Childs Nerv Syst (2007) 23:1135–1145 DOI 10.1007/s00381-007-0371-0

ORIGINAL PAPER

Endoscopic options in management of posterior third ventricular tumors S. V. Roopesh Kumar & Aaron Mohanty & Vani Santosh & Satyanarayana Satish & B. Indira Devi & Shanti Shankar Praharaj & Sastry V. R. Kolluri

Received: 18 February 2007 / Revised: 24 April 2007 / Published online: 31 May 2007 # Springer-Verlag 2007

Abstract Objective A spectrum of both radiosensitive and radioresistant lesions occurs in the region of the posterior third ventricle (PTV). Most of these are associated with hydrocephalus requiring a cerebrospinal fluid diversion procedure. The present study aims to assess the effectiveness of endoscopic biopsy and third ventriculostomy (ETV) in these patients. Materials and methods Twenty-four patients with PTV lesions with moderate to severe hydrocephalus were managed prospectively. All patients underwent ETV and biopsy of the lesion during the same procedure. The ETV could be performed in all patients, where as in 23, a successful biopsy could be obtained. The ETV was successful in 22 patients; it failed in 2 patients requiring shunt insertion. A positive biopsy was obtained in all the patients (pinealocytoma 4, pinealoblastoma 10, embryonal S. V. Roopesh Kumar : S. Satish : B. I. Devi : S. S. Praharaj : S. V. R. Kolluri Department of Neurosurgery, National Institute of Mental Health and Neurosciences (NIMHANS), Hosur Road, Bangalore, India A. Mohanty (*) Department of Neurosurgery, University of Texas Health Sciences Center at Houston, 6410 Fannin # 1020, Houston, TX 77030, USA e-mail: [email protected] V. Santosh Department of Neuropathology, National Institute of Mental Health and Neurosciences (NIMHANS), Hosur Road, Bangalore, India

cell carcinoma 1, germinoma 2, oligodendroglioma 1, astrocytoma 2, tuberculoma 4). All patients were subsequently managed with further surgery, radiation, and chemotherapy either alone or in combination depending on the pathology. One patient with a tumor bed hematoma required clot evacuation. The follow-up period ranged from 12 to 36 months. Two patients died, one during hospital stay with a tumor bed hematoma and another at 6 months follow-up due to extensive leptomeningeal spread. Conclusions The high yield of endoscopic biopsy (100%) and success of ETV (91%) emphasizes its role in management of the diverse group of PTV lesions in arriving at the optimal definitive management. Keywords Biopsy . Endoscopic third ventriculostomy . Pineal tumors . Posterior third ventricular tumors . Tuberculoma

Introduction A variety of pathologies varying from benign pineal cysts to malignant tumors like pinealoblastomas and germinomas occurs in the region of the posterior third ventricle. The complexity in accessing this anatomical location makes the management of these lesions a considerable challenge even for the skilled neurosurgeon. Some of the lesions, e.g., germinomas, are extremely radiosensitive making radiotherapy as the primary management option [8, 25]. On the other hand, the benign lesions require surgical excision, as radiation therapy is not an effective alternative [10, 19, 48, 55]. A histological diagnosis is considered essential before the radiation therapy, as the conventional neuroradiological evaluation cannot specifically diagnose the various tumor entities in this location [20, 27, 35].

1136

Most of the posterior third ventricular tumors obstruct the aqueduct and are associated with hydrocephalus requiring a CSF diversion procedure. Conventionally, ventriculoperitoneal shunts have been performed for the accompanying hydrocephalus. In the past decade, the widespread usage of neuroendoscopic techniques in managing obstructive hydrocephalus and intraventricular lesions has added a new dimension in management of these lesions. An endoscopic third ventriculostomy (ETV) is an effective procedure in managing hydrocephalus associated with pineal tumors, as the obstruction occurs at the level of the aqueductal inlet. During the same procedure, by slightly angulating the endoscope posteriorly, the tumor in the posterior third ventricular region is easily visualized and can be biopsied. A histologic biopsy can segregate the radiosensitive tumors in which a conventional craniotomy can be avoided, and radiotherapy may be administered. The radio-resistant or poorly radiosensitive lesions would require a craniotomy and tumor decompression.

Materials and methods The present study was aimed at assessing the efficacy of endoscopic biopsy and ETV for patients with pineal region tumors. From January 2000 to December 2003, patients with pineal tumors were prospectively included in the study. Subsequent to obtaining the biopsy and the ETV, further definitive management of the tumor in the postoperative period was decided basing on the tumor pathology. The present report does not aim to address the issues involving the efficacy of the various therapeutic options for management of pineal region tumors, which would be discussed in a subsequent report. The clinical data of these patients is summarized in Table 1. All the patients had a preoperative computed tomography (CT) scan, which demonstrated the tumor with a varying degree of hydrocephalus. Additionally, a preoperative magnetic resonance imaging (MRI) was performed in 16 patients. The patients with predominantly cystic lesions were excluded from the study. In addition, identifiable tectal region tumors were excluded from the study. In the present series, all patients were operated by the senior author (AM). An anterior frontal precoronal burr hole was placed in all patients. As both tumor biopsy and third ventriculostomy were planned to be performed during the same procedure with a single entry, the burr hole site was carefully planned. In patients with a preoperative MRI, it was located by identifying the midpoint of the line joining anterior margin of the tumor to the proposed site of the third ventriculostomy and then joining this point to the posterior rim of the foramen of Monro and extrapolating it

