Management pitfalls in the use of embolization for the treatment of severe epistaxis.

Abstract

Angiographic embolization for the treatment of severe recurrent epistaxis was added to the traditional treatment options--nasal packing, cauterization, and surgical vessel ligation--in 1974. Since then, clinical experience has shown that this procedure is safe and effective. When epistaxis cannot be controlled with cautery, nasal packing is the most common next step. As such, it is often performed by emergency physicians and other clinicians who are not otolaryngologists. We report two cases in which intranasal neoplasms were obscured as a result of a significant distortion of the normal anatomy. This distortion was secondary to emergency-room treatment of severe epistaxis by repeated nasal packing followed by angiographic embolization. Pre-embolization angiographic studies and subsequent postembolization endoscopic evaluations did not reveal the presence of the occult neoplasms because of the presence of inflammation and edema after treatment. Clinicians should be aware that nasal packing and embolization ca n obscure the underlying source of epistaxis, and follow-up radiologic studies and endoscopic evaluations are essential to avoid delays in diagnosis.

Introduction

Approximately 10% of all cases of epistaxis are severe enough to warrant the more aggressive interventions, such as posterior packing, operative vessel ligation, or angiography with embolization. The differential diagnosis of severe epistaxis includes conditions in a broad number of categories, including trauma, inflammation, hypertension, arteriosclerosis, foreign-body impaction, bleeding diathesis, vascular disorders, and neoplasms, as well as idiopathic causes.

Malignant tumors of the nose and paranasal sinuses account for only 0.5% of all tumors. (1) Although they are rare entities, they must be considered in the differential diagnosis of severe epistaxis. When no identifiable cause of epistaxis can be identified, it is important to consider whether previous treatment might have contributed to the difficulty in making a diagnosis. In this article, we describe two cases of severe epistaxis in which angiography and embolization not only failed to control the bleeding, but also obscured the underlying etiology.

Case reports

Patient 1. A 63-year-old man was admitted directly to the neuroradiology service of a major tertiary care teaching hospital to undergo angiographic embolization for severe recurrent epistaxis. His medical history was significant for emphysema, glaucoma, and cataracts. He had no known bleeding diathesis. He had an 80-pack-year smoking history and a history of alcohol abuse.

The patient had previously gone to the emergency department of an outside hospital for treatment of a severe episode of right-sided anterior and posterior epistaxis. Prior to this severe episode, the patient had experienced chronic epistaxis that had worsened over a period of 2 months prior to his visit to the emergency department. In the emergency department, he was initially treated with anterior nasal packing. When this failed to control the bleeding, hemostasis was achieved with an Epistat intranasal balloon.

Following balloon placement, the patient was referred directly to the interventional neuroradiology service at our institution for angiographic embolization. After angiography failed to demonstrate any active bleeding with the balloon in place, he underwent embolization of both distal internal maxillary arteries and the right facial artery with Gelfoam pledgets. The procedure was well tolerated and without complications. Postprocedurally, the patient was transferred to the otolaryngology service for observation. The nasal packing was removed on postprocedure day 2, and the patient was discharged on the third hospital day without further bleeding.

Within 2 weeks of discharge, the patient had begun to experience recurrent episodes of right-sided epistaxis. Nasal endoscopy in the office revealed the presence of granulation tissue and severe edema in the right nasal cavity secondary to the previous nasal packing. Initially, this granulation tissue was felt to be the source of the postembolization bleeding episodes, and the patient was managed with observation and serial examinations. After yet another episode of severe epistaxis that required anterior nasal packing in an outside emergency department, the patient was readmitted to our institution for further evaluation and management.

Contrast-enhanced computed tomography (CT) of the nose and paranasal sinuses detected a lesion in the right maxillary sinus that appeared to have eroded through the anterior wall of the maxillary sinus (figure 1). A biopsy of the mass was performed at the bedside via a sublabial approach. On pathologic examination, the mass was determined to be a moderately differentiated invasive squamous cell carcinoma, and was staged as a T3NOMO squamous cell carcinoma of the right maxillary sinus. The patient then underwent an uncomplicated right subtotal maxillectomy with negative pathologic margins. Following surgery, the patient received adjuvant radiation therapy and has remained disease-free without epistaxis after nearly 2 years of follow-up.

Patient 2. A 36-year-old woman was admitted to the otolaryngology service of a major tertiary care teaching hospital for debridement and possible endoscopic biopsy of the right nasal cavity secondary to severe recurrent epistaxis. Her medical history was notable only for systemic lupus erythematosis. She had no bleeding diathesis. She had experienced several episodes of severe rightsided epistaxis and had previously gone to the emergency department of an outside hospital, where she had undergone nasal packing with Merocel sponges for a period of 3 days. Upon emergency-ward evaluation, she was found to have persistent posterior epistaxis. The nasal packing was removed and replaced with an Epistat balloon, resulting in adequate control of the epistaxis. She was then referred for, and subsequently underwent, angiographic embolization.

