The number of people estimated to be infected with the hepatitis C virus (HCV) worldwide is 170 million.1 An estimated 33.2 million people are infected with HIV.2 Both of these viruses have been shown to infect lymphocytes3-5 and to cause serious damage to the peripheral and central nervous systems. The neurological manifestations of HIV infection have been well delineated in more than 20 years of research on HIV neuropathogenesis. However, the neurological manifestations of hepatitis C have been less well characterized and reported, in part because the virus has been structurally identified only for little more than a decade.
The most frequently encountered neurological problems are an HIV-associated distal predominantly sensory peripheral polyneuropathy (DSPN) and a progressive central cognitive-motor impairment called HIV-associated cognitive-motor complex (HIV-CMC). Other diagnostic terms for this condition include HIV-associated dementia and, when the deficits are less marked, minor cognitive-motor disorder, AIDS dementia complex, and HIV/AIDS encephalopathy.4,5
The pathogenesis of HIV neuropathy is less well understood. Before the introduction of antiretroviral drugs, investigators had reported a 30% to 35% prevalence of peripheral neuropathy in persons with AIDS. The majority had DSPN characterized by a significant or purely axonal pattern on electrophysiological studies. The major clinical symptoms experienced by persons with HIV peripheral neuropathy are pain, paresthesia, dysesthesia, numbness, cramps, and mild distal weakness. HIV-associated neuropathies are frequently so painful and difficult to manage that they can impair a person’s quality of life and ability to work and may necessitate long-term narcotic analgesia. Classic DSPN rarely requires confirmation by electromyography with nerve conduction velocity testing because the diagnosis is made clinically.6
HIV-related DSPN typically occurs in association with advanced HIV disease, low CD4+ cell counts, and high HIV RNA levels.7 However, some pesons with apparently good control of their HIV disease develop DSPN. Alternative explanations for the development of DSPN include the impairment of the nervous system by neurotoxins generated during HIV infection and the use of the antiretroviral “d-drugs,” which are the NRTIs didanosine (ddI), stavudine (d4T), and zalcitabine (ddC). These antiretrovirals are known to cause a neuropathy very similar to that seen in untreated HIV infection.8 However, symptoms in many affected patients do not improve when these drugs are withdrawn, leading to the hypothesis that d-drugs actually uncover a preexisting, subclinical HIV DSPN.
The best-known clinical sequelae of chronic hepatitis C are hepatic cirrhosis and hepatocellular carcinoma. However, HCV has also been implicated as a cause of autoimmune syndromes, cryoglobulinemia, and vasculitis, which, in turn, have been related to an increased incidence of stroke and peripheral neuropathy in HIV/HCV-coinfected persons.9-12
HIV penetrates the CNS early in the course of infection. The same may be true for HCV. In 1999, Fujita and colleagues13 reported a case of acute viral encephalitis with subsequent discovery of acute HCV infection. Two other investigative teams have reported the presence of HCV by quantitative polymerase chain reaction (PCR) analysis in the cerebrospinal fluid (CSF) of HCV-infected patients. It was postulated that the virus was of plasma origin because, in most of the cases, the HCV genotype was the same in both the patient’s plasma and CSF. The mechanism of transport of HCV from the systemic circulation into the CNS is likely to be similar to that of HIV because both viruses have been shown to infect monocytes, which cross the blood-brain barrier.14,15
Chronic active hepatitis is frequently associated with mixed cryoglobulinemia and has been shown to cause numerous CNS diseases, even in the absence of hepatic damage. The presumed mechanism is a vasculitis associated with cryoglobulinemia and multiple ischemic infarcts. Encephalopathy ranging from minor confusion to severe cogni-tive-motor dysfunction has been described.16-20
Previous attempts to ascertain the impact of HIV/HCV coinfection on the CNS have been sparse. In 2004, Ryan and colleagues21 reported statistically significant differences in the neuropsychiatric impact of hepatitis C on advanced HIV disease using the same neuropsychological test battery as used in our current study. In fact, their data did not show any statistically significant difference based on neuropsychological test scores only between HIV-infected subjects and those coinfected with HIV and HCV. The results reported showed merely trends in their data that were not significantly different.
