Antiviral Therapy 2015; 20: (doi: /IMP2926)

Transcription

1 Antiviral Therapy 2015; 20: (doi: /IMP2926) Original article Antiretroviral neuropenetration scores better correlate with cognitive performance of HIV-infected patients after accounting for drug susceptibility Massimiliano Fabbiani 1 *, Pierfrancesco Grima 2, Benedetta Milanini 1, Annalisa Mondi 1, Eleonora Baldonero 1, Nicoletta Ciccarelli 1, Roberto Cauda 1, Maria C Silveri 3, Andrea De Luca 1,4, Simona Di Giambenedetto 1 1 Institute of Clinical Infectious Diseases, Catholic University of Sacred Heart, Rome, Italy 2 Division of Infectious Diseases, S Caterina Novella Hospital, Galatina, Italy 3 Memory Clinic, Catholic University of Sacred Heart, Rome, Italy 4 Infectious Diseases Unit, Siena University Hospital, Siena, Italy *Corresponding author massimiliano.fabbiani@gmail.com Background: The aim of the study was to explore how viral resistance and antiretroviral central nervous system (CNS) penetration could impact on cognitive performance of HIV-infected patients. Methods: We performed a multicentre cross-sectional study enrolling HIV-infected patients undergoing neuropsychological testing, with a previous genotypic resistance test on plasma samples. CNS penetrationeffectiveness (CPE) scores and genotypic susceptibility scores (GSS) were calculated for each regimen. A composite score (CPE-GSS) was then constructed. Factors associated with cognitive impairment were investigated by logistic regression analysis. Results: A total of 215 patients were included. Mean CPE was 7.1 (95% CI 6.9, 7.3) with 206 (95.8%) patients showing a CPE 6. GSS correction decreased the CPE value in 21.4% (mean 6.5, 95% CI 6.3, 6.7), 26.5% (mean 6.4, 95% CI 6.1, 6.6) and 24.2% (mean 6.4, 95% CI 6.2, 6.6) of subjects using ANRS, HIVDB and REGA rules, respectively. Overall, 66 (30.7%) patients were considered cognitively impaired. No significant association could be demonstrated between CPE and cognitive impairment. However, higher GSS-CPE was associated with a lower risk of cognitive impairment (CPE-GSS ANRS odds ratio 0.75, P=0.022; CPE-GSS HIVDB odds ratio 0.77, P=0.038; CPE-GSS REGA odds ratio 0.78, P=0.038). Overall, a cutoff of CPE-GSS 5 seemed the most discriminatory according to each different interpretation system. Conclusions: GSS-corrected CPE score showed a better correlation with neurocognitive performance than the standard CPE score. These results suggest that antiretroviral drug susceptibility, besides drug CNS penetration, can play a role in the control of HIV-associated neurocognitive disorders. Introduction Since the introduction of combination antiretroviral therapy (cart), the incidence of HIV-associated dementia (HAD) has dramatically fallen [1]. In contrast, the prevalence of asymptomatic and milder forms of HIV-associated neurocognitive disorders (HAND) still remains high, even in patients on a stable and successful cart [2 4]. Several reasons could explain this observation, such as a lower mortality leading to aging and longer duration of HIV infection [5], comorbidities (HCV coinfection, cardiovascular and cerebrovascular diseases) [6 8], poor adherence [9], drug resistance but also a potential neurotoxicity [10] or a poor central nervous system (CNS) penetration of some antiretroviral agents [11,12]. To optimize cart regimens for patients with HAND, a CNS penetration-effectiveness (CPE) scoring system was first devised in 2008 [11] and then revised in 2010 [12]. It has been demonstrated that regimens including antiretroviral drugs with better CNS penetration, evaluated through the CPE score, are associated with lower viral load in the cerebrospinal fluid (CSF) [12]. Consequently, it has been postulated that regimens with a higher CPE score could be more effective for the treatment of patients with HAND. However, a relation between CPE and neuropsychological performance has not been constantly observed. In particular, some studies have 2015 International Medical Press (print) (online) 441

