Surgical Management and Outcomes of Cavernous Sinus Hemangiomas: A Single-Institution Series of 47 Patients

Original Article

Surgical Management and Outcomes of Cavernous Sinus Hemangiomas: A Single-Institution Series of 47 Patients Zong-Hao Li1-5, Zhen Wu1-4, Jun-Ting Zhang1-4, Li-Wei Zhang1-4

OBJECTIVE: The purpose of this study was to analyze the outcomes of cavernous sinus hemangiomas (CSHs) treated surgically, and to investigate factors that affect the gross total resection (GTR), newly developed or deteriorated cranial nerve injury (NDDCNI), and follow-up neurologic performance, and to further discuss the optimal treatment for CSHs.

-

CONCLUSIONS: Being treated by an experienced skull base surgeon favors the total removal of CSHs, whereas the invasion of the sella turcica does just the opposite. Increased tumor size is a risk factor for unfavorable followup KPS score. The invasion of the sella turcica was related to NDDCNI and unfavorable follow-up KPS score.

-

METHODS: Clinical data of 47 patients with CSHs treated surgically at our institution between 2012e2018 were retrospectively reviewed.

-

RESULTS: GTR was achieved in 26 (55.3%) patients. Significant relations were identified between the invasion of the sella turcica (odds ratio [OR] [ 0.012; 95% confidence interval [CI], 0.001e0.213; P [ 0.002), skull base ward (OR [ 27.838; 95% CI, 2.995e258.748; P [ 0.003), and GTR. The preoperative Karnofsky Performance Scale (KPS) score (OR [ 2.966, per 10 score increase; 95% CI, 1.136e 7.743; P [ 0.026) and the invasion of the sella turcica (OR [ 7.137; 95% CI, 1.282e39.726; P [ 0.025) were factors that significantly affected the incidence of NDDCNI. The average follow-up KPS score, which increased significantly compared with the pre (P < 0.001) and postoperative KPS scores (P < 0.001), was 89.1. Increased tumor size (OR [ 0.044, per 1cm increase; 95% CI, 0.004e0.477; P [ 0.010) was a risk factor for unfavorable follow-up KPS score.

-

Key words Cavernous sinus hemangioma - Extent of resection - Morbidity - Risk factor - Surgical management -

Abbreviations and Acronyms CI: Confidence interval CN: Cranial nerve CSH: Cavernous sinus hemangioma EETS: Endoscopic endonasal transsphenoidal FT: Frontotemporal FTOZ: Frontotemporal orbitozygomatic FTZ: Frontotemporal zygomatic GKRS: Gamma Knife radiosurgery GTR: Gross total resection ICA: Internal carotid artery KPS: Karnofsky Performance Scale MRI: Magnetic resonance imaging

WORLD NEUROSURGERY -: e1-e14, - 2018

INTRODUCTION

C

avernous sinus hemangiomas (CSHs) are extremely rare extra-axial vascular neoplasms arising within the confines of the cavernous sinus. The lesion accounts for approximately 3% of benign tumors and <2% of all tumors in the region of the cavernous sinus. CSHs can extend medially into the sella turcica, laterally into the middle cranial fossa, and superiorly toward the optic chiasm.1,2 Current treatment modalities include surgical removal,2,3 conventional radiotherapy,4 and stereotactic radiosurgery.5 However, the optimal management strategy for this lesion remains controversial.6 Because CSHs are benign, well-encapsulated tumors, and total removal could lead to a complete cure, surgical removal has always been the treatment of choice.1,3,7-9 Nevertheless, surgical treatment, especially total tumor removal, has long been considered as a great challenge because of the complex structures in the cavernous sinus; such as the internal carotid artery (ICA) and cranial nerves (CNs). The

NDDCNI: Newly developed or deteriorated cranial nerve injury NTR: Near-total resection OR: Odds ratio PR: Partial resection STR: Subtotal resection From the 1Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Beijing; 2China National Clinical Research Center for Neurological Diseases, Beijing; 3Center of Brain Tumor, Beijing Institute for Brain Disorders, Beijing; 4Beijing Key Laboratory of Brain Tumor, Beijing; and 5Department of Neurosurgery, Cangzhou Central Hospital, Cangzhou, Hebei, China To whom correspondence should be addressed: Li-Wei Zhang, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.11.015 Journal homepage: www.journals.elsevier.com/world-neurosurgery Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.

www.journals.elsevier.com/world-neurosurgery

e1

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

major difficulties encountered in surgery are the control of profuse bleeding and the protection of CNs. Given the paucity of CSHs, previous studies on the surgical treatment of CSHs are mainly series with a relatively small number of patients, and most of these studies, up to now, have been descriptive in nature.1-3,7-9 There is some lack of knowledge in the influential factors of the outcomes of patients with CSHs treated surgically. In this retrospective study, we presented a series of 47 patients with CSHs treated surgically at our institution. To our knowledge, this is the largest single-institution series on the surgical management of CSHs thus far. In addition to descriptive analysis, we performed univariate and multivariate analysis to investigate factors that affect the extent of tumor resection, newly developed or deteriorated cranial nerve injury (NDDCNI), and follow-up Karnofsky Performance Scale (KPS) score (evaluated at the last follow-up) of CSHs patients. We also compared the differences between preoperative, postoperative at discharge, and follow-up KPS scores. Moreover, we discussed the optimal management strategy for CSHs. METHODS Patient Selection The retrospective cohort included 47 consecutive cases of CSHs surgically treated in the Beijing Tiantan Hospital, Capital Medical University from January 2012 to April 2018. All lesions were pathologically confirmed as CSHs after surgery. Clinical and radiographic data were obtained from the patients’ clinical charts and radiological film, respectively. Follow-up information was obtained from outpatient visits or telephone interviews. The KPS score, a widely accepted scoring scale for patients with brain tumors, was used to evaluate the neurologic functional status. The neurologic assessments were conducted on admission, at discharge, and at the last follow-up. The study was approved by the Beijing Tiantan Hospital Research Ethics Committee. The clinical data, including age, sex, symptoms and signs, prior treatments, pre and postoperative KPS scores, surgical approach, the volume of blood loss, surgical morbidities, recurrence or progression, and radiotherapy were obtained retrospectively. Computed tomography and magnetic resonance imaging (MRI) with contrast enhancement were routinely performed preoperatively. Tumor characteristics such as location, extension or invasion and size were obtained from the preoperative MRI and operative records. The tumor size was evaluated as the tumor equivalent diameter (abc)1/3, where the a, b, and c represented the diameters as measured on axial, sagittal, and coronal MRIs, respectively. Management Strategy Surgical removal was recommended for all patients with a preoperative diagnosis of CSHs with the tumor equivalent diameter >1 cm, and no surgical contraindications in the neurosurgery department. The request of patients and their families was also an essential factor influencing the choice of management. Surgical approaches were selected based on tumor location, extension or invasion, size, previous surgery, patients’ requests, and surgeons’ preference. The frontotemporal (FT) approach was performed in 10 (21.3%) patients, with an additional zygomatic osteotomy (frontotemporal zygomatic approach [FTZ]) in 27 (57.4%)

