J Neurosurg (Pediatrics 5) 100:437441, 2004

Initial endoscopic management of pineal region tumors and associated hydrocephalus: clinical series and literature review
BAKHTIAR YAMINI, M.D., DANIEL REFAI, M.D., CHARLES M. RUBIN, M.D., AND DAVID M. FRIM, M.D., PH.D.
Sections of Pediatric Neurosurgery and Pediatric HematologyOncology, University of Chicago Childrens Hospital, University of Chicago, Illinois
Object. The authors report their experience in six patients with pineal tumors and associated hydrocephalus who underwent an endoscopic biopsy procedure and third ventriculocisternostomy (ETVC) in a single sitting. Methods. The ETVC was successfully performed without complication in all patients; however, a ventriculoperitoneal shunt was eventually required in four. Histological diagnosis was successfully established in four patients. The authors also reviewed the literature to assess reports involving ETVC and tumor biopsy sampling in patients with pineal tumors and hydrocephalus. A total of 54 cases, including those in this study, have been reported. Fifteen percent of the patients eventually required placement of a ventricular shunt. The transient complication rate was 15% with no death. A positive tissue diagnosis was established in 89% of the cases overall. Conclusions. The authors conclude that the endoscopic management of patients with pineal region masses and hydrocephalus may be a preferred initial strategy.

KEY WORDS pineal region tumor endoscopy hydrocephalus

third ventriculocisternostomy

anagement of pineal region tumors has changed considerably during the past quarter century. Conservative treatment involving shunt therapy and irradiation in the early 1970s evolved into a more aggressive approach with the advent of microneurosurgery.2,3 More recently, stereotactic and endoscopic technology has added a new dimension to the management of these tumors.1,20 The broad range of treatment modalities for these lesions is due to the diversity of tumor types found in the pineal region as well as the unique location of this structure (Fig. 1).21 The most common presenting signs and symptoms are related to raised ICP with hydrocephalus being present in approximately 90% of cases.6,21 Optimal treatment for a pineal region tumor may be aggressive resection, radiotherapy, chemotherapy, or a combination based on tumor histology.6,15,17 Obtaining an adequate tissue sample for histological diagnosis is an important aspect of treatment planning. Open craniotomy and microsurgical biopsy sampling can achieve this; however, the risk of permanent associated morbidity approaches 10%.6 Stereotactic biopsy sampling is a less invasive technique, but sampling error, due to the hetero-

geneous nature of tumors in the region, has been noted to be a significant problem. Additionally, this procedure can carry significant morbidity because of the numerous vascular structures in the vicinity of the pineal gland. Neither of these procedures, however, specifically addresses the hydrocephalus associated with masses in this area. For these reasons, endoscopy is emerging as the preferred initial management of these lesions.20 Patients presenting with hydrocephalus and a posterior third ventricular mass can undergo ETVC and an endoscopic tumor biopsy procedure in one sitting. Several case reports and short series involving this combined approach to pineal region tumors have been reported.4,5,7,8,11,13,14,1820 We present our experience in treating eight patients presenting with hydrocephalus and a pineal mass. We specifically discuss the use of endoscopy as an initial treatment modality and elaborate on several aspects evident in our study and in the literature. Clinical Material and Methods Patient Population Between 1997 and 2001, 11 patients with pineal region tumors and associated hydrocephalus presented for surgical treatment at the University of Chicago Childrens Hospital. In one patient a VP shunt had been placed at an outside facility and two patients had previously undergone
437

Abbreviations used in this paper: CSF = cerebrospinal fluid; ETVC = endoscopic third ventriculocisternostomy; ICP = intracranial pressure; MR = magnetic resonance; VP = ventriculoperitoneal.

