Highly active antiretroviral treatment in HIV infection:... : AIDS (original) (raw)
Introduction
Two major recent developments in HIV therapeutics, the use of quantitative HIV viral load testing and HIV-1 protease inhibitors, have greatly altered the course of HIV illness. Protease inhibitor regimens have been associated with marked reductions in peripheral viral load, elevations in CD4 cell counts and improved clinical outcome [1]. These developments have raised important questions regarding the neuropathological effects of HIV infection. One critical question is whether antiretroviral regimens including protease inhibitors, which may have low levels of blood–brain barrier (BBB) penetration, will benefit neuropsychological function [1]. The current recommendations for treatment of HIV infection with highly active antiretroviral therapy (HAART), including potent protease inhibitors, provide a framework for evaluating the effects of such treatment on neuropsychological function [2].
It has been assumed that antiretroviral medications must penetrate the BBB in order to exert a therapeutic effect on cognitive function, particularly since elevated cerebrospinal fluid (CSF) and brain levels of HIV RNA have been associated with neuropsychological impairment and dementia [3,4]. Some nucleoside analog reverse transcriptase inhibitors (NRTI), particularly zidovudine, traverse the BBB in sufficient quantity to reduce central nervous system (CNS) viral load and to improve neuropsychological function [5,6]. Despite the limited ability of the protease inhibitors to cross an intact BBB, evidence of gradual disruption of the BBB as HIV illness progresses [7] suggests that these medications may also enter the CNS in sufficient quantity to have a neuroprotective effect. However, contradictory findings regarding the integrity of the BBB suggest that the CNS is a ‘sanctuary site’ for HIV replication in the context of protease inhibitor therapy [3,8].
Given questions regarding the permeability of the BBB to protease inhibitors and the effect of HAART regimens on the neuropathological effects of HIV, concurrent data on neuropsychological functioning and viral load in the context of these treatments are needed. The current study was designed to investigate whether HIV-infected individuals taking HAART are less likely to show neuropsychological impairment compared with individuals taking less potent antiretroviral regimens.
Methods
Sample
Data for the current study were derived from the 18-month assessment in a longitudinal natural history study of homo-/bisexual men initiated in July 1995 to examine factors associated with psychosocial adjustment to AIDS. The study sample and procedures are described in more detail elsewhere [9]. Subjects were recruited from advertisements in community-based organizations and by word of mouth. While individuals at all stages of HIV infection were included in the sample, a particular effort was made to recruit those with symptomatic illness.
Measures
Laboratory measures
Laboratory tests were performed by Corning-Metpath (now Quest) Laboratory, New York, New York, USA, a widely used regional commercial service. Assays to determine CD4 lymphocyte count and peripheral HIV RNA viral load were performed. The lower limit of detection of the HIV RNA assay was 2.6 log10 copies/ml.
Medical symptom checklist
The medical symptom checklist adapted from earlier HIV-positive cohort studies [10], consists of 14 signs and symptoms commonly but not exclusively associated with HIV infection, such as night sweats, weight loss, and oral candidiasis. The total score measures number of physical symptoms currently present.
Depressive symptoms
The Beck Depression Inventory [11] was used as a self- report measure of depression. To avoid confounding somatic symptoms due to depression and HIV, only the 15 cognitive items were used.
Psychotropic medications
Because commonly-used psychotropic medications may affect cognitive performance, the use of antide- pressants, anxiolytics and psychostimulants was recorded.
Neuropsychological assessments
Neuropsychological tests, assessing the domains of attention, concentration, learning, memory, psychomotor speed and executive function were administered and scored by postdoctoral fellows in a National Institute of Mental Health-funded HIV research training fellowship (under the supervision of W.v.G.). The tests administered were the Trail-making tests A and B [12], the California Verbal Learning Test [13], the Grooved Pegboard Test [14], the Digit Symbol subtest of the Weschler Adult Intelligence Scale-revised [15], and the Stroop Color Word Test [16]. In addition, the North American Adult Reading Test [17] was given to a subset of 98 subjects to provide an estimate of pre- morbid intelligence. Subjects in the study cohort had not previously received neuropsychological testing.
