Management of pterygoid venous plexus hemorrhage during resection of a large juvenile nasopharyngeal angiofibroma: a review of 27 cases.
We retrospectively reviewed the cases of 27 patients who experienced intraoperative bleeding during resection of a large (Fisch type III or IV) juvenile nasopharyngeal angiofibroma (JNA). Of this group, 16 patients had a type III JNA and 11 had a type IV tumor. The degree of hemorrhaging during excision of these JNAs varied greatly among individual patients. The amount of blood lost ranged from 200 to 5,000 ml (mean: 1,800) in the type III cases and from 700 to 8,000 ml (mean: 2,850) in the type IV cases. In 5 of these cases, both intraoperative observations and imaging data suggested that an important factor in the blood loss was damage to the pterygoid venous plexus (PVP). The PVP communicates with the cavernous sinus, ophthalmic vein, maxillary vein, and facial vein; no valve exists between these veins. In patients with a large JNA, the PVP is usually compressed by or adherent to the tumor. When a PVP is seriously damaged during removal of a JNA, hemorrhaging can be very profuse. Therefore, a suitable surgical approach and appropriate hemostatic procedures should be used to prevent or manage PVP hemorrhage as effectively as possible. We also describe in greater detail 5 typical cases of JNA excision that did (n = 3) and did not (n = 2) involve PVP damage.
In recent years, we began to study whether damage to the pterygoid venous plexus (PVP) during the resection of large juvenile nasopharyngeal angiofibroma (JNAs) in the infratemporal fossa is an important factor in blood loss. Our analyses were based on intraoperative observations and advanced imaging methods. The key to success in removing a large JNA is effective control of bleeding. The PVP in the infratemporal fossa plays a crucial role in blood loss.
We describe a series of cases of profuse bleeding during the removal of large JNAs. To the best of our knowledge, no similar study has previously examined this issue.
Patients and methods
From October 9, 1981, through June 19, 2009, a total of 27 large JNAs were treated at our hospital. The 27 patients were made up of 25 males and 2 females, aged 8.5 to 44 years (mean: 19.1).
The designation of a JNA as "large" was based on the classification criteria developed by Fisch. (1) A total of 16 patients had a Fisch type III JNA, and 11 had a Fisch type IV JNA.
In addition to demographic data, we collected information on the amount of blood loss, specific causes of hemorrhage, and other factors. Our hospital's Institutional Review Board approved the use of these data for clinical reference.
In the 27 cases, the degree of hemorrhaging during excision of the large JNAs varied greatly among individual patients. The amount of blood loss ranged from 200 to 5,000 ml (mean: 1,800) in the type III cases and from 700 to 8,000 ml (mean: 2,850) in the type IV cases.
The differences in the amount of blood loss were related to several factors, including tumor size, embolization of the tumor-supplying branches of the external carotid artery (ECA), the type of surgical management, the duration of surgery, and damage to the PVP.
Here we review the clinical findings and imaging data in 5 typical cases. Patients 1, 2, and 3 had experienced PVP damage, and the estimated s amount of blood loss per t procedure was 2,500 to 8,000 ml. Patients 4 and 5 did not sustain any damage to the PVP, and they lost only 800 and 700 ml, respectively.
Patient 1. A 22-year-old man was admitted to our hospital with a right-sided JNA of 3 years' duration. He had previously undergone two operations elsewhere, both via a transpalatine approach, and neither had been successful. Magnetic resonance imaging (MRI) (figure 1, A) and computed tomography (CT) demonstrated the presence of a large tumor that involved the pterygopalatine foramen, pterygomaxillary fissure, infratemporal fossa, cheek area, sphenoid sinus, and orbital apex.
The tumor was removed via a combined antrum and infratemporal fossa-nasal cavity approach through an extended Caldwell-Luc incision (midfacial degloving incision). The semisolid tumor portion in the cheek area was easily separated. However, the tumor in the infratemporal fossa was difficult to separate from the surrounding tissue because of scar formation that had developed after the two previous operations. Therefore, the surgeon performed a quick and forceful dissection, after which dark-red venous blood immediately flowed profusely into the surgical cavity.