2 M Fabbiani et al. demonstrated that a higher CPE could be associated with better cognitive performance [13 16], while others have shown the opposite [17 19] or did not demonstrate any association [3,20,21]. Several reasons could explain these controversial findings, such as different study designs, the heterogeneity of the investigated populations, differences in neuropsychological batteries and poor adherence. Moreover, viral resistance could at least partly counterbalance the good neuropenetration of the prescribed drugs. In routine clinical practice a proportion of HIV-infected patients (especially if treatment-experienced) have limited treatment options and harbour viral subpopulations carrying mutations that can confer resistance to the prescribed drugs, thus potentially limiting their neuroeffectiveness. However, viral resistance has been rarely taken into account in previous studies investigating predictors of cognitive impairment. The aim of our study was to explore how viral resistance to antiretroviral drugs, together with drug CNS penetration, could impact on cognitive performance of HIV-infected patients. Methods Patients Patients were retrospectively selected from a database of adult (age 18 years) HIV-infected individuals with an available neuropsychological examination. All these subjects had been enrolled from 2009 to 2013 at two clinical centres in Italy (Rome and Galatina) in the setting of cross-sectional studies evaluating the prevalence of cognitive impairment among the HIV-infected population [6 8,10,16,21]. In order to select a sample representative of the entire clinic population, patients had been enrolled consecutively during routine outpatient visits. Only those characteristics that could lead to misinterpretation of HAND diagnosis were considered as exclusion criteria: history of neurological disorders (including cerebral stroke, head injury and neurological AIDS-defining events, with the exception of HAD), active psychiatric disorders (as major depression or any other psychotic disorder), alcoholism or drug abuse, decompensated liver diseases or cirrhosis, HCV treatment in the past 6 months and linguistic difficulties for non-native patients. All subjects had provided informed consent prior to enrolment. When multiple neuropsychological examinations were available for the same patient, only the first evaluation was considered, in order to avoid biases deriving from learning effect [22]. At the time of enrolment, demographic, clinical and laboratory variables were collected for each subject by patient interview (using a pre-defined form) and chart review. Among patients with an available neuropsychological examination, for the present study we selected only subjects on cart and with an available genotypic resistance test (GRT) performed on plasma samples collected close to neuropsychological testing in patients with detectable viraemia or at the time of last detectable HIV RNA in those with undetectable viraemia. Neuropsychological examination The neuropsychological battery included tasks exploring memory (immediate and delayed recall of Rey s words), attention (digit span forward), speed of mental processing (Wechsler Adult Intelligence Scale [WAIS] Digit Symbol), fine motor functioning (grooved pegboard for dominant and non-dominant hand) and language (letter fluency). The neuropsychological scores resulting from each test were transformed into normatively adjusted Z scores. A global composite Z score and composite Z scores for each domain (memory, attention, speed of mental processing, fine motor functioning and language) were then calculated. According to standard criteria, patients were considered as affected by cognitive impairment if they showed decreased cognitive function (-1 standard deviation [sd]) involving at least two ability domains among those explored: the profiles of asymptomatic neurocognitive impairment (ANI), mild neurocognitive disorders (MND) and HAD were