Dolutegravir: integrase resistance and the impact of the background regimen
Simon Collins, HIV i-Base
Two oral presentations and a poster provided updated information on the integrase inhibitor dolutegravir.
Results from the VIKING-3 study, in 183 treatment experienced patients with resistance to integrase inhibitors plus at least two other classes and viral load >400 copies/mL on current treatment. The study design continued the background regimens for eight days, switching a current integrase inhibitor to dolutegravir 50 mg twice-daily, and then optimising the background regimen for the next 24 weeks. 
This was an advanced group with median baseline CD4 of 140 cells/mm3 (range 19 – 1110), a median of 13 years previous ART (range 0.3 – 25 years) and 56% CDC Class C. In addition to integrase resistance at screening (n=124) or historical (n=59), at least two major mutations to RTIs and PIs were held by 79% and 70% of participants respectively, with 75% having >1 NNRTI major mutation. Prior ART included darunavir (73%), etravirine (56%), T-20 (49%) and maraviroc (32% – 62% having non-R5 tropic virus detected). Demographics included 23% women, 27% African-American and 21% HBV or HCV positive.
At day 8, mean viral load reduced by 1.4 log copies/mL (95% CI: 1.3, 1.5; p <0.001). At week 24, 63% patients with endpoint data (72/114) had suppressed viral load to <50 copies/mL, with 32% (37/114) non-responders and 5% (5/114) discontinuing due to side effects.
Response rates at both time points was closely related to the prior pattern integrase resistance, with lower responses reported in people with mutations on the Q148 pathway (approximate -1.0 log reductions at day 8 and 48% suppression to <50 copies/mL at week 24, if Q148 + two or more secondary mutations were also present, see Table 1.
Virologic response rates at week 24 were inversely correlated with to presence of Q148 + 2 additional primary mutations, and with fold change in sensitivity to dolutegravir but was not related to overall susceptibility score (OSS) of the drugs in the optimised background regimen, see Table 2. This is very unusual for a new drug used in treatment-experienced patients. As a mechanism for this is difficult to understand these finding will need to be confirmed with larger patient numbers.
|Primary INI mutations at baseline||N||mean (SD) VL change at day 8||% VL <50 week 24|
|Q148 + < 1 sec. mutations*||32||–1.1 (0.51)||69%|
|Q148 + > 2 sec. mutations*||20||–1.0 (0.81)||48%|
|> 2 primary mutations||8||–1.4 (0.76)||75%|
|No primary mutations||60||–1.6 (0.55)||95%|
* Secondary mutations:G140A/C/S, L741, E138A/K/T
|Derived integrase mutation group||OSS = 0||OSS = 1||OSS >2||Total|
|No Q148 *||100% (2/2)||83% (24/29)||76% (31/41)||79% (57/72)|
|Q148 + 1 **||100% (2/2)||43% (3/7)||36% (4/11)||45% (9/20)|
|Q148 + 2+**||50% (1/2)||0 (0/7)||0||11% (1/9)|
* 143, 156, 66, 96, historical evidence only
** G140A/C/S, E138A/K/T, L74I
People with only historical integrase resistance are likely to drive viral response at Day 8 higher as selective pressure is likely to take longer than one week to return, so it was significant that this group achieved one of the highest response rates at week 24. The breakdown of resistance patterns bypathway was not presented for these patients.
VIKING-3 also reported that overall drug susceptibility score (OSS) to the background regimen was not associated with week 24 response with 83%, 63%, 59% and 69% achieving <50 copies/mL for OSS 0, 1, 2 and >2, respectively.
Discontinuations due to side effects were uncommon, reported by 3% of participants (6/183). The most common drug-related side effects were diarrhoea, nausea and headache, each reported in 5% of patients.
