A promising and active field of comparative genomics con- sists in comparing two genomes by establishing a one-to-one correspon- dence (i.e., a matching) between their genes. This correspondence is usu- ally chosen in order to optimize a predefined measure. One such problem is the Exemplar Breakpoint Distance problem (or EBD, for short), which asks, given two genomes modeled by signed sequences of charac- ters, to keep and match exactly one occurrence of each character in the two genomes (a process called exemplarization), so as to minimize the number of breakpoints of the resulting genomes. Bryant [6] showed that EBD is NP-complete. In this paper, we close the study of the approximation of EBD by showing that no approximation factor can be derived for EBD considering non-trivial genomes - i.e. genomes that contain duplicated genes.