J Neurooncol (2009) 93:133–137 DOI 10.1007/s11060-009-9861-0
CLINICAL STUDY - PATIENT STUDY
Results of microsurgical treatment of medulla oblongata and spinal cord hemangioblastomas: a comparison of two distinct clinical patient groups Fabrice Parker Æ Nozar Aghakhani Æ Luis Gustavo Ducati Æ Adriano Yacubian-Fernandes Æ Mateus Violin Silva Æ Phillipe David Æ Stephane Richard Æ Marc Tadie
Received: 7 October 2008 / Accepted: 16 March 2009 Ó Springer Science+Business Media, LLC. 2009
Abstract Purpose To analyze the surgical outcome of a consecutive and single center series of medulla oblongata (MO) and spinal cord hemangioblastomas (HB). Patient and methods We retrospectively reviewed the medical charts of all MO and spinal HB patients operated on in our institution between 1985 and 2002. All patients had preand at least one post-operative MRI. McCormick classification was used to assess neurological status and functional outcome. Results Forty surgical procedures have been performed on 34 patients (19 females and 15 males, mean age of 41 years). Twenty-five (73%) patients had Von Hippel Lindau (VHL) disease confirmed by genetic screening, and nine patients had sporadic disease. Complete clinical, radiological, and genetic studies were done in all cases. The most frequent clinical symptom was pain (28 patient, 85%) followed by motor (42%) and sensitive deficits (42%). Fourteen lesions (19%) were located in the MO, 28 (38%) in the cervical spine, 25 (34%) in the thoracic spine, 4 (5%) in the lumbar spine and 3 (4%) in the Cauda Equina. In the VHL group, 15 patients (60%) presented multiple lesions and 10 a single neurological lesion (40%). A cyst was present in 23% of VHL patients and in 55% of the non-VHL group. A complete removal was achieved in 85% of all cases. No deaths related to surgery occurred. At the end of the follow-up period (mean
F. Parker N. Aghakhani (&) L. G. Ducati A. Yacubian-Fernandes M. V. Silva P. David M. Tadie Department of Neurosurgery, Biceˆtre University Hospital, 78 rue du general Leclerc, 94275 Le Kremlin-Biceˆtre Cedex, France e-mail:
[email protected] S. Richard Laboratoire de ge´ne´tique oncologique EPHE and CNRS FRE 2939, Bicetre University Hospital, University Paris XI, Bicetre, France
60 months) 50% of patients were stabilized, while the condition of 32.35% was improved and of 17.65% worsened. Comparing the clinical evolution considering the presence or not of VHL we have seen that there are no differences in terms of functional outcome between VHL and non-VHL groups. Conclusion Our results confirmed that surgery remains a safe and effective treatment for medulla oblongata and spinal hemangioblastoma. Only symptomatic lesion required surgical treatment. In other cases, especially in VHL patients, a close and regular follow-up (clinical and MRI) is necessary. Keywords Hemangioblastoma Medulla oblangata Spinal cord Von Hippel Lindau
Introduction Spinal cord hemangioblastomas correspond to 5% of intramedullary tumors [1–4]. They are associated with Von Hippel Lindau disease in 20–30% of the cases [5–10]. Pain syndrome, motor and sensitive deficits are the clinical symptoms usually presented. When located in the medulla oblongata, neck pain and cranial nerves palsy can be present [11–13]. Magnetic resonance imaging (MRI) suggests the diagnosis when the lesion has contrast enhancing associated or not to abnormal vessels in the subarachnoid space and intramedullary cysts [14, 15]. Complete microsurgical removal is the treatment of choice for these tumors [16–22]. The objectives of the present study were to analyze the results of the microsurgical treatment for the spinal cord and medulla oblongata hemangioblastomas and better understand the clinical and radiological presentation of these tumors. The early and late post-operative follow-ups were analyzed considering the occurrence of VHL.
