Neurology India Medknow Publications on behalf of the Neurological Society of India ISSN: 0028-3886 EISSN: 1998-4022 Vol. 52, Num. 1, 2004, pp. 32-35

Neurology India, Vol. 52, No. 1, January-March, 2004, pp. 32-35

Review Article

Critical appraisal of the international subarachnoid aneurysm trial (ISAT)

Sade B, Mohr G

Division of Neurosurgery, Sir Mortimer B. Davis-Jewish General Hospital, McGill University, Montreal, Quebec
Correspondence Address:Division of Neurosurgery, Sir Mortimer B. Davis- Jewish General Hospital, 3755 Chemin de la Côte Ste. Catherine, Montréal, H3T 1E2, Québec gmohr@neu.jgh.mcgill.ca

Code Number: ni04010

Abstract

Introduction

International Subarachnoid Aneurysm Trial (ISAT)[1] was designed as the first international prospective randomized trial aiming to compare the safety and efficacy of endovascular coiling with microsurgical clipping.
The first prospective randomized trial in this context was published by Koivisto et al.[2] Their single center study of 109 patients showed no difference in the clinical and neuropsychological outcomes of both treatment modalities at one year.

Impact of ISAT

The results of ISAT caught significant attention from both scientific and lay press. This single study impacted the management of patients with subarachnoid hemorrhage significantly all over the world and particularly in United Kingdom (UK), the major contributor of the study. The proportion of patients undergoing endovascular treatment increased from 34% to 54% in UK after this publication.[3] This increase was even more pronounced in centers from which the highest patient recruitment was provided for the study (from 49% to 87%). Loss of expertise and motivation in the neurosurgical practice has emerged as a potential outcome of the study.[4] The results of this trial have even been interpreted as the end of the aneurysm surgery era.[5] The report was heavily criticized by the neurosurgical societies,[6], [7] whereas it received support from neuroradiological community.[8]

Criticisms on ISAT
In this study, the ISAT report was analyzed critically. Methodological aspects of the trial including the patient selection criteria and assessment of the patients were evaluated. The outcome results and post-procedural re-bleeding rates were analyzed. Available neurosurgical and neuroradiological literature was reviewed and common criticisims regarding the methodology and interpretations of the results were put up together and stratified.

Selection biases: A total of 9559 patients were assessed in ISAT for eligibility and only 2143 (22.4%) were randomized. Although the study initially aimed to recruit 2500 patients, the steering committee stopped recruitment prematurely at 2143 patients. The contribution of the participant centers to the study varied between 1-44% of their real patient population, which seems to be due to the differences in the appreciation of clipping versus coiling in different centers.[9] This raised the question of whether there was an inclination to one of the treatment modalities in those centers with higher rates of patient recruitment.[10] It was stated that the major contributing centers had more expertise in endovascular treatment.[11] Lindsay reported the proportion of patients referred to endovascular treatment as 49% in the major participating centers of the study as compared to the average of 34% in UK.[3] UK provided 76.7% of the patients in this study and the most contributing 5 centers from UK recruited 50.9% of all the population of the study [Table - 1].

Outcome assessment: Modified Rankin Scale (mRS)[1],[12] was used to assess the functional outcome at two months, one year and annually after. The percentage of dependant and dead patients (defined by Rankin score 3-6) was 36.4% and 30.6% in the surgical group as compared to 25.4% and 23.7% in the endovascular group at two months and one year, respectively. It is important to note that the larger outcome difference at the two months evaluation decreases by one year.[13] At one year, there was a 22.5% relative and 6.9% absolute risk reduction in the endovascular group [Table - 2]. However, Harbaugh showed that the statistical significance between functional outcome of each treatment group exists only when the groups are pooled as 0-2 and 3-6.[14] The statistical difference disappears, if the groups are compared one by one, are pooled as 0-1 and 2-6 or 0-3 and 4-6. In fact, De Haan et al., proposed the pooling to be 0-3 and 4-5 for the use of large-scale multi-center trials.[15] The data were collected via a postal questionnaire and the ability of a postal questionnaire to differentiate Grade 2 (some restriction in lifestyle) from Grade 3 (significant restriction in lifestyle) is also questionable.[16]

The mRS has been previously described as a subjective scale with a possibility of misclassification.[17] In addition, patients may show a tendency to have more subjective complaints than their objective functional loss after subarachnoid hemorrhage and this may be reflected to the mRS because of its subjectivity 18. MRS may not correlate strongly with more objective scales in determining the functional outcome of patients with subarachnoid hemorrhage[19] and may have a lesser inter-observer agreement compared to other similar scales.[17]

