue consistent with adequate tumor coverage (ie, 1.5-2.0-cm margins around the clinically and radiographically evident tumor volume). The total midline dose ranged from 3,500 to 5,200 cGy, with a median of 4,800 cGy. (Tumor dose fractionation ranged from 20 to 29 fractions; median 23, mode 24.) Daily fraction size was typically 180200 cGy, with five fractions per week. As can be seen from the Table, there was no correlation in this series between dose and either initial tumor extent or tumor status (gross vs microscopic) at initiation of radiation therapy.
RESULTS All patients were alive and without evidence of disease at intervals ranging from 40 to 225 months (median, 264 . Radiology
136 months) after the start of radiation therapy (Table). Eleven of the 13 patients (85%) remained without evidence of recurrence since completion of radiation therapy. Two patients, both of whom had extensive stage III tumors involving the maxillary sinus and pterygomaxillary fossae, experienced tumor recurrence after radiation therapy. The first patient had received a dose of 5,000 cGy for gross recurrence after embolization and three resections, 18 months after the initial diagnosis. Forty-one months
after radiation therapy, he developed epistaxis and was found to have a recurrent mass within the nasal area. The second patient had received a dose of 5,200 cGy after gross total excision of recurrent tumor after resection. Twenty-nine months after radiation therapy, he also experienced epistaxis and a clinically evident tumor mass. Both patients were treated with embolization (with the addition of estrogens in the first patient), and both were without evidence of disease, 107 and 137 months, respectively, after embolization. The acute effects of toxicity of radiation therapy, principally inflammation and dryness of mucosal linings, were generally easily managed with analgesics, humidification, and topical lubricants. Only one patient required treatment interruption longer than 1 week. Most patients reported some degree of chronic nasal dryness, and two patients developed dental caries secondary to xerosto-
mia. Patient 8 (Table) underwent nasoseptoplasty and palatoplasty 7'/2 years after radiation therapy and 5 years after salvage embolization, to reconstruct the anatomic deficits left after an earlier excision; this patient experienced poor healing attributed to the radiation therapy. He continued to have occasional epistaxis not requiring transfusion and was without evidence of disease at age 26. No intellectual, neurologic, or visual deficits were discovered in these patients, nor were any additional neoplasms diagnosed.
DISCUSSION The choice of primary treatment for JNA remains controversial, with strong proponents for both radiation therapy (1,2) and surgery (3,5,10,11). Considerable disagreement exists in the literature regarding the relative merits and demerits of each type of therapy. Our experience confirms the efficacy of radiation therapy in JNA, with durable control attained in 85% of patients (11 of 13). Most of the patients in our series had advanced and/or recurrent tumors, and complete surgical excision was considered unlikely. The largest reported experience with modern megavoltage radiation therapy for JNA comes from the Princess Margaret Hospital in Toronto: Cummings et al (2) described 55 patients, 42 of whom were treated initially with radiation therapy, the other 13 treated with radiation therapy for recurrent tumor after excision. Radiation therapy consisted of a dose of 3,000-3,500 cGy in 14-16 fractions over 3 weeks, and all patients were fol-
July 1990
lowed up for a minimum of 3 years. Among the 55 patients, the tumor was controlled in 44 (80%) with the initial course of radiation therapy, with no dose response within this relatively narrow dose range. All 11 patients who experienced tumor regrowth were successfully treated with surgery or additional radiation therapy, for an ultimate tumor control rate of 100%. It was thought that most radiation therapy failures were attributable to geographic misses. Cataracts developed in two patients and hypopituitarism in one. Two patients developed second malignancies: a basal cell carcinoma of the facial skin within the radiation portal and a well-differentiated thyroid cancer, respectively. Both patients were treated surgically and showed no evidence of disease at the time of the report. These investigators performed a detailed analysis (1) of the efficacy and complications of both radiation therapy and surgery and found the relative mortality risks of the two types of therapy to be similar. They remain strong proponents of radiation therapy for