e2

www.SCIENCEDIRECT.com

patients, and an orbitozygomatic osteotomy (frontotemporal orbitozygomatic approach [FTOZ]) in 1 (2.1%) patient to minimize the retraction of the brain. The anterior subtemporal approach and the endoscopic endonasal transsphenoidal (EETS) approach were conducted in 5 (10.6%) and 4 (8.5%) patients, respectively. For large lesions, especially with the extension of the suprasellar region, the FT approach, FTZ approach, and FTOZ approach were preferred. The anterior subtemporal approach, a less aggressive craniotomy, was implemented in patients with relatively small lesions located in the posterior portion of the cavernous sinus. The EETS approach was used for lesions with the invasion of the sella turcica, but without significant extension to the middle cranial fossa (Figure 1). Bone removal was tailored to suffice tumor exposure and resection so that there would be less retraction of the brain during surgery. Furthermore, it is crucial to decrease the intracranial pressure at the beginning of the operation to reduce the potential damage of the brain retraction. Several methods, such as hyperventilation, osmotic diuresis, and cerebrospinal fluid drainage, were used to decrease the intracranial pressure when appropriate. Gross total resection (GTR) was a goal of surgery in all patients. Incomplete resection, when necessary, was recommended to avoid catastrophic complications. To avoid injury to surrounding CNs, hemostasis in the cavernous sinus was usually performed by using gelatin sponge tamponade rather than by coagulation. Postoperative computed tomography scans and MRI scans were routinely performed within 24 hours and 3 days, respectively, to evaluate the tumor resection (Figures 2 and 3). The extent of resection was determined based on both the intraoperative vision of the surgeons and findings on postoperative MRI. Incomplete resection was classified into 3 degrees: near-total resection (NTR, with 90%e99% tumor volume removal), subtotal resection (STR, with 80%e89% tumor volume removal), and partial resection (PR, <80% tumor volume removal). Adjuvant radiotherapy was recommended for patients with incomplete resection. Scheduled radiographic examination at 3, 6 and 12 months, and then once per year was recommended for all 47 patients. Recurrence was defined as the in situ reappearance of lesions after GTR. Progression was defined as the regrowth of residual lesions. After recurrence or progression, radiotherapy and surgical removal were recommended for lesions <3cm and lesions
3cm in size, respectively. Statistical Analysis Descriptive statistics were used to analyze the demographics, clinical data, and surgical information. Categorical variables were analyzed using the Fisher exact test or the c2 test, and univariate binary logistic regression analyses. According to where and by whom the patients were treated, the ward was dichotomized into skull base ward and non-skull base ward. The pre and postoperative and follow-up KPS scores were all dichotomized into 2 groups: favorable (KPS score
90) and unfavorable (KPS score <90). The age and lesion size were stratified into <46 or
46 years and <3 or
3 cm, respectively, but were directly used in univariate and multivariate logistic regression analyses. Factors with a P value <0.1 in the Fisher exact test or the c2 test or the univariate binary logistic regression analysis were entered into the multivariate binary logistic regression model to further investigate

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.11.015

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

Figure 1. Magnetic resonance images of tumor location for selecting the appropriate surgical approaches. Representative images for frontotemporal, frontotemporal orbitozygomatic or frontotemporal zygomatic approach: (A) axial contrast, (B) sagittal contrast, and (C) coronal contrast. Representative

the factors that influence the GTR, NDDCNI, and follow-up KPS score. The pre and postoperative and follow-up KPS scores were compared using the Wilcoxon signed rank test. All statistical analysis was performed using the IBM SPSS for Windows version 23.0 (International Business Machines Corp., Armonk, New York, USA). A P value <0.05 was considered statistically significant.

WORLD NEUROSURGERY -: e1-e14, - 2018

CAVERNOUS SINUS HEMANGIOMAS

images for anterior subtemporal approach: (D) axial contrast, (E) sagittal contrast, and (F) coronal contrast. Representative images for endoscopic endonasal transsphenoidal approach: (G) axial contrast, (H) sagittal contrast, and (I) coronal contrast.

RESULTS Patient Population A total of 47 patients with CSHs, including 11 (23.4%) men and 36 (76.6%) women, were reviewed. The mean and median age at diagnosis was 45.2 and 47 years (range, 20e66 years),

www.journals.elsevier.com/world-neurosurgery

e3

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Figure 2. Magnetic resonance imaging (MRI) and pathologic examination of a 48-year-old woman, who presented with a 1-month history of diplopia. Preoperative MRI shows a mass lesion at the right cavernous sinus: (A) T2-weighted axial, (B) axial contrast, (C) sagittal contrast, and (D) coronal

respectively. The average interval between the onset of initial symptoms and admission was 12 months (range, 1 monthe15 years). Headache as initial presentation occurred in 21 (44.7%) patients, followed by facial hypesthesia in 11 (23.4%) patients, and dizziness in 11 (23.4%) patients. Diminution of vision, ophthalmalgia, suppressed menstruation, epilepsy, and facial pain were observed in 10 (21.3%), 4 (8.5%), 3 (6.4%), 2 (4.3%), and 1 (2.1%) patients, respectively. CN III palsy was present in 10 (21.3%) patients, of whom 7 (14.9%) presented with ptosis and 5 (10.6%) with diplopia. Seven (14.9%) patients presented with CN VI palsy. Three (6.4%) patients were incidentally diagnosed via physical examination. Six (12.8%) patients were initially misdiagnosed with meningioma, followed by 2 (4.3%) patients with pituitary adenoma, and 1 (2.1%) patient with hemangiopericytoma. The average preoperative KPS score was 77.9 (range, 50e100) (Tables 1 and 2).

