We read with great interest the recent Personal View by Vermeulen and colleagues1 on cancer stem cells (CSCs). The authors elegantly outline benefits and limitations of assuming a hierarchical CSC model for tumour expansion, and report some promising new therapeutic approaches. They also discuss ingenious strategies to assess the efficacy of new drugs targeting CSCs. In addition to the therapeutic approaches cited by the authors, we believe that radiotherapy with highenergy charged particles should be considered to specifically target CSCs and improve therapy outcomes. 2 Particle beams of protons and carbon ions are becoming widely used in radiotherapy because of a favourable depth–dose distribution and a higher cell-killing efficiency (relative biological effectiveness [RBE]) compared with photons. 3 Experimental results show that RBE increases for radioresistant cell lines (figure A). Charged particles induce complex and clustered DNA damage, which is hard to repair and largely independent of the presence of reactive oxygen species. As a result, the mechanisms responsible for cell radioresistance (cell-cycle stage, high expression of radical scavengers, DNA damage-repair proteins, and antiapoptotic genes) are less effective after exposure to heavy ions. Hypoxia-induced radioresistance is also reduced after particle irradiation (figure B). Clinical data for radioresistant tumours—such as malignant melanoma, adenoid cystic carcinoma, osteosarcoma, chordoma, and chondrosarcoma—support these experimental results. 2 Several studies have reported that cancer cells of the tumour bulk are more sensitive to sparsely ionising radiation (x-rays or γ-rays) than their stem-cell counterparts. 4 Enrichment of tumour bulk with putative CSCs has been observed for many tumour types after irradiation, both in vitro and in vivo. Quiescence, faster resolution of double-strand breaks, strong cell-cycle checkpoints, and hypoxic conditions are hallmarks of several CSC types and have key roles in determining radioresistance and chemoresistance. Hypoxia is known to promote CSC radioresistance; low oxygen concentration is a hallmark of

CSC-like state and hypoxic regions in tumours are rich in CSCs. Energetic heavy ions are very effective against radioresistant and hypoxic tumours (figure B), and can overcome the intrinsic radioresistance of CSCs. The first data comparing the effectiveness of x-rays and carbon ions against colorectal CSCs were recently published. 5 Cells positive for putative CSC markers were enriched two to three times (depending on the dose) after exposure to x-rays, but remained unchanged or decreased after carbon-ion exposure. An in-vitro colony assay showed that carbon ions were twice as effective as x-rays in CSC cell killing, and results were confirmed with xenotransplantation assays. Delay in tumour growth increased after exposure to carbon ions, and no regrowth was noted after exposure to the highest dose (30 Gy). 5 These results are promising, and justify more studies to clarify the role of particle irradiation as prospective therapy against CSCs. Combined treatments with x-rays and a boost of heavy ions delivered to the highly hypoxic core of tumours should also be considered as a therapeutic strategy to target CSCs.