The Clinical Spectrum of Hepatitis C Virus in HIV Coinfection
Shiffman and Sterling report on a study finding no difference in progression of HCV between HCV/HIV coinfected vs HCV monoinfected. You should read the study report below to see the author's explanation as to why their findings differ from many other studies. Other studies find HIV accelerates HCV by 2-4 fold. He also finds a biopsy is important because a percentage of patients with normal ALT had advanced HCV. He also found higher HCV viral load was associated with more advanced HCV (23% advanced liver disease, 14% cirrhosis).
Summary: The biochemical, virologic, and histologic spectrum of hepatitis C virus (HCV) in 66 consecutive patients with HIV-HCV coinfection and 119 HCV controls was compared: 86% of coinfected patients had CD4 counts >200 cells/mm3, 51% had a normal alanine aminotransferase (ALT) value, the mean HCV RNA titer was 5.7log IU/mL, 92% of coinfected patients were of genotype 1, and the mean histologic activity index was 6.86 with advanced fibrosis in 32% of patients. The biochemical, virologic, and histologic findings of HCV in coinfected patients were similar to those observed in HCV controls. For both groups of patients, no clinical, biochemical, or virologic factors could reliably identify patients with advanced fibrosis or cirrhosis, underscoring the importance of liver biopsy in the evaluation of these patients. The spectrum of liver disease in coinfection includes a significant proportion of patients with normal ALT values, and excluding these patients from previous studies has led to an overestimation of HCV disease severity..., those with higher HCV RNA levels had more advanced fibrosis (43% vs. 17%; p = .047).
JAIDS Journal of Acquired Immune Deficiency Syndromes 2003; 32(1):30-37
*Richard K. Sterling; Melissa J. Contos; *Arun J. Sanyal; *Velimir A. Luketic; *R. Todd Stravitz; *Mary S. Wilson; A. Scott Mills; *Mitchell L. Shiffman
Section of Hepatology and Division of Pathology, Virginia Commonwealth University Health System/Medical College of Virginia Hospitals, Richmond, Virginia, U.S.A.
Hepatitis C virus (HCV) and HIV infections are global health concerns (1-7). Due to shared routes of transmission, coinfection with HIV and HCV is common. The prevalence of HCV in patients with HIV is dependent on risk behaviors
and is highest in intravenous drug users (60%-90%) and hemophiliacs (50%-70%) compared with homosexual men (4%-8%). Until recently, the clinical course of HCV in coinfected individuals was overshadowed by the high morbidity and mortality of HIV disease. With the introduction of highly active antiretroviral therapy (HAART) and its associated improvement in survival rates, HCV has now emerged as a significant viral pathogen with associated morbidity and mortality in patients with HIV-HCV coinfection.
Several investigators have examined the histopathology of HCV in patients with coinfection. Although some of these reports suggested that there is no difference in liver histology between coinfected patients and HCV controls, others have observed more inflammation and fibrosis in coinfected individuals. The discrepancies in these reports may be explained by the populations studied, because most were limited to a subset of coinfected patients, namely, those with abnormal liver chemistries or clinically advanced liver disease. As a result, the true spectrum of liver disease in coinfection has not been well defined. In addition, many of these studies were performed prior to HAART,
and the effect this may have on the natural history of HCV in HIV-coinfected patients remains controversial. The aims of this study were to describe the spectrum of liver disease in patients with HIV-HCV coinfection and to compare this with a group of HCV controls.
Inclusion criteria for both coinfected and control patients included presence of anti-HIV antibody (coinfected group only) and anti-HCV antibody, presence of
HCV RNA, platelet count of 80,000, and normal renal function. Most (92%) coinfected patients were on stable HAART therapy consisting of a combination of nucleoside reverse transcription inhibitors (NRTIs), non-NRTIs (NNRTIs), and protease inhibitors (PIs). Coinfected patients were excluded if they had evidence of decompensated advanced HIV (CD4 <50 cells/mm3, with a life expectancy of less than 1 year) or active or recent (within 3 months)
opportunistic infection. Patients in both groups were excluded if they had decompensated liver disease (i.e., bilirubin >4 mg/dL, albumin <2.5 g/L, prothrombin time prolonged >2 seconds, presence of ascites or hepatic encephalopathy), renal failure, presence of hepatitis B surface antigen, or another cause of liver disease.
