Hey my wife saw your note & asked me to write something. First off, we're really sorry you have to deal with this. Our kid was in a similar situation. Spetzler said not to do anything, but keep in mind he is a surgeon first. Your best option almost certainly is volume stage radiosurgery, and the place to go in our opinion is Pittsburgh with Dr. Dade Lunsford.
When you have a large, deep, AVM traditional surgery is indeed impossible, as you know. Moreover, traditional radiation is out because the area radiated is less than the volume of the AVM. You cannot radiate the larger area, because the overall dose would be too high.
Dr. Lunsford pioneered the idea of using the gamma knife to radiate one part of the AVM at a time, waiting 6 months between treatments. The radiation intensity is turned down so the protocol can be done safely - Lunsford has done tons of these, and has researched the optimal levels. Lunsford tries to do as few sessions as possible, for a "huge" AVM as you describe I bet he would suggest 3 sessions. Lunsford and those he has trained have had success with giant AVMs that previously were untreatable, and that many doctors currently would say cannot be treated.
Here are the downsides. First, there is the risk of a hemorrhage while you wait for the treatment to work (always true with radiation, the time and hence risk is extended with volume stage radiosurgery). Second, because the radiation is turned down, the odds that the protocol will work (complete obliteration of the AVM) are only about 50% (one shot radiation more like approaches 90%). Third, you didn't say where the AVM is, but if there is basal ganglia/thalamus involvement the protocol might have side effects involving some weakness on one side (hemiparesis is present in 10% of such patients, I think).
For my kid, I looked at the odds. If we did nothing, the odds of a hemorrhage approach 100% across the expected lifespan. Accordingly, an option with 50% success rate is preferred. If the protocol does not work, it can be repeated. The odds of success across two protocols I bet are like 75% (the other numbers I gave are based on my reading of the neurosurgery literature, 75% is based on my statistical calculation).
My kid was a great piano player before the treatment. His right hand doesn't work real well now (2.5 years after we finished the protocol), and he can't play with it anymore. But as I write he is playing "How Great Thou Art" on the piano downstairs with his left hand - using the sustain pedal to create an awesome effect. Also he can't write as a side effect, but he uses his iPad in school and gets better grades that I ever did. I completely (respectfully) disagree with dancermom who cautioned against getting opinions beyond Spetlzer. If you have a chance to get the kid healed, you go for it. Talking with Lunsford, the guy who does the seminal research on the treatment that may be effective and who has cured many huge AVMs, I think would be wise.
We don't know if the first round of my kid's treatment will work - every year after the protocol is finished an MRI is done to track progress. His AVM is shrinking, but we have a long way to go. If this round doesn't work completely, we'll do a second round to try to clean up what's left.
Your decision may be rolling the dice on an uncertain treatment option that may have side effects, or doing nothing. If you do nothing, I'm sorry to be blunt but a hemorrhage at some point is all but certain. In our case we felt like we should give our kid the best chance we could, even if that chance didn't have the best odds.
Finally, please do not let anyone embolize then try one-shot radiation. The data show this doesn't work. One guy suggested this to us, but I had read the neurosurgery literature and although the idea sounds great in theory, in practice it is dangerous and ineffective.
One more finally - your kid has an uphill battle. I would be very reluctant to chase ideas that have no basis in science or data - for example, cutting sugar. There is really no reason to think that this may do anything, and you want to basically party like maniacs now while you all are together because the future in uncertain. Cookies are a big part of partying for kids in my opinion. What you should do is use Tylenol instead of aspirin or ibuprofen (e.g., Advil) because these thin the blood, which will make a hemorrhage worse if one does occur. Also - avoid Omega 3 supplements - same reason (if your macrobiotic person mentions Omega 3 you know he/she is a quack).
Listen, good luck. We'll be praying for you.