Background There is an ongoing debate on the preferred treatment of middle cerebral artery (MCA) aneurysms. The purpose of this study was to assess the clinical and imaging outcomes comparing conventional coiling and clipping of unruptured and ruptured MCA aneurysms.
Methods We searched the electronic databases PubMed, EMBASE, and Cochrane from January 1990 to May 2014.
Results 51 studies were included in the analysis. Favorable outcome was reported in 97.0% and 77.1%, and in 97.2% and 72.8% of patients after coiling and clipping of unruptured and ruptured aneurysms, respectively. Death rates were 1.1% and 8.4% after coiling and 0.3% and 14.7% after clipping of unruptured and ruptured aneurysms, respectively. Initial adequate occlusion was obtained in 89.6% and 92.1% after coiling of unruptured and ruptured aneurysms, respectively. Only three studies on clipping reported on aneurysm occlusion during follow-up.
Conclusions Both coiling and clipping are procedures with low mortality and morbidity rates and, although it may seem that coiling is better for ruptured aneurysms and clipping for unruptured aneurysms, no firm conclusions can be drawn due to the variation in study design and lack of standardized reporting on MCA aneurysm treatments. Standardized observational studies from prospectively kept databases are needed to allow stronger conclusions to be drawn on what is the best treatment for MCA aneurysms. Comparable with aneurysms in other locations, a multidisciplinary approach is therefore recommended with selection of treatment modality based on the clinical condition of the patient and the morphological aspects of the aneurysm.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The International Subarachnoid Aneurysm Trial (ISAT) showed better independent survival at 1 year for patients with coiled ruptured intracranial aneurysms compared with clipping, but middle cerebral artery (MCA) aneurysms were relatively underreported.1 A recent systematic review showed that endovascular treatment of MCA aneurysms is feasible and effective in selected cases, but the periprocedural mortality and morbidity are not negligible, and studies in which MCA aneurysms were described as a subgroup of a total study population were not included.2 Another recent review favoring clipping of unruptured aneurysms also included papers solely reporting on stent assisted coiling or endovascular treatment of complex aneurysms, which are procedures which carry a higher risk of complications.3 It is therefore not surprising that the debate on the preferred treatment modality for MCA aneurysms continues. A worrying side effect of this lingering controversy is that individual patients do not always receive the best possible treatment, but are sometimes operated on only to maintain surgical experience or, in contrast, aneurysms are being stented and coiled when there is no neurosurgical backup.4 ,5
The purpose of this systematic review is to assess the clinical and imaging outcome rates of coiling and clipping of unruptured and ruptured MCA aneurysms during follow-up.
We searched the electronic databases PubMed, EMBASE, and Cochrane from January 1990 to May 2014. The key words and MESH terms ‘intracranial aneurysm*’, ‘mca’, ‘middle cerebral artery aneurysm*’, ‘coil/coils/coiling’, ‘clipping/clip ligation/clip application/clip reapplication’, and ‘endovascular procedures/endovascular embolization/embolization/therapeutic/therapeutic embolization’, were used in relevant combinations. The search was restricted to human studies in English, German, and Dutch. After combining the results, cross referencing was performed to search for additional studies.
Two authors (IJAZ and RvdB) screened all titles and abstracts for eligible studies. Studies describing patients with MCA aneurysms, both as the total patient sample or as a subgroup, were selected based on the following inclusion criteria: the study described more than 10 patients with either coiled or clipped MCA aneurysms; and the study reported at least one of the following items for the group of MCA patients—death, clinical status during follow-up, or follow-up imaging (DSA, MR angiography, or CT angiography) with aneurysm occlusion grade and/or retreatment rates.
Studies in which different aneurysm types were described and analyzed, such as complex, fusiform, giant, or mycotic aneurysms were included only when these aneurysms were subject to regular coil or clip treatment. This meant that series with parent vessel occlusion and more complex techniques, such as bypass surgery or flow diverters, which are usually reserved for more specialized, high volume centers, were excluded. Studies focused on stent assisted or balloon assisted coiling were also not included in this review for the same reason.
Studies without specific MCA data, case reports, book chapters, conference abstracts, comments, letters, reviews, and studies in children (<18 years) with aneurysms were excluded. In studies with overlapping data cohorts, we selected the study with the largest sample size or the most relevant data. Full papers were independently evaluated by two authors (IJAZ and RvdB). In the case of disagreement with respect to eligibility, consensus was reached.
