Dolutegravir linked to less sleep but not cardiovascular or other CNS side effects, or IRIS
Polly Clayden, HIV i-Base
Meta-analysis of randomised trials presented at EACS 2017 found higher risk of insomnia for dolutegravir (DTG) compared with other antiretrovirals (ARVs). But no difference for other CNS side effects. 
There was also no significant difference in the risk of cardiac serious adverse events (SAEs) between DTG and other ARVs. The risk of IRIS was low but the main trials excluded people with CDC stage C disease.
Observational data suggest higher risks of CNS adverse events for DTG, compared with other ARVs. There have been two case reports of myocarditis in people receiving DTG. And integrase inhibitors have been associated with IRIS in two cohort studies.
In response to these signals, Andrew Hill and Nikkita Mitchell from Liverpool University and Imperial College London, performed a meta-analysis to compare rates of each adverse event for DTG versus other ARVs – stratified by trial. They compared suicidality between DTG and efavirenz (EFV) and non-EFV controls in two separate analyses. This meta-analysis of randomised trials included 6,647 patient-years of follow-up.
For cardiac SAEs, the authors included trials: SINGLE, SAILING, FLAMINGO, SPRING-1, SPRING-2, ARIA, STRIIVING and NEAT SSAT 060. The analysis revealed 15/2,202 (0.7%) participants with cardiovascular SAEs receiving DTG compared with 8/2,215 (0.4%) receiving other ARVs (RR=1.69, NS). Only 1/25 (4%) cardiac SAE in SPRING-1 was considered to be related to DTG; 1 other cardiac SAE in the same trial was considered unlikely to be related. There was additional case information available for 19/23 participants with cardiac SAEs. Of these 17/19 (89%) had underlying cardiac risk factors.
When the authors looked at suicidality in the SINGLE and SPRING-1 trials, there were 5/465 participants with reported SAEs receiving DTG (1.1%) compared with 6/469 (1.3%) receiving EFV (RR=0.87, NS). In the SAILING, FLAMINGO, SPRING-2, ARIA, STRIIVING, SWORD and NEAT SSAT 060 trials, suicidality SAEs were reported for 15/2,250 participants receiving DTG (0.7%) compared with 9/2,257 receiving other ARVs (0.4%) (RR=1.58, NS). The authors found no significant differences in other CNS endpoints between DTG and other ARVs.
The risk of grade 1–4 insomnia was higher for DTG compared with other ARVs: 165/2,716 (6.1%) vs 124/2,727 (4.5%) (RR=1.30, p=0.02).
IRIS was seen in 1/414 participants receiving DTG compared with 2/419 participants receiving EFV in SINGLE, 6/354 receiving DTG compared with 3/361 receiving raltegravir (RAL) in SAILING, and 1/411 DTG compared with 0/411 receiving RAL in SPRING-2. No cases of IRIS were reported in SPRING-1, FLAMINGO, STRIIVING or NEAT SSAT 060. Although there was no significant difference in the risk of IRIS between DTG and other ARVs, none of the randomised trials included people with low CD4 counts where the risk is likely to be elevated.
Importantly the authors reported that the overall risk of adverse events was lower for DTG than EFV in the SINGLE trial, darunavir/ritonavir (DRV/r) in FLAMINGO, and atazanavir/ritonavir (ATV/r) in ARIA.
They added that other completed randomised DTG trials should be included in new safety analyses: DAWNING (n=627), SWORD 1 and 2 (n=1024), Gilead trial 1489 (n=629) and Gilead trial 1490 (n=645). And they stressed the importance of continued pharmacovigilance with regular meta-analysis to monitor safety.
Monitoring of new ARVs is particularly important for DTG, which will be used to treat millions of people in low- and middle-income settings, in programmes that have already begun or will do so in the next couple of years.
Overall these data are reassuring – particularly learning from the authors that the non-significant relative risk for suicidality was way off p=0.05 (test for overall effect: z=0.53, p=0.6).
The risk of IRIS was low, but event rates were low and the main trials excluded those at greatest risk. Data from ongoing closer-to-real-life trials such as ADVANCE and NAMSAL in South Africa and Cameroon will provide more information on IRIS risk with DTG.
This group will continue to re-evaluate DTG safety as data from completed randomised trials becomes available.
Of note, many posters at EACS reported lower rates of dolutegravir-related side effects with fewer differences to other integrase inhibitors.
The Dutch ATHENA cohort reported no differences in discontinuations between dolutegravir and elvitegravir. 
The Italian ICONA cohort of 1057 patients (approximately 600 were naive) reported 2.5% discontinuations due to side effects at one year. 
A UK study reported a discontinuation rate of 2.2% out of 181 (50 naive) using dolutegravir, with sleep changes manged by changing the timing of dosing. 
A German cohort of over 400 people starting dolutegravir-based ART reported 5.8% discontinuations linked to side effects at 27 months. 
Unless stated otherwise, references are to the programme and abstracts of the 16th European AIDS Conference (EACS), 25–27 October 2017. Milan.
- Hill A and Mitchell N. Meta-analysis of safety for DTG versus other ARVs in randomised trials: analysis of cardiovascular, CNS and IRIS endpoints. 16th EACS, 2017, Milan. Poster abstract PE12/17.
- Bollen P et al. Discontinuation of dolutegravir- and elvitegravir-containing cART in the Netherlands: incidence rates and risk factors. 16th EACS, 2017, Milan. Poster abstract PE12/10.
- Mondi A et al. Effectiveness and discontinuation rate of dolutegravir (DTG)-based regimens as either first line or switch antiretroviral therapy (ART): data from the Icona Cohort. 16th EACS, 2017, Milan. Poster abstract PE9/83.
- Migliorini D et al. Experience in using dolutegravir in experienced and treatment naïve patients attending two HIV clinics in Northamptonshire UK. 16th EACS, 2017, Milan. Poster abstract PE12/3.
- Lüftenegger D et al. Tolerability and persistence of dolutegravir-based regimens: second interim analysis of the prospective multicenter DOL-ART cohort. 16th EACS, 2017, Milan. Poster abstract PE9/52.