J. Neurosurg: Pediatrics / Volume 100 / May, 2004

B. Yamini, et al.
gonadotropin were obtained at the time of initial intraoperative ventricle puncture. Only one patient required emergency ventriculostomy prior to the endoscopic procedure. Spinal MR imaging was performed between 1 and 10 days postoperatively in all patients except two in whom low-grade astrocytoma was diagnosed. The followup period ranged from 1 to 68 months (mean 25.5 months) (Table 1). Two patients were lost to follow up after 1 month, and one patient died at 3 months.
Endoscopic Technique

FIG. 1. Sagittal T1-weighted MR image obtained in a patient with a pineal tumor and hydrocephalus.

craniotomy before treatment of the hydrocephalus. Eight of the patients with hydrocephalus were initially treated endoscopically, in six of whom a biopsy sample was obtained in the same sitting. There were five male and three female patients who ranged in age between 1 and 20 years (mean 8.75 years). In two patients biopsy specimens were not acquired (Table 1). In one patient (Case 8) a diagnosis of low-grade astrocytoma was made after endoscopic biopsy sampling several years previously at an outside institution, and in another patient (Case 4) the biopsy procedure was aborted because of excessive bleeding when the tumor surface was coagulated.
Preoperative Data

The presenting signs and symptoms are summarized in Table 1. Each patient underwent preoperative computerized tomography and MR imaging with and without infusion of contrast material (Fig. 1). The decision to proceed with the endoscopic biopsy procedure as initial treatment was based solely on the presence of a posterior third ventricular mass with dilated ventricles. Cerebrospinal fluid samples for -fetoprotein and human chorionic

A right-sided precoronal incision and burr hole are initially made at the midpupillary line, and an introducer is used to puncture the lateral ventricle. A 4.5- or 4.6-mm flexible endoscope or a 7-mm rigid endoscope is inserted through the introducer into the lateral ventricle. The ventricular landmarks including the choroid plexus, thalamostriate and septal veins, and foramen of Monro, are identified. The floor of the third ventricle can often be visualized through the dilated foramen of Monro, and the endoscope is advanced into the third ventricle. The ventriculocisternostomy is made using a blunt probe in the usual location anterior to the mammillary bodies and is then widened using a No. 3 French Fogarty balloon catheter. Electrocautery is not used during this part of the procedure. The scope is then directed toward the posterior aspect of the third ventricle, and the tumor is visualized (Fig. 2). The standard landmarks in this region are identified. Using biting forceps, two or more biopsy specimens are then obtained from different regions of the mass. Bleeding is controlled using copious irrigation or monopolar cautery. When hemostasis is achieved, the scope and sheath are removed. In all but three cases, prior to closure of the wound, either an external ventricular drain or a ventricular catheter connected to a subcutaneous reservoir was left in place. Results Of the eight patients treated endoscopically, six underwent ETVC and biopsy sampling and two underwent ETVC alone. Histological diagnosis was successfully es-

TABLE 1 Summary of treatment in eight patients with pineal region masses associated with hydrocephalus*
Case No. Age (yrs) Presenting Findings Final Pathological Diagnosis VP Shunt (postop day) Last Follow Up (mos) Neurological Deficit

Procedure

Biopsy Results

Condition

1 2 3 4 5 6 7 8

14 3 20 3 8 1 5 16

obtunded headache; somnolence headache somnolence, EOM palsy headache somnolence headache headache

pineoblastoma pineoblastoma pineoblastoma embryonal/ rhabdomyosarcoma teratoma astrocytoma pineoblastoma astrocytoma

ETVC/biopsy ETVC/biopsy ETVC/biopsy ETVC ETVC/biopsy ETVC/biopsy ETVC/biopsy ETVC

actual dx actual dx astrocytoma aborted normal ependymal actual dx actual dx none

no no yes, 16 no yes, 16 yes, 11 yes, 19 no

39 1 51 1 3 21 68 20

stable stable stable stable died stable stable stable

none none none none none none none none

* Dx = diagnosis; EOM = extraocular muscle.

