Re-exploration of the craniotomy after surgical treatment of unruptured intracranial aneurysms
- First Online:
- Cite this article as:
- Park, W., Ahn, J.S., Park, J.C. et al. Acta Neurochir (2014) 156: 869. doi:10.1007/s00701-014-2059-z
- 208 Views
Unplanned re-exploration of the craniotomy after surgical treatment of unruptured intracranial aneurysms (UIAs) is sometimes required, but the underlying causes and rates of these procedures are seldom reported. This study retrospectively analyzed the causes of such re-explorations to identify methods for decreasing their necessity.
From January 2000 to December 2011, 1,720 patients with a total of 1,938 UIAs underwent surgical treatment at our institution. From this cohort, 26 patients (1.5 %) with 38 UIAs required re-exploration. Clinical data, aneurysm characteristics, treatment methods, and the incidence and causes of re-exploration of the craniotomy were analyzed for these 26 patients.
Several causes of re-exploration were identified: compromised distal blood flow (eight patients, 0.47 %), hemorrhagic venous infarction (four patients, 0.23 %), brain retraction injury (three patients, 0.17 %), newly identified aneurysms (three patients, 0.17 %), bleeding from an incompletely clipped aneurysm (two patients, 0.12 %), epidural hematoma (two patients, 0.12 %), failed aneurysm clipping (two patients, 0.12 %) and other causes (two patients, 0.12 %). Annual re-exploration incidence rates ranged from 0 to 3.1 %. Annual incidence rates gradually decreased following the introduction of several intraoperative monitoring systems.
Precise surgical planning and careful operative techniques can reduce the incidence of unplanned re-exploration of the craniotomy. The introduction of various intraoperative monitoring systems can also contribute to a reduction in this incidence.
KeywordsCraniotomyRe-explorationSurgical treatmentUnruptured intracranial aneurysm
With the development of non-invasive neuroradiological techniques and an increasing frequency of patient requests for neurological screening using brain imaging, unruptured intracranial aneurysms (UIAs) have been detected at an increasing frequency. Recent studies using non-invasive neuroimaging techniques including magnetic resonance imaging (MRI) and computed tomography angiography (CTA) have revealed that 1.8 % to 5 % of the general population has UIAs [1, 20, 51].
The International Study of Unruptured Intracranial Aneurysms (ISUIA) has reported that in patients without a previous history of subarachnoid hemorrhage (SAH), UIAs less than 10 mm in diameter had a rupture rate of 0.05 % per year and UIAs larger than 10 mm in diameter had a rupture rate of less than 1 % per year . In addition, the ISUIA concluded that surgical treatment of UIAs less than 10 mm in diameter without previous history of SAH is inadvisable . However, Morita et al. have reported that an annual rupture rate was 1.69 % for UIAs of 7–9 mm in diameter in a Japanese cohort . Another systematic review of the literature has reported that the rupture rate is approximately 1 % per year for lesions 7–10 mm in diameter . Additionally, it has been reported that 17.8 % of ruptured aneurysms are less than 5 mm  and that 22.5 % and 38 % of ruptured aneurysms are less than 6 mm [35, 43]. Based on these results, conservative management for small UIAs is not always the best treatment option.
Since the advancement of endovascular techniques, many intracranial aneurysms have been treated using these methods. However, Park et al. have reported an angiographic recurrence rate after endovascular treatment of UIAs of 15.3 % and a major recanalization rate of 5.7 % . Ng et al. have reported a complete occlusion rate of the intracranial aneurysm of 46 % and a recanalization rate of 8 % after endovascular treatment . Other studies have also reported recurrences in 14 % and 33.6 % of aneurysm cases after treatment with detachable coils [12, 38]. In contrast, Taha et al. reported a total occlusion rate with surgery of 81.4 % versus 57.5 % with coiling , and David et al. reported a 98.5 % obliteration rate after surgical clipping of intracranial aneurysms .
However, neurosurgeons should be cautious in treating UIAs because the majority of affected patients do not complain of any neurological deficits. Several studies have reported mortality rates for UIA surgery of 0–3.8 % and overall morbidity rates post-surgery ranging from 2.2 to 21.0 % [2, 6, 17, 19, 24, 30, 33, 37, 52]. These mortality and morbidity rates must be reduced to justify the surgical treatment of asymptomatic UIAs. If the surgical risks are lower than the risks of aneurysm rupture, then surgical treatment of a UIA should be considered.
Unplanned re-exploration of the craniotomy after surgical treatment of UIAs is sometimes required but may lead to patient physical and emotional distress regardless of any surgery-associated morbidity and mortality. However, the causes and rates of unplanned re-exploration of the craniotomy after a first surgery for UIAs have seldom reported. In our current study, we retrospectively analyzed the causes and rates of unplanned re-exploration of the craniotomy after an initial surgery for the treatment of UIAs. And we discuss possible ways to decrease the frequency of unplanned re-explorations.
