PEER-REVIEW REPORTS

Surgical Management and Evaluation of Prognostic Factors Influencing Postoperative

Visual Outcome of Suprasellar Meningiomas
Chuan-Wei Wang1, Ying-Yu Li 2, Shu-Gan Zhu1, Yang Yang1, Hong-Wei Wang1, Jie Gong1, Yu-Guang Liu1

Key words 䡲 OBJECTIVE: To study the microsurgical technique and the prognostic factors
䡲 Meningioma
influencing postoperative visual outcome in patients with suprasellar meningi-
䡲 Microsurgery
䡲 Multivariate analysis omas (SMs).
䡲 Suprasellar tumor
䡲 Visual outcome 䡲 METHOD: The clinical materials of 45 patients with SMs treated microsurgi-
cally between January 2002 and October 2008 were analyzed retrospectively.
Abbreviations and Acronyms Patients received ophthalmologic and radiologic assessment before and after
ACPM: Anterior clinoid process meningioma
DSM: Diaphragma sellae meningioma
operation. Follow-up ranged from 12 to 93 months (median, 39 months). As far as
MRI: Magnetic resonance imaging monocular vision was concerned, univariate and multivariate statistical analysis
OC: Optic chiasma was performed among factors that might influence postoperative visual outcome.
ON: Optic nerve
PSM: Planum sphenoidale meningioma 䡲 RESULTS: In this group, the mean age of the patients was 51 years. Median
SM: Suprasellar meningioma tumor size was 3.5 cm. The duration of symptoms ranged from 10 days to 35 years
TSM: Tuberculum sellae meningioma
VA: Visual acuity
(median, 18 months). Total tumor resection was achieved in 40 cases (88.9%) and
VF: Visual field subtotal in 5 cases (11.1%). With respect to visual outcome of the 90 eyes, 54 eyes
(60%) improved, 24 (26.7%) remained unchanged, and 12 (13.3%) had worsened. It
From the 1Department of Neurosurgery, Qilu
Hospital of Shandong University, Jinan, P.R.
has been illustrated that the postoperative visual improvement was determined
China; and 2Department of Neurosurgery, Xintai Second by various factors, including age, recurrence, duration and severity of visual
People’s Hospital of Shandong Province, P.R. China disturbance, preoperative condition of optic disc, tumor size, location, peritu-
To whom correspondence should be addressed: moral edema, involvement with optic canal, arachnoid membrane interface, and
Yu-Guang Liu, M.D. [E-mail: [email protected]]
extent of tumor removal. However, the multivariate analysis showed that
Citation: World Neurosurg. (2011) 75, 2:294-302.
DOI: 10.1016/j.wneu.2010.08.021
recurrence, preoperative condition of optic disc, duration of impaired vision, and
Journal homepage: www.WORLDNEUROSURGERY.org
peritumoral edema are the most influential factors.
Available online: www.sciencedirect.com 䡲 CONCLUSIONS: Skillful microsurgical techniques are key to resect SMs
1878-8750/$ - see front matter © 2011 Elsevier Inc. while getting the desired visual outcome. The recurrence, preoperative condition
All rights reserved.
of optic disc, duration of impaired vision, and peritumoral edema might be the
INTRODUCTION most important factors influencing postoperative visual outcome.
Suprasellar meningiomas (SMs) account for
5%–10% of all intracranial meningiomas, in-
So far, there have been few reports of mul- Qilu Hospital of Shandong University. In con-
cluding meningiomas arising from the dura
tivariate analysis of the factors influencing sideration of visual outcome analysis, two
of tuberculum sellae, anterior clinoid process,
postoperative visual outcome, except that of cases were excluded because of obvious con-
diaphragma sellae and planum sphenoidale.
Zevgaridis et al. (41), who concluded that in- founding factors. One had measles when he
The impaired vision is the most common
tact brain–tumor interface and absence of se- was age 7 years, which had made him binoc-
symptom because of tumor compression of vere (⬍0.2) preoperative visual loss were the ular blind preoperatively, and the other died
the optic nerve (ON) and optic chiasma (OC), two significant influential factors. This study soon after surgery and the visual result could
or the extension of tumor tissue into optic focuses on a group of SM patients with visual not be tested. Clinical data of the remaining
canals and fossa orbitalis. Tuberculum sellae impairments. Many clinical variables were an- 45 patients were reviewed retrospectively.
meningioma (TSM) was first described by alyzed through statistical methods.
Stirling and Edin in 1897 (35), and the first
complete removal was performed in 1916 by Radiologic Examination and Follow-up
Cushing and Eisenhardt (9). Although many MATERIAL AND METHODS All patients were given computed tomogra-
studies on SMs show lower mortality and phy (CT) and/ or magnetic resonance imag-
higher complete excision rate (3, 11, 22), the Inclusion and Exclusion Criteria ing (MRI) scans. MRI was performed in 41
factors influencing visual improvement are Between January 2002 and October 2008, a patients preoperatively and 31 postopera-
still in dispute. total of 47 patients with SMs were treated in tively. After the discharge from hospital, pa-

294 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, DOI:10.1016/j.wneu.2010.08.021