To control the bleeding, petroleum jelly gauze was applied to the area. The total amount of blood loss was 8,000 ml. Based on our intraoperative observations and postoperative imaging, we believe that PVP damage might have been the major cause of the hemorrhage in this case.
The residual portion of the recurrent tumor (figure 1, B) in the pterygoid canal and infratemporal fossa was excised during a follow-up operation. MRI confirmed complete removal. While removing the small, hard, tail-like portion of the residual tumor from the infratemporal fossa (figure 1, B inset), the surgeon used a curette for a quick and forceful dissection. To our surprise, dark-red blood again started flowing. This time the total blood loss was 3,000 ml.
We concluded that the previously damaged PVP that had adhered to the tumor was the main factor resulting in the serious hemorrhage. At follow-up more than 2 years later, the patient was free of recurrence.
Patient 2. A 15-year-old boy presented with a recurrent JNA and epistaxis. Two years earlier, he had been treated with surgery and radiotherapy at a dose of 33.9 Gy. CT angiography showed an eroded petrous apex and the cavernous sinus portion of the internal carotid artery (ICA) on the left side (figure 2). The corresponding contralateral structures were normal (figure 2 inset). The PVP in the infratemporal fossa was difficult to differentiate from the tumor because of damage that had been caused during a previous operation. Three-dimensional digital subtraction angiography (DSA) revealed that the blood supply to the tumor was mainly from the ECA; some blood was also provided by the ICA by way of the ophthalmic artery and small transmitting branches of the cavernous sinus portion.
After embolization of the tumor-feeding vessels from the ECA, surgery was performed through a previous lateral rhinotomy. During separation of the tumor lobule in the sphenoid sinus, bright-red blood spurted from the lateral and inferior walls of the left sphenoid sinus. We believe that this bleeding might have emanated from the small ICA branches that supplied the tumor. Bipolar coagulation was used to stop the bleeding, and petroleum jelly strips were applied to maintain hemostasis.
The procedure resumed once the bleeding was controlled. In the infratemporal fossa, separation of the tumor from the surrounding tissue was difficult because of scar tissue that had formed after a previous surgery. Unfortunately, at the end of the operation, the surgeon neglected to remove the petroleum jelly strips slowly and carefully and, as a result, many hemostatic points were apparently stretched and broken. Consequently, a large amount of dark-red blood, which most likely represented a hemorrhage of the PVP, immediately flowed into the surgical cavity. A petroleum jelly gauze pocket was inserted into the surgical cavity, and this was followed by insertion of iodoform strips into the pocket to prevent postoperative hemorrhage and infection. The total amount of intraoperative blood loss was 2,500 ml.
Again, based on our intraoperative observations and the CT angiography, we believe that damage to the PVP might have been the main factor in the blood loss. The tumor was removed completely, and the patient remained healthy during follow-up.
Patient 3. A 13-year-old boy presented with a 2-year history of right-sided relapsing epistaxis and nasal obstruction and a 6-month history of progressive right-sided swelling of the cheek and vision loss. In addition to the swelling and proptosis, physical examination detected the presence of a mass in the nasopharynx and right nasal cavity. CT (figure 3) and MRI demonstrated a lobulated mass in the nasopharynx, right nasal cavity, infratemporal fossa, cheek area, and sphenoid sinus. The bones of the sella turcica and orbit were extensively eroded, and the cavernous sinus and ICA were compressed and displaced upward.
Two days after embolization of the tumor-feeding vessels from the ECA, a transtemporal intradural exploration was performed. No perforation of the dura matter or intradural tumor lobule could be found. Next, a combined antrum and infratemporal fossa-nasal cavity approach via an extended Caldwell-Luc incision was used to remove the tumor. The tumor lobe in the cheek area was easily separated from the surrounding tissue. During separation of the tumor in the infratemporal fossa, the surgeon performed a rapid and forceful dissection. Immediately, a large amount of dark-red venous blood flowed into the surgical cavity.
There was no way to check and treat the bleeding points, so it was necessary to stop the operation. The total amount of blood loss was 4,000 ml. According to our intraoperative observations, PVP damage might have been the most important factor in the hemorrhage.
