Impact of Postoperative Infarcts in Determining Outcome after Clipping of Anterior Communicating Artery Aneurysms
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.279675
Source of Support: None, Conflict of Interest: None
Keywords: Aneurysm, anterior communicating artery, cerebral infarcts, clipping, glasgow outcome scale, outcome, vasospasmKey Messages: The battle with ruptured aneurysm does not end with its clipping. The presence of symptomatic infarct is the sole important predictor of outcome in the patient after clipping of the aneurysm. Apart from traditional factors, an intraoperative rupture of an aneurysm is a major determinant of development of infarct and elective temporary clipping is always recommended.
Anterior communication (ACOM) artery aneurysms are the most commonly encountered aneurysms in neurosurgical practice accounting for quarter to one-third of all micro-surgically treated aneurysms. ACOM aneurysms are more prone to iatrogenic injuries than aneurysms at other locations because of the complex angioarchitecture of ACOM aneurysm and its numerous anatomic variants. Moreover, this complex is adjacent to numerous vital structures (such as hypothalamus and optic chiasm) supplied by many perforators in this area.
In India, surgical management of aneurysmal SAH was initially described by Prof. B. Ramamurthi, Dr. KV Mathai, Dr, Jacob Chandy, and Dr. PM Dalai in 1965.,, Since then, many new advanced instruments and techniques have emerged. From clipping of small saccular aneurysms to clipping of giant aneurysms, new techniques are being developed to improve patient outcome., Despite improvement in surgical and medical treatments, the outcome following subarachnoid hemorrhage (SAH) has shown only modest improvement. The outcome is principally determined by the severity of initial bleed, early rebleed, and delayed cerebral ischemia commonly attributed to vasospasm.,
One of the major determinants of outcome is postoperative ischemic complications attributable to vasospasm or vascular injury inflicted during surgery. Although many studies have documented ischemic complications, detailed studies regarding pattern incidence and their probable etiopathogenesis are limited in literature. The aim of this study was to assess radiological patterns of postoperative cerebral infarcts, their incidence, associated factors, clinical correlations, and outcome in patients who underwent clipping of ACOM artery aneurysms.
This study includes 118 patients from all age groups (range 15–75 years), who underwent clipping of ruptured ACOM artery aneurysm between July 2009 and June 2010 in the Department of Neurosurgery, Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh. Patients with unruptured ACOM artery aneurysm, aneurysms at other locations, multiple aneurysms, preoperative radiologic infarcts [on plain computed tomography (CT)], and ACOM aneurysms treated with other modality (e.g., endovascular intervention) were excluded. Demographic data such as name, age, sex, residence, presenting symptoms, and neurological status [World Federation of Neurologic Surgeon (WFNS) grade] were recorded.
All patients were managed according to the standard protocol of the institution. The diagnosis was established based on plain CT, CT angiogram (CTA) and/or digital subtraction angiogram (DSA). Diagnostic DSA was performed in patients with inconclusive findings on CTA or patients with doubtful angioarchitecture. All patients underwent surgical clipping and intraoperative events such as temporary clipping (TC), duration of TC, rescue versus elective TC, and intraoperative rupture (IOR) were recorded. TC was considered rescue if temporary clip was applied on the proximal vessel after IOR of the aneurysm, otherwise elective. All intraoperative findings (angioarchitecture, exclusion of perforators, brain edema) and events were noted by experienced operating neurosurgeons. At the time of this study, indocyanine green angiography (ICG) was not practiced in the institute. Postoperatively, neurological status was noted and plain CT of head was done within 24 h. The follow-up CT scan of head was performed on the 3rd to 5th postsurgery day, whereas the third scan was done at the time of discharge and/or clinical deterioration. All patients with suspected vasospasm were primarily evaluated with plain CT of head (to rule out any infarct). Patients with no infarcts but with clinical suspicion of vasospasm were further evaluated with DSA. Patients with radiological vasospasm received intraarterial nimodipine in the same setting. Transcranial Doppler (TCD) was used to detect and monitor vasospasm. Findings of TCD were not evaluated in this study in view of subjective nature of the investigation. Patients were followed up to three months and outcome was assessed as per Glasgow outcome scale (GOS).