Surgical outcome: In the ISAT report, 88% of the patients had good clinical grade (WFNS Gr1 and 2) and the aneurysm size was less than 10mm in 92%. In addition, 97% were anterior circulation aneurysms. All of these factors are accepted as good predictors for a favorable outcome.[20],[21] However, surgical outcomes of this study were criticized as being unfavorable as compared to other studies in spite of a favorable patient population [Table - 3] and experience level of the participating surgeons was proposed as an explanation for these relatively unfavorable results.[14],[23],[24]

Although there was a prerequisite of minimum 30 case experience for endovascular operators, no objective criteria were set for neurosurgeons. It has been shown that high case volume impacts the outcome in the treatment of aneurysms[25],[26] and this impact is shown to be even more pronounced in clipping than in coiling.[25] Furthermore, it has been stated that the dilution of microsurgical skills in aneurysm surgery was already becoming apparent in the UK, the main contributor of the study, before the results of ISAT due to the spread of the cases over a large number of surgeons.[27]

Angiographical considerations: In the study, 92.5% of the endovascular procedures were reported as "completed". However, in the absence of any angiographic outcome data, this expression does not give an idea about the obliteration of the aneurysm.[23],[28] Therefore, this creates some ambiguity concerning the efficacy of this treatment modality. It is important to remember that repeat procedure rate in this study was 12.7% in the endovascular group as compared to 3.2% in the surgical group [Table - 2]. Post-operative angiogram was not obligatory in the microsurgical group. This may underscore the real number of suboptimal clipping, which may be another reason for the unfavorable outcome results in the surgical group.[9],[10] Macdonald et al, reported a 4% of residual filling after surgery and in an additional 4%, uncomplete clipping required re-operation.[29] Residual filling after microsurgical clipping was reported as 5.2% in a review of 1569 aneurysms.[30]

Long term angiographic results of clipping has also been investigated. David et al., reported 91.8% complete obliteration of the aneurysm after clipping.[31] This study had more than four years follow-up and the overall recurrence was 1.5%. However, in the presence of a residual neck, this rate increased significantly. Tsutsumi et al., reported 2.9% overall recanalization rate after clipping with an average follow-up of nine years.[32] This rate was 2.4% for aneurysms which were completely clipped [Table - 4].

The follow-up in this study is one year. It is well accepted that one year follow-up is not enough to evaluate the results of endovascular treatment.[6],[7],[33] Raymond et al., reported a 33.6% rate of overall aneurysm recurrence who had undergone coil treatment.[34] Interestingly, only 46.9% of these recurrences were detected in six months, whereas it required 36 months to detect 96.9% of the recurrences. In the same study of 356 patients with 383 aneurysms, only 38.3% of the patients had complete obliteration at one year and 15.3% of the initially completely obliterated patients recanalized after three years. Murayama et al., reported their results of 916 aneurysms in 818 patients.[35] Their overall complete obliteration rate was 55%. Even in the small size-small neck group this rate was 75.4%. In this series, 489 had an average of 11 months follow-up. They had a recanalization rate of 5.1% in the small size-small neck group and 20.9% in the overall patients [Table - 4]. In the presence of a residual aneurismal neck after the initial coiling, different studies demonstrated that only 25-46% of theses aneurysms thrombose, whereas 28-49% undergo recanalization.[36],[37]

Post-procedural re-bleeding and outcome: Post-procedural re-bleeding from the target aneurysm at one year was 1.3% with no re-bleeding after one year in the microsurgical group and 3.2% with two cases of re-bleeding after one year in the endovascular group . These results show a 2.6 times higher re-bleeding rate at one year in coiled patients as compared to clipped ones. Time of re-bleeding after coiling has been a controversial topic in the literature. Thornton et al., reported no re-bleeding after six months.[35] However, Brilstra et al., documented that nine of the 16 patients re-bled between six months and three years.[38]

Based on the findings of ISAT and assuming that 70% of the patients with re-bleeding would have a mRS Grade 3 or worse, Ogilvy calculated that the difference in functional outcome disappears by six years and the outcome of microsurgical group becomes more favorable with statistical significance by 10 years.[39]

Conclusion

Results of ISAT may be applicable to patients who have good WFNS grades, ICA or ACA aneurysms with a size of 10mm or less and in whom both modalities can be considered as therapeutic options. However, the risk reduction in dependency provided by endovascular treatment must be evaluated with reservation in the presence of an increased risk of repeat procedure and post-procedural bleeding. An optimum outcome assessment should include a clinical outcome assessment with a universally accepted scale and long-term angiographic outcome.

The collaboration of cerebrovascular neurosurgeons and endovascular operators as well as training of neurosurgeons in endovascular techniques would result in the availability of both treatment modalities to patients, eventually raising the quality of care.

References

Copyright 2004 - Neurology India

The following images related to this document are available:

Photo images