We conducted a study to determine whether HIV/HCV-coinfected persons have a significantly higher prevalence and/or severity of HIV-CMC and DSPN than in those infected with HIV alone.
The study population and associated frozen archival plasma, serum, and CSF samples were drawn from a well-characterized, HIV-positive cohort with AIDS enrolled in an ongoing longitudinal tracking and fluid, tissue, and brain banking repository study known as the National Neurological AIDS Bank (NNAB) funded by the NIH as part of The National NeuroAIDS Tissue Consortium.22 The main criterion for the selection of study subjects was having a greater likelihood of mortality within 2 years of enrollment as determined by an HIV medicine specialist. Subjects were evaluated every 6 months, and the data and archived specimens from the subject’s latest visit in the NNAB repository study were evaluated.
The prevalence and severity of HIV-CMC and DSPN in HIV-infected subjects were compared with those in subjects who were coinfected with HIV and HCV. The presence of HIV-CMC was correlated with the HIV RNA and HCV RNA levels in both plasma and CSF. Baseline characteristics of the subjects were determined using a standardized questionnaire, which covered their medical, neurological, psychiatric, risk behavior, and medication histories, as well as by a standardized neurological and neuropsychological battery well used in NIH HIV neurological studies.22,23
Specific laboratory research analysis was conducted on the subjects’ frozen archival blood (plasma and serum) and CSF from the NNAB. Plasma and CSF samples were prepared, and HIV RNA levels were determined using the Roche Amplicor HIV Monitor Test, version 1.5 (microwell plate), and HCV RNA levels using the Roche Amplicor HCV Monitor Test, version 2.0 (microwell plate). In addition, serum was tested for the presence of cryoglobulins and CSF was tested for glutamine levels. The CNS and peripheral nervous system diagnoses were linked cross-sectionally as a single time point from each subject’s most recent blood and CSF results, including testing for HCV. Other tests performed at 6-month intervals included measurement of CD4+ cell count; analysis of CSF and plasma HIV RNA levels by Amplicor HIV-1 Montior Test, version 1.5 (ultrasensitive); assessment of glycosylated hemoglobin level in the blood; and toxicological screens of urine.
At each study visit, subjects were evaluated neurologically (by a neurologist with special training in neuroAIDS) and neuropsychologically (by either a neuropsychologist or a technician trained to administer the neuropsychological test battery under the supervision of a neuropsychologist with expertise in HIV/AIDS). Within 2 months of the study visit, each subject was again evaluated both neurologically and neuro-psychologically during a case conference attended by the examining neurologist and the supervising neuropsychologist. The subject was assigned a composite neurocognitive diagnosis and a Memorial Sloan-Kettering Cancer Center AIDS Dementia Complex (MSK ADC) score by both the neurologist and neuropsychologist and given a final diagnosis for the presence of HIV-associated DSPN by the examining neurologist. Subjects with diabetic neuropathy as defined by a clinical chart review that identified a diagnosis of diabetes mellitus, with a prescription of an oral antidiabetic agent or insulin, or with an abnormally elevated glycosylated hemoglobin level were excluded from the neuropathy analysis. Similarly, subjects with a history of alcoholism were excluded from the study. Individuals with a confounding CNS diagnosis such as prior or current cryptococcal meningitis, cerebrovascular accident, cerebral toxoplasmosis, primary brain lymphoma, history of significant head trauma with prolonged loss of consciousness, and those acutely intoxicated on drugs or alcohol were also excluded. This clinical review was conducted with all examiners blinded to the knowledge of HCV status as frozen archival plasma and CSF specimens were tested for HCV, serum cryoglobulins, and CSF glutamine levels after the neurocognitive and DSPN diagnoses were assigned.
The neuropsychological battery consisted of the following tests within 7 cognitive domains (the test details and documentation have been described elsewhere23):
1. Verbal fluency: Controlled Oral Word Association Test.
2. Speed of information processing: WAIS-III Digit Symbol and Symbol Search subtests and Trail Making Test (TMT) Part A.
3. Learning: Hopkins Verbal Learning Test-Revised (HVLT-R), Total Trial 1-3 Recall, Brief Visuospatial Memory Test-Revised (BVMT-R).