In the postoperative period, patients were routinely monitored and managed in the neurosurgery intensive care unit (ICU). Blood pressure (BP) was maintained above baseline (mean arterial pressure about 20 mmHg above baseline). Triple H therapy was administered to patients with clinical and/or radiological evidence of vasospasm (on DSA). Radiologic and/or refractory vasospasm was treated with intraarterial nimodipine. Patients developing severe vasospasm and symptomatic infarcts with mass effect were managed with decompressive craniectomy.
The statistical analysis was carried out using Statistical Package for Social Sciences (SPSS Inc., Chicago, IL, USA; version 15.0 for Windows). For all quantitative variables, data were checked for skewness. For all quantitative variables, mean, median, and standard deviation were calculated. Qualitative or categorical variables were described as frequencies and proportions. Proportions were compared using Chi-square or Fisher's exact test whichever was applicable. All statistical tests were two-sided and performed at a significance level of α = 0.05.
The median age of our patients was 48 years (range 15–75 years); 76 patients were males and 42 females. Overall, 65 (55.08%) patients presented in WFNS Grade I. After resuscitation, 70 (59.32%) patients improved to Grade I. One patient from WFNS Grade I deteriorated to Grade IV, whereas five patients from Grade II and one patient from Grade IV improved to Grade I. No patient with WFNS Grade V underwent surgery. A total of 82 (69.50%) patients had Fisher Grade III SAH on admission CT head. The mean interval between ictus and admission was 3.32 (1–24) days and mean time of clipping after ictus was 4.47 (0.5–21) days. The delay was attributed to late referral.
Headache was the most common presenting symptom (99.1%). Approximately 65 (55.08%) patients presented with history of loss of consciousness, 15 (12.71%) with seizures, and 15 (12.71%) with focal neurological deficit. Around 29 (24.57%) patients had IOR and 67 (56.77%) patients needed TC (mean duration 5.46 min). Nine of 29 (31.3%) patients developed infarcts in multiple vascular territories. These were most commonly located in anterior cerebral artery (ACA) territory (n = 11, unilateral = 5, bilateral = 6). Five patients had isolated infarcts in middle cerebral artery (MCA) territory. All nine patients with multiple-vessel involvement had infarct in MCA territory. One patient had exclusive posterior cerebral artery (PCA) territory infarct, whereas three other with multiple-vessel involvement also had infarcts in PCA region. Six patients had infarcts in areas supplied by deep perforators (three exclusive deep perforator infarcts and three others deep with other vascular territories) [Table 1].
Factors affecting development of infarcts
Neurological status (WFNS) grade before surgery had significant (P = 0.012) impact on the development of postoperative infarct with rising trend in the percentage of infarcts with worsening WFNS grade. Patients with prior history of seizures (either at the time of ictus or in hospital) had higher chances of developing postoperative infarcts than those without seizures (46.66% vs. 21.35%). This difference was borderline significant (P = 0.051). Patients with preoperative focal deficits had higher (P = 0.010) chances of developing postoperative infarcts than those without neurological deficits (53.33% vs. 20.38%). Hypertension was the most common comorbidity (noticed on historical basis) but there was no significant difference on development of postoperative infarcts in such patients. Seven of 24 (29.6%) patients developed postoperative infarcts when compared with 22 of 94 (23.46%) patients (P = 0.793). About 13 patients (11.01%) had intracerebral hematoma (ICH) on admission CT scan. Four (30.76%) patients who had ICH on admission CT scan developed postoperative infarcts when compared with 25 of 105 (23.80%) patients of the cohort without ICH (P = 0.733) [Table 2]. ICH or retraction injuries were not considered as infarcts. Infarcts developing only in a vascular territory were included in the analysis.
The majority of the patients presented with Fisher Grade III (69.51%) SAH. Although patients with Grade IV SAH had higher incidence of infarcts, this difference was not statistically significant. Around 37.93% patients with IOR developed postoperative infarcts when compared with 20.22% of those without IOR (P = 0.054). The development of postoperative infarcts was irrespective of TC duration of more or less than 5 min (P = 0.322). We could not find any significant difference in incidence of postoperative infarcts in patients who underwent elective or rescue clipping (P = 0.439). TC was applied once in 54 of 67 (83.07%) patients and more than once (highest up to six times) in 13 of 67 (16.93%) patients. About 12 (22.64%) patients, in whom TC was applied once, developed postoperative infarcts when compared with 5 (35.7%) patients in whom TC was applied more than once (P = 0.138) [Table 2].