Surgical Results One patient had been treated with radiosurgery 4 years before surgery. Two patients had undergone surgical treatment at other hospitals before the surgery at our institution. The remaining 44 patients were treated for the first time. The median intraoperative

e4

www.SCIENCEDIRECT.com

contrast. Postoperative MRI shows that the mass lesion was completely resected: (E) axial contrast, (F) sagittal contrast, and (G) coronal contrast. (H) Postoperative pathologic examination (hematoxylin-eosin, 200) established the diagnosis of cavernous hemangioma.

blood loss was 650 ml (range, 100e3500 ml). GTR was achieved in 26 (55.3%) patients, with NTR, STR, and PR in 13 (27.7%), 2 (4.3%), and 6 (12.8%) patients, respectively (Tables 1 and 2). Postoperative headache, dizziness, ophthalmalgia, epilepsy, and facial pain were completely relieved in all patients. Suppressed menstruation persisted in all 3 patients. Of the 11 patients with facial hypesthesia, 1 (9.1%) relieved completely, 7 (63.6%) improved significantly, and 3 (27.3) remained unchanged. Of the 11 patients with CN III palsy, deterioration was observed in 5 (45.4%) patients, followed by no change in 4 (36.4%) patients, improvement in 1 (9.1%) patient, and complete recovery in 1 (9.1%) patient. CN VI palsy was completely alleviated in 1 (14.3%) patient, with 5 (71.4%) patients unchanged, and 1 (14.3%) patient worsened. CN III palsy, which was the most frequent surgical morbidity, occurred in 21 patients, followed by CN VI palsy in 11 patients. The average postoperative KPS score at discharge was 80.6. No significant difference was observed between the pre and postoperative KPS scores at discharge (P ¼ 0.053). Factors Influencing GTR Characteristics of patients with GTR were compared with those without to identify factors that significantly affect GTR. Univariate

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.11.015

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

Figure 3. Magnetic resonance imaging (MRI) and pathologic examination of a 51-year-old woman, who presented with a 2-year history of headache with diplopia and diminution of vision. Preoperative MRI shows a mass lesion at the right cavernous sinus with the invasion of the sella turcica and middle cranial fossa: (A) T2-weighted axial, (B) axial contrast, (C) sagittal

analysis identified significant associations between the invasion of the sella turcica (P < 0.001, Fisher exact test; P ¼ 0.004, univariate logistic regression), skull base ward (P ¼ 0.001, Fisher exact test; P ¼ 0.002, univariate logistic regression), and GTR. In contrast, factors such as age, sex, left or right side, size, and preoperative KPS score were not significantly related to GTR (Table 3). Furthermore, the predictors with P < 0.1 in the univariate analysis were entered into the multivariable binary logistic regression model. Significant associations were identified between the invasion of the sella turcica (odds ratio [OR] ¼ 0.012; 95% CI, 0.001e0.213; P ¼ 0.002), skull base ward (OR ¼ 27.838; 95% confidence interval [CI], 2.995e258.748; P ¼ 0.003), and GTR (Table 4). Factors Influencing NDDCNI Characteristics of patients with NDDCNI were compared with those without to identify potential risk factors for NDDCNI. Univariate analysis revealed that no factors were significantly related to NDDCNI (Table 5). According to the findings of the univariate analysis, the invasion of the sella turcica (P ¼ 0.136, Fisher exact test; P ¼ 0.074, logistic regression) and preoperative KPS score (P ¼ 0.17, Fisher exact test; P ¼ 0.066, logistic regression) were entered into the multivariate binary logistic regression model.

WORLD NEUROSURGERY -: e1-e14, - 2018

CAVERNOUS SINUS HEMANGIOMAS

contrast, and (D) coronal contrast. Postoperative MRI shows that the mass lesion was completely resected: (E) axial contrast, (F) sagittal contrast, and (G) coronal contrast. (H) Postoperative pathologic examination (hematoxylin-eosin, 200) established the diagnosis of cavernous hemangioma.

The multivariate analysis revealed that the preoperative KPS score (OR ¼ 2.966, per 10 score increase; 95% CI, 1.136e7.743; P ¼ 0.026) and the invasion of the sella turcica (OR ¼ 7.137; 95% CI, 1.282e39.726; P ¼ 0.025) were factors affecting the incidence of NDDCNI (Table 6). Follow-Up Follow-up evaluations were performed on all 47 patients by clinic consultation or telephone interviews. The mean and median follow-up duration was 48.83 and 56 months (range, 1e76 months), respectively. During the follow-up period, complete recovery and significant improvement were observed in 14 (29.8%) and 26 (55.3%) patients, respectively; 1 (2.1%) patient died of gastric cancer 1 year and a half later; the remaining 6 (12.8%) patients had no change. Although 20 (42.6%) and 8 (17.0%) patients still complained of CN III disturbance and CN VI disturbance at the most recent follow-up evaluation, respectively, the majority of these patients improved significantly after discharge from hospital. A total of 13 (61.9%) patients with incomplete resection underwent radiotherapy; 8 (61.5%) with Gamma Knife radiosurgery (GKRS) and 5 (38.5%) with conventional radiotherapy. After radiotherapy, 12 (92.3%) residual CSHs

www.journals.elsevier.com/world-neurosurgery

e5

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Table 1. Data of the 47 Patients with Cavernous Sinus Hemangiomas in the Presented Series Case Number

Sex/Age (Year)

Preoperative Symptoms

CN Deficit

Tumor Size (cm)