4. Memory: HVLT-R Delayed Recall, BVMT-R Delayed Recall.
5. Executive functions: Wisconsin Card Sorting Test perseverative responses, TMT Part B.
6. Attention and working memory: WAIS-III Letter-Number Sequencing, PASAT-50.
7. Motor: Grooved Pegboard Test dominant and nondominant hand performances, Wide Range Achievement Test-3.
A reading subtest was administered as an estimate of premorbid verbal intellectual functioning. As part of the neuropsychological evaluation, participants also completed self-report questionnaires to assess their mood state (ie, Beck Depression Inventory) as well as to document their perception of cognitive difficulties and their degree of independence in completing everyday activities. The Patient's Assessment of Own Functioning Inventory is a relatively brief questionnaire that assesses noticed difficulties with memory, language and communication, motor skills, sensory perception, and higher-level cognitive and intellectual functions. The Activities of Daily Living skills questionnaire is a modified version of the Lawton and Brody Activities of Daily Living measure, which assesses the subject's degree of independence in performing a variety of tasks involved in independent living, ranging from self-hygiene and performing housework to managing finances and medications. The specific neuropsychological testing battery and documentation of this method of arriving at a reliable diagnosis using interrater reliability has been described.23 A composite neurological/neuropsychological diagnostic score was assigned and matched with an MSK ADC rating.5,23 All of these diagnoses were entered into a central database.
The results of the neurological examinations and the NNAB study laboratory tests and a review of subjects' medical records were used to determine the presence or absence of DSPN in each subject. The diagnoses of HIV-associated DSPN and mixed HIV/NRTI toxicity-related DSPN were condensed to DSPN. Subjects with HIV-CMC were further subclassified as follows: subclinical equivocal (defined as not impaired enough to meet the set diagnostic criteria [given below] for minor cognitive-motor disorder) or HIV-associated dementia. The severity of dementia was ranked according to diagnostic criteria set by the American Academy of Neurology and according to the MSK ADC.
Statistical analyses of inter-method and inter-subject differences in detection rates for HIV RNA and HCV RNA were performed using a logistic regression model. Alternatively, non-parametric analyses suitable for nominal data, such as evaluations of contingency coefficients, were applied to the data.
Between- and within-group differences in HIV-1 RNA level (plasma and CSF levels; subjects with HIV-CMC and DSPN vs subjects without HIV-CMC and DSPN) were evaluated by submitting the data to profile analyses of repeated measures (using an ANOVA or MANOVA model as appropriate with follow-up alpha-corrected pairwise univariate analyses where appropriate.
Of the total 159 subjects enrolled, the mean age was 42.6 years (SD 7.8 years); the median age was 41.6 years. There were 131 men, 28 of whom were coinfected. There were 28 women, 3 of whom were coinfected. There were no significant differences between the HIV-alone and HIV/HCV-coinfected groups with regard to all demographic features tested, including age, race, ethnicity, gender, risk factor, level of education, current or previous drug use, and history of minor head trauma. There were also no statistically significant differences between the two groups for plasma and CSF HIV RNA levels and CD4+ cell count.
Of the 159 total subjects tested for hepatitis C, 31 (20%) tested positive for HCV RNA by quantitative PCR using the Roche Amplicor HCV Monitor, version 2.0. There were no statistically significant between-group differences for a DSPN diagnosis, but the 2 groups were different on the basis of disproportionate population samples within each group (128 vs 31 subjects, respectively). To compare study groups with an equal number of subjects for the neurocognitive analyses, 31 of the 128 HIV-alone subjects were randomly selected for the between-group comparisons.
Data analysis revealed a trend showing that HIV/HCV-coinfected subjects faired less well by the neuropsychological Impairment T score and Global Deficit Score. However, there was a statistically significant difference in the MSK ADC rating (P = .035) when the composite neurological/neuropsychological scores were compared and a statistically significant between-group difference for HIV-CMC diagnosis when the neurocognitive diagnosis was applied (P = .01) (Figures 1 and 2).
Figure 1. Memorial Sloan-Kettering Cancer Center AIDS Dementia Complex (MSK ADC) score rating. Of 31 randomly matched HIV-infected subjects without hepatitis C virus (HCV) infection and HIV/HCV-coinfected subjects paired by MSK ADC score rating, HIV/HCV-coinfected subjects were statistically significantly more impaired neurocognitively than HIV-infected subjects without coinfection (using ANOVA). (MSK ADC score rating: 0.0 = no impairment, 0.5 = subclinical/equivocal impairment, 1.0 = HIV-associated dementia.)
