The HIV-associated neurocognitive disorders (HAND) are regarded as a direct manifestation of HIV-1 infection of the central nervous system (CNS). They are characterized by acquired impairment in cognitive functioning, disturbances in memory, attention, processing speed, and fine motor functions.1 Neuropsychologically, there is a characteristic impairment in psychomotor speed, memory, and sustained attention, and neuropsychological (NP) test batteries have been routinely included as primary outcome criterion to monitor neurological response to highly active antiretroviral treatment (HAART) in most published trials.2,3 The optimal treatment for HAND has not been established, but there is strong evidence that HAART can improve the cognitive dysfunction.4,5 The general consensus is that the ability of antiretroviral (ARV) drugs to reach the CNS is a critical factor for patients' neurological response.6 Several ARV drugs have been shown to be more neurologically active than others on the basis of their capacity to reduce cerebrospinal fluid (CSF) HIV RNA to undetectable levels7,8 and improve cognition.2,3 Moreover, it has been shown that virologic suppression in the CSF is associated with significant NP improvement.9 However, although HAND should be treated with ARV drugs with good central nervous system penetration (CP), how to identify a neuroactive HAART regimen remains controversial. A number of cross-sectional and prospective studies have investigated the effects of neuroactive HAART regimens on CSF viral load (VL) and on cognitive functions. In almost all published studies, zidovudine, stavudine, abacavir, lamivudine (3TC), indinavir, efavirenz, and nevirapine were considered ARV drugs with good CP,9-16 generally on the bases of their CSF concentration from human studies. CSF is closer to the brain parenchyma than any other body fluid, and CSF ARV drug concentration is relatively easy to measure. However, CSF drug concentration can vary widely between individuals. Moreover, how well CSF drug concentration reflects drug concentration within the brain parenchyma is unclear. Clinical studies are confined to examining drug concentration in human CSF, and we know little about brain penetration of ARV drugs. Recently, when trying to develop a clinically useful approach to estimating CNS effectiveness, the CHARTER Study Group proposed a new ranking system based not only on CSF drug concentration but also on the chemical properties of ARV drugs and on clinical studies on drug ability to reduce CSF VL and improve cognition.17,18 For these reasons, we evaluated changes in cognition during HAART as a function of 2 different scoring systems to measure ARV drug efficacy in treating HAND: the CP reference score derived from the existing literature9-16 and the CNS penetration-effectiveness (CPE) score proposed by the CHARTER Study Group.17,18 We hypothesized that neuroeffectiveness as calculated with the CPE CHARTER score could be more likely to show an association with changes in NP performance than the alternative CP reference score.
Study Design and Setting
This was an observational cohort study of patients followed at the National Institute for Infectious Diseases Lazzaro Spallanzani, (Rome, Italy), which provides care for HIV infection to more then 4000 patients. The study was conducted in the context of usual clinical care according to an internal protocol approved by the local ethics committee. Between June 1999 and December 2006, patients with HAND or at risk for HAND treated with HAART underwent serial NP test batteries as part of their clinical management.
Patients were eligible if they had either symptoms of suspected HAND or symptomatic HIV infection and/or CD4 cell count below 200/μL. Exclusion criteria were current or past opportunistic infections or tumors of the CNS, non-HIV-related major neurological or psychiatric disorders, current use of illicit drugs, or sedative-hypnotics. After excluding potentially confounding clinical conditions that could affect cognition through brain magnetic resonance imaging studies, neurological examinations, and laboratory testing, patients were referred for NP evaluation.
Patients underwent NP test batteries before HAART initiation or HAART change and at follow-up. The HAART regimen was prescribed as best judged by the primary medical care provider according to national guidelines.
All subjects underwent comprehensive baseline and follow-up clinical evaluations. A general physical examination was performed every 3 months. A general medical history was taken, and current and past medications were recorded. Laboratory measures were performed every 3 months and were targeted to detect HIV-related complications, drug-related adverse events, and any clinical event that could affect neurocognitive functions. CD4 cell counts and plasma HIV-1 RNA were also performed every 3 months. Virological response to HAART was defined as durable suppression if the patient achieved and maintained plasma VL below 50 copies per milliliter or virological failure if the patient either failed to achieve undetectable VL or experienced a confirmed virological rebound to more than 1000 copies per milliliter after having achieved undetectable VL.