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

tients were followed up from 12 to 93 Generally, the base and the feeding artery of Statistical Analysis
months (median, 39 months). Forty-one the tumor were coagulated with bipolar co- Univariate and multivariate statistical anal-
patients received CT whereas 21 patients re- agulation first. After the tumor feeding ar- ysis were done among factors that might
ceived MRI during follow-up. tery was blocked, the shriveled tumor could influence postoperative visual outcome.
be coped with under exsanguinate operative Statistical analysis was performed with the
field, and en bloc resection of tumor could use of ␹2 test, Fisher’s exact test and logistic
Ophthalmologic Examination and be achieved. However, sometimes the tu- regression. SPSS 13.0 statistical software
Evaluation mor was too large and pervasive to get its was used to deal with data and make 0.05
Ophthalmologic examination consisted of real base confirmed and the major feeding the boundary of statistical significance. For
testing the patient’s visual acuity (VA) with arteries could not be interrupted immedi- logistic statistical analysis, entry and re-
Snellen E decimal chart, Humphrey visual ately. In this situation, we preferred to first moval probability values for stepwise were
Field Analyser (HFA储) perimetry for visual coagulate the tumor capsule and the abnor- 0.10 and 0.10.
field (VF), and ophthalmoscope examina- mal arteries creeping on it, simultaneously
tion of optic disc condition for every single dissecting the tumor tissue nearby the base
eye. A total of 90 eyes in the 45 patients were progressively. After tumor debulking was
evaluated. Improvement in VA was defined managed in piecemeal removal, the base of RESULTS
as a change of ⱖ1 line. Improvement in VF tumor could be determined and the major
was considered significant if VF improved feeding artery be blocked. In most cases, Symptoms and Signs
with five or more points of P ⬍5% or with a the dura where tumor attached was coagu- This group was composed of 36 women and
cluster of three or more points of P ⬍5% lated or incised. If the optic canal was in- 9 men. The mean age of the patients was 51
(17). Improvement in visual outcome was volved, the fibrous ligament would be years (ranging from 23 to 75 years). Dura-
defined by two conditions: both in VA and opened and the bone canal be drilled with tion of symptoms was from 10 days to 35
VF or improvement in one of the two items continuous irrigation to avoid heat damage years (median, 18 months). Visual impair-
and no change in the other. The situation to the optic apparatus. If the tumor invaded ment was the most common initial symp-
that one of the two items improved but the into orbit, the paries superior orbitae and tom. Forty-two patients (93.3%) com-
other deteriorated was not counted. As far lateral wall of orbit should be opened for plained of visual problems, but the
as monocular vision was concerned, univar- resecting the tumor. ophthalmologic findings showed visual
iate and multivariate statistical analysis There are some surgical key points for
problems in all patients. The clinical symp-
were performed among factors that might visual improvement. Arachnoid mem-
toms and signs are shown in Table 1.
influence postoperative visual outcome. branes in subarachnoid cistern that cover
above the surface of the tumor, the optic
apparatus, and surface of cerebral lobe
should be preserved. Excessively radical Radiologic Features
Surgical Techniques manipulation should be avoided and a little The tumor appeared as mass-like shadow,
All operations were carried out under gen- remnant tumor on the optic apparatus and hyperdense on 31 (68.9%) of 45 CT scans,
eral anaesthesia using microsurgery tech- arteries will be permitted if tumor tissue hypodense on 14 (31.1%), and enhanced
nique. Unilateral subfrontal approach was adhere closely to the neighboring vital with contrast administration on 35 scans
performed in 12 patients, bilateral subfron- structures. Arterial branches should be pro- (77.8%). Mean tumor diameter was 3.6 cm
tal approach in 8, frontal-interhemispheric tected carefully, such as the posterior eth- (range 1.2– 8 cm). Calcification was present
approach in 4, pterional approach in 16, moidal, ophthalmic, superior hypophysial, in 6 (13.3%), peritumoral edema in 17
frontal–pterional approach in 4, and trans- and anterior communicating artery (40). (37.8%), shift of midline structure in 12
sphenoidal approach in 1. The choice of When close to the ON and OC, tumor resec- (26.7%), and compression to cerebral ven-
surgical strategies depended on several fac- tion should be done with low-power bipolar tricle in 10 (22.2%). Most tumors showed
tors, including the tumor size, location, di- coagulation or just be performed by Cavit- sharply demarcated margin, a dural tail and
rection of tumor growth, relationships ron ultrasonic aspiration.
among tumor and the adjoining structures, generally isointensity to the grey matter on
dural attachment of the tumor, etc. The side T1 WI and mild hyperintensity on T2 WI,
of marked visual deterioration was selected Pathologic Studies with strong contrast enhancement on MRI
for incision, whereas the nondominant According to WHO classification (2007) of scans. The shape of the tumors was round-
hemispheric side was preferred in bilateral tumors of the central nervous system (21), ish (22 cases), lobulated (3), or irregular
involvement. Prophylactic treatment with meningothelial meningioma was con- (11). Invasion into the unilateral cavernous
steroids and anticonvulsants was used in firmed in 24 cases, fibrous (fibroblastic) in sinus was present in 5 patients, optical ca-
the cases in which seizures occurred preop- 7, transitional in 7, angiomatous in 2, nal involvement in 17 eyes with 5 further
eratively or obvious peritumoral edema was psammomatous in 1, metaplastic in 2, and extended into the orbital wall or orbital
found. Mannitol, nimodipine, and papaver- atypical in 2. The former six groups corre- compartment, and involvement with acces-
ine were administered perioperatively. Af- sponded to WHO grade 0, while the last sory nasal cavity in 5 patients. Hyperostosis
ter the tumor was exposed, the ON and in- “atypical” group corresponded to WHO of the sellar base of the skull bone was pres-
ternal carotid artery should be identified. grade I. ent in 13 patients.