On multivariate analysis of the factors studied for development of infarcts, none of the variable could reach the significant level, but infarcts were approximately 3.4 times more common in patients with focal neurological deficit and those with seizures. IOR was associated with 2.5 times higher chances of development of infarcts. The timing of surgery (time interval between ictus and surgery) did not have any impact on development of infarcts.
In all, 29 of 118 (24.57%) patients developed postoperative radiological infarcts. On 3-month follow-up, 17 (58.62%) patients with infarcts died, which was significantly higher (P = 0.0001) in comparison to 13 (14.60%) deaths in patients without infarct (n = 89). On analyzing the outcome according to the location of infarct, outcome was fairly better in unilateral ACA territory infarct. Only one of five (20%) of unilateral ACA territory patients died, whereas four of five (80%) had good outcome (living independently with some motor deficits). Among patients with bilateral ACA territory infarcts, two of six (33.33%) died. Out of four (66.67%) surviving patients, two patients (33.33%) are living independently. Among patients with exclusive deep perforator infracts, four of six (66.67%) are surviving and living independently. All patients with infarcts in multiple (n = 9) and PCA territory died. This difference was statistically significant (P< 0.006) on univariate analysis [Table 3].
Influence of age and sex on outcome
We did not find any statistical correlation between age, sex, and postoperative outcome at 3 months of follow-up interval using GOS scale (P = 0.418 for age; P = 0.368 for sex) [Figure 1] and [Figure 2].
Timing of infarct and GOS at discharge and 3 months of follow-up
About 17 of 29 (58.62%) patients developed infarcts between 2 and 5 days after surgery and 8 of 17 patients (47.05%) died in same hospital admission. Time of development of infarct had not affected outcome significantly. Five of 9 (17.25%) patients developed infarct within 24 h of surgery and three of five (60%) of these died in the same hospital admission. Time of infarct has not affected outcome significantly (P = 0.715) neither at discharge [Figure 3]a nor at 3 months of follow-up (P = 0.430) [Figure 3]b.
Development of infarcts versus GOS at discharge and 3 months of follow-up
Around 23 of 118 (19.5%) patients died in the same hospital admission, and 14 of 23 (60.86%) patients had postoperative infarcts. The remaining nine (39.14%) patients did not have infarcts but died because of other reasons (four patients died within 1–2 days after surgery). In all, 74 of 118 (62.72%) patients were discharged in GOS 5. Around 14 of 118 (11.86%) patients were discharged in dependent status (GOS 2 and 3). Development of infarct is highly significant for outcome (P < 0.0001) [Figure 4]a.
At 3 months of follow-up, seven more patients died. Hence, 30 of 118 (25.4%) patients of study population expired. Three of seven patients developed postoperative infarct and were discharged in fully dependent status (GOS 2 and 3). Four of seven patients had no radiological infarcts and two of them were discharged in GOS 5. At 3 months of follow-up, 17 of 29 (58.62%) patients with infarcts died, whereas 13 of 89 (14.60%) patients without infarcts died. Only 2 of 29 (6.9%) patients with postoperative infarcts were living independently at 3 months of follow-up when compared with 75 of 89 (84.26%) patients without infarct [Figure 4]b.
Even though aneurysmal SAH is one of the most fatal intra-cranial pathologies, its natural history is not well understood. We intervene to prevent fatal complication such as rebleed and vasospasm. Many complications are described in management of aneurysmal SAH. [Figure 5] shows some common complications.
The outcome following aneurysmal SAH is primarily determined by rebleeding and vasospasm leading to delayed cerebral ischemia. Vasospasm and postoperative infarcts are the two major determinants of the functional outcome following aneurysm clipping., A few studies have reported incidence of cerebral infarcts following SAH, but detailed analysis of the location of infarcts, factors responsible for the development of infarcts, and outcome has not been performed.
Mortimer et al. studied the postoperative injury or infarction patterns in anterior communicating artery aneurysm patients. They found basal forebrain and basal ganglia as the common regions involving infarcts. Common patterns in his study are tabulated below in [Table 4].
