570936
research-article2015
AORXXX10.1177/0003489415570936Annals of Otology, Rhinology & LaryngologyMarston et al
Case Report
A Unique Case of Bilateral Recurrent Sphenoid Sinus Cerebrospinal Fluid Leaks: Primary Acquired Leak Within the Lateral Sphenoid Sinus Recess, Followed by a Leak via Sternberg’s Canal
Annals of Otology, Rhinology & Laryngology 1–5 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0003489415570936 aor.sagepub.com
Alexander P. Marston, MD1, Jamie J. Van Gompel, MD2, Matthew L. Carlson, MD1, and Erin K. O’Brien, MD1
Abstract Objectives: This case presents a previously undescribed clinical scenario of spontaneous cerebrospinal fluid (CSF) leaks secondary to a lateral sphenoid sinus recess skull base dehiscence and contralateral Sternberg’s canal. This case report aims to characterize the presentation and successful management of these lesions. Methods: The electronic medical record was used to collect information pertaining to the patient’s clinical history. Results: The patient was a middle-aged, obese female with persistent clear rhinorrhea as her only presenting symptom. Neuroradiologic studies localized the defect to the lateral sphenoid sinus recess. CSF opening pressures were within normal limits, but radiographic findings were consistent with elevated intracranial pressure. After an endoscopic transnasal transsphenoidal approach failed to resolve the CSF leak, a transpterygoid approach facilitated CSF leak resolution. The patient then did well for the following 2 years, but later developed a CSF leak through a contralateral Sternberg’s canal. An endoscopic suprapterygoid procedure and ventriculoperitoneal shunt placement led to CSF leak resolution. Conclusion: This case demonstrates one of the only published examples of a sphenoid sinus CSF leak secondary to Sternberg’s canal as it was originally described in the literature. Wide endoscopic surgical exposure and intracranial pressure management ultimately led to CSF leak resolution. Keywords spontaneous cerebrospinal fluid leak, Sternberg’s canal, idiopathic intracranial hypertension, transpterygoid, lateral sphenoid sinus recess
Introduction Cerebrospinal fluid (CSF) leaks can be classified as idiopathic or secondary to iatrogenic, traumatic, inflammatory, neoplastic, and congenital etiologies. Idiopathic, or spontaneous, CSF leaks commonly occur in the setting of elevated intracranial pressure with a high percentage of these patients meeting diagnostic criteria for idiopathic intracranial hypertension (IIH).1 Although middle cranial fossa skull base dehiscences of the sphenoid sinus are rare, it is reported that the lateral sphenoid sinus is the most common site of spontaneous CSF leak in patients with intracranial hypertension.2 Sternberg’s canal is a lateral craniopharyngeal canal remnant and a rare site of encephalocele formation and CSF leak located medial to the foramen rotundum within the sphenoid sinus.3 The sphenoid bone arises from the ossification of multiple cartilaginous components and a lateral craniopharyngeal canal may persist in the setting of incomplete fusion
between the presphenoid and basisphenoid plates with the greater wing of the sphenoid.4 Although rarely symptomatic, Sternberg’s canal was originally described in up to 4% of adults.5 However, a more recent anatomical assessment of normal controls revealed that the incidence of Sternberg’s canal may be even more infrequent.3 Computed tomography (CT) and magnetic resonance (MR) imaging with and without cisternography can be used to identify the precise location of the CSF leak. Surgical repair is the gold standard of treatment for CSF leaks of the 1
Department of Otorhinolaryngology, Mayo Clinic, Rochester, Minnesota, USA 2 Department of Neurosurgery, Mayo Clinic, Rochester, Minnesota, USA Corresponding Author: Erin K. O’Brien, MD, Department of Otorhinolaryngology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905, USA. Email: [email protected]
Downloaded from aor.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 14, 2015
2
Annals of Otology, Rhinology & Laryngology
Figure 1. Coronal CT image demonstrating the lateral sphenoid sinus recess (star) and pneumatized pterygoid process (plus sign). The left-sided Sternberg’s canal (thin arrow) with an associated encephalocele is located medial to foramen rotundum (block arrow). Vidian canal is depicted with an arrowhead.
sphenoid sinus. Options for surgical repair include transcranial, transfacial, and endoscopic transnasal approaches. The excellent visualization and limited morbidity achieved with endoscopic approaches has led to the increased utilization of these techniques. Spontaneous CSF leaks are reported to have the highest recurrence rate when compared to all other etiologies of CSF leak2, and long-term treatment success is dependent on a robust surgical repair and control of intracranial pressure postoperatively. The case presented in this report describes a spontaneous CSF leak within the lateral sphenoid sinus followed by a contralateral leak via Sternberg’s canal.