differentiated on the basis of the degree of cognitive impairment and interference with everyday life activities (evaluated through the Instrument Activities of Daily Living [IADL] scale) [23]. The Zung Self-Rating Depression Scale was also administered to measure mood status. Neuropenetration and viral resistance Neuropenetration/neuroeffectiveness was evaluated through the CPE score, calculated for each antiretroviral regimen according to rules proposed in the 2010 revised version [12]. Briefly, a score of 1 (tenofovir, zalcitabine, nelfinavir, ritonavir, saquinavir/ritonavir, saquinavir, tipranavir/ritonavir, enfuvirtide), 2 (didanosine, lamivudine, stavudine, etravirine, atazanavir/ritonavir, atazanavir, fosamprenavir), 3 (abacavir, emtricitabine, delavirdine, efavirenz, darunavir/ritonavir, fosamprenavir/ritonavir, indinavir, lopinavir/ritonavir, maraviroc, raltegravir) and 4 (zidovudine, nevirapine, indinavir/ritonavir) was assigned to each antiretroviral drug according to its increasing neuropenetration or neuroeffectiveness. CPE score was then obtained by summing the scores of each drug in the regimen. Viral resistance was evaluated by interpreting plasma GRT according to ANRS V , HIVDB version and Rega V8.0.2 algorithms, all available through the Stanford HIV database site [24]. The genotypic susceptibility score (GSS) was calculated by assigning a International Medical Press

3 Neuropenetration scores, resistance and cognitive performance score of 0 (resistance), 0.5 (intermediate susceptibility) or 1 (susceptible) to each drug in relationship to the susceptibility predicted by each algorithm. The GSS of the regimen was then calculated by summing the scores of each prescribed drug, as previously described [25]. A composite score accounting for both neuropenetration and viral resistance (CPE-GSS) was then constructed by multiplying CPE and GSS (according to ANRS V , HIVDB version and Rega V8.0.2) for each drug and then summing the scores of drugs in the regimen. Using this approach, CPE- GSS was equal to CPE if the virus was susceptible to each drug in the regimen, but it was lower if a certain degree of resistance to at least one of the prescribed drugs was observed. As an example, in a patient treated with tenofovir/emtricitabine + atazanavir/ritonavir and harbouring a virus with M184V mutation (which confers resistance to emtricitabine), the scores were calculated as follows: CPE = (tenofovir = 1) + (emtricitabine = 3) + (atazanavir/ritonavir = 2) = 6; GSS = (tenofovir = 1) + (emtricitabine = 0) + (atazanavir/ritonavir = 1) = 2; CPE-GSS = [(CPE tenofovir = 1) (GSS tenofovir = 1)] + [(CPE emtricitabine = 3) (GSS emtricitabine = 0)] + [(CPE atazanavir/ritonavir = 2) (GSS atazanavir/ritonavir = 1)] = 3. Statistical analyses Descriptive statistics were calculated for quantitative variables (mean, sd, 95% CI, median, IQR, range) and qualitative variables (absolute and percent frequencies). Categorical variables were compared using c 2 test or, when appropriate, Fisher s exact test. Variables associated with cognitive impairment (both global and in each domain) were investigated by multivariable logistic regression analysis. For each analysis, we performed four separate models where CPE, CPE-GSS ANRS, CPE- GSS HIVDB or CPE-GSS REGA were adjusted for clinically relevant covariates. A two-tailed P-value of less than 0.05 was considered statistically significant. All analyses were performed using the SPSS Version 13.0 software package (SPSS Inc., Chicago, IL, USA). Results Population characteristics Overall, neuropsychological testing was available for 299 patients; among these, 18 and 66 subjects were excluded since they were not taking antiretroviral drugs or did not have an available GRT, respectively. Finally, a total of 215 patients were included in the study; the main characteristics of the enrolled population are reported in Table 1. At the time of neuropsychological examination, the majority of patients (n=121, 56.3%) were