For data analysis, neuropsychological test scores were used in two ways: raw scores were used as continuous measures, and subjects were also rated as neuropsychologically impaired or unimpaired based on their performance relative to published normative data. Subjects were classified as neuropsychologically impaired if they scored ≥ 2 SD in the impaired direction relative to age-matched population-based norms [13,15,16,18] on two or more neuropsychological tests.
Subjects were also assessed for factors other than HIV that may affect neuropsychological performance, including limited English fluency, learning disability, concurrent substance abuse, history of head trauma with sustained loss of consciousness (> 30 min), CNS opportunistic infections, neurosyphilis, seizures, meningitis or encephalitis.
Antiretroviral treatment potency
In order to compare neuropsychological test scores and neuropsychological impairment by antiretroviral regimen potency, subjects were divided into two antiretroviral potency groups based on the most recent treatment recommendations of the International AIDS Society–USA Panel [2]. Subjects taking two or more NRTI with or without nevirapine or delavirdine (non-NRTI) with ritonavir, indinavir or nelfinavir for 1 month or more were classified as taking HAART (n = 69). Subjects taking other regimens were classified as HAART-negative. The validity of this potency classification was supported by the fact that over the 18 months between study entry and neuropsychological assessment, HAART-positive subjects had a significantly greater mean reduction in HIV RNA than HAART-negative subjects [HAART-positive: mean reduction, 1.4 log10 copies/ml; HAART-negative: mean reduction, 0.51 log10 copies/ml; F = 23.25, degrees of freedom (df) = 1, P < 0.0001].
Procedure
Between January and June 1997, subjects were seen for about 3 h, during which self-rating scales, neuropsychological assessments and blood tests were completed. Medical information was elicited by clinicians under the supervision of a physician who worked in the hospital HIV/AIDS clinic (S.J.F.). Written informed consent was obtained under the Institutional Review Board guidelines of Cornell University Medical College. Subjects were paid US$ 40 for each study visit.
Statistical analysis
HAART groups were compared on demographic, laboratory, depressive and HIV symptom measures, neuropsychological test scores and prevalence of neuropsychological impairment. Between-group comparisons of categorical variables were performed using χ2 analysis. Multivariate analysis of variance (MANOVA) was used to test for overall differences between HAART groups on neuropsychological test scores, followed by analysis of variance (ANOVA) to investigate differences on individual neuropsychological tests. Logistic regression was used to determine the relative contribution of demographic characteristics, laboratory markers, depressive and HIV symptom measures, psychotropic medication use and HAART category to neuropsychological impairment.
Results
Demographics and HIV illness measures
The 130 subjects had a mean age of 41 years (SD, 8), 42% were non-white (20% Hispanic, 16% African American, 4% mixed race, 1% Asian, 1% Native American) and 87% had some post-high-school education. Mean CD4 cell count of the sample was 296 × 106/l (SD, 227), mean HIV RNA was 3.8 log10 copies/ml (SD, 1 log10), and 106 (82%) subjects had an AIDS diagnosis. All but three men were candidates for HAART under current treatment guidelines [2]. Sixty-nine (53%) patients were taking HAART and 61 (47%) were taking less potent regimens. There were no significant differences in demographic characteristics (each group reflecting the entire sample), estimated premorbid intelligence or depressive symptoms between the HAART-positive and HAART-negative groups (Table 1). There were no group differences with respect to the frequency of antidepressant, anxiolytic or psychostimulant use. Whereas HAART-positive subjects had significantly lower mean CD4 cell counts and were more likely to have AIDS than HAART-negative subjects, those taking HAART were not more symptomatic and were more than twice as likely to have undetectable viral load.
Demographic and HIV disease characteristics by highly active antiretroviral therapy (HAART) group and for the total sample.
Frequency and duration of antiretroviral treatment
Of subjects taking any protease inhibitors, 58% were taking indinavir, 20% were taking saquinavir, 14% were taking ritonavir, and 8% were taking nelfinavir, which had just been released. Some subjects were taking more than one protease inhibitor. Of the 69 HAART-positive subjects, 18 took HAART for 1–6 months (five for 1–2 months, the rest for 3–5 months), 40 took HAART for 6–12 months, and 11 took HAART for > 12 months. Of the 61 HAART-negative men, 20 were taking none or one of the antiretrovirals, 27 were taking two antiretrovirals, nine were taking three NRTI or two NRTI and saquinavir, and five were taking two or more NRTI with one or more non-NRTI.
