Temporal changes of sleep disturbances and their associations with CYP450 2B6 polymorphism and plasma drug level in HIV patients on efavirenz

Neurological and sleep disturbances were commonly reported among HIV patients on efavirenz (EFV), the pharmacokinetic pattern varies with different CYP450 2B6 G516T genotypes. This prospective study aims to detect temporal changes and differences in the profile of these adverse reactions and their relation to plasma EFV level and host genotypes. HIV patients of Chinese ethnicity on an EFV‐containing HAART regimen were recruited from a specialist clinic. Blood for CYP2B6 G516T genotypes was taken. A questionnaire assessing adverse neurological problems and sleep disturbance was administered, alongside testing for plasma EFV levels at baseline, 4 weeks, 8 months and 1 year intervals after treatment. Analysis was performed using χ 2 and t‐test. A total of 64 patients (31 GG, 27 GT, 6 TT genotypes, 59 male, and 5 female, mean age of 41±9.9) were recruited. At 4 weeks after EFV, 49 (76%) gave a history of any one of the neurological side effects: dizziness, headache and drunk feeling. Sleep disturbances were common: bizarre dream (45%), nightmares (35%), waking at night (73%), poor sleep quality (31%), nocturia (84%) and difficulty in falling asleep (67%). The mean plasma EFV level of GG genotype was 2.8 µg/ml and 3 µg/ml, GT genotype was 3.8 µg/ml and 3.9 µg/ml, at 4 weeks and 1 year respectively. The mean plasma EFV level of TT genotype was 11.9 µg/ml and 9.7 µg/ml at 4 weeks and 1 year respectively. There was no significant variation of drug level within each genotype over time (p>0.08), while EFV level of TT was significantly higher at all time points (p<0.01). Overall, nightmares and difficulty to fall asleep were significantly related to the plasma EFV level (p=0.021 and 0.017 respectively). However, the sleep quality, nocturnal awakening, nocturia or requirement of sleeping pills was not significantly associated with EFV level (p=0.28, 0.06, 0.1 and 0.5 respectively). When the side effects were separately evaluated according to time points, they all became insignificantly related to plasma EFV level at 12 weeks. In conclusion, very high plasma EFV level was observed in TT genotype (9.4% of patients). There was no relationship between genotype GG/GT and the occurrence of neurological side effects. While selected sleep disturbances like nightmares and difficulty to fall asleep were associated with plasma EFV levels, the general sleep quality was not significantly affected. The influence of plasma EFV levels on side effects diminished over time.