|Home | About | Journals | Submit | Contact Us | Français|
Various surgical approaches for the removal of sellar region lesions have previously been described. This study aimed to evaluate the reliability and safety of the frontolateral approach (FLA) to remove sellar region lesions.
We presented a retrospective study of 79 patients with sellar region lesions who were admitted and operated by the FLA approach from August 2011 to August 2015 in Department of Neurosurgery of Beijing Tian Tan Hospital. We classified FLA into three types, compared the FLA types to the areas of lesion invasion, and analyzed operation bleeding volume, gross total resection (GTR) rate, visual outcome, and mortality.
Seventy-nine patients were followed up from 2.9 to 50.3 months with a mean follow-up of 20.5 months. There were 42 cases of meningiomas, 25 cases of craniopharyngiomas, and 12 cases of pituitary adenomas. The mean follow-up Karnofsky Performance Scale was 90.4. GTR was achieved in 75 patients (94.9%). Two patients (2.5%) had tumor recurrence. No patients died perioperatively or during short-term follow-up. Three patients (3.8%) with craniopharyngioma died 10, 12, and 23 months, respectively, after surgery. The operative bleeding volume of this study was no more than that of the other approaches in the sellar region (P = 0.783). In this study, 35 patients (44.3%) had visual improvement after surgery, 38 patients (48.1%) remained unchanged, and three patients’ visual outcome (3.8%) worsened.
FLA was an effective approach in the treatment of sellar region lesions with good preservation of visual function. FLA classification enabled tailored craniotomies for each patient according to the anatomic site of tumor invasion. This study found that FLA had similar outcomes to other surgical approaches of sellar region lesions.
Sellar region lesions are deeply seated and are in close proximity to neurovascular structures of the anterior skull base, such as the optic nerve, optic chiasm, pituitary stalk, and hypothalamus. Mass effect and compression of these structures can lead to a wide variety of clinical symptoms, including visual defects, polydipsia, polyuria, acromegaly, gigantism, amenorrhea, and galactorrhea.
Resection of sellar region lesions has been a challenge for neurosurgeons for nearly a century. In 1908, Krause first described the treatment of sellar region lesions by using a superior orbital subfrontal approach. Tandler and Ranzi applied the same approach to superior sellar region lesions in 1920. In 1971, Wilson pioneered the concept of limited exposure in cerebral surgery by minimizing any unnecessary surgical interference. Yasargil and Fox further modified the frontal, lateral pterion approach of Dandy by drilling at the sphenoidal ridge in 1975. In 1978, Brock and Dietz reported the first successful treatment of anterior circulation aneurysms via the small frontolateral approach (FLA). Samii used the subfrontal lateral space to resect olfactory groove meningiomas. In 1999, Perneczky et al. presented the “Keyhole” concept, and modified “frontolateral approach” to “frontolateral keyhole” approach. In 2008, Samii and Gerganov reported the experiences of FLA applied in sellar region lesions and described it as a simple and safe approach. Moreover, Romani et al. reported the experiences of using FLA, which was called as a lateral supraorbital approach, in 52 patients with tuberculum meningioma in 2012.
The most frequent microsurgical approaches to remove sellar region lesions are pterional, fronto-orbital and fronto-orbitozygomatic, frontolateral, and bifrontal. FLA has both features and exposure advantages of subfrontal and pterional approach, and many neurosurgeons apply this approach in sellar region lesions. In this article, we retrospectively reviewed the outcomes of 79 patients with sellar region lesions using the FLA.
One hundred and seven patients with sellar region lesions were operated by FLA in Department of Neurosurgery of Beijing Tian Tan Hospital from August 2011 to August 2015. Twenty-eight patients were lost to follow-up and a total of 79 patients were included in this study. All cases were analyzed retrospectively. Pre- and post-operative clinical conditions were expressed using the Karnofsky Performance Scale (KPS). Visual acuity and visual fields were tested by a neuro-ophthalmologist before and after surgery. Both the Goldmann and Octopus perimetry systems were used for visual field testing. Operative bleeding volume was collected from the operation records.
