Psych Impairment in Coinfected Women

 

“Neuropsychological functioning in a cohort of HIV & HCV-infected women in WIHS”

 

“….This is the first investigation to evaluate the status of neurocognitive function in a large sample of women who are not infected with HCV or HIV, or who are infected with one or both of these viruses. We found that HCV-positive women were significantly more likely to demonstrate abnormal NP (neuropsychological impairment) results in comparison with HCV-negative women (48.5 versus 31.7%, respectively). These results are consistent with others who have looked at the NP effects of HCV… NP impairment will be a continuing concern in the optimal management of patients with HCV and HIV disease.”

 

AIDS:  Volume 19(15)  14 October 2005  p 1659-1667

 

Richardson, Jean La; Nowicki, Marekb; Danley, Kathleena; Martin, Eileen Mc; Cohen, Mardge Hd; Gonzalez, Raulc; Vassileva, Jasminc; Levine, Alexandra Me
>From the aDepartment of Preventive Medicine
bDepartment of Pediatrics
eDepartment of Medicine, Keck School of Medicine, University of Southern California, Los Angeles, California
cDepartment of Psychiatry, University of Illinois College of Medicine, Chicago
dDepartment of Medicine, Cook County Hospital, Chicago, Illinois, USA

 

“….Of the 220 women, 70 were HCV-positive/HIV-positive…. The risk of NP impairment was higher for HCV-positive women in comparison with HCV-negative women (OR, 2.03)…. and also higher for HIV-positive women in comparison with HIV-negative women (OR, 2.09)…. Using the non-infected group (HIV-negative/HCV-negative) as the control, risks of NP impairment for women mono-infected with either virus were elevated and those co-infected with both were significantly more likely to be abnormal…. Those who were dually infected were more likely to have abnormal overall NP tests results in which the CD4 cell count was over or under 200 cells (OR, 3.48 and OR, 5.38, respectively). In addition, HIV-positive women off ART therapy were more likely (OR, 7.03) to have abnormal NP test results than those not infected with either virus….”

 

Risk factors for overall abnormal NP results
We conducted a series of bivariate analyses to examine the association between HCV positive serostatus and abnormal NP results in combination with HIV disease status, treatment status, clinic site, age, education, estimated verbal IQ, CES-D, sedating medications within 24 h of testing, history of head injury, ethnicity, and lifetime history of substance use. With the exception of age, the odds of impairment associated with HCV remained significant after controlling for these variables. Similarly, after controlling in a multivariate analysis for all of these variables except age, the odds of NP impairment was OR, 3.07 (95% CI, 1.29-7.35) for those co-infected in comparison with those not infected with either virus. However, after adding age to the model, the OR dropped to 1.97 (95% CI, 0.79-4.94). We then tested the interaction between age and HIV/HCV status and found the interaction term was of borderline significance (Wald chi-square = 7.46, P = 0.059). These results suggested that the relationship between age and HIV/HCV status on NP performance required more detailed examination.

 

HCV and age as associated with overall abnormal NP results
To test for the combined effect of co-infection and age, we computed odds ratios and confidence intervals across age and infection status controlling for education, IQ, CESD and ethnicity. These results indicate that abnormal NP performance was related to both age and infection status. Finally, we stratified by age to more clearly understand the interaction. For those aged under 40 years who were HCV negative/HIV positive, and for those HCV-positive/HIV-positive, there was an increased odds of NP impairment in comparison with those uninfected with either virus [(OR, 3.92; 95% CI, 1.15-13.30); OR, 4.67 (95% CI, 1.22-17.84), respectively]. For those aged over 40 years, there were no significant differences in NP impairment based upon infection status, however, the cell sizes were small, thus introducing greater instability.