123
134
J Neurooncol (2009) 93:133–137
Materials and methods
Table 2 Modification of McCormick classification in early postoperative period (48 h post-surgery)
A retrospective study based on clinical, radiological, and genetical analysis of patients with spinal and medulla oblongata hemangioblastomas was done. From 71 patients who underwent microsurgical removal of medulla oblongata or spinal cord HB in our hospital from 1985 to 2002, 34 patients who had complete (brain and spine) pre- and post-operative MRI studies and at least 3 years clinical and neuroradiological follow-up were selected for this study. We applied the McCormick classification [23] (Tables 1, 2, 3, 4) to assess the neurological function in our patients. This assessment was performed before surgery, in early post-surgical period (within 48 h after surgery), and at the last time of follow-up. The site of the lesion in neural axis and its volume (estimated by the method proposed by Wanebo et al. [10]), and the presence of cyst were analyzed. The first post-operative MRI was performed during the first three months following surgery.
Pre-operative grading
Post-operative grading I
I
Total
19
Pre-operative grading
I I
Pre-operative grading
I
13
2
2
17
II
4
2
2
8
III
6
2
8
IV
1
Total
24
1 4
6
0
Stability: 50%, improvement: 32.4%, worsening: 17.7%
123
Total
34
1 2
40
8
11
II
III
IV
Total 3
1
Results
IV
10 10 1
19
3
7
III
1
Post-operative grading
Total
II
1
Table 3 Non-VHL patients’ final outcome according to McCormick classification
The Kruskal–Wallis non-parametric test, the Fisher’s exact test and the Spearman’s correlation test were used to perform the statistical analysis.
I
Total
Stability: 95%, worsening: 5%; 40 operation performed in 34 patients
3 1
Post-operative grading
8
10
III IV
Pre-operative grading
IV
III IV
Statistical analysis
Table 1 Final outcome of all patients according to McCormick classification
III
19
II
II
Among 71 cases we selected, 34 patients were operated on from 1985 to 2002. The average age of the patients was 41 years (range 19–75 years), 19 were females (58%) and 15 were males (42%). Twenty-five (73%) patients had the diagnosis of VHL confirmed by genetic screening. In the VHL sub-group, 15 patients (63%) were women and 13 (54%) patients presented familial history of HB or other typical lesions. There was no difference in the average age comparing VHL group to non-VHL patients. All patients were symptomatic at the time of operation. The most
II
1 1
1
1
4 1 0
9
Stability: 55%, improvement: 45% Table 4 VHL patients’ final outcome according to McCormick classification Post operative grading I
II
III
I
10
2
2
II
4
1
III IV
3
Total
17
3
IV
Total 14
2
7
1 1
4 0
0
25
5
Stability: 48%, improvement: 28%, worsening: 24%
frequent clinical symptom was pain, presented in 28 patients (85%). Motor deficit (42%) and sensitive deficits (42%) were also frequent. MRI studies (Fig. 1) were able to identify 74 lesions—14 lesions (19%) located in the medulla oblongata, 28 (38%) in the cervical spine, 25 (34%) in the thoracic spine, 4 (5%) in the lumbar spine, and 3 (4%) in the cauda equina or neural root. In the non-VHL group, we always found a single lesion. In the VHL group, 15 patients (60%) presented multiple lesions and 9 patients (40%) a single lesion. A cyst was present in 23% of VHL patients and in 55% of the non-VHL group. The average volume for the operated lesions was 1.5 ml for the VHL group and 2.0 ml for the non-VHL group. The follow-up period ranged from 36 to 204 months (mean 60 months). During the follow-up period, 42% of VHL patients
J Neurooncol (2009) 93:133–137
135
Fig. 1 a Magnetic resonance images. Sagittal T1-weighted image after intravenous administration of gadolinium showing a T7 hemangioblastoma. b Preoperative angiography demonstrating hypervascular tumor. c Intra-operative view of the same tumor
