Experimental inactivation cross sections are compared to vacancy production in inner shell (K-shell) of so-called strategic C, N, O atoms of the DNA: in a trial calculation, atoms on the DNA backbone or just contiguous to it are considered as strategic. Various processes are considered: ionization and electron capture by the primary particle, ionization by secondary electrons. The last phenomenon is shown to be important for highly charged and medium velocity ions. A strong similarity is observed between variations versus LET of inactivation and K-vacancy cross sections, especially with regards to LET values for which they maximize: for K-ionization these maxima occur when the ion impact velocity roughly matches the K electron velocity. It is shown that the "K-hypothesis" allows a quantitative fit of experimental inactivation cross sections if a 4% mean lethal efficiency is attributed to ion-induced K-vacancies in strategic C, N, O atoms.