Childs Nerv Syst (2007) 23:1135–1145

to the calvarium. In patients without an MRI, the burr hole was placed approximately 2 cm anterior to the conventional precoronal burr hole. The side of the burr hole was determined depending on the tumor extension to the left or right side, with the burr hole being placed opposite to the side of the predominant tumor location. This was performed to facilitate a direct vision with the rigid endoscope. In tumors with a predominant central location, the right prefrontal region was chosen. After the burr hole was made and the dura opened, the ventricle was cannulated with a 12-Fg peel-away sheath, and the cerebrospinal fluid (CSF) was collected for analysis. This included evaluation for malignant cells in the cytospin specimens. Subsequently, a Gaab’s rigid neuroendoscope was introduced into the lateral ventricle. The foramen of the Monro was identified, and the endoscope was advanced into the third ventricle. As almost all the patients required a CSF diversion on a priority basis, initially a third ventriculostomy was performed. This avoided a possible postponement of ETV in the event of bleeding from the tumor surface, if the tumor was biopsied early in the procedure. The floor was penetrated with 3-Fg Fogarthy’s catheter, and the same was used to dilate the opening to an acceptable size. The endoscope was then advanced to visualize the basal cisterns and the basilar artery to ensure that the Liliequest’s membrane was also fenestrated. The floor of the third ventricle was inspected for evidence of any tumor seedling. Upon a satisfactory completion of the procedure, the endoscope with the sheath was gently angulated posteriorly to visualize the tumor. The tumor usually was readily identified in the posterior III ventricular region obstructing the aqueduct (Fig. 1a). Extreme care was taken during the procedure not to injure the columns of the fornix while angulating the scope posteriorly. The tumor capsule was coagulated with the bipolar forceps, and the tumor was biopsied using 2-mm forceps. Multiple biopsies from various regions of the tumor surface were obtained to avoid a sampling error or a negative biopsy. A minimum of three biopsies was obtained during each procedure. Any minor bleeding during the procedure was managed by coagulation and continuous irrigation till the bleeding stopped. The endoscope was then gently withdrawn, and the third ventriculostomy site was evaluated for patency (Fig. 1b). The endoscope track was also visualized for any bleeding. The scalp was closed with continuous galeal stitches and interrupted skin stitches to ensure a watertight closure. A CT scan was performed in the first postoperative day to ensure that there was no intraventricular bleed or tumor bleed. The efficacy of third ventriculostomy was assessed by resolution of clinical symptoms in the postoperative period. In patients in whom the ETV failed, a ventriculoperitoneal shunt was performed. Further treatment was based on the histopathological diagnosis. Patients with

Pt no.

Age/ sex

Presentation

Duration

20/F

Increased ICP and fever

1 year

2

19/M

Increased ICP

4 months

3

18/M

Increased ICP Parinaud’s

1 week

4

11/F

Increased ICP

6 weeks

5

9/F

Increased ICP

2 months

6

32/M

Increased ICP

3 months

7

58/M

Increased ICP

6

8

11/M

Increased ICP Parinaud’s

1 week

9

35/F

Increased ICP Parinaud’s

6 months

10

2/M

Incessant cry+lower limb weakness

1 months

11

30/M

Increased with altered sensorium

6 weeks

12

33/M

Increased ICP Parinaud’s

6 months

13

45/M

Increased ICP with Parinaud’s

4 months

14

30/F

Increased ICP with Parinaud’s

1 week

15

35/F

Increased ICP and Parinaud’s

3 months

16

20/M

9 months

17

12/M

Increased ICP Parinaud’s and decreased cognition Increased ICP

18

22/F

Increased ICP

1 month

19

18/M

Increased ICP with Parinaud’s

2 months

20

17/M

MCP with Parinaud’s

6 months

2 months

MRI

Appearance

Hydro Calcification cephalus

Iso-dense enhancing Iso-dense Enhancing Iso-dense enhancing Iso-dense enhancing Hyper-dense enhancing Hyper-dense enhancing Hyper-dense enhancing Iso-dense enhancing Iso-dense enhancing

+

−

+

−

+

−

+

+

+

−

+

−

+

−

+

+

+

−

Hyper-dense enhancing Hypo-dense enhancing

+

+

+

−

Iso-dense enhancing Hypo-dense enhancing Iso-dense enhancing Hypo-dense enhancing Hyper-dense enhancing Iso-dense enhancing Iso-dense enhancing Iso-dense enhancing Iso-dense

+

+

+

−

+

+

+

+

+

−

+

−

+

−

+ +

ETV

Biopsy HPE diagnosis

Complication Tumor markers

CSF cytology

Further treatment

+

TB

−

−

−

ATT

−(STB) TB

−

−

−

ATT

+

TB

−

−

−

ATT

+

PB

−

−

+

CSRT+CT

+

PB

−

−

−

CRT+CT

Yes

+

GM

−

−

−

CSRT+CT

Yes

+

PC

−

−

−

Surgery

Yes

+

EC

−

+

−

Yes

+

PC

−

−

Yes

+

PB

IVH(tumor bed hematoma) −

CRT (stereotactic) +CT Surgery

−

−

CT

Yes (failed requiring shunt) Yes

−

ASTRO II

−

−

−

Surgery+RT

+

PC

−

−

−

Surgery

Yes

+

−

−

−

CRT + CT

Yes

+

OLIGO III PB

−

−

−

CRT+CT

Yes

+

PC

−

−

−

Surgery

Yes

+

PB

−

−

−

CRT+CT

Yes

+

TB

−

−

−

ATT

T2 Yes inversion − Yes T2 Yes inversion No Yes seedling − Yes No seedling − No seedling No seedling No seedling −

No seedling − No seedling No seedling No seedling −

Yes

+

PB

−

−

+

CSRT+CT

−

No seedling −

Yes

+

PB

−

−

−

CRT+CT

−

No

Yes

+

PB

−

−

+

CRT+CT

1137

1

CT

Childs Nerv Syst (2007) 23:1135–1145

Table 1 Clinical data of patients

− + 1 month 18/M 24

Raised ICP

1 month 56/F 23

Abnormal behavior

1 month Increased ICP 45/M 22

TB Tuberculoma; PB pinealoblastoma; PC pinealocytoma; GM germinoma; EC embryonal cell carcinoma; ASTROII astrocytoma gr II; ASTROIII astrocytoma gr III; OLIGOIII oligodedroglioma gr III; ICP intracranial pressure; STB stereotactic biopsy; ATT antitubercular therapy; CSRT craniospinal radiotherapy; CRT cranial radiotherapy; CT chemotherapy; ETV endoscopic third ventriculostomy

CSRT Yes

+

GM

+ − −

Surgery+CSRT − ASTRO III − + Yes

No seedling Seedling −

− +

+

Yes

+

PB

− − −

CRT+CT − − − 17/M 21

Increased ICP

4 months

enhancing Iso-dense enhancing Hyper-dense enhancing Iso-dense enhancing Iso-dense enhancing

+

+

seedling No seedling −

Yes

+

PB

CSF cytology Complication Tumor markers Biopsy HPE diagnosis ETV MRI Hydro Calcification cephalus CT Appearance Duration Presentation Age/ sex Pt no.