Neuropsychological function by HAART group
In the HAART-negative group, 45% were neuropsychologically impaired, compared with 22% of the HAART-positive group (χ2 = 14.6, df = 1, P < 0.0001; Table 2). The MANOVA of combined neuropsychological test means by HAART group was highly significant (Hotelling's trace: F = 3.6, df = 10, P < 0.0001). ANOVA of individual neuropsychological test means by HAART group (Table 2) revealed that HAART-positive subjects performed significantly better than HAART-negative subjects on the California Verbal Learning Test total, long-delay free recall and learning slope subtests, as well as the Grooved Pegboard and Digit Symbol tests. To ensure that HAART group differences on neuropsychological tests could not be accounted for by differences in CD4 cell count or HIV RNA levels, analyses were repeated using these measures as covariates. Results were essentially unchanged (data not shown). Thus, fewer HAART-positive subjects than HAART-negative subjects were impaired in the domains of attention, concentration, learning, memory and psychomotor speed.
Neuropsychological test performance by highly active antiretroviral therapy (HAART) group and for the total sample.
The duration of HAART beyond 1 month of treatment did not significantly affect the prevalence of neuropsychological impairment: four (22%) of those taking HAART for 1–6 months were neuropsychologically impaired, compared with eight (20%) taking HAART for 6–12 months, and three (27%) taking HAART for > 12 months (χ2 = 0.272, df = 2, P = 0.872).
HAART groups were compared on the prevalence of potential non-HIV neurocognitive confounders. In addition, comparisons of HAART groups on demographic characteristics, HIV illness measures, neuropsychological impairment and neuropsychological test scores were repeated after excluding subjects with these factors (n = 34). HAART-positive and HAART-negative subjects did not differ on the prevalence of potential neuropsychological confounders (Table 1). Of the 96 subjects without potential confounders, 52 (54%) were taking HAART and 32 (33%) were neuropsychologically impaired, similar proportions to the entire sample. In this subsample, all HAART group comparisons did not differ from those seen for the entire sample (data not shown).
Given the lack of baseline neuropsychological test data, group comparisons were conducted on laboratory and medical data from the baseline assessment 18 months prior to neuropsychological testing. At study baseline, only one subject was taking HAART and two subjects had undetectable virus. HAART-positive compared with HAART-negative subjects had significantly lower mean CD4 cell counts at baseline (156 ± 138 versus 356 ± 261 × 106/l; unequal variances, −5.3, df = 85, P < 0.0001) and higher mean viral load levels (5.1 ± 0.83 log10 versus 4.5 ± 0.92 log10 copies/ml; t = 3.8, df = 123, _P_ < 0.0001). Furthermore, CD4 cell counts for HAART-positive subjects rose significantly from base-line to the neuropsychological assessment (baseline, 156 ± 138 × 106/l; neuropsychological assessment, 256 ± 164 × 106/l; t = −6.56, df = 67, _P_ < 0.001), whereas mean CD4 cell count for HAART-negative subjects was unchanged (baseline, 356 ± 261 × 106/l; current assessment, 335 ± 274 × 106/l; t = 0.935, df = 59, not significant). Although both HAART-positive and HAART-negative subjects experienced significant reductions in mean viral load (HAART-positive: baseline, 5.1 ± 0.83 log10 copies/ml; current assessment, 3.7 ± 1.1 log10 copies/ml; t = 10.2, df = 66, _P_ < 0.0001; HAART-negative: baseline, 4.5 ± 0.92 log10 copies/ml; current assessment, 4.0 ± 1.0 log10 copies/ml; t = 4.3, df = 57, _P_ < 0.0001), the mean 1.4 log10 reduction for HAART-positive subjects (1.1 log10 for HAART 1–6 months, 1.7 log10 for HAART 6–12 months, 0.8 log10 for HAART > 12 months) was significantly greater than the mean 0.51 log10 drop for HAART-negative subjects (F = 23.25, df = 1, P < 0.0001).