During the immediate postembolization period, she did well following the removal of the balloon, and she experienced no further epistaxis. However, after discharge, she experienced several more episodes of right-sided epistaxis, and she complained of a worsening nasal obstruction on the right side. On endoscopic examination of the right nasal cavity in the office, extensive crusting and fibrinoid exudate were noted, which prevented a thorough endoscopic visualization. The patient was followed with serial endoscopic debridement, but both the extensive crusting and exudate were difficult to clear.

The patient subsequently experienced several recurrent bouts of epistaxis from the right side. Initially, it was felt that the recurrent epistaxis was either secondary to a persistent necrotic turbinate (which was the result of the repeated nasal packing and embolization on the affected side) or, less likely, a neoplasm. CT with contrast showed a large enhancing soft-tissue mass in the right posterosuperior nasal cavity and attenuation of the adjacent bone, but no evidence of bony destruction (figure 2).

The patient was taken to the operating room for endoscopic biopsy of the right nasal mass. After an extensive debridement of organized thrombus and some granulation tissue, a large soft-tissue mass was visualized in the posterior nasal vault, medial to the superior and middle turbinates and seemingly attached to the septum. Upon removal of the biopsy specimen, the tumor began to bleed excessively. The visualized portion of the tumor was debulked, and the bleeding was controlled (with some difficulty) by postoperative nasal packing. Pathologic examination revealed that the mass was a lobular capillary hemangioma that was marked by extensive ulceration, hemorrhage, and necrosis. The patient was scheduled for a complete tumor removal.

In the operating room, all nasal packing was removed and the patient's right nasal cavity was examined endoscopically. The highly vascular lesion was identified medial to the superior and middle turbinates, and it was indeed attached to the septum. Endoscopic ligation of the sphenopalatine artery was performed, and the tumor was carefully debulked with suction cautery and bipolar forceps. Once the tumor's site of origin could be seen, a septal flap was initiated anterior to the mass, which lifted the tumor off the vomer and delivered it in its entirety with good mucosal margins. There were no complications, and the patient tolerated the procedure well. She has had no further episodes of epistaxis after 18 months of follow-up.

Discussion

The role of angiographic embolization as a treatment for intractable epistaxis has become increasingly important since the technique was first described by Sokoloff et al in 1974. (2) As our experience with this modality has increased and the technology has evolved, complication rates have fallen and success rates have correspondingly improved. (3) The technique has even been recommended as the primary treatment for cases of severe epistaxis. (4)

In the two cases described here, both patients initially went to outside emergency departments for treatment of severe epistaxis. After both failed treatment with posterior nasal packing, they were referred to a major tertiary care teaching hospital specifically for angiography and embolization as the primary treatment. Following embolization, both patients were followed up by the otolaryngology service. The cause of the epistaxis in each case was not appreciated on endoscopic examination because of the extreme distortion of the nasal anatomy, which had been caused by both the packing and the embolization procedure.

Well-known complications of nasal packing include pressure necrosis of the nasal septum, alar rim, and columella. (5) In addition, the presence of eschar, granulation tissue, or even infection after packing can cause a distortion of the normal anatomy of the nasal cavity. For these reasons, it can be difficult to appreciate the presence of an intranasal tumor by either simple visual inspection or office endoscopic examination.

Apart from nasal packing, angiography with embolization can also cause mucosal necrosis and slough, leading to the formation of additional granulation tissue and delayed healing. Manipulation of granulation tissue and intranasal crusting is often avoided in the postembolization state for fear of spurring recurrent bleeding. Our two patients received appropriate otolaryngologic follow-up after embolization, but their intranasal tumors were difficult to appreciate because of the intranasal changes caused by packing and embolization. Such a scenario presents a considerable challenge to otolaryngologists, who often see patients with recurrent epistaxis only after they have failed treatment with nasal packing and angiographic embolization.

Neoplasms are a rare cause of epistaxis in general, and in cases of severe recurrent epistaxis, they represent only 1 to 9% of identifiable causes. (4,6) Because as many as 70 to 86% of cases of epistaxis can be of idiopathic origin, further diagnostic investigation of these patients has typically not been pursued initially, and they are generally referred for angiographic embolization. (7-9)

The initial treatment of epistaxis in the emergency department is focused on the immediate control of the hemorrhage. Once placed, nasal packs are usually left in the nose for several days, obviating an initial comprehensive (endoscopic) visual inspection of the nasal cavity. With failure rates ranging from 26 to 52% for posterior epistaxis treated with packing, rebleeding might not necessarily trigger an intensive effort to identify a specific etiology. (4) By the time a patient has been referred to the otolaryngologist after repeated attempts at packing, it can be very difficult to promptly identify the presence of a tumor.