presented new lesions and 54% a growing lesion. In the VHL group, 83% presented extra-spinal lesions: kidney (62%), retinal (50%), cerebellum (46%), pancreas (33%), epidydimus (4%), and gastric (4%). Thirteen patients (42%) have had diagnostic angiography before the surgery. Indication was based on: large volume of the lesion, recurrence after first surgery and the anterior located lesions (close to anterior spinal artery). Among these patients, four have been embolized. No complication related to this procedure was noted. Forty surgical procedures have been done on those 34 patients; a complete removal-based on analysis of the first post-operative MRI was achieved in 85% of cases. No infectious or general complications or deaths related to surgery were occurred. In the early post-operative period (first 48 h), we had 15% neurological deterioration (two motor deficits and four sensitive posterior column deficits). In the late post-operative period (36 months at least), sensorial deficits had disappeared but the motor deficits were still present in two (6%) VHL patients. Among the 12 patients operated for medulla oblongata lesions, none presented cranial nerve dysfunction or respiratory complication. Patients’ McCormick classification in pre-operative, early and late post-operative periods are represented in Tables 1, 2, 3, 4. There was an important rate of improvement in the long term period—48% McCormick 1 in the pre-operative period versus 61% in the late post-operative period (at least 6 months after surgery). Comparing the clinical evolution according to the presence or not of VHL, we have seen that there are no differences in terms of functional outcome between VHL and non-VHL after each surgery whenever long-term follow-up showed that VHL patients were worsened compared to the non-VHL group (P = 0.02) due to the appearance of new lesions and/or multiple surgical procedures (Table 2). Additionally, there was a statistically significant difference related to the appearance of new lesions and growing of ancient lesions in the VHL group (P = 0.03).
Discussion The spinal cord presentation of the HB corresponds to approximately 5% of all spinal cord tumors and is more
frequent in the cervical and dorsal spine [24, 25]. The average age for the diagnosis of spinal HB is in the 4th decade of life [2, 7]. They can be associated with the VHL disease in approximately 25% of cases [6–9]. The most frequent clinical presentation is pain syndrome and our data is in accordance with the data from the literature [16, 20]. We report a series of 34 surgically treated spinal HB. Surgery permitted a total removal of the lesion in 85% of cases and this with no mortality. Incomplete procedure was noted in cases of second surgery at the same site, and anterior location closed to the anterior artery. At the end of follow-up period, only 17% of patients were worsened according to McCormik’s classification. The majority of these patients were VHL patients and worsening was the result of multiple surgical procedures. The high rate of VHL patients in our series is due to our close collaboration with the national registry of VHL and Prof. Richard. We observed a predominance of neck pain over the symptoms related to the posterior column disturbances or cranial nerve dysfunction when they are located in the medulla oblongata, which differs from the data in the literature [11, 25, 26]. Although other modalities of treatment stereotaxic radiosurgery [27–31] have been recently proposed, surgery is still the treatment of choice for spinal and brainstem HB [13, 16–18, 22, 32–34]. The pre-operative angiography is important to clarify anatomical data of the vessels in the vicinity of the tumor, helping the surgical planning [35–38] and, with new techniques currently available (micro-catheters and guide wires), some lesions can be safely catheterized and embolized pre-operatively. Small lesions posteriorly located do not required angiography. Our experience concerning the benefit of pre-operative embolization is similar to Eskridge et al. [35] who reported that, after the embolization of nine lesions, the blood loss at the surgery was reported to be less important than expected by experienced surgeons and manipulation or removal of tumors was also subjectively easier. In our experience, preoperative embolization facilitates tumor dissection and removal and reduces blood loss. We think that embolization may be proposed to patients with a large and anterior situated lesion. Lunardi et al. [3] evaluated a series of 18 patients harboring sporadic HB under surgical treatment and presented 61% of clinical improvement and a peri-operative