 

Introduction
Large national prevalence studies have shown that 1.8% of the US population is infected with hepatitis C virus (HCV) [1]. HCV co-infection may be found in 16-30% of HIV-infected patients and in 60-90% of HIV-infected injection drug users [2]. Previous research has documented the prevalence and characteristics of neuropsychological (NP) impairment associated with HIV infection [3-6]. However, few studies have examined NP impairment among those infected with HCV or among those co-infected with both viruses.

 

HCV-infected patients have a lower quality of life and experience more depression and fatigue than controls [7-11], and the few existing studies suggest that persons infected with HCV demonstrate an increase in NP impairment. Cordoba et al. [10] found impaired attention, executive function and motor performance in patients with decompensated cirrhosis. Forton et al. [9,11] showed that HCV-infected patients were impaired on cognitive tasks (i.e. concentration, speed of memory processes) when compared with an HCV-clear group, independent of history of injection drug use (IDU), depression and/or fatigue. Using a sensitive electrophysiologic test of cognitive processing with minimal sociocultural and demographic biases, Kramer et al. [8] found slight but significant neurocognitive impairments in concentration and memory processing among HCV-infected individuals and 17% exhibited grossly abnormal scores. There was no appreciable impact of previous IDU or severity of liver diseases on the relationship between NP impairment and HCV; however, those with abnormal scores were significantly older than those with normal scores.

 

One possibility is that these NP effects are directly related to infection of the central nervous system by HCV. HCV viral sequences and viral replicative forms have been detected in autopsy tissue of brain [12]. HCV has been reported to replicate in macrophages and lymphocytes, particularly in HIV-1-infected patients; these cells could certainly carry the virus from the systemic circulation into the brain [13].

 

NP impairment is greatest among those HIV-infected patients who are not treated with antiretroviral therapy (ART) [5,14-18]. Psychological distress, older age, and verbal intelligence quotient (IQ) have been associated with an increased risk of NP impairment [5,19,20]; however, these factors do not account completely for the NP deficits observed in HIV-infected patients [4,6,21].

 

These studies illustrate important covariates that should be examined and controlled in any examination of persons who are co-infected with HIV and HCV. To our knowledge, only four studies [22-25] have examined the issue of NP function in HCV and HIV co-infected individuals. Letendre et al. [22] demonstrated that HCV contributed to impaired NP performance, independent of HIV serostatus and methamphetamine use. Neurocognitive impairments were evident in learning, memory and motor skills, but not in verbal, abstraction, perception or attention skills. Hilsabeck et al. [23] showed that HCV-infected patients with other comorbidities, including HIV, performed worse on tests than those with HCV alone or in those with chronic liver disease alone. Ryan et al. [24] compared performance on NP examination for patients co-infected (HIV-positive/HCV-positive) with patients who were infected by HIV alone (HIV-positive/HCV-negative). The impairment rates were similar across domains except for executive functioning in which the co-infected patients scored worse. Martin et al. [25] reported that polydrug users co-infected with HIV and HCV showed significantly slower reaction time in comparison with mono-infected and uninfected patients.

 

The current data were collected from women enrolled in the Women's Interagency HIV Study (WIHS), a large investigation of disease progression in women living with HIV/AIDS. WIHS is an ongoing NIH-funded multi-center study of approximately 2000 HIV-positive and 550 HIV-negative women with centers in Los Angeles, San Francisco, Chicago, New York, and Washington, DC [26]. Neuropsychological findings from the WIHS have been reported previously [5]. In this study, we report additional data that compare NP performance among women who are co-infected with HCV and HIV, those infected with only one of these viruses, and those seronegative for both viruses.

 

Abstract
Objective: To evaluate the neurocognitive function in 220 women enrolled in the Women's Interagency HIV Study (WIHS), a study of disease progression in women living with HIV/AIDS and in HIV-negative controls.

 

Methods: We evaluated the prevalence of abnormal neuropsychological (NP) results in hepatitis C virus (HCV)-positive compared with HCV-negative women in combination with HIV serostatus.