Methods and materials
Location of 1,938 UIAs which were treated by surgery and location of 38 UIAs in 26 patients who experienced re-exploration of the craniotomy
Aneurysms in 26 patients
No. aneurysms (n=1,938)
No. aneurysms (n=38)
The characteristics of all patients in our series who experienced unplanned re-exploration, including properties of the UIAs, kinds of surgical modalities, annual incidence of re-exploration, reasons for re-exploration and patient neurological status during treatment were evaluated. The size of the UIA prior to treatment was determined in each case using transfemoral cerebral angiography (TFCA) or computed tomographic angiography (CTA) and was reported as the maximum diameter. Intraoperative micro-Doppler ultrasound was utilized for all patients who underwent aneurysm surgery. At our institution, indocyanine green (ICG) angiography during aneurysm surgery has been used since 2009, and intraoperative monitoring of somatosensory evoked potentials (SSEP) and motor-evoked potentials (MEP) has been used during aneurysm surgery since 2010. Brain computed tomography (CT) and CTA were performed immediately postoperatively for all patients, with additional imaging modalities including TFCA and/or MRI utilized as needed. Clinical outcomes were analyzed using the modified Rankin score (mRS) at 6 months after surgery based on patient medical records.
Illustrative case 1
Illustrative case 2
Patient population and clinical presentation
Among the 1,720 patients treated at our hospital with 1,938 UIAs identified by retrospective record review, 26 cases (1.5 %) with 38 UIAs met our study inclusion criteria. This patient group consisted of 19 women and seven men with a mean (\( \pm \)SD) age of 55.0 \( \pm \) 8.9 years. Twenty-four of these patients were asymptomatic and the remaining two patients were symptomatic prior to their first surgery. One symptomatic patient presented with a left homonymous hemianopsia due to optic track compression by a giant aneurysm of the basilar tip, whilst the other symptomatic case presented with seizure resulting from the mass effect of a giant aneurysm at the MCA bifurcation. The 24 asymptomatic patients were all found to have UIAs during examination for chronic headache, dizziness, or minor head trauma.
Aneurysm characteristics and treatment methods
The locations of the 38 UIAs requiring re-exploration are summarized in Table 1. The mean lesion size was 8.7 mm (range, 3–50 mm). Seventeen patients had single aneurysms, seven patients had two aneurysms, one patient had three aneurysms, and one patient had four aneurysms. A total of 29 craniotomies were performed for the 26 patients in this cohort. Twenty-four patients underwent a single craniotomy. One patient underwent two craniotomies in a single operation for multiple aneurysms while the other final patient underwent three craniotomies in a single operation. Surgical methodologies included 25 pterional approaches, two interhemispheric approaches, one far lateral approach and one orbitozygomatic approach. During surgery, 32 clippings of the aneurysm, two wrappings of the aneurysm, one trapping of the aneurysm with bypass surgery, and one trapping without bypass were performed. Obliteration of the two thrombosed giant aneurysms failed at the first surgery due to their immense size and the hard consistency of the lesions. The two patients with these thrombosed giant aneurysms, therefore, underwent re-exploration of the craniotomy using another treatment strategy.
Incidence and cause of the re-exploration of the craniotomy
Six (0.35 %) of our patients needed re-exploration of the craniotomy due to hemorrhage. Among these cases, four (0.23 %) patients required re-exploration and hematoma evacuation with or without extended craniectomy to address a large amount of hemorrhagic venous infarction, and the remaining two patients (0.12 %) required re-exploration and hematoma evacuation due to large epidural hematoma (EDH).
Postoperative TFCA revealed newly found aneurysms in three patients (0.17 %) that required re-exploration and clipping. For these three cases, only CTA, but not TFCA, was performed prior to their first surgical procedure. One patient (0.06 %) underwent re-exploration and successful removal of a cottonoid. The other patient (0.06 %) complained of a visual disturbance in their left eye after clipping of a right ophthalmic artery aneurysm and wrapping of a left ophthalmic artery aneurysm. When re-exploration surgery was performed in this case, previously placed wrapping materials that were compressing the left optic nerve were removed and the aneurysm was carefully re-wrapped.