Table 1 (continued)

CRT+CT

Childs Nerv Syst (2007) 23:1135–1145 Further treatment

1138

extremely or relatively radiosensitive tumors (germinoma, pinealoblastoma) were considered for radiotherapy and chemotherapy. A craniospinal MRI and tumor markers in serum and CSF were performed when deemed appropriate. Poorly radiosensitive and radio-resistant tumors (pinealocytomas, low-grade gliomas) were considered for craniotomy and tumor decompression. Inflammatory granulomas were started on appropriate chemotherapeutic agents. The follow-up period ranged from 2 months to 3 years with a mean of 13 months. CT scans were performed at intervals to look for the extent of tumor regression and the ventricular size.

Results Twenty-four patients were operated during the study period (Table 1). Of these, 11 were in the pediatric age group (≤18 years). The predominant clinical presentation was raised intracranial pressure with some patients demonstrating signs suggestive of localization in the posterior third ventricular region. The third ventriculostomy could be performed in all the patients. In five patients, there was evidence of tumor seedling in the third ventricular floor. Four of these patients had pinealoblastomas, and the rest had a germinoma. The number of biopsies varied between 3 and 7 patients. All the specimen was sent for evaluation by permanent histopathology. In 23 of the 24 patients, a biopsy from the tumor could be obtained. In one patient, the biopsy could not be performed. This patient was operated in the initial part of the study period, and as the operating surgeon anticipated risk injury to the forniceal walls during the posterior angulation of the endoscope, the procedure was abandoned. This patient subsequently underwent stereotactic biopsy of the lesion. In the later part of the study period, we used a thin Gaab’s rigid fiberscope in four patients with narrow foramen of Monro to obtain the biopsy. A mild forniceal ependymal contusion was observed in six patients at the end of the procedure. In none, however, was there any breach of the ependymal lining to suggest significant forniceal injury. In two patients, the ETV failed in the early postoperative period requiring a VP shunt insertion. One of them had a tuberculoma in the posterior third ventricular region, and another had a pinealoblastoma. In the rest, ETV functioned well, relieving the symptoms of raised intracranial pressure. A histological diagnosis could be achieved in all the patients in whom biopsy was attempted (Table 1). There were ten pinealoblastomas (Fig. 2a-c), four pinealocytomas, two germinomas (Fig. 3a,b), two astrocytomas (grades II and III), one oligodendroglioma (grade III), and one

Childs Nerv Syst (2007) 23:1135–1145

1139

embryonal cell carcinoma. The rest four had histological features suggestive of tuberculoma (Fig. 4a-c). CSF obtained during the procedure was sent for analysis for malignant cells. In only four patients, it showed evidence of tumor cells. However, no definite diagnosis could be obtained with the CSF cytology. In one patient (case 9), the postoperative CT scan demonstrated a tumor hematoma with intraventricular hemorrhage (Fig. 5). She required an external ventricular drainage for the hydrocephalus. The histology of this

patient revealed pinealocytoma. She underwent a craniotomy for the evacuation of hematoma and removal of the residual tumor. However, subsequently, she developed aspiration pneumonia requiring ventilatory support and died a week later. The patients with a histological diagnosis suggestive of germinoma received cranial irradiation. The patients with pinealoblastomas were considered for craniospinal irradiation and chemotherapy. The youngest patient in this group, who was 2 years old, received only chemotherapy. Patients with a histological diagnosis of pinealocytomas and low-grade gliomas underwent occipital craniotomy, transtentorial approach, and decompression of the tumor. The child with embryonal cell carcinoma initially underwent several cycles of chemotherapy. The CSF tumor marker levels showed considerable decrease in beta human chorionic gonadotrophin (βHCG) and alpha fetoprotein (αFP) levels. He was subsequently administered radiotherapy. The patients with low- and intermediate-grade gliomas (cases 11, 23) were operated, and the tumor was decompressed adequately. The patient with oligodendroglioma grade 3 refused further surgery and was considered for radiation and chemotherapy. Antitubercular therapy consisting of isoniazid, rifampicin for 18 months and pyrazinamide for the initial 3 months was administered to the patients with tuberculomas. Six of the 24 patients underwent definitive surgery for tumor decompression (Table 2). In all these patients, the initial biopsy results concurred reasonably with the final histopathological diagnosis. The patients were followed up for a period of 12– 36 months. Follow-up CT scans were performed in all the patients at 3–6 month intervals. It showed complete to neartotal tumor resolution in six patients (two germinomas, three pinealoblastomas, and one embryonal cell carcinoma) and a considerable decrease in the size of the lesion in all four

Fig. 2 a Preoperative CT scan demonstrating a contrast-enhancing lesion in posterior third ventricle with hydrocephalus. b Histopathological examination of the biopsy material revealing evidence of

pinealoblastoma. c Follow up CT scan at 1 year after chemotherapy and radiotherapy revealing no significant residual tumor. The hydrocephalus has resolved

Fig. 1 a Endoscopic view of the tumor at the foramen of Monro. The tumor is seen posterior to the mass intermedia. b Endoscopic view of the third ventricle after completion of biopsy and third ventriculostomy. The third ventriculostomy site can be visualized at 12 o’clock position

1140

Childs Nerv Syst (2007) 23:1135–1145

Fig. 3 a Preoperative CT scan showing a large posterior third ventricular tumor with hydrocephalus. This tumor upon biopsy was proved to be a germinoma. b Follow-up CT scan at 1 year post radiation and chemotherapy. The tumor has completely resolved with radiation and chemotherapy

patients with tuberculomas. All the rest except one showed partial tumor resolution and are being followed up closely. One child with pinealoblastoma presented 3 months later with paraparesis. A craniospinal MRI revealed extensive seedling in the spinal subarachnoid space. The parents refused further treatment, and she died 3 months later.