Case Description The patient was a 48-year-old healthy female with a body mass index (BMI) of 26 kg/m2 who presented with unilateral right-sided clear rhinorrhea. She denied headache, vision changes, and balance difficulty. Biochemical analysis with beta-2 transferrin testing confirmed the presence of CSF. CT imaging showed extensive pneumatization into the lateral sphenoid sinus recesses and pterygoid processes bilaterally (Figure 1). CT cisternography localized an encephalocele and active CSF leak to the right lateral sphenoid sinus recess (Figure 2). In addition, there was a left lateral sphenoid sinus skull base dehiscence medial to foramen rotundum consistent with a Sternberg’s canal (Figures 1, 3, and 4). MR imaging demonstrated an empty sella and bilaterally dilated Meckel’s caves consistent with a history of elevated intracranial pressure. Despite these radiographic signs, the opening pressure at the time of CT cisternography was found to be 17 cm H2O. After lumbar drain placement, an endoscopic transnasal transsphenoidal approach was used for reduction of the right-sided encephalocele and
Figure 2. Axial CT cisternogram image depicting the rightsided lateral sphenoid sinus recess encephalocele and active CSF leak.
repair of the CSF leak by overlay closure with fat and fascia lata. This initially led to CSF leak resolution; however, 3 months later the patient was found to have a recurrent leak at the same site. The revision surgical procedure was an endoscopic transnasal transpterygoid approach that allowed for improved visualization and more precise multilayered closure. Perioperative lumbar drains were utilized with the primary and secondary surgical repairs and after a 24 hour clamping trial, the drains were removed between the third and fifth postoperative day. Medical management of intracranial pressure was not administered after the primary repair. Following the revision procedure, the patient was instructed to take acetazolamide; however, the medication was discontinued after 3 months due to intolerance of the side effects. The patient did well for the following 2 years, but later developed a CSF leak from the Sternberg’s canal within the left sphenoid sinus. At this time, the patient’s BMI had increased to 30 kg/m2, but the CSF opening pressure was again within normal limits. Throughout the patient’s clinical course, multiple opening pressures were less than 20 cm H2O with all measurements obtained in a standardized fashion in the lateral decubitus position. Multilayered closure and resolution of the CSF leak was achieved with an endoscopic suprapterygoid approach. A ventriculoperitoneal shunt was then placed on postoperative day 5 due to the history of recurrent CSF leaks and suspicion of IIH.
Surgical Technique The endoscopic transnasal transpterygoid approach used in this case was performed with a dual surgeon, bimanual, and binostril technique. Intrathecal fluorescein was utilized to assist in identifying the skull base defect and confirm cessation of the CSF leak. A medial maxillectomy was completed
Downloaded from aor.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 14, 2015
3
Marston et al
was repaired with underlay adipose and fascia lata grafts. On the left, fat was placed using a “bath plug” technique within the Sternberg’s canal followed by several overlapping layers of fibrin-glue, nasal mucosal tissue, fascia, and fat.
Discussion
Figure 3. 3D coronal image exhibiting the Sternberg’s canal (thin arrow) located medial to foramen rotundum (block arrow). Vidian canal is shown at the arrowhead.