One author (IJAZ) extracted and processed the relevant data according to the PRISMA statement.6 Data on the total number of (patients with) MCA aneurysms, coiled/clipped aneurysms, and ruptured/unruptured aneurysms were extracted. Furthermore, clinical data were extracted including neurological status on admission, death, cause of death, clinical status during follow-up, intraprocedural ruptures, ischemia, post-treatment hemorrhages, and time to clinical follow-up.
Neurological status on hospital admission was extracted if measured by the World Federation of Neurological Society subarachnoid hemorrhage grading scale (WFNS) or the Hunt and Hess score (HH).7 ,8 Ischemia was defined as thromboembolic periprocedural complications, ischemia on imaging, or clinical ischemia. Clinical status during follow-up was extracted if measured by the modified Rankin Scale (mRS) or the Glasgow Outcome Scale (GOS).9 ,10 In concordance with other studies, we defined a favorable clinical outcome as a modified Rankin Scale score of 0–2 or a GOS score of 4–5 (French GOS).11 All other outcomes, including death, were considered unfavorable.
Imaging follow-up data included time to imaging during follow-up, occlusion rate, and retreatment rate. Completely occluded aneurysms and aneurysms with a neck remnant were considered adequately occluded.12 If the data were unclear, the authors were contacted for additional information.
Assessment of quality
The Newcastle–Ottawa quality assessment scale for cohort studies (NOS Scale) and the Cochrane Collaboration's risk of bias assessment tool for randomized trials were used to assess the risk of bias.13 ,14 The assessment was done independently by two authors (IJZ and DV). As there were no studies in this review comparing cohorts, both the items ‘selection of the non-exposed cohort’ and ‘comparability’ of the NOS Scale were not applicable and therefore excluded from the scale for this study. These modifications resulted in a maximum score of 6 points instead of 9. Follow-up was considered adequate when clinical and imaging follow-up was performed in 80% or more of the included patients. Follow-up was considered long enough if it was performed at least 3 months after baseline. In the case of disagreement between authors with respect to risk of bias scores, consensus was reached.
When possible, clinical outcome data were reported in four subgroups: coiled unruptured aneurysms, coiled ruptured aneurysms, clipped unruptured aneurysms, and clipped ruptured aneurysms. Where possible, imaging outcome for coiled aneurysms was divided between ruptured and unruptured aneurysms.
Percentages for the different clinical outcome items were calculated based on the number of patients eligible for the specific item. Percentages for imaging outcome items were calculated based on the number of aneurysms eligible for the specific item. Mean percentages and 95% CIs were calculated for all outcomes.
The search yielded 11 598 studies. After removing 3913 duplicates, 7685 studies remained (Embase 2570, Pubmed 4923, and Cochrane 192— figure 1 in supplementary appendix). After screening titles and/or abstracts, 7449 studies were excluded, leaving 236 full text papers for review. Finally, 51 studies were included in the analysis (tables 1 and 2; references in supplementary appendix).
The most frequent reasons for exclusion were no (clear) description of MCA outcome data and too small sample size.
All included studies, except one randomized controlled trial, were cohort studies. Ten studies were based on prospective data. The risk of bias in the cohort studies was high, with no study receiving the maximum of 6 points on the NOS Scale and seven studies receiving 1 point (table 1).
The follow-up period ranged from time to discharge to 108 months.
Favorable outcome was reported in 97.0% and 97.2% of cases after coiling and clipping, respectively. Death rate was 1.1% in patients treated by coiling and 0.3% in patients treated by clipping (table 3). In the coiled group, this included one death caused by a myocardial infarction 3 months after treatment that was not likely caused by the treatment or hospital stay. Ischemia was the main cause of death in both groups (see table in supplementary appendix).
Ischemia was reported in 10.4% of cases after coiling and in 2.9% of cases after clipping. Post-treatment hemorrhages were reported in 3 of 495 patients (0.6%) after coiling and documented as negative in two clip studies including 152 patients. Clinical outcome after a post-treatment hemorrhage was favorable in one patient, unfavorable in one patient, and unknown in the other. Initial occlusion rates were unknown in these patients.
The follow-up period ranged from time to discharge to 108 months.