the primary treatment of JNA. Other authors (6-9) also suggest the efficacy of radiation therapy. Jereb et al (7) described 66 patients treated with primary radiation therapy over a 42-year period, during which treatment techniques and policies varied widely. Brachytherapy and both kilovoltage and megavoltage external beam radiation therapy were used. Among this heterogeneous group of patients, the tumor was controlled in 75% (47 of 63 patients with adequate followup). In two more recent series (8,9), totaling 14 patients who were treated with megavoltage radiation, durable local control in 92% (12 of 13) of those receiving a dose of 3,000-3,500 cGy over 3 weeks was reported. The 14th patient received only 1,000 cGy, and treatment failed. Economou et al (6) recently updated a study with 14 patients who had intracranial extension and were treated with initial radiation therapy. Only three required further therapy, for a control rate of 79%. Failures were attributed to inadequate dose (see below) or volume. Of another four patients who were irradiated for tumor recurrence after initial excision, one required further therapy. Morbidity due to treatment was not addressed in this report. In these series, the reported toxic effects of radiation therapy for JNA were generally not severe. Acute reactions seldom required treatment interruptions. Common chronic effects include xerostomia and nasal dryness. More serious effects such as osteitis, necrosis, or retardation of facial development appear ex-
Volume 176 * Number 1
tremely uncommon with megavoltage techniques, and the reported frequencies of second malignancies in series with long follow-up have been 0%-4% (1,2,79). According to Cummings' analysis (1), the risk of severe morbidity and mortality appears approximately equivalent for radiation therapy and surgery. Fatalities due to surgical complications are often immediate, whereas radiation-induced tumors are not seen for many years. In surgical series, control rates in the same range as those found in radiation therapy series (70%-100%) were generally reported. With the improved ability to stage the tumor radiographically, with the use of preoperative embolization, and with improved surgical technique, recent reports (3,10) show 100% control rates for lesions without intracranial extension. Operative blood loss, morbidity, and mortality have also decreased significantly compared with previous experience. Because of this excellent tumor control with minimal morbidity, surgery is the preferred treatment for patients with early lesions. Indeed, some groups are investigating the use of intra-/extracranial surgical approaches for selected intracranial lesions (3,5). This aggressive treatment has the potential for significant morbidity, and radiation therapy is generally recommended (6,10-14) for these more extensive JNAs. The radiation therapy dose response of JNA remains uncertain. In our series, all patients received doses greater than or equal to 3,500 cGy, and only two received doses less than 4,000 cGy (both of these patients show no evidence of disease). The two patients in whom radiation therapy was unsuccessful received 5,000 and 5,200 cGy, respectively. Our local control rate appears roughly equivalent to that of Cummings et al (2), who used doses of 3,000 cGy and 3,500 cGy in 14-16 fractions, with no evidence of dose response. On the other hand, Economou et al (6) reported more failures in patients receiving doses less than or equal to 3,600 cGy and recommended doses above that level. The question of dose response is somewhat obfuscated by the likelihood that some failures are due to geographic misses (2). In conclusion, our experience confirms the effectiveness of radiation therapy in the control of JNA. In patients with early lesions, surgery is the usual treatment and offers a high likelihood of control with low morbidity. Radiation therapy should, however, be considered for the occasional patient with early JNA whose operative morbidity or mortality might be expected to be higher, on medical or religious grounds. For treatment of many patients with extensive tumors, especially those with intracranial extension, we believe that the risk-benefit ratio favors radiation therapy. We also recommend radiation therapy for patients with recur-
rent (especially multiply recurrent) JNA after resection. Until more precise information is available, we currently recommend doses greater than or equal to 3,600 cGy, in 200-cGy daily fractions, with possibly higher doses (4,000-4,500 cGy) for extensive disease. U Acknowledgment: We thank LuAnn Russell for her help with the preparation of this manuscript.