Figure 4. Oblique coronal image through Sternberg’s canal (thin arrow) situated medial to foramen rotundum (block arrow). Dotted line through the axial image at lower left demonstrates the coronal slice.
followed by widely opening the natural sphenoid ostium. The medial aspect of the encephalocele was encountered after completion of the sphenoidotomy; however, complete exposure of the lateral sphenoid sinus recess necessitated the transpterygoid approach. A skull base drill was used to drill the posterior maxillary sinus wall revealing the pterygopalatine fossa and its neurovascular contents. The maxillary artery was controlled with endoscopic clips and branches of the maxillary and vidian nerves were carefully preserved. The anterolateral wall of the sphenoid sinus was removed providing complete access to the lateral sphenoid sinus recess. The encephalocele was reduced and the defect
The published literature suggests that middle-aged, obese females with intracranial hypertension are at risk for development of spontaneous CSF leaks. In a study by Woodworth et al,2 46 of the 56 (77%) patients who underwent spontaneous CSF leak repair were women and the average age at presentation was 61 years. Of the patients in the study, 82% were obese (BMI >30 kg/m2) and 45 of 48 (94%) with recorded lumbar drain pressures had evidence of intracranial hypertension. Due to the demographic characteristics of this patient population in the setting of obesity and elevated intracranial pressure, the authors postulate that spontaneous CSF leaks represent a variant of IIH. Schlosser and colleagues1 investigated 11 patients with spontaneous CSF leaks and found that 8 (72%) met the strict requirements for IIH by the modified Dandy criteria, with the remaining 3 demonstrating a high degree of clinical suspicion for the disease. The patient in the present case is similar in that she is a middle-aged, overweight female. Although our patient’s opening pressures were consistently less than 20 cm H2O, the radiographic signs, such as an empty sella, and the recurrent nature of the CSF leaks are likely consistent with a diagnosis of IIH. In addition to the risk factors secondary to the patient’s sex, age, and body habitus, the structural variation of the sphenoid sinus likely predisposed her to CSF leak development. In the present case, pneumatization of the pterygoid process and lateral aspect of the sphenoid sinus likely resulted in an inherently weak and vulnerable skull base. In the setting of even normal CSF pressure spikes, the cumulative forces exerted on the skull base may have led to progressive thinning, arachnoid pit development, skull base dehiscence, and eventual CSF leak.6,7 In a study by Shetty et al6 investigating the sinus anatomy of patients with a history of spontaneous sphenoid sinus CSF leaks, the authors describe that 10 of 11 (91%) patients had extensive pneumatization of the lateral recess of the sphenoid sinus compared to 23 of 100 (23%) age-matched controls. With 9 of 11 (82%) CSF leaks occurring laterally within the extensively pneumatized sphenoid sinus, the Shetty et al group theorized that the lateral sphenoid sinus recess, when present, is most vulnerable to the vector forces of elevated intracranial pressure. Sphenoid sinus CSF leaks through Sternberg’s canal have been reported; however, the incidence with which this canal is present in the general population and its precise anatomic description is of recent debate. Per the original anatomic
Downloaded from aor.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 14, 2015
4
Annals of Otology, Rhinology & Laryngology
studies published by Sternberg8 in the late 19th century, the canal was thought to be present in approximately 4% of adults. A recent study by Barañano et al3 reinvestigated this estimate by examining 1000 high-resolution CT scans. Using the definition that Sternberg’s canal must be medial to both the superior orbital fissure and foramen rotundum, only 1 of the 1000 patients was identified to have a Sternberg’s canal. In a report by Illing et al,9 59 patients with a history of 77 spontaneous lateral sphenoid sinus recess CSF leaks were evaluated and all were found to have skull base defects located lateral to V2. Due to the infrequency with which Sternberg’s canal was identified and the fact that prior Sternberg’s canal series have included skull base dehiscences lateral to foramen rotundum,7,10 it is now believed that the etiology of spontaneous sphenoid sinus CSF leak is primarily due to de novo skull base dehiscence in the setting of elevated intracranial pressure.2,3,9 With the assumption that Sternberg’s canal must be located lateral to V2, this case demonstrates one of the only published examples of a symptomatic Sternberg’s canal apart from the report authored by Schick et al.11 The absence of elevated CSF opening pressures may have been due to the fact that increased intracranial pressure resolved by way of the active CSF leak.1 Alternatively, the CSF opening pressure measurements may not have captured an accurate representation of the patient’s intracranial pressure. This was demonstrated in a study by Reh et al12 using continuous CSF pressure monitoring in which a subset of spontaneous CSF leak patients with mean CSF pressures