Surgical Approach

Extent of Resection

1

F/47

Headache, suppressed menstruation

none

3.8

FTZ

NTR

2

F/32

Headache, dizziness, diminution of vision, diplopia

II, III

3.2

FTZ

NTR

3

F/40

Headache, diminution of vision, suppressed menstruation

II

2.4

ETTS

PR

4

F/55

Facial hypesthesia

V

1.5

AST

GTR

5

M/45

Epilepsy, diplopia, ptosis

III

3.4

FTZ

NTR

6

F/48

Facial pain

V

2.5

FTZ

GTR

7

F/57

Headache, facial hypesthesia, diplopia, ptosis

III, V

6.3

FT

GTR

8

M/40

Dizziness, blindness, ptosis

II, III

7.0

FTZ

GTR

9

F/54

Diminution of vision

II

3.4

FTZ

GTR

10

F/34

Headache, dizziness

none

3.2

FTZ

GTR

11

M/41

None

none

1.8

AST

NTR

12

F/58

Epilepsy

none

4.3

FT

NTR

13

F/20

Headache, diplopia, ptosis

III, VI

1.8

AST

GTR

14

F/36

Headache, ophthalmalgia

none

1.8

AST

GTR

15

F/46

Headache

none

1.5

FT

PR

16

M/48

Facial hypesthesia, ptosis, diplopia

III, V, VI

2.7

AST

GTR

17

F/24

Headache, dizziness, diminution of vision, suppressed menstruation

II

5.6

FTZ

GTR

18

F/60

Headache, dizziness

none

3.5

FT

NTR

19

F/65

None

none

4.5

FTZ

STR

20

F/52

Headache, diminution of vision

II

3.4

FTZ

NTR

21

F/38

Headache, ophthalmalgia, diplopia, ptosis

III, VI

3.7

FT

GTR

22

M/49

Headache

none

3.3

FTZ

NTR

23

F/41

Dizziness, diminution of vision, diplopia

II

2.9

FT

NTR

24

F/54

Ophthalmalgia, diminution of vision

II

3.1

FTZ

GTR

25

M/38

Facial hypesthesia, diplopia

26

F/50

Headache, diminution of vision, facial hypesthesia

27

F/50

Headache, dizziness, diplopia

28

F/46

Facial hypesthesia, nausea, diminution of vision

29

F/51

Facial hypesthesia

30

M/41

31

M/36

32

F/51

33 34

V, VI

3.8

FTZ

GTR

II, III, V

4.3

FTZ

NTR

VI

2.6

FT

GTR

II, V

4.3

FTZ

GTR

V

3.3

FTZ

GTR

Dizziness, diminution of vision

II

2.0

FTZ

GTR

Facial hypesthesia

V

5.0

FTZ

NTR

Headache, diminution of vision, diplopia

II, III

4.7

FT

GTR

F/40

Headache, dizziness, diminution of vision, facial hypesthesia

II, V

4.0

FTZ

STR

F/36

Diminution of vision

II

1.1

ETTS

GTR

35

F/45

Diminution of vision

II

4.9

FTZ

PR

36

F/52

Diminution of vision

II

2.2

FTOZ

GTR

37

F/48

Diplopia

VI

3.0

FT

GTR

38

F/25

Facial hypesthesia

V

4.4

FTZ

GTR Continues

e6

www.SCIENCEDIRECT.com

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.11.015

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Table 1. Continued Case Number 39

Sex/Age (Year)

Preoperative Symptoms

CN Deficit

Tumor Size (cm)

Surgical Approach

Extent of Resection

II, V

3.5

ETTS

PR

F/66

Diminution of vision, facial hypesthesia

40

F/23

Diminution of vision, diplopia

II, III

4.7

FT

PR

41

M/35

Headache

none

4.6

FTZ

PR

42

F/54

Diplopia

VI

1.3

ETTS

GTR

43

F/45

Headache, ophthalmalgia, ptosis

III

4.5

FTZ

GTR

44

M/49

None

none

3.9

FTZ

GTR

45

M/57

Dizziness

none

6.0

FTZ

NTR

46

F/47

Headache, dizziness, nausea, vomiting

none

5.1

FTZ

NTR

47

F/54

Headache, dizziness, nausea, vomiting

none

3.6

FTZ

GTR

CN, cranial nerve; F, female; FTZ, frontotemporal zygomatic; NTR, near-total resection; EETS, endoscopic endonasal transsphenoidal; PR, partial resection; AST, anterior subtemporal; GTR, gross total resection; M, male; FT, frontotemporal; STR, subtotal resection; FTOZ, frontotemporal orbitozygomatic.

were successfully controlled; the other 1 (7.7%) patient developed progression and later underwent a salvage resection; fortunately, the patient recovered completely after the radical removal of the recurrent lesion. Eight (38.1%) patients with incomplete resection declined adjuvant radiotherapy and were followed with scheduled observation, of whom 1 had progression but still refused further treatment. The average KPS score at the last follow-up evaluation (follow-up KPS score), which increased significantly compared with the pre (P < 0.001) and postoperative KPS scores at discharge (P < 0.001), was 89.1 (range, 0e100) (Figure 4). Factors Influencing Follow-Up KPS Score Characteristics of patients with favorable follow-up KPS score (KPS score
90) were compared with those with unfavorable follow-up KPS score (KPS score 80) to investigate factors that influence the follow-up neurologic performance. Univariate analysis indicated that the size (P ¼ 0.041, Fisher exact test; P ¼ 0.007, per 1 cm increase, univariate logistic regression), the invasion of the sella turcica (P ¼ 0.041, Fisher exact test), and the preoperative KPS score (P ¼ 0.009, per 10 score increase, univariate logistic regression analysis) may be significantly related to the follow-up KPS score (Table 7). Multivariate binary logistic regression analysis revealed that increased tumor size (OR ¼ 0.044, per 1 cm increase; 95% CI, 0.004e0.477; P ¼ 0.010) was a risk factor for unfavorable follow-up KPS score. However, no significant association was observed between the preoperative KPS score and the follow-up KPS score in the multivariate analysis (Table 8). DISCUSSION CSHs are benign vascular tumors. These lesions are different from intra-axial cavernous malformations that are true vascular malformations in the brain parenchyma.10 CSHs occur predominantly in middle-aged women. Bansal et al.6 reported 22 patients with CSHs in their 14-year experience, only 4 patients were men. Tang et al.11 described 53 patients with CSHs treated with GKRS and there were 15 (28.3%) male and 38 (71.7%) female patients, with a mean age of 52 years old. In the current report, the mean

WORLD NEUROSURGERY -: e1-e14, - 2018

and median age at diagnosis was 45.2 years and 47 years, respectively; only 12 (25%) patients were men, with a female to male preponderance of 3.27:1, which is consistent with previous studies. Although the exact etiology of CSHs remains unclear, estrogen may be associated with the pathogenesis of this lesion. The expanding masses may compress and distort CNs coursing through the cavernous sinus and further result in neurologic morbidities—which resembled other benign neoplasms in this region—such as pituitary adenomas and meningiomas. However, these lesions typically grow slowly and the onset of symptoms is usually insidious; some patients harbored a large tumor but presented with mild symptoms. Zhou et al.12 reported the interval between the initial symptoms and admission ranged from 3 monthse10 years, with an average of 2.5 years. Similar findings were observed in the current series that the mean interval between the initial symptoms and admission was 23.5 months, and the longest interval even reached up to 15 years; 3 patients were asymptomatic and were diagnosed incidentally. Surgical Treatment A variety of surgical approaches have been used in previous literature. The most often used approaches were the FT approach, FTZ approach, and FTOZ approach in the majority of previous reports.1-3,6-9,12 In our series, similarly, these approaches were applied in up to 80.9% of all patients. The major difficulties encountered in surgery are controlling profuse bleeding and avoiding potential CN injuries. Several methods such as proximal ICA control,1,12 preoperative radiotherapy, and intraoperative hypotension13 were described in previous studies to be effective in reducing bleeding. CSHs typically have a pseudocapsule, and a potential plane exists between the tumor and the surrounding structures.1 Choosing the plane to expose and resect CSHs may help reduce bleeding and facilitate the protection of surrounding CNs, perhaps lowering postoperative morbidities.3 Because the surrounding CNs, especially CN III, are very sensitive to injuries such as stretching, compressing, and thermal injury, the dissection near these structures should be as meticulous as

www.journals.elsevier.com/world-neurosurgery

e7

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Table 2. Characteristics of 47 Patients with Cavernous Sinus Hemangiomas Clinical Characteristics