on protease inhibitor-based regimens, while 55 (25.6%), 18 (8.4%) and 21 (9.8%) individuals were on non-nucleoside reverse transcriptase inhibitorbased, integrase inhibitor-based and other regimens, respectively. Most subjects (n=189, 87.9%) had HIV RNA<50 copies/ml and median CD4 + T-cell counts were relatively high (545 cells/µl, IQR ). The characteristics of subjects excluded from the study (n=84) were similar to those of patients enrolled (Additional file 1), except for a slightly older age (median 47 versus 44 years; P=0.023) and longer time from HIV infection (median 12.8 versus 9.9 years; P=0.003); specifically, a similar proportion of cognitive impairment was observed in the two groups (23.8% versus 30.7%; P=0.298). Neuropenetration scores Median CPE was 7 (IQR 6 7, range 4 15) with 206 (95.8%) patients showing a CPE 6 and 160 (74.4%) a CPE 7. A total of 46 (21.4%), 57 (26.5%) and 52 (24.2%) subjects were on an antiretroviral regimen including at least one not fully active drug (intermediate or resistant) according to ANRS, HIVDB and REGA algorithms, respectively; as a consequence, in these patients GSS correction decreased the CPE value. Overall, median CPE-GSS ANRS, CPE-GSS HIVDB and CPE- GSS REGA were 7 (IQR 6 7, range 1 11, n=179 [83.3%] with a score 6), 7 (IQR 6 7, range 1 11, n=173 [80.5%] with a score 6) and 7 (IQR 6 7, range 2 11, n=174 [80.9%] with a score 6), respectively. Cognitive performance Scores obtained by HIV-1-infected patients on each neuropsychological test are shown in Table 2. A higher proportion of pathological performances was observed in tasks exploring memory (impairment in delayed and Table 1. Population characteristics (n=215) Characteristic Value Male 166 (77.2) Age, years 44 (38 50) Education, years 13 (8 13) Non-Italian born 19 (8.8) Past injecting drug users 45 (20.9) HCV coinfection 48 (22.3) Past AIDS-defining events 37 (17.2) Time from HIV diagnosis, years 9.9 ( ) Time from starting last cart regimen, months 16 (7 34) Past virological failure 84 (39.1) Time between GRT and cognitive testing, months 35 (15 72) HIV RNA<50 copies/ml 189 (87.9) CD4 + T-cell count, cells/µl 545 ( ) CD4 + T-cell count at nadir, cells/µl 211 (87 296) Values are expressed as n (%) or median (IQR). cart, combined antiretroviral therapy; GRT, genotypic resistance testing. Antiviral Therapy

4 M Fabbiani et al. Table 2. Scores obtained by HIV-infected patients on each neuropsychological test Proportion of impaired Mean Z score (sd) patients, n (%) Memory Immediate recall of Rey words (1.20) 66 (30.7) Delayed recall of Rey words (1.19) 92 (42.8) Attention Digit span (forward) (0.52) 22 (10.2) Speed of mental processing WAIS digit symbol (0.91) 35 (16.3) Fine motor functioning Grooved pegboard dominant hand (1.99) 62 (28.8) Grooved pegboard non-dominant hand (1.88) 51 (23.7) Language Letter fluency (1.16) 51 (23.7) Zung Depression Scale 35.6 (10.5) a 23 (10.7) a Raw scores. WAIS, Wechsler Adult Intelligence Scale. immediate recall of Rey s words 42.8% and 30.7%, respectively), followed by tasks exploring fine motor functioning (impairment in pegboard for dominant and non-dominant hand 28.8% and 23.7%, respectively) and language (impairment in letter fluency 23.7%). Overall, 66 (30.7%) patients were considered cognitively impaired according to standard criteria [23], all showing a profile of ANI (IADL was 7 in all subjects). Investigating each single domain, 33.5% (n=72) of patients were impaired in memory, 25.6% (n=55) in fine motor functioning, 23.7% (n=51) in language, 16.3% (n=35) in speed of mental processing and 10.2% (n=22) in the attention domain. Relation between neuropenetration scores and cognitive performance We explored if the neuropenetration score (CPE), resistance scores (GSS) and scores combining neuropenetration and resistance data (CPE-GSS) were associated with cognitive impairment by means of multivariable logistic regression analysis. We performed seven separate models, each including one