Clinical Neurology and Neurosurgery 128 (2015) 78–83
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Management of ruptured posterior fossa arteriovenous malformations Elsa Magro a , Jonathan Chainey a , Chiraz Chaalala a , Hosam Al Jehani a,b , Jean-Yves Fournier a,c , Michel W. Bojanowski a,∗ a b c
Division of Neurosurgery, Division of Surgery, Centre Hospitalier de l’Université de Montréal, Montréal, QC, Canada Department of Neurosurgery, King Fahad University Hospital, Dammam University, Dammam, Saudi Arabia Department of Neurosurgery, Cantonal Hospital of St. Gall, CH-9007 St. Gall, Switzerland
a r t i c l e
i n f o
Article history: Received 19 August 2014 Received in revised form 16 October 2014 Accepted 9 November 2014 Available online 15 November 2014 Keywords: Arteriovenous malformation Posterior fossa Ruptured Management
a b s t r a c t Objective: Posterior fossa arteriovenous malformations (pAVMs) are rare and because of their location at or close to vital structures, their treatment remains challenging despite overall improvements in the management of cerebrovascular lesions. We reviewed our recent series of ruptured pfAVMs in search of guiding principles in the management of these complex lesions. Methods: This is a retrospective series of consecutive patients admitted for a ruptured pfAVM from 2002 to 2013. We analyzed clinical and radiological data, as well as initial and definitive management. Outcome was assessed using the modified Rankin Scale (mRS) at 6 months. Results: The study included 34 patients (19 women and 15 men). Upon admission, 79% of patients presented with an intra-cerebellar hematoma, 42% of which required urgent drainage. Hydrocephaly was also present in 82% of patients, 56% of which required emergency ventriculostomy. There was an aneurysm associated with the AVM in 47% of cases. In 38% of the cases, the aneurysm was the source of the hemorrhage. Only 68% of patients were amenable to undergo treatment of the AVM: 24% exclusively by surgery, 9% by embolization, 3% by radiosurgery, and 32% using combined means. Five patients died within the first week: one as a direct result of the severity of the hemorrhage, and the other four due to re-bleeding before treatment. Outcome was favorable (mRS 0–2) in 71% of patients. Conclusion: Patients with a ruptured pfAVM are often comatose upon admission, requiring emergency live-saving surgical treatment. An associated aneurysm is often the source of bleeding which if dealt with immediately, offers time to plan the most appropriate strategies to eliminate the AVM. Nevertheless, early re-bleeding is frequent, and a cause of concern as it often leads to death. Despite the gravity of the clinical condition upon admission, outcome is favorable for those amenable to treatment. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Posterior fossa arteriovenous malformations (pfAVMs) are rare, representing only 5–15% of all AVMs, with an incidence in the population of 2 per 10,000 [1–5]. In contrast with supra-tentorial AVMs, pfAVMs present with hemorrhage much more frequently, varying between 72% and 92% depending on the series [1,2,6–8]. A few studies have suggested that the posterior fossa location is an independent risk factor for rupture [4,9,10]. Because of the presence
∗ Corresponding author at: Division of Neurosurgery, Hôpital Notre Dame, 1560 Sherbrooke Est, Montréal, Quebec H2L 4M1, Canada. Tel.: +1 514 890 8000; fax: +1 514 412 7816. E-mail addresses: [email protected] (E. Magro), [email protected] (J. Chainey), [email protected] (C. Chaalala), [email protected] (H.A. Jehani), [email protected] (J.-Y. Fournier), [email protected] (M.W. Bojanowski). http://dx.doi.org/10.1016/j.clineuro.2014.11.007 0303-8467/© 2014 Elsevier B.V. All rights reserved.
of vital structures in this confined area, they are life-threatening [1–3,11]. Despite of the availability and improvements in multimodal treatments, their management remains challenging [6,7]. The objective of this study is to review our recent series of ruptured pfAVMs in search of guiding principles in the management of these complex lesions. 2. Materials and methods We reviewed retrospectively a consecutive series of patients admitted for a ruptured pfAVM in our neurosurgical department, between March 2002 and August 2013. Clinical and radiological data were analyzed as well as treatment modalities. AVMs were classified according to the Spetzler-Martin scale [12]. They were considered eloquent if they involved the deep cerebellar nuclei, the brainstem or the cerebellar peduncles. Venous drainage was defined as deep, when at least one vein drained into the deep
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
venous system. Associated aneurysms were classified as intranidal, prenidal and remote according to Redekop et al. [13] Outcome was assessed using the modified Rankin Scale (mRS) at 6 months.
Table 3 Management in emergency.
Hematoma Evacuation Aneurysm Embolization Hydrocephaly Ventricular drain
3. Results 3.1. Patients Between March 2002 and August 2013, 34 patients (19 women and 15 men), with a mean age of 46 years (range from 6 to 73 years), were admitted for a ruptured pfAVM. 3.2. Clinical presentation Eighteen patients (53%) were admitted with impaired consciousness: eight had a glasgow coma scale (GCS) between 13 and 9 and 10, a score inferior to 9, thus requiring intubation. Nine of the 24 patients not requiring intubation on admission, presented with neurological deficits and two with trigeminal neuralgia (Table 1). Initial CT scan revealed a hemorrhage in all cases: a cerebellar hematoma in 27 patients (79%), a subarachnoid hemorrhage (SAH) in 27 patients (79%), an intra-ventricular hemorrhage (IVH) in 18 patients (53%), and a hydrocephaly in 28 patients (82%). As for the 10 intubated patients, 9 presented with a hematoma and all of them with hydrocephaly (Table 1). 3.3. AVM features An AVM was located in a cerebellar hemisphere in 17 cases, in the vermis in 8 cases, in both these locations in 2 cases, and in the brainstem in 7. The Spetzler-Martin grade was as follows: 11 patients were classified as grade I, 14 patients were grade II, and 9 grade III. No AVM graded IV or V was observed. A deep venous drainage was present in 14 of the 34 AVMs (Tables 2 and 5). Table 1 Clinical and radiological initial presentation. Initial presentation
Total, n = 34
Intubate, n = 10
Initial GCS 15–14 13–9 8–3 Neurological deficits Ataxia Dysmetry Dysarthria Diplopia Trigeminal neuralgia Hemorrhage SAH IVH Hematoma Hydrocephaly
16 8 10 9 8 3 1 4 2 34 27 (79%) 18 (53%) 27 (79%) 28 (82%)
– – 10
79
Total, n = 34
Intubate, n = 10
27 (79%) 14 (42%) 16 (47%) 12(35%) 28(82%) 19 (56%)
9 7 9 7 10 10
Conventional angiogram performed on all 34 patients revealed an aneurysm in 16 (47%) of them, 9 of whom had multiple aneurysms, for a total of 32 aneurysms. The distribution of these aneurysms was as follows: 24 pre-nidals, 5 intra-nidals, and 3 aneurysms unrelated to the AVM. Based on the location of the hemorrhage as seen on the CT-scan it was possible to conclude that the source of the bleeding was the aneurysm in 13 cases (38%) and the AVM for all other cases. Of the eighteen patients who did not have an associated aneurysm, 3 (16%) had a GOS of less than 9, whereas of the 16 patients with an aneurysm, 7 (47%) had a GOS of less than 9. Of the 10 patients who were intubated upon admission, 7 (70%) had an aneurysm responsible for the bleeding. 3.4. Management Because of the life-threatening clinical condition upon admission, 14 patients (42%) required an urgent evacuation of a cerebellar hematoma (Table 3). Drainage of cerebrospinal fluid (CSF) was necessary in 19 patients (56%), including in all those who were comatose. In cases (13) where the source of hemorrhage was a ruptured aneurysm, aneurysms in 12 patients (35%) were embolized and one was clipped while evacuating a hematoma. The mean time between admission and embolization was 16 h (range from 1 h to 39 h). As for the AVMs (Table 4), 23 patients (68%) were amenable to treatment: 8 patients (24%) underwent surgery alone, 3 patients (9%) underwent embolization alone, one patient (3%) received radiosurgery alone, and 11 patients (32%) were treated with a combination of these modalities. Of these 23 treated patients, in 21, treatment resulted in complete exclusion of the AVM. A residual AVM was seen in two patients, one who had been treated by radiosurgery alone, and one who had undergone embolization followed by radiosurgery. In six patients no attempt was made to treat the AVM: in 5 of them the AVMs were located in the brainstem, and 1 patient with a grade II AVM chose not to be treated. Case examples from our series are illustrated in Figs. 1 and 2. 3.5. Complications
10 7 7 9 10
Abbreviations: GCS, glasgow coma scale; SAH, subarachnoid hemorrhage; IVH, intraventricular hemorrhage.