References 1. Cummings BJ. Relative risk factors in the treatment of juvenile nasopharyngeal angiofibroma. Head Neck Surg 1980; 3:2126. 2. Cummings BJ, Blend R, Keane T, et al. Primary radiation therapy for juvenile nasopharyngeal angiofibroma. Laryngoscope 1984; 94:1599-1605. 3. Antonelli AR, Cappiello J, Lorenzo DD, Donajo CA, Nicolai N, Orlandini A. Diagnosis, staging and treatment of juvenile nasopharyngeal angiofibroma. Laryngoscope 1987; 97:1319-1325. 4. Jafek BW, Nahum AM, Butler RM, Ward PH. Surgical treatment of juvenile nasopharyngeal angiofibroma. Laryngoscope 1973; 83:707-720. 5. Jafek BW, Krekorian EA, Kirsch WM, Wood RP. Juvenile nasopharyngeal angiofibroma: management of intracranial extension. Head Neck Surg 1979; 2:119128. 6. Economou TS, Abemayor E, Ward PH. Juvenile nasopharyngeal angiofibroma: an update of the UCLA experience, 19601985. Laryngoscope 1988; 98:170-175. 7. Jereb J, Anggard A, Baryd I. Juvenile nasopharyngeal angiofibroma. Acta Radiol 1970; 9:302-310. 8. Vadivel SP, Bosch A, Jose B. Juvenile nasopharyngeal angiofibroma. J Surg Oncol 1980; 15:323-326. 9. Sinha PP, Aziz HI. Juvenile nasopharyngeal angiofibroma: a report of seven cases. Radiology 1978; 127:501-505. 10. Waldman SR, Levine HL, Astor F, Wood BG, Weinstein M, Tucker HM. Surgical experience with nasopharyngeal angiofibroma. Arch Otolaryngol 1981; 107:677682. 11. Biller HF, Sessions DG, Ogura JH. Angiofibroma: a treatment approach. Laryngoscope 1974; 84:695-706. 12. Biller HF. Juvenile nasopharyngeal angiofibroma. Ann Otol 1978; 78:630-632. 13. Chandler JR. Goulding R. Moskowitz L, Quencer RM. Nasopharyngeal angiofibromas: staging and management. Ann Otol Rhinol Laryngol 1984; 93:322-329. 14. Neel HB, Whicker JH, Devine KD, Weiland LH. Juvenile angiofibroma: review of 120 cases. Am J Surg 1973; 126:547-556.
Radiology * 265
Juvenile Nasopharyngeal Angiofibroma: Efficacy of Radiation TherapyL From 1962 to 1984, 13 patients with juvenile nasopharyngeal angiofibroma (JNA) were treated with megavoltage radiation therapy. Followup ranged from 40 to 255 months (median, 136 months). Two patients received radiation therapy as the initial treatment; the other 11 patients had undergone unsuccessful previous surgical treatment (median, three resections). Gross tumor was evident at the start of radiation therapy in seven patients, and orbital, sphenoid sinus, or intracranial extension was noted in eight of 13 (62%). Doses ranged from 3,600 to 5,200 cGy (median, 4,800 cGy in daily fractions of 180-200 cGy). Tumor was controlled in 11 patients (85%) after irradiation. Two patients were treated with embolization for residual mass; both remained asymptomatic and without evidence of tumor 134 and 83 months after embolization, respectively. With the exception of xerostomia and caries, no significant chronic morbidity was seen. This review and other studies demonstrate that megavoltage radiation therapy is an effective and appropriate treatment for advanced and recurrent JNA; its routine use for early tumors remains controversial. Index terms: Angiofibroma, 10.314, 22.314, 23.314, 263.314 * Nasopharynx, neoplasms, 263.314 * Nasopharynx, therapeutic radiology, 263.1299 * Orbit, neoplasms, 22.314 * Orbit, therapeutic radiology, 22.1299 * Paranasal sinuses, neoplasms, 23.314 * Paranasal sinuses, therapeutic radiology, 23.1299 * Therapeutic radiology, in infants and children * Therapeutic radiology, postoperative
Radiology 1990; 176:263-265
JUVENILE nasopharyngeal