 

Results: NP impairment was significantly higher for HCV-positive women in comparison with HCV-negative women [odds ratio (OR), 2.03; 95% confidence interval (CI), 1.17-3.51]. Women co-infected with HCV and HIV demonstrated greater abnormal NP performance than those not infected with either, particularly if there was evidence of CD4 T-lymphocyte immunosuppression [> 200 × 106 CD4 cells/l (OR, 3.48; 95% CI, 1.49-8.15) and ≤ 200 × 106 CD4 cells/l (OR, 5.38; 95% CI, 1.46-19.84)]. Women who were HCV-positive/HIV-positive and not taking antiretroviral therapy (ART) were more likely (OR, 7.03; 95% CI, 2.63-18.82) to demonstrate NP impairment than those who were HCV-negative/HIV-negative. In analyses controlling separately for education, intelligence quotient, depression, sedating drug use, head injury, ethnicity, and history of substance use, HCV continued to significantly predict NP impairment. The HCV effect did not reach significance when controlling for age in bivariate or multivariate analyses although the odds ratio for NP abnormalities in HCV-infected patients was only slightly reduced (ORs above 1.9). After testing for an interaction between age and infection status, we conducted age-stratified analysis and showed a significant effect of infection status for those aged under 40 years.

 

Conclusions: The effect of aging on co-infected populations will require further study. This study has demonstrated the association of HCV with the risk of neurocognitive impairment in women living with HIV/AIDS and suggests that co-infection has an additive effect.

 

Discussion
This is the first investigation to evaluate the status of neurocognitive function in a large sample of women who are not infected with HCV or HIV, or who are infected with one or both of these viruses. We found that HCV-positive women were significantly more likely to demonstrate abnormal NP results in comparison with HCV-negative women (48.5 versus 31.7%, respectively). These results are consistent with others who have looked at the NP effects of HCV [7-10,25].

 

These data also demonstrate that HCV in combination with HIV results in substantially increased odds of NP impairment. For those women who were dually infected with HCV and HIV regardless of CD4 category, the odds of NP impairment were significantly elevated. The test for an interaction between HIV and HCV was not significant and the combined effect of HIV and HCV status on NP impairment appears to be additive. We also showed a significant increase in NP impairment for those who were co-infected and not on HIV antiretroviral therapy. These results are similar to those reported by others [22-24]. Due to our cross-sectional design, we must limit our conclusions to the observation of a strong statistical association between NP function and HCV, while suggesting that the combined effect of HCV and HIV is greater than either alone in terms of NP impairment.

 

Bivariate analyses were carried out in order to investigate the impact of each factor individually on the relationship between HCV and NP disorders. The relationship held up on analyses controlling for each of these variables: HIV serostatus, education, ethnicity, history of head injuries, use of potentially sedating medications in the past 24 h, lifetime drug use, psychological distress, estimated verbal IQ as predictors, and for site of data collection. Only after controlling for age did HCV become non-significant.

 

Poorer NP performance has been associated with age among patients with HIV in previous research [39,40]. However, prior research does not address the issue of co-infection. The significant interaction between age and infection status suggest the association of HCV serostatus with NP impairment may be moderated by age. The relationship between HCV and NP impairment was significant among those under 40 years of age and those over 40 years old had elevated NP impairment in all categories. However, there were limited numbers of women over age 40 years; thus these analyses are less stable. Further, the impact on NP performance of the combined effects of antiretroviral therapy and disease progression in both young and aging HIV/HCV patients will require further study. An additional potential limitation is that the control group was younger and better educated than either of the HCV-positive categories although these variables were controlled in statistical analysis. Although IQ as measured by the Quick test was not different between any groups, the differences in education may have had an effect on literacy and thereby have influenced the study results.

 

Previous studies of HCV and NP performance have reported evidence of impaired psychomotor and working memory function [8,9,11]; our findings are similar, in that we found significant differences between HCV groups on the Color Trails, Grooved Pegboard, and Symbol Digit tests. These timed tests require cognitive tasks of dividing and set shifting. Consistent with Kramer et al. [8] there were no differences on memory tests.

