| |
Undetectable HIV May Slow Fibrosis in HCV/HIV Coinfected Patients
|
| |
| |
"Slower fibrosis progression in HIV/HCV-coinfected patients with successful HIV suppression using antiretroviral therapy"
Journal of Hepatology
Article in Press Oct 2005
Norbert Bräua, Mirella Salvatorea, Carlos F. Ríos-Bedoyab, Alberto Fernández-Carbiac, Fiorenzo Paronettoa, José F. Rodríguez-Orengod, Maribel Rodríguez-Torresbe, for the Puerto Rico-New York Hepatitis Study Group. Norbert Bräu co-initiated the study, collected and analyzed data and wrote the paper. Maribel Rodríguez-Torres co-initiated the study, collected and analyzed data, and critically reviewed the paper.
Early HAART & early HCV Therapy are considerations fir HCV/HIV coinfected patients.
Authors conclude: .....This study shows that suppression (<400 c/ml) of HIV infection through HAART is associated with a slower HCV-induced liver fibrosis progression rate, and that HIV viremia but not CD4+ cell count independently predicts FPR, a phenomenon not previously described.....
"..... HIV/HCV-coinfected patients with undetectable HIV RNA through HAART have a slower FPR (fibrosis progression rate) than those with any HIV RNA level and an FPR similar to HCV-monoinfected individuals....
..... Patients with any detectable HIV viral load had a significantly higher necroinflammatory activity than HIV-negative patients (7.41 vs. 6.53, P=0.012) and than HIV-positive patients with HIV viral load <400copies/mL (7.41 vs. 6.52, P=0.035)...
.... fibrosis was more advanced in HIV/HCV-coinfected patients with any detectable HIV viral load compared to those with undetectable HIV RNA level (3.09 vs. 2.65, P=0.041)...."
Authors: Rapidly declining CD4+ cell count and high HIV viral load are suggested indications to initiate HAART which need to be weighed against the potential of long-term toxicity and patient choice. Based on our data, HIV/HCV-coinfection could be added to this list of considerations to initiate HAART in an asymptomatic patient. HAART would then be initiated for the purpose of slowing down HCV-related fibrosis progression when CD4+ cells are stable between 200/mm3 and 500/mm3, and stable HIV viral load would not prompt initiation of HAART for immunological reasons. However, the potential benefits of such an approach would need to be weighed against the cost and inconvenience of taking HIV medications as well the potential toxicity of long-term HAART such as rare but serious mitochondrial toxicity [34], hepatotoxicity [35], lipodystrophy, and the metabolic syndrome.
The ideal approach to slow and even reverse liver fibrosis progression in HIV/HCV-coinfected patients is, of course, by achieving a sustained virologic response (SVR) through anti-HCV therapy. A recent trial using peginterferon alfa-2a plus ribavirin showed an SVR rate of 40% in HIV/HCV coinfection, 29% in genotype 1 and 62% in genotypes 2 or 3 [13]. But many coinfected patients do not achieve an SVR, cannot tolerate combination therapy of pegylated interferon and ribavirin, have medical contraindications against therapy, or do not want to be treated [37].
Ongoing prospective studies of HIV/HCV-coinfected patients will further clarify factors that influence fibrosis progression based on multiple liver biopsies and longitudinal data collection. Meanwhile, it may be worthwhile considering HAART for asymptomatic HIV/HCV-coinfected patients with CD4+ cells <500/mm3 (regardless of HIV viral load) who cannot clear HCV infection with antiviral therapy.
BACKGROUND: The current study was undertaken to examine whether 5 years into the HAART era fibrosis progression is still accelerated in HIV/HCV-coinfected patients compared to HCV-monoinfected patients. HAART can suppress HIV RNA levels, raise CD4+ cell counts, and restore cellular immune function. Our study tested the hypothesis that successful suppression of HIV viral load through HAART is also able to attenuate the accelerating influence of HIV infection on HCV-induced fibrosis progression.