In our current study, we analyzed 26 of the 1,720 patients (1.5 %) who underwent surgery for UIAs at our institution as they required an unplanned re-exploration of the craniotomy after their first surgery. For these 26 patients, compromised arterial blood flow (eight patients, 0.47 %) was the most common cause of unplanned re-exploration. This ischemic complication is one of the most common causes of morbidity and mortality following brain aneurysm surgery [8, 10, 30, 33, 34, 41, 47]. Several intraoperative techniques have been developed to prevent surgery-related ischemic damage associated with aneurysm surgery: direct vascular monitoring using micro-Doppler ultrasound, ICG angiography, intraoperative CT angiography, and intraoperative digital subtraction angiography; and functional monitoring using electroencephalography, electrocorticography, MEP, and SSEP . At our institution, only micro-Doppler ultrasound had been used during aneurysm surgery up to 2008. Micro-Doppler ultrasound is a safe, time-efficient, and cost-efficient technique for detecting parent artery stenosis or occlusion after clipping [5, 13]. However, resulting measurements vary depending on the arterial diameter, vessel wall thickness, probe angle, and insonation angle. Furthermore, removal of the retractor after clipping may lead to brain shifting and subsequent changes in the position of the clip in relation to adjacent vessels [5, 13]. To overcome the shortcomings of micro-Doppler ultrasound, ICG angiography has been used at our institution since 2009, and MEP and SSEP have been used since 2010. MEP is more reliable than SSEP. But, muscle twitching during stimulation for MEP monitoring can interfere with surgical maneuvers. Therefore, MEP monitoring is not suitable for continuous monitoring. We usually continuously monitor SSEP during surgery for intracranial aneurysms and intermittently monitor MEP when SSEP is changed. In addition, even if the SSEP is not changed, we also monitor MEP after manipulation of arteries around aneurysm, every 3 min after temporary clipping of a parent artery, after clipping of the aneurysm and after closure of dura mater and cranial bone flap. These changes have had positive outcomes in that no patients at our institution since 2010 have required an unplanned re-exploration of the craniotomy after surgery for UIAs due to compromised arterial blood flow.
Hemorrhagic venous infarction (four patients, 0.23 %) was the second most common cause of unplanned re-exploration of the craniotomy in our present study cohort. These patients underwent aneurysm clipping using a pterional approach. Venous infarction associated with the pterional approach can be classified into two groups according to the venous anatomy: the superficial Sylvian vein and its branches in the Sylvian fissure, and the bridging veins and their pathways around the ICA . Sacrifice or stretching damage of the superficial Sylvian vein and its branches may cause venous infarction. Among the many bridging veins surrounding the ICA, small branches from the frontal base or temporal tip to the superficial Sylvian vein can typically be safely ligated . However, the first segment of the basal vein of Rosenthal and its variable anastomosis can exist around the ICA, and ligation of these veins may produce venous complications [44–46]. Sylvian fissure dissection and approach of the skull base thus demand the utmost care and accurate anatomical understanding to ensure a safe procedure and positive outcome. In our experience, the veins located on the surgical trajectory towards an aneurysm should be substantially retracted and occasionally must be sacrificed using minimal Sylvian dissection during aneurysm surgery. In addition, it is also difficult to observe aneurysm and arteries around aneurysm from various angles with minimal opening of the Sylvian fissure.
Several orbito-frontal venous tributaries which cross the Sylvian fissure need to be sacrificed during wide opening of the Sylvian fissure on the frontal side of the veins. This sacrifice of the veins could bring about venous infarction of the frontal lobe. Therefore, since 2007 in our hospital, we have used a distal transsylvian approach in order to minimize sacrifice of the bridging veins during a wide opening of the Sylvian fissure as described by Kazumata et al. . We acquire the precise anatomy of the venous drainage system using TFCA before surgery. Skeletonization of the main stem of Sylvian veins is then performed with consideration of variations in the venous anatomy. Although the distal transsylvian approach with skeletonization of the main stem of Sylvian veins is a time consuming procedure, it can preserve the venous structures and allow for easily displacement of the Sylvian veins, its tributaries and the bridging veins surrounding the ICA. Moreover, this procedure provides a wide field of vision.