angiofibroma (JNA) is a relatively uncommon and histologically benign neoplasm that primarily affects male adolescents. This locally aggressive tumor characteristically expands into the paranasal sinuses, pterygomaxillary and infratemporal areas, and the orbit and intracranial fossae. Serious morbidity and even death may occur from hemorrhage or intracranial extension. Both surgery and radiation therapy are reported to be highly effective in the treatment of these tumors, but the relative merits of the procedures remain controversial. Some authors (1,2) espouse primary radiation therapy for essentially all lesions, including the very early ones, whereas others (3-5) prefer a primary surgical approach even for the most extensive tumors. Each approach has its strong proponents who have produced analyses of efficacy, toxicity, and cost that support their respective positions. The high rates of tumor control obtained with radiation therapy at some institutions (2,6-9) have not been the universal experience, and some authors (3,10) question the efficacy of this modality in the treatment of JNA. The effectiveness and toxicity of radiation therapy are addressed in this report, which presents the Mallinckrodt Institute experience with modern megavoltage radiation therapy as primary or adjuvant treatment for histologically proved JNA. PATIENTS AND METHODS Twenty patients with JNA were referred to the Mallinckrodt Institute of Radiology, Washington University Medical
I From the Radiation Oncology Center, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St Louis. Received December 26, 1989; revision requested February 19, 1990; revision received March 12; accepted March 19. Address reprint requests to C.A.P., Suite 5500, 4939 Audubon Ave, St Louis, MO 63110. 2 Current address: Department of Radiation Oncology, Memorial Medical Center, Springfield,
Ill.
¢ RSNA, 1990
Center, St Louis, between 1962 and 1984. Five patients were treated surgically and received no radiation therapy. Two patients were irradiated at outside institutions. The remaining 13 patients, who were treated with radiation therapy at our institution, are the subject of this report. Follow-up is current in all patients and ranges from 40 to 255 months, with a median of 136 months. All patients were boys. Age at diagnosis ranged from 11 to 17 years, with a median of 14 years. The most common symptoms at presentation nasal obstruction, reported by 11 pa-
were
tients (85%), and epistaxis, also reported by 11 patients (85%); however, exophthalmos/periorbital edema and auditory changes were also seen in two patients (15%) each. Diagnosis was histologically confirmed in all patients, on the basis of previous excision in 11, and on the basis of biopsy alone in two. The majority of patients (10 of 13) in this series were evaluated by means of angiography to define the blood supply to the tumor and also its intracranial extent. Skull radiographs (five patients) and tomograms (four patients) were used to assess tumor extent during the 1960s, with computed tomography (CT) (four patients) and magnetic resonance imaging (one patient also underwent CT) used more recently. The extent of tumor, determined radiographically before initial treatment, is shown in the Table. Tumors were staged retrospectively according to the following system (3): stage I, confined to the nasopharynx and/or nasal fossae; stage II, extending into the sphenoid sinus and/or pterygomaxillary fossae; stage III, extending beyond stage II limits into the maxillary sinus, ethmoid sinuses, orbits, infratemporal fossae, cheeks, or palate; and stage IV, extending into the intracranial fossae. Of the 13 patients treated with radiation therapy, 11 had undergone unsuccessful surgical treatment, with a mean of three resections each (range, one to five). Gross recurrent tumor was evident at the start of radiation therapy in seven of
Abbreviation: JNA geal angiofibroma.