Fibrosis progression rate
Fig. 1 shows the fibrosis progression rates, measured in Ishak fibrosis units per year (IshFU/yr), in different subgroups. There was no difference in FPR between HIV-positive and HIV-negative patients with chronic hepatitis C (0.136 vs. 0.128 IshFU/yr, P=0.29). However, when HIV-positive patients were divided by HIV viral load, only HIV-infected patients with undetectable HIV RNA (<400copies/mL) had the same FPR as HIV-negative patients (0.122 vs. 0.128 IshFU/yr, P=0.52). By contrast, patients with any detectable HIV viral load had a more rapid FPR than HIV-negative hepatitis C patients (0.151 vs. 0.128 IshFU/yr, P=0.015) and than coinfected patients with undetectable HIV RNA (0.151 vs. 0.122 IshFU/yr, P=0.013). When HIV viral load was divided into three subgroups (<400; 400-99,999; ≥100,000copies/mL), each subgroup with a higher viral load had a more rapid FPR than the one with a lower viral load.
Fig. 1. Fibrosis progression rate by HIV status and HIV viral load patients with chronic hepatitis C. IshFu/yr, Ishak fibrosis units per year; HIV RNA level in copies per mL; K=1,000.
ABSTRACT
Background/Aims
HIV/HCV-coinfected patients reportedly have a faster fibrosis progression rate (FPR) than HCV-monoinfected patients. This study examined whether HIV suppression through highly active antiretroviral therapy (HAART) attenuates this accelerated fibrosis progression.
Methods
In two hepatitis C centers, a retrospective analysis identified 656 consecutive treatment-naïve HCV-infected patients who had undergone a liver biopsy, had a presumed date of HCV infection, and had been tested for HIV, 274 of them HIV-positive (95.2% on HAART) and 382 HIV-negative. The primary outcome measure was the FPR, defined as Ishak fibrosis score [0-6] over estimated duration of HCV infection.
The primary outcome measure of the study was the comparison of the FPR between HIV/HCV-coinfected and HCV-monoinfected patients. Secondary outcome measures were comparison of the FPR within the HIV/HCV-coinfected group by HIV viral load and CD4+ cell count. Other outcome measures were Ishak necroinflammation grade, Ishak fibrosis stage, and proportion of patients with bridging fibrosis and cirrhosis.
Results
Among HIV/HCV-coinfected patients, 51.2% had undetectable HIV RNA (<400copies/mL). There was no difference in FPR between HIV/HCV-coinfected and HCV-monoinfected patients (0.136 vs. 0.128 Ishak fibrosis units/year, P=0.29).
However, HIV/HCV-coinfected patients with any detectable HIV viral load >400copies/mL had a faster FPR (0.151) than HCV-monoinfected patients (0.128, P=0.015) and than HIV/HCV-coinfected patients with undetectable plasma HIV RNA (0.122, P=0.013) who in turn had the same FPR as HCV-monoinfected subjects (0.128, P=0.52).
An accelerated FPR in HIV viremic patients was seen with CD4+ cells <500/mm3 (0.162 vs. 0.123, undetectable HIV RNA, P=0.005) but not with CD4+ cells >500/mm3 (0.118 vs. 0.121, P=0.89).
In multivariable linear regression analysis of HIV/HCV-coinfected patients, log10 HIV RNA level, necroinflammation, and age at HCV infection were independently correlated to FPR, but not alcohol use or CD4+ cell count (r2=0.45 for model).
Conclusions
HIV/HCV-coinfected patients with undetectable HIV RNA through HAART have a slower FPR than those with any HIV RNA level and an FPR similar to HCV-monoinfected individuals.