Discussion Pineal region tumors comprise 0.4–1% of all intracranial tumors in adults and 5–10% in pediatric population [10, 22, 23, 25]. These pose a considerable challenge even to the skilled neurosurgeon in view of its proximity to the vital neural and vascular structures. This issue is further

Fig. 4 a Preoperative MRI scan lobulated lesion in the posterior hydrocephalus. b Upon endoscopic be a tuberculoma. The prominent

showing an enhancing multithird ventricular region with biopsy, the lesion was found to Langhans giant cells are well

complicated by the diversity of lesions peculiar to the pineal region and the associated hydrocephalus. The objectives of managing these tumors are to establish CSF diversion and effectively treat the primary tumor [30, 31, 35]. An accurate histological diagnosis is essential for the later [20, 27, 35]. In past few decades, there have been changing trends in the management of pineal region tumors [23, 27, 30]. As the early experience with open surgery in the pineal region was extremely discouraging, a number of patients were managed empirically with radiotherapy and shunting procedures [45]. With the advent of microsurgical techniques and improved neuroimaging, there has been a resurgence of open surgery for pineal region tumors in the past two decades [20, 54]. Depending on the location of the lesion and displacement of major veins, either an occipital transtentorial or an infratentorial supracerebellar approach has been practiced [3, 4]. Although in recent years a reduction in the overall mortality has been witnessed, the morbidity continues to remain high in patients harboring highly malignant tumors and those with prior radiotherapy [5, 18, 31, 47]. Surgical resection is considered to be curative in benign pineal region tumors, which comprise one third of the lesions in this location [27]. In the remaining two thirds of patients harboring malignant lesions, there has been considerable debate concerning the efficacy of cytoreductive surgery [12, 21, 26, 35, 47]. Several reports in the recent years have demonstrated the efficacy of non-resective management of such lesions by a combination of radio- and chemotherapy [1, 26, 35, 39]. However, a histological diagnosis is considered to be essential before considering such measures [27, 35]. There are several issues concerning the need for establishing the accurate histopathological diagnosis before

demonstrated (arrows). c CT scan at 1 year follow-up. The lesion resolved near totally with anti-tubercular treatment. The hydrocephalus resolved with ETV

Childs Nerv Syst (2007) 23:1135–1145

1141

Fig. 5 Postoperative CT scan of a patient with significant intratumoral bleed after endoscopic biopsy

entertaining major therapeutic decisions. The establishment of an accurate tissue diagnosis is essential due to the diversity of lesions occurring in the pineal region [25, 54]. In recent years, high-resolution CT and MRI have proven to be extremely sensitive in detecting the pineal tumors. However, they have a low specificity in predicting the accurate histological diagnosis [3, 11, 50]. Additionally, although with a cytospin evaluation of the lumbar CSF, tumor cells may be identified [23, 25], a histological diagnosis is often not possible [13]. Moreover, a lumbar puncture in the presence of obstructive hydrocephalus is hazardous due to the risk of herniation [51]. Stereotactic biopsy has been proven useful in the past to obtain a histological diagnosis in the posterior third ventricular region. In two large series of patients with pineal region tumors, the diagnostic yield by stereotactic biopsy was as high as 94% [28, 43]. The disadvantages include a relatively higher risk of bleeding, which is attributed to the multiple pial surface breaches during the biopsy, proximity to the adjacent deep venous structures, and lack of tamponade effect. An 8% morbidity associated with stereotactic biopsy of pineal region tumors has been reported, which appears to be considerably higher than the 3% risk evident in stereotactic biopsy of other regions. In addition, an increased mortality has been reported with stereotactic biopsy in pineal region tumors (1.3–1.9%) compared to other regions (0.81%) [28, 43]. Relatively higher risks of complications have been reported with firm Table 2 Concurrence in pathological diagnosis between the initial endoscopic biopsy and the definitive surgery Case no.

Pathology obtained at endoscopic biopsy

Pathology obtained at microsurgery

7 9 11 12 15 24

Pinealocytoma Pinealocytoma Low-grade astrocytoma Pinealocytoma Pinealocytoma Anaplastic astrocytoma

Pinealocytoma Pinealocytoma grade 2 Astrocytoma grade 2 Pinealocytoma Pinealocytoma Anaplastic astrocytoma grade 3

tumors (teratoma, pinealocytoma, astrocytoma) as compared to other tumors [43]. Furthermore, the option of stereotactic biopsy warrants the requirement of a CSF diversion procedure as an additional procedure. Various tumor markers in serum and CSF have been used for arriving at a diagnosis in the lesions occurring in the pineal region. Elevated αFP and βHCG levels in serum and CSF have been considered by some to be sufficient enough to entertain the diagnosis of malignant nongerminomatous germ cell tumors in the presence of a mass in the posterior third ventricle [19]. However, obtaining a CSF sample by lumbar puncture preoperatively runs the risk of precipitating a tonsillar herniation [51]. Moreover, several others consider tumor markers to lack specificity and do not recommend to be used as the sole criteria for establishing a diagnosis [14, 17, 38]. Hence, it has been considered that a combination of biopsy, neuroimaging, and tumor markers is required for optimal diagnosis and management of pineal region tumors [18]. In the present series, one patient with an embryonal cell carcinoma (case 8) had evidence of elevated tumor markers in the serum and CSF. On last evaluation, the child is on remission with undetectable markers in the serum and the CSF.