Spectrum of Liver Disease in HIV-Hepatitis C Virus Coinfection Compared with Hepatitis C Virus Controls
The baseline patient data reflects our primary indigent inner city HIV clinic population (mean age = 42 years, 76% male, and 85% African American). Most (92%) patients were taking HAART, including NRTIs in 90%, NRTIs in 46%, and PIs in 44%. The median and mean HIV loads were 2.5 and 1.9log copies/mL; 47% of patients were negative for HIV RNA on HAART therapy. The median and
mean CD4 counts were 503 and 555 cells/mm3 (range: 52-2332 cells/mm3), and 14% of patients had CD4 counts under 200 cells/mm3. The median and mean ALT values were 65 and 80 U/L (range: 18-278 U/L), and 51% of patients had normal ALT values at the time of evaluation. The median and mean HCV RNA level was 5.7log IU/mL, and 56% of patients had values greater than 850,000 IU/mL (the upper limit for the Amplicor assay without dilution). The majority of coinfected patients were of genotype 1 (92%). The mean HAI was 6.86, with a mean inflammatory score of 5.4 and a fibrosis score of 1.4. Advanced fibrosis was observed in 32% of patients (bridging fibrosis in 18% and cirrhosis in 14%). 72% of control patients (HCV) had genotype 1 vs 92% of coinfected patients.
Half of those with coinfection had a normal ALT value at the time of evaluation. When stratified by ALT, no significant differences were observed in clinical, demographic, virologic, or histologic parameters between coinfected patients and HCV controls. In patients with coinfection, we observed no significant differences in CD4 count, HIV or HCV RNA titer, or use or type of HAART between those with normal and elevated ALT values. Patients with normal ALT values in both groups had less advanced fibrosis compared with patients with an elevated ALT value.
One third of coinfected patients had advanced fibrosis. When comparing those with and without advanced fibrosis, there were no differences in mean ALT value (85 vs. 77 U/L), proportion with a normal ALT value (43% vs. 53%), HCV
RNA or HIV RNA level, CD4 count, CD4 percentage, or proportion of those with a CD4 count <200 cells/mm3. Those with advanced fibrosis did have higher necroinflammatory scores (6.5 vs. 4.8; p = .022). Biochemical and virologic testing could not accurately identify which patients had advanced fibrosis on liver biopsy. Furthermore, we observed no significant differences in the percentage of those on NRTIs (89 vs. 97), NNRTIs (57 vs. 44), or PIs (63
vs. 48) in those with or without advanced fibrosis, respectively.
When stratifying patients by HCV RNA level, no difference in patient demographics, HIV RNA level, proportion negative for HIV RNA, total CD4 count, percentage with a CD4 count <200 cells/mm3, and serum ALT values were observed between those with low (<850,000 IU/mL) vs. high >850,000 IU/mL) HCV RNA titers. There were differences in mean HAI (5.75 vs. 7.72; p = .027) between these groups attributed more to fibrosis (0.96 vs. 1.8; p =
.019) than to inflammation (4.79 vs. 5.89; p = .089) scores. As a result, those with higher HCV RNA levels had more advanced fibrosis (43% vs. 17%; p = .047). Although those coinfected patients with higher HCV RNA titers had more
significant histology on univariate analysis, we were unable to demonstrate by multivariate logistic regression analysis any predictive factor, including clinical, patient demographics, serum ALT value, HIV RNA or HCV RNA level, or
CD4 count, that predicted hepatic fibrosis.
