One potential problem with this approach is that nucleoside inhibitors like AZT first need to be phosphorylated by cellular kinases. Only after they have been converted to a triphosphate form can they be incorporated into viral DNA by reverse transcriptase. Therefore modifying the drug to resist phosphorylation would be likely to render the drug inactive.
An alternative strategy would be to design a nucleotide analog that, once incorporated into DNA, cannot be removed by pyrophosphoroylsis. To the best of my knowledge, such compounds have not been described yet.
Another interesting idea is to design chain terminators that do not block polymerization right after they are added, but rather block polymerase activity only after a few additional nucleotides have been added to the inhibitor. Such
delayed chain terminator drugs may make it difficult for the enzyme to remove them via pyrophosphorolysis.