Number (%)

Sex

Clinical Characteristics

Number (%)

Diameter (cm) Range

1.1e7.0

Female

36 (76.6)

Mean

3.56

Male

11 (23.4)

Median

3.5

Age (years)

Extent of resection

Range

20e66

Gross total resection

26 (55.3)

Mean

45.2

Near-total resection

13 (27.7)

Median

47

Duration of symptom Median

12 months

Range

1 monthe15years

Initial manifestation

Subtotal resection

2 (4.3)

Partial resection

6 (12.8)

Blood loss (ml) Range

100e3500

Mean

971.3

Headache

21 (44.7)

Median

Facial hypesthesia

11 (23.4)

Radiotherapy

Dizziness

11 (23.4)

Diminution of vision

10 (21.3)

650

Gamma knife

8

Conventional radiotherapy

5

Ophthalmalgia

4 (8.5)

Suppressed menstruation

3 (6.4)

Range

1e76

Epilepsy

2 (4.3)

Mean

48.83

Facial pain

1 (2.1)

Median

Blindness

1 (2.1)

Oculomotor nerve palsy

11 (23.4)

Ptosis

7 (14.9)

Diplopia

5 (10.6)

Abducent nerve palsy Diplopia

7 (14.9)

Asymptomatic

3 (6.4)

Range

50e100

Mean

77.9

Median

80

Postoperative KPS score Range

70e90

Mean

80.6

Median

80

Follow-up KPS score Range

0e100

Mean

89.1

Median

90 Continues

www.SCIENCEDIRECT.com

Follow-up duration (months)

Death

56 1

Values are number (%) except where indicated otherwise, % is the percentage of the proportion in this group. KPS, Karnofsky Performance Scale.

7 (14.9)

Preoperative KPS score

e8

Table 2. Continued

possible. In our view, it is better to stay as far away from the CN III as possible in surgery. Because the CN VI is typically hidden within the substance of CSHs, we preferred piecemeal resection to en bloc removal when removing large lesions. As Suri et al.3 described, the lesion was rapidly debulked and the feeding artery, that is usually the meningohypophyseal trunk from the ICA, was exposed and coagulated in the early course of surgery. In this way, the tumor vascularity and blood loss were markedly reduced. Then, the lesion could be slowly dissected away from the CN VI and ICA. One of the reasons why the blood loss during surgery was high in the current series is that we preferred piecemeal resection to en bloc removal when removing large lesions. To reduce blood loss, we used an autologous blood recovery machine during surgery. We think it is reasonable to temporarily control the proximal ICA during surgery to reduce blood loss when the bleeding is profuse and difficult to control; meanwhile, care must be taken to avoid cerebral ischemic events. We think that both the control of the proximal ICA and intraoperative hypotension can reduce the intraoperative bleeding.

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.11.015

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Table 3. Univariate Analysis Showing the Influencing Factors for Gross Total Resection GTR Variable

Yes

Univariate Logistic Regression No

Age (years) 45

11

10


46

15

11

Sex Male

5

6

Female

21

15

Side Right

18

13

Left

8

8


3cm

15

17

<3cm

11

4

Yes

12

20

No

14

1

Size

Invasion of sella turcica

Skull base ward Yes

20

6

No

6

15

Preoperative KPS score

P

4

4

80

22

17

P

0.716

Continuous variable

0.979 (0.925e1.306)

0.457

0.505

Male versus female

0.595 (0.153e2.317)

0.454

0.598

Right versus left

1.462 (0.437e4.889)

0.538

0.121

per 1 cm increase

0.767 (0.489e1.203)

0.247

<0.001*

Yes versus no

0.043 (0.005e0.368)

0.004*

0.001*

Yes versus no

8.333 (2.238e31.033)

0.002*

per 10 score increase

1.139 (0.642e2.022)

0.657

1


90

OR (95% CI)

GTR, gross total resection; OR, odds ratio; CI, confidence interval; KPS, Karnofsky Performance Scale. *P values are statistically significant.

Given the benign biologic feature of CSHs, GTR, when safe, is preferable. Suri et al.3 described 7 cases of CSHs, with GTR achieved in 85.7% of patients. Yin et al.,8 Li et al.,9 and Zhou et al.12 reported that GTR was obtained in 81.8%, 100%, and 60% of patients, respectively. In the current report, GTR was attempted in all 47 patients, with GTR and NTR achieved in a total of 39 (83.0%) patients, which was consistent with previous studies. The most common surgical morbidities are CN palsy, especially the CN III and CN VI; however, fortunately, most of these complications are transient. When a patient has a CN III palsy, it is difficult to identify whether the patient has a CN IV palsy at the same time. Moreover, some patients in the current series were followed up via telephone, so it is difficult to evaluate the exact recovery of the CN IV during the follow-up period. Therefore, the damage of the CN IV was not evaluated in our study. Suri et al.3 reported that 71.4% of patients developed transient CN palsy, and 14.3% of patients developed permanent CN VI palsy. Yin et al.8 reported CN III deficit and CN VI deficit occurred in 36.4% and 22.7% of patients, respectively. Li et al.9 reported transient

WORLD NEUROSURGERY -: e1-e14, - 2018

CN dysfunction occurred in 66.7% of patients, with a persistent CN VI palsy in 25% of patients. Similar findings were found in our series, 21 (44.7%) and 11 (23.4%) patients developed CN III and VI palsy, respectively. During the follow-up period, 14 (29.8%) patients recovered completely and 26 (55.3%) patients Table 4. Multivariate Analysis Showing the Influencing Factors for Gross Total Resection Variable

OR

95% CI

P

0.012

0.001e0.213

0.002*

27.838

2.995e258.748

0.003*

Invasion of sella turcica Yes versus no Skull base ward Yes versus no

OR, odds ratio; CI, confidence interval. *P values are statistically significant.

www.journals.elsevier.com/world-neurosurgery

e9

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Table 5. Univariate Analysis Showing Influencing Factors for Newly Developed or Deteriorated Cranial Nerve Injury NDDCNI Variable