438

J. Neurosurg: Pediatrics / Volume 100 / May, 2004

Endoscopy in pineal region tumors

literature in showing the efficacy of initial endoscopic management in cases in which pineal region masses are associated with hydrocephalus.19 In addition to sampling CSF, the goal in the endoscopic treatment of pineal tumors is twofold. The first and primary objective is the diversion of CSF. As has been demonstrated in a large series of patients with pineal tumors, approximately 90% present with hydrocephalus and require placement of a shunt for CSF diversion.20 As documented, however, the incidence of shunt malfunction in this population is as high as 20%.20 Additionally, unless complete resection is achieved, it is unusual to reverse the hydrocephalus microsurgically. Third ventriculocisternostomy in patients with aqueductal stenosis may avoid the need for a shunt, and in general, the results are very good, with approximately a 60 to 80% long-term success rate overall (that is, shunt independence). In 125 patients with hydrocephalus followed for a mean of 28 months, a subgroup of 33 with triventricular hydrocephalus due to tumor-induced aqueductal compression was associated with a patency rate of almost 85% after third ventriculostomy.12 There were nine patients with pineal tumors; however, in this and other reports involving large series of ETVC-treated patients, biopsy samples were not obtained at the same time. Nevertheless, in 10 smaller series of patients with pineal tumors and hydrocephalus who underwent endoscopic biopsy sampling and ETVC, only four of 48 patients subsequently required insertion of a VP shunt (Table 2). In our series, four of the six patients in whom a biopsy specimen was obtained subsequently required early shunt therapy. The cause of this high failure rate in our series is unknown but could be due to the occlusion of the ventriculostomy by intraventricular debris formed when these tumors were sampled. The two patients who did not undergo a biopsy procedure did not require placement of a shunt. Additionally, two of our shunt-treated patients required placement of the shunt after having undergone craniotomy. This most likely converted the initial obstructive hydrocephalus to an absorptive hydrocephalus due to the release of protein and blood into the CSF. In addition, in one of our shunt-treated patients, craniospinal seeding of tumor along the meninges was demonstrated on MR imaging. Finally, although our ETVC technique is very similar to others, it is generally our policy to monitor ICP with an external ventricular drain or a ventricular catheter and a noninvasive telemonitor. This is very sensitive to changes in ICP and may have encouraged us into a more aggressive approach with respect to subsequent shunt therapy. The second goal in the endoscopic management of these patients is tissue procurement for diagnosis. In the series reported by Fukushima,9 he obtained biopsy specimens in 21 patients, four of whom harbored pineal region masses. A correct diagnosis, however, was made in only one of the four patients. Nevertheless, in more recent studies, tumor diagnosis has been successful in almost all cases (Table 2).4,11,13,18,19 We found that the histological diagnosis was correct in four of six patients. In one of these patients, examination of the endoscopically obtained tissue showed a low-grade astrocytoma, but the final pathological diagnosis, determined after open craniotomy, was pineoblastoma. This
439

FIG. 2. Diagram of the ETVC and biopsy trajectory.

tablished in four of the six patients (Table 1). In Case 3 the endoscopic and final pathological diagnoses differed. In Case 5, evaluation of the biopsy sample was not diagnostic. As noted previously, one attempted biopsy procedure was aborted because excessive bleeding obscured the view after monopolar cautery had been used to coagulate the surface of the tumor. Although ETVC was successful in all patients, a VP shunt was placed in four patients between 11 and 19 days (mean 15.5 days) after the procedure. The two patients lost to follow up were seen for 1 month postoperatively and were discharged without hydrocephalus. Subsequent care was rendered at an outside institution. The patient in Case 8 did not undergo a biopsy procedure, and the ETVC was a repeated procedure because the initial ETVC, performed 5 years previously, had become occluded. Pathological diagnosis, subsequent treatment, and clinical outcome at last follow up are summarized in Table 1. Five patients required subsequent craniotomy for tumor resection. There were no deaths directly attributable to the endoscopic procedures and no significant new postoperative neurological deficits were seen. Discussion In 1973, Fukushima, et al.,10 provided the first modern description of the use of an endoscope in the treatment of pineal tumors. Prior to this, the endoscope was mainly used to coagulate the choroid plexus in the treatment of hydrocephalus. In the early 1970s, endoscopic biopsy sampling of pineal lesions was unpopular because surgeons feared inducing uncontrollable bleeding. Since then, adequate techniques have been developed for hemostasis, and the ventriculoscope has been used for various neurosurgical conditions including hydrocephalus, cysts, and tumors.12 Currently, in many centers with neurosurgeons who are skilled in endoscopy, patients who present with pineal tumors and hydrocephalus are initially managed endoscopically. Our findings are consistent with the
J. Neurosurg: Pediatrics / Volume 100 / May, 2004