A detailed NP assessment was performed through a battery of 15 standardized NP tests administered by a trained neuropsychologist (P.B.). Tests were selected to include a spectrum of different cognitive domains: mental flexibility, concentration and speed of mental processing, memory, visuospatial and constructional abilities, and fine motor functioning. The NP scores resulting from each test were transformed into normatively adjusted z scores as previously reported.19 The following composite z scores were obtained: global NPZ-8 score, global NPZ-4 score, memory z score, concentration and speed of mental processing z score, mental flexibility z score, and fine motor functioning z score.19 Each z score was adjusted so that negative values indicated below-average performance. Patients were classified as neurocognitively impaired or unimpaired based on their performance relative to age and sex-adjusted normative data. Neurocognitive impairment was defined as performing below 1 SD from the normative mean on at least 2 NP tests or 2 SDs below the mean on at least 1 test. The diagnosis of neurocognitive impairment also required the exclusion, by neurological examinations, laboratory measures, and brain imaging studies, of other conditions that could explain the finding.19 The American Academy of Neurology criteria20 were used to determine whether the degree of impairment met the criteria for HIV-associated dementia. NP assessments were performed at baseline and repeated at follow-up as best judged by the primary medical care provider. For the purpose of this study, only patients who had at least 3 serial NP testing examinations were included.
Changes in NP Test Scores and Main Outcome Measures
Changes in summary NPZ-8, NPZ-4, memory, concentration and speed of mental processing, mental flexibility, and fine motor functioning z scores from baseline to the first and the second NP follow-up examination were calculated for each subject so that positive values indicated improvement and negative values indicated deterioration in cognitive functions. Changes in NP testing between baseline and first and second follow-up NP examination were considered as the primary outcome criteria.
HAART Regimens and CP Scores
Patients' HAART regimens were reviewed, and 2 CNS HAART penetration scores were calculated for each patient: the CP reference score and the CPE score proposed by the CHARTER Study Group.
The CP reference score defined as a summary score of 1 (high penetration: zidovudine, stavudine, abacavir, 3TC, indinavir, efavirenz, and nevirapine)9-16 and 0 (low penetration: remaining ARV drugs) for each individual drug in the combination. Because 3TC was not considered as a neuropenetrating drug in 2 studies,9,13 an additional analysis was performed ranking 3TC as a nonpenetrating drug. Moreover, an additional analysis was performed ranking lopinavir/ritonavir as a penetrating drug.
According to Letendre et al,17,18 the CPE CHARTER score was defined as a summary score of 1 (high penetration: zidovudine, abacavir, delavirdine, nevirapine, amprenavir-ritonavir, fosamprenavir-ritonavir, atazanavir-ritonavir, indinavir-ritonavir, and lopinavir-ritonavir), 0.5 (intermediate penetration: stavudine, 3TC, emtricitabine, efavirenz, amprenavir, fosamprenavir, atazanavir, and indinavir), and 0 (low penetration: tenofovir, didanosine, zalcitabine, nelfinavir, saquinavir, saquinavir-ritonavir, ritonavir, tipranavir-ritonavir, and enfuvirtide) for each individual ARV drug in the combination.
For patients who underwent changes in the HAART scheme due to treatment switches, CP scores were first calculated for each individual HAART combination. Subsequently, composite CP scores were obtained by adjusting the score of each individual combination for the duration of exposure to that combination.
Results using the 2 scoring systems were compared with changes from baseline to the first and the second follow-up NP examination in normatively adjusted NP z scores. For the purpose of this analysis, regression analyses were used to assess the association of changes over the study period in NP test z scores with the 2 different CNS HAART penetration scores as calculated from HAART regimens received by each individual patient. A regression analysis was also used to assess the association of changes in CD4 count and plasma VL with the 2 CNS HAART penetration scores over the study period.