In our study, incidence of infarcts following clipping of ruptured ACOM artery aneurysm is 24.57%, which is slightly lower than the previous reported studies. A high rate (40%–60%) of hypodense lesion consistent with cerebral infarction was evident on the follow-up CT scans among survivors at either 3 months or 1 year after aneurysmal SAH., Juvela et al. have also reported a high incidence (65%) of ischemic lesions in ruptured aneursymal patients on follow-up scan at 3 months. Rabinstein et al. reported 24%–35% incidence of cerebral infarction caused by vasospasm when defined by CT, and may be as high as 81% when magnetic resonance imaging (MRI) was used for diagnosis. These studies involve all cases of SAH. Diagnosis of infarction on CT scan rather than on MRI might be responsible for lower incidence in our series in comparison to other published literature. Although useful, getting MRI as a protocol in sick patients is technically demanding.
Hijdra et al. in their study found clinical delayed cerebral ischemia in 56 cases out of their 176 cases recruited for study (56/176 patients; 31.8%). Rabinstein et al. in their study of patterns of brain infarcts following aneurysmal SAH noted radiological cerebral infarctions in 56 patients (39% of the study population). Patients with single infarctions frequently had ischemia in the territory of the ruptured aneurysm (22/28 patients; 79%). Wani et al. studied the role of preoperative diffusion-weighted imaging (DWI) or diffusion-weighted MRI in patients of anterior communicating artery aneurysm. They found 50% (8 of 16) cases with postoperative infarcts. In their study, among the patients who had developed postoperative infarcts, 87.5% patients showed the presence of preoperative DWI abnormalities. Umredkar et al. in their study of cerebral infarcts following aneurysmal SAH noted infarctions in 69 patients (39.65% of the study population). Various patterns of infarcts have been tabulated from literature in [Table 5].
We compared our results with Umredkar et al.'s study and found that ACA infarcts are more common following anterior communicating artery aneurysm clipping [Table 6]. Fisher grade did not correlate with the development of cerebral infarction in our study, in either univariate or multivariate analysis. A few other studies have also questioned the predictive value of Fisher grade.,
On univariate analysis, WFNS grade was a significant risk factor for functional outcome and development of infarcts. There was 1.5 times greater risk of infarct with poor WFNS grade. In accordance to our findings, previous studies have also reported the apparent trend of developing more infarcts in poor grade patients. However, the difference was not statistically significant.,
About 46.66% patients with history of seizures were 3.4 times more likely to development of postoperative infarcts. Although data in previous studies are limited, these findings are consistent with previous analyses, which also suggest that seizures at the onset of ictus are independent predictors of outcome, probably because of ischemic insult to brain. About 53.33% of patients with preoperative focal deficits developed postoperative infarcts, which were significantly higher when compared with patients without deficits. Previous studies have indicated that focal neurological deficits are mainly because of vasospasm leading to decreased blood flow in the corresponding area. Hence, it can be assumed that these patients already had vasospasm or cerebral ischemia, which was not evident on admission CT scan. For such patients, MRI might prove to be a better investigation for realistic evaluation and prognostication.
Our study indicates that IOR is a weakly significant (P = 0.054) factor for development of postoperative infarcts. Sandalcioglu et al. did not find any statistically significant impact of IOR on patients' outcome, but they observed a trend of increasing morbidity and mortality when IOR occurred. We observed that patients with IOR were 2.5 times more likely to developing postoperative infarcts and poor outcome.
