Since you were a cancer researcher in a previous life, may I ask you a question I thought of commenting on another's comment?
If cancer cells have high metabolic requirements, couldn't you attack them by:
1. lowering the caloric intake of the patient
2. Making the glucose the patient intakes mildly radioactive, say with a beta or an alpha emitter [1]. As the patient is strictly kept on a low caloric diet, the radioactive glucose is consumed and expelled quickly (i.e. doesn't accumulate in the adipose (?) tissue).
Not so radioactive that the person is harmed but radioactive enough to pack an extra punch to the metabolically starved, and therefore stressed, cancer cells (who are drawing more of the glucose to themselves).
That should target cancer more specifically, and I guess it can be done in tandem with other techniques.
Is something similar done? I know that's how cancer is imaged, and the way I see it (I studied optics in a previous life) if you can image it, you should be able to destroy it.
[1] Say replacing the hydrogens in glucose with tritium. Hydrogen hopping will ensure that the water produced by metabolizing the sugar is homogeneously mixed, and therefore expelled with all the water loss mechanisms. According to wikipedia, TO2 has a half life of 12 days. I'd imagine you can lower that by pumping fluids into the patient.