RESULTS
Baseline patient characteristics
Among the 925 HCV-infected persons seen at our centers, 656 patients were included in the analysis after excluding patients without liver biopsy (n=63), with prior HCV therapy (106), lack of HIV RNA and CD4+ cell data within 12 months of the biopsy (29), unreliable history of alcohol intake (40), no estimated date of HCV infection (24), and coinfection with hepatitis B virus (6) or Schistosoma mansoni (1). A liver biopsy was routinely recommended to all patients at initial evaluation, regardless of comorbidity (including hemophilia) or stage of HIV disease. Sixty-three patients declined the biopsy, and their baseline characteristics did not differ from those who underwent a biopsy.
The 656 study subjects included 382 (58.2%) HCV-monoinfected and 274 (41.8%) HIV/HCV-coinfected patients. Their demographics, historical and laboratory data are displayed in Table 1. HIV/HCV-coinfected patients were on average 3.5 years younger than HCV-monoinfected persons, and within the HIV-infected group, patients with detectable HIV RNA (n=133, 48.5%) were younger than individuals with HIV RNA <400copies/mL (n=141, 51.5%). HCV risk factors were more commonly IDU and high-risk sexual behavior in HIV/HCV-coinfected patients, but transfusion of blood or blood products was more commonly a risk factor in HCV-monoinfected patients. Both groups on average consumed the same amount of alcohol over the last 5 years (36.4g/d), but more HIV/HCV-coinfected individuals were completely abstinent (59.9%) than HCV-monoinfected patients (51.0%, P=0.025), and they consumed three or more alcoholic drinks (>50g) per day less commonly than monoinfected individuals (17.2% vs. 27.2%, P=0.003).
Among the HIV-positive patients with chronic hepatitis C, 95.2% were receiving HAART for a median time period of 3.6 years (interquartile range, 2.2-4.2 years) with 87% of patients receiving HAART for one year or more. HAART consisted of a protease inhibitor (PI) plus nucleoside analogues (NAs) in 52% of patients, of a non-nucleoside reverse transcriptase inhibitor (NNRTI) plus NAs in 18%, and of three or more NAs in 30%. There was no difference in duration of HAART between patients with and without suppression of HIV RNA to undetectable levels. Among the 141 patients with HIV RNA <400copies/mL, only five (3.5%) were not on HAART and had never been treated before.
The FPR was also more rapid in HIV-infected patients with a CD4+ cell count of <200/mm3 than in those with CD4+ cells of 200/mm3 or more (0.161 vs. 0.131 IshFU/yr, P=0.041), and the same was seen with a CD4+ cell cutoff of 350/mm3. When HIV-infected subjects were subdivided into CD4+ cells <500 and ≥500/mm3, the FPR was more rapid for any HIV viral load compared to undetectable HIV RNA in patients with <500 CD4+ cells per mm3, but not in those with CD4+ cells of 500/mm3 or more (Fig. 2).
Fig. 2. Fibrosis progression rate by CD4+ cell count in HIV-infected patients with chronic hepatitis C. IshFu/yr, Ishak fibrosis units per year; HIV RNA level in copies per mL; CD4+cells per mm3
A linear regression analysis was performed to examine whether log10 HIV RNA level or absolute CD4+ cell count and other baseline factors were correlated with FPR. In HIV/HCV-coinfected patients, log10 HIV RNA load and CD4+ cell count, but not treatment with HAART, were significantly correlated with FPR in univariate analysis. In all patients, a significant correlation with FPR existed for necroinflammation, age at HCV infection, ALT level and alcohol consumption (Table 2). There was no correlation between the duration of HAART and FPR, both in patients with and without HIV viremia.
In multivariable linear regression analysis for all patients (HIV positive or negative), only Ishak necroinflammation score, age at HCV infection, and alcohol consumption were independently correlated to FPR. The same was seen in HCV monoinfection only. In HIV/HCV-coinfected patients, log10 HIV RNA level, Ishak necroinflammation score, and age at HCV infection were independently correlated to FPR, but not CD4+ cell count and alcohol consumption (Table 3).