 

The mean CES-D score, identifying psychological distress, was not significantly different between HCV-positive women and HCV-negative women as has been previously reported [7,8]. Research has shown a high rate of depression among HIV-positive women but also among their risk-matched controls [41]. Our findings indicated psychological distress was associated with an increased risk of NP impairment, which persisted after controlling for HCV status. HCV status was also significant after controlling for psychological distress, indicating that each of these factors independently contribute to NP impairment. Prior studies also show that psychological distress does not account completely for NP deficits in persons living with HIV [4,6,19-21]. We also note that because the data were collected prior to the use of pegylated IFN-α with or without ribavirin for the treatment of HCV disease, and no women reported taking these medications at the time of NP testing, our results are not confounded by depression or other neuropsychiatric complications associated with these therapies [42].

 

We did not assess either HCV or HIV viral load in the cerebrospinal fluid (CSF). A significant relationship between viral load in CSF and neurocognition is most evident in patients with advanced HIV-associated dementia [43] and we excluded such women from this study. Prior research has also raised the possibility of HCV infection of the brain through infected macrophages/microglia cells [9] but we did not test for HCV-RNA levels either in CSF or in blood. Assessment of hepatic abnormalities due to HCV by liver enzyme testing may be unreliable in HIV co-infected individuals on antiretroviral therapy as these drugs commonly lead to liver enzyme elevations; therefore liver enzyme analysis were not used in our assessment of potential causes of NP impairment [44]. Thus the study is limited in its ability to assess severity of liver disease as a factor in NP performance. Severe liver disease may be more likely to occur in persons co-infected with HIV and HCV [45] and greater fibrosis is associated with poorer NP performance [23]. These biological abnormalities associated with co-infection, including the roles of liver function and CNS infection, on NP performance will require further investigation.

 

The limitations found in speed of information processing requiring divided attention or set shifting have significant implications for impairment in performance of daily activities. The ability to concentrate, perform multiple tasks, and learn new information can lead to interference in driving, self care, employment, and adherence with therapy regimens for HIV and/or HCV. When co-infection occurs, physicians should be especially sensitive to the possibility that the patient may be at increased risk of problems with treatment adherence.

 

This study is among the first to report evidence of increased risk of NP deficit in subjects dually infected with HIV and HCV and the first to examine these relationships among women. NP impairment will be a continuing concern in the optimal management of patients with HCV and HIV disease.

 

Results
Of the 220 women, 70 were HCV-positive/HIV-positive, 27 were HCV-positive/HIV-negative, 75 were HCV-negative/HIV-positive and 48 were HCV-negative/HIV-negative. Of the HIV-positive women 78 (52.7%) were receiving antiretroviral therapy (ART) at testing and 70 (47.3%) were not. No HCV-positive participant was taking pegylated interferon-alpha (IFN-α) therapy with or without ribavirin for HCV as it was not available as standard treatment during the time data were collected.

 

Table 1 shows demographic data for all subjects grouped by HCV and HIV serostatus. As shown, the groups differ significantly on mean age, education, history of closed head injury, lifetime drug use, and sedating drug use in the past 24 h. Groups differed, as expected, on CD4 cell count, with the HIV seropositive groups (with or without HCV coinfection), being significantly lower. There were no differences in reported psychological distress as measured by CES-D scores, IQ as estimated by the Quick test, or ethnicity.

 

NP test results
In comparison with HCV negative, HCV-positive subjects had higher proportions of abnormal scores on the Color Trails 1 (P < 0.0001), Color Trails 2 (P = 0.006), Grooved Pegboard (dominant hand) (P ≤ 0.0001), Grooved Pegboard (non-dominant hand) (P = 0.0013), and the Symbol Digit test (P = 0.0022).

