Published online before print June 13, 2013, doi: 10.3174/ajnr.A3652
AJNR 2013

T.E. Darsaut
University of Alberta Hospital
Mackenzie Health Sciences Centre
Division of Neurosurgery, Department of Surgery
Edmonton, Alberta, Canada

J. Raymond
Centre Hospitalier de l’Université de Montréal
Notre-Dame Hospital
Department of Radiology
Montreal, Quebec, Canada

We would like to air our concerns regarding 2 recent publi­cations in the American Journal of Neuroradiology: a meta­analysis of coil embolization versus clipping for ruptured aneu­rysms¹ and the accompanying editorial entitled “Best Evidence: Comments on Meta-Analysis of Coiling versus Clipping,” by Sellar and White.²

Let us begin with the original research paper. The authors admit conflating the results of 2 randomized trials (the Kuopio study³ and the International Subarachnoid Aneurysm Trial [ISAT]⁴) with a study that used a different design (the Barrow Ruptured Aneurysm Trial [BRAT]⁵). Whether it is valid to “aver­age ” the results of randomized trials with those of a prerandom­ized trial such as BRAT (with a high rate of cross-overs from coiling to clipping) in a meta-analysis remains unclear. Because they chose not to follow the Preferred Reporting Items for Sys­tematic Reviews and Meta-Analyses (PRISMA) guidelines for re­porting meta-analyses,⁶ the scientific appraisal of their report is problematic.

For this to be followed by an editorial entitled “Best Evidence” seems to be overstating the case. Skepticism, the mother of sci­ence, tells us not to uncritically accept those opinions that just happen to correspond with our own. One all-too-common prob­lem with the interpretation of scientific studies is the extrapola­tion of findings to patients who were not included in the original studies. The sweeping application of ISAT results to all aneurysms is a good example of this problem. There can be no support, statistical or other, for such inferences.⁷

The authors of the editorial failed to make important dis­tinctions: the free statement “unequivocal evidence exists that the results of coiling are superior to those of clipping in the treatment of aneurysms” is a sweeping statement, unjustified by evidence.

First, the editorialists failed to distinguish ruptured from unruptured aneurysms, whereas there are many reasons to sus­pect that treatment of the 2 conditions may differ. Surgical clipping of unruptured intracranial aneurysms may very well be superior to coiling, for all we know, at the time of writing this comment. The best treatment of unruptured aneurysms remains to be seen, and a trial addressing the question has been launched.⁸

Second, Lanzino et al¹ mentioned, but the authors of the edi­torial failed to emphasize, that the Kuopio and ISAT populations were only a small portion of all patients with ruptured aneurysms. This does not mean these trials were biased, as is often claimed in the surgical literature,⁹ only that it affects the generalizability of the study results to future patients. In the BRAT, a study initially designed to palliate this problem and to include all aneurysms, a significant proportion of patients allocated to coiling were crossed over to clipping. If the “intent to coil ” BRAT group results were in line with ISAT results, almost 40% of patients of that group were deemed to be better treated surgically, for reasons of “clinical judgment” that remain difficult to specify. It is important to em­phasize that we do not know for sure that clipping was best for the patients who were crossed over. These patients were included in the analysis, but no one knows for sure what the best treatment for these cases was, or is.

One important point is that ISAT, a trial interrupted more than 10 years ago, weighs heavily in the overall results of the meta­analysis. Who were these patients in ISAT? They were good-grade patients bearing small anterior circulation ruptured aneurysms with anatomic characteristics favorable for coiling. The small amount of proof we have won in our field is supported by a single large, properly conducted trial, which showed that for patients with aneurysmal subarachnoid hemorrhage for whom coiling and clipping were considered good options 10–18 years ago, random allocation to coiling led to better outcomes compared with surgery.

Many patients are still offered clipping despite ISAT results, but there are many more patients who today are only offered coiling, though they would not have been included in the original ISAT study. Admittedly, and, as mentioned by Sellar and White,² technical developments now permit endovascular treatment for aneurysms that decades ago would not have been considered for coiling. Again, no one knows for sure if these patients would not be better served with a surgical approach. However, rather than threatening clinicians who consider clipping for some cases to be at risk of litigation or suggesting that the expert opinion of a coiler could protect them from such litigation, we propose that we have yet to complete the task of gathering the evidence for all patients not included in the original ISAT study. For patients with rup-tured aneurysms for whom we still do not know what is best between clipping and coiling, optimal treatment is to offer a trial comparing the 2 treatment options. For those who properly re-main uncertain, ISAT-II is underway.¹⁰