Table 2 AVM features: location and Spetzler-Martin grade. Location/Spetzler-Martin grade
I
II
III
IV/V
CH V CHV Brainstem Total
7 4 0 0 11
9 3 1 1 14
1 1 1 6 9
0
Abbreviations: CH, cerebellar hemispheres; V, vermis; CHV, cerebellar hemisphere and vermis.
In this series, 5 patients (15%) re-bled after initial hemorrhage. In four of them, the re-bleeding occurred soon after admission, before initial treatment, all of them between the second and the third day. For these 4 patients, the reason for re-bleeding was an associated aneurysm in 2 cases and the AVM itself in two others (Table 5). Concerning the 2 AVMs that re-bled, both were located in the cerebellar hemisphere, one with deep venous drainage (grade II) and the other without (grade I). All four patients died as a result of re-bleeding. As for the fifth patient, re-bleeding occurred months after the initial treatment, while waiting for definitive treatment. The reason for delaying definitive treatment was the severity of the patient’s neurological status after the rupture of the AVM. This one patient who re-bled improved gradually (mRS 3). Of the 19 patients who required an external ventricular drain, one bled in the tract while inserting the drain. This patient had a favorable outcome (mRS 2). Two other patients contracted meningitis resulting from the drain, and were successfully treated. Six patients required a ventriculoperitonal shunt (VPS). Two patients experienced a stroke: one
80
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
Table 4 Management of AVM. Grade
n
Surgery alone
Embolization alone
Radiosurgery alone
Combined
Observation
Death
I II III IV/V Total
11 14 9 0 34
5 3 – – 8 (24%)
2 1 – – 3 (9%)
– – 1
2 6 3 – 11 (32%)
0 1 5 – 6 (18%)
2 3 – – 5 (14%)
1 (3%)
Fig. 1. Case illustration of a Spetzler-Martin grade 3 AVM of right cerebellar hemisphere in a 27-year-old patient admitted comatose (glasgow coma scale 3) witch required an urgent evacuation of a cerebellar hematoma. The patient improved and the residual AVM was approached by surgical resection after two stages of endovascular embolization. Final outcome was favorable (mRS 2). (A) Axial head CTscan on admission showing a cerebellar hematoma with compression of the fourth ventricle. (B) Head CTscan after drainage of the hematoma. (C) Sagittal head angioCTscan and (D) and (E) cerebral angiogram in lateral and antero-posterior view, respectively, showing an AVM of the inferior part of the right cerebellar hemisphere, vascularized by the right posterior inferior and anterior inferior cerebellar arteries, and with a deep venous drainage. (F) Cerebral angiogram in lateral view showing complete cure of the AVM after endovascular embolization and surgery.
was asymptomatic and seen on imagery post-embolization, one presented dysmetria due to hemispheric cerebellar stroke after conventional angiogram. Both patients had a favorable outcome. 3.6. Clinical outcome In this series, favorable outcome (mRS 0–2) was achieved in 24/34 patients (71%). Of the 10 patients who were intubated upon admission, 4 had a favorable outcome (mRS 0–2). The follow-up period for survivors (n = 29) in this series ranges from 6 months to 110 months (mean 31 months). There were unfortunately 5 deaths, all of them within the first week. One patient died as a direct result of the severity of the hemorrhage, and the other four, as a direct result of re-bleeding. 4. Discussion PfAVMs are rare and located at or close to critical areas of the brain [1–5]. Unlike supra-tentorial AVMs, which may manifest with epilepsy, pfAVMs do not, thus diagnosis is most often done following a rupture [14]. Because of their location, rupture may lead to a life-threatening situation. In fact, in our series, 53% of patients presented with some level of impaired consciousness, half of whom required intubation. As a direct result of the hemorrhage, cerebellar hematoma was found in 79% of the patients,
intra-ventricular hemorrhage in 53%, and hydrocephaly in 82%. Therefore, it is highly likely that most patients admitted for ruptured pfAVMs, will require urgent surgical intervention. In our series of ruptured pfAVMs, we found a high incidence of associated aneurysms. These associated aneurysms were the cause of bleeding in 38% of our patients. The literature also suggests that pfAVMs are more frequently associated with aneurysms than are supra-tentorial AVMs [2,7,11,15,16,6,17] and that these aneurysms are often the cause of hemorrhage [2,18,19]. In our series, there was a reciprocal relation between the severity of the patients’ clinical condition upon admission and the source of the hemorrhage. In fact, 70% of our intubated patients had an associated aneurysm. Da Costa et al. [11], in their series of pfAVMs observed that the presence of aneurysms was an independent factor of poor mRS score at followup. A recent study by Abla et al. [20] also suggested that posterior fossa location and associated aneurysm were significant predictors of severe outcome. Of the 13 patients in our series who presented with an associated aneurysm responsible for the hemorrhage, 12 of them were embolized urgently. Meisel et al. [21] noted that associated aneurysms should be the first targets of endovascular treatment because they seem to increase the risk of re-bleeding. In our series, of the four patients where early re-bleeding occurred, two of them had an associated aneurysm, which was the cause of both the initial hemorrhage and the re-bleeding. This would lead us to conclude
Table 5 Database of the series. GCS
Type/L
Hemorrhage/Hydro
Drain H/CSF
Aneurysm/type/B/M
W/50 W/28 M/63 W/29 M/52 W/64 W/71 M/29 W/50 M/20 W/50 W/16 M/61 W/19 M/37 M/62 M/58 M/32 W/50 W/58 W/46 W/6 M/27 W/44 W/49 W/59 W/48 M/62 W/15 M/73 M/62 M/68 W/55 M/62
15 14 13 14 15 14 15 15 7 3 13 12 15 13 15 15 15 15 14 13 13 3 3 15 13 15 6 8 10 8 3 14 3 3
I/CH I/CH I/V I/CH II/CH/dp II/V/dp I/CH I/CH III/PM/dp II/V II/CH II/CH III/PCA/dp III/CHV/dp III/P/dp I/V III/M/dp III/M/dp II/CH/dp II/PCA II/CH I/CH II/CH/dp I/V II/CHV III/M III/CH/dp II/CH III/V/dp I/CH II/V/dp II/CH/dp II/CH I/V
SAH SAH/Hematoma SAH/Hematoma SAH/Hematoma SAH/IVH/Hematoma/hydro SAH/Hematoma SAH/Hematoma/hydro Hematoma/Hydro SAH/HIV/hydro Hematoma/Hydro SAH/IVH/Hematoma/hydro Hematoma SAH SAH/IVH/Hematoma/hydro SAH/HIV/hydro SAH/IVH/Hematoma/hydro SAH/HIV/Hydro Hematoma SAH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro HIV/Hematoma/hydro HIV/Hematoma/hydro SAH/HIV/hydro Hematoma SAH SAH/Hematoma/hydro SAH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro
No No No No Yes/EVD No No No EVD/VPS Yes/EVD EVD No No Yes/EVD No No/EVD/VPS EVD/VPS No Yes/no No Yes/EVD Yes/EVD Yes/EVD/VPS EVD/VPS No No Yes/EVD Yes/EVD Yes/EVD/VPS No/EVD Yes/EVD Yes/EVD Yes/EVD Yes/EVD
No No No No 1/intra/yes/surgery 2/pre/no/obs No No No 1/pre/yes/embo 4/pre/yes/embo No 1/pre/yes/embo No No 2/pre/yes/embo No No 4/2pre, 2 intra/yes/embo No No 2/pre/yes/embo No 4/3pre, 1un/yes/embo 1/un/no/obs No 2/pre/yes/embo No 1/intra/yes/embo 1/un/no/obs 1/pre/yes/embo No 3/pre/yes/embo 2/pre, intra/yes/embo