Favorable outcome was reported in 77.1% of cases after coiling and in 72.8% after clipping. In the studies reporting on WFNS and HH on admission, 20.9% (168/802) of patients with coiled ruptured aneurysms and 35.3% (114/323) of patients with clipped ruptured aneurysms were in a poor initial neurological condition (WFNS/HH grade 4–5). Death rate was 8.4% and 14.7% in patients treated by coiling and clipping, respectively (table 3). It was not possible to determine the exact cause of death in these patients. Ischemia was reported in 13.1% after coiling and in 16.7% after clipping. Post-treatment hemorrhages were reported in 10 (1.6%) of 618 patients after coiling. Clinical outcome was favorable in 3 of these 10 (30.0%) patients, unfavorable in 3 (30.0%) patients, and unknown in 4 (40%). In 4 (40.0%) of these 10 patients, there was a neck remnant after coiling, in 1 (10.0%) the initial occlusion rate was 100%, and in the remaining 5 (50.0%) the occlusion rate was unknown.
In five studies on clipped ruptured aneurysms, 7 (4.5%) of 157 patients had a post-treatment hemorrhage at follow-up. All other clip studies did not report on this. Clinical outcome was unfavorable in 6 (85.7%) of these 7 patients and unknown in 1 patient. The initial occlusion rates of these aneurysms were unknown.
Imaging outcome in all aneurysms
The follow-up period ranged from time to discharge to 84 months.
Initial adequate occlusion was achieved in 89.6% and 92.1% of cases after coiling of unruptured and ruptured aneurysms, respectively (table 4). Only three studies on clipping (Choi et al, Diaz et al, and Kunert et al) described 98.0%, 100%, and 96.2% adequate aneurysm occlusion, respectively, during follow-up with CT angiography or DSA.
Retreatments in all aneurysms
In 16 studies with 1286 patients with coiled aneurysms, 88 (6.8%) retreatments were reported (table 4). No deaths were reported after retreatment of recanalized aneurysms. Clip studies did not report on retreatments.
This review shows that favorable outcome rates are high, after both coiling and clipping of unruptured and ruptured MCA aneurysms, and that mortality is low. The data in this review should however be interpreted with care as the risk of bias was high for the majority of the studies.
There seems to be an advantage for clipping of unruptured aneurysms with respect to death rate. Favorable outcome rates after coiling and clipping of unruptured aneurysms were comparable to the 95.6% and 95.9% rates reported in the International Study of Unruptured Intracranial Aneurysms (ISUIA) study, whereas death rates for coiling and clipping were lower than the reported death rate for all aneurysm locations in the ISUIA study (3.4% for coiling and 2.7% for clipping, respectively).15 The post-treatment hemorrhage rate after coiling of unruptured aneurysms was comparable with that reported in the Unruptured Cerebral Aneurysm Study (UCAS) study, varying between 0.23% and 1.56% in unruptured MCA aneurysms of 3–9 mm,16 but higher than the per year rupture rate of 0.1% in all untreated unruptured aneurysms <7 mm in the ISUIA study. It is therefore unclear whether treatment of unruptured aneurysms has an advantage over a wait and see policy. A prospective randomized trial to determine the best strategy in unruptured aneurysms comparing coiling and clipping with conservative management seems warranted, although recent endeavors have proved unsuccessful.17
In this review it was not possible to separate periprocedural thromboembolic events from clinically apparent ischemia or ischemic changes reported on imaging studies. Although ischemia is reported more frequently after coiling than after clipping of unruptured aneurysms, this did not affect the favorable outcome rate, possibly because not all reported thromboembolic events led to clinical sequelae. The higher frequency of thromboembolic complications might be explained by an increased likelihood to detect thromboemboli events during the coiling procedure.
Furthermore, follow-up MR imaging, including T2 weighted and three-dimensional time of flight MR angiography images, are more routinely performed after coiling to assess the occlusion rate of the coiled aneurysm, so ischemic changes are more readily detected.
In ruptured aneurysms, there seems to be an advantage of coiling with respect to clinical outcome rate and post-treatment hemorrhage rate. However, the lower death rate after coiling than after clipping is probably related to the fact that neurological status on admission was worse in the clipping group. The reason could be that patients with large intracranial hematomas carrying a worse prognosis are probably more often operated on for the purpose of decompression and/or hematoma evacuation, but we could not confirm this in our review as most studies did not report on this subject. An additional factor related to poor outcome could be that morphologically complex aneurysms are often unsuitable for coiling, and therefore more prone to be clipped, during often technically challenging procedures.
