Prevalence and Associated Factors of Neurocognitive Impairment in HIV-Positive Patients on Effective Efavirenz/Emtricitabine/Tenofovir Disoproxil Fumarate Treatment

Editor: Neurocognitive impairment (NCI) is still a common complication in HIV-infected patients. ¹ The frequency of NCI in HIV-positive patients successfully treated (HIV load <40 copies/mL for at least 6 months) with fixed-dose efavirenz/emtricitabine/tenofovir disoproxil fumarate (EFV/FTC/TDF) was assessed. This antiretroviral combination is still one of the most commonly used worldwide and the latest WHO guidelines ² recommend it as first-line therapy. Baseline therapeutic drug monitoring of EFV plasma levels, complete neuropsychologic examination, and adherence assessment were performed. The Frascati criteria ³ were applied to classify the impairment, and the neuropsychological z-score for eight tests (NPZ8 score) was calculated. Overall, 127 HIV-positive patients were enrolled: 116 (91.3%) were men, median age 43 years (interquartile range [IQR] 38–50); nadir and baseline CD4 cells were 300/mm³ (IQR 244–373) and 684/mm³ (IQR 554–852), respectively (Supplementary Table S1; Supplementary Data are available online at www.liebertpub.com/aid); >95% adherence was detected in 93% patients. NCI was found in 25 (19.7%) patients: 2 subjects with mild NCI, 15 asymptomatic NCI, and 8 with “non-HIV-related impairment.” The latter condition was defined as any neurocognitive impairment not fulfilling the Frascati criteria. The prevalence of HIV-associated neurocognitive disorders (HANDs) was 13.4% (11.8% symptomatic HANDs). At multivariable logistic regression, NCI was significantly associated with older age (odds ratio [OR] 2.34 per 10 years increase; 95% confidence interval [CI] 1.13–4.85, p = .02) and lower educational level (OR 0.69 per 1-year increase; 95% CI 0.57–0.84, p < .001). The probability of having NCI was significantly lower in patients with >95% adherence (odds ratio [OR] 0.27, 95% CI 0.13–0.57, p = .05). Considering the 17 patients with HANDs, NCI was significantly associated with education only (OR 0.74, 95% CI 0.61–0.90, p = .03). At univariate analysis carried out on the whole study population, a significant correlation between higher EFV plasma levels and worse neurocognitive performance was observed (Fig. 1). At multivariable linear regression, a better NPZ8 score was significantly associated with higher educational level (beta = 0.06, 95% CI 0.02–0.10, p = .008 per 1-year increase), whereas a worse performance was more weakly associated with higher EFV plasma levels (beta = −0.06, 95% CI −0.12 to −0.01, p = .024 per each 500 ng/mL increase). The latter association was also observed, but to a lesser extent, when considering the 17 HAND cases only (beta = −0.06, 95% CI −0.12 to 0.00, p = .05 per each 500 ng/mL increase).

OPEN IN VIEWER

FIG. 1.

Association between neurocognitive performance (NPZ8) and efavirenz (EFV) plasma levels in the whole study population.

Some limitations regarding our study need to be mentioned: first, as the study population did not include a control group, comparison of observed neurocognitive performance with that of antiretroviral therapy (ART)-treated subjects not receiving EFV was not possible. Second, only a small number of women were enrolled. However, no significant association was found between gender and NCI, and the gender distribution in the study mirrored proportions found in our HIV/AIDS clinic. Third, it cannot completely be ruled out that distinction of HIV-related and non-HIV-related NCI based on the Frascati criteria may have led to some minor extent of misclassification. Finally, it should be noted that nowadays patients treated with EFV are different than those who received this type of ART in the past. In fact, in most settings in the world, the availability of newer and better tolerated antiretrovirals has led to the selection of subjects with good tolerability of the drug. Nonetheless, the findings of our study remain of major interest for those geographical areas where antiretroviral options are fewer.

We found that, among HIV-infected patients on successful combined antiretroviral therapy (cART) with EFV/FTC/TDF, almost 20% had minor forms of NCI, which in two-thirds were related to HIV. This prevalence of HANDs is lower than those previously reported, ¹ but it is noteworthy because our study population was on stable cART, had a durable control of HIV in plasma, and reported high adherence to antivirals. Education was the only factor independently associated with HIV- and non-HIV-related NCI. Of note furthermore, higher EFV plasma concentrations were associated with NCI and correlated with lower NPZ8 score. EFV is one of the available antiretrovirals with better central nervous system (CNS) penetration but concerns about its neurotoxicity are emerging. ⁴ In fact, as high EFV plasma concentrations carry an increased risk of CNS adverse reactions, EFV plasma concentration monitoring may be considered as a useful tool to reduce adverse events occurrence.