References

1. Lanzino G, Murad MH, d’Urso PI, et al. Coil embolization versus clipping for ruptured intracranial aneurysms: a meta-analysis of prospective controlled published studies. AJNR Am J Neuroradiol 2013 Apr. 11. [Epub ahead of print]
2. Sellar RJ, White P. Best evidence: comments on meta-analysis of coiling versus clipping. AJNR Am J Neuroradiol 2013 Apr 4. [Epub ahead of print]
3. Koivisto T, Vanninen R, Hurskainen H, et al. Outcomes of early endovascular versus surgical treatment of ruptured cerebral aneurysms: a prospective randomized study. Stroke 2000;31:2369–77
4. Molyneux A, Kerr R, Stratton I, et al. International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 2002;360:1267–74
5. McDougall CG, Spetzler RF, Zabramski JM, et al. The Barrow Ruptured Aneurysm Trial. J Neurosurg 2012;116:135–44
6. Moher D, Liberati A, Tetzlaff J, et al, the PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: the PRISMA Statement. BMJ 2009;339:b2535
7. Ludbrook J. Analysis of 2×2 tables of frequencies: matching test to experimental design. Int J Epidemiol 2008;37:1430–35
8. Darsaut TE, Findlay JM, Raymond J, CURES Collaborative Group. The design of the Canadian UnRuptured Endovascular versus Surgery (CURES) trial. Can J Neurol Sci 2011;38:236–41
9. Raymond J, Kotowski M, Darsaut TE, et al. Ruptured aneurysms and the International Subarachnoid Aneurysm Trial (ISAT): what is known and what remains to be questioned. Neurochirurgie 2012;58:103–08
10. Darsaut TE, Jack A, Kerr RS, et al, for the ISAT-II Collaborative Group. International subarachnoid aneurysm trial – ISAT Part II: study protocol for a randomized controlled trial. Trials 2013;14:156

Reply

Published online before print June 13, 2013, doi: 10.3174/ajnr.A3663
AJNR 2013

G. Lanzino^a
aDepartment of Neurosurgery

A.A. Rabinstein^b
bDepartment of Neurology
Mayo Clinic
Rochester, Minnesota

We thank Drs Dursault and Raymond for their interest in our meta-analysis of prospective studies comparing clipping with coiling in patients with ruptured aneurysms. The goals of treatment for ruptured aneurysms are the following: 1) to protect the patient from aneurysm rebleeding, and 2) to minimize com­plications while achieving goal 1. Based on our interpretation of the literature, there is firm evidence that for a given aneurysm amenable to coiling, endovascular treatment is associated with better functional outcomes.

We agree that the results of the International Subarachnoid Aneurysm Trial (ISAT)¹ weigh heavily on the finding of our anal­ysis, but we also found it intriguing that as shown by the Table accompanying our study, the results of the 3 different trials ana­lyzed are strikingly similar, despite methodologic differences. This finding argues that the positive association between coiling and better outcomes is real.

The Barrow Ruptured Aneurysm Trial (BRAT),² a reaction to the results of the ISAT, reached the same overall conclusions. This outcome is remarkable because BRAT has a strong bias in favor of surgery. The design of BRAT allowed the assignment of aneu­rysms that were not ideal for endovascular treatment to be, none­theless, assigned to coiling, which penalized endovascular treat­ment. Nevertheless, the results of BRAT show, in agreement with ISAT, that functional outcome at 1 year was better after coiling rather than after surgical treatment, both in the intent-to-treat and in the as-treated analyses.²

The argument that ISAT is not representative of the overall population of patients with ruptured aneurysms is true from a pure methodologic perspective but weak from a pragmatic and clinical point of view. It is true that more than 90% of patients randomized in ISAT were patients in good clinical grade with small anterior circulation aneurysms. However, patients with poor clinical grade and posterior circulation aneurysms were al­ready being preferentially treated with coiling even beyond ISAT. On the other hand, most small MCA aneurysms can be still more effectively treated with an open surgical approach.

The main issue of treatment of ruptured intracranial aneu­rysms is long-term durability, not in terms of percentage of aneurysm occlusion but in terms of true risk of rerupture during long-term follow-up. This issue is still open, but so far the documented long-term risk of rerupture has been low and does not seem to negate the initial benefit of coiling, except in the very young.

More than a decade after ISAT, it is time to move beyond the issue of coiling versus clipping.³ The 2 technologies are comple­mentary, but when feasible, endovascular treatment is associated with better outcomes. Unfortunately, there continues to be wide variability in the percentage of patients treated with coiling or surgery, and this variability is not related to scientific evidence but too often to personal egos, turf battles, and finances. For years, we have heard about absurd situations where, in “reputable” institu­tions, treatment of the ruptured aneurysm is based on consider­ations that have nothing to do with the severity of the clinical presentation or the characteristics of the aneurysm, but rather with personal preferences, convenience, or frank bias.

The availability of 2 valid therapeutic options and improve­ment in neurocritical care has dramatically ameliorated the prog­nosis of most patients with aneurysmal SAH. In our unit, we ex­pect every patient with grades I-IV aneurysmal SAH (except those with large intraparenchymal hematomas and those who had an unwitnessed SAH with prolonged loss of consciousness) to return to a normal and productive life. If this outcome is not the case, it is often because of mistakes made along the way and not because of the SAH itself, which we have conveniently blamed for less optimal outcomes in the past.

References

1. Molyneux A, Kerr R, Stratton I, et al. International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovas¬cular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 2002;360:1267–74
2. McDougall CG, Spetzler RF, Zabramski JM, et al. The Barrow Rup¬tured Aneurysm Trial. J Neurosurg 2012;116:135–44
3. Lanzino G. The Barrow Ruptured Aneurysm Trial. J Neurosurg 2012;116:133–34, discussion 134