WORLD NEUROSURGERY 75 [2]: 294-302, FEBRUARY 2011 www.WORLDNEUROSURGERY.org 295

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

Table 1. Symptoms and Signs of the 45 Patients with Suprasellar Meningiomas

Symptoms and Signs No. of Patients Percentages Symptoms and Signs No. of Patients Percentages

Visual acuity decrease 43 95.6 Ocular proptosis 3 6.7

Monocular 10 23.3 Ocular movement disorder 3 6.7
Binocular 33 76.7 Seizures 2 4.4
Visual field defect 32 71.1 Mental deficits 1 2.2
Monocular 12 37.5 Polydipsia or polyuria 1 2.2
Binocular 20 62.5 Lactopoiesis and menstruation 1 2.2
irregularis
Headache or dizziness 19 42.2 Paroxysmal ache of cheek 1 2.2
Hyposmia 6 13.3 Incidental head trauma 1 2.2

Surgical Outcome and Complications died for other diseases not relevant to Univariate Analysis on Postoperative
One of the 47 patients died after surgery SM. And 2 were lost to contact. We used Visual Outcome
(overall mortality 2.13%) and was thus ex- the Karnofsky Score method just like Regarding monocular eye, visual outcome
cluded from the statistical analysis for the Bassiouni (3), but our classifications improved in 54 eyes (60%), remained un-
sake of sample homogeneity. Simpson were a little different. Nineteen patients changed in 24 (26.7%), and got worsened in
Grade I–III resection was achieved in 40 of (46.3%) were able to carry on normal ac- 12 (13.3%). Univariate analyses of different
45 cases (88.9%) (Figures 1 and 2) and tivity and work (Karnofsky score ⫽ 80 – variables were made to search influential
Simpson Grade IV–V in 5 (I in 13, II in 20, III 100). Sixteen patients (39%) had varying preliminary factors. In the general vari-
in 7, IV in 4, and V in 1). The mean operative degrees of visual deficits and were unable ables, there was an obvious negative rela-
time was 4.62 hours (range 3–9 hours). The to work but could care for most of his tionship between age and postoperative vi-
estimated blood loss (EBL) was 385 mL personal needs (Karnofsky score ⫽ 50 – sual improvement but was not significant in
(range 100 –1200 mL). The comparison be- 70). Six patients (14.6%) were disabled gender and preoperative headaches. In the
tween pre- and postoperative images of and required special assistance because recurrence group, only one eye had visual
SMs was shown in Figures 1 and 2. The of severe neurologic deficits (Karnofsky improvement, six eyes were unchanged,
complications are explained in Table 2. score ⬍40). Recurrences were observed and three worsened. Both the duration and
in three cases (recurrence rate ⫽ 7.3%), severity of preoperative visual disturbance
Patient Follow-up respectively, at 23, 56, and 68 months had statistically significant influence on
In the 45 patients, 43 were followed up for postoperatively. Two of them accepted re- postoperative vision improvement. In con-
a mean period of 39 months, in which 2 operation in our hospital. sideration of the preoperative condition of
the optic disc, postoperative visual im-
provement in atrophy group was conspicu-
ously worse than the normal group and
fuzzy group (Table 3).
Tumor variables were also a serial of im-
portant factors. Visual outcome had a sig-
nificant relationship with tumor size and
texture. The tumors with firm texture were
correlated with a worse visual outcome than
soft ones. The group of tumor maximal di-
ameter ⬍3 cm had better postoperative vi-
sion than the two group of 3–5 cm and ⬎5
cm. Visual improvement in the group with-
out invasion into optic canal and orbit ex-
celled the other two groups, whereas only
one improvement in five occurred when tu-
mor intruded into the orbital wall and com-
Figure 1. Preoperative MRI image of one patient Figure 2. Postoperative MRI image of the
partment. Tumor location, referring to not
(median sagittal section) showing a meningioma same patient (median sagittal section) only the location of tumor base but also
with wide base involving the dura of tuberculum showing complete removal of tumor and direction of tumor growth, was a significant
sellae and diaphragma sellae. The arrow shows good protection of surrounding normal factor influencing vision recovery. The best
the encasement of arteries by tumor. structures.
visual outcome occurred in meningiomas

296 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, DOI:10.1016/j.wneu.2010.08.021

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

Table 2. Complications

Complications No. of Patients Percentages Outcomes at Discharge

Hyponatremia 7 15.6 5 cured, 2 improved

Hypokalemia 3 6.7 1 cured, 1 improved
Diabetes insipidus 2 4.4 All improved but uncured
Transient urinary volume increase 6 13.3 All cured with or without hydrochlorothiazide and desmopressin
Hyposmia 4 8.9 All improved but uncured
Anosmia 2 4.4 Uncured
Cerebrospinal fluid rhinorrhea 2 4.4 1 cured, 1 received skull-base repairing surgery and got cured
Subcutaneous hydrops 3 6.7 All cured
Intracranial infection 4 8.9 All cured
Mild ptosis 2 4.4 1 cured, 1 improved
Mood disorder 1 2.2 Uncured
Epidural hematoma 1 2.2 Cured
One patient got subarachnoid hemorrhage (intruding into ventricle) soon after surgery. This patient died of herniation and subsequent respiration and circulation failure, and was excluded from
data analysis because there was no visual test.