=
juvenile nasopharyn-
263
Data from 13 JNA Patients Treated with Radiation Therapy No. of Prior
Extension Beyond Nasopharynx and
Patient
Age (y)*
Nasal Cavity
Stage'
Surgical Procedures
1 2 3 4 5
14
17 14 13 14
S None IC PMF None
II I IV II I
2 4 4 5 1
6 7
13 11
PMF M, PMF, orbit
II
III
0 3t
8
14
S, M, PMF,
III
9 10 11 12
13 13 12 15
II
13
14
S, PMF None E,S, PMF, S, M, E, PMF, orbit S, M, PMF, orbit, IC
Tumor Status at XRT G G G
Dose
(cGy/fx/d) 3,660/21/29 4,140/22/37
Statust NED (173) NED (119)
Remarks Caries
Microscopic G G
NED (255) NED (127) NED (204)
Caries Jehovah's
G
5,018/26/37 4,962/24/31 5,000/25/38 5,000/29/45 5,000/25/36
NED (179) Recurrence (41), NED (107 [after salvage
Estrogen therapy
2
Microscopic
5,200/26/38
Poor healing, reconstructive surgery
51 1 1t
Microscopic
I III
III
3
Microscopic
IV
01
G
5,000/25/35 3,500/17/22 4,816/24/49 4,500/25/34 4,000/20/28
Recurrence (29), NED (137 after embolization) NED (157) NED (75) NED (50) NED (44) NED (40)
Estrogen therapy
G G
embolization])
Witness
Note.-E ethmoid sinuses, fx = no. of treatment fractions, G gross residual tumor (at initiation of radiation therapy), IC = intracranial, M = maxillary sinus, NED - alive without evidence of disease, PMF = pterygomaxillary fossa, S - sphenoid sinus, XRT = radiation therapy. * At initial presentation (reference 3). t Months after termination of radiation therapy in parentheses. t Initial treatment included embolization.
these patients, as well as in the two patients who received irradiation as initial treatment after biopsy alone. Four patients received elective irradiation after gross excision of recurrent tumor; in all these patients, the radiation therapy was
delivered after two or more recurrences of the tumor. The interval from initial diagnosis to irradiation in the 11 patients with unsuccessful surgery ranged from 9 to 72 months, with a median of 20 months. Prior to radiation therapy, four patients had also been treated with embolization at the time of initial arteriography, and one of these patients underwent a second embolization at the time of recurrence. Two patients in this series received estrogen treatment. All patients were treated with megavoltage radiation (cobalt-60, 25 MV) through opposed lateral portals. Simulation and custom blocking were used to ensure minimal irradiation of normal tissue consistent with adequate tumor coverage (ie, 1.5-2.0-cm margins around the clinically and radiographically evident tumor volume). The total midline dose ranged from 3,500 to 5,200 cGy, with a median of 4,800 cGy. (Tumor dose fractionation ranged from 20 to 29 fractions; median 23, mode 24.) Daily fraction size was typically 180200 cGy, with five fractions per week. As can be seen from the Table, there was no correlation in this series between dose and either initial tumor extent or tumor status (gross vs microscopic) at initiation of radiation therapy.
RESULTS All patients were alive and without evidence of disease at intervals ranging from 40 to 225 months (median, 264 . Radiology
136 months) after the start of radiation therapy (Table). Eleven of the 13 patients (85%) remained without evidence of recurrence since completion of radiation therapy. Two patients, both of whom had extensive stage III tumors involving the maxillary sinus and pterygomaxillary fossae, experienced tumor recurrence after radiation therapy. The first patient had received a dose of 5,000 cGy for gross recurrence after embolization and three resections, 18 months after the initial diagnosis. Forty-one months
after radiation therapy, he developed epistaxis and was found to have a recurrent mass within the nasal area. The second patient had received a dose of 5,200 cGy after gross total excision of recurrent tumor after resection. Twenty-nine months after radiation therapy, he also experienced epistaxis and a clinically evident tumor mass. Both patients were treated with embolization (with the addition of estrogens in the first patient), and both were without evidence of disease, 107 and 137 months, respectively, after embolization. The acute effects of toxicity of radiation therapy, principally inflammation and dryness of mucosal linings, were generally easily managed with analgesics, humidification, and topical lubricants. Only one patient required treatment interruption longer than 1 week. Most patients reported some degree of chronic nasal dryness, and two patients developed dental caries secondary to xerosto-
mia. Patient 8 (Table) underwent nasoseptoplasty and palatoplasty 7'/2 years after radiation therapy and 5 years after salvage embolization, to reconstruct the anatomic deficits left after an earlier excision; this patient experienced poor healing attributed to the radiation therapy. He continued to have occasional epistaxis not requiring transfusion and was without evidence of disease at age 26. No intellectual, neurologic, or visual deficits were discovered in these patients, nor were any additional neoplasms diagnosed.