123
136
mortality of 5%. Pietila et al. [19] referred 30 patients operated by spinal HB. Five patients presented post-operative neurological deterioration, 19 were unchanged and 6 patients improved their conditions. Xu et al. [15] reviewed the clinical and radiological data of 13 cases of intramedullary HB and found a rate of 100% of associated syringomyelia and, in the late post-operative period (3– 48 months), 12 patients (92.4%) had improved their condition and only 1 (7,6%) was worse than his initial condition. Cristante et al. [16] published, in 1999, a series of 19 patients, being 6 VHL and the others sporadic (with a total of 22 surgeries), operated for spinal HB. They evaluated the functional status by a scale similar to McCormick’s scale at the early post-operative (worse in 22.7%, unchanged in 59.1%, and improved in 18.2% of the patients), and at the late post-operative period (9.1% was still worse, 50% were unchanged, and 40.9% better than the pre-operative one). Our results are similar to the literature [16–20, 32, 34] and we did not show any difference when considering VHL disease as a distinct group. The surgical results for brainstem HB in patients with VHL disease in 12 patients who underwent 13 operations were presented by Weil et al. [26], and no neurological deterioration was seen after brainstem surgery. They concluded that the preoperative neurological function was the best predictor of long-term outcome and that brainstem HB can be removed safely when they become symptomatic or when the tumor has reached an important size. Wang et al. [34] showed an improvement rate in 61.5% of the patients after complete surgical resection of intramedullary hemangioblastomas. The main question is when to operate a spinal HB. This is a thorny question essentially for VHL patients with multiple lesions. We agree with Resche et al. [7] who concluded that only symptomatic and growing lesions should be operated. Non-symptomatic and stable lesions require conservative treatment and may be carefully followed by regular clinical and MRI follow-up examination. This opinion is based on our own experience and the fact that many authors still recommend conservative treatment for all asymptomatic hemangioblastomas [16, 32]. There is no evidence that treatment of non-symptomatic and stable hemangioblastoma is beneficial for patients. Good results obtained by microsurgical techniques allowed us to propose surgical removal of hemangioblastoma in case of growing lesions or symptomatic patients. Regarding the natural history of the lesions in the VHL group, there is not a consensus among the studies [6–8]. Our results do not allow us to calculate a rate of predictability for the appearance of new lesions or for the growing of previous lesions. Wanebo et al. [10] have done an extensive revision of this theme and they show the evidence of the growing cysts
123
J Neurooncol (2009) 93:133–137
being associated with the HB in producing the symptoms and being responsible for the surgery indication. In spinal HB, the largest tumors are more frequently associated with cysts. The cyst can grow more than the tumor but these tumors grow in an unpredictable manner with periods of rapid growth and periods of stabilization. They could not find factors that would predict the development of a cyst and the need for treatment. Conway et al. [5] concluded that, in 15 years of follow-up, 80% of the patients can develop a new lesion.
Conclusion Surgery is a safe and effective treatment for the medulla oblongata and spinal HB presenting acceptable morbidity and good outcomes. There was no difference between sporadic and the VHL group comparing the surgical results whenever multiple surgical procedure required by VHL patients made them worse compared to sporadic HB patients. For the VHL group, even for asymptomatic patients, a strict and lifelong follow-up is essential. Periodic image monitoring by IRM (once a year) allows the identification of new lesions or the regrowth of previous ones and the development of cysts or syringomyelia.