restricted to the regions of diaphragma sel- DISCUSSION rent cases were irregularly shaped, and
lae and planum sphenoidale, and the worst Of late, articles on SMs have been numer- three exceeded 4 cm. All the five experi-
in complex bases that creep across a wide ous. Several methods were used to evaluate enced a long duration before the second
location. Peritumoral edema was a negative the factors influencing postoperative visual surgery, the longest of which lasted 15
factor. However, tumor blood supply and outcome (2, 3, 7, 11, 12, 16, 22, 23, 25, 26, years. We thought that the optic apparatus
pathologic type did not have close relation- 28, 33, 36, 41). However, some classifica- of the recurrent cases might be more frag-
ships with postoperative visual outcome tions about factors might not be consider- ile.
(Table 3). There were no differences be- ate enough. For example, in many articles, Patient age and duration of and severity
tween groups within different surgical ap- the variable “age” was divided into only two of visual decrease were controversial factors
proaches, but the degree of tumor removal, groups, younger than or above 50 or 60 in many articles. Some authors found pa-
integrality of arachnoid membrane inter- years of age. In this paper, the data were tients younger than age 60 years (25), 50
face, and disposal of optic canal indeed had analyzed via univariate and multivariate sta- years (11) or 40 years (28), with duration less
significant influence. EBL and the time of tistical analysis. In previous reports, post- than 6 months or 2 years (2, 28), and visual
operation were found insignificant to visual operative vision improved in 36%– 80% of loss not exceeding 75% (3) had an obvious
outcome (Table 3). patients, were unchanged in 0%– 63% and better visual outcome, but some others ar-
worse in 0%–20% (3, 7, 11, 12, 25, 26). In gued that there were not necessarily signif-
our cohort, the rates were 60%, 26.7%, and icant differences (3, 22, 33). Our univariate
Multivariate Analysis on Postoperative
13.3%, respectively. analysis showed that these three factors
Visual Outcome
Aforesaid univariate analysis showed some were all negatively related with the postop-
potential factors influencing visual out- General Variables erative visual outcome. However, we argue
come. But that was not enough. We found For gender factor, although female domi- that there should not be an exact age, dura-
that recurrence, preoperative condition of nance has been acknowledged in SMs (25, tion, or severity boundary but a tendency.
optic disc, duration of visual decrease, and 36), no statistically significant effect of gen- There were no progressive sequences
peritumoral edema were the most influen- der on visual outcome was found in our data among subgroups of these three factors ac-
tial factors by multivariate logistic analysis (P ⫽ 0.432) and others’ (25). Five of the 45 cording to our data (seen in Table 3).
(Table 4). However, other factors that had patients were admitted to our hospital with Gloomy visual outcome of the older may be
statistical significance in the univariate a diagnosis of recurrent meningioma, with related to poor tolerance of ischemia (29).
analysis, including age, severity of preoper- worse postoperative visual outcome than Long duration and severe visual decrease
ative visual loss, tumor size, extent of me- nonrecurrent ones (P ⫽ 0.001). In addition, always mean worse harm to the optic appa-
ningioma resection, location, involvement the three new recurrence cases after surgery ratus.
with optic canal or orbit, arachnoid plane, also had bad visual outcome at the time of Optic disc pallor was present with a high
and degree of tumor removal, were elimi- follow-up. As Nakasu et al. (24) said, recur- rate up to 92% (11) or almost all (25).
nated through selection by the multivariate rent meningioma was always huge or irreg- Bassiouni et al. (3) suggested that the pres-
logistic analysis. ularly shaped. In our data, four in five recur- ence of optic disc pallor did not preclude

WORLD NEUROSURGERY 75 [2]: 294-302, FEBRUARY 2011 www.WORLDNEUROSURGERY.org 297

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

complete postoperative recovery of vision

except when it was accompanied by a VA Table 3. Univariate Analysis of Factors on Visual Improvement
lower than 0.25. However, some authors Improved/ Improved
correlated the presence of optic disc pallor Variables Unimproved Percentage P Value
(28) with a worse postoperative outcome.
Our data showed a statistical significance Age (years) 0.001
among three groups (P ⫽ 0.000). Visual im- ⬍40 8/2 80.0
provement rate of fuzzy disc group was
40–50 28/6 82.4
close to normal disc group, and the two
were much better than the atrophy group. 50–60 14/18 43.8
Atrophy optic disc always means severe and ⬎60 5/9 35.7
long-time destruction to ON or OC so that Gender 0.432
the nerves are rather difficult to get salvage.
In previous studies, headache or dizzi- Male 10/8 55.6
ness were reported in different incidence Female 44/28 61.1
rates, from 4.8% to even 85.7% (3, 6, 7, 12, Recurrence 0.001
14, 26, 31-33). Some authors demonstrated
No 53/27 66.3
that headache was the second most com-
mon symptom. In our series, headaches or Yes 1/9 10.0
dizziness was present in 19 patients Duration (months) 0.02
(42.2%). Although nearly half of our cases ⬍6 26/8 76.5
presented with headaches, whether head-
6–12 4/4 50.0
aches existed or not did not influence visual
outcome (P ⫽ 0.415). 12–24 11/5 68.8
⬎24 13/19 40.6

Tumor Variables Preoperative vision 0.015

Several authors found that a better visual ⬎0.8 16/5 76.2
outcome was associated with tumors less 0.4–0.8 14/7 66.7
than 3 cm (12, 25, 28) or 4 cm (27) in diam-
0.1–0.4 14/6 70.0
eter or 15 mm3 in volume (25). In other
studies, however, postoperative visual out- ⬍0.1 10/18 35.7
come was not influenced by tumor size (2, Headache 0.415
11, 33, 41). Our findings showed that visual No 25/15 62.5
recovery was good (83.3%) when tumor
Yes 29/21 58.0
size was less than 3 cm and postoperative
vision of both the other two subgroups were Optic disc 0.000
sustained at a low level. Large SMs always Normal 23/5 82.1
mean the surgically unfavorable stage de- Fuzzy 21/7 75.0
fined by Cushing (9). When peritumoral
edema existed, it always meant a bad visual Atrophy 10/24 29.4
outcome (25). This might be due to the fact Tumour size (cm) 0.005
that peritumoral edema had highly signifi- ⬍3 25/5 83.3
cant lower cerebral blood flow and volume
3⬃5 20/20 50.0
in neural tissue adjacent to the tumor (5,
37). In our series, only 36.7% of 15 patients ⬎5 9/11 45.0
who had clear peritumoral edema had vi- Peritumoral edema 0.003
sion improvement, in comparison with the No 43/17 71.7
nonedema group of 71.7%. Peritumoral
Yes 11/19 36.7
edema may increase difficulty and risk of
surgery, raising the chance of hurting optic Tumor texture 0.019
nerves. Firm 31/25 55.4
Tumor texture was not considered in Soft 23/11 67.6
many previous reports (3, 25, 26, 33). Some
authors (11, 12, 15) argued a soft tumor Tumor blood supply 0.357
might be related with visual improvement. Rich 42/30 58.3
In our study, there was a significant differ- Poor 12/6 66.7
ence (P ⫽ 0.019) between firm and soft tu-
Continues