Brain retraction injury (three patients, 0.17 %) was found to be another common cause of unplanned re-exploration of the craniotomy among our study subjects. Two patients who had a giant aneurysm of the distal ICA and one patient who had a giant aneurysm of the posterior communicating artery underwent surgery to obliterate their aneurysm. During each surgery, the duration of temporary clipping was less than 4 min at a time and distal flow through the clipping site was verified by micro-Doppler ultrasound. Immediate postoperative CTA did not reveal any unusual findings. However, these patients displayed progressive consciousness deterioration and hemiparesis, and brain CT showed progressive brain swelling with infarction at the ipsilateral hemisphere. Retrospective reviews of these operations showed that the surgeons applied brain retractors for a prolonged period with high pressure during treatment of the giant aneurysm. We thus concluded that the brain swelling with infarction that occurred in these patients was caused by brain retraction injury. Brain retraction injury is not infrequent, with several studies reporting its incidence during UIA surgery of 2.5–5 % [4, 30, 39]. Brain retraction injury is caused by the retractor blade’s focal pressure on the brain, which causes direct damage or produces a local deformation of the brain tissue and subsequent development of a reduction or cessation of regional perfusion [54, 55]. Appropriate patient positioning during surgery, drainage of the cerebrospinal fluid, extensive removal of the skull base and wide dissection of the Sylvian fissure are effective ways to minimize brain retraction injury. Sometimes, the addition of mannitol during UIA surgery may be helpful during brain retraction. We have selectively used low dose mannitol in our hospital when the brain is not sufficiently relaxed with drainage of the cerebrospinal fluid. Andrews et al. have previously recommended that brain retraction should be limited to 15 min maximum at a pressure of less than 40 mmHg, with a 5 min recovery period between retractions and the use of multiple, narrow blades that may be less injurious than a single, wide blade . Kaido et al. provided experimental evidence that intermittent release of brain retractor could reduce brain tissue damage caused by brain retraction .
Bleeding from an incompletely clipped aneurysm (two patients, 0.12 %) also led to unplanned re-exploration of the craniotomy. The neurological status was normal in these cases, but CTA revealed small residual sacs immediately after surgery. Several hours later, these two patients experienced consciousness deterioration and CT revealed extensive SAH. Several studies have reported a residual lesion in 3.9–8.2 % of all surgically clipped aneurysms [14, 16, 26, 42, 49]. Thornton et al. reported a bleeding rate from the residual lesion of 0.4 % for all surgically clipped aneurysms , and the other studies have reported 0.5–1.5 % annual bleeding risk from the residual lesion[14, 16]. Small dog ear-shaped residual lesions are the most commonly found [14, 22, 42], but broad-based residual lesions tend to rupture more frequently than the small dog-ear shaped residual lesions [7, 14]. Based on our own experience, there are several factors leading to residual lesion occurrence after surgical aneurysm clipping including insufficient exposure of the aneurysm and its neck, severe atherosclerotic change in the broad aneurysm neck, aneurysm neck in close contact with perforating vessels, and a distal artery or significant perforating artery arising from the neck or dome of the aneurysm. In our current study, one aneurysm had a large diameter of 15 mm that ruptured during surgery. Consequently, the aneurysm neck was insufficiently exposed and a broad-based lesion remained. The other large aneurysm among our study patients had a diameter of 18 mm and a broad atherosclerotic neck. In addition, the anterior choroidal artery originated from the neck. Subsequently, a broad-based lesion also remained. We speculated that relatively weak points of the aneurysms might still remain after clipping and that hemodynamic pressure might congregate at the weak points leading to eventual rupture of the residual lesions.
In our institution, almost all intracranial aneurysm patients have undergone TFCA before surgery. For a very small minority of these patients, CTA alone and not TFCA was performed prior to surgery. The CTA revealed aneurysms and lesions which looked like infundibulums of posterior communicating arteries in three patients. These aneurysms were treated surgically but the infundibulums were ignored. Postoperative TFCA revealed that the lesions which looked like infundibulums before surgery turned out to be aneurysms. Hence, three patients (0.17 %) experienced re-exploration of the craniotomy due to such newly found aneurysms.
TFCA is still considered to be the gold standard in the diagnosis of cerebral aneurysms. However, TFCA can cause serious complications [11, 15, 18] and CTA is easier to perform and less invasive. As CTA techniques have developed, the sensitivity and specificity of this method for detecting cerebral aneurysms have rapidly approached those of TFCA [18, 50, 53]. Matsumoto et al. have reported that 100 consecutive patients with SAH due to a ruptured intracranial aneurysm were successfully treated using CTA only and they concluded that this method could replace TFCA in the diagnosis of ruptured aneurysms . However, CTA still has several noteworthy limitations. First, it does not clearly describe small perforating arteries around an aneurysm [27, 31]. Second, CTA does not provide any dynamic information on the cerebral circulation including collateral flow [9, 27, 31]. Third, small aneurysms less than 2–3 mm can sometimes not be detected on CTA [9, 25, 53]. Finally, the aneurysms near the skull base may not be identified clearly by CTA due to bony structures [3, 27]. For these several reasons, we have used TFCA as a standard method in our hospital for the diagnosis of cerebral aneurysm.
UIAs are being detected with increasing frequency due to improvements in the quality and availability of non-invasive neuroradiological techniques. Based on previously published reports and the collective experience at our institution, precise surgical planning and careful operative techniques can reduce the incidence of unplanned re-exploration of the craniotomy after surgical treatment of UIAs. The use of various intraoperative monitoring techniques can also help to reduce the incidence of unplanned re-exploration of the craniotomy.
Conflict of Interest