Management of hydrocephalus The associated hydrocephalus requires a CSF diversion in more than 90% of the patients [44]. Conventionally, a ventriculoperitoneal shunt has been performed in these patients. However, the association of shunt-related complications and occurrence of shunt dependency have been a constant problem with these patients. It is presumed that the increased protein content in the CSF and the cellular debris results in a higher incidence of shunt malfunction [1, 15, 27, 46]. Further, in approximately 10% of the patients with pinealoblastomas and malignant germ cell tumors, there is a risk of abdominal and pelvic metastasis [5, 25]. Although, millipore filters have been advocated for prevention of the metastasis, a higher risk of shunt malfunction has been associated with these [2]. In addition, there is an increased risk of infection associated with foreign body placement in this subset of patients who become immunocompromised by the subsequent adjuvant chemotherapeutic regimes as part of the treatment protocol. ETV has been considered as an effective alternative to placement of ventriculoperitoneal shunt in obstructive hydrocephalus [27, 29, 36, 40, 46, 53]. Essentially, a third ventriculostomy opens up the obstructed CSF pathways proximal to the obstruction at the aqueductal inlet caused by the tumor. The fenestration site in the tuber cinerium is usually thinned out in such cases of subacute onset of

1142

hydrocephalus, although not to an extent seen in chronic obstruction. However, the position of basilar artery is usually well visualized, thus permitting the surgeon to place the fenestration anterior to the bifurcation site. In the present series, the authors did not encounter any significant difficulty in performing the procedure. The prepontine cistern was well visualized after the fenestration in all the patients to confirm the adequacy of the procedure. In tumor-associated hydrocephalus, the shortterm and long-term success rates of ETV have been as high as 95 and 83%, respectively [29]. Specifically, in pineal tumors, the success rate of third ventriculostomy has been reported to be between 91 and 94% [40, 53]. In a recent report compiling the reported cases of ETV in pineal region tumors, a 15% (8 of 54) failure requiring a ventriculoperitoneal shunt placement was noted [58]. A failed ETV can manifest in the early postoperative period or at a later date [33]. The early failure is considered due to non-absorption of the CSF at the arachnoid granulation. In addition, non-perforation of the Liliequist’s membrane (the ‘second membrane’) has been also reported to cause an early failure [6]. In the present series, in 2 of the 24 patients (8%), the ETV failed in the early postoperative period. One of these patients had a pinealoblastoma, and the other harbored a tuberculoma in the pineal region. In all the patients at surgery, adequate effort was made during the procedure to visualize the basal cisterns and the basilar artery, thus excluding non-fenestration of the Liliequist’s membrane as a cause for the failure. Hence, in both these patients, non-absorption of the CSF was considered to be the cause of failure of third ventriculostomy. Although technically an easy procedure, ETV has been associated with several unique complications. The potential complications include injury to the basilar artery, transient and permanent deficits resulting from injuries to the fornix, disturbances of electrolytes, and hemiparesis. In the present series, although we had evidence of mild forniceal ependymal contusion in six patients, we did not encounter any significant complications in the postoperative period as a result of the procedure. A small percentage of patients with pineal tumors do not have significant hydrocephalus at the time of presentation. In the absence of hydrocephalus, an endoscopic approach to these lesions would be associated with significant risks and may not be advisable.

Role of endoscopic biopsy During the endoscopic procedure, by inclining the endoscope posteriorly to the region of aqueduct, the tumor is usually encountered, and a biopsy can be taken under direct vision [15, 27, 29, 36, 40]. In view of the tumor

Childs Nerv Syst (2007) 23:1135–1145

heterogeneity, it is preferable to do biopsy from multiple sites [20, 21, 26, 31]. The preferred area of biopsy has been considered to be in the midline in upper part of the tumor with an aim to avoid the lower and lateral part of the tumor. This diminishes the risk of worsening of existing hemiparesis in large infiltrative pineal tumors, as the pyramidal tract while descending to cerebral peduncles courses close to the third ventricle [15]. The minor bleeding from the tumor surface usually subsides with irrigation. McArthur et al. [29] reported a positive yield in 61% using a flexible endoscope. Subsequently, a 94% positivity rate was reported by Pople et al. [40]. A few other reports have confirmed the efficacy of endoscopic biopsy in posterior third ventricular tumors [15, 36, 53, 58]. In the present study, a positive biopsy was obtained in all (100%) the patients in whom the biopsy was attempted. In one patient, the biopsy was not attempted, as forniceal injury was anticipated by the operating surgeon while angling the scope posteriorly in the third ventricle. We believe our protocol of obtaining multiple biopsies from several sites contributed to the high positive yield witnessed in the present series. In addition, conservation of all the specimens for the final histopathological evaluation was also of considerable benefit for the neuropathologist in arriving at the diagnosis. The major complication encountered during biopsy is bleeding, which can be particularly troublesome in view of the ineffective tamponade effect within the ventricles [15]. In addition, hypervascularity and abrupt decrease in intracranial pressure during the endoscopic procedure have been described as the other possible causes [57]. A recent report describes the occurrence of intra-tumoral hemorrhage in a pineal germinoma after neuroendoscopic tumor biopsy [57]. In the present series, one patient of pinealocytoma (case 9) had a significant bleeding after biopsy requiring an external ventricular drainage. She required craniotomy and microsurgical decompression due to the associated tumor bed hematoma. Although the tumor could be significantly decompressed, she developed further complications and died of her illness later. Two prior studies have found a higher incidence of spinal dissemination among patients undergoing biopsy for germinoma in comparison to un-biopsied patients [24, 56]. However, several subsequent studies have not substantiated this [31, 34]. In the present study, one patient of pinealoblastoma (case 4) developed spinal metastasis 3 months after ETV and biopsy. This patient had tumor dissemination along the floor of the third ventricle as evident during the endoscopic procedure, which however, was not apparent in the preoperative imaging studies. Interestingly, in five patients (four pinealoblastoma, one germinoma), there was evidence of tumor seedling in the third ventricular floor evident at the time of the neuroendoscopic procedure,