Re-bleed
Time H/T
Initial
Definitive
Time I/D
Result
mRS
9 days 2 years 5 months 6 months 12 days
Embo Surgery Surgery Surgery Surgery Obs Surgery Surgery Obs Embo Embo Embo RadioS Embo (2×) Obs Embo Obs Obs Embo Embo Surgery Embo Embo (2×) Embo Surgery Obs Embo Embo Embo
Surgery
10 months
CR CR CR CR CR
0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 2 2 2 3 3 3 3 3 6 6 6 6 6
1 month 20 days 18 months 40 days 2 months 6 months 3 months 1 day
AVM 2 y
45 days 5 months 1 month 9 days 3 months 20 days 2 years 5 months 20 days 15 days
AVM 2 d AVM 3 d Aneur 2 d Aneur 2 d
CR CR Surgery Surgery Surgery
2 days 20 days 3 months
Surgery
11 months
CR CR CR PO CR CO
Surgery RadioS
16 days 1 month
Surgery Surgery
15 days 20 days
Surgery Surgery Surgery
10 days 20 days Same day
CR PO CR CO CR CR CR CR CR CR
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
S/A
Abbreviations: S, sex; W, women; M, men; A; age; GCS, glasgow coma scale; CH cerebellar hemispheres, V vermis, CHV cerebellar hemisphere and vermis, M, mesencphal; P, pons; PCA, pontocerebellar angle; dp, deep venous drainage; SAH subarachnoid hemorrhage, IVH intraventricular hemorrhage, Hydro, hydrocephaly; Drain H, evacuation hematoma; Drain CSF, drainage of cerebrospinal fluid; VPS, ventriculoperitoneal shunt; EVD, external ventricular drain; Aneur; number of associated aneurysms; type, intra as intranidal, pre as prenidal and un as unrelated; B, bleeding related to aneurysm; M, aneurysm management; Re-bleed, as re-bleeding due to and delay since the first bleeding; Time H/T, delay between hemorrhage and initial treatment of the AVM; Initial, modality of initial treatment of the AVM; definitive, modality of definitive treatment of the AVM when multimodality; Time I/D, delay between initial and definitive treatment when multimodality; CR, complete resection; CO complete occlusion; PO partial occlusion; mRS, modified ranking scale at 6 months.
81
82
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
Fig. 2. Case illustration of a vermian and right paravermian Spetzler-Martin grade 2 AVM in a 20-year-old patient admitted comatose (glasgow coma scale 3) and operated on urgently upon admission for drainage of a cerebellar hematoma. The cerebral angiogram showed a pre-nidal aneurysm embolized the day of admission. The patient improved neurologically and the AVM was subsequently approached via endovascular embolization followed by complete microsurgical resection. Final outcome was favorable (mRS 0). (A and B) Axial head CTscan, on admission and after drainage of the hematoma, respectively. (C and D) Initial cerebral angiogram antero-posterior and lateral view, respectively, showing the vermian and right paravermian AVM vascularized by the right superior cerebellar artery and the pre-nidal aneurysm (white arrow) localized on its feeding artery. (E and F) Cerebral angiogram antero-posterior and lateral view, respectively, after endovascular occlusion of the aneurysm (white arrow). (G and H) Cerebral angiogram antero-posterior and lateral view, respectively, after partial endovascular embolization of the AWM. (I and J) Cerebral angiogram antero-posterior and lateral view, respectively, after complete microsurgical resection of the AVM.
that whenever a patient is admitted for a ruptured pfAVM, we must be vigilant to search for an aneurysm on the angiogram, since it is highly likely that it could be responsible for the hemorrhage and that there is an associated high risk of re-bleeding, thus requiring prompt treatment. Most of the time, the associated aneurysms can be successfully treated by endovascular means, thus reducing the urgency for definitive AVM treatment. When there is no associated aneurysm, besides the initial bleeding of an AVM, it is not yet known whether there are other characteristics that could indicate the potential for its early rebleeding. In our series, 2 patients without an associated aneurysm, re-bled early. This seems to be a higher rate compared to what has been found for supra-tentorial AVMs. This may suggest that the posterior fossa location itself may be a risk factor for early rebleeding. Although our series is small, this observation has also been reported by others [18]. Therefore, when there is no associated aneurysm responsible for the initial hemorrhage, it may be appropriate to consider early definitive treatment of ruptured pfAVMs. Even though it may be justified to postpone definitive treatment in cases of non-life-threatening intra-cerebral hematoma or brain swelling, our series suggests that this delay could result in fatal re-bleeding. A larger series may provide us with more substantial evidence. There are several options available in definitive treatment of pfAVMs: surgical resection, embolization, radiation, and a combination of these modalities. Despite these various options, in our series, 32% of patients, including those who died, could not benefit from definitive treatment at all; in most of them, the AVMs were located in the brainstem. There are controversies as to which treatment modality to choose since all treatments pose some degree of risk [6,7,15,16,18,22,23]. Generally, because of the risk of re-bleeding, the goal is to eliminate the AVM with minimal delay. Therefore, in considering the various options, in our series radiotherapy was reserved for those patients where the characteristics of the AVM or the patient’s clinical condition did not allow for safe definitive treatment by either surgery or embolization. As for AVMs in the brainstem, we preferred to abstain from treating them and none of
them re-bled. Other authors have proposed radiotherapy in such cases with or without prior embolization [7,6,24]. In our series, surgery alone was the treatment of choice for grade I patients with small AVMs where dissection is usually straightforward, unless it was thought that embolization alone could offer a potential cure. For grade II and III patients where the AVMs were larger, when possible we embolized first to reduce the flow, and followed with surgery, except for one case, which was followed by radiotherapy. The pre-operative use of embolization is debatable because the possible increased risk due to an added treatment. According to some authors, cerebellar grades I and II should be treated by surgery alone [7,6,25,26]. In fact, in our series, the only complication resulting from treatment occurred following embolization. However, the authors believe that pre-operative embolization can actually add to the success of surgery and decrease the risk of complications. It is not known how many complications due to surgery were prevented by pre-operative embolization. More studies are needed to determine its role. Despite the location of the hemorrhage and the severity of the patients’ clinical condition