Postoperative imaging showed tumor residue in the nasal cavity and sphenoid sinus, and another operation was undertaken to remove the tumor completely (figure 3 inset). At 5 years of follow-up, the patient remained free of recurrence.
Patient 4. A 17-year-old boy with a 1-year history of nasal obstruction on the left side was admitted to our hospital. CT indicated that a tumor had spread into the left sphenoid sinus, pterygopalatine fossa, pterygomaxillary fissure, and infratemporal fossa (figure 4, A). DSA indicated that small branches arising from the cavernous part of the left ICA were supplying blood to the tumor. The tumor-feeding vessels from the ECA were embolized, and 2 days later the mass was removed via a combined antrum and infratemporal fossa-nasal cavity approach through an extended Caldwell-Luc incision. The small transmitting branches of the cavernous part of the left ICA that supplied blood to the tumor were coagulated near the tumor surface. The tumor lobule in the infratemporal fossa was separated under direct visualization without damage to the PVP in the fatty pad. The 5.5 x 5.0 x 4.5-cm tumor was removed en bloc (figure 4, B inset).
Postoperative CT showed that the PVP and the lateral and internal pterygoid muscles were intact (figure 4, B). The intraoperative blood loss was only 800 ml. The patient was in good health at the 3-year follow-up examination.
Patient 5. A 16-year-boy with right-sided relapsing epistaxis and progressive nasal obstruction of 2 years' duration presented for evaluation of a 2-week history of right-sided severe headache and vision loss. MRI identified a large, lobulated tumor with extensive spread into the infratemporal fossa (figure 5, A). MRI also showed that the posterior wall of the maxillary sinus was compressed forward.
The tumor-feeding vessels from the ECA were embolized, and 3 days later a combined antrum and infratemporal fossa-nasal cavity approach through an extended Caldwell-Luc incision was used to remove the tumor under direct visualization and with careful dissection. The intraoperative blood loss was 700 ml.
The tumor, which measured 6 x 7 x 8 cm, was removed completely (figure 5, B inset). A postoperative MRI revealed no surgical damage to the PVP or lateral pterygoid muscle (figure 5, B). The patient remained healthy following the operation.
The PVP in the infratemp oral fossa lies partly between the temporalis and lateral pterygoid muscles and the lateral and internal pterygoid muscles. The PVP is embedded in a fatty pad, and it communicates with the deep facial vein, maxillary vein, and cavernous sinuses. No valve exists between any of these veins, and they communicate freely with each other. (2,3) In our limited experience, when a PVP is severely damaged, the venous blood from the intracranial and facial region flows profusely into the surgical field.
In 2007, Yang described a patient who had a squamous cell carcinoma in the maxillary sinus that had caused extensive bone erosion. (4) During removal of the lesion at the site of the temporal muscle and lateral pterygoid muscle, a large amount of dark-red blood from various sites flowed profusely into the surgical cavity. Intraoperative observations indicated that the bleeding represented a hemorrhage of the PVP in the infratemporal fossa.
In 1999, Guinto et al described their use of the zygomatic-transmandibular approach in the treatment of 10 patients with massive tumors that were localized in the infratemporal fossa and parapharyngeal space. (5) On drilling into the middle fossa, they noted profuse bleeding around the third vertebra, which was caused by the numerous veins in the pterygoid plexus around the superior head of the lateral pterygoid muscle.
Several factors can affect intraoperative blood loss during the removal of JNAs. Liu et al reported that the patient's age, the duration of surgery, and the staging of the lesion significantly influence intraoperative bleeding. (6) They pointed out the difficulty of accurately determining the involvement of any single factor in blood loss. We agree that determining the exact amount of intraoperative blood loss at any site is difficult. Moreover, sites of tumor origin that are rich in blood vessels and small branches of the ICA that supply blood to the tumor might also be important factors in intraoperative blood loss.
To the best of our knowledge, no similar study of profuse PVP hemorrhage in INA resection has been previously reported in the literature. We believe that attention should be paid to this specific problem in clinical practice. Preoperative image analyses, including identification of the relationship between a tumor and a PVP, are important.