Yes

Univariate Logistic Regression No

Age (years) 45

15

6


46

21

5

Sex Female

29

7

Male

7

4

Side

P

24

7

Left

12

4


3cm

26

6

<3cm

10

5

Yes

27

5

No

9

6

Size

Invasion of sella turcica

Skull base ward Yes

20

6

No

16

5

Preoperative KPS score

Continuous variable

1.016 (0.953e1.084)

0.622

0.256

Male versus female

0.422 (0.096e1.856)

0.254

Right versus left

1.143 (0.279e4.683)

0.853

0.292


3cm versus <3cm

2.500 (0.613e10.202)

0.202

0.136

Yes versus no

3.600 (0.882e14.691)

0.074

1

Yes versus no

1.042 (0.268e4.045)

0.953

per 10 score increase

2.050 (0.954e4.406)

0.066

0.17


90

8

0

80

28

11

P

0.505

1

Right

OR (95% CI)

NDDCNI, newly developed or deteriorated cranial nerve injury; OR, odds ratio; CI, confidence interval; KPS, Karnofsky Performance Scale.

improved significantly. No improvement was observed in 4 patients, and a possible explanation might be that these 4 patients had a short follow-up period, and might have significant improvement if given an extended follow-up period. A significant reason for incomplete tumor resection and CN injuries in our series was the profuse intraoperative bleeding. We believe that measures reducing Table 6. Multivariate Analysis Showing Influencing Factors for Newly Developed or Deteriorated Cranial Nerve Injury Variable

OR

95% CI

P

2.966

1.136e7.743

0.026*

7.137

1.282e39.726

0.025*

Preoperative KPS score per 10 score increase Invasion of sella turcica Yes versus no

OR, odds ratio; CI, confidence interval; KPS, Karnofsky Performance Scale. *P values are statistically significant.

e10

www.SCIENCEDIRECT.com

Figure 4. Box plot illustrating the pre and postoperative and follow-up KPS score. Postoperative KPS score at discharge was similar to the preoperative KPS score. The follow-up KPS score was better than the pre and postoperative KPS score. Filled circles indicate outliers.

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.11.015

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Table 7. Univariate Analysis Showing Influencing Factors for Follow-Up Karnofsky Performance Scale Score Follow-Up KPS Score Variable

‡90

£80

Age (years) 45

17

4


46

21

5

Sex Male

9

2

Female

29

7

Side Right

23

8

Left

15

1

<3cm

15

0


3cm

23

9

Yes

23

9

No

15

0

Size

Invasion of sella turcica

Skull base ward Yes

22

4

No

16

5

GTR Yes

22

4

No

16

5

NDDCNI Yes

30

6

No

8

3

Preoperative KPS score

Univariate Logistic Regression P

7

1

80

31

8

P

1

Continuous variable

1.002 (0.934e1.075)

0.956

1

Male versus female

1.086 (0.191e6.191)

0.926

0.138

Right versus left

0.192 (0.022e1.693)

0.137

0.041*

per 1 cm increase

0.080 (0.013e0.495)

0.007*

0.041*

Yes versus no

NA

NA

0.486

Yes versus no

1.719 (0.398e7.431)

0.468

0.486

Yes versus no

1.719 (0.398e7.431)

0.468

0.419

Yes versus no

1.875 (0.382e9.197)

0.438

per 10 score increase

3.970 (1.401e11.255)

0.009*

1


90

OR (95% CI)

KPS, Karnofsky Performance Scale; OR, odds ratio; CI, confidence interval; NA, not available; GTR, gross total resection; NDDCNI, newly developed or deteriorated cranial nerve injury. *P values are statistically significant.

the intraoperative bleeding make the surgical field clearer, and could further facilitate the total tumor resection, and reduce the intraoperative CN injuries. EETS Approach in the Treatment of CSHs The EETS approach is a rapidly developing technique, providing a wide surgical field and broad lateral vision. In addition, the EETS approach provides access to the cavernous sinus through the medial wall that contains no CNs.14 In recent years, this approach has been extensively applied in the management of lesions located in the sellar region, such as pituitary adenomas and craniopharyngiomas.15 However, the role of the EETS approach

WORLD NEUROSURGERY -: e1-e14, - 2018

in the treatment of CSHs remains undetermined. Rare reports have been published on the implementation of the EETS approach in the management of CSHs thus far. Das et al.16 and Hori et al.17 reported surgeries via the transsphenoidal approach for cavernous hemangiomas located in the sellar region. They suggested that it was really a challenge to remove the highly vascular lesion via this approach. Recently, Al-Sharydah et al.18 reported a case of intrasellar cavernous hemangioma, with an STR via the EETS approach. The authors suggested that the EETS approach was a feasible option for debulking intrasellar cavernous hemangiomas, with fewer surgical complications. Similarly, Chibbaro et al.19 recommended that the EETS

www.journals.elsevier.com/world-neurosurgery

e11

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

Table 8. Multivariate Analysis Showing Influencing Factors for Follow-Up Karnofsky Performance Scale Score Variable

OR

95% CI

P

0.044

0.004e0.477

0.010*

5.21

0.991e27.399

0.051

Size per 1 cm increase Preoperative KPS score per 10 score increase

OR, odds ratio; CI, confidence interval; KPS, Karnofsky Performance Scale. *P < 0.05 considered statistically significant.

approach is practicable for CSHs located mainly in the sellar region, and PR or biopsy followed by radiotherapy may be a safe and effective strategy for these lesions. In our series, the EETS approach was used in 4 patients, and GTR was acquired in 2 patients with a relatively small lesion, with no surgery-related CN injuries. Further studies with more patients are needed to investigate the application of the EETS approach in treating CSHs. Factors Related to Outcomes It is clinically significant to investigate the factors that influence the GTR, NDDCNI, and follow-up KPS score of patients with CSHs. However, little was described in available literature. The current study suggested that lesions with the invasion of the sella turcica were less likely to be totally removed; patients, operated by experienced skull base surgeons, had a higher possibility of GTR; the size of the lesion did not influence the GTR. These results are not surprising and could be explained by the fact that it was more difficult to expose lesions located medial to the ICA when performing surgery via the FT, FTZ, and FTOZ approach; experienced skull base surgeons who are more skilled and more familiar with the anatomy of the cavernous sinus can achieve higher rates of GTR; tumors of large size are not always companied by the invasion of the sella turcica, and lesions located in the middle cranial fossa are easier to remove compared with that located in the sellar region. Additionally, we think that the application of intraoperative MRI may reduce the omission of lesions and facilitate the GTR. In our series, according to the multivariate analysis, patients that harbored tumors with the invasion of sella turcica suffered more NDDCNIs, which further illustrated that CSHs with the invasion of sella turcica are more difficult to remove than those without. In addition, the preoperative KPS score was significantly associated with the NDDCNI. A possible explanation for this finding is that patients with preoperative cranial neuropathy are less likely to develop a new CN palsy. For instance, it seems impossible for a patient with complete CN III palsy to develop a new CN III palsy in operation. As for the long-term performance, what significantly influenced the follow-up KPS score was the size of lesions according to the multivariate analysis. Patients with a relatively small lesion tended toward a favorable follow-up neurologic status. Interestingly, the extent of tumor removal and the skull base ward did not reach statistical significance. In other words, the extent of tumor resection, whether complete or not, did not affect patients’ long-term

