Treatment adaptation after hepatitis B virus (HBV) treatment failure relies on genotypic resistance testing. However, the results of such tests are not always consistent with treatment response. These discrepancies may be due to differences in resistance levels between isolates with the same genotypic resistance testing profiles. We explored this hypothesis by investigating six cases of entecavir treatment failure with an integrative strategy combining genotypic and phenotypic resistance testing, medical record review and therapeutic drug monitoring. Among isolates with genotypic reduced susceptibility to entecavir, one displayed a higher level of resistance to entecavir (mean fold change in entecavir IC50 of 1 508 ± 531 vs. 318 ± 53, p=0.008). This isolate harbored a substitution (rt250L) at a position reported to be associated with resistance (rt250V). Reversion to wild-type amino acid at this position partially restored susceptibility to entecavir, confirming that the rt250L mutation was responsible for the high level of resistance to entecavir. This is the first description of entecavir treatment failure associated with selection of the rt250L mutation without other entecavir resistance mutations. One isolate with genotypic resistance to entecavir, harboring the rt173L mutation, displayed a lower level of resistance than the other, harboring the rt202G mutation (mean fold change of 323 ± 124 vs. 6 036 ± 2 100, p=0.20). These results suggest that isolates harboring the rt250L mutations should be considered resistant to entecavir, whereas isolates harboring the rt173L mutations should be considered to display reduced susceptibility to entecavir. An integrative approach to antiviral drug resistance in HBV would provide a more accurate assessment of entecavir treatment failures and help to improve the accuracy of genotypic testing algorithms.