After coiling and clipping of ruptured aneurysms, the favorable outcome rates were comparable with the overall favorable outcome rates in the ISAT study (76.3% and 69.4%, respectively), suggesting that the risk profile and treatment of MCA aneurysms, which are subject to regular coil or clip treatment, does not differ from aneurysms in other locations.
The post-treatment hemorrhage rate in patients with coiled ruptured MCA aneurysms in this review seemed lower than the overall post-treatment hemorrhage rate within 1 year for ruptured aneurysms in the ISAT trial (2.6%) and the Cerebral Aneurysm Rerupture After Treatment (CARAT) study (3.4%) but this might be caused by the more variable time to follow-up.
The post-treatment hemorrhage rate in patients with clipped ruptured MCA aneurysms was higher than the overall rate for clipped ruptured aneurysms in the ISAT trial (1.0%) and the CARAT study (1.3%). The limitation in the ‘coil’ and ‘clip’ data is that not all studies commented on post-treatment hemorrhage. Assuming that the lack of complete reporting on this implies that there were no post-treatment hemorrhages, the percentages may be equal. Again, the lack of standardization in reporting on aneurysm repair hinders solid conclusions.
Intraprocedural aneurysm ruptures were only described in coil studies and were less frequent in unruptured than in ruptured aneurysms. None of the studies on clipping reported on intraoperative ruptures. During clipping, an intraoperative hemorrhage can usually be controlled effectively by temporary clipping of the afferent artery, in combination with removal of the extravasated blood. After stabilization, the aneurysm can be definitively occluded. Possible negative side effects of these hemorrhages are difficult to determine and are therefore probably not registered as a complication.
Follow-up imaging after coiling lacked a strict time interval, which makes calculations on short to mid term follow-up (<1 year) difficult, if not impossible. However, the adequate occlusion rate was high initially and seemed stable during follow-up. This review also showed that follow-up imaging was not routinely performed after clipping. Even in a recently published review including single center, single surgeon data, follow-up imaging was performed in only 22% of patients.18
Limitations of this review are that most studies were retrospective and no study described the neurological condition before admission. This additional information can prove crucial when interpreting the final neurological condition during follow-up. Overall, there was a lack of standardization on reporting without a clear description of patient selection, clinical status on admission, morphological aspects and size of the aneurysms, follow-up period, clinical outcome, and a clear description of complications. For example, the reported clinical and imaging follow-up periods were highly variable. To improve comparison of future studies on aneurysm treatment, standard reporting should be instigated according to recently published guidelines, as this will allow for better comparison of coiling versus clipping of unruptured and ruptured aneurysms, with more solid conclusions on what is the best possible treatment modality.19 ,20 Without these improvements, it will not be possible to draw definitive conclusions from a future systematic review on MCA aneurysm treatment.
Bias plays a major role in treatment selection. Local availability, skill set, and experience of the operator will mainly determine what treatment choice is made for the individual patient. In addition, financial incentives may bias treatment selection. Some advocate a rather rigid clipping strategy for MCA aneurysms because morbidity and mortality are low, and it gives neurosurgeons the opportunity to maintain their clipping skills.4 On the other hand, others treat MCA aneurysms solely endovascularly, including stent assisted coilings, with much higher complication rates.5
The best policy for the individual patient is probably to be treated in a large neurovascular center, with a high level of expertise in both coiling and clipping, that uses a multidisciplinary approach to customize the best possible treatment based on the individual patient's clinical condition and aneurysm morphology.
Both coiling and clipping are procedures with low mortality and morbidity rates, and although it may seem that coiling is preferable for ruptured MCA aneurysms and clipping for unruptured MCA aneurysms, no definitive conclusions can be drawn from this systematic review due to the variation in study design and lack of standardization in reporting on patients’ clinical status and on the morphological aspects of the MCA aneurysms after treatment. Standardized observational studies from prospectively kept databases are needed to allow stronger conclusions to be made on what is the best treatment for MCA aneurysms.
Comparable with aneurysms in other locations, a multidisciplinary approach in large volume neurovascular centers is therefore recommended with selection of the optimal treatment modality based on the clinical condition of the patient and the morphological aspects of the aneurysm.
We thank J Daams of the AMC Medical Library for his contribution to the literature search.
Contributors All authors included fulfill the criteria of authorship. There is no one who fulfills the criteria but has not been included as an author.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.