Seventy-nine patients underwent magnetic resonance imaging (MRI) before and after surgery. Forty-two patients underwent an additional computed tomography (CT) or magnetic resonance angiography because of the close relation between the lesion and major vessels including the internal carotid artery (ICA), anterior cerebral artery (ACA), middle cerebral artery (MCA), anterior communicating artery (AcomA) or cavernous sinus. The diameter of the lesion and affected structures was recorded from preoperative MRI while the extent of lesion resection was measured by comparing pre- and post-operative MRI and the surgical records. All patients underwent CT within 1 day after surgery.
Patients were operated in a supine position with the head rotated to the contralateral side at an angle of 30°–45°. FLA was classified into three types based on the anatomic location of lesion invasion [Figure 1].
After the bone flap was performed, the dura was opened in a C-shaped, semilunar fashion with its base toward the supraorbital rim, followed by opening of the Sylvian fissure and the carotid cisterns for cerebrospinal fluid (CSF) drainage.
Short-term follow-up data were available for patients at 6 months and 1 year after surgery and included a clinical examination and MRI scan. The patients who underwent surgery <6 months received follow-up by telephone. Long-term follow-up was performed every 1–2 years thereafter and data were collected either through outpatient records or through home visits.
All descriptive and statistical analyses were performed using IBM SPSS software version 22.0 (IBM Corporation, New York, USA) and NCSS version 10 (Copyright© 1983–2016, NCSS, LLC, Utah, USA, All Rights Reserved). The FLA type and anatomic region of tumor invasion were compared by cross-tabulation through a Pearson's Chi-squared test, and age, operative bleeding volume, and tumor size were compared using Student's t-test. Age, operative bleeding volume, and tumor size were shown as mean ± standard deviation (SD). A P < 0.05 was considered statistically significant.
Seventy-nine patients were included in this study (29 males, 50 females), with the mean age of 44.9 ± 11.7 years. Unilateral visual acuity defect was found in 16 patients, and bilateral visual acuity defect was in 35 patients. Forty-three patients had visual field defect. There were 42 cases of meningiomas, 25 craniopharyngiomas, and 12 pituitary adenomas [Figures [Figures22–4]. Gross total resection (GTR) was achieved in 75 patients (94.9%) and STR in four patients (5.1%).
Surgical approach and characteristics of tumors
Twenty patients had undergone FLA Type I, the mean operative bleeding volume was 360.0 ± 303.8 ml: eight patients had tumor size ≤3 cm and 12 patients had tumor size >3 cm. Tumors invaded the greater wing of the sphenoid bone in one case, anterior clinoidal process in one case, olfactory groove or sphenoidal platinum in ten cases, cavernous sinus in two cases, and the third ventricle in eight cases. There were 11 cases of meningioma, four cases of craniopharyngioma, and five cases of pituitary adenomas. GTR was achieved in 19 patients. One patient with pituitary adenoma received subtotal resection (STR) because of tumor invasion of the cavernous sinus.
Twenty-six patients had undergone FLA Type II, the mean operative bleeding volume was 292.9 ± 217.1 ml: 16 patients had tumor size ≤3 cm and ten patients had tumor size >3 cm. Tumors invaded the greater wing of sphenoid bone in one case, anterior clinoidal process in ten cases, olfactory groove or sphenoidal platinum in eight cases, cavernous sinus in three cases, and the third ventricle in three cases. There were 13 cases of meningioma, ten cases of craniopharyngioma, and three cases of pituitary adenoma. GTR was achieved in 25 patients. One meningioma patient received STR because of the tumor invasion of the cavernous sinus.
Thirty-three patients had undergone FLA Type III, the mean operative bleeding volume was 475.0 ± 148.1 ml: 16 patients had tumor size ≤3 cm and 17 patients had tumor size >3 cm. Tumor invaded the greater wing of sphenoid bone in 12 cases, anterior clinoidal process in ten cases, cavernous sinus in 14 cases, and the third ventricle in ten cases. About 18 cases were meningioma, 11 cases were craniopharyngioma, and four cases were pituitary adenoma. GTR was achieved in 31 patients. Two patients received STR: one meningioma patient with tumor invasion of the cavernous sinus and anterior clinoidal process and one pituitary adenoma patient with tumor invasion of the cavernous sinus.
