Fast fibrosis progression between repeated liver biopsies

Fast fibrosis progression between repeated liver biopsies in patients coinfected with human immunodeficiency virus/hepatitis C virus

"Fibrosis progresses with high frequency in HIV/HCV-coinfected patients over a period of time of 3 years."

Hepatology Oct 2009

Juan Macías 1, Juan Berenguer 2, Miguel A. Japón 3, José A. Girón 4, Antonio Rivero 5, Luis F. López-Cortés 6, Ana Moreno 7, Mercedes González-Serrano 8, José A. Iribarren 9, Enrique Ortega 10, Pilar Miralles 2, José A. Mira 1, Juan A. Pineda 1 *

1Infectious Diseases Unit, Hospital Universitario de Valme, Seville, Spain

2Infectious Diseases Unit, Hospital Universitario Gregorio Marañón, Madrid, Spain

3Pathology Department, Hospital Universitario Virgen del Rocío, Seville, Spain

4Infectious Diseases Unit, Hospital Universitario Puerta del Mar, Cadiz, Spain

5Infectious Diseases Unit, Hospital Universitario Reina Sofía, Cordoba, Spain

6Infectious Diseases Department, Hospital Universitario Virgen del Rocío, Seville, Spain

7Infectious Diseases Unit, Hospital Universitario Ramón y Cajal, Madrid, Spain

8Infectious Diseases Unit, Hospital Universitario Virgen de la Victoria, Malaga, Spain

9Infectious Diseases Unit, Hospital Donostia, San Sebastian, Spain

10Infectious Diseases Unit, Hospital General Universitario de Valencia, Valencia, Spain

email: Juan A. Pineda (japineda@telefonica.net)

from Jules: although this study and others find undetectable HIV viral load is protecting of fibrosis progression, still 42% of patients with undetectable HIV RNA progressed 1 or more stages within the 3 year period that paired biopsies were performed in this study, suggesting that HIV can accelerate HCV disease progression despite successful HAART.

Table 4. Factors Associated with Observed Fibrosis Progression 1 or More Stages Between Liver Biopsies
Picture 18.png

Picture 21.png

Abstract


A few studies have assessed the observed fibrosis progression between serial liver biopsies (LB) in human immunodeficiency virus (HIV) / hepatitis C virus (HCV)-coinfected patients. Approximately half of the patients progressed at least one fibrosis stage over a short period of time. The risk factors for this fast progression need clarification. Because of this, we evaluated the observed fibrosis progression rates of HIV/HCV-coinfected patients and the risk factors for accelerated progression. Overall, 135 HIV-infected patients with positive serum HCV RNA, without other possible causes of liver disease, who underwent two LB, separated at least by 1 year, were included in this retrospective cohort study. The median (Q1-Q3) time between both LBs was 3.3 (2.0-5.2) years. Patients showed the following changes in fibrosis stage: regression >/=1 stage: 23 (17%), no change: 52 (39%), progression 1 stage: 38 (28%), and progression >/=2 stages: 22 (16%). Seventeen (13%) patients had cirrhosis in the second biopsy. Factors independently associated with progression >/=1 stage were undetectable plasma HIV RNA during the follow-up (relative risk [RR] [95% confidence interval, 95% CI] 0.61 [0.39-0.93], P = 0.03), moderate-to-severe lobular necroinflammation (1.77 [1.16-2.7], P = 0.009), time between biopsies (1.11 [1.08-1.2], P = 0.01), and end of treatment response to anti-HCV therapy (0.41 [0.19-0.88], P = 0.02). Conclusion: Fibrosis progresses with high frequency in HIV/HCV-coinfected patients over a period of time of 3 years. Absent-to-mild lobular necroinflammation at baseline, achievement of response with anti-HCV treatment, and effective antiretroviral therapy are associated with slower fibrosis progression.

Article Text


The natural course of chronic hepatitis C virus (HCV) is accelerated in human immunodeficiency virus (HIV) infection.[1] Low CD4 cell counts, lack of exposure to antiretroviral therapy (ART), and advanced liver disease are risk factors for liver events and mortality in cohort studies.[1-5] Particularly, CD4 cell recovery and suppression of HIV replication after starting highly active antiretroviral therapy (HAART) are predictors of survival free of complications of cirrhosis.[5]


Most data on liver fibrosis progression in HIV and HCV coinfection come from cross-sectional studies that relied on a single liver biopsy.[6-11] In those studies, the date of HCV infection is usually estimated as the date of the first use of injection drugs. At HCV infection, it is assumed that liver fibrosis is absent. The majority of the duration of HCV infection was not observed, and factors that could influence fibrosis progression were only collected during a relatively short period of the entire course of HCV infection. These assumptions are probably one reason for the conflicting results of those studies, especially regarding the effect of ART on fibrosis progression.[6-11]