DISCUSSION The choice of primary treatment for JNA remains controversial, with strong proponents for both radiation therapy (1,2) and surgery (3,5,10,11). Considerable disagreement exists in the literature regarding the relative merits and demerits of each type of therapy. Our experience confirms the efficacy of radiation therapy in JNA, with durable control attained in 85% of patients (11 of 13). Most of the patients in our series had advanced and/or recurrent tumors, and complete surgical excision was considered unlikely. The largest reported experience with modern megavoltage radiation therapy for JNA comes from the Princess Margaret Hospital in Toronto: Cummings et al (2) described 55 patients, 42 of whom were treated initially with radiation therapy, the other 13 treated with radiation therapy for recurrent tumor after excision. Radiation therapy consisted of a dose of 3,000-3,500 cGy in 14-16 fractions over 3 weeks, and all patients were fol-
July 1990
lowed up for a minimum of 3 years. Among the 55 patients, the tumor was controlled in 44 (80%) with the initial course of radiation therapy, with no dose response within this relatively narrow dose range. All 11 patients who experienced tumor regrowth were successfully treated with surgery or additional radiation therapy, for an ultimate tumor control rate of 100%. It was thought that most radiation therapy failures were attributable to geographic misses. Cataracts developed in two patients and hypopituitarism in one. Two patients developed second malignancies: a basal cell carcinoma of the facial skin within the radiation portal and a well-differentiated thyroid cancer, respectively. Both patients were treated surgically and showed no evidence of disease at the time of the report. These investigators performed a detailed analysis (1) of the efficacy and complications of both radiation therapy and surgery and found the relative mortality risks of the two types of therapy to be similar. They remain strong proponents of radiation therapy for the primary treatment of JNA. Other authors (6-9) also suggest the efficacy of radiation therapy. Jereb et al (7) described 66 patients treated with primary radiation therapy over a 42-year period, during which treatment techniques and policies varied widely. Brachytherapy and both kilovoltage and megavoltage external beam radiation therapy were used. Among this heterogeneous group of patients, the tumor was controlled in 75% (47 of 63 patients with adequate followup). In two more recent series (8,9), totaling 14 patients who were treated with megavoltage radiation, durable local control in 92% (12 of 13) of those receiving a dose of 3,000-3,500 cGy over 3 weeks was reported. The 14th patient received only 1,000 cGy, and treatment failed. Economou et al (6) recently updated a study with 14 patients who had intracranial extension and were treated with initial radiation therapy. Only three required further therapy, for a control rate of 79%. Failures were attributed to inadequate dose (see below) or volume. Of another four patients who were irradiated for tumor recurrence after initial excision, one required further therapy. Morbidity due to treatment was not addressed in this report. In these series, the reported toxic effects of radiation therapy for JNA were generally not severe. Acute reactions seldom required treatment interruptions. Common chronic effects include xerostomia and nasal dryness. More serious effects such as osteitis, necrosis, or retardation of facial development appear ex-
Volume 176 * Number 1
tremely uncommon with megavoltage techniques, and the reported frequencies of second malignancies in series with long follow-up have been 0%-4% (1,2,79). According to Cummings' analysis (1), the risk of severe morbidity and mortality appears approximately equivalent for radiation therapy and surgery. Fatalities due to surgical complications are often immediate, whereas radiation-induced tumors are not seen for many years. In surgical series, control rates in the same range as those found in radiation therapy series (70%-100%) were generally reported. With the improved ability to stage the tumor radiographically, with the use of preoperative embolization, and with improved surgical technique, recent