Necroinflammation grade, fibrosis stage, and proportion cirrhosis
The mean Ishak necroinflammation grade was no different between HIV-positive and HIV-negative patients with chronic hepatitis C (Table 4). However, as with FPR, this only held true for HIV/HCV-coinfected patients with undetectable HIV RNA level compared with HCV-monoinfected patients (6.52 vs. 6.53, P=0.95). Patients with any detectable HIV viral load had a significantly higher necroinflammatory activity than HIV-negative patients (7.41 vs. 6.53, P=0.012) and than HIV-positive patients with HIV viral load <400copies/mL (7.41 vs. 6.52, P=0.035). Multivariable linear regression analysis identified ALT level as the only variable independently correlated with necroinflammation score (beta=0.013, P<0.001, r2=0.108 for model, all patients), and the same result was seen in the HIV/HCV-coinfected subgroup.
The mean Ishak fibrosis score (Table 4) was the same in HCV-monoinfected and all HIV/HCV-coinfected patients (2.88 vs. 2.86, P=0.88). However, fibrosis was more advanced in HIV/HCV-coinfected patients with any detectable HIV viral load compared to those with undetectable HIV RNA level (3.09 vs. 2.65, P=0.041). Multivariable linear regression analysis found serum ALT level (beta=+0.005, P<0.001) and age at biopsy (beta=+0.025, P=0.007) to be independently correlated with fibrosis stage (r2=0.061) when all patients (HCV and HIV/HCV) were considered. In HIV/HCV-coinfected patients, by contrast, only ALT level (beta=+0.003, P=0.028) was independently correlated with fibrosis stage (r2=0.031). There was no difference between HIV/HCV-coinfected and HCV-monoinfected patients in the proportion of patients with minimal fibrosis, bridging fibrosis, or cirrhosis. The rates were also similar between HIV-infected patients with and without detectable plasma HIV RNA. Multivariable logistic regression analysis of all subjects found necroinflammation score ≥10 (OR, 11.1; 95% CI, 5.8-21.2; P<0.001) and daily alcohol intake of ≥100 grams per day (OR, 4.54; 95% CI, 1.8-11.8; P=0.002) to be independently associated with cirrhosis. In HIV/HCV-coinfected patients, only necroinflammation was independently predictive of cirrhosis (OR, 5.9; 95% CI, 2.4-14.0).
DISCUSSION
This study shows that suppression of HIV infection through HAART is associated with a slower HCV-induced liver fibrosis progression rate, and that HIV viremia but not CD4+ cell count independently predicts FPR, a phenomenon not previously described. With the understanding that HCV-related fibrosis progression is not linear over time but tends to accelerate with age [18], it is assumed here that this reduced fibrosis progression rate represents the average fibrosis progression over the entire period of HCV infection with slower progression in some time periods and faster progression in others. The study data suggest that periods with slower progression are the ones when plasma HIV RNA is suppressed to undetectable levels. The study is limited by its retrospective nature and the reliance on patient recall for the time estimate of initial HCV infection. Furthermore, liver biopsies may underestimate stage of fibrosis in small samples (as is the case for all studies of liver histology), but in our centers only biopsy samples with sufficient size and number of portal tracts were accepted for interpretation. However, even within these limitations, it appears that the FPR rates obtained here reflect the natural history of HCV-related fibrosis progressing with a reasonable accuracy. Our study found a fibrosis progression rate of 0.128 Ishak fibrosis units per year in HCV-monoinfected patients, a rate virtually identical to those reported in two other studies using the Ishak score. Ghany and coworkers measured an FPR of 0.12 IshFU/yr in 123 untreated HCV-monoinfected subjects with paired biopsies (mean 3.7 years apart) [19], and Mohsen and colleagues measured 0.13 IshFU/yr in 153 patients with chronic hepatitis C without HIV using our FPR model with a single biopsy and the estimated duration of HCV infection [17].