of the seven scores (CPE, GSS ANRS, GSS HIVDB, GSS REGA, CPE-GSS ANRS, CPE-GSS HIVDB or CPE-GSS REGA ) adjusted for other clinically relevant variables (age, education, past injecting drug use, HCV coinfection, years from HIV diagnosis, CD4 + T-cell count nadir, HIV RNA<50 copies/ml, Zung depression score) associated with cognitive impairment at univariate analysis (Table 3). No significant independent association could be demonstrated between CPE or any GSS and global cognitive impairment (HAND). However, each higher GSS-corrected CPE was associated with a lower risk of global cognitive impairment. The proportion of cognitively impaired patients according to different CPE-GSS cutoffs is shown in Table 4. Overall, a cutoff of CPE-GSS 5 seemed the most discriminatory according to each different interpretation system. A sensitivity analysis was performed in the subgroup of patients with HIV RNA<50 copies/ml (n=189). In this subpopulation, no relationship could be demonstrated between cognitive impairment and CPE (adjusted odds ratio [aor] 0.85, 95% CI 0.63, 1.14; P=0.270), CPE-GSS HIVDB (aor 0.84, 95% CI 0.65, 1.09) or CPE-GSS REGA (aor 0.83, 95% CI 0.64, 1.08); although not statistically significant, a trend toward an association was observed for CPE-GSS ANRS (aor 0.80, 95% CI 0.61, 1.04; P=0.096). Exploring single cognitive domains, no significant relation between CPE or any CPE-GSS score and impaired performance was observed, although some weak associations were observed between CPE-GSS scores and memory impairment, particularly using resistance interpretation by ANRS. Discussion It has been suggested that regimens including antiretroviral drugs with increased neuropenetration/neuroeffectiveness could be more effective in the treatment of patients with HAND. However, studies investigating this issue have shown controversial results [3,13 21]. Viral resistance could at least partially limit the neuroeffectiveness of antiretroviral drugs. However, few studies have investigated how resistance could impact on cognitive performance of cart-treated patients. In our study, we performed a cross-sectional analysis in HIV-infected patients enrolled during clinical outpatient care. We observed a moderate prevalence of cognitive impairment (30.7%), a proportion slightly lower than that observed in other cohorts (ranging from 35% to 52%, including 4 12% of patients affected by MND/HAD) [2 4,10,26 29]. This could be ascribed to the fact that all patients in our population were neuroasymptomatic and most were virologically suppressed; however, difficulties in accessing outpatient health-care services and the possibility of receiving home care for patients with more severe cognitive dysfunction (a service available in A Gemelli Hospital, Rome, Italy) could be additional explanations. Indeed, in neuro-asymptomatic HIV-infected patients the prevalence of cognitive impairment has been reported as low as 19% [20]. In our population, the most commonly involved domain was memory, in accordance to other reports suggesting a shift toward an increased cortical involvement in the cart era [4,26]. In our population, no association was observed between the standard CPE and cognitive performance, both globally and for single cognitive domains International Medical Press

5 Neuropenetration scores, resistance and cognitive performance Table 3. Association between neuropenetration scores and cognitive impairment (multivariable logistic regression analysis) Speed of Fine motor Global cognitive Attention mental processing functioning Language impairment (HAND) Memory impairment impairment impairment impairment impairment aor aor aor aor aor aor (95% CI) P-value (95% CI) P-value (95% CI) P-value (95% CI) P-value (95% CI) P-value (95% CI) P-value CPE (0.63, 1.10) (0.63, 1.07) (0.88, 1.65) (0.65, 1.20) (0.86, 1.42) (0.67, 1.19) GSS ANRS (0.26, 1.15) (0.32, 1.26) (0.69, 4.12) (0.58, 2.85) (0.92, 4.00) (0.27, 1.22) GSS HIVDB (0.31, 1.17) (0.45, 1.51) (0.52, 2.50) (0.56, 