Childs Nerv Syst (2007) 23:1135–1145

which was not evident in the preoperative neuroimaging studies. A similar finding has been recently reported by Reddy et al. [42]. There has been considerable debate regarding prioritization of the procedures during an ETV and tumor biopsy [9, 15, 16, 29, 35, 40]. Oka et al. [37] advocate tumor biopsy before third ventriculostomy to minimize the potential dissemination of tumor cells into the basal cisterns during irrigation after biopsy and also to prevent the collapse of third ventricle while taking biopsy. Additionally, some authors have raised concerns of bleeding during the tumor biopsy, blocking the ventriculostomy if the ETV is performed earlier than the biopsy [59]. However, several others disagree with this issue and prefer ventriculostomy before biopsy, as even a mild hemorrhage during biopsy can prevent adequate visualization of third ventricular floor, rendering the subsequent ventriculostomy unsafe [6, 15, 16, 29, 40]. In the present study, the authors feel that as the primary objective is to relieve the intracranial hypertension and to avoid a shunt placement, it is always preferable to do third ventriculostomy initially followed by biopsy, thus eliminating the concerns of nonperformance of an ETV in a vascular tumor that bleeds during the procedure. In the past, both rigid and flexible endoscopes have been used for the biopsy from the posterior third ventricular region [7, 9, 15, 29]. The advantage of rigid endoscope is its superior optics with excellent illumination and larger working channels enabling larger specimens to be obtained at surgery. In addition, better orientation within the third ventricle is of considerable help to the surgeon during the procedure [7, 15]. However, the requirement of precise positioning of the burr hole while using the rigid endoscope sometimes is disadvantageous to the inexperienced surgeon. This is extremely important, as the rigid endoscope has to be maneuvered through the foramen of Monro to the pre-mamillary region for the ETV and subsequently posteriorly to visualize the tumor. This maneuver can injure the fornix if adequate care is not exercised during the procedure. To overcome this, flexible scopes have been used, with considerable decreased risk of forniceal injury. This facilitates easy navigation into third ventricle through a small foramen thus reducing the precision with which the burr hole has to be placed. However, the suboptimal illumination and the risk of impaired orientation in the III ventricle is certainly disadvantageous with a flexible scope [15, 29]. To obviate this, placement of two burr holes with different trajectories have been advocated: one each for the ETV and the tumor biopsy [59]. We have placed the burr hole more anterior than the conventional precoronal burr hole to have a slightly posterior trajectory, which is in between the third ventriculostomy and the anterior edge of the tumor. In patients with preoperative MRI, this was achieved by identifying the midpoint of the anterior margin

1143

of the tumor and the site of the third ventriculostomy and extrapolating the line along the posterior margin of the foramen of Monro to the calvarium. In a recent communication, Chernov et al. [9] reported the usage of bilateral prefrontal burr holes for approaching the tumor and performing the third ventriculostomy. These authors used the contralateral access for continuous irrigation during the procedure while using the flexible endoscope to access the posterior third ventricular region for the biopsy and third ventriculostomy. As the size of the foramen of Monro is an important consideration, we preoperatively assess adequacy of the foramen of Monro visually by carefully reviewing the available neuroimaging studies. With all the patients in the present series, we initially proceeded with a Gaab’s rigid endoscope with a working port. While this neuroendoscope was adequate for most of the patients, in four patients with a mildly dilated foramen of Monro, the procedure was carried out with a thinner diameter rigid Gaab’s fiberscope to avoid injuring the fornix and the walls of the foramen of Monro. Comparing the endoscopic and the stereotactic techniques, endoscopic techniques have been considered to be superior to stereotactic techniques [17, 59]. This is because the biopsy is performed under direct visualization avoiding vascular structures, and multiple biopsies can be obtained from different sites during the procedure thus reducing the tumor sampling error. Also, the CSF obtained during the procedure can be used for evaluation of tumor markers. In addition, a CSF diversion can be performed through the same route thus avoiding a further procedure. In the present series, definitive surgery was carried out in six patients. All these patients had evidence of either a pinealocytoma or low- or intermediate-grade gliomas. In all these patients, the results of the tissue obtained at initial biopsy matched reasonably with the final histopathology obtained at the definitive microsurgery (Table 2). The present study included four patients of tuberculomas in the posterior third ventricular region. Tuberculomas are inflammatory granulomas and respond well to antitubercular chemotherapy [13, 41, 52]. However, a biopsy is considered essential before starting a prolonged chemotherapeutic regimen, as the neuroradiological studies have a relatively low predictive value [32, 49]. These lesions are usually avascular and, hence, are associated with a minimal risk of bleeding during the biopsy. In all the four patients, the lesions showed a considerable decrease in size during follow-up reflecting the efficacy of antitubercular therapy in these lesions. In the present study, we have not attempted to address the efficacy of the specific modes of treatment for these patients with the various subgroup of tumors. We are aware of the controversies in management for the specific tumors

1144

in this location, which have been further accentuated with the advancement of radiation therapy and availability of radiosurgery. We are carefully following these patients and would address this issue in a subsequent report. From the present series, it is apparent that an endoscopic approach to the pineal region tumors achieves the CSF diversion by a third ventriculostomy and establishes a histological diagnosis by obtaining a biopsy during the same procedure under direct vision. The subsequent therapy is based on the histopathology. The high positive yield of biopsy (100%) in the present series confirms the superiority of endoscopic procedure over the other procedures. The procedure is safe when performed in expert hands and is not associated with any significant morbidity.