research-article2015
AORXXX10.1177/0003489415570936Annals of Otology, Rhinology & LaryngologyMarston et al
Case Report
A Unique Case of Bilateral Recurrent Sphenoid Sinus Cerebrospinal Fluid Leaks: Primary Acquired Leak Within the Lateral Sphenoid Sinus Recess, Followed by a Leak via Sternberg’s Canal
Annals of Otology, Rhinology & Laryngology 1–5 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0003489415570936 aor.sagepub.com
Alexander P. Marston, MD1, Jamie J. Van Gompel, MD2, Matthew L. Carlson, MD1, and Erin K. O’Brien, MD1
Abstract Objectives: This case presents a previously undescribed clinical scenario of spontaneous cerebrospinal fluid (CSF) leaks secondary to a lateral sphenoid sinus recess skull base dehiscence and contralateral Sternberg’s canal. This case report aims to characterize the presentation and successful management of these lesions. Methods: The electronic medical record was used to collect information pertaining to the patient’s clinical history. Results: The patient was a middle-aged, obese female with persistent clear rhinorrhea as her only presenting symptom. Neuroradiologic studies localized the defect to the lateral sphenoid sinus recess. CSF opening pressures were within normal limits, but radiographic findings were consistent with elevated intracranial pressure. After an endoscopic transnasal transsphenoidal approach failed to resolve the CSF leak, a transpterygoid approach facilitated CSF leak resolution. The patient then did well for the following 2 years, but later developed a CSF leak through a contralateral Sternberg’s canal. An endoscopic suprapterygoid procedure and ventriculoperitoneal shunt placement led to CSF leak resolution. Conclusion: This case demonstrates one of the only published examples of a sphenoid sinus CSF leak secondary to Sternberg’s canal as it was originally described in the literature. Wide endoscopic surgical exposure and intracranial pressure management ultimately led to CSF leak resolution. Keywords spontaneous cerebrospinal fluid leak, Sternberg’s canal, idiopathic intracranial hypertension, transpterygoid, lateral sphenoid sinus recess
Introduction Cerebrospinal fluid (CSF) leaks can be classified as idiopathic or secondary to iatrogenic, traumatic, inflammatory, neoplastic, and congenital etiologies. Idiopathic, or spontaneous, CSF leaks commonly occur in the setting of elevated intracranial pressure with a high percentage of these patients meeting diagnostic criteria for idiopathic intracranial hypertension (IIH).1 Although middle cranial fossa skull base dehiscences of the sphenoid sinus are rare, it is reported that the lateral sphenoid sinus is the most common site of spontaneous CSF leak in patients with intracranial hypertension.2 Sternberg’s canal is a lateral craniopharyngeal canal remnant and a rare site of encephalocele formation and CSF leak located medial to the foramen rotundum within the sphenoid sinus.3 The sphenoid bone arises from the ossification of multiple cartilaginous components and a lateral craniopharyngeal canal may persist in the setting of incomplete fusion
between the presphenoid and basisphenoid plates with the greater wing of the sphenoid.4 Although rarely symptomatic, Sternberg’s canal was originally described in up to 4% of adults.5 However, a more recent anatomical assessment of normal controls revealed that the incidence of Sternberg’s canal may be even more infrequent.3 Computed tomography (CT) and magnetic resonance (MR) imaging with and without cisternography can be used to identify the precise location of the CSF leak. Surgical repair is the gold standard of treatment for CSF leaks of the 1
Department of Otorhinolaryngology, Mayo Clinic, Rochester, Minnesota, USA 2 Department of Neurosurgery, Mayo Clinic, Rochester, Minnesota, USA Corresponding Author: Erin K. O’Brien, MD, Department of Otorhinolaryngology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905, USA. Email: [email protected]
Downloaded from aor.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 14, 2015
2
Annals of Otology, Rhinology & Laryngology
Figure 1. Coronal CT image demonstrating the lateral sphenoid sinus recess (star) and pneumatized pterygoid process (plus sign). The left-sided Sternberg’s canal (thin arrow) with an associated encephalocele is located medial to foramen rotundum (block arrow). Vidian canal is depicted with an arrowhead.
sphenoid sinus. Options for surgical repair include transcranial, transfacial, and endoscopic transnasal approaches. The excellent visualization and limited morbidity achieved with endoscopic approaches has led to the increased utilization of these techniques. Spontaneous CSF leaks are reported to have the highest recurrence rate when compared to all other etiologies of CSF leak2, and long-term treatment success is dependent on a robust surgical repair and control of intracranial pressure postoperatively. The case presented in this report describes a spontaneous CSF leak within the lateral sphenoid sinus followed by a contralateral leak via Sternberg’s canal.