Serious PVP hemorrhages can generally be avoided, even when the PVP has been compressed by the tumor, if the dissection proceeds carefully along the pseudocapsule plane, such as was accomplished in our patients 4 and 5.
The risk of PVP hemorrhage can be reduced by following hemostatic measures such as tilting the head upward by 30[degrees], controlling hypotension, and using advanced fine hemostatic instruments. For a large JNA with extensive spread in a patient with an intact dura mater, a combined antrum and infratemporal fossa-nasal cavity approach through an extended Caldwell-Luc incision is a minimally invasive and suitable approach. (7,8) This approach allows for dissection of the tumor under direct visualization with the use of both hands, which avoids the need to use a finger or instrument for blind, rapid removal. With this approach, every bleeding point can be seen and treated carefully.
Based on our experience, if the separated portion of the tumor blocks the surgeon's vision and affects the management of the PVP, it is better to cut this portion after tightly suturing the cut margin. In this way, a wider operative field can be achieved for complete removal of the tumor via the use of delicate, effective hemostatic procedures.
(1.) Fisch U. The infratemporal fossa approach for nasopharyngeal tumors. Laryngoscope 1983;93(1):36-44.
(2.) Pickering PT, Howden R, eds. Gray's Anatomy. 15th ed. New York: Barnes & Noble; 1995:573-4.
(3.) Bannister LH, Berry MM, Collins P, et al, eds. Gray's Anatomy. The Anatomical Basis of Medicine and Surgery. 38th ed. New York: Churchill Livingstone; 1995:1577-8.
(4.) Yang Z. The doctors' valuable clinical experiences. In: Ping WJ, Yang W, eds. Otolaryngology-Head and Neck Surgery. Beijing: Military Medical Scientific Publishers; 2007:139-40.
(5.) Guinto G, Abello J, Molina A, et al. Zygomatic-transmandibular approach for giant tumors of the infratemporal fossa and parapharyngeal space. Neurosurgery 1999;45(6):1385-98.
(6.) Liu L, Wang R, Huang D, et al. Multiple factors analysis of intraoperative bleeding and recurrence of juvenile nasopharyngeal angiofibromas [in Chinese]. Zhonghua Er Bi Yah Hou Ke Za Zhi 2001;36(3):220-3.
(7.) Yi ZX, Li ZC, Cheng JM, et al. Huge nasopharyngeal angiofibroma with intracranial extension: Change in the dura mater and choice of surgical management. J Laryngol Otol 2007;121 (11):1108-12.
(8.) Lin C, Li ZC, Cheng JM, et al. Pathological features and clinical managements of nasopharyngeal angiofibroma [in Chinese]. Zhonghua Er Bi Yan Hou Tou Jing Wai Ke Za Zhi 2008;43(10):736-6.
Lin Chang, MD; Yi Zixiang, MD; Fang Zheming, MD, Lin Gongbiao, MD; Li Zhichun, MD; Zhang Rong, MD; Zhou Aidong, MD; Lan Shuzhan, MD
From the Department of Otolaryngology (Dr. Chang, Dr. Zixiang, Dr. Gongbiao, Dr. Zhichun, Dr. Rong, Dr. Aidong, and Dr. Shuzhan) and the Department of Imaging (Dr. Zheming), First Affiliated Hospital of Fujian Medical University, Fuzhou, People's Republic of China.
Corresponding author: Yi Zixiang, MD, Department of Otolaryngology, First Affiliated Hospital of Fujian Medical University, 20 Chazhong Rd., Fuzhou 350005, People's Republic of China. Email: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||ORIGINAL ARTICLE|
|Author:||Chang, Lin; Zixiang, Yi; Zheming, Fang; Gongbiao, Lin; Zhichun, Li; Rong, Zhang; Aidong, Zhou; Shuzh|
|Publication:||Ear, Nose and Throat Journal|
|Article Type:||Clinical report|
|Date:||Apr 1, 2013|
|Previous Article:||Chondromyxoid fibrorna of the mastoid portion of the temporal bone: MRI and PET/CT findings and their correlation with histology.|
|Next Article:||Propranolol in the treatment of upper airway hemangiomas.|