A total of 305 consecutive patients with symptoms of suspected HAND or with symptomatic HIV infection and/or CD4 cell count below 200/μL were eligible. Patients with current or past opportunistic infections or tumors of the CNS (n = 42), with non-HIV-related major neurological or psychiatric disorders (n = 10), and those using illicit drugs or sedative-hypnotics (n = 68) were excluded. The remaining 185 patients represent this study group.
Baseline Patient Characteristics
Baseline demographic and clinical characteristics are summarized in Table 1. Of the 185 patients included in this analysis, 138 subjects (74.6%) were men, 69 (37.3%) had Centers for Disease Control and Prevention stage C disease, and 93 (50.3%) had impaired NP performance, meeting the criteria for HIV dementia in 11 cases (5.9%). Ten patients (5.4%) had experienced 1 previous HAART regimen for a median of 18 months and the remaining 175 subjects (94.6%) were HAART naive.
HAART Regimens and CP Scores
Patients received standard 3-drug ARV regimens as prescribed by the medical care provider, according to current guidelines. Only 1 regimen with 3 nucleoside reverse transcriptase inhibitor (NRTI) regimens, zidovudine-3TC-abacavir, was used and received by 8 patients. The remaining regimens were 1 protease inhibitor plus 2 NRTIs and 1 nonnucleoside reverse trancriptase inhibitor plus 2 NRTIs. Among protease inhibitors, the most commonly used drug was indinavir, which was prescribed to 90 patients (48.6%). Among nonnucleoside reverse trancriptase inhibitors, the most commonly used drug was efavirenz, which was prescribed to 61 subjects (33.0%). Between baseline and the first and between baseline and the second follow-up NP evaluation, patients received a mean of 1.8 (range 1-2) and 2.8 (range 1-17) different HAART regimens, respectively. The first and the second follow-up NP examinations occurred after a mean of 20 (SD ± 5.6) months and after a mean of 39 (SD ± 6.2) months, respectively.
Between baseline and the first follow-up NP examination and between baseline and the second follow-up NP examination, mean CP reference scores were 1.53 (range 0-3) and 1.56 (range 0-3), respectively. Between baseline and the first follow-up NP examination and between baseline and the second follow-up NP examination, mean CPE CHARTER scores were 1.65 (range 0-3) and 1.55 (range 0-3), respectively.
Changes in NP Test, CD4, and Plasma VL From Baseline to Time Points
Changes in NP test, CD4 count, and plasma VL from baseline to time points are reported in Table 2. Significant improvements in NPZ-4 global scores were observed both at the first (+0.44; P = 0.0141) and second follow-up (+0.53; P = 0.0369) NP examination. Moreover, composite memory z scores improved significantly at the second follow-up NP examination (+0.39; P = 0.0107). Patients also showed a significant increase in CD4 cell count and a significant decrease in plasma HIV RNA at both time points. At the first and second follow-up NP examination, 72 and 76 patients had detectable plasma VL. Overall, 76 of 185 patients showed evidence of virological failure, whereas 109 patients achieved and maintained durable virological suppression.
Correlation Between the 2 CNS HAART Penetration Scores and Changes in NP Test, CD4, and Plasma VL
The correlation between the 2 CNS HAART penetration scores and changes in NP test z scores, CD4 count, and plasma VL over the study period are reported in Table 3. Higher CPE CHARTER scores, consistent with greater neuroeffectiveness, were correlated to greater improvements at the first and follow-up NP examination in concentration and speed of mental processing (r 2 = 0.0306; P = 0.0264) and mental flexibility (r 2 = 0.0268; P = 0.0442) summary z scores. Similarly, higher CPE CHARTER scores were correlated to greater improvements at the second follow-up NP examination in NPZ-4 (r 2 = 0.0280; P = 0.0283), NPZ-8 (r 2 = 0.0486; P = 0.0071), concentration and speed of mental processing (r 2 = 0.0571; P = 0.0046), and mental flexibility (r 2 = 0.0345; P = 0.0262) summary z scores (Table 3) (Fig. 1). By contrast, higher estimates of neuroeffectiveness, as calculated by the alternative ranking system, the CP reference score, did not show such correlation at any time point (Table 3) (Fig. 1). The same negative results were obtained when, within the CP reference scoring system, 3TC was considered as a nonpenetrating drug and/or when lopinavir/ritonavir was considered as a penetrating drug (data not shown in tables).