Conclusions Pineal region tumors, which for long were considered a formidable challenge to neurosurgeons, are now being increasingly managed by minimally invasive procedures with the aid of endoscopy. This appears to be a promising tool to achieve all the objectives of managing such lesions. With increasing knowledge about the biology of tumors harboring in pineal region, most of the malignant tumors in the future could possibly be managed with multimodality approach using radiotherapy and chemotherapy, restricting microsurgical resection to a small subset of benign lesions.

References 1. Barlas O, Bayindir C, Imer M, Ayan I, Darendeliler E (2000) Non-resectional management of pinealoblastoma. Minim Invasive Neurosurg 43:163–170 2. Berger M, Baumeister B, Geyer JR (1991) The risks of metastases from shunting in children with primary central nervous system tumors. J Neurosurg 74:872–877 3. Bruce JN (1993) Management of pineal region tumors. Neurosurg Q 3:103–119 4. Bruce JN, Ogdon AT (2004) Surgical strategies for treating patients with pineal region tumors. J Neuro Oncol 69:221–236 5. Bruce JN, Stein BM (1995) Surgical management of pineal region tumors. Acta Neurochir (Wien) 134:130–135 6. Buxton N, Vloeberghs M, Punt J (1998) Liliequist’s membrane in minimally invasive endoscopic neurosurgery. Clin Anat 11:187–190 7. Caemaert J, Abdullah J, Calloauw L (1994) A multipurpose cerebral endoscope and reflections on technique and instrumentation in endoscopic neurosurgery. Acta Neurochir (Wien) 61 (Suppl):49–53 8. Calaminus G, Bamberg M, Baranzelli MC (1994) International germ cell tumors: a comprehensive update of the European data. Neuropediatrics 25:26–32 9. Chernov MF, Kamikawa S, Yamane F, Ishihara S, Kubo O, Hori T (2006) Neurofiberscopic biopsy of tumors of the pineal region and

Childs Nerv Syst (2007) 23:1135–1145

10.

11. 12. 13. 14. 15. 16.

17.

18. 19. 20. 21.

22.

23. 24.

25. 26.

27. 28. 29.

30.

posterior third ventricle: indications, technique, complications and results. Neurosurgery 59:267–277 d’Andrea AD, Packer RJ, Rorke LB, Bilaniuk LT, Sutton LN, Bruce DA, Schut L (1987) Pineocytomas of childhood. A reappraisal of natural history and response to therapy. Cancer 59:1353–1357 Decampo MP, Davis MD (1991) The role of CT in diagnosis and management of childhood pineal region tumors. Australas Radiol 35:336–339 Diez B, Balamaceda C, Matsutami M, Weiner HL (1999) Germ cell tumors of the CNS in children: recent advances in therapy. Childs Nerv Syst 15:578–585 Domingo Z, Peter JC (1989) Intracranial tuberculomas. An assessment of a therapeutic drug trial in 35 children. Pediatr Neurosci 15:161–166 Edwards MS, Davis RL, Laurent JP (1995) Tumor markers and cytologic features of cerebrospinal fluid. Cancer 56:1773–1777 Ferrer E, Santamarta D, Garcia-Fructuoso G, Caral L, Rumia J (1997) Neuroendoscopic management of pineal region tumours. Acta Neurochir (Wien) 139:12–21 Gaab M (2006) Comments to the article by Chernov et al. Neurofiberscopic biopsy of the tumors of the pineal region and posterior third ventricle: indications, technique, complications and results. Neurosurgery 59:277 Gangemi M, Maiuri F, Colella G, Buonamassa S (2001) Endoscopic surgery for pineal region tumors. Minim Invasive Neurosurg 44:70–73 Hasegawa T, Kondziolka D, Hadjipanayis CG, Flickinger JC, Lunsford LD (2003) Stereotactic radiosurgery. Pediatr Neurosurg 38:329–333 Herrmann HD, Winkler D, Westphae M (1992) Treatment of tumors of the pineal region and posterior part of III ventricle. Acta Neurochir (Wien) 116:137–146 Hoffman HJ, Yoshida M, Becker LE, Hendrick EB, Humphrey RP (1994) Pineal region tumors in childhood. Experience at the hospital for sick children. Pediatr Neurosurg 21:91–104 Hyunnam D, Chu BK, Shin HJ, Ahn HS, Kim JJ, Wang KC (1999) Treatment of intracranial nongerminomatous malignant germ cell tumors in children: the role of each treatment modality. Childs Nerv Syst 15:185–191 Itoyama Y, Kochi M, Kuratsuy JI, Takamura S, Kitano I, Marubayashi T, Uemura S, Ushio Y (1995) Treatment of intracranial nongerminomatous malignant germ cell tumors producing α-fetoprotein. Neurosurgery 36:459–466 Jakachi R (1999) Pineal and nonpineal supratentorial primitive neuro ectodermal tumors. Childs Nerv Syst 15:586–591 Jenkin D, Berry M, Chan H (1990) Pineal region germinomas in childhood treatment considerations. Int J Radiat Oncol Biol Phys 18:541–545 Jennings MT, Gelman R, Hochberg F (1985) Intracranial germ cell tumors: natural history and pathogenesis. J Neurosurg 63:155–167 Knierim DS, Yamada S (2003) Pineal tumors and associated lesions: the effect of ethnicity on tumor type and treatment. Pediatr Neurosurg 38:307–323 Konovalov AN, Pitskhelavri DI (2003) Principles of treatment of pineal region tumors. Surg Neurol 59:250–268 Kreth FW, Schatz CR, Pagenstecher A, Faist M, Volk B, Ostergtag CB (1996) Stereotactic management of lesions of the pineal region. Neurosurgery 39:280–291 Macarthur DC, Buxton N, Punt J, Vloeberghs M, Robertson IJA (2002) The role of neuroendoscopy in the management of brain tumors. Br J Neurosurg 16:465–470 Macarthur DC, Buxton N, Vlorberghs M, Punt J (2001) The effectiveness of neuroendoscopic interventions in children with brain tumors. Childs Nerv Syst 17:589–594