Our findings reinforce abundant data demonstrating that HIV suppression through HAART can reduce or resolve the adverse effects of HIV infection. Randomized controlled trials have shown that initiation of HAART is associated with a suppression of HIV viremia, an increase of CD4+ cell count as well as reduced morbidity from opportunistic infections and improved mortality [20]. HAART can also reverse or ameliorate non-immunological effects of HIV infection such as anemia [21], AIDS dementia [22], and opportunistic malignancies such as Kaposi sarcoma, lymphoma, and cervical cancer [23-26].
Based on the broad salutary effects of HIV suppression through HAART, it is biologically plausible that suppression of HIV infection with successful antiretroviral therapy might also reverse the accelerating effect of HIV on fibrosis progression in chronic hepatitis C. This idea is now supported by our data. HIV/HCV-coinfected patients have the same FPR as HIV-negative patients with hepatitis C if their HIV viral load is suppressed to undetectable levels. However, coinfected patients with HIV viremia ≥400copies/mL continue to have a more rapid fibrosis progression than HCV-monoinfected persons. These fibrosis progression rates are associated with differences in necroinflammatory activity, which is similar between HCV monoinfection and HIV/HCV coinfection where plasma HIV RNA is undetectable but significantly increased in coinfected patients with ongoing HIV viremia. Prospective studies have shown that high activity of necroinflammation leads to more rapid progression to cirrhosis [1], and in our study (as in others) necroinflammation was highly and independently correlated with FPR.
Our data contrast with those of previous reports documenting an accelerated fibrosis progression rate in HIV/HCV-coinfected patients compared to HCV-monoinfected persons [4,16,17,27]. However, these previous studies did not distinguish HIV/HCV-coinfected patients between those with undetectable vs. detectable plasma HIV RNA. Our study indicates that whether a given cohort of HIV/HCV-coinfected patients has a more rapid FPR than HCV-monoinfected subjects primarily depends on the percentage of coinfected patients with undetectable HIV RNA. It is, therefore, likely that among the HIV/HCV-coinfected patients of the previous studies, many had uncontrolled HIV viremia which contributed to a higher FPR in the coinfected group. Our study does support previous data [4,16,17] showing that age at HCV infection, daily alcohol intake, and necroinflammation are independent predictors of FPR in HIV/HCV-coinfected persons.
Our data suggest that HIV viremia rather than low CD4+ cell count leads to a more rapid fibrosis progression from HCV. Unlike the previous studies, we did not find an independent correlation between FPR and CD4+ cell count when controlled for HIV viral load, necroinflammation and age at HCV infection. Even when HIV viral load is excluded from the linear regression model, CD4+ cell count still does not independently correlate with FPR (data not shown). Furthermore, in our study HAART itself was not correlated with FPR, but instead successful HAART (undetectable HIV RNA) was. This is explained by the fact that HAART does not always lead to undetectable HIV RNA due to viral resistance or suboptimal adherence to HIV medication. On the other hand, almost all HIV/HCV-coinfected patients with undetectable HIV viral load (<400copies/mL) were concurrently receiving HAART (96.5%). The remaining 3.5% of anti-HIV(+) subjects had no HIV viremia without ever having been treated. Their CD4+ cell counts were all normal (690-1100/mm3), and they probably spontaneously cleared HIV during the initial acute infection [28].
The impact of HAART on FPR in HIV/HCV-coinfected persons was described in one study by Benhamou and colleagues which was conducted from 1995 to 1998, i.e. around the time of introduction of HIV protease inhibitors (PIs) and HAART (1996). The FPR was faster in patients who were not on HAART than in patients receiving HAART with a PI-containing regimen. Also, the study found CD4+ cells <200/mm3 but not plasma HIV RNA ≥200copies/mL to be independently associated with FPR. Conversely, our data show that HIV viral load but not CD4+ cell count is independently associated with FPR. This discrepancy may be explained by the fact that in Benhamou's study, only one-third of coinfected patients were receiving HAART for a median duration of 14 months, and HIV RNA levels may not have been completely suppressed in the early phase of therapy.