298 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, DOI:10.1016/j.wneu.2010.08.021

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

mor. Firm tumor is always composed of

Table 3. Continued more parenchymal tissue than soft one. We
Improved/ Improved found that some firm meningiomas were
Variables Unimproved Percentage P Value too rigid to be cut, accidentally leading to an
increasing risk for injury of visual apparatus
Location* 0.000 and its blood supply, whereas soft tumor
None 2/2 50.0 was a favorable role for visual outcome be-
cause of its availability of being removed.
TSM 36/16 69.2
The importance of interruption of the tu-
ACPM 1/3 25.0 mor basal blood supply and hollowing of
PSM 3/1 75.0 the tumor have been proposed by many au-
DSM 11/3 78.57
thors (3, 16, 20, 33). However, not many
authors mentioned the influence of the dif-
Complex 2/10 16.7 ferent levels of tumor blood supply on vi-
Involvement with optic canal and orbit 0.04 sual outcome. Our data showed that visual
No involvement 49/24 67.1 outcome in the poor-blood-supply group
was better, but there was no significant dif-
Optic canal 5/7 41.6
ference between them (P ⫽ 0.396). We
Orbit 1/4 20.0 thought although abundant blood supply
Integrality of arachnoid membrane interface 0.006 feeding tumor might reduce blood supply
Yes 46/21 68.7 of neighboring neuronal structures, inter-
ruption of tumor’s feeding arteries and
No 8/15 34.8
strict but not excessive bipolar electrocoag-
Degree of tumor removal 0.001 ulation might contribute to the satisfied
I 21/5 80.8 outcome.
II 23/17 57.5
III 9/5 64.3 Surgical Variables
IV–V 1/9 10.0 After Simpson first introduced the degree of
meningioma resection (34), many authors
Approach 0.375
had extended different modifications about
Subfrontal(unilateral) 14/10 58.3 the Simpson’s degree (3, 11, 22). There were
Subfrontal(bilateral) 6/10 37.5 different views which either supported ad-
Frontal–interhemispheric 6/2 75.0 vantageous effect of total resection (25) or
not (33). Our data showed significant dif-
Pterional 21/11 65.6
ference among degrees of tumor resection
Frontal–pterional 5/3 62.5 (P ⫽ 0.001). On one hand, tumor resection
Transsphenoidal 2/0 100.0 should be as complete as possible to pre-
Estimated blood loss (mL) 0.151
vent recurrence (11, 25, 41). On the other
hand, too radical tumor removal should not
⬍300 13/5 72.2 be performed when the tumor invaded into
300–600 28/14 66.7 the cavernous sinus or closely adhered to
60–900 8/10 44.4 the important arteries, ONs, and/or the pi-
tuitary stalk, and a slight tumor tissue rem-
⬎900 5/7 41.7
nant is permitted.
Time of operation (hours) 0.328 Some authors (33, 41) investigated that
⬍4 15/11 57.7 tumor location had no effect on outcome,
4–5 23/9 71.9 whereas Andrews et al. (2) argued that tu-
mors limited to the tuberculum sellae were
5–6 10/7 58.8
the most frequent completely resected and
6–7 4/5 44.4 correlated with postoperative visual recov-
⬎7 2/4 33.3 ery. Our findings indicated different view-
points. Visual outcome of the cases with
*Regarding monocular vision as research object, tumor locations in our series were classified into six groups. “None” means the
contralateral side of unilateral anterior clinoid process meningioma; TSM, tuberculum sellae meningioma; PSM, planum sphenoidale
complex and wide basal location were obvi-
meningioma; ACPM, anterior clinoid processs meningioma; DSM means diaphragma sellae meningioma that is restricted to the ously worse than others’. Anterior clinoid
diaphragma sellae region; “Complex” means wide, extensive, or unclear locations and growth direction, such as large and irregular tumor process meningioma (ACPM) was associ-
that creeps across te planum sphenoidale, tuberculum sellae and diaphragma sellae, and even the olfactory groove. ated with a bad visual recovery (25%), be-
cause it was liable to destroy the anterior

WORLD NEUROSURGERY 75 [2]: 294-302, FEBRUARY 2011 www.WORLDNEUROSURGERY.org 299

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

Table 4. Variables in the Equation of Multivariate Analysis

95.0% C.I. for EXP(B)

B SE Wald Significance Exp(B) Lower Upper

Recurrence ⫺2.168 1.159 3.497 .061 .114 .012 1.110

Duration 6.623 .085
Duration (1) ⫺1.445 1.025 1.984 .159 .236 .032 1.760
Duration (2) ⫺.446 .874 .261 .610 .640 .115 3.550
Duration (3) ⫺1.873 .781 5.743 .017 .154 .033 .711
Opticdisc 6.021 .049
Opticdisc (1) ⫺1.116 .815 1.877 .171 .327 .066 1.617
Opticdisc (2) ⫺1.828 .746 6.008 .014 .161 .037 .693
Edema ⫺1.403 .618 5.161 .023 .246 .073 .825
Constant 3.226 .914 12.445 .000 25.170
B, standard partial regression coefficient; SE, standard error; Wald, value of Wald test; C.I., confidence interval.