The majority of coinfected patients had CD4 counts greater than 200 cells/mm3, and almost half had an undetectable HIV RNA level. When coinfected patients were stratified by CD4 cell count, we observed no significant differences in mean HIV titers (2.8 vs. 1.8log copies/mL), percentage with undetectable HIV RNA (33 vs. 49), ALT values (76 vs. 80
U/L), proportion with normal ALT values (44 vs. 51), HCV RNA titers (5.8 vs. 5.7log copies/mL), HAI (5.8 vs. 7.0), or proportion of advanced fibrosis (33 vs. 32) in those with a CD4 count <200 cells/mm3 compared with those with a
CD4 count >200 cells/mm3, respectively. When patients were further stratified by HIV RNA status, we also found no differences in clinical, biochemical, or virologic parameters between those with a high and low CD4 count with or
without detectable HIV RNA (Table 4). Finally, we found no correlation between CD4 count, ALT values, and total HAI or its inflammatory or fibrosis score.
Several reports have described hepatic histopathology in patients with HIV. Most of these reports were published prior to the identification of HCV and in patients with abnormal liver chemistries, however. The most common findings were macrovesicular steatosis, granulomas, and cholestasis. Histopathologic findings typical of viral hepatitis were not commonly seen. Cirrhosis was rarely if ever seen in patients with only HIV infection.
Several reports have also described the liver histologic findings observed in patients with HIV-HCV coinfection. Some studies found no difference in histologic severity between coinfected patients and those with only HCV, whereas others reported more significant inflammation and/or fibrosis in patients with coinfection. The heterogeneity of these findings may be explained in part by the subpopulations of patients studied, the
immunologic status of HIV, the exclusion of patients with normal ALT values, the degree of alcohol consumption, and the age at infection. These factors have been shown to affect the natural history of HCV.
The present study evaluated consecutive coinfected patients identified by HCV RNA screening irrespective of liver chemistries and compared this group with a cohort of HIV-uninfected HCV controls. This allowed us to obtain a more
realistic perspective of HCV infection in patients with HIV in the era of HAART. The majority of the coinfected cohort had well-controlled HIV on HAART as demonstrated by a mean CD4 count of 555 cells/mm3. Only 14% had a CD4
count less than 200 cells/mm3, and almost half of the patients had undetectable HIV RNA levels. Although advanced fibrosis was observed in 32% of those with coinfection, no significant histologic differences were observed between coinfected patients and HCV controls. The 14% prevalence of cirrhosis observed in coinfected patients in the current report is similar to that reported by others. No clinical, demographic, biochemical, or virologic parameters
could accurately identify patients with significant fibrosis. Liver biopsy is therefore required to determine which patients with coinfection have significant liver histopathology.
There have been several recent studies of liver histology in HCV patients with normal ALT values. In general, 5% to 12% of such patients have significant fibrosis. In the current report, 51% of coinfected patients had a normal ALT value, which was similar to our control population (47%) and that reported by others. Although those with normal ALT values had a lower HAI and a lower proportion of advanced fibrosis compared with coinfected patients with an elevated ALT value, 23% of coinfected patients with normal ALT values had advanced fibrosis, including 14% with cirrhosis. This underscores the importance of liver biopsy in coinfected populations.
The exact mechanism for the pathogenesis of HCV is not known. Because HCV is not believed to be cytopathic, the liver injury associated with chronic disease is thought, in part, to be related to the host immune system. Consequently, HCV may have a different natural history in HIV infection. Analysis of studies that have examined the impact of immunologic status on coinfection has yielded mixed results. Some studies have described less inflammation
in those with lower CD4 counts, whereas others observed more significant histopathology. Furthermore, it has been proposed that HIV itself may affect the virulence of HCV through either direct viral interference
or interactions mediated through CD4-positive cells.
The use of combination antiretroviral therapy for HIV is associated with restoration of the immune system and increases in both CD4+ and CD8+ lymphocytes. Since HAART was introduced, a significant number of HCV-related
liver deaths have been observed in patients with well-controlled HIV disease. Liver chemistries have been known to increase in coinfected patients treated with HAART. There are limited data on liver histology in coinfected patients while on HAART. In the short term, liver histology can worsen. In the long term, HAART has been associated with a lower rate of progression to cirrhosis. The exact mechanism of worsening HCV during HAART is unknown, because serum HCV levels have been reported to decrease, increase, or remain
unchanged. Although our study was not powered to determine the effect of HAART, we observed no differences in the use of NRTIs, NNRTIs, or PIs with respect to serum ALT values or liver histology.
