ARTICLE IN BRIEF
An independent data and safety monitoring board found a three-times higher event rate — strokes and time to death — in those patients who received interventional treatment (any combination of surgery, endovascular embolization, or radiotherapy) for unruptured arteriovenous malformations.
In May, the National Institute of Neurological Disorders and Stroke halted enrollment of ARUBA (A Randomized Trial of Unruptured Brain Arteriovenous Malformations) after a preplanned interim analysis by an independent data and safety monitoring board found a much higher event rate in those patients who received interventional treatment (any combination of surgery, endovascular embolization, or radiotherapy) for unruptured AVMs, compared with those who received only conservative medical management. The trial was more than halfway through enrollment.
To date, there is no standardized course of treatment for unruptured AVMs, which are abnormal connections between arteries and veins that may lead to bleeding in the brain upon rupture. ARUBA — led by clinical investigators J.P. Mohr, MD, Daniel Sciarra professor of neurology at Columbia University Neurological Institute; and Christian Stapf, MD, attending neurologist at Hôpital Lariboisière in Paris and adjunct assistant professor of neurology in the Division of Stroke at Columbia University — set out to compare long-term outcomes in the two groups of patients with unruptured AVMs, using a primary endpoint of time to death or stroke. The ARUBA data coordinating center is headed by Alan Moskowitz, MD, of Mount Sinai Hospital in NY. On April 15, after a mean follow-up of 33 months with 224 participants enrolled at 39 sites worldwide, the data and safety monitoring board declared an event rate in the intervention group that was more than three times higher than the medical management group. [At press time, the data on numbers of deaths and stroke were not available pending publication in a peer-reviewed journal.]
ARUBA participants were randomized to receive either symptomatic medical management alone (leaving the AVM untouched, but treating the headaches, seizures, or other coexisting medical conditions) or symptomatic management and invasive therapy (any combination of surgery, endovascular embolization, or radiotherapy). Functional assessment was carried out at the time of randomization, before the intervention, and 48-hour post-intervention, and for all participants at one month, and at six-month intervals throughout the follow-up period.
“The intent was to randomize 400 participants and follow them for five to 10 years; however, because of the higher than expected event rate in the interventional arm compared with medical management, the [data and safety monitoring board] has recommended stopping enrollment while continuing to follow all participants to determine whether the difference in stroke and death in the two arms changes over time. The study was not powered to detect differences in efficacy among the three interventional approaches,” according to the NINDS news alert.
These findings were presented at the European Stroke Conference in London on May 31.
Based on currently available data, Kevin Sheth, MD, chief of the Division of Neurocritical Care and Emergency Neurology and director of the Neurosciences Intensive Care Unit at Yale New Haven Hospital, told Neurology Today, “medical therapy for AVMs is superior to interventional therapy for patients with unruptured AVM.” However, the question remains: “Do these differences in rate of stroke and death between the study arms persist over time? Or, as the years pass, does the rate of stroke and death actually go up in the placebo arm, so that the difference between them becomes much less?” Dr. Sheth is also a member of the editorial advisory board for Neurology Today.
Dileep R. Yavagal, MD, director of interventional neurology and co-director of endovascular neurosurgery at the University of Miami & Jackson Memorial Hospitals, said it was not surprising that patients who undergo surgery or interventional treatment would have more events in the short term because of the procedure itself.
“The reason to treat AVMs which have not ruptured is to prevent morbidity over a much longer period of time — especially for younger patients — to prevent them from having a hemorrhage in the next five to ten years,” said Dr. Yavagal, who is also an associate professor of clinical neurology and neurosurgery and a faculty member of the Interdisciplinary Stem Cell Institute at the University of Miami. The study's duration — at 33 months — is too short to see the true difference in outcomes, he said.