Serial liver biopsy studies may have some advantages over single liver biopsy studies. Changes in fibrosis stage are observed between two dates, and fibrosis at the initial date is known. Thus, the estimations and assumptions of cross-sectional studies on fibrosis progression are not made. Additionally, factors that could influence fibrosis progression are observed during the period of time between biopsies, and are more easily and reliably gathered. Serial liver biopsy studies in HIV/HCV coinfection have shown unexpected high rates of fibrosis progression over short periods of time.[12-14] However, these studies are scarce and have drawbacks. Some of them are limited due to small sample size.[12][13] The largest cohort study on repeated liver biopsies mostly included African-American men,[14] which may not be representative of other populations.


Given the limited information on the characteristics of patients at higher risk for rapid fibrosis progression, studies on this topic are needed. The results of these studies would allow us to act on the factors associated with faster progression, and, thus, ultimately to achieve a better long-term outcome of liver disease. Due to these issues, we aimed this study at evaluating the rates and risk factors of observed liver fibrosis progression in HIV/HCV-coinfected patients.


Abbreviations


ALT, alanine aminotransferase; ART, antiretroviral therapy; CDC, Centers for Disease Control and Prevention; ETR, end of treatment response; eFPR, estimated fibrosis progression rate; FPR, fibrosis progression rate; FU, fibrosis units; HAART, highly active antiretroviral therapy; HCV, hepatitis C virus; HIV, human immunodeficiency virus; SVR, sustained virological response


Results


Baseline Characteristics.


A liver biopsy was performed in 2,285 (25%) of 9,257 patients from the cohort during the study period. Of those with liver biopsy, 156 (7%) patients underwent two liver biopsies. Twenty-one patients were excluded from the analysis. Eleven of them showed cirrhosis in the first liver biopsy. Four patients had inadequate biopsy samples due to fragmentation or length shorter than 10 mm. Six had liver biopsies taken with a purpose different than screening for HCV treatment, namely, evaluation of fever in patients with AIDS. Finally, 135 patients were included in the study cohort.


The characteristics of 135 study patients at the time of the first liver biopsy compared with the 2,129 patients with a single liver biopsy are shown in the Table 1. Most patients were men with a past habit of injection drug use. The majority of the cohort was treated with HAART at baseline (Table 1). Eighty-two (75%) of the 110 subjects on HAART during the follow-up showed plasma HIV RNA levels below the limit of detection in more than 70% of the determinations. The median (1st quartile, Q1; 3rd quartile, Q3) CD4 cell counts of patients without HAART were 495 (258-587) and of patients with HAART, it was 500 (338-699) (P = 0.3). The median (Q1-Q3) nadir CD4 cell counts of patients without HAART was 346 (195-409) and of those with HAART, it was 194 (98-312) (P = 0.03). The distribution of individual antiretroviral drugs and of treatment regimens prescribed during the follow-up is represented in the Table 2. The median (Q1-Q3) length of the initial biopsy was 17 (15-25) with a range of 12-30 mm. The median (Q1-Q3) length of the second biopsy was 18 (16-25) with a range of 15-30 mm.

Fibrosis Progression.


The median (Q1-Q3) period of time between biopsies was 3.3 (2.0-5.2) years. Fibrosis stages at first and second liver biopsies are displayed in the Table 3. At the second biopsy, 52 (39%) patients did not show changes in the fibrosis stage, and in 23 (17%) subjects a regression of one or more stages was observed. Fibrosis progression of one stage was detected in 38 (28%) individuals, and 22 (16%) patients progressed two or more stages. Seventeen (13%) patients had cirrhosis in the second biopsy. Similarly, 25 (32%) of the 78 patients with stage 0 or 1 in the baseline biopsy progressed one stage, and 19 (24%) increased their fibrosis stage by two or more points. Six (7.7%) individuals with absent or mild fibrosis at the initial biopsy showed cirrhosis in their final biopsy. Fibrosis progression according to exposure to anti-HCV therapy is displayed in Fig. 1.