Case Description The patient was a 48-year-old healthy female with a body mass index (BMI) of 26 kg/m2 who presented with unilateral right-sided clear rhinorrhea. She denied headache, vision changes, and balance difficulty. Biochemical analysis with beta-2 transferrin testing confirmed the presence of CSF. CT imaging showed extensive pneumatization into the lateral sphenoid sinus recesses and pterygoid processes bilaterally (Figure 1). CT cisternography localized an encephalocele and active CSF leak to the right lateral sphenoid sinus recess (Figure 2). In addition, there was a left lateral sphenoid sinus skull base dehiscence medial to foramen rotundum consistent with a Sternberg’s canal (Figures 1, 3, and 4). MR imaging demonstrated an empty sella and bilaterally dilated Meckel’s caves consistent with a history of elevated intracranial pressure. Despite these radiographic signs, the opening pressure at the time of CT cisternography was found to be 17 cm H2O. After lumbar drain placement, an endoscopic transnasal transsphenoidal approach was used for reduction of the right-sided encephalocele and
Figure 2. Axial CT cisternogram image depicting the rightsided lateral sphenoid sinus recess encephalocele and active CSF leak.
repair of the CSF leak by overlay closure with fat and fascia lata. This initially led to CSF leak resolution; however, 3 months later the patient was found to have a recurrent leak at the same site. The revision surgical procedure was an endoscopic transnasal transpterygoid approach that allowed for improved visualization and more precise multilayered closure. Perioperative lumbar drains were utilized with the primary and secondary surgical repairs and after a 24 hour clamping trial, the drains were removed between the third and fifth postoperative day. Medical management of intracranial pressure was not administered after the primary repair. Following the revision procedure, the patient was instructed to take acetazolamide; however, the medication was discontinued after 3 months due to intolerance of the side effects. The patient did well for the following 2 years, but later developed a CSF leak from the Sternberg’s canal within the left sphenoid sinus. At this time, the patient’s BMI had increased to 30 kg/m2, but the CSF opening pressure was again within normal limits. Throughout the patient’s clinical course, multiple opening pressures were less than 20 cm H2O with all measurements obtained in a standardized fashion in the lateral decubitus position. Multilayered closure and resolution of the CSF leak was achieved with an endoscopic suprapterygoid approach. A ventriculoperitoneal shunt was then placed on postoperative day 5 due to the history of recurrent CSF leaks and suspicion of IIH.
Surgical Technique The endoscopic transnasal transpterygoid approach used in this case was performed with a dual surgeon, bimanual, and binostril technique. Intrathecal fluorescein was utilized to assist in identifying the skull base defect and confirm cessation of the CSF leak. A medial maxillectomy was completed
Downloaded from aor.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 14, 2015
3
Marston et al
was repaired with underlay adipose and fascia lata grafts. On the left, fat was placed using a “bath plug” technique within the Sternberg’s canal followed by several overlapping layers of fibrin-glue, nasal mucosal tissue, fascia, and fat.
Discussion
Figure 3. 3D coronal image exhibiting the Sternberg’s canal (thin arrow) located medial to foramen rotundum (block arrow). Vidian canal is shown at the arrowhead.