reports (3,10) show 100% control rates for lesions without intracranial extension. Operative blood loss, morbidity, and mortality have also decreased significantly compared with previous experience. Because of this excellent tumor control with minimal morbidity, surgery is the preferred treatment for patients with early lesions. Indeed, some groups are investigating the use of intra-/extracranial surgical approaches for selected intracranial lesions (3,5). This aggressive treatment has the potential for significant morbidity, and radiation therapy is generally recommended (6,10-14) for these more extensive JNAs. The radiation therapy dose response of JNA remains uncertain. In our series, all patients received doses greater than or equal to 3,500 cGy, and only two received doses less than 4,000 cGy (both of these patients show no evidence of disease). The two patients in whom radiation therapy was unsuccessful received 5,000 and 5,200 cGy, respectively. Our local control rate appears roughly equivalent to that of Cummings et al (2), who used doses of 3,000 cGy and 3,500 cGy in 14-16 fractions, with no evidence of dose response. On the other hand, Economou et al (6) reported more failures in patients receiving doses less than or equal to 3,600 cGy and recommended doses above that level. The question of dose response is somewhat obfuscated by the likelihood that some failures are due to geographic misses (2). In conclusion, our experience confirms the effectiveness of radiation therapy in the control of JNA. In patients with early lesions, surgery is the usual treatment and offers a high likelihood of control with low morbidity. Radiation therapy should, however, be considered for the occasional patient with early JNA whose operative morbidity or mortality might be expected to be higher, on medical or religious grounds. For treatment of many patients with extensive tumors, especially those with intracranial extension, we believe that the risk-benefit ratio favors radiation therapy. We also recommend radiation therapy for patients with recur-
rent (especially multiply recurrent) JNA after resection. Until more precise information is available, we currently recommend doses greater than or equal to 3,600 cGy, in 200-cGy daily fractions, with possibly higher doses (4,000-4,500 cGy) for extensive disease. U Acknowledgment: We thank LuAnn Russell for her help with the preparation of this manuscript.
References 1. Cummings BJ. Relative risk factors in the treatment of juvenile nasopharyngeal angiofibroma. Head Neck Surg 1980; 3:2126. 2. Cummings BJ, Blend R, Keane T, et al. Primary radiation therapy for juvenile nasopharyngeal angiofibroma. Laryngoscope 1984; 94:1599-1605. 3. Antonelli AR, Cappiello J, Lorenzo DD, Donajo CA, Nicolai N, Orlandini A. Diagnosis, staging and treatment of juvenile nasopharyngeal angiofibroma. Laryngoscope 1987; 97:1319-1325. 4. Jafek BW, Nahum AM, Butler RM, Ward PH. Surgical treatment of juvenile nasopharyngeal angiofibroma. Laryngoscope 1973; 83:707-720. 5. Jafek BW, Krekorian EA, Kirsch WM, Wood RP. Juvenile nasopharyngeal angiofibroma: management of intracranial extension. Head Neck Surg 1979; 2:119128. 6. Economou TS, Abemayor E, Ward PH. Juvenile nasopharyngeal angiofibroma: an update of the UCLA experience, 19601985. Laryngoscope 1988; 98:170-175. 7. Jereb J, Anggard A, Baryd I. Juvenile nasopharyngeal angiofibroma. Acta Radiol 1970; 9:302-310. 8. Vadivel SP, Bosch A, Jose B. Juvenile nasopharyngeal angiofibroma. J Surg Oncol 1980; 15:323-326. 9. Sinha PP, Aziz HI. Juvenile nasopharyngeal angiofibroma: a report of seven cases. Radiology 1978; 127:501-505. 10. Waldman SR, Levine HL, Astor F, Wood BG, Weinstein M, Tucker HM. Surgical experience with nasopharyngeal angiofibroma. Arch Otolaryngol 1981; 107:677682. 11. Biller HF, Sessions DG, Ogura JH. Angiofibroma: a treatment approach. Laryngoscope 1974; 84:695-706. 12. Biller HF. Juvenile nasopharyngeal angiofibroma. Ann Otol 1978; 78:630-632. 13. Chandler JR. Goulding R. Moskowitz L, Quencer RM. Nasopharyngeal angiofibromas: staging and management. Ann Otol Rhinol Laryngol 1984; 93:322-329. 14. Neel HB, Whicker JH, Devine KD, Weiland LH. Juvenile angiofibroma: review of 120 cases. Am J Surg 1973; 126:547-556.
Radiology * 265