upon admission, aggressive treatment can result in favorable outcome in most patients. Outcome can be improved by taking measures to avoid re-bleeding and by better identifying which treatment modalities are best suited for which patients, in order to reduce the risk of complications following treatment. There are presently no uniform treatment modalities in the literature. The risks of complications following treatment has been found to be significant in the recent ARUBA study [27]. Although ARUBA study dealt with un-ruptured AVMs, we must keep in mind its results regarding the rate of complications due to treatment since it might also be applicable to ruptured AVMs. Future studies are needed to clarify this point. We recognize that the small size of this series and its retrospective nature pose certain limitations. However, bias selection in greatly reduced since this series represents all patients admitted in our neurosurgical unit during the study period. Although some similar conclusions might have been drawn in various other papers, by no means are they well established. Due to the importance of the potential implications raised in these studies, it
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
would advisable to continue documenting data leading to guiding principles in the management of ruptured pfAVMs. 5. Conclusion Patients with a ruptured pfAVM are often comatose upon admission, requiring emergency live-saving surgical treatment. An associated aneurysm is often the source of bleeding which if dealt with immediately, offers time to plan the most appropriate strategies to eliminate the AVM. Nevertheless, early re-bleeding is frequent, and a cause of concern as it often leads to death. More studies are needed to better define which treatment modalities are most appropriate and when. Despite the gravity of the clinical condition upon admission, outcome is favorable for those amenable to treatment. References [1] Batjer H, Samson D. Arteriovenous malformations of the posterior fossa: clinical presentation, diagnostic evaluation and surgical treatment. Neurosurg Rev 1986;9:287–96. [2] Drake CG, Friedman a H, Peerless SJ. Posterior fossa arteriovenous malformations. J Neurosurg 1986;64:1–10. [3] Arnaout OM, Gross BA, Eddleman CS, Bendok BR, Getch CC, Batjer HH. Posterior fossa arteriovenous malformations. Neurosurg Focus 2009;26:E12. [4] Khaw AV, Mohr JP, Sciacca RR, Schumacher HC, Hartmann A, Pile-Spellman J, et al. Association of infratentorial brain arteriovenous malformations with hemorrhage at initial presentation. Stroke 2004;35:660–3. [5] Berman MF, Sciacca RR, Pile-Spellman J, Stapf C, Connolly ES, Mohr JP, et al. The epidemiology of brain arteriovenous malformations. Neurosurgery 2000;47:389–96, discussion 397. [6] Lawton MT, Hamilton MG, Spetzler RF. Multimodality treatment of deep arteriovenous malformations: thalamus, basal ganglia, and brain stem. Neurosurgery 1995;37. [7] Kelly ME, Guzman R, Sinclair J, Bell-Stephens TE, Bower R, Hamilton S, et al. Multimodality treatment of posterior fossa arteriovenous malformations. J Neurosurg 2008;108:1152–61. [8] Bowden G, Kano H, Tonetti D, Niranjan A, Flickinger J, Lunsford LD. Stereotactic radiosurgery for arteriovenous malformations of the cerebellum. J Neurosurg 2014;120:583–90. [9] Hernesniemi JA, Dashti R, Juvela S, Väärt K, Niemelä M, Laakso A. Natural history of brain arteriovenous malformations: a long-term follow-up study of risk of hemorrhage in 238 patients. Neurosurgery 2008;63:823–9, discussion 829–831. [10] Wilkins RH. Natural history of intracranial vascular malformations: a review. Neurosurgery 1985;16:421–30. [11] Da Costa L, Thines L, Dehdashti a R, Wallace MC, Willinsky Ra, Tymianski M, et al. Management and clinical outcome of posterior fossa arteriovenous
[12] [13]
[14]
[15]
[16] [17]
[18] [19]
[20]
[21]
[22]
[23]
[24]
[25]
[26] [27]
83
malformations: report on a single-centre 15-year experience. J Neurol Neurosurg Psychiatry 2009;80:376–9. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986;65:476–83. Redekop G, TerBrugge K, Montanera W, Willinsky R. Arterial aneurysms associated with cerebral arteriovenous malformations: classification, incidence, and risk of hemorrhage. J Neurosurg 1998;89:539–46. Garcia Monaco R, Alvarez H, Goulao A, Pruvost P, Lasjaunias P. Posterior fossa arteriovenous malformations. Angioarchitecture in relation to their hemorrhagic episodes. Neuroradiology 1990;31:471–5. Maruyama K, Koga T, Niranjan A, Kondziolka D, Flickinger JC, Lunsford LD. Radiosurgery for brainstem arteriovenous malformation. Prog Neurol Surg 2013;27:67–72. Matsumura H, Makita Y, Someda K, Kondo a. Arteriovenous malformations in the posterior fossa. J Neurosurg 1977;47:50–6. Kouznetsov E, Weill A, Ghostine J, Gentric J, Raymond J, Roy D. Association between posterior fossa arteriovenous malformations and prenidal aneurysm rupture: potential impact on management. Neurosurg Focus 2014;37: 1–4. Samson D, Batjer H. Arteriovenous malformations of the cerebellar vermis. Neurosurgery 1985;16:341–9. Mpotsaris a, Loehr C, Harati A, Lohmann F, Puchner M, Weber W. Interdisciplinary clinical management of high grade arteriovenous malformations and ruptured flow-related aneurysms in the posterior fossa. Interv Neuroradiol 2010;16:400–8. Abla A, Nelson J, Rutledge C, Young WL, Kim H, Lawton MT. The natural history of AVM hemorrhage in the posterior fossa: comparison of hematoma volumes and neurological outcomes in patients with ruptured infra- and supratentorial AVMs. Neurosurg Focus 2014;37:1–6. Meisel HJ, Mansmann U, Alvarez H, Rodesch G, Brock M, Lasjaunias P. Cerebral arteriovenous malformations and associated aneurysms: analysis of 305 cases from a series of 662 patients. Neurosurgery 2000;46:793–800, discussion 800–2. Rodríguez-Hernández A, Kim H, Pourmohamad T, Young WL, Lawton MT. Cerebellar arteriovenous malformations: anatomic subtypes, surgical results, and increased predictive accuracy of the supplementary grading system. Neurosurgery 2012;71:1111–24. Thines L, Dehdashti AR, da Costa L, Tymianski M, ter Brugge KG, Willinsky Ra, et al. Challenges in the management of ruptured and unruptured brainstem arteriovenous malformations: outcome after conservative, single-modality, or multimodality treatments. Neurosurgery 2012;70:155–61, discussion 161. Massager N, Régis J, Kondziolka D, Njee T, Levivier M. Gamma knife radiosurgery for brainstem arteriovenous malformations: preliminary results. J Neurosurg 2000;93(Suppl. 3):102–3. Sinclair J, Kelly ME, Steinberg GK. Surgical management of posterior fossa arteriovenous malformations. Neurosurgery 2006;58:ONS189–201, discussion ONS201. Viale GL, Pau A, Viale ES. Surgical treatment of arteriovenous malformations of the posterior fossa. Surg Neurol 1979;12:379–84. Mohr JP, Parides MK, Stapf C, Moquete E, Moy CS, Overbey JR, et al. Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. Lancet 2013;6736:1–8.
