| |
Effect of antiretroviral drugs on liver fibrosis in HIV-infected patients with chronic hepatitis C: harmful impact of nevirapine
|
| |
| |
AIDS: Volume 18(5) 26 March 2004 pp 767-774
Macías, Juana,b; Castellano, Victorc; Merchante, Nicolása; Palacios, Rosa Ba; Mira, José Aa; Sáez, Carmend; García-García, José Aa; Lozano, Fernandob; Gómez-Mateos, Jesús Mb; Pineda, Juan Aa,b
From the aServicio de Medicina Interna, bUnidad de Enfermedades Infecciosas, cServicio de Anatomia Patológica, Hospital Universitario de Valme, and dServicio de Anatomia Patológica, Hospital Universitario Virgen del Rocio, Seville, Spain.
This was a cross-sectional study of HIV/HCV co-infected patients from a cohort of HIV-infected patients prospectively followed at the Infectious Diseases Unit of the Hospital Universitario de Valme, at Seville, southern Spain.
Brief Summary: the authors say: "In this study, we found that the use of HAART regimens including nevirapine is associated with an increased degree of liver fibrosis in HIV-infected patients with chronic hepatitis C. On the contrary, patients on PI as the backbone of potent antiretroviral therapy are more likely to show less liver fibrosis... Chronic hepatitis C has emerged as a serious concern in the HAART era. Coinfected patients can show more unwanted drug reactions. This may complicate the choice of a safe HAART regimen in coinfected patients. The results of this study suggest that HAART including PI could be more advantageous in terms liver fibrosis progression than nevirapine-based regimens in this setting. Whether the associations found in this study have a clinical impact need to be confirmed by randomized prospective studies with hard clinical end-points, such as development of decompensated cirrhosis or death attributable to liver failure".
"...The estimated fibrosis progression rate is defined as the ratio between the fibrosis stage and the estimated duration of HCV infection in years. In this model it is assumed that the patient has no liver fibrosis the day of infection and that the fibrosis progression rate is constant...For example, for a patient with fibrosis stage 2 (F2) and a 10-year duration of HCV infection, the fibrosis progression rate is 0.2 (some might say that fibrosis rate may not be constant during certain periods of infection, thus paired biopsies could better reflect fibrosis progression rate)..."
The authors say about this: "the development of fibrosis does not follow a continuous linear progression. Thus, the stage of liver fibrosis may have been reached a long time before, for example, due to a severe acute liver damage followed by a prolonged period of inactivity. This fact limits the worth of the estimated liver fibrosis progression rate. The only alternative design to achieve a more reliable assessment of the rate of fibrosis progression would be a prospective follow-up with sequential liver biopsy withholding anti-HCV therapy. This study would be very difficult to carry out, mainly due to lack of patient acceptance and ethical concerns. Because of this, most surveys where factors associated with liver fibrosis progression have been investigated were cross-sectional studies in patients with estimated or known date of infection."
"... Sixty-two (40.8%) patients had never received HAART; 47 (75.8%) of them were antiretroviral naive; 90 (59.2%) patients had been prescribed HAART before liver biopsy; 78 (86.7%) of them were prescribed PI at any time before liver biopsy...(so, 25% of patients had ART before HAART)"
"...Advanced liver fibrosis was observed in 21 (29%) and 38 (48%) patients with ALT levels below and above 100 IU/ml, respectively (P = 0.02)"
"...Six (29%) women and 37 (28%) men reported daily alcohol intake > 50 g/day (P = 0.9)..."
"...A significantly greater proportion of patients who had received PI-based therapy showed liver fibrosis below F3 stage (49% vs 31%)... Similarly, 15 (75%) patients at any time on efavirenz-based regimens showed non-advanced liver fibrosis. However, patients who had ever received regimens containing nevirapine demonstrated advanced liver fibrosis more frequently (56% vs 36%)... Eleven (64.7%) of 17 patients who had used nevirapine for more than 1 year showed advanced liver fibrosis (AOR, 3.5; 95% CI, 1.2-10.6; P = 0.03)..."
"...Eight (47%) of 17 subjects who used nevirapine for more than 1 year showed a faster fibrosis progression rate of > 0.2 (AOR, 3.4; 95% CI, 1.8-6.5; P < 0.001)..."
"Age at infection, CD4 cell counts at liver biopsy, exposure to PI at any time of the follow-up and having ever used nevirapine were independently associated with faster fibrosis progression": factors--CD4 count <250 at liver biopsy (0.03), 52% of patients with <250 CD4s at biopsy had faster fibrosis progression rate meaning 48% did not; 19% of patients on PI therapy had faster fibrosis progression rate (<0.001, AOR 0.39); 44% of patients on nevirapine had faster rate of fibrosis progression (<0.001, AOR 3.82); 15% of patients on efavirenz had faster fibrosis progression rate (0.7).
"...The levels of ALT increased 2.5-fold from the baseline levels in 21 (27%) of 78 patients during their follow-up on PI and in 19 (76%) of 25 patients during their follow-up on nevirapine (P < 0.001)..."