Figure 2. Primary neurocognitive composite scores. Of 31 randomly matched HIV-infected subjects without hepatitis C virus (HCV) infection and HIV/HCV-coinfected subjects, there was a statistically significant between-group difference based on a composite primary neurocognitive diagnosis, with the coinfected group being more impaired (using ANOVA). (HAD, HIV-associated dementia; MCMD, minor cognitive-motor disorder.)
The overall predominance of DSPN was high in both groups. While this appeared to be higher in the HIV/HCV-positive group (71% vs 59%), this was merely a trend. The results are illustrated in the Table and Figure 3. No serum specimens tested positive for cryoglobulins, and all CSF glutamine levels were within normal limits.
Figure 3. Neuropathy scores. There was a trend toward a greater prevalence of distal predominantly sensory peripheral polyneuropathy (DSPN) in subjects coinfected with HIV and hepatitis C virus (HCV) (using ANOVA). (0 = no DSPN, 1 = equivocal DSPN, 2 = possible DSPN, 3 = probable DSPN, 4 = definite DSPN.)
This study found no statistically significant differences between subjects coinfected with HIV and HCV and those with HIV infection alone when applying neuropsychological scores independent of a subject’s clinical review. However, when both the neurologist and neuropsychologist, who were unaware of the subjects’ HCV status, arrived at a composite diagnosis (neurocognitive diagnosis) and assigned a correlative MSK ADC rating, there was a statistically significant difference between the coinfected group and the HIV-alone group. One possible explanation for this finding is the variability in neuropsychological test results, even for the same subject over time. Specifically, highly educated persons may be cognitively impaired but score well on formal neuropsychological testing; whereas, poorly educated, poorly acculturated persons with low literacy may score poorly even when they function normally and have little if any cognitive and motor deficits.
Future studies require an increase in the sample size per group. Although our results show a trend toward a potentiation of neurological manifestations in coinfected subjects, the data generated using current neuropsychological testing batteries alone were not statistically significant. The composite ratings, while showing a significant effect of coinfection on the CNS, are more subject to neurological examiner bias but still yield a more accurate picture of each person’s ability to function in the real world because the neuropsychological tests are inclusive of neither neurocognitive status at baseline nor regularity in activities of daily living, and these tests were standardized using a highly acculturated, English-fluent population. Future studies need to focus on including HIV-negative persons who are infected with HCV, so that we may better understand the neurological manifestations of HCV monoinfection in relation to HIV/HCV coinfection and to what is already known about the neurological manifestations of HIV.
Despite the fact that cryoglobulinemia is a reported etiological factor for both cerebrovascular disease and peripheral neuropathy in hepatitis C, we did not find this abnormality in any of the subjects tested. We can only assume that it is fairly rare and may pose less of a problem than some investigators have postulated.11
Hepatitis C can and does cause neurological disease in both the central and peripheral nervous systems. Some of the homogeneity between the HIV-alone-infected group and the coinfected group in our study may be attributed to the fact that all subjects had late-stage HIV infections, where the effects of either virus was so substantial as to overshadow the effects of each virus when occurring with coinfection. This is especially true of DSPN where the prevalence in our study was close to 60% or greater in HIV-infected persons with and without hepatitis C. In the literature, the prevalence has historically been in the 30% to 35% range; the higher prevalence rate in our study is likely because our study subjects were specifically chosen to have a greater likelihood of mortality within 2 years. Again, to fully understand the effects of HIV infection and hepatitis C, either alone or together, on HIV-CMC and DSPN, we must begin to study monoinfected and coinfected persons longitudinally from the beginning of their infection.
Acknowledgments: H.A.A. was supported in part by NS-38841. We gratefully acknowledge the additional support for this project provided by Roche Diagnostic Corporation for their payment of the costs associated with HCV testing in plasma and CSF and for serum cryoglobulin and CSF glutamine testing not included as part of NNAB.
Other than that indicated in the Acknowledgments, no potential conflict of interest relevant to this article was reported by the authors.