Figure 4. Oblique coronal image through Sternberg’s canal (thin arrow) situated medial to foramen rotundum (block arrow). Dotted line through the axial image at lower left demonstrates the coronal slice.
followed by widely opening the natural sphenoid ostium. The medial aspect of the encephalocele was encountered after completion of the sphenoidotomy; however, complete exposure of the lateral sphenoid sinus recess necessitated the transpterygoid approach. A skull base drill was used to drill the posterior maxillary sinus wall revealing the pterygopalatine fossa and its neurovascular contents. The maxillary artery was controlled with endoscopic clips and branches of the maxillary and vidian nerves were carefully preserved. The anterolateral wall of the sphenoid sinus was removed providing complete access to the lateral sphenoid sinus recess. The encephalocele was reduced and the defect
The published literature suggests that middle-aged, obese females with intracranial hypertension are at risk for development of spontaneous CSF leaks. In a study by Woodworth et al,2 46 of the 56 (77%) patients who underwent spontaneous CSF leak repair were women and the average age at presentation was 61 years. Of the patients in the study, 82% were obese (BMI >30 kg/m2) and 45 of 48 (94%) with recorded lumbar drain pressures had evidence of intracranial hypertension. Due to the demographic characteristics of this patient population in the setting of obesity and elevated intracranial pressure, the authors postulate that spontaneous CSF leaks represent a variant of IIH. Schlosser and colleagues1 investigated 11 patients with spontaneous CSF leaks and found that 8 (72%) met the strict requirements for IIH by the modified Dandy criteria, with the remaining 3 demonstrating a high degree of clinical suspicion for the disease. The patient in the present case is similar in that she is a middle-aged, overweight female. Although our patient’s opening pressures were consistently less than 20 cm H2O, the radiographic signs, such as an empty sella, and the recurrent nature of the CSF leaks are likely consistent with a diagnosis of IIH. In addition to the risk factors secondary to the patient’s sex, age, and body habitus, the structural variation of the sphenoid sinus likely predisposed her to CSF leak development. In the present case, pneumatization of the pterygoid process and lateral aspect of the sphenoid sinus likely resulted in an inherently weak and vulnerable skull base. In the setting of even normal CSF pressure spikes, the cumulative forces exerted on the skull base may have led to progressive thinning, arachnoid pit development, skull base dehiscence, and eventual CSF leak.6,7 In a study by Shetty et al6 investigating the sinus anatomy of patients with a history of spontaneous sphenoid sinus CSF leaks, the authors describe that 10 of 11 (91%) patients had extensive pneumatization of the lateral recess of the sphenoid sinus compared to 23 of 100 (23%) age-matched controls. With 9 of 11 (82%) CSF leaks occurring laterally within the extensively pneumatized sphenoid sinus, the Shetty et al group theorized that the lateral sphenoid sinus recess, when present, is most vulnerable to the vector forces of elevated intracranial pressure. Sphenoid sinus CSF leaks through Sternberg’s canal have been reported; however, the incidence with which this canal is present in the general population and its precise anatomic description is of recent debate. Per the original anatomic
Downloaded from aor.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 14, 2015
4
Annals of Otology, Rhinology & Laryngology
studies published by Sternberg8 in the late 19th century, the canal was thought to be present in approximately 4% of adults. A recent study by Barañano et al3 reinvestigated this estimate by examining 1000 high-resolution CT scans. Using the definition that Sternberg’s canal must be medial to both the superior orbital fissure and foramen rotundum, only 1 of the 1000 patients was identified to have a Sternberg’s canal. In a report by Illing et al,9 59 patients with a history of 77 spontaneous lateral sphenoid sinus recess CSF leaks were evaluated and all were found to have skull base defects located lateral to V2. Due to the infrequency with which Sternberg’s canal was identified and the fact that prior Sternberg’s canal series have included skull base dehiscences lateral to foramen rotundum,7,10 it is now believed that the etiology of spontaneous sphenoid sinus CSF leak is primarily due to de novo skull base dehiscence in the setting of elevated intracranial pressure.2,3,9 With the assumption that Sternberg’s canal must be located lateral to V2, this case demonstrates one of the only published examples of a symptomatic Sternberg’s canal apart from the report authored by Schick et al.11 The absence of elevated CSF opening pressures may have been due to the fact that increased intracranial pressure resolved by way of the active CSF leak.1 Alternatively, the CSF opening pressure measurements may not have captured an accurate representation of the patient’s intracranial pressure. This was demonstrated in a study by Reh et al12 using continuous CSF pressure monitoring in which a subset of spontaneous CSF leak patients with mean CSF pressures