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Management of ruptured posterior fossa arteriovenous malformations Elsa Magro a , Jonathan Chainey a , Chiraz Chaalala a , Hosam Al Jehani a,b , Jean-Yves Fournier a,c , Michel W. Bojanowski a,∗ a b c
Division of Neurosurgery, Division of Surgery, Centre Hospitalier de l’Université de Montréal, Montréal, QC, Canada Department of Neurosurgery, King Fahad University Hospital, Dammam University, Dammam, Saudi Arabia Department of Neurosurgery, Cantonal Hospital of St. Gall, CH-9007 St. Gall, Switzerland
a r t i c l e
i n f o
Article history: Received 19 August 2014 Received in revised form 16 October 2014 Accepted 9 November 2014 Available online 15 November 2014 Keywords: Arteriovenous malformation Posterior fossa Ruptured Management
a b s t r a c t Objective: Posterior fossa arteriovenous malformations (pAVMs) are rare and because of their location at or close to vital structures, their treatment remains challenging despite overall improvements in the management of cerebrovascular lesions. We reviewed our recent series of ruptured pfAVMs in search of guiding principles in the management of these complex lesions. Methods: This is a retrospective series of consecutive patients admitted for a ruptured pfAVM from 2002 to 2013. We analyzed clinical and radiological data, as well as initial and definitive management. Outcome was assessed using the modified Rankin Scale (mRS) at 6 months. Results: The study included 34 patients (19 women and 15 men). Upon admission, 79% of patients presented with an intra-cerebellar hematoma, 42% of which required urgent drainage. Hydrocephaly was also present in 82% of patients, 56% of which required emergency ventriculostomy. There was an aneurysm associated with the AVM in 47% of cases. In 38% of the cases, the aneurysm was the source of the hemorrhage. Only 68% of patients were amenable to undergo treatment of the AVM: 24% exclusively by surgery, 9% by embolization, 3% by radiosurgery, and 32% using combined means. Five patients died within the first week: one as a direct result of the severity of the hemorrhage, and the other four due to re-bleeding before treatment. Outcome was favorable (mRS 0–2) in 71% of patients. Conclusion: Patients with a ruptured pfAVM are often comatose upon admission, requiring emergency live-saving surgical treatment. An associated aneurysm is often the source of bleeding which if dealt with immediately, offers time to plan the most appropriate strategies to eliminate the AVM. Nevertheless, early re-bleeding is frequent, and a cause of concern as it often leads to death. Despite the gravity of the clinical condition upon admission, outcome is favorable for those amenable to treatment. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Posterior fossa arteriovenous malformations (pfAVMs) are rare, representing only 5–15% of all AVMs, with an incidence in the population of 2 per 10,000 [1–5]. In contrast with supra-tentorial AVMs, pfAVMs present with hemorrhage much more frequently, varying between 72% and 92% depending on the series [1,2,6–8]. A few studies have suggested that the posterior fossa location is an independent risk factor for rupture [4,9,10]. Because of the presence
∗ Corresponding author at: Division of Neurosurgery, Hôpital Notre Dame, 1560 Sherbrooke Est, Montréal, Quebec H2L 4M1, Canada. Tel.: +1 514 890 8000; fax: +1 514 412 7816. E-mail addresses: [email protected] (E. Magro), [email protected] (J. Chainey), [email protected] (C. Chaalala), [email protected] (H.A. Jehani), [email protected] (J.-Y. Fournier), [email protected] (M.W. Bojanowski). http://dx.doi.org/10.1016/j.clineuro.2014.11.007 0303-8467/© 2014 Elsevier B.V. All rights reserved.
of vital structures in this confined area, they are life-threatening [1–3,11]. Despite of the availability and improvements in multimodal treatments, their management remains challenging [6,7]. The objective of this study is to review our recent series of ruptured pfAVMs in search of guiding principles in the management of these complex lesions. 2. Materials and methods We reviewed retrospectively a consecutive series of patients admitted for a ruptured pfAVM in our neurosurgical department, between March 2002 and August 2013. Clinical and radiological data were analyzed as well as treatment modalities. AVMs were classified according to the Spetzler-Martin scale [12]. They were considered eloquent if they involved the deep cerebellar nuclei, the brainstem or the cerebellar peduncles. Venous drainage was defined as deep, when at least one vein drained into the deep
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
venous system. Associated aneurysms were classified as intranidal, prenidal and remote according to Redekop et al. [13] Outcome was assessed using the modified Rankin Scale (mRS) at 6 months.
Table 3 Management in emergency.
Hematoma Evacuation Aneurysm Embolization Hydrocephaly Ventricular drain
3. Results 3.1. Patients Between March 2002 and August 2013, 34 patients (19 women and 15 men), with a mean age of 46 years (range from 6 to 73 years), were admitted for a ruptured pfAVM. 3.2. Clinical presentation Eighteen patients (53%) were admitted with impaired consciousness: eight had a glasgow coma scale (GCS) between 13 and 9 and 10, a score inferior to 9, thus requiring intubation. Nine of the 24 patients not requiring intubation on admission, presented with neurological deficits and two with trigeminal neuralgia (Table 1). Initial CT scan revealed a hemorrhage in all cases: a cerebellar hematoma in 27 patients (79%), a subarachnoid hemorrhage (SAH) in 27 patients (79%), an intra-ventricular hemorrhage (IVH) in 18 patients (53%), and a hydrocephaly in 28 patients (82%). As for the 10 intubated patients, 9 presented with a hematoma and all of them with hydrocephaly (Table 1). 3.3. AVM features An AVM was located in a cerebellar hemisphere in 17 cases, in the vermis in 8 cases, in both these locations in 2 cases, and in the brainstem in 7. The Spetzler-Martin grade was as follows: 11 patients were classified as grade I, 14 patients were grade II, and 9 grade III. No AVM graded IV or V was observed. A deep venous drainage was present in 14 of the 34 AVMs (Tables 2 and 5). Table 1 Clinical and radiological initial presentation. Initial presentation
Total, n = 34
Intubate, n = 10
Initial GCS 15–14 13–9 8–3 Neurological deficits Ataxia Dysmetry Dysarthria Diplopia Trigeminal neuralgia Hemorrhage SAH IVH Hematoma Hydrocephaly
16 8 10 9 8 3 1 4 2 34 27 (79%) 18 (53%) 27 (79%) 28 (82%)
– – 10
79
Total, n = 34
Intubate, n = 10
27 (79%) 14 (42%) 16 (47%) 12(35%) 28(82%) 19 (56%)
9 7 9 7 10 10
Conventional angiogram performed on all 34 patients revealed an aneurysm in 16 (47%) of them, 9 of whom had multiple aneurysms, for a total of 32 aneurysms. The distribution of these aneurysms was as follows: 24 pre-nidals, 5 intra-nidals, and 3 aneurysms unrelated to the AVM. Based on the location of the hemorrhage as seen on the CT-scan it was possible to conclude that the source of the bleeding was the aneurysm in 13 cases (38%) and the AVM for all other cases. Of the eighteen patients who did not have an associated aneurysm, 3 (16%) had a GOS of less than 9, whereas of the 16 patients with an aneurysm, 7 (47%) had a GOS of less than 9. Of the 10 patients who were intubated upon admission, 7 (70%) had an aneurysm responsible for the bleeding. 3.4. Management Because of the life-threatening clinical condition upon admission, 14 patients (42%) required an urgent evacuation of a cerebellar hematoma (Table 3). Drainage of cerebrospinal fluid (CSF) was necessary in 19 patients (56%), including in all those who were comatose. In cases (13) where the source of hemorrhage was a ruptured aneurysm, aneurysms in 12 patients (35%) were embolized and one was clipped while evacuating a hematoma. The mean time between admission and embolization was 16 h (range from 1 h to 39 h). As for the AVMs (Table 4), 23 patients (68%) were amenable to treatment: 8 patients (24%) underwent surgery alone, 3 patients (9%) underwent embolization alone, one patient (3%) received radiosurgery alone, and 11 patients (32%) were treated with a combination of these modalities. Of these 23 treated patients, in 21, treatment resulted in complete exclusion of the AVM. A residual AVM was seen in two patients, one who had been treated by radiosurgery alone, and one who had undergone embolization followed by radiosurgery. In six patients no attempt was made to treat the AVM: in 5 of them the AVMs were located in the brainstem, and 1 patient with a grade II AVM chose not to be treated. Case examples from our series are illustrated in Figs. 1 and 2. 3.5. Complications
10 7 7 9 10
Abbreviations: GCS, glasgow coma scale; SAH, subarachnoid hemorrhage; IVH, intraventricular hemorrhage.