Finally, no association was seen between CD4 cell increase and plasma VL reductions with both the CPE CHARTER and the CP reference scores at any time point (Table 3).
Correlation Between the 2 CNS HAART Penetration Scores and Changes in NP Test in Impaired and Unimpaired Subjects
We further investigated whether the correlation between HAART penetration scores and changes in cognition differed between impaired and unimpaired subjects. These results are reported in Table 4. Among the 93 NP-impaired patients, higher CPE CHARTER scores, consistent with greater neuroeffectiveness, were correlated to greater improvements at the first follow-up NP examination in NPZ-4 (r 2 = 0.0759; P = 0.022), NPZ-8 (r 2 = 0.0943; P = 0.012), concentration and speed of mental processing (r 2 = 0.1160; P = 0.007), and mental flexibility (r 2 = 0.1497; P = 0.004) summary z scores. Similarly, higher CPE CHARTER scores were correlated to greater improvements at the second follow-up NP examination in NPZ-4 (r 2 = 0.2293; P = 0.004), NPZ-8 (r 2 = 0.1705; P = 0.001), concentration and speed of mental processing (r 2 = 0.1271; P = 0.003), and mental flexibility (r 2 = 0.1511; P = 0.003) summary z scores (Table 4). By contrast, higher estimates of neuroeffectiveness, as calculated by the alternative ranking system, the CP reference score, did not show such correlation at any time point (data not shown in tables).
Among the 92 NP-unimpaired subjects, higher CPE CHARTER scores, consistent with greater neuroeffectiveness, were correlated to greater improvements at the first follow-up NP examination (r 2 = 0.0459; P = 0.027) and at the second follow-up NP examination (r 2 = 0.0562; P = 0.017) in fine motor functioning z scores (Table 4). By contrast, higher estimates of neuroeffectiveness as calculated by the alternative ranking system, the CP Reference score, did not show such correlation at any time point (data not shown in tables).
Finally, no association was seen between the CPE CHARTER score and CD4 cell increase and plasma VL reductions in both NP-impaired and NP-unimpaired patients (data not shown in tables).
Correlation Between the 2 CNS HAART Penetration Scores and Changes in NP Test in Patients With and Without Virological Suppression
We further investigated the correlation between HAART penetration scores and changes in cognition in patients with and without virological suppression. Among the 109 patients with virological suppression, higher CPE CHARTER scores, consistent with greater neuroeffectiveness, were correlated to greater improvements at the first follow-up NP examination in NPZ-4 (r 2 = 0.0754; P = 0.014), NPZ-8 (r 2 = 0.0763; P = 0.014), concentration and speed of mental processing (r 2 = 0.0929; P = 0.008), and fine motor functioning (r 2 = 0.0449; P = 0.049) summary z scores. Similarly, at the second follow-up NP examination, higher CPE CHARTER scores were correlated to greater improvements in NPZ-4 (r 2 = 0.0693; P = 0.016) and NPZ-8 (r 2 = 0.0619; P = 0.026) summary z scores (Table 4). By contrast, higher estimates of neuroeffectiveness as calculated by the alternative ranking system, the CP reference score, did not show such correlation at any time point (data not shown in tables). Among the 76 patients who were not virologically suppressed, no correlation between both HAART penetration scores and changes in cognition were observed at any time (Table 4).
Forty-six of 76 HAART-failing patients had genotypic drug resistance testing data available. For patients with drug resistance data available, the CPE CHARTER score was recalculated by ignoring drugs for which resistance had developed. However, even after adjusting for drug resistance data, no correlation between both HAART penetration scores and changes in cognition were observed at any time (data not shown in tables).