The significance of HCV RNA titer in patients with coinfection remains unclear. In patients with chronic HCV alone, HCV titer does not correlate with histologic severity. Some studies have demonstrated that HCV RNA levels were higher in those with coinfection compared with HCV controls, especially in those patients with low CD4 counts. Others, however, including the current study, found no difference in HCV RNA titers in patients with HIV-HCV coinfection and HCV controls. When coinfected patients were stratified by HCV RNA level
(<850,000 vs. >850,000 IU/mL), however, the current study did find a relationship between HCV RNA titer and HAI, fibrosis, and the proportion of patients with advanced fibrosis. This was similar to findings of previous reports. HCV genotype has been shown to correlate with liver disease severity in some but not all studies as well as with HIV disease progression. The majority of our patients with coinfection were of genotype 1 (92%); therefore, we were not able to assess the effect of HCV genotype on liver histology or HIV severity.
Antibodies to HCV are two to three times more common among African Americans than whites in the United States. In the state of Virginia, African Americans account for over half of the HIV-positive population (Virginia
Department of Heath, 2001). In our population of coinfected patients, African Americans make up the majority (85%). Whether the natural history of HCV in African Americans differs from that in whites remains controversial. In
the current report, which included patients on stable HIV therapy, we found no increase in histologic severity or the proportion with advanced fibrosis compared with matched HCV controls. When stratified by race, we also observed no differences in African Americans with or without HIV. Due to the small proportion of whites in our coinfected patient population, we were not able to adequately assess the difference in histologic severity between races.
There are several possible explanations for the discrepancy between our results and those of other studies. We included patients in both the study and control groups regardless of ALT values. Excluding these patients in
other studies, which may represent half of the coinfected population, may have led to an overestimation of the histologic severity of HCV in this population. Our coinfected population is representative of our urban HIV clinic, which consists predominantly of African Americans and men who acquired HIV through illicit drug use. Whether our results our applicable to other coinfected patients such as white homosexual men remains to be tested. Thirdly, there may have been selection bias in our coinfected patients referred for biopsy. We excluded those with obvious hepatic dysfunction (jaundice, ascites, or coagulopathy) as well as those with significant thrombocytopenia. By doing so, we may have underestimated the overall proportion with cirrhosis. Those with decompensated liver disease or significant thrombocytopenia were also excluded from the control group, however, which should have minimized any referral bias differences comparing the spectrum of liver disease due to chronic HCV in those with and without HIV coinfection. We also excluded HIV patients with advanced HIV disease (persistent CD4 count <50 cells/mm3). Although doing so may have biased our results, liver biopsy in these individuals rarely affects their clinical management and therefore is not routinely performed.
Fourth, although all patients were abstinent for at least 6 months prior to biopsy, our results did not take into account lifetime alcohol consumption. This was true for both coinfected patients and controls, however, which should have minimized any differences. Another limitation is the dynamic processes of HIV and HCV infections. Consequently, the single point in time laboratory values used without knowledge of nadir CD4 levels may have limited our analysis. Finally, we did not take into account duration of disease in either group; because most of our patients were not able to give a precise date of infection, such data were not included.
In summary, we found that the spectrum of HCV disease in patients with HIV-HCV coinfection on HAART includes a significant proportion of patients with normal ALT values. By including these patients, the histologic picture of
HIV-HCV coinfection is similar to that observed in HCV controls. Biochemical and virologic measures were not predictive of liver histology. Consequently, liver biopsy is needed to identify those coinfected patients with significant
liver disease. Further long-term prospective studies are needed to determine the clinical impact of antiretroviral therapy on the natural history of HCV.