 

Eighty-six subjects (39.1%) were classified as abnormal on the basis of having two or more unique NP scores in the impaired range. The prevalence of abnormal NP test results for the HCV-positive (48.5%) women was significantly higher than for the HCV-negative women (31.7%) (chi-squared = 6.39, P ≤ 0.01).

 

HCV status in combination with HIV status and HIV treatment status
The risk of NP impairment was higher for HCV-positive women in comparison with HCV-negative women (OR, 2.03; 95% CI, 1.17-3.51) and also higher for HIV-positive women in comparison with HIV-negative women (OR, 2.09; 95% CI, 1.15-3.81). Using the non-infected group (HIV-negative/HCV-negative) as the control, risks of NP impairment for women mono-infected with either virus were elevated and those co-infected with both were significantly more likely to be abnormal (OR, 3.77; 95% CI, 1.66-8.57). The test for an interaction between HIV and HCV was non-significant (Wald chi-square = 0.006, P = 0.94). Those who were dually infected were more likely to have abnormal overall NP tests results in which the CD4 cell count was over or under 200 × 106 cells/l (OR, 3.48; 95% CI.1.49-8.15 and OR, 5.38; 95% CI, 1.46-19.84, respectively). In addition, HIV-positive women off ART therapy were more likely (OR, 7.03; 95% CI, 2.63-18.82) to have abnormal NP test results than those not infected with either virus.

 

Method
Research participants
Detailed information on the national WIHS design and methods is reported elsewhere [26]. Between April 1995 and April 1997, we tested 231 English-speaking women enrolled in the WIHS at the Los Angeles or Chicago centers, of whom 220 had complete data. The participants met the following criteria: no history of AIDS-defining neurological conditions, including clinical dementia [The parent WIHS project excludes women with evidence of clinical dementia. Consequently study data pertain to the spectrum of cognitive dysfunction in women whose impairment fell short of frank dementia.]; no history of schizophrenia, bipolar disorder or epilepsy; and no evidence of alcohol intoxication at testing. All women were ambulatory, without acute illness and all testing was conducted on an outpatient basis. Approximately 5% of the women approached refused to participate in the study. Participants received US$ 35 incentive. The study received Institutional Review Board approval from each participating institution and all women gave written informed consent.

 

We queried participants about types and duration of street drugs use. We constructed an index of lifetime drug use by summing the years a subject used cannabis, opiates, and cocaine and years of abusive alcohol use, defined as three or more drinks daily. Subjects were classified as having a low (accumulated 0-9 years), medium (10-30 years), or high (31-121 years) drug use history based on tertile cutpoints. Participants also reported whether they had taken any medication with potentially sedating side-effects, such as antihistamines, in the 24 h prior to testing.

 

Measures and procedures
Neurocognitive testing
All NP tests were administered by a doctoral level clinical psychologist or by a master's level psychometrician supervised by a board-certified neuropsychologist. Criteria for NP test selection included; sensitivity to HIV-related cognitive/motor impairment [6,27], brevity of administration and minimal dependence on literacy or language ability wherever feasible. Tests used at both sites included Color Trails 1 & 2 to measure divided attention, set-shifting, and psychomotor functioning [28]; WHO/UCLA Auditory Verbal Learning Test of multiple trial auditory verbal list learning under immediate and delayed recall conditions [28]; Grooved Pegboard to measure psychomotor speed and fine motor control [29]; Symbol Digit Modalities Test of written coding that requires psychomotor speed, memory, attention, and concentration [30]; Visual Reproduction Subtest of the Wechsler Visual Memory Scale-Revised for immediate and delayed recall of four geometric designs [31] and Mental Alternations Test, an oral analogue of the Trail Making Test designed to assess divided attention and set shifting with minimal demand on motor function [E.L. Teng, The Mental Alternations Test (MAT). unpubl. manuscript. 1994].

 

We obtained an estimated verbal IQ using the Quick Test [32]. We administered the Centers for Epidemiologic Studies Depression Scale (CES-D) as an index of psychological distress [33].