In our two cases, once the initial emergency-room evaluation and treatment failed to control the epistaxis, both patients were referred for angiographic embolization. In both cases, the angiographic procedures failed to definitively control the epistaxis or reveal the presence of the tumors. The overall failure rate of angiographic embolization for intractable epistaxis has been reported to range from 11 to 25%, (8,9) and recurrent epistaxis following angiographic embolization has been reported to occur anywhere from 1 to 12 months after the initial procedure. (6,10) Most recurrences are probably the result of an inability to embolize feeding vessels from the ethmoidal arteries out of concern for possible reflux of embolic material into the internal carotid circulation. (10) Another potential cause of failure is that inappropriate angiogenesis might continue to provide vessels with a predilection to hemorrhage, especially in cases that have a neoplastic origin. (11)

It is important to note that the angiographic procedure itself failed to reveal the presence of the tumors in both of the cases described here. At the initiation of the procedure, once the catheter tip has been advanced to the external carotid system, diagnostic angiography is undertaken to confirm the anatomy and visualize potential bleeding vessels. (12) Once this diagnostic procedure has been accomplished, embolization proceeds. Because the presence of either an acute inflammatory or neoplastic process can increase blood flow to the site, cause dilation of feeding vessels, and produce parenchymal staining, differentiation between inflammation and neoplasia can be difficult or impossible. (13) A characteristic tumor blush, caused by the radiopaque dye filling the tortuous microvasculature of the tumor parenchyma, is considered diagnostic for the presence of the tumor. However, the lack of a discernible blush is not necessarily proof of the absence of a tumor. In both of our cases, no tumor blush was apprec iated prior to embolization, and only after contrast-enhanced cross-sectional imaging was obtained were the tumors identified.

Because of the intense inflammation and distortion of the normal nasal anatomy that occurs following nasal packing and angiography with embolization, endoscopic inspection of the nasal cavity can be delayed and might fail to reveal the presence of an occult neoplastic process. Patients with severe epistaxis who are treated primarily with embolization must be followed closely with serial examinations and cross-sectional imaging to exclude neoplastic etiologies. In cases of continued epistaxis despite embolization, a neoplasm should be suspected as the cause until proven otherwise.

References

(1.) Tariq M. Gluckman P, Thebe P. Metastatic testicular teratoma of the nasal cavity: A rare cause of severe intractable epistaxis, J Laryngol Otol 1998;112:1078-81.

(2.) Sokoloff J, Wickbom I, McDonald D, et al. Therapeutic percutaneous embolization in intractable epistaxis. Radiology 1974;111:285-7.

(3.) Moreau S, De Rugy MG, Babin E, et al. Supraselective embolization in intractable epistaxis: Review of 45 cases. Laryngoscope 1998;108:887-8.

(4.) Elahi MM, Parnes LS, Fox AJ, et al. Therapeutic embolization in the treatment of intractable epistaxis. Arch Otolaryngol Head Neck Surg 1995;121:65-9.

(5.) Maves MD, Stevens C. Vascular tumors of the head and neck. In: Bailey BJ, ed. Head and Neck Surgery-Otolaryngology. 2nd ed. Philadelphia: Lippincott-Raven, 1998:1820-2.

(6.) Pollice PA, Yoder MG. Epistaxis: A retrospective review of hospitalized patients. Otolaryngol Head Neck Surg 1997;117:49-53.

(7.) Tseng EY, Narducci CA, Willing SJ, Sillers MJ. Angiographic embolization for epistaxis: A review of 114 cases. Laryngoscope 1998;108:615-9.

(8.) Leppanen M, Seppanen S, Laranne J, Kuoppala K. Microcatheter embolization of intractable idiopathic epistaxis. Cardiovase Intervent Radiol 1999;22:499-503.

(9.) Barlow DW, Deleyiannis WB, Pinczower EF. Effectiveness of surgical management of epistaxis at a tertiary care center. Laryngoscope 1997;107:21-4.

(10.) Riche MC, Chiras J, Melki JP, Merland JJ. [The role of embolisation in the treatment of severe epistaxis. A review of 54 cases], J Neuroradiol 1979;6:207-20.

(11.) Folkman J, Watson K, Ingber D, Hanahan D. Induction of angiogenesis during the transition from hyperplasia to neoplasia. Nature 1989;339:58-61.

(12.) Alvi A, Joyner-Triplett N. Acute epistaxis. How to spot the source and stop the flow. Postgrad Med 1996;99(5):83-90, 94-6.

(13.) Valji K. Vascular and Interventional Radiology. Philadelphia: W.B. Saunders, 1999:38-58.

From the Harvard Medical School, Boston (Dr. Ortiz); and the Division of Otolaryngology, Brigham and Women's Hospital, Boston, and the Department of Otology and Laryngology, Harvard Medical School (Dr. Bhattacharyya).

Reprint requests: Neil Bhattacharyya, MD, Division of Otolaryngology, 333 Longwood Ave., Boston, MA 02115. Phone: [phone]; fax: [phone].