The trial has been controversial from the get-go, he continued; in his own department, the specialists who treat AVMs feel that “it is very predictable that the trial was halted at this stage, and it does not warrant a change in current clinical practice. The study did not help with defining the right course of care in the treatment of unruptured AVMs.” As the ARUBA investigators continue to follow up with volunteers who are already enrolled, Dr. Yavagal said, more useful information may be gained. “However, the numbers are going to be less [than the study initially intended], so the power of the study is diminished.”
Robert D. Brown Jr., MD, chair of the department of neurology, John T. and Lillian Mathews professor of neuro-science, and professor of neurology at the College of Medicine at the Mayo Clinic, called ARUBA a “landmark study… the first randomized clinical trial of arteriovenous malformations that has assessed the comparison of medical management to the procedure that was felt to be best for that particular patient.” [Dr. Brown was involved in administrative aspects of ARUBA during its early stages.]
However, since it was stopped early, he stressed that there will need to be careful consideration of “the AVM characteristics of the 224 participants entered into the study, and those people not entered into the study at the participating centers. This will clarify whether the AVMs in the medical management group and surgical/interventional treatment groups differed in any substantial manner, and which people were not entered into the study at the participating centers during the same time period.” Dr. Brown said that this will also allow an understanding of how generalizable these data are to clinical practice.
One large criticism of the trial has been the lack of consensus in the interventional treatment arm on patient or treatment selection, leaving the decision up to the individual institution, commentators told Neurology Today. There were a variety of different interventions included — surgery, endovascular embolization, radiotherapy, or any combination of those three. In the future, Dr. Yavagal said, “I think a trial is needed to focus on better patient selection and a more codified treatment protocol.”
Agreeing, Dr. Sheth said, “One of the challenges and opportunities with AVMs is that there have been multiple clinical approaches used in clinical practice.” He added that the motivation for including all of the different intervention approaches in the ARUBA study was to increase enrollment, increase generalizability, and to look at multiple procedures. However, “when we're looking for proof-of-principle, the question is often whether it is better to have a high-specificity, uniform approach so that we can learn something about whether or not we should be treating [unruptured AVMs] at all, and whether we should be treating with a particular procedure or not?” For example, the advantage of selecting candidates for treatment who are most at risk for hemorrhage and randomizing them to one predefined interventional therapy versus conservative management, he said.
These are big and expensive studies to do, Dr. Sheth pointed out. “So, I think that there is some reluctance to go after high specificity because you can't generalize. The downside of not using a high specificity approach is that you are sometimes left with a negative study after all of that work and lots of unanswered questions.”
We've learned more about the natural history of AVMs over the course of the ARUBA study, he said, which may have an effect on outcome differences as well. “We also need to see what the long-term functional outcome difference is between those two study arms.” As more of the data are elucidated, Dr. Sheth is hoping to find answers to some of these lingering questions. It will be interesting to know if the event rates differ depending upon which of the three approaches was used and whether a combination of those approaches was used, he said.
Despite the trial being prematurely halted, Dr. Yavagal believes that, from the current results and the expected follow-up results, neurologists will better understand “how we can decrease the morbidity of treating unruptured AVMs.”
At this point, Dr. Yavagal stated that he remains “conservative” in his treatment of unruptured AVMs, and that this is unlikely to change. “If the published details will show the efficacy of one treatment over another over time, then I may consider altering my practice,” he added.
Still, what may change, he said, is a referring physician's likelihood of sending the patient on for interventional treatment. “The legal liability of doing this becomes a lot higher now with the results of the ARUBA trial, and there will likely be a dampening in enthusiasm of the referring physicians — who may not be treating AVMs, but are diagnosing them — for sending the patient on to surgical therapy,” Dr. Yavagal explained.
Potential changes to treatment may also hinge on whether any amendments will be made to reimbursement for the interventional procedures, said Dr. Sheth. “If there are changes to reimbursement, I think people will be a little less aggressive. But if there are no changes to reimbursement, then I think that there may not be as many significant changes to practice.”