Abstract & Commentary

Resistance to HIV integrase inhibitors

By Dean L. Winslow, MD, FACP, FIDSA, Chief, Division of AIDS Medicine, Santa Clara Valley Medical Center; Clinical Professor, Stanford University School of Medicine

Synopsis: Integrase inhibitors (INI) have a low genetic barrier to resistance. Raltegravir (RAL) and elvitegravir (EVG) share extensive cross-resistance, whereas S/GSK 1349572 appears to have less cross-resistance. Three common genetic pathways to integrase inhibitor resistance have been identified.

Sources: Blanco JL, Varghese V, Rhee SY, et al. HIV-1 integrase inhibitor resistance and its clinical implications. J Infect Dis 2011;203:1204-1214.

Raltegravir (RAL) received FDA approval in 2007 and is now used in treatment of both treatment-experienced and treatment-naïve patients. Two other integrase inhibitors, elvitegravir (EVG) and S/GSK 1349572, are in late-stage clinical development. In patients receiving RAL, three common pathways associated with INI resistance have been observed: 1) Q148HRK+/-G140SA, 2) N155H+/-E92Q, and 3) Y143CR+/-T97A. It has been observed that INI resistance can be rapidly selected in vitro and virological failure on INI-containing regimens often occurs within the first several months of therapy. Significant cross-resistance between the INIs exists. Q148HRK-substituted viruses generally display a > 150 times increase in IC50 to both RAL and EVG with a 3-8 times increase in IC50 to S/GSK 1349572. 155H-substituted viruses generally display a 10-150 times increase in IC50 to RAL and EVG with a 1-3 times increase in IC50 to S/GSK 1349572. 143CR-substituted viruses show from a 3 to > 150 times increase in IC50 to RAL, but only a 1.5-2 times increased IC50 to EVG and no effect on in vitro susceptibility to S/GSK 1349572.


HIV integrase represents the most 3' gene product of the HIV pol gene. It performs two basic functions: 1) following reverse transcription, integrase (IN) cleaves the conserved GT nucleotides from the 3' ends of the double-stranded HIV-1 cDNA leaving CA overhangs (3' processing reaction), and 2) IN remains bound to the 3' ends of the cDNA, circularizing it, complexes with a host protein, lens epithelial-derived growth factor (LEDGF), translocates to the nucleus where it catalyzes the insertion of the viral 3'-hydroxy ends of the cDNA on to the phosphodiester bonds of the host genomic DNA (strand transfer reaction).

HIV-1 IN integrase inhibitors are structurally diverse molecules that interfere with the strand transfer reaction by binding a divalent metal cation (Mg++ or Mn++) and a hydrophobic region for binding within the catalytic domain, displacing viral DNA in the active site. Crystal structures of IN bound to various inhibitors exist and have aided in drug discovery. Historically, Merck scientists originally identified INI as an attractive potential target for an antiretroviral in the late 1980s. Despite a number of compounds falling out of development for various reasons, Merck persisted and their diketo compound, RAL, eventually received FDA approval in 2007, initially for use in treatment-experienced patients and later for treatment-naïve patients.

RAL has proven to be a very valuable drug over the last 4 years. When combined with other potentially active agents, it often allows patients with multidrug-resistant HIV to fully suppress their HIV RNA levels. It is also a very safe drug and has fewer drug-drug interactions than do HIV protease inhibitors and non-nucleoside reverse transcriptase inhibitors.

One of the minor downsides of RAL is the requirement for twice daily dosing; unfortunately, once-daily dosing of RAL in clinical trials resulted in lower efficacy than twice-daily dosing. Also, when RAL is used in patients in combination with fewer than two other active antiretroviral agents, relatively rapid virological failure ensues and is associated with INI resistance-associated substitutions as described above.

EVG can be dosed once daily, but requires pharmacologic boosting with either ritonavir or Gilead's proprietary boosting agent, cobicistat. These EVG combinations do have the downside of more drug-drug interactions, and although cobicistat causes elevation of serum creatinine, it appears this is a specific tubular effect and does not represent true nephrotoxicity. Unfortunately, this creatinine elevation results in loss of the utility of serum creatinine to be reliably used to assess renal function changes.

The extensive and high-level cross resistance between RAL and EVG suggest that one would not be able to sequence these two agents. While S/GSK 1349572 does not appear to share high-level cross-resistance with RAL and EVG in vitro, there is little clinical experience using the GSK integrase inhibitor following virological failure on RAL or EVG-containing regimens.