Neurological side effects with integrase inhibitors cause low rates of discontinuation in clinical practice
Simon Collins, HIV i-Base
An analysis from use of integrase inhibitors in clinical practice reported significantly higher rates of discontinuations related to side effects than seen in clinical studies and includes neurological complications with dolutegravir.
This was a retrospective analysis presented in an oral presentation at the 18th International Workshop on Comorbidities and Adverse Drug Reactions in HIV by Esteban Martinez from University of Barcelona. 
The study included antiretroviral naive and experienced patients who received a first integrase-inhibitor based ART with at least one follow up visit, with the analysis looking for those who switched treatment within the first year.
Baseline characteristics of 1061 patients in the overall cohort included mean age 45, >80% men, and >50% were gay men. Median CD4 count at diagnosis was >350 (IQR approximately 200 to 550) cells/mm3. Sensitivity analyses were run to allow for the great use of earlier integrase inhibitors.
The incidence of side-effect related discontinuations was 12.7% for raltegravir (71/557), 8/1% for elvitegravir (26/332) and 12/3% for dolutegravir (26/212). Based on incidence of 270, 167 and 263 events per 1000 patient years of follow up for raltegravir, elvitegravir and dolutegravir respectively, the unadjusted IRR compared to raltegravir was 0.62 (95%CI 0.39 to 0.97) for elvitegravir and 0.97 (95%CI 0.62 to 1.52) for dolutegravir, showing non-significant differences between drugs (p=0.821).
A further analysis for each drug compared discontinuations for naive compared to experienced patients. Although there were not statistically significant differences for raltegravir (unadjusted IRR 0.64 [95%CI 0.36 to 1.12, p=0.10) or elvitegravir (unIRR 1.66 [95%CI 0.53 to 2.53], p=0.70), with dolutegravir, treatment-experienced patients were more likely to discontinue compared to those who were naive (unIRR 3.11 [95%CI 1.03 to 9.39], p=0.03).
Although just over one-third of discontinuations overall (n=44/123) were due to side effects rather than other reasons, there were significant differences between drug: 28% with raltegravir (20/71), 62% with elvitegravir (16/26) and 31% with dolutegravir (8/26); p=0.008). So although elvitegravir was discontinued less frequently, when it was stopped this was more likely to be related to side effects than other reasons.
When categorised by type of side effect, multiple organs were involved but reports of CNS-related symptoms (disorientation, mood changes, sleep disturbance) were notable for similarity to side effects related to efavirenz, albeit at a much lower incidence. See Table 1.
This was reported with each drug but was significantly more important with dolutegravir (88%; 7/8) compared to raltegravir (35%; 7/20) and elvitegravir (19%; 3/16); p=0.005.
In adjusted analysis, only older age associated with overall discontinuations for any reason (adj HR 1.04 (95%CI 1.02 to 1.07; p=0.0007).
|Raltegravir (n=20; 3.6%)||Elvitegravir (n=16; 5.0%)||Dolutegravir (n=8; 3.8%)||p-value|
|CNS (n=17)||7 (35%)||3 (19%)||7 (88%)||0.005|
|Muscular (n=12)||3 (15%)||6 (38%)||3 (38%)||0.244|
|Digestive (n=11)||7 (35%)||4 (25%)||0||0.179|
|Skin/mucoses (n=6)||4 (20%)||2 (13%)||0||0.456|
|Systemic (n=5)||2 (10%)||0||3 (38%)||0.224|
|Number of organ systems|
|1||17 (85%)||16 (100%)||5 (63%)||0.066|
|>1||3 (15%)||0||3 (37%)|
Although discontinuations were higher than in clinical trials, integrase inhibitors still generally have fewer side effects compared to drugs in other classes. This study is important for highlighting the range of problems that can occur.
All drugs look safest when first approved because by definition this occurs when data are limited to only several thousand people. Post marketing reports of side effects nearly always show more complicated problems with wider use.
Anecdotal reports of neurological problems with dolutegravir shortly after approval were sufficient for i-Base to add these side effects in online patient information. This was largely reported from people who switched away from dolutegravir within the first weeks of treatment.
A similar sized study at IAS 2015 reported discontinuation rates of 4.6% for raltegravir (24/522), 8.6% for elvitegravir (26/301) and 3.1% for dolutegravir (9/299). 
A letter published ahead of print in September 2016 in AIDS also reported higher rates of discontinuation of dolutegravir compared to clinical trials due to side effects. 
In this Dutch cohort of 556 patients, 102 of which were ARV-naive. Overall, 85/556 patients (15.3%) stopped dolutegravir, none due to virological failure, with 76/85 (13.7%) due to intolerability. Main symptoms were: insomnia and sleep disturbance (5.6%), gastrointestinal complaints (4.3%) and neuropsychiatric symptoms e.g. anxiety, psychosis and depression (4.3%). Dolutegravir was switched more frequently in combinations that included abacavir (adjusted RR 1.92 95%CI 1.09-3.38, p-log-rank 0.01).
- Padilla M et al. Tolerability of integrase inhibitors in clinical practice. 18th International Workshop on Comorbidities and Adverse Drug Reactions in HIV, 12-13 September 2016, New York. Oral abstract O25.
- Lepik KJ et al. Adverse drug reactions associated with integrase strand transfer inhibitors (INSTI) in clinical practice: post-marketing experience with raltegravir, elvitegravir-cobicistat and dolutegravir. IAS 2015, Toronto. Poster abstract TuPEB258.
- de Boer M et al. Intolerance of dolutegravir containing cART regimens in real life clinical practice. AIDS (2016). Published online: September 24, 2016. doi: 10.1097/QAD.0000000000001279.