Results from the international Aruba study

Dear Friends,

Results from the Aruba study (a multi center study) are just published in the Lancet (Mohr at al published on line, november 20). I imagine that many of us were waiting to know what is the statistically better decision in the case of an unruptured BAVM: to wait and see or to decide for interventional therapy. A comment is published in Nature Review Neurology this week. I will give you a short summary.
223 patients were enrolled, the follow up was 33 months. About half have a medical management (if they have pain, or seizures...), the others have medical management + interventional therapy (neurosurg, radiotherapy and/or embolization); this was a random study. Patients with prior bleeding were discarded from the study.
The risk of bleeding in the patients with only medical management was 2.2 % per year, that is considered rather low.
Strokes (hemorragic or ischemic) occured in 45 our 119 patients with interventional therapy, and in 12 out 109 patients with only clinical management.
The results suggest that in unruptured BAVM natural history is of lower risk than interventional therapy.
A further follow up (5 years) of the patients is planned.

Personal comment: unfortunately, only 223 patients (the authors were hoping 3 or 4 fold more) and only 33 months. Even if the risk of bleeding is considered as low (2.2% per year), most people agree to consider that it largely increases during the first year after the event. So, it is important, if it is decided to wait and see, to have a very regular follow up, to be ready to change his mind in case of... this is my own, personal, experience.

I spoke to my surgeon about this as I took it to mean that not treating was best option. However she said that the results would only be truly reflected after the complete lifetime of the participants and the fact that the time to date was only 33 months was misleading. I am still unsure what to do for the best!! Unfortunately the study did not differentiate between types of treatment eg., gamma knife, emboli station, etc.

this is interesting. i had my avm treated 20 years after it initially bled. apart from seizures i had no symptoms from it. after two courses of stereotactic radiosurgery i was told my avm was gone but i still have seizures which im followed up for. i have had no follow up for the avm since they sent the letter to me unlike others on this site. i sometimes wonder if it willl come back or thre cd be another one. do i need follow up and do i need to be concerned.

Linda, my neuro surgeon said the same thing. He said there weren't enough patients in the study and that 33 months wasn't enough time...The reason it only lasted for 33 months is that the federal goverment dropped the study for some reason..

I am not a specialist, but I know that late onset re-permeabilization of AVM that have been obliterated might happen. Maybe you will feel safer if your ask to your neurologist an follow up with MRI every 5 years? It's not so much and you will feel good,

Read the entire article.
Published online November 20, 2013. It is eight pages.

I have one question, as I can't access the study at The Lancet; you write 223 patient were enrolled, but the sum of 109 patient without and 119 with intervention is 228. Were 5 patients excluded in the final results or is that a typo?

(By the way, using a contingency table I calculated phi^2=0.0956 assuming a statistical population of 228 patients, which doesn't impress or scare me much, but without having access to the real data I wouldn't count on it. phi^2<0.25 would indicated that there's weak correlation between strokes and intervention if the null hypothesis is 'there's no connection'.) Cheers, Alex.

Hi, you are right. I apologize for the typo error. The correct numbers are 114 (interventional) and 109 (medical).

Hi Christina,

thanks, I already found a longer abstract and learned that a logrank-test was used, the study was stoped preliminary because medical management was found superior.
Will get a a full copy of the study and go through it, from experience (in other areas, I'm not medically trained) I know that abstracts are ofteen misleading.
Will have a follow-up with my neurosurgeon soon anyway.

Bye, Alex.

Here is my issue with the study…too few people involved and not long term enough. I want to see a study with more people and something that is at least 25 years old! Unfortunately, there still is no risk free way to get rid of an AVM. It is up to the individual how to deal with it. The great thing about this page is that we will support your decision on however you decide to handle it!

Yes, the population is small, but it's the only randomized study including a full population instead of a sample. The quality of such studies is generally better because the data includes all aspects and are not skewed through, for example, leaving out data because they were not available or unfavorable.

In the last few hours I had the chance to have a brief overview of the study and will look more deeply into that later. I'll post an update later, first have to look into certain technical aspects of their analysis (need to look up hazard-functions again, I'm not working often with these).

And never mind my phi-coeffecient calculation above, the study used chi^2, which looks much more suited.

Have a nice evening,

I think randomizing avm treatment or no treatment is a terrible idea, because brain avms are not a uniform entity but have endless variability in size, location, and complexity. Not to mention that the patients themselves have variability in age, gender, genetic makeup, and overall health. I don't see that it is possible to control for all the other variables, so cannot see how such a small sample size could give us any significant information.

A registry makes so much more sense to me than a randomized study. Over time, patterns would emerge that could be used to improve the standard of care.

To give the study some credit, they took different AVMs into account; some relevant aspects were:

  • modified Rankin-score
  • Spetzler-Martin grading
  • Morphology: Size, <3 cm, side, location: lobar, infratenorial; eloquent
  • venous drainage pattern
  • Aneurisms

They *did* a comparison of the randomized study using a full population (plot A in figure 2) and collected *samples* from other studies (plot B, figure 2) and compared them; the outcome for patients from the randomized study is much better than from the samples. This could indicate three things: 1. The randomized study's population wasn't big enough 2. The samples from other studies are not representative ("skewed") 3. Both. However, the resulting survival functions(*) seem to aggregate the data again so without having raw data you can't really tell under which circumstances events occurred; did the surgery/embolization go wrong? Are events correlated to combined therapy?

There seems to be an appendix to the study I couldn't find yet, will try to get my hands on it.


(*) "survival function" is a statistical term, originally from medical studies, but it's not to be taken literally; it means the probability, that a certain event - here stroke or death - might occur.

I fully agree with you, although I doubt that this is feasible due to different regulations in different countries. Where "feasible" means "there is a disturbing lack of political will".

I spoke with my doc about the study last summer when I went for my yearly follow-up. One of the biggest problems they had with the study was finding participants. Because it was a randomized study, the participants didn't have a choice in what treatment, if any, they were going to have. For me personally, I would want to have total choice in what sort of treatment, if any, I would be receiving.

I am one of those "medical management" people, however because I have no symptoms as of yet, medical management for me means going for a yearly MRI to make sure everything is status quo. My AVM is bordering on a grade 4 so proceeding with surgery or radiation would be quite risky and as my doc tells me, right now "i look so good!".

AVMs are very hard anomalies to study with regards to treatment. I just keep on hoping that they come up with new and improved treatments with less of a risk of deficit.

When my AVM was dtected, I went and see 3 medical teams in Paris (my city)... and send also the MRI and angio images to 2 groups in USA. One French group proposed to me to participate to the study, but I declined beacuse I could not imagine that a computer will decide that I will be treated or not. But it was clear that, when the computer decided that you will be treated, it will be with the most appropriate interventional treatment. The computer only indicates in what group you will be: interventional versus medical management. Despite 4 groups told me that I must be treated (the 5th was the one who proposed that I participate to Aruba Trial), my personal choice was the same as yours: statu quo, wait and see. I must say that proposed treaments were not the same from one group to another one: gamma knife, surgery, embolzation, radiation and all combinations!!!. Finally, it seems to me that they proposed what they used to do in their hospital, depending on their equipment ... When a small hemmorrage was detected in my brain, I decided without any doubt to be treated and choosed a French Group on the following criteria : a large experience ... and the quality of the associated anesthesist and intensive unit care!

Interesting . Anything we can learn or think prompts another mind to question , research and perhaps take us a step closer to answers .

For those of you that would like to know about Study's for Brain AVM's, I was sent this from the U.S. National Institutes of Health: