Higher Rates of Neuropsychiatric Adverse Events Leading to Dolutegravir Discontinuation in Women and Older Patients

C Hoffmann; T Welz; M Sabranski; M Kolb; E Wolf; H-J Stellbrink; C Wyen


HIV Medicine. 2017;18(1):56-63. 

In This Article


We identified 1704 patients who had initiated 1950 INSTI-based therapies. A total of 228 patients had received two INSTIs and nine had received three INSTIs. In total, 21% (208) of patients were started on an INSTI as first-line therapy. The proportion of patients starting INSTI first-line therapy was higher for elvitegravir (23%) and dolutegravir (21%), compared with raltegravir (13%). Among treatment-experienced patients, the proportion of patients switching their regimen to an INSTI therapy because of neuropsychiatric AEs (almost exclusively caused by efavirenz) was similar for raltegravir (9%) and dolutegravir (8%) but higher for elvitegravir (23%).

Duration of follow-up and reasons for discontinuation are shown in Table 1. As expected, follow-up for raltegravir (European licensing 2007) was longer than for elvitegravir (2013) or dolutegravir (January 2014) and more patients died while on raltegravir (n = 32). No deaths were considered to be drug-related. Discontinuation because of virological failure was very low for all INSTIs. Of patients exposed to raltegravir, almost a third switched to other regimens for simplification.

In total, an AE leading to discontinuation was observed in 122 of 1950 INSTI exposures (6.3%). The estimated rates of any AEs leading to discontinuation within 12 and 24 months, respectively, were 7.6% and 12.3% for elvitegravir (n = 287), 7.6% and 9.3% for dolutegravir (n = 985), and 3.3% and 3.9% for raltegravir (n = 578). Discontinuation rates were highest for elvitegravir (as Stribild©), mainly because of renal AEs, probably attributable to TDF/cobicistat. Neuropsychiatric AEs leading to discontinuation were reported more frequently with dolutegravir. The estimated rates of neuropsychiatric AEs leading to discontinuation within 12 and 24 months were 5.6% and 6.7% for dolutegravir, 0.7% and 1.5% for elvitegravir, and 1.9% and 2.3% for raltegravir, respectively. The Kaplan–Meier curves for INSTI exposure and discontinuations because of neuropsychiatric events are shown in Figure 1.

Figure 1.

Discontinuation because of neuropsychiatric adverse events for all integrase strand transfer inhibitors (INSTIs) within 24 months of initiation (all other events censored). EVG, elvitegravir; RAL, raltegravir; DTG, dolutegravir.

Neuropsychiatric AEs leading to discontinuation among 49 of 985 patients started on dolutegravir were further analysed. The median time between dolutegravir start and discontinuation was 3.1 months and 38 of 49 (78%) patients had stopped dolutegravir within 6 months. The most frequent symptoms (multiple symptoms possible but no temporal sequence documented) included insomnia and sleep disturbances as well as dizziness and painful paraesthesia. No symptoms were life-threatening or led to hospitalization and most symptoms disappeared quickly after discontinuation of dolutegravir. In 32 of 37 (86%) patients followed for at least 3 months after dolutegravir discontinuation, the subsequent antiretroviral regime was tolerated and effective. In six patients who had interrupted dolutegravir, neuropsychiatric AEs recurred in all six cases upon re-exposure. Of note, three of these patients were also simultaneously re-exposed to abacavir.

Table 2 shows the main characteristics of patients on dolutegravir and those discontinuing dolutegravir because of AEs. Considering the entire observation period, 6.8% of patients discontinued dolutegravir due to any AE and 5.0% due to neuropsychiatric AEs. Female patients and patients older than 60 years had an increased risk of dolutegravir discontinuation: 15.7% and 14.3% discontinued for any AE, and 10.0% and 11.0% for neuropsychiatric AEs, respectively. Treatment-experienced patients initiating abacavir at the same time had a higher risk of discontinuation because of any AE (11.5% of patients) or because of neuropsychiatric AEs (9.1% of patients) than patients who did not initiate abacavir simultaneously (P = 0.014 and P = 0.016, respectively). A similar but nonsignificant trend was also seen in treatment-naïve patients. Of note, human leucocyte antigen (HLA)-B*57 testing had been performed in all patients prior to abacavir initiation. In the multivariate Cox regression model, female gender, age > 60 years, simultaneous initiation of abacavir, and dolutegravir initiation in 2016 remained significantly associated with dolutegravir discontinuation (Table 3). These associations remained consistent when patients starting dolutegravir in 2016 were excluded. There were no statistically significant interactions between any of the variables in the final model. There was no association between dolutegravir discontinuation and treatment centre, ethnicity, treatment line (first-line vs. treatment experienced), prior regimen, reason for switch or CD4 T-cell count. Figures 2a–c show the Kaplan–Meier curves reflecting the time on dolutegravir stratified by gender, age group, concomitant initiation of abacavir and treatment line.

Figure 2.

Relationship between neuropsychiatric adverse events (all other events censored) and time on dolutegravir for (a) female (f) vs. male (m) patients, (b) older (> 60 years old) vs. younger patients, (c) patients on abacavir (ABC) vs. those not initiating ABC with dolutegravir, and (d) first-line vs. treatment-experienced (TE) patients (ABC vs. no ABC initiated with dolutegravir). DTG, dolutegravir.