 

Laboratory studies
HIV serostatus was confirmed using Food and Drug Administration (FDA)-approved enzyme-linked immunosorbent assay testing and if reactive, an FDA approved western blot HIV-1 confirmatory assay. Absolute CD4 lymphocyte count was determined by immunophenotyping and flow cytometric analysis. Laboratories performing flow cytometry testing were participating in the NIAID DAIDS Flow Cytometry Quality Assessment Program and followed Guidelines for Flow Cytometric Immunophenotyping [34]. The CD4 cell count has been shown to correlate with clinical AIDS diagnosis and disease progression in the WIHS cohort [35]. CD4 cell count was stratified between those participants with CD4 cell count > 200 × 106 cells/l and those with CD4 cell count ≤ 200 × 106 cells/l (the CDC definition for immunologic AIDS).

 

Ultrafrozen plasma from individual patients were tested for the detection and quantitation of HIV RNA copies [36,37] (viral load) using nucleic acid, sequence-based amplification (NASBA) which employs an isothermic RNA amplification method. Laboratories performing the NASBA testing were participants in the NIAID AIDS Program Virology Quality Assurance HIV RNA Proficiency Program of the NIH [38]. Viral load was detectable at ≥ 4000 copies /ml using the NASBA method. A stratification of viral load was used that represented approximately equivalent number of women per category: viral load less than 4000 copies/cm3 (the assay cut off of detection), viral load 4000 to 35 000 copies/cm3, and viral load over 35 000 copies/cm3.

 

At enrollment, all participants were screened for the presence of antibodies to HCV using a second generation enzyme immunoassay (HCV EIA 2.0; Abbott Laboratories, Abbott Park, Illinois, USA). Enrollee plasma samples were stored at -70°C. In 1999, the last plasma specimen placed in repository for each of the initially HCV seronegative women was tested for antibody to HCV using a third-generation enzyme immunoassay (HCV EIA 3.0; Ortho-Diagnostic Systems, Raritan, New Jersey, USA).

 

Overview of statistical analyses
We administered and scored all NP tests according to standardized procedures and used raw scores consistent with our previously published study [5]. We z-transformed each participant's raw NP scores using means and standard deviations for the HIV negative control group. We then classified a z-transformed NP test score as impaired if it fell at least one standard deviation below the mean of the control group [3,4,16]. We classified a participant's overall NP test results as abnormal if there were two or more unique test scores falling in the impaired range. 'Unique' was defined as derived from separate tests. We evaluated the prevalence of z-transformed NP test scores in the impaired range by chi-squared analyses. Normed scores were not ideal for comparison because existing norms were derived using subject populations who differed significantly in ethnicity from our study group. We used raw scores for comparison from uninfected women in the WIHS study, who were of similar ethnic composition, drawn from the same geographic regions and who use the same medical care system.

 

For categorical variables, chi-squared analyses were conducted to compare distributions of covariates between the four groups differentiated by HCV and HIV status. Comparison of means within the same four groupings was conducted using Duncan multiple range test allowing means for each of the four groups to be compared and tested for differences between any two groups.

 

We employed a series of bivariate analyses that predicted prevalence of overall NP test results, calculating odds ratios (ORs) and 95% confidence intervals (CIs) (i.e. P ≤ 0.05) using maximum likelihood estimates from logistic regression models. We examined the combined effect of HCV and HIV as well as HCV and untreated HIV. We also examined the relationship of NP impairment and HCV infection after controlling individually for these variables: education, age, ethnicity, history of head injuries, use of potentially sedating medications in the past 24 h, lifetime drug use, psychological distress (CES-D), estimated verbal IQ, and testing site. We employed multivariate logistic regression, adjusted for the same variables, to evaluate the risk for overall NP test results associated with HCV/HIV status. We also tested for interactions between HIV and HCV and between age and HIV/HCV infection on NP function. In order to test for possible additive effects of co-infection and age together, we computed odds ratios and confidence intervals across age and infection status and within age strata.