Secondly, a poster was presented that combined 48 week results from the Phase 3 SPRING-2 and SINGLE studies of dolutegravir in treatment-naive patients was notable for the stratification of responses by baseline viral load and background nucleosides, given that abacavir/3TC is currently not recommended in guidelines at VL >100,000 copies/mL (though contended by GlaxoSmithKline). 
SPRING-2 randomised 822 patients to dolutegravir 50 mg once-daily or raltegravir 400 mg twice-daily, plus investigator-selected NRTIs (tenofovir/FTC or abacavir/3TC) finding dolutegravir non-inferior. SINGLE randomised 833 patients to once-daily regimens of either dolutegravir 50 mg + abacavir/3TC or efavirenz/tenofovir/FTC (Atripla) finding the integrase combination superior based on fewer side effects.
While cross study comparisons of different combinations in different populations have only limited value, pulling results from high baseline viral load may be useful in the context of the considerable caution reported in ACTG-5202. These results are detailed in Table 2.
|Dolutegravir + RTIs||Raltegravir + RTIs||Dolutegravir + abacavir/3TC||Efavirenz/tenofovir/FTC|
|Baseline viral load <100,000 copies/mL|
|ABC/3TC||115/132 (87%)||110/125 (885)||253/280 (90%)||—|
|TDF/FTC||152/165 (92%)||154/170 (91%)||—||236/288 (83%)|
|Baseline viral load >100,000 copies/mL|
|ABC/3TC||30/37 (81%)||32/39 (82%)||111/134 (83%)||—|
|TDF/FTC||64/77 (83%)||55/77 (71%)||—||100/131 (76%)|
Results were also presented from SPRING-2 on responses by futher stratification at viral load >100k copies/mL. In the <100k group viral suppression to <50 copies/mL at week 48 was achieved by 88% (225/257) vs 91% (306/335) in the abacavir/3TC vs tenofovir/FTC groups respectively. Responses were 86% (36/42) vs 82% (72/88) ar 100k-250k, 81% (13/16) vs 76% (29/38) in the 250k-500k and 72% (13/18) vs 64% (18/28) in the >500k groups respectively. The numbers of patients in these analysis become very low.
Finally, a second oral presentation reported results from in vitro passaging of increasing doses of dolutegravir from 0.05 nM (25% of the EC50) to 50-100 nM over six months, in MT-2 cells and PBMCs. 
The most common integrase mutations varied by HIV subtype: R263K followed by H51Y in subtype B and A/G; G118R followed by H51Y in subtype C.
R263K alone resulted in approximately 3 -6-fold level of resistance in culture, a 30% drop in levels of recombinant integrase strand transfer activity, as well as an approximate 20 -30% loss in viral replicative capacity.
H51Y had little impact alone, but together withR263K together with H51Y increased dolutegravir resistance to 15-fold, with approximately 50% loss in both viral replication capacity and integrase strand transfer activity.
The lack of differences in the high viral load analyses (though these were smaller numbers than ACTG 5202) nevertheless show some support that in a clinical trial setting, when dolutegravir is used as the third drug, abacavir/3TC does not show the same potency issues as with NNRTIs or protease inhibitors.
This is important given the single pill formulation of dolutegravir/abacavir/3TC that is already in Phase 3 studies.
- Nichols G et al. Antiviral activity of dolutegravir in subjects with failure on an integrase inhibitor-based regimen: week 24 phase 3 results from VIKING-3. 11th International Congress on Drug Therapy in HIV, 11-15 November 2012, Glasgow. Oral abstract O232.
- Eron J et al. Dolutegravir treatment response by baseline viral load and NRTI backbone in treatment-naive HIV-infected individuals. 11th International Congress on Drug Therapy in HIV, 11-15 November 2012, Glasgow. Poster abstract P204.
- Mesplede T et al. Selection in culture of HIV resistance to dolutegravir by mutations at integrase positions R263K and H51Y that diminish viral replication fitness. 11th International Congress on Drug Therapy in HIV, 11-15 November 2012, Glasgow. Oral abstract O233.