The FLA type and region of tumor invasion were compared through cross-tabulation by Pearson's Chi-squared test. In the first analysis, five regions (greater wing of the sphenoid bone, anterior clinoidal process, olfactory groove or sphenoidal platinum, cavernous sinus, and third ventricle) were compared to FLA type. A P value, P1=0.000, was achieved but five cells (33.3%) have expected count <5 wherein the minimum expected count was 3.11 (Chi-square value: 28.980). In the second analysis, the greater wing of the sphenoid bone and anterior clinoidal process were merged as one region for analysis. The resulting of four regions was compared to FLA type. A P value, P2=0.002, was achieved and two cells (16.7%) have expected counts <5 wherein the minimum expected count was 4.44 (Chi-square value: 21.352). The result showed FLA Types I, II, and III had significant differences in the regions of tumor invasion.
In the study by Zhang et al. (as control group), the average operative bleeding volume of thirty patients was 375.0 ml. The mean operative bleeding volume in this group was 389.9 ± 289.9 ml [Table 1]. There were no significant differences for gender, age, tumor size, and patient numbers achieved GTR between our study and Zhang et al.'s study (all P > 0.05). The operative bleeding volume of this study was no more than that of Zhang et al.'s study (P = 0.783).
Seventy-nine patients were followed up from 2.9 to 50.3 months with a median follow-up of 20.5 months. Among the 79 patients, 64 patients (81.0%) had a normal life (KPS: 90–100), 12 (15.2%) had moderate disabilities (KPS: 60–80), and 3 (3.8%) were dead (KPS: 0). The median follow-up KPS was 90.4. Shallow postoperative forehead wrinkles were present in two cases that underwent FLA Type II. One case undergoing FLA Type II had transient CSF leakage immediately after operation because of a large frontal sinus, which was subsequently resolved after 10 days of continuous lumbar drainage. One patient with pituitary adenoma, who underwent FLA Type I, suffered a skin infection caused by an obstructive frontal sinus mucocele 38 months after surgery. One patient with meningioma who received FLA Type III had temporalis atrophy. Four patients with craniopharyngioma (1 in FLA Type II and 3 in FLA Type III) had polydipsia and polyuria, which resolved 7 months after surgery in one patient, and remained in three patients at follow-ups of 21, 26, and 27 months after surgery, respectively. One meningioma patient undergoing FLA Type III had an episode of epilepsy 20 months after surgery. One patient undergoing FLA Type II lost olfactory function on the operative side, which did not recover 20 months after surgery.
Tumor recurrences were found in two patients: one patient with tuberculum sellae meningioma recurred 34 months after surgery and underwent the operation again; and another patient with craniopharyngioma recurred 20 months after surgery and received radiation therapy. The other two patients received radiotherapy after surgery, and no recurrence was found during follow-up.
No patient died perioperatively or during short-term follow-up (<6 months). Three patients (3.8%) with craniopharyngioma died from severe hypopituitarism 10, 12, and 23 months after surgery, respectively.
In this study, 35 patients (44.3%) had visual improvement after surgery, 38 patients (48.1%) remained unchanged, and three patients’ visual outcome (3.8%) worsened.
In this study, FLAs were performed on sellar region lesions. According to the different anatomic region of tumor invasion, we classified FLA into three types. FLA Types I, II, and III had significant differences in tumor invaded areas. We were able to tailor the procedure to each patient according to the specific anatomic location of tumor invasion. FLA was more powerful in tran-orbital-cranial lesions and more flexible combined with other approaches. In the research of Maier et al. and Adawi and Abdelbaky, the FLA were also suitable for tumors of the posterior orbit. Yamada et al. also reported the combination of an orbitozygomatic approach and Le Fort I osteotomy under frontolateral craniotomy in the treatment of a large juvenile nasopharyngeal angiofibroma in an 8-year-old boy. In the report of Gerganov et al., they focused on the feasibility and efficacy of microsurgical resection of extensive craniopharyngiomas using an FLA. Metwali et al. introduced the surgical technique of medial optic nerve mobilization for better inspection and preservation of the pituitary stalk via an FLA.
The median operative bleeding volume of this study showed no more than the subfrontal approach. Only one patient in this study had the operative bleeding volume of 2000 ml. The meningioma of this patient was larger than 3 cm (6.5 cm in diameter) and invaded the anterior clinoid process and the bottom of the third ventricle. Ipsilateral ACA and AcomA were encased. The FLA Type III was applied in this patient, with the intent to gain more exposure of the Sylvian fissure. However, bleeding from the AcomA was difficult to coagulate because of the poor exposure from the cerebral longitudinal fissure.
