Background:  Although integrase inhibitors are highly effective in the management of drug-resistant HIV, some patients fail to achieve durable viral suppression. The long-term consequences of integrase inhibitor failure have not been well defined. 

Methods:  We identified 29 individuals who exhibited evidence of incomplete viral suppression on a regimen containing an integrase inhibitor (24 raltegravir and 5 elvitegravir). Genotypic and phenotypic resistance testing was performed at regular intervals in patients with detectable viremia (>50 copies RNA/mL). 

Results:  Prior to initiating the integrase inhibitor containing regimen, baseline CD4⁺ T cell count and plasma HIV RNA levels were 46 cells/mm³ and 4.61 log₁₀ copies/mL, respectively. Subjects had been on 13 prior antiretroviral drugs (4 prior drug classes). Subjects were followed for a median 13.2 months. At first failure, the most common resistance pattern was wild type (no known mutations), followed in order of frequency by G140S+Q148H/K/R, N155H, and Y143R/C. The G140S+Q148H/K/R resistance pattern typically occurred early, was associated with high-level phenotypic resistance, and remained stable over time. N155H was associated with lower levels of resistance and tended to wane during long-term failure. Y143R/C occurred in the absence of other mutations, was associated with high-level resistance, and tended to be observed after long-term failure. Wild-type failure was common in patients with known partial adherence or in patients with low, intermittent viremia (<1000 copies/mL). Despite evidence of integrase inhibitor failure, patients appeared to have a persistent immunologic benefit, with a median change in CD4⁺ T cell count of +65, +75, +96, and +31 cells/mm³ at months 3, 6, 9, and 12 of documented failure, respectively. However, there was evidence of continued evolution of resistance over time in subjects who remained on a failing integrase inhibitor containing regimen. 

Conclusions:  Long-term evolution of integrase resistance is complex, with various patterns emerging and waning over time.  In this cohort of treatment-experienced subjects failing an integrase inhibitor containing regimen, wild type failure was not uncommon, especially in the context of partial adherence or low-level viremia. Collectively, these data suggest that the genetic barrier to resistance to integrase inhibitors may be higher than previously assumed.