e12

www.SCIENCEDIRECT.com

neurologic performance. Possible explanations for the result might be that the lesion grew slowly and patients with incomplete resection were routinely recommended for radiotherapy that may have postponed the recurrence or progression of the lesion. Because it remains unknown how long the tumor-controlling effect of radiation treatment will persist, significant differences in neurologic performance between the GTR group and the incomplete resection group may be observed with an extended follow-up period. In addition, the multivariate analysis showed that the preoperative KPS score did not affect patients’ follow-up performance. It is not surprising, and the possible explanation for this result is that the preoperative KPS score is mainly determined by the preoperative symptoms of patients, but the majority of these symptoms completely recovered after surgery. Optimal Management Strategy Becasue CSHs are benign tumors, the ideal treatment is total tumor removal with the preservation of neurologic function. In the current series, no significant difference was observed between the pre and postoperative KPS scores at discharge, which could be recognized as, to some extent, an acceptable result. Moreover, during the follow-up period, the neurologic performance improved in 85.1% of patients, and the average KPS score at the last follow-up even reached up to 89.1%. GTR was a significant goal of the surgery in the current study. However, NDDCNI after surgery was common in our series and worthy of attention. It was disappointing that a total of 36 (76.6%) patients developed NDDCNI. An aggressive surgical strategy is not a flawless treatment option. In recent years, preserving neurologic function rather than achieving GTR has been becoming the top priority of treatment, that means a more effective and safer management strategy is worth exploring. At the same time, radiotherapy was demonstrated to be effective in decreasing the size of CSHs. In 1999, GKRS was initially used as an adjuvant treatment for the residual lesion after an STR.5 Since then, an increasing number of reports on CSHs treated with conventional radiotherapy,4 CyberKnife radiosurgery,20 and GKRS21-23 have been published. Park et al.4 reported 10 patients treated with conventional radiotherapy and found the volume of tumors decreased in all patients, with an average reduction rate of 72.9%. Similar results were also observed in patients treated with GKRS. Wang et al.22 observed a significant volume reduction in 84.3% of patients receiving GKRS. A systematic review and meta-analysis of stereotactic radiosurgery in CSHs revealed that remarkable tumor shrinkage was achieved in 67.8% of patients.24 The complications of radiotherapy, as described in previous studies, are usually mild. Wang et al.20 reported that 10 of the 31 patients in their study developed temporary headache, and 5 experienced vomiting. These symptoms resolved completely after steroid administration. However, due to the paucity of this lesion, few controlled studies focused on comparing the efficacy and safety of radiosurgery with that of surgical resection. Bansal et al.6 suggested that although radiosurgery seems to have fewer morbidities compared with open surgery, it is too early to conclude that radiosurgery is better than open surgery. Srinivas et al.25 suggested that both surgical resection and GKRS were effective modalities to treat CSHs; microsurgery is the preferred treatment modality, especially for large tumors; GKRS can supplement microsurgery wherever possible. In summary,

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.11.015

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

controversy still exists regarding which treatment modality is better in the treatment of CSHs. Surgical management remains an important option for patients with CSHs; if properly performed, favorable outcomes can be achieved via surgical resection. Although GTR is a significant goal of the surgery, what we should always be concerned about during the resection of CSHs is not only the total tumor removal, but also the avoidance of complications, especially CN injuries, because the potential benefits of GTR might be counterbalanced by the risk of neurologic deficits. When the bleeding is difficult to control and the surgical field is not clear enough to distinguish the adjacent CNs, it is difficult to remove the tumor totally, and easy to cause CN injuries. Because the effectiveness of radiotherapy for CSHs is increasingly elucidated, a more conservative debulking or decompressive approach without causing new CN injuries followed by radiosurgery may be more appropriate than aggressive surgical resection. It would be interesting and of great significance that radiosurgery cases be compared with direct surgery ones. Data of patients undergoing radiation therapy and surgery should be collected in the future to perform a controlled study with a longer follow-up period, comparing the efficacy and safety of the 2 treatment modalities. Strategy for Incomplete Resection In the current series, of the 13 patients with incomplete resection who underwent radiotherapy, 12 (92.3%) patients were successfully controlled, and 1 (7.7%) patient developed progression; of the 8 patients with incomplete resection who declined adjuvant radiotherapy, 1 patient had progression. It seems that there were not significant differences between these 2 groups. However, as reported in previous studies, radiotherapy could not only control the growth of tumors, but also reduce or even eliminate the preoperative symptoms and signs in patients. Additionally, the volume of the cases in our study is not large. Thus, we think that in order to further investigate the necessity and efficacy of adjuvant radiotherapy for tumors with incomplete resection, more patients

REFERENCES 1. Linskey ME, Sekhar LN. Cavernous sinus hemangiomas: a series, a review, and an hypothesis. Neurosurgery. 1992;30:101-108. 2. Goel A, Muzumdar D, Sharma P. Extradural approach for cavernous hemangioma of the cavernous sinus: experience with 13 cases. Neurol Med Chir (Tokyo). 2003;43:112-118 [discussion: 119]. 3. Suri A, Ahmad FU, Mahapatra AK. Extradural transcavernous approach to cavernous sinus hemangiomas. Neurosurgery. 2007;60:483-488 [discussion: 488-489]. 4. Park S, Yoon SM, Lee S, Park JH, Song SY, Lee SW, et al. Role of fractionated radiotherapy in patients with hemangioma of the cavernous sinus. Radiat Oncol J. 2017;35:268-273. 5. Iwai Y, Yamanaka K, Nakajima H, Yasui T. Stereotactic radiosurgery for cavernous sinus cavernous hemangioma–case report. Neurol Med Chir (Tokyo). 1999;39:288-290.