2.27) (0.88, 3.15) (0.38, 1.44) GSS REGA (0.32, 1.26) (0.40, 1.38) (0.59, 2.98) (0.69, 3.13) (0.97, 3.88) (0.28, 1.12) CPE-GSS ANRS (0.58, 0.96) (0.66, 1.03) (0.84, 1.53) (0.72, 1.21) (0.96, 1.55) (0.70, 1.13) CPE-GSS HIVDB (0.61, 0.99) (0.72, 1.10) (0.79, 1.42) (0.73, 1.22) (0.93, 1.48) (0.74, 1.18) CPE-GSS REGA (0.61, 0.99) (0.69, 1.06) (0.78, 1.40) (0.76, 1.25) (0.98, 1.56) (0.69, 1.11) Each model was adjusted for age, education, past injecting drug use, HCV coinfection, years from HIV diagnosis, CD4 + T-cell count nadir, HIV RNA<50 copies/ml and Zung depression score. Bold values represent statistically significant P-values. aor, adjusted odds ratio; CPE, central nervous system penetration effectiveness; GSS, genotypic susceptibility score. Table 4. Proportion of cognitively impaired patients according to different CPE-GSS cutoffs Score Cutoff Impaired < cutoff versus cutoff P-value CPE-GSS ANRS % versus 27.4% % versus 27.8% % versus 27.4% CPE-GSS HIVDB % versus 27.6% % versus 27.8% % versus 27.7% CPE-GSS REGA % versus 27.2% % versus 27.5% % versus 27.6% Bold values represent statistically significant P-values. CPE, central nervous system penetration effectiveness; GSS, genotypic susceptibility score. However, it should be noticed that the standard CPE could probably not reflect the real neuroeffectiveness of the regimens, since nearly a quarter of patients were prescribed at least one antiretroviral drug to which the virus showed a certain degree of resistance. This occurrence is quite common in routine clinical practice, especially in treatment-experienced patients. On this basis, we constructed a composite score, the CPE-GSS, which integrated information about neuropenetration/neuroeffectiveness and viral resistance, estimated by three popular interpretation algorithms. The composite score showed a better correlation with global cognitive impairment. Notably, the type of resistance interpretation did not seem do deeply influence the association with cognitive impairment, although a slightly better association was found between CPE- GSS ANRS and memory impairment. The association between a higher CPE-GSS and lower risk of global cognitive impairment seems to suggest a beneficial role of neuropenetrating drugs on cognitive abilities. In particular, regimens with a CPE-GSS 5 could be prescribed to treat or prevent HAND. The association between CPE-GSS and cognitive impairment could not be demonstrated in patients with suppressed viral load. However, in this subgroup aors were similar to the overall study population and a trend toward an association was observed for CPE-GSS ANRS. It is plausible that the lower number of patients included in this sensitivity analysis could have reduced the power to detect a significant effect of CPE-GSS. Nonetheless, in the total population the association between CPE- GSS and cognitive impairment was demonstrated after adjusting for HIV RNA as a covariate. Our data are apparently in disagreement with a recent large study which assessed close to 62,000 HIVinfected treatment-naive patients starting cart and concluded that cart regimen with a high CPE score increased the risk of HAD [19]. The authors suggested that this unexpected finding could be related to the potential neurotoxicity of antiretroviral drugs; however, viral resistance was not taken into account. In that study, most regimens with high CPE comprised drugs no longer recommended as first-line therapy (for example, nevirapine, indinavir, zidovudine) for lower tolerability, adherence and virological efficacy, which, in turn, can lead to development of resistance. Whether Antiviral Therapy