Quirishi and coworkers published data suggesting that HIV/HCV-coinfected patients receiving HAART have a lower liver-related mortality that those being treated with HIV nucleoside analogues only and those not on antiretroviral therapy [7]. However, subsequent discussions have raised questions about survivorship bias and the appropriateness of the Kaplan-Meier analysis for liver-related deaths and called the validity of the study results into question [29-31].
When examining the actual fibrosis stage on biopsy rather than the dynamic fibrosis progression, we observed no difference between HIV/HCV-coinfected and HCV-monoinfected patients. However, coinfected patients with undetectable HIV RNA had a slightly lower Ishak fibrosis stage than those with uncontrolled HIV viremia, but there was no difference in the rate of minimal fibrosis, bridging fibrosis, or cirrhosis. These findings contrast with previous studies that all found HIV/HCV-coinfected patients to have a more advanced fibrosis stage and a higher frequency of cirrhosis than HCV-monoinfected individuals [4,16,17,32]. However, those studies did not provide the status of HIV viremia in relationship to the fibrosis stage.
Our data suggest a potential role for HAART in asymptomatic HIV/HCV-coinfected patients with CD4+ cells <500/mm3 to slow down liver fibrosis progression, not just in those with advanced immunodeficiency. Initiation of HAART has always been recommended for patients with symptomatic HIV disease, but recommendations on initiation of HAART in asymptomatic HIV patients have been evolving based on more recent data. In a recent report, a panel of the International AIDS Society-USA recommended HAART initiation in asymptomatic patients with CD4+ cells <200/mm3 and endorsed individualized decisions to start therapy in asymptomatic patients with 200 or more CD4+ cells per mm3[33]. Rapidly declining CD4+ cell count and high HIV viral load are suggested indications to initiate HAART which need to be weighed against the potential of long-term toxicity and patient choice. Based on our data, HIV/HCV-coinfection could be added to this list of considerations to initiate HAART in an asymptomatic patient. HAART would then be initiated for the purpose of slowing down HCV-related fibrosis progression when CD4+ cells are stable between 200/mm3 and 500/mm3, and stable HIV viral load would not prompt initiation of HAART for immunological reasons. However, the potential benefits of such an approach would need to be weighed against the cost and inconvenience of taking HIV medications as well the potential toxicity of long-term HAART such as rare but serious mitochondrial toxicity [34], hepatotoxicity [35], lipodystrophy, and the metabolic syndrome [36].
The ideal approach to slow and even reverse liver fibrosis progression in HIV/HCV-coinfected patients is, of course, by achieving a sustained virologic response (SVR) through anti-HCV therapy. A recent trial using peginterferon alfa-2a plus ribavirin showed an SVR rate of 40% in HIV/HCV coinfection, 29% in genotype 1 and 62% in genotypes 2 or 3 [13]. But many coinfected patients do not achieve an SVR, cannot tolerate combination therapy of pegylated interferon and ribavirin, have medical contraindications against therapy, or do not want to be treated [37].
Ongoing prospective studies of HIV/HCV-coinfected patients will further clarify factors that influence fibrosis progression based on multiple liver biopsies and longitudinal data collection. Meanwhile, it may be worthwhile considering HAART for asymptomatic HIV/HCV-coinfected patients with CD4+ cells <500/mm3 (regardless of HIV viral load) who cannot clear HCV infection with antiviral therapy.
Introduction
Chronic hepatitis C leads to various degrees of necroinflammation and fibrosis in the liver. Fibrosis can ultimately progress to cirrhosis, which may lead to liver failure, hepatocellular carcinoma, and death. Progression to cirrhosis is faster with higher necroinflammatory activity on histology [1]. Two factor have been associated with more rapid fibrosis progression induced by the hepatitis C virus (HCV), heavy alcohol consumption and coinfection with the human immunodeficiency virus (HIV). Progression to cirrhosis is more frequent in HIV/HCV coinfection than in HCV monoinfection [2,3]. In HIV/HCV-coinfected patients, fibrosis progression is further accelerated with advanced immunodeficiency [4]. In addition to histological progression, clinical progression to liver decompensation or death was also shown to be accelerated in HIV/HCV-coinfected patients [5]. These studies were all conducted before highly active antiretroviral therapy (HAART) had a major impact on morbidity and mortality in HIV-infected patients. One preliminary study found that HIV/HCV-coinfected patients who are taking HAART have a slower fibrosis progression than those who are not [6], and a recent study suggested that HAART may reduce liver-related mortality in HIV/HCV-coinfected persons [7].
