I received a Google Alert this morning about your thread, and as a researcher and PhD candidate who has been studying oxytocin for the past 5 years, I felt it necessary to share my knowledge and to caution you against the use of oxytocin as a treatment for behavioral symptoms of ASD (or any other disorder for that matter). Currently, there are a handful of academic studies that have investigated the effects of intranasally administered oxytocin, and some of these have shown quite promising effects, citing improvements in social cognition (Hollander et al., 2007), emotion recognition (Guastella et al., 2010), and social behavior (Andari et al., 2010). While it is tempting to jump to the conclusion that oxytocin is a cure-all for symptoms associated with ASD, there are a number of issues with its use therapeutically.
Before describing these issues, let me give you a brief description of how intranasal administration of oxytocin works (or is thought to work). To administer this hormone, you have to squirt a large quantity of the substance (from 24-40 IU is typical for research) into each nostril a number of times. In our research on oxytocin, we typically do 4 sprays in each nostril a total of four times per nostril. This is an awful lot of fluid, and if you think about it, intranasal administration of anything is really not the most enjoyable experience, as a majority of the substance drips down the back of the throat! We administer this quantity in hopes that some of it will enter the brain to have the behavioral effects we are investigating. Because oxytocin is a large molecule, it cannot cross the blood brain barrier, although there is some evidence that a bit of it does get into cerebral spinal fluid. Thus, the the hope is that if enough is administered, it will get into the brain to act as a neurotransmitter and exert behavioral effects (oxytocin as a hormone circulating in the blood does not directly affect behavior, although it can have anxiolytic effects which could in turn affect behavior). Anyway, now that I've given you a brief summary of administration, on to the potential issues with its use as a treatment.
The first issue is the absence of clinical research testing the effects of this hormone for use as a long-term treatment. To date, published academic research conducted on oxytocin has only tested the effects of single administration rather than repeated use. Further, these academic studies are often conducted on small samples of specific groups, meaning the results may not generalize to larger and more diverse samples. Currently, academic research is really in the early stages of testing oxytocin as a treatment, and results are in need of replication before consideration for clinical testing. Having noted these things, oxytocin available via prescription, but only because it is an integral part of labor onset (if you were induced, you received the synthetic form of oxytocin called pitocin) and milk letdown, but note that oxytocin has been medically tested and approved regarding these things.
The second issue is that there is little to no knowledge about how long these behavioral effects would last in alleviating symptoms of ASD. Research has shown that when OT is administered intranasally, it completely clears the body within three hours. Thus, effects would only be expected to last for a short time after this. There's word that a pharmaceutical company is working on a longer-lasting synthetic form of oxytocin; however, as I mentioned, it's not even clear how much intranasally-administered oxytocin enters the brain, so it seems as though there's some basic "how it works" knowledge that should be acquired before turning to use of longer-lasting forms.
The third issue with the use of oxytocin as a long-term treatment is that regular administration of any hormone will, over time, lead to decreased natural production of that substance in the body. Flooding your son's system (which, as you say may already produce lower than normal amounts of oxytocin) will cause it to turn down production of the hormone even more. This means that over time you would have to continually increase the dosage of the prescription for the same effects (as with any drug), and without the treatment you might see even more severe behavioral symptoms. In the same vein, lower baseline circulating levels of oxytocin could occur for a number of system issues: too few oxytocin receptors, dysfunctional receptors, or dysfunctional neuronal projections for oxytocin release. We currently have little to no knowledge about what each of these system issues looks like in humans (they've been well studied in rodents); therefore, it is premature to assume that low levels of oxytocin in the blood means we just need to add more to the body.
Based on all of these details, I would strongly advise against the use oxytocin as a treatment, at least given our current state of knowledge regarding its long-term use. It is possible that with more research and discovery, oxytocin may be deemed a useful treatment in the future, although I predict that they only effectual method will be via gene therapy or receptor-based treatments. I am not an MD, and I highly encourage you to look to other resources regarding oxytocin as a treatment; however, I do hope that you take time to cautiously consider your doctor's recommendation, and I hope the information I provided was helpful!