We found no significant impact of TC in development of infarcts. However, when TC was used for a longer duration (>5 min), infarcts were seen more frequently (37.55%) in comparison to 21.56% patients with TC for shorter time (<5 min); this different could not reach to statistically significant level. Similarly, infarcts were more frequent (38.46%) when TC was applied more than once, in comparison to patient in whom TC was applied only once. There was no significant difference in between incidence of infarcts in patients with rescue clipping and elective clipping. Previous studies conclude that 70% patients develop postoperative infarcts on first postoperative day when TC duration exceeds 10 min duration, but after first day TC did not affect occurrence of infarct.,
We observed development of infarcts in two different time windows: the early-onset infarcts (<48 h) and delayed (after 48 h) of surgery. In our study, out of all patients showing postoperative infarcts, 17.25% patients developed infarcts within 48 h and 80% of these happened in ACA territory. These infarcts were attributed to direct vessel injury during surgery, permanent/temporary occlusion of artery, and retraction injury. Such infarcts were mostly single, deep, or cortical in distribution and were located in distribution of ACA territory (in the vessel bearing aneurysm). About 82.75% of patients had delayed onset infarcts, which were mostly diffuse, patchy, and etiological because of diffuse process such as vasospasm. The difference in mortality was not significant for timing of postoperative infarct, but location of infarcts significantly affected the outcome. Unilateral or bilateral ACA involvement was either because of direct vessel injury during surgery or because of localized vasospasm. Patients with only deep perforator infarcts had shown better outcome. Isolated vasospasm of the perforating vessels has been reported. Patients with multiple infarcts had 100% mortality at 3-month follow-up. Patient with late infarcts secondary to vasospasm had worse outcome. Previous studies have also reported that late onset of infarcts had higher mortality because of vasospasm., However, the exact details of patterns of infarcts in ACOM artery aneurysm and their etiology have not been studied separately in any previous study.
The mechanism of these infarcts seems to differ from well-defined vasospasm. Moreover, preliminary data from human studies indicate that auto-regulatory responses get impaired after SAH, and microcirculatory changes manifested by prolonged cerebral circulation time may lead to decrease in regional cerebral blood flow and/or microembolism. Autopsy studies performed in patients after SAH reveal that the majority of infarcts are widespread, scattered small wedge-shaped, or laminar infarcts. This pattern is more consistent with small thrombo-emboli than with large arterial spasm. Stoltenberg-Didinger, who studied 156 aneurysmal SAH patients (who died without surgery), found that 76% had evidence of small infarcts; fewer than 6% had large territorial infarcts. Small infarcts were often located immediately beneath the thickest subarachnoid clot. These findings prompted Neil-Dwyer et al. to suggest diffuse micro-angiopathy as the major cause of delayed ischemic neurologic deficits (DINDs) after aneurysmal SAH.
Development of postoperative infarct is the sole important factor for outcome, which was found significant in univariate and multivariate analyses. Out of the total 30 deaths, 58.62% patients died because of infarcts. Only 6.88% patients with infarcts improved to GOS 5 in comparison to 84.26% patients without any infarct. Overall, mortality was 25.42%. These findings are consistent with previous studies in which mortality rate has varied between 15% and 45%., The factors affecting outcome in patients with aneurysmal SAH with cerebral infarcts are reviewed in [Table 7].
Limitations of the study
We purposely excluded patients with preoperative infarcts because the highlight of this study was to evaluate the correlation between surgical factors (such as TC and IOR) and the chances of development of infarcts. For the very same reason, we did not correlate the findings in the comparable population treated with endovascular coiling. This internal exclusion may erroneously attribute vasospasm-induced infarcts to surgical handling. At the time of this study, ICG was not done and hence intraoperative findings are based on operating surgeon's notes, which may suffer from subjective bias. MRI rather than CT might be helpful in early detection of vascular insults and can affect the management protocol.
The battle of a ruptured aneurysm does not end with its clipping. Multiple factors, that is, WFNS grade, presence or absence of focal deficits at presentation, history of seizures at the time of ictus, TC, and IOR are responsible for development of cerebral infarction following clipping of ACOM artery aneurysm. Most of these infarcts are because of vasospasm or direct injuries. Since IOR is one of the factors contributing to postoperative infarcts, we recommend elective TC to prevent IOR. There is high incidence of infarcts after surgical clipping of ACOM artery aneurysm. This is probably because of angioarchitecture variation and complexity of the anatomy in this region. Infarcts in the aneurysmal territory were more common than infarcts in remote territories because of more chances of local injury to the parent vessel or perforator during surgery. The presence of symptomatic infarct is the sole important predictor of outcome in the patient after clipping of the aneurysm. Patients with infarcts secondary to vasospasm perform poor when compared with single-vessel territory infarcts. After surgery, prevention of vasospasm is recommended to reduce the incidence of vasospasm-related infarcts to improve the outcome.,,,
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7]