This is a very good question. Any new therapeutic would certainly have to be specific for the cancer stem cells and not harm the good stem cells. So that may involve some form of targeting scheme, perhaps using nano-technology. But in my opinion, the even larger hurdle to success is that many CSC tend to reside in the deep hypoxic/necrotic niches of a tumor, where drugs can not easily penetrate. Therapeutic strategies that work to limit blood flow to a tumor may in reality be making the tumor more resistant, by increasing areas for CSC expansion. We may need a paradigm shift, working to get better blood flow into the tumor to deliver drugs. Just some food for thought....
There are plenty of drugs thought to target cancer stem cells already in clinical trials, so it is clearly possible. Here is a link of clinical trials in the US, many of which are testing cancer stem cell-targeted therapies:
In Japan, sulfasalazine, the conventional drug for patients with rheumatoid arthritis and ulcerative colitis, is now under the clinical trial for advanced lung cancer without driver gene mutation such as Ras or ALK fusion gene , tonuge cancer and gastric cancer with high level of CD44v8-10 expression. The agent works as xCT inhibitor and now currently available for glioma patients to pass through blood-brain barrier (BBB). CD44v8-10 helps xCT stabilization at the cellular membrane, which is why the heterogeneity of CD44v8-10 expression reflects the dependency to xCT-mediated GSH synthesis in cancer stem-like cells as compared with CD44v-negative cells. Furthermore, brain tumor cells are mesenchymal and lack of CD44v expression, but highly express and localize xCT transporter independently CD44v, so that the modification of sulfasalazine to pass through BBB is expected to exhibit therapeutic effect.