Childs Nerv Syst (2007) 23:1135–1145 31. Matsutani M, Sano K, Takakura K, Fujimaki T, Nakamura O, Funata N, Seto T (1997) Primary intracranial germ cell tumors: a clinical analysis of 153 histologically verified cases. J Neurosurg 86:446–455 32. Mohanty A, Santhosh V, Anandh B, Kolluri VR, Vasudev MK, Hegde T, Shankar SK (1999) Diagnostic efficacy of stereotactic biopsies in intracranial tuberculomas. Surg Neurol 52:252–257 33. Mohanty A, Vasudev MK, Sampath S, Radhesh S, Sastry Kolluri VR (2002) Failed endoscopic third ventriculostomy in children: management options. Pediatr Neurosurg 37:304–309 34. Neuwelt EA, Glasberg M, Frenkel E, Clark WK (1979) Malignant pineal region tumors: a clinical pathological study. J Neurosurg 51:597–607 35. Nicholson JC, Punt J, Hale J, Saran F, Calaminus G (2002) Neurosurgical management of paediatric germ cell tumours of the central nervous system—a multi disciplinary team approach for the new millennium. Br J Neurosurg 16:93–95 36. Oi S, Shibata M, Tominaga J, Honda Y, Shinoda M, Takei F, Tsugane R, Matsuzawa K, Sato O (2000) Efficacy of neuroendoscopic procedures in minimally invasive preferential management of pineal region tumors: a prospective study. J Neurosurg 93:245–253 37. Oka K, Kin Y, Go Y (1999) Neuroendoscopic approach to tectal tumors: a consecutive series. J Neurosurg 91:964–970 38. Packer RJ, Sutton LN, Rosenstock JG (1984) Pineal region tumors of childhood. Pediatrics 74:97–102 39. Patil AA, Good R, Bashir R, Etemadrezaie H (1995) Nonresective management of pineoblastoma: a case report. Surg Neurol 44:386–391 40. Pople IK, Athanasiou C, Sandeman R, Coakham HB (2001) The role of endoscopic biopsy and third ventriculostomy in the management of pineal region tumors. Br J Neurosurg 15:305–311 41. Rajshekhar V, Chandy MJ (1997) Tuberculomas presenting as isolated intrinsic brain stem masses. Br J Neurosurg 11:127–133 42. Reddy AT, Wellons JC 3rd, Allen JC, Fiveash JB, Abdullatif H, Braune KW, Grabb PA (2004) Refining the staging evaluation of pineal region germinoma using neuroendoscopy and the presence of preoperative diabetes insipidus. Neuro Oncol 6:127–133 43. Regis J, Bouillot P, Volat FR, Branger DF, Dufoer H, Peragut JC (1996) Pineal region tumors and the role of stereotactic biopsy: review of the mortality, morbidity and diagnostic rates in 370 cases. Neurosurgery 39:907–914 44. Reiger A, Rainov NG, Brucke M, Marx T, Holz C (2000) Endoscopic third ventriculostomy is the treatment of choice for

1145

45. 46. 47.

48. 49.

50. 51. 52. 53. 54. 55. 56. 57. 58.

59.

obstructive hydrocephalus due to pediatric pineal tumors. Minim Invasive Neurosurg 43:83–86 Richard MW, Laws ER (1985) Shunting and irradiation of pineal tumors. Clin Neurosurg 32:384–396 Robinson S, Cohen AR (1997) The role of neuroendoscopy in the treatment of pineal region tumors. Surg Neurol 48:360–367 Sawamura Y, De Tribolet N, Ishii N, Abe H (1997) Management of primary intracranial germinomas: diagnostic biopsy or radical resection? J Neurosurg 87:262–266 Sawamura Y, DeTribolet N (2002) Neurosurgical management of pineal tumors. Adv Tech Stand Neurosurg 27:217–243 Selvapandian S, Rajashekar V, Chandy MJ, Idikula J (1994) Predictive value of computer tomography based diagnosis of intracranial tuberculomas. Neurosurgery 35:845–850 Smirniotopoulos JG, Rushing EJ, Mena H (1992) Pineal region masses, differential diagnosis. Radiographics 12:577–596 Steinbok P, Cochrane DD (2001) Pineal region and intracranial germ cell tumors. In: Mclone DG (ed) Pediatric neurosurgery: surgery of the developing nervous system. W.B. Saunders, Philadelphia, pp 711–724 Tandon PN, Bhargava S (1985) Effect of medical treatment on intracranial tuberculoma—a CT study. Tubercle 66:85–97 Teo C, Young R II (1999) Endoscopic management of hydrocephalus secondary to tumors of the posterior third ventricle. Neurosurg focus DOI 7: Article (2) Vaquero J, Ramiro J, Martinez R, Bravo G (1992) Neurosurgical experience with tumors of pineal region at Clinica Puerta de Hierro. Acta Neurochir (Wien) 116:23–32 Vaquero J, Ramiro J, Martinez R, Coca S, Bravo G (1990) Clinicopathological experience with pineocytomas: report of five surgically treatment cases. Neurosurgery 27:612–619 Wara WM, Jenkin KDT, Evans A (1979) Tumors of the pineal and suprasellar region. Children’s cancer study group treatment results, 1960–1975. Cancer 43:698–701 Wong TT, Yen SH, Ho DM, Chang FC, Chang KP (2003) Pineal germinoma with intratumoral hemorrhage after neuroendoscopic tumor biopsy. Childs Nerv Syst 19:769–772 Yamini B, Refai D, Rubin CM, Frim DM (2004) Initial endoscopic management of pineal region tumors and associated hydrocephalus: clinical series and literature review. J Neurosurg 100(5 Suppl Pediatrics):437–441 Yurtseven T, Ersahin Y, Demirta E, Mutluer S (2003) Neuroendoscopic biopsy for intraventricular tumors. Minim Invasive Neurosurg 46:293–299