Picture 19.png
Picture 20.png

Median (Q1-Q3) observed FPR between liver biopsies was 0 (0-0.23) FU/year, whereas eFPR before the first biopsy was 0.077 (0-0.13) FU/year. The correlation between observed FPR and eFPR was low (rho = 0.28, P = 0.001). Among patients with fibrosis progression, median (Q1-Q3) observed FPR was 0.36 (0.19-0.6) FU/year. For patients with fibrosis stage 0 at first biopsy who showed progression, median (Q1-Q3) observed FPR was 0.46 (0.19-0.53) FU/year. Thus, if this FPR remained constant the median (Q1-Q3) time to cirrhosis of these individuals would be 8.7 (7.5-21.1) years.


Factors Associated with Liver Fibrosis Progression.


The univariate analysis of associations with fibrosis progression is presented in the Table 4. HAART was prescribed to 110 (81%) patients between liver biopsies. Patients exposed to HAART were less likely to progress in the univariate analysis. Among these individuals on HAART, 43 (39%) were only prescribed protease inhibitor-based regimens and 34 (31%) only took nonnucleoside analog-based combinations during the follow-up. No increase in the fibrosis stage between biopsies was observed in 26 (61%) patients taking only protease inhibitors as the third drug in their ART, compared with 9 (64%) of 14 patients having only nevirapine-based regimens (P = 0.8), and with 12 (60%) of 20 individuals taking only efavirenz-based HAART (P = 0.9)

After multivariate analysis, moderate to severe necrosis and inflammation in the initial biopsy, time between liver biopsies, ETR to anti-HCV therapy, and plasma HIV RNA below the detection limit in more than 70% of the determinations during the follow-up were independent predictors of fibrosis progression (Table 5). Variables independently associated with FPR higher than 0.1 FU/year were moderate to severe necrosis and inflammation in the initial biopsy (RR [95% CI] 1.75 [1.17-2.61], P = 0.02), ETR to anti-HCV therapy (RR [95% CI] 0.34 [0.14-0.86], P = 0.007), and plasma HIV RNA below the detection limit in 70% of the determinations during the follow-up (RR [95% CI] 1.66 [1.13-2.42], P = 0.009). In a subanalysis restricted to patients without anti-HCV treatment between liver biopsies, time between liver biopsies (per year, RR [95% CI] 1.1 [1.03-1.18], P = 0.003), moderate to severe necrosis and inflammation in the initial biopsy (RR [95% CI] 1.6 [1.03-2.33], P = 0.04), and plasma HIV RNA below the detection limit in 70% of the determinations during the follow-up (RR [95% CI] 1.68 [1.11-2.53], P = 0.01) were independent predictors of fibrosis progression.

Discussion



In this study, we found a high frequency of fibrosis progression among HIV/HCV-coinfected patients. Over a period of 3 years, 44% of patients progressed one or more stages of fibrosis. Progression of fibrosis was predicted by the degree of necrosis and inflammation in the first biopsy. ETR with anti-HCV therapy and suppression of HIV replication with potent ART also slow or prevent fibrosis progression.


The rates of fibrosis progression observed in the present cohort are in agreement with a previous report.[14] In that cohort study, 46% of patients experienced an increase of one or more stages in the Ishak fibrosis score (0-6 stages) over a period of 2.9 years.[14] A quarter of the patients progressed two or more Ishak stages. Importantly, 9% of patients developed cirrhosis, and 4.6% of patients with mild fibrosis, i.e., Ishak score equal or less than 2, harbored cirrhosis in the second biopsy.[14] In our cohort study we found that 13% of our study patients presented cirrhosis in the final biopsy, and 7.7% of patients with initial absent or mild fibrosis bore cirrhosis in the final biopsy. By contrast, serial liver biopsy studies in HCV-monoinfected patients show progression of two or more Ishak stages in 10% of patients over similar periods of time.[18][19] Development of cirrhosis was seen in 14% of HCV-monoinfected individuals, and in 1.6% of those with mild fibrosis at first biopsy.[18]


We observed expected associations with fibrosis progression. High degrees of necrosis and inflammation at the initial liver biopsy and achievement of ETR were predictors of progression. First, necroinflammatory activity has been reported as a risk factor for fibrosis progression in HCV-monoinfected patients.[18] Moreover, elevated AST serum levels, presumably indicative of liver inflammation, have also been associated with fibrosis progression in HIV/HCV-coinfected subjects.[14] Second, we found that achievement of ETR was associated with a lower probability of fibrosis progression. Regression of liver fibrosis has been observed in HIV-infected and uninfected subjects with chronic hepatitis C who achieve SVR[20][21] and in HCV-monoinfected patients treated with pegylated interferon plus ribavirin without SVR.[22]


Suppression of HIV replication through HAART was associated with a slower estimated fibrosis progression rate in a single liver biopsy study.[9] However, paired liver biopsy studies in coinfected patients have not found any relation between HAART and fibrosis progression.[12-14] On the contrary, in a prospective cohort of ART-naïve patients, HAART-mediated control of HIV RNA was a protective factor for liver-related death.[5] We found that fibrosis progression between liver biopsies was independently related with HIV RNA suppression both in the global study group and in the patients without anti-HCV treatment. A possible mechanism underlying the association of HIV suppression and fibrosis progression is the effect of HIV on fibrogenesis. In an in vitro model, either gp120 or inactivated HIV enhanced TGF-beta 1 expression in both uninfected and HCV-infected hepatocytes.[23] This cytokine is a pivotal mediator of fibrogenesis in the liver.