B. Yamini, et al.
TABLE 2 Summary of data obtained in review of reported cases involving endoscopic biopsy sampling and ETVC for pineal region masses*
Authors & Year No. of Cases Biopsy Definitive for Diagnosis (%) No. of VP Shunts (%) Complications (%)

Ellenbogen & Moores, 1997 Ferrer, et al., 1997 Robinson & Cohen, 1997 de Divitiis, 1998 Gaab & Schroeder, 1998 Decq, et al., 2000 Oi, et al., 2000 Gangemi, et al., 2001 Haw & Steinbok, 2001 Pople, et al., 2001 present series total

1 4 4 4 1 4 6 5 1 18 6 54

1 (100) of 1 3 (75) of 4 3 (75) of 4 3 (75) of 4 1 (100) of 1 4 (100) of 4 6 (100) of 6 5 (100) of 5 1 (100) of 1 17 (94) of 18 4 (67) of 6 48 (89) of 54

0 1 1 0 0 0 1 0 0 1 4/6 8 (15) of 54

0 3 (hem, increased mem loss & hemiparesis temp) 0 1 (hem) 0 0 0 1 (transient dysphagia & rt hemiparesis) 0 2 (hem, temp amnesia) 0 8 (15) of 54 temp 0 of 54 permanent

* Hem = hemorrhage; mem = memory; temp = temporary.

was one of our earliest cases, in which the biopsy sample was obtained using a 1.5-mm biting forceps. The small sample size may well have contributed to the misdiagnosis. In the other patient, evaluation of the tissue was nondiagnostic; in this case, the tumor did not penetrate the ependymal wall and could not be seen during the endoscopic procedure. Although the endoscopic biopsy procedure showed only ependymal wall, subsequent open craniotomy revealed a diagnosis of teratoma. In cases such as this in which there is no clear exophytic mass, it may be best to perform the biopsy procedure by using a combination of endoscopy and frameless stereotaxy. Our results are not as encouraging as those in the literature, where in 44 of the 48 cases in which endoscopic biopsy procedures were performed a histological diagnosis was achieved (Table 2). This most likely is due to the aforementioned reasons. Although it is quite likely that the actual correct diagnosis rate is closer to 90%, the most recently published data are very encouragingonly one misdiagnosis has been reported since 2000.5,13,14,18,19 The gold standard for establishing a diagnosis is open craniotomy; however, as previously noted, the permanent morbidity rate associated with this approach can be as high as 10%.6 Additionally, some tumor types may not require microsurgical resection. Stereotactic biopsy sampling is the other minimally invasive method for establishing diagnosis. Even in experienced hands, however, the mortality rate can be as high as 2%.16 Although some authors have reported a good overall correct diagnosis rate,16 others have noted poor results when performing stereotactic biopsy procedures with significant sampling error.6 The ability to obtain multiple biopsy specimens from different regions of the tumor gives the endoscopic technique a theoretical advantage over the stereotactic technique. As can be seen from this review, however, the problem of misdiagnosis also occurs with endoscopic biopsy sampling. Hopefully, this will be minimized as endoscopic techniques and tools are refined. This problem may be an unavoidable consequence of the nature of the tumor types found in the region of the pineal gland. Of
440