and longer duration of follow-up is needed. We recommended that for patients with STR, observation can also be an option, with radiation only reserved for salvage therapy, if necessary. Strategy for Asymptomatic Patients As little is known about the natural history of CSH, for asymptomatic patients, in order to avoid the complication of surgery, it is reasonable to give these patients close observation with serial imaging until the lesion shows growth or the patients become symptomatic. Limitations This study is limited by its retrospective nature and the duration of follow-up. The single-center experience, patient referral pattern, and neurosurgeons’ preference may also lead to selection bias. As CSHs grow slowly, a longer period of follow-up is warranted to further investigate the long-term performance and factors affecting tumor recurrence. This study is also limited by the volume of cases. CONCLUSIONS We presented the largest series to date of CSHs managed surgically in a Chinese patient population. The results demonstrated in this study suggested that being treated by an experienced skull base surgeon favors the total removal of the tumors, whereas the invasion of the sella turcica does just the opposite; moreover, increased tumor size does not affect GTR, but is a risk factor for unfavorable follow-up KPS score; the invasion of the sella turcica is related to NDDCNI and unfavorable follow-up KPS score. Surgical management remains an important option for patients with CSHs. To investigate the optimal treatment strategy, prospective controlled studies with more patients and an extended follow-up period are warranted in the future.

6. Bansal S, Suri A, Singh M, Kale SS, Agarwal D, Sharma MS, et al. Cavernous sinus hemangioma: a fourteen year single institution experience. J Clin Neurosci. 2014;21:968-974. 7. Dou Y, Meng Q, Yan Z, Xu J, Che S, Jiao Y, et al. Diagnosis and microsurgical treatment of cavernous sinus hemangioma. Artif Cells Blood Substit Immobil Biotechnol. 2010;38:109-112. 8. Yin YH, Yu XG, Xu BN, Zhou DB, Bu B, Chen XL. Surgical management of large and giant cavernous sinus hemangiomas. J Clin Neurosci. 2013;20:128-133. 9. Li MH, Zhao JL, Li YY, Zeng CH, Xu GS, Hong T. Extradural transcavernous approach to cavernous sinus cavernous hemangiomas. Clin Neurol Neurosurg. 2015;136:110-115. 10. Gonzalez LF, Lekovic GP, Eschbacher J, Coons S, Porter RW, Spetzler RF. Are cavernous sinus hemangiomas and cavernous malformations different entities? Neurosurg Focus. 2006;21:e6. 11. Tang X, Wu H, Wang B, Zhang N, Dong Y, Ding J, et al. A new classification and clinical results of Gamma Knife radiosurgery for cavernous sinus

WORLD NEUROSURGERY -: e1-e14, - 2018

hemangiomas: a report of 53 cases. Acta Neurochir (Wien). 2015;157:961-969 [discussion: 969]. 12. Zhou LF, Mao Y, Chen L. Diagnosis and surgical treatment of cavernous sinus hemangiomas: an experience of 20 cases. Surg Neurol. 2003;60:31-36 [discussion: 36-37]. 13. Ohata K, El-Naggar A, Takami T, Morino M, ElAdawy Y, El-Sheik K, et al. Efficacy of induced hypotension in the surgical treatment of large cavernous sinus cavernomas. J Neurosurg. 1999;90:702-708. 14. Prajapati HP, Jain SK, Sinha VD. Endoscopic versus microscopic pituitary adenoma surgery: an institutional experience. Asian J Neurosurg. 2018;13:217-221. 15. Bal E, Oge K, Berker M. Endoscopic endonasal transsphenoidal surgery, a reliable method for treating primary and recurrent/residual craniopharyngiomas: nine years of experience. World Neurosurg. 2016;94:375-385. 16. Das S, Ang LC, Ramsay D. Intrasellar cavernous hemangioma presenting as pituitary adenoma: a report of two cases and review of the literature. Clin Neuropathol. 2018;37:64-67.

www.journals.elsevier.com/world-neurosurgery

e13

ORIGINAL ARTICLE ZONG-HAO LI ET AL.

CAVERNOUS SINUS HEMANGIOMAS

17. Hori S, Hayashi N, Nomoto K, Sato H, Hayashi T, Nagai S, et al. Cavernous sinus cavernous hemangioma largely extending into the sella turcica and mimicking pituitary adenoma: case report. Neurol Med Chir (Tokyo). 2010;50:330-332.

21. Lee CC, Sheehan JP, Kano H, Akpinar B, Martinez-Alvarez R, Martinez-Moreno N, et al. Gamma Knife radiosurgery for hemangioma of the cavernous sinus. J Neurosurg. 2017;126: 1498-1505.

18. Al-Sharydah AM, Al-Suhibani SS, Al-Jubran SA, AlAbdulwahhab AH, Al-Bar M, Al-Jehani HM, et al. Endoscopic management of atypical sellar cavernous hemangioma: a case report and review of the literature. Int J Surg Case Rep. 2018;42:161-164.

22. Wang Y, Li P, Zhang XJ, Xu YY, Wang W. Gamma Knife surgery for cavernous sinus hemanginoma: a report of 32 cases. World Neurosurg. 2016;94:18-25.

19. Chibbaro S, Cebula H, Ganau M, Gubian A, Todeschi J, Lhermitte B, et al. Multidisciplinary management of an intra-sellar cavernous hemangioma: case report and review of the literature. J Clin Neurosci. 2018;52:135-138. 20. Wang X, Zhu H, Knisely J, Mei G, Liu X, Dai J, et al. Hypofractionated stereotactic radiosurgery: a new treatment strategy for giant cavernous sinus hemangiomas. J Neurosurg. 2018;128:60-67.

e14

www.SCIENCEDIRECT.com

cavernous sinus hemangiomas: an institutional experience. J Neurol Surg B Skull Base. 2017;78: 399-407.

Conflict of interest statement: The authors declare that the article content was composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Received 6 June 2018; accepted 3 November 2018

23. Park CK, Choi SK, Kang IH, Choi MK, Park BJ, Lim YJ. Radiosurgical considerations for cavernous sinus hemangioma: long-term clinical outcomes. Acta Neurochir (Wien). 2016;158:313-318. 24. Wang X, Mei G, Liu X, Dai J, Pan L, Wang E. The role of stereotactic radiosurgery in cavernous sinus hemangiomas: a systematic review and metaanalysis. J Neurooncol. 2012;107:239-245.

Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.11.015 Journal homepage: www.journals.elsevier.com/worldneurosurgery Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.

25. Srinivas D, Sarma P, Shukla D, Bhat D, Pandey P, Somanna S, et al. Multimodality management of

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.11.015