Table 2 AVM features: location and Spetzler-Martin grade. Location/Spetzler-Martin grade
I
II
III
IV/V
CH V CHV Brainstem Total
7 4 0 0 11
9 3 1 1 14
1 1 1 6 9
0
Abbreviations: CH, cerebellar hemispheres; V, vermis; CHV, cerebellar hemisphere and vermis.
In this series, 5 patients (15%) re-bled after initial hemorrhage. In four of them, the re-bleeding occurred soon after admission, before initial treatment, all of them between the second and the third day. For these 4 patients, the reason for re-bleeding was an associated aneurysm in 2 cases and the AVM itself in two others (Table 5). Concerning the 2 AVMs that re-bled, both were located in the cerebellar hemisphere, one with deep venous drainage (grade II) and the other without (grade I). All four patients died as a result of re-bleeding. As for the fifth patient, re-bleeding occurred months after the initial treatment, while waiting for definitive treatment. The reason for delaying definitive treatment was the severity of the patient’s neurological status after the rupture of the AVM. This one patient who re-bled improved gradually (mRS 3). Of the 19 patients who required an external ventricular drain, one bled in the tract while inserting the drain. This patient had a favorable outcome (mRS 2). Two other patients contracted meningitis resulting from the drain, and were successfully treated. Six patients required a ventriculoperitonal shunt (VPS). Two patients experienced a stroke: one
80
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
Table 4 Management of AVM. Grade
n
Surgery alone
Embolization alone
Radiosurgery alone
Combined
Observation
Death
I II III IV/V Total
11 14 9 0 34
5 3 – – 8 (24%)
2 1 – – 3 (9%)
– – 1
2 6 3 – 11 (32%)
0 1 5 – 6 (18%)
2 3 – – 5 (14%)
1 (3%)
Fig. 1. Case illustration of a Spetzler-Martin grade 3 AVM of right cerebellar hemisphere in a 27-year-old patient admitted comatose (glasgow coma scale 3) witch required an urgent evacuation of a cerebellar hematoma. The patient improved and the residual AVM was approached by surgical resection after two stages of endovascular embolization. Final outcome was favorable (mRS 2). (A) Axial head CTscan on admission showing a cerebellar hematoma with compression of the fourth ventricle. (B) Head CTscan after drainage of the hematoma. (C) Sagittal head angioCTscan and (D) and (E) cerebral angiogram in lateral and antero-posterior view, respectively, showing an AVM of the inferior part of the right cerebellar hemisphere, vascularized by the right posterior inferior and anterior inferior cerebellar arteries, and with a deep venous drainage. (F) Cerebral angiogram in lateral view showing complete cure of the AVM after endovascular embolization and surgery.
was asymptomatic and seen on imagery post-embolization, one presented dysmetria due to hemispheric cerebellar stroke after conventional angiogram. Both patients had a favorable outcome. 3.6. Clinical outcome In this series, favorable outcome (mRS 0–2) was achieved in 24/34 patients (71%). Of the 10 patients who were intubated upon admission, 4 had a favorable outcome (mRS 0–2). The follow-up period for survivors (n = 29) in this series ranges from 6 months to 110 months (mean 31 months). There were unfortunately 5 deaths, all of them within the first week. One patient died as a direct result of the severity of the hemorrhage, and the other four, as a direct result of re-bleeding. 4. Discussion PfAVMs are rare and located at or close to critical areas of the brain [1–5]. Unlike supra-tentorial AVMs, which may manifest with epilepsy, pfAVMs do not, thus diagnosis is most often done following a rupture [14]. Because of their location, rupture may lead to a life-threatening situation. In fact, in our series, 53% of patients presented with some level of impaired consciousness, half of whom required intubation. As a direct result of the hemorrhage, cerebellar hematoma was found in 79% of the patients,
intra-ventricular hemorrhage in 53%, and hydrocephaly in 82%. Therefore, it is highly likely that most patients admitted for ruptured pfAVMs, will require urgent surgical intervention. In our series of ruptured pfAVMs, we found a high incidence of associated aneurysms. These associated aneurysms were the cause of bleeding in 38% of our patients. The literature also suggests that pfAVMs are more frequently associated with aneurysms than are supra-tentorial AVMs [2,7,11,15,16,6,17] and that these aneurysms are often the cause of hemorrhage [2,18,19]. In our series, there was a reciprocal relation between the severity of the patients’ clinical condition upon admission and the source of the hemorrhage. In fact, 70% of our intubated patients had an associated aneurysm. Da Costa et al. [11], in their series of pfAVMs observed that the presence of aneurysms was an independent factor of poor mRS score at followup. A recent study by Abla et al. [20] also suggested that posterior fossa location and associated aneurysm were significant predictors of severe outcome. Of the 13 patients in our series who presented with an associated aneurysm responsible for the hemorrhage, 12 of them were embolized urgently. Meisel et al. [21] noted that associated aneurysms should be the first targets of endovascular treatment because they seem to increase the risk of re-bleeding. In our series, of the four patients where early re-bleeding occurred, two of them had an associated aneurysm, which was the cause of both the initial hemorrhage and the re-bleeding. This would lead us to conclude
Table 5 Database of the series. GCS
Type/L
Hemorrhage/Hydro
Drain H/CSF
Aneurysm/type/B/M
W/50 W/28 M/63 W/29 M/52 W/64 W/71 M/29 W/50 M/20 W/50 W/16 M/61 W/19 M/37 M/62 M/58 M/32 W/50 W/58 W/46 W/6 M/27 W/44 W/49 W/59 W/48 M/62 W/15 M/73 M/62 M/68 W/55 M/62
15 14 13 14 15 14 15 15 7 3 13 12 15 13 15 15 15 15 14 13 13 3 3 15 13 15 6 8 10 8 3 14 3 3
I/CH I/CH I/V I/CH II/CH/dp II/V/dp I/CH I/CH III/PM/dp II/V II/CH II/CH III/PCA/dp III/CHV/dp III/P/dp I/V III/M/dp III/M/dp II/CH/dp II/PCA II/CH I/CH II/CH/dp I/V II/CHV III/M III/CH/dp II/CH III/V/dp I/CH II/V/dp II/CH/dp II/CH I/V
SAH SAH/Hematoma SAH/Hematoma SAH/Hematoma SAH/IVH/Hematoma/hydro SAH/Hematoma SAH/Hematoma/hydro Hematoma/Hydro SAH/HIV/hydro Hematoma/Hydro SAH/IVH/Hematoma/hydro Hematoma SAH SAH/IVH/Hematoma/hydro SAH/HIV/hydro SAH/IVH/Hematoma/hydro SAH/HIV/Hydro Hematoma SAH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro HIV/Hematoma/hydro HIV/Hematoma/hydro SAH/HIV/hydro Hematoma SAH SAH/Hematoma/hydro SAH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro SAH/IVH/Hematoma/hydro
No No No No Yes/EVD No No No EVD/VPS Yes/EVD EVD No No Yes/EVD No No/EVD/VPS EVD/VPS No Yes/no No Yes/EVD Yes/EVD Yes/EVD/VPS EVD/VPS No No Yes/EVD Yes/EVD Yes/EVD/VPS No/EVD Yes/EVD Yes/EVD Yes/EVD Yes/EVD
No No No No 1/intra/yes/surgery 2/pre/no/obs No No No 1/pre/yes/embo 4/pre/yes/embo No 1/pre/yes/embo No No 2/pre/yes/embo No No 4/2pre, 2 intra/yes/embo No No 2/pre/yes/embo No 4/3pre, 1un/yes/embo 1/un/no/obs No 2/pre/yes/embo No 1/intra/yes/embo 1/un/no/obs 1/pre/yes/embo No 3/pre/yes/embo 2/pre, intra/yes/embo