Relationship of demographic and HIV disease variables to neuropsychological status
To determine whether neuropsychological impairment was related to demographic factors, degree of immuno-suppression, peripheral viral load, or psychotropic medication use, neuropsychologically unimpaired subjects (n = 82) were compared with neuropsychologically impaired subjects (n = 48) on age, ethnicity, education, mean CD4 cell counts, mean HIV RNA levels, proportion with undetectable virus and proportion taking antidepressants, anxiolytics or psychostimulants. The neuropsychologically unimpaired subjects were more likely than impaired subjects to have some post-highschool education (94 versus 75%; χ2 = 9.5, df = 1, P < 0.01) and less likely to be taking anxiolytic medications (4 versus 25%; χ2 = 13.5, df = 1, P < 0.0001). In repeated analyses on the 96 subjects without neuropsychological confounders, neuropsychologically unimpaired (n = 64) compared with neuropsychologically impaired subjects (n = 32) had significantly lower mean HIV RNA levels (3.6 ± 1.1 log10 versus 4.0 ± 1.0 log10 copies/ml; F = 5.3, df = 1, P < 0.05), had a higher proportion with undetectable virus (46 versus 18%; χ2 = 8.4, df = 1, P < 0.01), were more likely to be white (70 versus 43%; χ2 = 6.36, df = 1, P < 0.02), and were again more educated and less likely to be taking anxiolytics.
Predictors of neuropsychological impairment
The logistic regression model predicting neuropsychological impairment by demographics (age, ethnicity, education), depressive symptoms, HIV illness markers (CD4 cells, HIV RNA levels), psychotropic use and HAART grouping had an overall predictive accuracy of 78% (χ2 = 46.1, df = 8, P < 0.0001). Significant predictors of neuropsychological status were educational level (Wald statistic = 5.5, df = 1, P < 0.02), ethnic minority status (Wald statistic = 4.8, df = 1, P < 0.03), anxiolytic medication use (Wald statistic = 10.3, df = 1, P < 0.01) and HAART grouping (Wald statistic = 14.3, df = 1, P < 0.001). The model for the 96 subjects without neuropsychological confounders had a predictive accuracy of 85% (χ2 = 45.5, df = 8, P < 0.0001). While ethnicity and anxiolytic use were again significant predictors of neuropsychological status in this model, HAART status remained a highly significant predictor.
Discussion
The current findings provide preliminary evidence to suggest that HAART, compared with less potent antiretroviral regimens, may have a superior neuroprotective benefit in HIV. The specific domains of cognitive function that were superior for subjects taking HAART included attention, concentration, learning, memory and psychomotor speed, all of which may be affected by HIV [19] and may be improved with NRTI treatment [5,6]. Even after accounting for the effects of ethnicity, education and anxiolytic medication use, HAART continued to be a robust predictor of neuropsychological status.
The benefits of HAART on neuropsychological function may be seen as early as 1–6 months after therapy initiation, concurrent with reductions in plasma viremia. In fact, the lower rate of neuropsychological impairment seen among subjects taking HAART compared with those taking less potent regimens existed in the context of reduced peripheral viral load, despite more advanced immunosuppression both crosssectionally and 18 months prior to neuropsychological assessment.
Possible explanations for these findings are that potent protease inhibitors (indinavir, ritonavir, nelfinavir) in HAART regimens penetrate an intact BBB more than previously thought, that they traverse a compromised BBB [8], or that even low CNS concentrations of the protease inhibitors benefit neuropsychological function. Alternatively, reduced peripheral viral load in the context of HAART may decrease HIV trafficking into the CNS, thus diminishing the neuropathological effects of the virus. This is supported by the fact that subjects in this study without potential neuropsychological confounders and without neuropsychological impairment had lower peripheral viral load levels and were more likely to have undetectable viral load than those without neuropsychological confounders but with neuropsychological impairment. In this context, peripheral viral load measurement may play a role in the clinical monitoring of the risk for neuropsychological impairment in symptomatic HIV illness. However, this contrasts with earlier reports that cognitive impairment in advanced HIV illness is independent of plasma viral load [3]. Clearly, this issue warrants further study.
Several limitations to this study must be considered. It is possible that HAART-positive subjects were more cognitively intact prior to antiretroviral treatment and were therefore more likely to be prescribed these complex regimens. In addition, given the naturalistic quality of this study, there is lack of uniformity in the composition and duration of antiviral regimens, and adherence was not systematically assessed. Nonetheless, the current findings are particularly encouraging because they indicate that heterogeneous HAART regimens may exert a neuroprotective effect. Finally, the lack of CSF viral load and antiretroviral quantification make it impossible to clarify the etiology of the HAART effect on neuropsychological function.