This study may have a few limitations. First, the association between fibrosis progression and HIV RNA suppression might be confounded by a number of covariates. Patients with effective HAART could be more compliant with follow-up and treatments, and therefore be more likely responders to anti-HCV therapy. However, the association between HIV RNA suppression and fibrosis progression was adjusted by that confounder using a conservative multivariable method.[17] HIV RNA suppression was found to be a predictor of fibrosis progression independent of response to anti-HCV treatment or liver inflammation. In addition, this association was confirmed in the subgroup of patients without anti-HCV treatment. Second, as expected, the length of time between the first and the second liver biopsies accounted for fibrosis progression. Time between liver biopsies might have confounded other associations found in the present study. However, the main multivariate analysis was adjusted by time between liver biopsies. In addition, we performed a second analysis with FPR, which integrates time between biopsies, as the outcome variable. The results of this analysis confirmed the main analysis results. Third, patients who undergo serial liver biopsies are a highly selected group. Individuals that accept to undergo a second biopsy are usually highly compliant with follow-up and treatments. Most probably, these individuals are not representative of the general coinfected population. Patients with paired liver biopsies were younger and less exposed to alcohol than those with a single biopsy. However, both groups were similar in factors potentially involved in fibrosis progression such as CD4 cell counts, HIV RNA suppression, or baseline necroinflammatory activity. Thus, the study patients might be representative of HIV/HCV-coinfected individuals eligible for liver biopsy.


The risk factors for progression identified in the present study could have clinical implications. First, we identified, as others,[18] necrosis and inflammatory activity as a risk factor for fibrosis progression. Unfortunately, this information can only be obtained through liver biopsy samples. The development of noninvasive tests has focused on the evaluation of fibrosis, but little attention has been paid to noninvasive prediction of liver inflammation.[24] Second, ART prescription for HCV/HIV-coinfected patients with CD4 cell counts greater than 350/L is considered a possible option under current guidelines.[25] The impact of immunosuppression on liver mortality is one reason for this recommendation.[2-5] In the present study, lack of control of HIV replication was a risk factor for progression. This is an additional reason to consider earlier ART initiation. Third, efforts should be made to provide anti-HCV treatment to coinfected patients and to optimize treatment schedules to achieve maximal rates of SVR. Anti-HCV treatment is the best-proven means of preventing chronic hepatitis C progression and complications.[5][20][21][26][27] Fast fibrosis progression in HIV/HCV-coinfected patients observed in this study and by other authors,[12-14] and specifically progression of patients with initial mild fibrosis, supports that pegylated interferon plus ribavirin should be indicated in this population regardless of the fibrosis stage.


In conclusion, HIV/HCV-coinfected patients experience fast fibrosis progression frequently. Coinfected patients with increased inflammation in the liver biopsy are at risk for accumulation of fibrosis, and anti-HCV treatment should be started in them. Currently, pegylated interferon plus ribavirin is recommended in coinfected patients with significant fibrosis. Follow-up and repeated biopsy is suggested for those with mild fibrosis. However, given the high rate of fibrosis progression over short periods of time in HIV/HCV-coinfected patients with mild fibrosis, anti-HCV treatment should also be considered in them. Finally, HAART should be started as early as possible in coinfected patients in whom HCV has not been eradicated.

Patients and Methods

Abstract Patients and Methods Results Discussion Acknowledgements References


Patients.


This was a multicohort study gathering patients from two Spanish prospective cohorts of patients with HIV and HCV coinfection. One of the cohorts was a collaboration of 11 centers, and the other cohort recruited patients from five centers from January 1997.[4][5] All individuals were evaluated at clinical visits scheduled every 3 months. Nine university hospitals participating in these cohorts followed 9,257 HIV/HCV-coinfected individuals from January 1986 to January 2008. This was a retrospective analysis of those patients that fulfilled the following criteria: (1) Active HCV infection as confirmed by serum HCV RNA; (2) Two or more liver biopsies, obtained as part of the evaluation of HCV infection to assess prognosis and establish indication for treatment, separated by at least 1 year; (3) No other concomitant causes of liver disease. Patients with cirrhosis in the first liver biopsy were excluded.