6 M Fabbiani et al. resistance might explain their findings remains an unresolved issue. Theoretically, the use of CPE-GSS could at least partly overcome this bias. We acknowledge that our study has some limitations because uncontrolled biases can occur in cross-sectional surveys performed in routine clinical practice. Viral resistance was evaluated in plasma samples and this could not accurately reflect HIV-1 resistance in the CNS compartment in each patient: indeed partially independent evolution of viral quasispecies can occur in plasma and CSF with different resistance patterns [30,31]. Evaluation of resistance pattern in CSF could further improve the predictive value of the composite score. However, the study of large populations with an available viral resistance test performed on CSF samples is problematic. Additional controlled longitudinal studies are needed to confirm our findings and before a causal relationship between GSS-CPE scores can be firmly established. In conclusion, GSS-corrected CPE score showed a better correlation with neurocognitive performance than the standard CPE score. The GSS-corrected CPE score could be a simple parameter that integrates both neuropenetration drug properties and viral resistance data; designing regimens with a CPE-GSS 5 could be useful for optimizing the management of patients at risk of or affected by cognitive impairment. However, CPE-GSS needs to be validated in prospective longitudinal studies before it could be used in routine clinical practice. Our results represent a proof of concept that antiretroviral drug susceptibility, besides drug CNS penetration, can play a role in the control of HAND. Acknowledgements No specific funding was received for this study. Preliminary data were presented at the 20th Conference on Retroviruses and Opportunistic Infections, 3 6 March 2013, Atlanta, GA, USA. Disclosure statement MF received speakers honoraria from Bristol Myers Squibb, Merck Sharp & Dohme and Janssen-Cilag. RC has been advisor for Gilead, Janssen-Cilag and Basel Pharmaceutical, received speakers honoraria from ViiV, Bristol Myers Squibb, Merck Sharp & Dohme, Abbott, Gilead and Janssen-Cilag and research support from Fondazione Roma. ADL received speaker s honoraria and fees for attending advisory boards from ViiV Healthcare, Gilead, Abbott Virology, Janssen-Tibotec, Siemens Diagnostics and Monogram Biosciences. SDG received speakers honoraria and support for travel meetings from Gilead, Bristol Myers Squibb, Abbott, Boehringer Ingelheim, Janssen-Cilag and GlaxoSmithKline. All the other authors declare no competing interests. Additional file Additional file 1: A table showing the comparison of clinical characteristics between enrolled patients and those excluded from the study can be found at Fabbiani_Add_file.pdf References 1. d Arminio Monforte A, Cinque P, Mocroft A, et al. Changing incidence of central nervous system diseases in the EuroSIDA cohort. Ann Neurol 2004; 55: Robertson KR, Smurzynski M, Parsons TD, et al. The prevalence and incidence of neurocognitive impairment in the HAART era. AIDS 2007; 21: Simioni S, Cavassini M, Annoni JM, et al. Cognitive dysfunction in HIV patients despite long-standing suppression of viremia. AIDS 2010; 24: Heaton RK, Franklin DR, Ellis RJ, et al. HIV-associated neurocognitive disorders before and during the era of combination antiretroviral therapy: differences in rates, nature, and predictors. J Neurovirol 2011; 17: Antiretroviral Therapy Cohort Collaboration. Life expectancy of individuals on combination antiretroviral therapy in high-income countries: a collaborative analysis of 14 cohort studies. Lancet 2008; 372: Ciccarelli N, Fabbiani M, Grima P, et al. Comparison of cognitive performance in HIV or HCV mono-infected and HIV-HCV co-infected patients. Infection 2013; 41: Fabbiani M, Ciccarelli N, Tana M, et al. Cardiovascular risk factors and carotid intima-media thickness are associated with lower cognitive performance in HIV-infected patients. HIV Med 2013; 14: Grima P, Fabbiani M, Ciccarelli N, et al. Increased ophthalmic artery resistance index is associated with cognitive impairment in HIV-infected patients. J Infect 2012; 65: Lovejoy TI, Suhr JA. The relationship between neuropsychological functioning and HAART adherence in HIV-positive adults: a systematic review. J