New data from an analysis of patients who would have been eligible for the ARUBA trial—which previously indicated medical management is likely favourable over intervention in unruptured arteriovenous malformation (AVM) treatment—have produced results contradicting this seminal study, and look set to reignite debates over the role for surgical approaches in these cases. Following the presentation of their findings at this year’s Society of NeuroInterventional Surgery (SNIS) annual meeting (31 July–4 August, San Diego, USA), Anas Alrohimi and Nina Moore (both Cleveland Clinic, Cleveland, USA) spoke to NeuroNews to provide a closer look at the study in question and its wider implications.
You described ARUBA as “controversial” in your SNIS presentation—would you be able to outline the study’s limitations and, as such, why its findings were potentially lacking in generalisability?
AA: Despite ARUBA being a prospective randomised controlled trial, there were limitations in the study design that compromise the ability to generalise the results into real-world practice. Almost 50% of the screened, eligible patients refused participation and 20% of patients at active centres underwent intervention outside the trial. Additionally, only 18 randomised subjects in the interventional arm (19%) underwent microsurgical resection with/without other modalities, with the rest undergoing embolisation, radiation, or some combination of the two without reporting complete obliteration rates. The majority of patients underwent radiosurgery—for which the results are not expected to be seen in the short term. However, the follow-up duration was relatively short at an average of 32.9 months. This represents a significant flaw in the trial’s methodology, as the short follow-up duration was insufficiently long for radiated AVMs to obliterate.
In addition, brain AVMs are a heterogenous disease and one treatment modality cannot be applied to all, but the ARUBA trial considered the intervention arm as a single entity while, in fact, the distribution of interventions (microsurgery, embolisation, radiosurgery, or a combination) is unbalanced. Further, the significant outcomes in the interventional arm are driven by a high rate of haemorrhagic strokes, which raises speculation over whether this outcome is related to the imbalance in the distribution of interventions. Finally, multiple subsequent cohorts of ARUBA-eligible patients treated with embolisation, microsurgery and radiosurgery—alone or in combination—have demonstrated lower overall risks of stroke or death in their treatment arm versus in ARUBA.
How much of a contrast do your recent findings represent as compared to the original ARUBA trial results from nearly a decade ago?
NM: In this study, we had a remarkably lower complication rate for patients that met ARUBA trial inclusion criteria, with complication rates being 8.7% from the NeuroVascular Quality Initiative-Quality Outcomes Database (NVQI-QOD) versus the 30.7% seen in the ARUBA trial treatment arm, which was statistically significant (p<0.001). This NVQI-QOD lower rate was better than the medical management complication rate of 10.1% seen in the ARUBA trial as well.
What do you think is the likely cause of this discrepancy?
NM: There may be multiple reasons for this discrepancy. During the enrolment for ARUBA, over half of the screened participants refused enrolment, as the trial randomised patients to treatment versus the medical arm—and these patients that refused enrolment may have been better surgical candidates. Additionally, the centres putting data in the NVQI-QOD are specialised centres that handle complex cerebrovascular disease routinely. The methods of treatment in our study also varied in distribution from the ARUBA trial, with a larger percentage undergoing open surgical resection, which likely played a part in there being lower complication rates. Endovascular therapy is making rapid strides in improving on all fronts with new technology, and a greater wealth of experience and research.
Between the three individual interventions assessed in your research—embolisation, microsurgery and radiosurgery—were there any interesting observations that may warrant additional investigation or perhaps have wider implications for AVM treatments?
AA: Our cohort is divided into three groups; microsurgery with/without embolisation (75/173, 43%), embolisation (61/173, 35%), and radiosurgery (37/173, 22%). Our analysis did not show significant differences between the three groups in the risk of stroke and death over five years, functional outcomes, and other neurological and systemic complications. However, the microsurgery-with/without-embolisation group achieved a higher rate of complete AVM obliteration on follow-up imaging modalities. I think our results raise many questions that may warrant additional investigations, such as: what is the safety and efficacy of each interventional modality? What are the AVM characteristics that warrant intervention and timing? Is targeting the AVM flow and high-risk features sufficient, or is complete obliteration the goal?
How significant are these findings in relation to the neurointerventional treatment of unruptured AVMs? Are they alone likely to trigger a shift in clinical practice, or are further studies needed to fully disprove the initial results of ARUBA?
NM: These findings are important for all treatment options for unruptured AVMs—open surgical resection, radiosurgery, and endovascular embolisation. There is a cumulative risk of rupture for cerebral AVMs over a patient’s lifespan. The ARUBA trial had a relatively short follow-up time period and, with our findings in mind, leaving an unruptured AVM alone is likely a greater risk than treatment of the AVM by any method. This study has a larger number of ARUBA-eligible patients (173) than the treatment arm of the ARUBA trial (105) and should refute the complication rates seen in ARUBA—as demonstrated by the NVQI-QOD specialised cerebrovascular centres—producing ‘real-world data’ from clinical practice.
Is any further research now planned to bolster this recent study—either from the Cleveland Clinic team, or additional studies from other centres you may be aware of?
AA: With the known fact that brain AVMs have an accumulative annual risk of rupture, which can lead to significant morbidity and mortality, the results of ARUBA suggesting the superiority of symptomatic management over intervention were disappointing. Our study is the largest cohort for ARUBA-eligible patients following the original trial, and it showed significantly better outcomes than ARUBA despite heterogeneity of modalities and treatment paradigm at 18 comprehensive stroke centres that collaborated via the NVQI platform. I am not aware of any current, large-scale study comparing the safety and efficacy of AVM interventions.
Our analysis is encouraging and expands the opportunity to explore this entity further. Dr Moore and her team at the Cerebrovascular Center of the Cleveland Clinic took this to a different level by studying the biodynamic and structural mechanics of AVMs—using her expertise and background in biomedical engineering. There is also a significantly growing interest in the molecular and genetic background of AVMs, and applying precision medicine by using targeted medical therapies. There are many aspects of brain AVMs that are yet to be investigated, but the field is evolving and we are determined to make steady progress to provide the best care to our patients.