particular note, although germinomas are one of the most common tumor types found in this region,21 none was demonstrated in our series. This most likely is due to the small sample size and does not reflect a special referral pattern. There were no significant deficits associated with our endoscopic biopsy procedures. One intraoperative hemorrhage occurred during monopolar coagulation of the tumor surface prior to the biopsy procedure. Since then, we have refrained from using cautery and instead use copious warm irrigation for hemostasis. In reviewing the literature, there were no deaths attributable to the endoscopic procedures. There was a 15% transient morbidity rate involving either hemorrhage or other transient neurological deficit (Table 2).4,5,7,8,11,13,14,1820,22 There is no report of any permanent morbidity in the literature. In conclusion, endoscopic biopsy sampling with ETVC has been reported in 54 patients to date and, combined, appears to be safe. This approach has now evolved to the point where it should be considered in the initial management of patients presenting with pineal region masses and associated hydrocephalus.
References 1. Barat JL, Benabid A, Blond S, et al: [Stereotaxic biopsies of pineal tumors. Comments on their risk and implication apropos of 370 cases.] Neurochirurgie 40:39, 1994 (Fr) 2. Bruce JN, Stein BM: Surgical management of pineal region tumors. Acta Neurochir 134:130135, 1995 3. Chapman PH, Linggood RM: The management of pineal area tumors: a recent reappraisal. Cancer 46:12531257, 1980 4. de Divitiis O: Provision of a neuroendoscopy service. The Southampton experience. J Neurosurg Sci 42:137143, 1998 5. Decq P, Le Guerinel C, Sakka L, et al: [Endoscopic surgery of third ventricle lesions.] Neurochirurgie 46:286294, 2000 (Fr) 6. Edwards MS, Hudgins RJ, Wilson CB, et al: Pineal region tumors in children. J Neurosurg 68:689697, 1988 7. Ellenbogen RG, Moores LE: Endoscopic management of a pineal and suprasellar germinoma with associated hydrocephalus: technical case report. Minim Invasive Neurosurg 40: 1316, 1997

J. Neurosurg: Pediatrics / Volume 100 / May, 2004

Endoscopy in pineal region tumors

8. Ferrer E, Santamarta D, Garcia-Fructuoso G, et al: Neuroendoscopic management of pineal region tumours. Acta Neurochir 139:1221, 1997 9. Fukushima T: Endoscopic biopsy of intraventricular tumors with the use of a ventriculofiberscope. Neurosurgery 2: 110113, 1978 10. Fukushima T, Ishijima B, Hirakawa K, et al: Ventriculofiberscope: a new technique for endoscopic diagnosis and operation. Technical note. J Neurosurg 38:251256, 1973 11. Gaab MR, Schroeder HW: Neuroendoscopic approach to intraventricular lesions. J Neurosurg 88:496505, 1998 12. Gangemi M, Donati P, Maiuri F, et al: Endoscopic third ventriculostomy for hydrocephalus. Minim Invasive Neurosurg 42:128132, 1999 13. Gangemi M, Maiuri F, Colella G, et al: Endoscopic surgery for pineal region tumors. Minim Invasive Neurosurg 44:7073, 2001 14. Haw C, Steinbok P: Ventriculoscope tract recurrence after endoscopic biopsy of pineal germinoma. Pediatr Neurosurg 34:215217, 2001 15. Jooma R, Kendall BE: Diagnosis and management of pineal tumors. J Neurosurg 58:654665, 1983 16. Kreth FW, Schatz CR, Pagenstecher A, et al: Stereotactic management of lesions of the pineal region. Neurosurgery 39: 280291, 1996 17. Oi S, Matsumoto S: Controversy pertaining to therapeutic modalities for tumors of the pineal region: a worldwide survey of different patient populations. Childs Nerv Syst 8:332336, 1992 18. Oi S, Shibata M, Tominaga J, et al: Efficacy of neuroendoscopic procedures in minimally invasive preferential management of pineal region tumors: a prospective study. J Neurosurg 93: 245253, 2000 19. Pople IK, Athanasiou TC, Sandeman DR, et al: The role of endoscopic biopsy and third ventriculostomy in the management of pineal region tumours. Br J Neurosurg 15:305311, 2001 20. Robinson S, Cohen AR: The role of neuroendoscopy in the treatment of pineal region tumors. Surg Neurol 48:360367, 1997 21. Stein BM, Bruce JN: Surgical management of pineal region tumors (honored guest lecture). Clin Neurosurg 39:509532, 1992 22. Yamini B, Goumnerova LC, Frim DM: Endoscopic approach to noncommunicating fluid spaces in the shunted patient. Pediatr Neurosurg 31:237241, 1999

Manuscript received October 16, 2003. Accepted in final form November 3, 2003. Address reprint requests to: Bakhtiar Yamini, M.D., Section of Neurosurgery, University of Chicago Childrens Hospital, MC 4066, 5841 South Maryland Avenue, Chicago, Illinois 60637. email: byamini@surgery.bsd.uchicago.edu.

J. Neurosurg: Pediatrics / Volume 100 / May, 2004

441