Re-bleed
Time H/T
Initial
Definitive
Time I/D
Result
mRS
9 days 2 years 5 months 6 months 12 days
Embo Surgery Surgery Surgery Surgery Obs Surgery Surgery Obs Embo Embo Embo RadioS Embo (2×) Obs Embo Obs Obs Embo Embo Surgery Embo Embo (2×) Embo Surgery Obs Embo Embo Embo
Surgery
10 months
CR CR CR CR CR
0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 2 2 2 3 3 3 3 3 6 6 6 6 6
1 month 20 days 18 months 40 days 2 months 6 months 3 months 1 day
AVM 2 y
45 days 5 months 1 month 9 days 3 months 20 days 2 years 5 months 20 days 15 days
AVM 2 d AVM 3 d Aneur 2 d Aneur 2 d
CR CR Surgery Surgery Surgery
2 days 20 days 3 months
Surgery
11 months
CR CR CR PO CR CO
Surgery RadioS
16 days 1 month
Surgery Surgery
15 days 20 days
Surgery Surgery Surgery
10 days 20 days Same day
CR PO CR CO CR CR CR CR CR CR
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
S/A
Abbreviations: S, sex; W, women; M, men; A; age; GCS, glasgow coma scale; CH cerebellar hemispheres, V vermis, CHV cerebellar hemisphere and vermis, M, mesencphal; P, pons; PCA, pontocerebellar angle; dp, deep venous drainage; SAH subarachnoid hemorrhage, IVH intraventricular hemorrhage, Hydro, hydrocephaly; Drain H, evacuation hematoma; Drain CSF, drainage of cerebrospinal fluid; VPS, ventriculoperitoneal shunt; EVD, external ventricular drain; Aneur; number of associated aneurysms; type, intra as intranidal, pre as prenidal and un as unrelated; B, bleeding related to aneurysm; M, aneurysm management; Re-bleed, as re-bleeding due to and delay since the first bleeding; Time H/T, delay between hemorrhage and initial treatment of the AVM; Initial, modality of initial treatment of the AVM; definitive, modality of definitive treatment of the AVM when multimodality; Time I/D, delay between initial and definitive treatment when multimodality; CR, complete resection; CO complete occlusion; PO partial occlusion; mRS, modified ranking scale at 6 months.
81
82
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
Fig. 2. Case illustration of a vermian and right paravermian Spetzler-Martin grade 2 AVM in a 20-year-old patient admitted comatose (glasgow coma scale 3) and operated on urgently upon admission for drainage of a cerebellar hematoma. The cerebral angiogram showed a pre-nidal aneurysm embolized the day of admission. The patient improved neurologically and the AVM was subsequently approached via endovascular embolization followed by complete microsurgical resection. Final outcome was favorable (mRS 0). (A and B) Axial head CTscan, on admission and after drainage of the hematoma, respectively. (C and D) Initial cerebral angiogram antero-posterior and lateral view, respectively, showing the vermian and right paravermian AVM vascularized by the right superior cerebellar artery and the pre-nidal aneurysm (white arrow) localized on its feeding artery. (E and F) Cerebral angiogram antero-posterior and lateral view, respectively, after endovascular occlusion of the aneurysm (white arrow). (G and H) Cerebral angiogram antero-posterior and lateral view, respectively, after partial endovascular embolization of the AWM. (I and J) Cerebral angiogram antero-posterior and lateral view, respectively, after complete microsurgical resection of the AVM.
that whenever a patient is admitted for a ruptured pfAVM, we must be vigilant to search for an aneurysm on the angiogram, since it is highly likely that it could be responsible for the hemorrhage and that there is an associated high risk of re-bleeding, thus requiring prompt treatment. Most of the time, the associated aneurysms can be successfully treated by endovascular means, thus reducing the urgency for definitive AVM treatment. When there is no associated aneurysm, besides the initial bleeding of an AVM, it is not yet known whether there are other characteristics that could indicate the potential for its early rebleeding. In our series, 2 patients without an associated aneurysm, re-bled early. This seems to be a higher rate compared to what has been found for supra-tentorial AVMs. This may suggest that the posterior fossa location itself may be a risk factor for early rebleeding. Although our series is small, this observation has also been reported by others [18]. Therefore, when there is no associated aneurysm responsible for the initial hemorrhage, it may be appropriate to consider early definitive treatment of ruptured pfAVMs. Even though it may be justified to postpone definitive treatment in cases of non-life-threatening intra-cerebral hematoma or brain swelling, our series suggests that this delay could result in fatal re-bleeding. A larger series may provide us with more substantial evidence. There are several options available in definitive treatment of pfAVMs: surgical resection, embolization, radiation, and a combination of these modalities. Despite these various options, in our series, 32% of patients, including those who died, could not benefit from definitive treatment at all; in most of them, the AVMs were located in the brainstem. There are controversies as to which treatment modality to choose since all treatments pose some degree of risk [6,7,15,16,18,22,23]. Generally, because of the risk of re-bleeding, the goal is to eliminate the AVM with minimal delay. Therefore, in considering the various options, in our series radiotherapy was reserved for those patients where the characteristics of the AVM or the patient’s clinical condition did not allow for safe definitive treatment by either surgery or embolization. As for AVMs in the brainstem, we preferred to abstain from treating them and none of
them re-bled. Other authors have proposed radiotherapy in such cases with or without prior embolization [7,6,24]. In our series, surgery alone was the treatment of choice for grade I patients with small AVMs where dissection is usually straightforward, unless it was thought that embolization alone could offer a potential cure. For grade II and III patients where the AVMs were larger, when possible we embolized first to reduce the flow, and followed with surgery, except for one case, which was followed by radiotherapy. The pre-operative use of embolization is debatable because the possible increased risk due to an added treatment. According to some authors, cerebellar grades I and II should be treated by surgery alone [7,6,25,26]. In fact, in our series, the only complication resulting from treatment occurred following embolization. However, the authors believe that pre-operative embolization can actually add to the success of surgery and decrease the risk of complications. It is not known how many complications due to surgery were prevented by pre-operative embolization. More studies are needed to determine its role. Despite the location of the hemorrhage and the severity of the patients’ clinical condition