Robert Harbaugh, MD

As we are all well aware, the International Subarachnoid Aneurysm Trial (ISAT) was recently published in the Lancet¹. ISAT, a prospective, randomized trial comparing the functional outcomes of patients with ruptured intracranial aneurysms treated either with clipping or coiling, has been trumpeted as the definitive article on the safety of clipping versus coiling for ruptured intracranial aneurysms. The following message summarizes a letter to the editor of the Lancet and a position statement on ISAT from the AANS, CNS and the AANS/CNS Section on Cerebrovascular Surgery.

The results published in the Lancet article demonstrate that, for this particular subset of aneurysm patients cared for in these particular centers, patients treated with coiling fared better at one year than those treated with clipping based on an evaluation using one specific outcomes measure. The purpose of our letter, our position statement and this Chairman's Message is to indicate points that we believe warrant additional emphasis and clarification.

First, the investigators decided to evaluate the functional status of patients using a modified Rankin Scale score. It is important to keep in mind the rather subtle differences that exist between adjacent scores (Table 1). Of greater concern, the ISAT investigators analyzed the outcomes in two groups; scores of 0-2 and scores of 3-6. They only reported the statistical analysis comparing clipped versus coiled patients for scores of 3-6. For this group of patients, a statistically significant difference between clipping and coiling was seen at one year. If the ISAT investigators had analyzed the outcomes by looking at groups 0-1 or 0-3, no statistically significant difference between clipping and coiling would have been found. In fact, if we designate the groups as was done in ISAT (0-2 and 3-6) but subject the 0-2 group rather than the 3-6 group to statistical analysis, no statistically significant difference is found. It is only by doing the statistical analysis on the group of patients with scores of 3-6 that one can find a significant difference between clipping and coiling in this study. This is a pretty shaky foundation on which to build a revolution in the treatment of ruptured aneurysms.

It should also be noted that most ISAT centers were located in Europe (particularly England), Australia and Canada. Only two patients were entered into the study from a single U.S. center. The results from ISAT may not be applicable to patients in the United States where practice patterns, particularly in reference to the degree of sub-specialization of neurovascular surgeons in major centers, are different. We believe that a carefully planned and executed randomized trial in the United States would be of value.

Another important but unreported piece of information is how many practitioners in the ISAT performed craniotomies for aneurysm clipping and how many practitioners performed endovascular procedures for aneurysm coiling. The absolute risk reduction for coiling compared to clipping at one-year follow-up is only 6.9 percent. If the number of coiling cases per endovascular practitioner is significantly greater than the number of clipping cases per neurosurgical practitioner, better outcomes in the coiled patients could be completely explained by a difference in practitioner experience and expertise. The number of neurosurgical and endovascular practitioners in the study and the number of procedures each performed should be published.

Physicians and surgeons involved in the ISAT felt that one form of treatment was preferred in almost 80 percent of patients for whom records are available. Of 9,559 patients with ruptured intracranial aneurysms assessed for eligibility, only 2,143 were randomized. In those not randomized, more patients underwent clipping than coiling as treatment for their ruptured aneurysms. In other words, over the course of this trial, neurovascular teams in the participating centers felt that surgery was the best option for a majority of patients with ruptured aneurysms who were not randomized. Therefore, if an experienced vascular neurosurgeon recommends clipping as the best option for a patient, that patient should continue to be offered surgery as the treatment of choice. The results of ISAT do not apply to such patients, as they were not evaluated in the randomized trial.

We also await with interest the long-term follow-up data on these patients. It is crucial to determine whether or not coiling will be as effective as clipping in preventing re-bleeding over each patient's lifetime. During the relatively short follow-up of the interim ISAT report, 2.6 percent of endovascular patients suffered a hemorrhage following treatment compared to 0.9 percent of surgical patients. In addition, 139 patients treated by coiling required further treatment compared to 31 patients treated by clipping. Although re-bleeding rates more than one year after treatment have been low in both groups, if a differential rate of re-bleeding persists over time, the modest 6.9 percent absolute risk reduction with coiling at one year will disappear. As the authors note, these patients need to be followed for many years before legitimate conclusions can be drawn about whether coiling or clipping is the safer treatment for patients with ruptured intracranial aneurysms.

The ISAT report is an important step in defining the roles of endovascular and microsurgical treatment of patients with ruptured intracranial aneurysms. The concerns noted above are raised to remind all of us that much more study is needed to develop definitive medical evidence on this issue. To extrapolate the early results of this study to all patients with ruptured aneurysms would be a misinterpretation of the ISAT data and a serious disservice to our patients and our profession.