Yes, I would like to add that with MLST you are typing a limited number of loci, which apparently all your strains share. On the other hand, with the PFGE you are visualizing differences in other loci (not included in the MLST).
It is possible. PFGE is more discriminatory than MLST: PFGE can detect chromosomal rearrangements. In contrast, MLST is more appropriate for strain phylogeny and large-scale epidemiology. Indeed, determination of phylogenetic relationships among strains with distinct PFGE patterns may be imprecise. For example, strains belonging to the same ST did not always cluster in a single branch based on cluster analysis of PFGE patterns. According by Johnson et al (2007), PFGE detects changes in the molecular clock better than MLST does. PFGE and MLST may be considered complementary methods, which are appropriate for studies at distinct scales, i.e. local epidemiology versus global population structure, respectively.