clinoid process, result in subtotally re- evidence. Five of the 12 eyes involving optic lateral approach (23), and endonasal trans-
moval, invade into unilateral optic canal canal but without orbital involvement got sphenoidal approach (8, 10, 16, 18, 19). No
and make severe compression on the ON (3, visual recovery (41.7%), whereas another patient received radiation therapy as a pri-
4). Patients with TSM, PM, and diaphragma four eyes underwent visual deterioration. mary treatment. Six kinds of approaches
sellae meningioma (DSM) were apt to get We supposed that optic canal unroofing can were used in our series. Many authors rec-
visual recovery (respectively 69%, 75%, decompress ON and release extensive space ommended some certain kind of approach
78.6%) than the former two groups. Optic for manipulation, but the radical optic canal as the optimal choice for their patients (11,
canal involvement in SMs had been re- unroofing and surgical interference might 14, 22, 23, 33). We thought the choice
ported in many articles (2, 14, 16, 33). Some do harm to the microvasculature in the should be determined according to several
authors considered that patients with tu- arachnoid membranes, which crawl in the factors, such as patients’ condition, the tu-
mor invasion into the optic canal had worse optic canal, especially when the tumor ad- mor size and location, direction of tumor
visual recovery (2, 30, 33) and suggested hered closely to ON. invasion, the surgeons’ preference, etc.
aggressive skull base surgery, routine un- The preservation of integrality of the Unilateral ACPM and smaller (⬍3cm) me-
roofing of the optic canal (30), and com- arachnoid membrane interface has been sal TSM can be resected via a unilateral sub-
plete tumor removal together with any dura emphasized by many authors (11, 41). An frontal approach. For planum sphenoidale
involved (33). However, Bassiouni et al. (3) intact arachnoid membrane was correlated meningioma (PSM) and larger (⬎5 cm)
observed tumor encroachment only in 5 of with a better visual outcome (25, 33, 41). TSM that has a tendency to develop into the
62 patients (8%), opened the optic canal in Our data also showed the same result (P ⫽ anterior cranial fossa, a bilateral subfrontal
2 of them, and did not suggest routine un- 0.006). It is necessary to pay attention to the approach is a better choice. For a median-
roofing. Our results showed that optic canal anatomy and function of the subarachnoid size (3–5 cm) tumor, the two aforesaid ap-
or orbit invasion occurred in 15 of 45 pa- cistern (11). We suppose the material causes proaches are both available, depending on
tients (33.3%) and two bilaterally, corre- of unsatisfactory visual outcome in cases many factors such as patients’ physical con-
sponding to 17 of 90 eyes, in which the tu- without intact arachnoid membrane may be dition, tumor location, etc. Subfrontal ap-
mor further extended into the orbital two aspects: encasement of ON and arteries proach is a good access to the contralateral
compartment in 5 eyes. The extent of optic that induced difficulty in preserving the ar- ON as well as the ipsilateral ON. The surgi-
canal invasion was influential to visual teries or microvessels feeding ON, tumor cal field toward the tuberculum sellae and
prognosis (P ⫽ 0.04). The falciform mem- invasion, and infiltration into ON. diaphragma sellae is also an unobstructed
branous part of the optic canal was opened view. Here, we emphasized the frontal-in-
and the bony part of the optic canal was terhemispheric approach. All of the four pa-
drilled away using a minute diamond drill, Surgical Approaches tients (three TSM and one PSM) operated
and then the tumor was removed. However, Many surgical approaches had been de- via this approach got vision improvement
we do not recommend unroofing as a con- scribed, including a unilateral frontal or bi- postoperatively, without severe complica-
ventional procedure. It should be per- lateral frontal approach (2, 9, 22, 23, 26), tions except for one temporary diabetes in-
formed when optic canal involvement has pterional approach (3, 4, 14, 39), or pterion- sipidus and one mild hyposmia. Superior
been confirmed intraoperatively, or at least al–frontotemporal approach (11, 23, 25, 26, sagittal sinus was not sutured or only a little
highly suspected by preoperative radiologic 33), supraorbital approach (1, 13), fronto- part of the front cerebral falx was cut off,

300 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, DOI:10.1016/j.wneu.2010.08.021

