Nicholson [pointing a gun at DiCaprio]: I taste a rat!
DiCaprio: Those are you're new implants, boss!
Nicholson [putting gun away]: Oh yeah. It's going to take some time to get used to these...

Agreed, the perfect solution is the one that generates teeth already.

The issue with fibroblasts in adult human tissue, is that they don't form a blastema upon "injury". (unless you are a reptile, or certain kinds of fish.) I mention fibroblasts, because they are the creators and components of the extracellular matrix, which helps undifferentiated cells to understand where in the host they have found themselves.

In mammals, fibroblasts "regenerate" damage as scar tissue, instead of forming the blastema. In regenerating lizards and the like, they form this structure, which essentially regrows the missing or damaged tissue using stemcells from the host's blood supply, which migrate to the blastema, attach, then begin regrowing the damaged or amputated feature following the embryonic blueprint, using HOX gene activation of the supporting fibroblast matrix as a signpost.

The ideal solution, would be to collect a sample of tissue from a recently extracted diseased molar, culture it in a petri dish, and use a chemical cocktail to force it to become a blastema, which would then be reinserted into the jaw later. This ensures proper HOX activation for the site from the tissue culture, which helps ensure that the resulting tissue from the blastema will not only be a tooth, but also the CORRECT tooth.

It is important to note that the location in the body from which the fibroblasts harvested to create these blastema is critical in determining "what" will grow. Several experiments were performed on salamanders, where a lesion was purposefully created, then the blastema translocated to a different location on the host surgically. The result was the induction of grown limbs in inappropriate places, (such as tails and legs in the middle of the back), at the sites of translocation. Once the blastema has formed, it has already begun the developmental program for what will be produced. It is believed this is due to the activation state of bodyplan HOX genes in the fibroblasts involved in the blastema's creation, according to several gene expression assays performed.

This means that the tooth formed by a dental blastema would be highly dependent upon where in the mouth the cell culture was taken, and the presence of scar tissue being extent or not. It would be a very good idea to write down that information when taking your samples for culture, and not mixing the samples up on implantation. :D

It is personal conjecture time, but personally, I think that a cultured then frozen blastema could be later reintroduced as a grown tooth bud after the bone tissue has healed in the extraction channel, and after the in vitro blastema has had sufficient time to decide which way it will make its root structure, to reduce the risks of "serious complication", which needless to say, would require very invasive surgery to correct, as well as for it to develop diagnostic criteria for ensuring proper orientation and clocking for insertion. (The transplanted tissue bud would be about the size of a grain of milo, or smaller at this stage of development. Just enough to know which way it is pointed, and to get some diagnostic data for proper insertion from.) This way the introduced blastema would grow and integrate with the jaw in the appropriate fashion, though it would be a good idea to monitor its growth to ensure it was properly inserted, and is not going to cause an impaction later. (Somehow I doubt most insurance would cover the added expenses over that of a prosthetic device though, and they bitch enough about those. As such, I dont see this happening any time soon, but I don't see a major obstacle against it biologically/technologically. Bureaucratically is another issue entirely!)

Even if the resulting dental crown is abnormal, this at least produces a healthy root structure, (at least in theory), which would allow surgical correction with more traditional techniques, as required.

Naturally, this process needs to be performed in animals many times to work out all the risks of complicat