References 1. Brotchi J (2004) Surgical treatment of intramedullary tumors. Experience with 316 cases. Bull Mem Acad R Med Belg 159(5– 6):335–339 discussion 339-41 2. Emery E, Hurth M, C Lacroix-Jousselin et al (1994) Les he´mangioblastomes intra-rachidiens. A propos d’une serie re´cente de 20 cas. Neurochirurgie 40:165–173 3. Lunardi P, Cervoni L, Maleci A et al (1993) Isolated haemangioblastoma of spinal cord : report of 18 cases and a review of the literature. Acta Neurochir (Wien) 122:236–239. doi:10.1007/ BF01405535 4. Spetzger U, Bertalanffy H, Huffmann B et al (1996) Hemangioblastomas of the spinal cord and the brainstem: diagnostic and therapeutic features. Neurosurg Rev 19(3):147–151. doi:10.1007/BF00512042 5. Conway JE, Chou D, Clatterbuck R et al (2001) Hemangioblastomas of the central nervous system in von Hippel-Lindau syndrome and sporadic disease. Neurosurgery 48:55–63. doi:10.1097/00006123-200101000-00009 6. Joerger M, Koeberle D, Neumann HP et al (2005) Von HippelLindau disease—a rare disease important to recognize. Onkologie 28(3):159–163. doi:10.1159/000083860 7. Resche F, Moisan JP, Mantoura J et al (1993) Haemangioblastoma, Haemangioblastomatosis and von Hippel-Lindau Disease. Adv Tech Stand Neurosurg 20:199–277 8. Richard S, Martin S, David P et al (1998) Maladie de von HippelLindau et he´mangioblastomes du syste`me nerveux central. Applications cliniques des progre`s ge´ne´tiques. Neurochirurgie 44(4):258–266
J Neurooncol (2009) 93:133–137 9. Sora S, Ueki K, Saito N et al (2001) Incidence of von HippelLindau disease in hemangioblastoma patients: the University of Tokyo Hospital experience from 1954–1998. Acta Neurochir (Wien) 143(9):893–896. doi:10.1007/s007010170019 10. Wanebo JE, Lonser RR, Glenn GM et al (2003) The natural history of hemangioblastomas of the central nervous system in patients with von Hippel-Lindau disease. J Neurosurg 98(1):82– 94 11. Fukushima T, Sakamoto S, Iwaasa M et al (1998) Intramedullary hemangioblastoma of the medulla oblongata–two case reports and review of the literature. Neurol Med Chir (Tokyo) 38(8):489– 498. doi:10.2176/nmc.38.489 12. Jelsma RK, Jelsma LF, Johnson GS (1993) Surgical removal of brainstem astrocytomas and hemangioblastomas: report of three cases and review. Surg Neurol 39(6):494–510. doi:10.1016/00903019(93)90037-2 Review 13. Zhou LF, Du G, Mao Y et al (2005) Diagnosis and surgical treatment of brainstem hemangioblastomas. Surg Neurol 63(4):307–315. doi:10.1016/j.surneu.2004.07.038 discussion 315-6 14. Trost HA, Seifert V, Stolke D (1993) Advances in diagnosis and treatment of spinal hemangioblastomas. Neurosurg Rev 16(3):205–209. doi:10.1007/BF00304329 15. Xu QW, Bao WM, Mao RL et al (1994) Magnetic resonance imaging and microsurgical treatment of intramedullary hemangioblastoma of the spinal cord: clinical study. Neurosurgery 35:671–676. doi:10.1097/00006123-199410000-00013 16. Cristante L, Herrmann HD (1999) Surgical management of intramedullary hemangioblastoma of the spinal cord. Acta Neurochir (Wien) 141:333–340. doi:10.1007/s007010050308 17. Lee DK, Choe WJ, Chung CK et al (2003) Spinal cord hemangioblastoma: surgical strategy and clinical outcome. J Neurooncol 61(1):27–34. doi:10.1023/A:1021297622944 18. Maira G, Amante P, Denaro L et al (2001) Surgical treatment of cervical intramedullary spinal cord tumors. Neurol Res 23(8):835–842. doi:10.1179/016164101101199432 19. Pietila TA, Stendel R, Schilling A et al (2000) Surgical treatment of spinal hemangioblastomas. Acta Neurochir (Wien) 