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

and the contralateral frontal lobe was pre- long period despite its compression on the
served better while the operational field was optic apparatus (25). Furthermore, pa- REFERENCES
still large. Pterional approach is fit to DSM, tients’ optic nerves or optic chiasma might 1. Al-Mefty O: Supraorbital-pterional approach to
ACPM, and most TSM, in particular those be extremely sensitive to radical surgical skull base lesions. Neurosurgery 21:474-477, 1987.
that grow into the parasellar space. This manipulation or any other disadvantageous 2. Andrews BT, Wilson CB: Suprasellar meningiomas:
approach can provide the shortest distance stimulus after long-term compression by the effect of tumor location on postoperative visual
to the sella region, with a slight stretch or mass lesions. We found intraoperatively outcome. J Neurosurg 69:523-528, 1988.
drag of the cerebral lobe, while preventing that patients with preoperative optic disc
opening of the frontal sinus and protecting 3. Bassiouni H, Asgari S, Stolke D: Tuberculum sellae
atrophy also had obvious lesions of optic meningiomas: functional outcome in a consecutive
the olfactory tract. However, the undersur- apparatus, including visibly pale and atro- series treated microsurgically. Surg Neurol 66:37-
face of the ON and OC are not as well visu- phied nerves, tumor infiltration, or ON im- 44; discussion 44-45, 2006.
alized as in the subfrontal approach (20). pingement by tumor compression. The 4. Bassiouni H, Asgari S, Sandalcioglu IE, Seifert V,
For the tumors of complex or wide base, a state of preoperative optic disc can reflect Stolke D, Marquardt G: Anterior clinoidal meningi-
frontal–pterional approach that possesses the pathologic condition of optic apparatus omas: functional outcome after microsurgical re-
mutual advantages is also a good choice. section in a consecutive series of 106 patients. Clin-
clearly and directly. ical article. J Neurosurg 111:1078-1090, 2009.
Since Weiss (38) first described the ex-
According to previous studies and our
tended transsphenoidal approach in 1987,
data, we suppose the risk factors of recur- 5. Bitzer M, Klose U, Geist-Barth B, Nägele T, Schick
there have been many reports on extended F, Morgalla M, Claussen CD, Voigt K: Alterations in
rent meningioma to visual outcome also in-
endonasal transsphenoidal removal of su- diffusion and perfusion in the pathogenesis of peri-
clude abnormal structures around the tu- tumoral brain edema in meningiomas. Eur Radiol
prasellar and parasellar lesions, including
mor left over after the first surgery and the 12:2062-2076, 2002.
TSMs (8, 10, 16, 18, 19). We had limited
experience in this approach when dealing fragile ON that could not suffer more dam-
6. Chi JH, McDermott MW. Tuberculum sellae menin-
with SMs, and we did not suggest it as a age. Clear brain edema around SMs may giomas. Neurosurg Focus 14:e6, 2003.
routine strategy for SMs. One patient with induce further injury to visual apparatus be-
sides tumor direct compression. This may 7. Chicani CF, Miller NR: Visual outcome in surgically
binocular visual decrease was operated on treated suprasellar meningiomas. J Neuroophthal-
via this approach because of misdiagnosis be related to the ON impingement at the mol 23:3-10, 2003.
as pituitary adenoma. Although the patient falciform ligament and the compression on
got visual recovery postoperatively, it was the draining veins which induced the cere- 8. Cook SW, Smith Z, Kelly DF: Endonasal transsphe-
noidal removal of tuberculum sellae meningiomas:
difficult to arrest bleeding and remove the bral tissue swelling. Ischemia, necrosis of technical note. Neurosurgery 55:239-244; discus-
tumor completely via an endonasal ap- neurons, and demyelination of optic nerves sion 44-46, 2004.
proach, because of abundant blood supply may be the main pathologic changes under
the influence of edema and compression. 9. Cushing H, Eisenhardt L: Suprasellar meningio-
and the hard texture of meningioma plus an
mas. In: Cushing H, Eisenhardt L, eds. Meningio-
inadequate operative field, unlike the soft However, the accurate reason is still uncer- mas. Baltimore: Charles C. Thomas; 1938:224-249.
pituitary adenoma. Moreover, the repair of tain.
skull base might be more difficult than in a 10. de Divitiis E, Cavallo LM, Cappabianca P, Esposito F:
Extended endoscopic endonasal transsphenoidal
pituitary adenoma because of more open approach for the removal of suprasellar tumors: Part
portions of the sella bone. This patient de- 2. Neurosurgery 60:46-58; discussion 58-59, 2007.
veloped CSF rhinorrhea after surgery for a CONCLUSION
long time. After all, our experience on this 11. Fahlbusch R, Schott W: Pterional surgery of menin-
No surgical approach can be considered
giomas of the tuberculum sellae and planum sphe-
approach with respect to SMs is too little. perfect for the patients of SMs, but for every noidale: surgical results with special consideration
single individual, some kind of approach of ophthalmological and endocrinological out-
comes. J Neurosurg 96:235-243, 2002.
can be most suitable. Skillful microsurgical
Multivariate Analysis
technique is the key to total tumor removal 12. Galal A, Faisal A, Al-Werdany M, El Shehaby A, Lotfy
Multivariate analysis located quite obvious
with good outcome. In view of the signifi- T, Moharram H: Determinants of postoperative vi-
statistical significance with regard to the in- sual recovery in suprasellar meningiomas. Acta
fluence of duration of visual symptoms on cant influential factors, univariate analysis
Neurochir (Wien) 152:69-77, 2010.
visual recovery, and it was also thought to brought into recurrence, duration of visual
be a main factor influencing visual outcome decrease, preoperative condition of optic 13. Hernesniemi J, Ishii K, Niemelä M, Smrcka M, Kivi-
disc, peritumoral edema, age, severity of vi- pelto L, Fujiki M, Shen H: Lateral supraorbital ap-
in a previous article (11). Because of some proach as an alternative to the classical pterional
other diseases that can also impair vision as sual disturbance, tumor size, location, in- approach. Acta Neurochir Suppl 94:17-21, 2005.
comorbidities, and the slow and insidious volvement with optic canal, arachnoid
growth pattern of meningioma (6), SMs can membrane interface and extent of tumor re- 14. Jallo GI, Benjamin V: Tuberculum sellae meningio-
mas: microsurgical anatomy and surgical tech-
remain undiscovered or be misdiagnosed moval, but multivariate analysis confirmed
nique. Neurosurgery 51:1432-1439; discussion
for a long time. In our series, 10 cases were the first four factors as the most important 1439-1440, 2002.
misdiagnosed as cataract, optic neuritis, agents. However, this is not to say other
and other ophthalmopathy for a long period factors unimportant. Clinical significances 15. Kim TW, Jung S, Jung TY, Kim IY, Kang SS, Kim SH:
Prognostic factors of postoperative visual outcomes
from 5 months to even 10 years. Even large of the four items still need to be testified in tuberculum sellae meningioma. Br J Neurosurg
meningiomas may be asymptomatic for a with more researches. 22:231-234, 2008.