Antiviral Therapy.


Patients received HAART according to the availability of drugs and the recommendations from a panel of experts in force. Particularly, the indication of starting HAART changed during the study period due to the accumulating evidence on drug toxicity and risk of HIV progression. However, the final decision on starting HAART and the specific drug combination was arrived at by the caring physician.


Anti-HCV treatment was prescribed usually considering the presence of fibrosis in the liver biopsy. Pegylated interferon plus ribavirin became the standard of care from 2001. Standard interferon alone or in combination with ribavirin was available before. Management of anti-HCV therapy was carried out according to accepted guidelines during the study period. End of treatment response (ETR) was defined as undetectable serum HCV RNA at the planned date of treatment termination. Sustained virological response (SVR) was defined as undetectable serum HCV RNA 24 weeks after completing treatment.


Assessment of Fibrosis.


Liver biopsy specimens were centrally read and scored blindly by two independent pathologists using the Scheuer classification for staging fibrosis and for grading inflammation and necrosis.[16] The necroinflammatory components of the Scheuer index include lobular inflammation and necrosis (0-4) and portal inflammation (0-4). The Scheuer score stages fibrosis from 0-4 (0, absent fibrosis; 1, portal fibrotic expansion; 2, extension of fibrosis to the lobule, but with few septa; 3, bridging fibrosis with numerous septa, with architectural distortion without cirrhosis; 4, cirrhosis). Steatosis was scored based on the proportion of hepatocytes with fat (0, absent steatosis; 1, less than 33%; 2, 33%-66%; 3, more than 66%). To validate the adequacy of the biopsy, we recorded the length of each biopsy. The median length of the initial biopsy was 17 (range, 5-30) mm. The median length of the second biopsy was 18 (range, 6-30) mm. Forty-three (14%) of 312 biopsies obtained from 156 patients had less than 15 mm, the size generally used to define an adequate biopsy. The agreement between the two pathologists was excellent for staging fibrosis ( k= 0.82; agreement 86.8%). Discordant fibrosis staging was resolved by consensus reading.


Statistical Analysis.


The primary outcome variable of this study was worsening of fibrosis, defined as a 1-point increase or more in the fibrosis stage. The secondary outcome variable was the fibrosis progression rate (FPR), calculated as the numerical difference between the fibrosis stage found in the first and second biopsy divided by the time between biopsies in years. FPR was expressed as fibrosis units (FU) per year. The factors that were assessed for association with progression in fibrosis were sex, age, age at HCV infection, risk group, daily alcohol intake, alanine aminotransferase (ALT) level, HCV genotype, serum HCV RNA level, treatment against HCV infection, response to anti-HCV treatment, Centers for Disease Control and Prevention (CDC) stage, CD4 cell counts at liver biopsies, nadir CD4 cell counts, plasma HIV RNA levels at liver biopsies and during the follow-up, exposure to ART, necroinflammatory grade, steatosis grade, and time between liver biopsies. Exposure to ART was classified as either lack of ART, non-HAART, or HAART. Clinical and laboratory data were collected every 6 months, except for HCV genotype and serum HCV RNA. The estimated fibrosis progression rate (eFPR) was calculated as the numerical fibrosis stage found in the first biopsy divided by the duration of HCV infection time in years.


Associations were first examined by univariate analysis. Student's t test or the Mann-Whitney U test, if appropriate, were used to compare continuous variables. The chi square test was applied to compare categorical variables; if the expected frequency of at least one cell was less than 5 the Fisher test was used. Correlations between observed FPR and eFPR were assessed using the Spearman test. Variables with a level of association equal or less than 0.2 in the univariate analysis were entered in a Poisson regression model with robust error variance, considering the time of exposure as 1. The relative risk (RR, 95% confidence interval [95% CI]) of the variables independently associated with fibrosis progression could be estimated in this way.[17] This approach was used because when the outcome event is common (incidence of 10% or more), odds ratios, calculated by a logistic regression model, may largely overestimate the risk.[17] FPR was categorized by the value of 0.1 FU per year to introduce it as the outcome variable in a Poisson regression model with robust error variance.


Statistical analysis was carried out with the Stata statistical software package (Stata 9.0, College Station, TX).


Ethical Aspects.


The study was performed according to the Helsinki Declaration and was approved by the Ethics Committee of Hospital Universitario de Valme.