142:879– 886. doi:10.1007/s007010070073 20. Roonprapunt C, Silvera MV, Setton A et al (2001) Surgical management of isolated hemangioblastomas of the spinal cord. Neurosurgery 49:321–328. doi:10.1097/00006123-20010800000012 21. Sandalcioglu IE, Gasser T, Asgari S et al (2005) Functional outcome after surgical treatment of intramedullary spinal cord tumors: experience with 78 patients. Spinal Cord 43(1):34–41. doi:10.1038/sj.sc.3101668 22. Yasargil MG, Antic J, Laciga R et al. (1976) The microsurgical removal of intramedullary spinal hemangioblastomas. Report of twelve cases and a review of the literature. Surg Neurol (3):141– 148 23. McCormick PC, Torres R, Post KD et al (1990) Stein BM: Intramedullary ependymoma of the spinal cord. J Neurosurg 72:523–532
137 24. Sanford RA, Smith RA (1986) Hemangioblastoma of the cervicomedullary junction Report of three cases. J Neurosurg 64(2):317–321 25. Montaut J, Picard L, Roland J et al (1975) Intra and parabulbomedullary hemangioblastomas. Neurochirurgie 21(7):517–526 26. Weil RJ, Lonser RR, DeVroom HL et al (2003) Surgical management of brain stem hemangioblastomas in patients with von Hippel-Lindau disease. J Neurosurg 98:95–105 27. Chandler HC, Friedman WA (1994) Radiosurgical treatment of a hemangioblastoma. Neurosurgery 34:353–355. doi:10.1097/ 00006123-199402000-00021 28. Krishnan KG, Schackert G (2006) Outcomes of surgical resection of large solitary hemangioblastomas of the craniocervical junction with limitations in preoperative angiographic intervention: report of three cases. Zentralbl Neurochir 67(3):137–143. doi:10.1055/s-2006-933360 29. Pan L, Wang EM, Wang BJ et al (1998) Gamma knife radiosurgery for hemangioblastomas. Stereotact Funct Neurosurg 70(Suppl 1):179–186. doi:10.1159/000056420 30. Patrice SJ, Sneed PK, Flickinger JC et al (1996) Radiosurgery for hemangioblastoma: results of a multiinstitutional experience. Int J Radiat Oncol Biol Phys 35(3):493–499. doi:10.1016/S03603016(96)80011-3 31. Ryu SI, Kim DH, Chang SD (2003) Stereotactic radiosurgery for hemangiomas and ependymomas of the spinal cord. Neurosurg 15(5):E10 32. Lonser RR, Oldfield EH (2005) Microsurgical resection of spinal cord hemangioblastomas. Neurosurgery 57(4)(Suppl):372–376. doi:10.1227/01.NEU.0000176849.76663.E4. discussion 372-6 33. Van Velthoven V, Reinacher PC, Klisch J et al (2003) Treatment of intramedullary hemangioblastomas, with special attention to von Hippel-Lindau disease. Neurosurgery 53(6):1306–1313. doi:10.1227/01.NEU.0000093497.81390.29 discussion 1313-4 34. Wang C, Zhang J, Liu A et al (2001) Surgical management of medullary hemangioblastoma Report of 47 cases. Surg Neurol 56(4):218–226. doi:10.1016/S0090-3019(01)00590-0 discussion 226-7 35. Eskridge JM, McAuliffe W, Harris B et al (1996) Preoperative endovascular embolization of craniospinal hemangioblastomas. AJNR Am J Neuroradiol 17:525–531 36. Ohtakara K, Kuga Y, Murao K et al (2000) Preoperative embolization of upper cervical cord hemangioblastoma concomitant with venous congestion–case report. Neurol Med Chir (Tokyo) 40(11):589–593. doi:10.2176/nmc.40.589 37. Standard SC, Ahuja A, Livingston K et al (1994) Endovascular embolization and surgical excision for the treatment of cerebellar and brain stem hemangioblastomas. Surg Neurol 41:405–410. doi:10.1016/0090-3019(94)90035-3 38. Tampieri D, Leblanc R, TerBrugge K (1993) Preoperative embolization of brain and spinal hemangioblastomas. Neurosurgery 33(3):502–505. doi:10.1097/00006123-199309000-00022 discussion 505
123