The current findings support the need for clinical trials investigating the impact of HAART regimens on neuropsychological status. Such studies would optimally include pre- and post-treatment neuropsychological testing, functional status assessments, CSF and peripheral viral load assays, and CSF antiretroviral quantification.
Acknowledgements
The authors thank the study participants, S. Cohen (study coordinator), and Z. Rosen for his statistical assistance.
References
1. Deeks SG, Smith M, Holodniy M, Kahn JO: HIV-1 protease inhibitors: a review for clinicians. JAMA 1997, 277:145–153.
2. Carpenter CCJ, Fischl MA, Hammer SM, et al.: Antiviral therapy for HIV infection in 1997: updated recommendations of the International AIDS Society–USA Panel. JAMA 1997, 277:1962–1969.
3. Ellis RJ, Hsia K, Spector SA, et al.: Cerebrospinal fluid human immunodeficiency virus type 1 RNA levels are elevated in neurocognitively impaired individuals with acquired immunodeficiency syndrome. Ann Neurol 1997, 42:679–688.
4. McArthur JC, McClernon DR, Cronin MF, et al.: Relationship between human immunodeficiency virus-associated dementia and viral load in cerebrospinal fluid and brain. Ann Neurol 1997, 42:689–698.
5. Schmitt FA, Bigley JW, McKinnis R, et al.: Neuropsychological outcome of zidovudine (AZT) treatment of patients with AIDS and AIDS-related complex. N Engl J Med 1988, 319:1573–1578.
6. Brouwers P, Hendricks M, Lietzau JA, et al.: Effect of combination therapy with zidovudine and didanosine on neuropsychological functioning in patients with symptomatic HIV disease: a comparison of simultaneous and alternating regimens. AIDS 1997, 11:59–66.
7. Power C, Kong P, Crawford TO, et al.: Cerebral white matter changes in acquired immunodeficiency syndrome dementia: alterations in the blood–brain barrier. Ann Neurol 1993, 34:339–350.
8. Krivine A, Esmilaire L, Lebon P, et al.: HIV-RNA in CSF and permeability of the blood-brain barrier. Fourth Conference on Retroviruses and Opportunistic Infections. Washington, DC, January 1997 [abstract 143].
9. Rabkin JG, Ferrando SJ, Jacobsberg LB, Fishman B: Prevalence of Axis I disorders in an AIDS cohort: a controlled study. Comp Psychiatry 1997, 38:146–154.
10. Rabkin JG, Williams JB, Remien RH, Goetz R, Kertzner R, Gorman JM: Depression, distress, lymphocyte subsets and human immunodeficiency virus symptoms on two occasions in HIV positive homosexual men. Arch Gen Psychiatry 1991, 48:111–119.
11. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J: An inventory for measuring depression. Arch Gen Psychiatry 1961, 4:53–63.
12. Reitan R: Validity of the Trail Making Test as an indication of organic brain damage. Percept Mot Skills 1958, 8:271–276.
13. Delis DC, Kramer JH, Kaplan E, Ober BA: California Verbal Learning Test, Research Edition. Cleveland: Psychological Corporation; 1983.
14. Lezak MD: Neuropsychological Assessment, 3rd Edn. New York: Oxford University Press; 1983:683–684.
15. Wechsler D: Wechsler Adult Intelligence Scale–Revised. San Antonio: The Psychological Corporation/Harcourt Brace Jovanovich; 1985:32–33.
16. Golden CJ: Stroop Color and Word Test: A Manual for Clinical and Experimental Uses. Illinois: Stoetling Company; 1978.
17. Blair JR, Spreen O: Predicting premorbid IQ: a revision of the National Adult Reading Test. Clin Neuropsychologist 1989, 3:129–136.
18. Bornstein RA: Normative data on selected neuropsychological measures from non-clinical samples. J Clin Psychol 1985, 41:651–659.
19. Janssen RS and the working group of the American Academy of Neurology AIDS Task Force: Nomenclature and research case definitions for neurologic manifestations of human immunodeficiency virus-type 1 (HIV-1) infection. Neurology 1991, 41:778–785.
Keywords:
HIV; AIDS; neuropsychological function; protease inhibitors; highly active antiretroviral therapy
© Lippincott-Raven Publishers.