There is a general consensus that the primary objective in treating patients with HAND is to maximally suppress HIV replication within the CNS and that CNS HAART penetration is crucial to achieve this goal.6 However, there is no widely accepted approach to estimating CNS HAART penetration. Several studies have addressed the relationship between penetration of ARV drugs into the CNS and neurocognitive response.9-13 In these studies, zidovudine, stavudine, abacavir, indinavir, efavirenz, nevirapine,9-13 and 3TC10-12 were considered ARV drugs with good CP. These studies provided only partly overlapping definitions of CNS-penetrating drugs because 3TC was considered as a neuropenetrating drug in 3 out10-12 of 5 of them.9-13 Generally speaking, the overall theorized neuroeffectiveness of the regimen was estimated by counting the number of penetrating drugs. Overall, these studies gave conflicting results. An association between neuropenetrating HAART regimens and improvement in cognitive functions was reported in 3 of9-11 5 studies,9-13 regardless of the inclusion of 3TC among the neuropenetrating ARVs.
A potential explanation of these inconsistent results is the absence of a widely accepted definition of CNS HAART penetration. To increase our understanding of the impact of CP of ARV drugs within the CNS and neurocognitive response to HAART, we evaluated changes in cognition during HAART as a function of 2 different scoring systems to measure ARV drug efficacy in treating HAND. We compared the CPE ranking system proposed by the CHARTER Study Group with the system that was generally used in previously published analyses9-16 and that we called, for our convenience, the CP reference score. The main result of our study was that higher CPE scores, consistent with greater neuropenetration as assessed by the ranking system proposed by the CHARTER Study Group, were associated with greater improvements in NP performance. Notably, the association with changes in cognitive performance was seen over the entire study period.
By contrast, higher estimates of neuropenetration scores calculated with the alternative CP reference score did not show such correlation over the entire study period.
Furthermore, we assessed whether the correlation between HAART penetration scores and changes in cognition differed between impaired and unimpaired subjects. We found that in NP-impaired patients higher CPE scores were associated with greater short-term and long-term improvements in NP performance in most z scores. By contrast, among unimpaired subjects, the association was seen only in measures exploring fine motor functions. This observation suggests that the CPE scoring system might be clinically relevant, especially in impaired patients.
Moreover, we assessed the correlation between HAART penetration scores and changes in cognition according to virological response to HAART. Interestingly, the correlation between CPE score improvements in NP performance was noted exclusively among patients who achieved sustained virological suppression. This observation suggests that the identification of a neuroactive HAART regimen could represent a treatment goal in patients with virological suppression.
Finally, no association was found between CD4 cell count increase and plasma VL decrease with either CP scores over the entire study period. These data suggest that the use of HAART regimens with higher neuropenetration score could have a positive effect on cognitive functions without influencing the systemic effects of the regimen.
Some study limitations are worth mentioning. Observational analyses could be hampered by a variety of confounding factors such as severity, duration, and reversibility of cognitive impairment. Moreover, patients were recruited from a single HIV clinic, which may limit the generalizability of our findings to other populations.
In summary, our data indicate that the CPE score proposed by the CHARTER Study Group could represent a step forward toward identifying a clinically useful approach in estimating HAART ability to improve cognition.
Our data could be useful in identifying CNS-targeted neuroactive HAART regimens particularly for patients with HAND. Prospective data in this field are warranted.
The authors thank Andrea Baker for editing the article. Author Contributions: Valerio Tozzi: study design and performance, data analysis, data interpretation, primary article preparation, and article review and editing. Pietro Balestra: NP assessment, NP data analysis, and interpretation. Maria Flora Salvatori: statistical analysis, data analysis, and data interpretation. Chrysoula Vlassi: study performance and data analysis. Giuseppina Liuzzi: study performance and data analysis. Maria Letizia Giancola: study performance and data analysis. Marinella Giulianelli: study performance and data analysis. Pasquale Narciso: study design and performance, data analysis, data interpretation, and primary article preparation. Andrea Antinori: study design and performance, data analysis, data interpretation, primary article preparation, and article review and editing.