Among the 79 patients, the median follow-up KPS was 90.4. Forehead wrinkles were shallow postoperatively in two patients who received FLA Type II, and the superior orbital nerve was bruised during the formation of the bone flap from the high-speed milling tool in the early stage of this group. From prior studies, CSF leakage ranged from 4% to 33%.[16,17,18,19,20,21] In this study, only one patient with FLA Type II had transient CSF leakage after operation because of a large frontal sinus, which was resolved after 10 days of continuous lumbar drainage.
One patient with pituitary adenoma in the FLA Type I group suffered skin infection caused by an obstructive frontal sinus mucocele 38 months after surgery. In FLA Type, I or Type II with large frontal sinuses, the treatment of the opened frontal sinus was difficult. In this study, we tried the following methods to decrease the CSF leakage and infection rate: first, the hard dura was sutured in a water-tight manner; second, the mucosa of the frontal sinus had to be completely removed and the smashed temporalis muscle was used to fill the sinus cavity; finally, a galea aponeurotica-pericranium flap with an intact blood supply was used to seal the opening of the sinus cavity.
In FLA Type III, the incision and retraction of the temporalis muscle were still larger than those in Types I and II. One patient with meningioma receiving FLA Type III had temporalis atrophy in the early stage. The overdose coagulation with mono-polar or bipolar on temporalis muscle and the decreased blood supply were the main reasons for temporalis atrophy.
Three patients (3.8%) with craniopharyngioma remained polydipsic and polyuric at follow-up, which was similar to the results of Shi et al.'s study. In Shi et al.'s study, two patients out of 110 craniopharyngioma cases (1.8%) still presented with polydipsia and polyuria 2 years after surgery. The incidences of postoperative epilepsy after tuberculum sellae meningiomas (TSM) surgery in the literature ranges from 2% to 11%.[16,19,23] In this study, only one patient with meningioma had a single episode of epilepsy at 20 months after surgery, the incidence rate was 1.3%, which was lower than previously reported.
Tumor recurrence was found in two patients with one tuberculum sellae meningioma and one craniopharyngioma, respectively. In previous literature,[24,25,26,27,28,29] the recurrence rate of craniopharyngiomas was from 9.3% to 41%, while the recurrence rate in this study was only 4.0% (1/25). The recurrence rates of meningioma previously reported were from 0 to 20% and it was 2.4% (1/42) in this study.[30,31,32,33,34]
In the literatures,[29,30,31,32,33,35] the mortality of craniopharyngioma was reported from 0 to 16.7%. In this study, the mortality of craniopharyngioma was 12% (3/25), and all three patients died from severe hypopituitarism during follow-up. All of them were living in other provinces outside of Beijing, and their situations were changing rapidly. If they were received hormone replacement therapy in time, their prognosis might be better.
The visual deterioration rate in this study was significantly lower than previously reported, and this might be attributed to good exposure of the optic nerve, anterior clinoidal process, and carotid artery of the FLA. In the report of Chokyu et al., 29 (90.6%) of 32 patients showed improved visual outcomes after TSM operations. However, the visual improvement rate in this study was 44.3% (35/79), the reasons might be due to the different timing of operation and tumor size.
The GTR rate and visual deterioration rate varied from 71% to 100% and 0% to 20%, respectively, in previous literature, and in this study, GTR was achieved in 75 patients (94.9%) and visual deterioration rate was 3.8%. The mortality was 3.8% in this study, similar to previously reported.[37,38,39,40,41,42,43,44] Because of potential differences in patient age, tumor size, and different pathological component distribution of each report, we could not confirm FLA was better, which need further study in the future to confirm it. However, at least FLA was comparable with other approaches of sellar region lesions.
In conclusion, we found that FLA was a safe and effective approach for the treatment of sellar region lesions with good preservation of visual function. We were able to tailor the procedure to each patient according to the specific anatomic location of tumor invasion. This study found that FLA had similar outcome to other surgical approaches of sellar region lesions.
This study was supported by grants from Natural Science Foundation of Beijing Municipality, China (No. 7142052) and Beijing Municipal Science and Technology Commission (No. Z131107002213179).
There are no conflicts of interest.
Edited by: Xin Chen