Well, RA is different from a cartilage tear, of course. But overall, my takeaway from the whole "stem cell rejuvenation" gig is that the marketing is probably ahead of the science. Classical error made by injured, fragile people looking for healing. Maybe injected stem cells migrate and cause tumors elsewhere? That's FDA's concern, I think. I'm seeing an ortho in a few weeks. See what the (biased) university-connected docs have to pitch with regard to meniscal tear regeneration. I don't expect much. Boo-hoo
There is no hope of stem cell marketing ever being accurate: most therapists are full of exaggerated claims and obtuse testimonials; while overregulated conventional docs say it doesn't work because they don't want to get sued if it fails, but won't get rewarded if it succeeds. But that's OK because stem cells don't care about truth-in-advertising. So I would look at the noisy-but-ample data on all this, and ignore both camps otherwise.
I don't think the FDA is concerned about health at all, but I think you're correct that their overriding putative "concern" about stem cell therapy is unregulated tissue growth. But again, if the reproduction of senescent stem cells were the only requirement for cancer, then babies would be loaded with tumors, because they're formed from decades-old stem cells, just like the ones that might fix your cartilage. And statistically, I think we've had enough test cases (tens of thousands) of BMSC and adipose therapies in the good old overregulated USA that if the cancer risk were wildly increased, we would have seen the lawyers partying by now. (Yeah, the theoretical increased risk might for example peak 20 years out, but that would still probably result in a detected increase in short term malignancies, especially because the cancer risk is widely advertised to patients in the popular press.) But this doesn't seem to be happening.
What does seem to be happening with respect to mortality risk is a steady stream of incompetent therapists having procedural accidents resulting in injury and death, mainly in dodgy third world clinics. But if the therapy is successfully delivered, then logically, better tissue perfusion due to IV (but not cartilagenous) stem cell injection should actually improve organ function and immune access to deeply buried cells, thereby reducing the probability of cancer, and likely all-cause mortality. Bottom line: I understand that we won't be able to prove the safety and effectiveness of stem cell therapy for several human lifetimes yet, so we're forced to decide on the most logically appealing argument and risk-reward tradeoff, and behave accordingly. Welcome to life with noisy data.
So because stem cell therapy can have such global effects, the best method of analysis may be something along the lines of what I'm trying to do along with my friend: if you take a risk, publish as much data as you're willing to share, so that eventually we can achieve a roughly accurate understanding of cause-and-effect. Clinical trials are almost invariably too narrow in their focus (and to be sure, their selection criteria) to see the majority of either risks or benefits, so I suspect that most of our understanding of this field will come from retroactive data mining of published before-and-after measurements, which are inevitably tainted data sources, but are certainly worth more than "we don't know" when taken in the aggregate. I don't want this to turn into a discussion of artificial intelligence, but suffice to say that the most significant missing piece of engineering in stem cell research is probably an IBM-Watson-like machine which can read massive amounts of untrusted data about therapy effects, and come to some statistically supported conclusions using forensic/Bayesian techniques to eliminate the BS.
So I for one would be thrilled to have any statistically significant data that emerges from your case, good or bad.