1. Antinori A, Arendt G, Becker JT, et al. Updated research nosology for HIV-associated neurocognitive disorders. Neurology
2. Sidtis JJ, Gatsonis C, Price RW, et al. Zidovudine treatment of the AIDS dementia complex: results of a placebo-controlled trial. AIDS Clinical Trials Group. Ann Neurol
3. Price RW, Yiannoutsos CT, Clifford DB, et al. Neurological outcomes in late HIV infection: adverse impact of neurological impairment on survival and protective effect of antiviral therapy. AIDS Clinical Trial Group and Neurological AIDS Research Consortium study team. AIDS
4. Ferrando S, Van Gorp W, McElhiney M, et al. Highly active antiretroviral treatment in HIV infection: benefits of neuropsychological function. AIDS
5. Tozzi V, Balestra P, Galgani S, et al. Positive and sustained effects of highly active antiretroviral therapy on HIV 1-associated neurocognitive impairment. AIDS
6. McArthur JC, Brew BJ, Nath A. Neurological complications of HIV infection. Lancet Neurol
7. Polis MA, Suzman DL, Yoder CP, et al. Suppression of cerebrospinal fluid HIV burden in antiretroviral naive patients on a potent four-drug antiretroviral regimen. AIDS
8. Marra CM, Lockhart MS, Zunt JR, et al. Changes in CSF and plasma HIV-1 RNA and cognition after starting potent antiretroviral therapy. Neurology
9. Letendre SL, McCutchan JA, Childers ME, et al. Enhancing antiretroviral therapy for human immunodeficiency virus cognitive disorders. Ann Neurol
10. Cysique LA, Maruff P, Brew BJ. Antiretroviral therapy in HIV infection: are neurologically active drugs important? Arch Neurol
11. Cysique LA, Maruff P, Brew B. Variable benefit in neuropsychological function in HIV-infected HAART-treated patients. Neurology
12. Giancola ML, Lorenzini P, Balestra P, et al. Neuroactive antiretroviral drugs do not influence neurocognitive performance in less advanced HIV-infected patients responders to highly active antiretroviral therapy. J Acquir Immune Defic Syndr
13. Evers S, Rahmann A, Schwaag S, et al. Prevention of AIDS dementia by HAART does not depend on cerebrospinal fluid drug penetrance. AIDS Res Hum Retroviruses
14. Eggers C, Hertogs K, Stürenburg HJ, et al. Delayed central nervous system virus suppression during highly active antiretroviral therapy is associated with HIV encephalopathy, but not with viral drug resistance or poor central nervous system drug penetration. AIDS
15. Antinori A, Giancola ML, Grisetti S, et al. Factors influencing virological response to antiretroviral drugs in cerebrospinal fluid of advanced HIV-1 infected patients. AIDS
16. De Luca A, Ciancio BC, Larussa D, et al. Correlates of independent HIV-1 replication in the CNS and of its control by antiretrovirals. Neurology
17. Letendre S, Marquie-Beck J, Capparelli E, et al. Validation of the CNS penetration-effectiveness rank for quantifying antiretroviral penetration into the central nervous system. Arch Neurol
18. Clifford DB. HIV-associated neurocognitive disease continues in the antiretroviral era. Top HIV Med
19. Tozzi V, Balestra P, Bellagamba R, et al. Persistence of neuropsychological deficits despite long-term HAART in patients with HIV-related neurocognitive impairment. Prevalence and risk factors. J Acquir Immune Defic Syndr
20. American Academy of Neurology. Nomenclature and research case definitions for neurologic manifestations of human immunodeficiency virus-type 1 (HIV-1) infection: report of a Working Group of the American Academy of Neurology AIDS Task Force. Neurology
Keywords:© 2009 Lippincott Williams & Wilkins, Inc.
central nervous system; highly active antiretroviral therapy; HIV dementia; neurocognitive disorders