upon admission, aggressive treatment can result in favorable outcome in most patients. Outcome can be improved by taking measures to avoid re-bleeding and by better identifying which treatment modalities are best suited for which patients, in order to reduce the risk of complications following treatment. There are presently no uniform treatment modalities in the literature. The risks of complications following treatment has been found to be significant in the recent ARUBA study [27]. Although ARUBA study dealt with un-ruptured AVMs, we must keep in mind its results regarding the rate of complications due to treatment since it might also be applicable to ruptured AVMs. Future studies are needed to clarify this point. We recognize that the small size of this series and its retrospective nature pose certain limitations. However, bias selection in greatly reduced since this series represents all patients admitted in our neurosurgical unit during the study period. Although some similar conclusions might have been drawn in various other papers, by no means are they well established. Due to the importance of the potential implications raised in these studies, it
E. Magro et al. / Clinical Neurology and Neurosurgery 128 (2015) 78–83
would advisable to continue documenting data leading to guiding principles in the management of ruptured pfAVMs. 5. Conclusion Patients with a ruptured pfAVM are often comatose upon admission, requiring emergency live-saving surgical treatment. An associated aneurysm is often the source of bleeding which if dealt with immediately, offers time to plan the most appropriate strategies to eliminate the AVM. Nevertheless, early re-bleeding is frequent, and a cause of concern as it often leads to death. More studies are needed to better define which treatment modalities are most appropriate and when. Despite the gravity of the clinical condition upon admission, outcome is favorable for those amenable to treatment. References [1] Batjer H, Samson D. Arteriovenous malformations of the posterior fossa: clinical presentation, diagnostic evaluation and surgical treatment. Neurosurg Rev 1986;9:287–96. [2] Drake CG, Friedman a H, Peerless SJ. Posterior fossa arteriovenous malformations. J Neurosurg 1986;64:1–10. [3] Arnaout OM, Gross BA, Eddleman CS, Bendok BR, Getch CC, Batjer HH. Posterior fossa arteriovenous malformations. Neurosurg Focus 2009;26:E12. [4] Khaw AV, Mohr JP, Sciacca RR, Schumacher HC, Hartmann A, Pile-Spellman J, et al. Association of infratentorial brain arteriovenous malformations with hemorrhage at initial presentation. Stroke 2004;35:660–3. [5] Berman MF, Sciacca RR, Pile-Spellman J, Stapf C, Connolly ES, Mohr JP, et al. The epidemiology of brain arteriovenous malformations. Neurosurgery 2000;47:389–96, discussion 397. [6] Lawton MT, Hamilton MG, Spetzler RF. Multimodality treatment of deep arteriovenous malformations: thalamus, basal ganglia, and brain stem. Neurosurgery 1995;37. [7] Kelly ME, Guzman R, Sinclair J, Bell-Stephens TE, Bower R, Hamilton S, et al. Multimodality treatment of posterior fossa arteriovenous malformations. J Neurosurg 2008;108:1152–61. [8] Bowden G, Kano H, Tonetti D, Niranjan A, Flickinger J, Lunsford LD. Stereotactic radiosurgery for arteriovenous malformations of the cerebellum. J Neurosurg 2014;120:583–90. [9] Hernesniemi JA, Dashti R, Juvela S, Väärt K, Niemelä M, Laakso A. Natural history of brain arteriovenous malformations: a long-term follow-up study of risk of hemorrhage in 238 patients. Neurosurgery 2008;63:823–9, discussion 829–831. [10] Wilkins RH. Natural history of intracranial vascular malformations: a review. Neurosurgery 1985;16:421–30. [11] Da Costa L, Thines L, Dehdashti a R, Wallace MC, Willinsky Ra, Tymianski M, et al. Management and clinical outcome of posterior fossa arteriovenous
[12] [13]
[14]
[15]
[16] [17]
[18] [19]
[20]
[21]
[22]
[23]
[24]
[25]
[26] [27]
83
malformations: report on a single-centre 15-year experience. J Neurol Neurosurg Psychiatry 2009;80:376–9. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986;65:476–83. Redekop G, TerBrugge K, Montanera W, Willinsky R. Arterial aneurysms associated with cerebral arteriovenous malformations: classification, incidence, and risk of hemorrhage. J Neurosurg 1998;89:539–46. Garcia Monaco R, Alvarez H, Goulao A, Pruvost P, Lasjaunias P. Posterior fossa arteriovenous malformations. Angioarchitecture in relation to their hemorrhagic episodes. Neuroradiology 1990;31:471–5. Maruyama K, Koga T, Niranjan A, Kondziolka D, Flickinger JC, Lunsford LD. Radiosurgery for brainstem arteriovenous malformation. Prog Neurol Surg 2013;27:67–72. Matsumura H, Makita Y, Someda K, Kondo a. Arteriovenous malformations in the posterior fossa. J Neurosurg 1977;47:50–6. Kouznetsov E, Weill A, Ghostine J, Gentric J, Raymond J, Roy D. Association between posterior fossa arteriovenous malformations and prenidal aneurysm rupture: potential impact on management. Neurosurg Focus 2014;37: 1–4. Samson D, Batjer H. Arteriovenous malformations of the cerebellar vermis. Neurosurgery 1985;16:341–9. Mpotsaris a, Loehr C, Harati A, Lohmann F, Puchner M, Weber W. Interdisciplinary clinical management of high grade arteriovenous malformations and ruptured flow-related aneurysms in the posterior fossa. Interv Neuroradiol 2010;16:400–8. Abla A, Nelson J, Rutledge C, Young WL, Kim H, Lawton MT. The natural history of AVM hemorrhage in the posterior fossa: comparison of hematoma volumes and neurological outcomes in patients with ruptured infra- and supratentorial AVMs. Neurosurg Focus 2014;37:1–6. Meisel HJ, Mansmann U, Alvarez H, Rodesch G, Brock M, Lasjaunias P. Cerebral arteriovenous malformations and associated aneurysms: analysis of 305 cases from a series of 662 patients. Neurosurgery 2000;46:793–800, discussion 800–2. Rodríguez-Hernández A, Kim H, Pourmohamad T, Young WL, Lawton MT. Cerebellar arteriovenous malformations: anatomic subtypes, surgical results, and increased predictive accuracy of the supplementary grading system. Neurosurgery 2012;71:1111–24. Thines L, Dehdashti AR, da Costa L, Tymianski M, ter Brugge KG, Willinsky Ra, et al. Challenges in the management of ruptured and unruptured brainstem arteriovenous malformations: outcome after conservative, single-modality, or multimodality treatments. Neurosurgery 2012;70:155–61, discussion 161. Massager N, Régis J, Kondziolka D, Njee T, Levivier M. Gamma knife radiosurgery for brainstem arteriovenous malformations: preliminary results. J Neurosurg 2000;93(Suppl. 3):102–3. Sinclair J, Kelly ME, Steinberg GK. Surgical management of posterior fossa arteriovenous malformations. Neurosurgery 2006;58:ONS189–201, discussion ONS201. Viale GL, Pau A, Viale ES. Surgical treatment of arteriovenous malformations of the posterior fossa. Surg Neurol 1979;12:379–84. Mohr JP, Parides MK, Stapf C, Moquete E, Moy CS, Overbey JR, et al. Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. Lancet 2013;6736:1–8.