WORLD NEUROSURGERY 75 [2]: 294-302, FEBRUARY 2011 www.WORLDNEUROSURGERY.org 301

 PEER-REVIEW REPORTS
CHUAN-WEI WANG ET AL. POSTOPERATIVE VISUAL OUTCOME OF SUPRASELLAR MENINGIOMAS

16. Kitano M, Taneda M, Nakao Y: Postoperative im- tuberculum sellae meningiomas. Acta Neurochir 35. Stirling JW, Edin M: Tumor of the meninges in the
provement in visual function in patients with tuber- (Wien) 147:1121-1130; discussion 1130, 2005. region of the pituitary body, pressing on the chi-
culum sellae meningiomas: results of the extended asma. Ann Ophthalmol 6:15-16, 1897.
transsphenoidal and transcranial approaches. J 26. Puchner MJA, Fischer-Lampsatis RCM, Herrmann
Neurosurg 107:337-346, 2007. HD, Freckmann N: Suprasellar meningiomas— 36. Symon L, Rosenstein J: Surgical management of su-
neurological and visual outcome at long term fol- prasellar meningioma. Part 1: The influence of tu-
17. Landers J, Sharma A, Goldberg I, Graham S: A com- low-up in a homogeneous series of patients treated mor size, duration of symptoms, and microsurgery
parison of perimetric results with the Medmont and microsurgically. Acta Neurochirurgica 140:1231- on surgical outcome in 101 consecutive cases. J Neu-
Humphrey perimeters. Br J Ophthalmol 87: 1238, 1998. rosurg 61:633-641, 1984.
690-694, 2003.
27. Raco A, Bristot R, Domenicucci M, Cantore G: Me- 37. Vignes JR, Sesay M, Rezajooi K, Gimbert E, Liguoro
18. Laufer I, Anand VK, Schwartz TH: Endoscopic, en- ningiomas of the tuberculum sellae. Our experience D: Peritumoral edema and prognosis in intracranial
donasal extended transsphenoidal, transplanum tr- in 69 cases surgically treated between 1973 and meningioma surgery. J Clin Neurosci 15:764-768,
anstuberculum approach for resection of supra- 1993. J Neurosurg Sci 43:253-260; discussion 260- 2008.
sellar lesions. J Neurosurg 106:400-406, 2007. 262, 1999.
38. Weiss M: The transnasal transsphenoidal ap-
19. Laws ER, Kanter AS, Jane JA, Dumont AS: Extended 28. Rosenstein J, Symon L: Surgical management of su- proach. In Apuzzo MLJ, ed. Surgery of the Third
transsphenoidal approach. J Neurosurg 102:825- prasellar meningioma. Part 2: prognosis for visual Ventricle. Baltimore: Williams & Wilkins; 1987:
827, 2005. function following craniotomy. J Neurosurg 61: 476-494.
642-648, 1984.
20. Li X, Liu M, Liu Y, Zhu S: Surgical management of
39. Yasargil MG, Antic J, Laciga R, Jain KK, Hodosh RM,
Tuberculum sellae meningiomas. J Clin Neurosci
29. Rupp-Montpetit K, Moody ML: Visual loss as a com- Smith RD: Microsurgical pterional approach to an-
14:1150-1154, 2007.
plication of non-ophthalmologic surgery: a review eurysms of the basilar bifurcation. Surg Neurol
of the literature. AANA J 72:285-292, 2004. 6:83-91, 1976.
21. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK,
Burger PC, Jouvet A, Scheithauer BW, Kleihues P:
30. Sade B, Lee JH: High incidence of optic canal in- 40. Yasargil MG: Microneurosurgery. Vol. 1. Stuttgart;
The 2007 WHO classification of tumours of the cen-
volvement in tuberculum sellae meningiomas: ra- New York: Georg Thieme Verlag; Thieme Medical
tral nervous system. Acta Neuropathol 114:97-109,
tionale for aggressive skull base approach. Surg Publishers; 1987.
2007.
Neurol 72:118-123; discussion 123, 2009.
22. Margalit N, Kesler A, Ezer H, Freedman S, Ram Z: 41. Zevgaridis D, Medele RJ, Muller A, Hischa AC,
Tuberculum and diaphragma sella meningioma— 31. Schankin CJ, Ferrari. U, Reinisch VM, Birnbaum T, Steiger HJ: Meningiomas of the sellar region pre-
surgical technique and visual outcome in a series of Goldbrunner R, Straube A: Characteristics of brain senting with visual impairment: impact of various
20 cases operated over a 2.5-year period. Acta Neu- tumour-associated headache. Cephalalgia 27:904- prognostic factors on surgical outcome in 62 pa-
rochir (Wien) 149:1199-1204; discussion 1204, 911, 2007. tients. Acta Neurochirurgica 143:471-476, 2001.
2007.
32. Schankin CJ, Krumbholz M, Sostak P, Reinisch VM,
23. Nakamura M, Roser F, Struck M, Vorkapic P, Samii Goldbrunner R, Straube A: Headache in patients Conflict of interest statement: The authors declare that the
M: Tuberculum sellae meningiomas: clinical out- with a meningioma correlates with a bone-invasive article content was composed in the absence of any
come considering different surgical approaches. growth pattern but not with cytokine expression. commercial or financial relationships that could be
Neurosurgery 59:1019-1028; discussion 1028-1029, Cephalalgia 30:413-424, 2010. construed as a potential conflict of interest.
2006. received 12 January 2010; accepted 27 August 2010
33. Schick U, Hassler W: Surgical management of tu-
24. Nakasu S, Nakasu Y, Nakajima M, Matsuda M, berculum sellae meningiomas: involvement of the Citation: World Neurosurg. (2011) 75, 2:294-302.
Handa J: Preoperative identification of meningio- optic canal and visual outcome. J Neurol Neurosurg DOI: 10.1016/j.wneu.2010.08.021
mas that are highly likely to recur. J Neurosurg 90: Psychiatry 76:977-983, 2005. Journal homepage: www.WORLDNEUROSURGERY.org
455-462, 1999.
Available online: www.sciencedirect.com
34. Simpson D: The recurrence of intracranial meningi-
25. Pamir MN, Ozduman K, Belirgen M, Kilic T, Ozek omas after surgical treatment. J Neurol Neurosurg 1878-8750/$ - see front matter © 2011 Elsevier Inc.
MM: Outcome determinants of pterional surgery for Psychiatry 20:22-39, 1957. All rights reserved.

302 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, DOI:10.1016/j.wneu.2010.08.021