Patient selection
The study retrospectively examined 925 consecutive patients with chronic hepatitis C (with or without HIV coinfection), compensated liver disease and no drug-induced hepatotoxicity who were evaluated for possible therapy at two viral hepatitis referral centers (San Juan, Puerto Rico and Bronx, New York) between January 1999 and December 2002. Patients were included in the analysis if they were HCV treatment-naïve, had no other cause of liver disease, had an estimated date of HCV infection, had been tested for HIV antibody, and had a liver biopsy scored according to the Ishak classification [8]. HIV/HCV-coinfected patients were included in the study if they had CD4+ cell count and plasma HIV RNA level measured within 1 year of the liver biopsy.
Data collection and outcome measures
Historical data that were routinely obtained during the initial clinical evaluation at both sites included demographics, risk factors for hepatitis C virus (HCV) infection, the estimated time of initial HCV infection, i.e. the time from the onset of the main risk factor (blood transfusion, injection-drug use [IDU], high-risk sexual behavior, etc.), and average daily alcohol consumption in the 5 years prior to the evaluation. The estimated onset of risk behavior and alcohol consumption was obtained from the patients through structured interviews. In HIV-infected patients, the current HAART regimen (if any) was recorded. Laboratory values included serum alanine aminotransferase (ALT) level, plasma HCV RNA level (viral load), HCV genotype, and hepatitis B serology as well as CD4+ cell count and HIV viral load in HIV-positive patients. For the purpose of defining HIV viral load as a continuous variable, HIV RNA levels of <400copies/mL were arbitrarily defined as 200copies/mL (median between lower limit of detection and zero), and likewise HIV RNA levels of <50copies/mL were defined as 25copies/mL. Each liver biopsy was scored according to the Ishak method [8] by a hepatopathologist (A.F. and F.P.) who was blinded to the patient's HIV status. Liver biopsy specimens were only read if the hepatopathologist considered them adequate for interpretation, which typically included a minimum length of 15mm and at least 12 visible portal tracts. In patients with uninterpretable biopsy specimens, a liver biopsy was repeated. The Ishak scoring system semi-quantitatively measures the grade of necroinflammation (0-18) and the stage of fibrosis (0-6) and has been utilized in a variety of studies of natural history [9,10] and therapy of chronic hepatitis C [11-14]. The fibrosis progression rate (FPR) was calculated as the ratio of the Ishak fibrosis score (in Ishak fibrosis units, IshFU) divided by the estimated duration of HCV infection in years (time from estimated HCV infection to biopsy). This model was first introduced by Poynard and colleagues [15] and has since been used in multiple studies examining the natural history of HCV-related fibrosis progression including comparison of FPR between HIV/HCV-coinfected and HCV-monoinfected patients [4,16,17]. The model represents the average progression of liver fibrosis over time, regardless of different rates of fibrosis progression in different time intervals and does not claim to represent linear progression. It has indeed been shown that fibrosis progression is not linear over time but accelerates somewhat with increasing duration of HCV infection [18]. The primary outcome measure of the study was the comparison of the FPR between HIV/HCV-coinfected and HCV-monoinfected patients. Secondary outcome measures were comparison of the FPR within the HIV/HCV-coinfected group by HIV viral load and CD4+ cell count. Other outcome measures were Ishak necroinflammation grade, Ishak fibrosis stage, and proportion of patients with bridging fibrosis and cirrhosis.
|
|
| |
| |
|
|
|