Behav Med 2009; 32: Ciccarelli N, Fabbiani M, Di Giambenedetto S, et al. Efavirenz associated with cognitive disorders in otherwise asymptomatic HIV-infected patients. Neurology 2011; 76: 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 2008; 65: Letendre S. Central nervous system complications in HIV disease: HIV-associated neurocognitive disorder. Top Antivir Med 2011; 19: Cysique LA, Vaida F, Letendre S, et al. Dynamics of cognitive change in impaired HIV-positive patients initiating antiretroviral therapy. Neurology 2009; 73: Tozzi V, Balestra P, Salvatori MF, et al. Changes in cognition during antiretroviral therapy: comparison of 2 different ranking systems to measure antiretroviral drug efficacy on HIV-associated neurocognitive disorders. J Acquir Immune Defic Syndr 2009; 52: Smurzynski M, Wu K, Letendre S, et al. Effects of central nervous system antiretroviral penetration on cognitive functioning in the ALLRT cohort. AIDS 2011; 25: International Medical Press

7 Neuropenetration scores, resistance and cognitive performance 16. Ciccarelli N, Fabbiani M, Colafigli M, et al. Revised central nervous system neuropenetration-effectiveness score is associated with cognitive disorders in HIV-infected patients with controlled plasma viraemia. Antivir Ther 2013; 18: Marra CM, Zhao Y, Clifford DB, et al. Impact of combination antiretroviral therapy on cerebrospinal fluid HIV RNA and neurocognitive performance. AIDS 2009; 23: Kahouadji Y, Dumurgier J, Sellier P, et al. Cognitive function after several years of antiretroviral therapy with stable central nervous system penetration score. HIV Med 2013; 14: Caniglia EC, Cain LE, Justice A, et al. Antiretroviral penetration into the CNS and incidence of AIDS-defining neurologic conditions. Neurology 2014; 83: Garvey L, Surendrakumar V, Winston A. Low rates of neurocognitive impairment are observed in neuroasymptomatic HIV-infected subjects on effective antiretroviral therapy. HIV Clin Trials 2011; 12: Ciccarelli N, Fabbiani M, Baldonero E, et al. Effect of aging and human immunodeficiency virus infection on cognitive abilities. J Am Geriatr Soc 2012; 60: Bartels C, Wegrzyn M, Wiedl A, Ackermann V, Ehrenreich H. Practice effects in healthy adults: a longitudinal study on frequent repetitive cognitive testing. BMC Neurosci 2010; 11: Antinori A, Arendt G, Becker JT, et al. Updated research nosology for HIV-associated neurocognitive disorders. Neurology 2007; 69: HIValg program. HIV Drug Resistance Database, Stanford University, Stanford. (Accessed 4 February 2014.) Available from JSierra?action=hivalgs Accepted 15 December 2014; published online 17 December De Luca A, Cingolani A, Di Giambenedetto S, et al. Variable prediction of antiretroviral treatment outcome by different systems for interpreting genotypic human immunodeficiency virus type 1 drug resistance. J Infect Dis 2003; 187: Cysique LA, Maruff P, Brew BJ. Prevalence and pattern of neuropsychological impairment in human immunodeficiency virus-infected/acquired immunodeficiency syndrome (HIV/AIDS) patients across pre- and post-highly active antiretroviral therapy eras: a combined study of two cohorts. J Neurovirol 2004; 10: Cysique LA, Maruff P, Brew BJ. Antiretroviral therapy in HIV infection: are neurologically active drugs important? Arch Neurol 2004; 61: Cysique LA, Brew BJ. Prevalence of non-confounded HIV-associated neurocognitive impairment in the context of plasma HIV RNA suppression. J Neurovirol 2011; 17: Heaton RK, Clifford DB, Franklin DR, Jr., et al. HIVassociated neurocognitive disorders persist in the era of potent antiretroviral therapy: CHARTER Study. Neurology 2010; 75: Bergroth T, Ekici H, Gisslén M, Hagberg L, Sönnerborg A. Difference in drug resistance patterns between minor HIV-1 populations in cerebrospinal fluid and plasma. HIV Med 2009; 10: Bierhoff M, Boucher CA, Fibriani A, Ten Kate RW. Ongoing HIV replication in cerebrospinal fluid under successful monotherapy. Antivir Ther 2013; 18: Antiviral Therapy