"We report, for the first time to our knowledge, more advanced stage of liver fibrosis and faster liver fibrosis progression among coinfected patients exposed to nevirapine. A previous report on this issue included too few patients on NNRTI to draw any conclusion. Nevirapine is a well-characterized hepatotoxic agent. Nevirapine-associated severe hepatotoxicity is more common in patients with hepatitis C infection. There are two possible presentations of toxicity related to nevirapine: an early idiosyncratic reaction and a late onset cumulative toxicity both of which might be implicated in the worsening liver fibrosis among HIV-infected patients with chronic hepatitis C. We found that patients with exposure to nevirapine over 1 year were more likely to show advanced liver fibrosis and faster liver fibrosis progression. These data support that the cumulative nevirapine hepatotoxicity is involved in the development of liver fibrosis in HIV/HCV-coinfected patients. On the other hand, efavirenz was not associated with liver fibrosis progression. Thus, the effect of nevirapine is not related to the NNRTI drug class, but rather to individual characteristics of this drug".
"We found that PI-based antiretroviral regimens were associated with non-advanced liver fibrosis. We also found a lower rate of fibrosis progression among patients exposed to PI, in agreement with a previous report. HAART including efavirenz showed a similar tendency that might have not reached the significance level due to lack of statistical power."
"..There is no clear explanation for the protective effect of regimens including PI as backbone drugs for liver fibrosis in HIV/HCV coinfection... HIV infection seems to alter the pattern of cytokine response to HCV antigens. Therefore, HIV infection can induce quantitative and qualitative changes in the immune response to HCV infection. The immune reconstitution associated with HAART may partially reverse these alterations in the immune response to HCV.."
"We did not find a significant association between sex, alcohol intake and estimated duration of HCV infection and liver fibrosis. The duration of HCV infection, as estimated in this study, has not correlated with fibrosis in other studies. This is difficult to explain, but it is probably the result of inaccurate estimation of duration and the limited duration of most surveys and lack of linearity in fibrosis progression over time."
"The time of exposure to antiretroviral drugs was not associated with liver fibrosis progression, except for nevirapine. Again, this finding should be a consequence in part of the non-linear evolution of fibrosis; likewise, the different degree of liver injury in patients at starting HAART could also partly explain this observation."
Introduction
Dual infection with HIV and hepatitis C virus (HCV) is frequent, particularly in areas where injecting drug use is the main means of transmission for HIV. Mutual interactions between the two infections have been described. Thus, the course of chronic hepatitis C seems to be accelerated by HIV coinfection.
However, this more rapid progression of chronic hepatitis C was overshadowed by the high rate of mortality caused by HIV-related opportunistic diseases before the introduction of protease inhibitors (PI) in clinical practice. Highly active antiretroviral therapy (HAART) effectively prevents AIDS. Because of this, liver failure is now a leading cause of morbidity and mortality among HIV/HCV-coinfected patients.
HAART regimens do not suppress HCV replication. Instead, they are associated with transient flares of HCV replication. The latter may increase the liver damage in chronic hepatitis C. Moreover, many antiretroviral drugs commonly used in HAART combinations are hepatotoxic. For these reasons, HAART could be associated with more severe liver damage and, as a consequence, with progression of liver fibrosis. On the other hand, HAART-related immune restoration could contribute to lessen the liver damage associated with HCV infection.
There are very few and contradictory data on the effect of HAART on liver fibrosis in HIV/HCV co-infected patients. In this regard, one study showed that PI-based regimens were associated with less fibrosis progression [9]. The duration of antiretroviral therapy was associated with the stage of fibrosis in another study, but the duration of PI therapy was not [10]. Finally, HAART was not associated with any effect on fibrosis progression in two further surveys [11,12]. On the other hand, the effect of non-nucleoside analogues (NNRTI) on liver fibrosis is largely unknown. To clarify these issues, we evaluated the associations between the use of different antiretroviral drugs and the liver fibrosis of patients with HIV and HCV infections.
Patients and methods
This was a cross-sectional study of HIV/HCV co-infected patients from a cohort of HIV-infected patients prospectively followed at the Infectious Diseases Unit of the Hospital Universitario de Valme, at Seville, southern Spain. The patients' records are abstracted and entered in a database at baseline and at each scheduled visit. For the present study, data were recovered from the database and, whenever necessary, directly from the patients records.
Patients
Nine hundred and thirty-one patients have been evaluated for HIV infection at the Unit since January 1985 to March 2003. HCV infection status was determined in 776 patients. Five hundred and fifty-nine (72%) of them showed HCV infection. All HIV/HCV co-infected patients who underwent a liver biopsy were included in this study, if the date of infection was known or could be estimated and if they had not received treatment for HCV infection previously.
Since the majority of injecting drug users become infected by HCV during their first year sharing needles, the year of infection was estimated as the first year when the patient was an injecting drug user sharing the injecting equipment. Patients were excluded if there was evidence of a coexisting liver disease. Most liver biopsies were taken in the Unit with the aim of establishing the prognosis and indicating therapy for chronic hepatitis C. Usually, patients were proposed to undergo liver biopsy if they were compliant with the previous follow-up at the Unit and did not show advanced HIV infection. Active drug using or alcoholism were factors that negatively influenced the decision to biopsy. One hundred and fifty-two (24.4%) patients fulfilled the inclusion criteria. The earliest liver biopsy of a patient included in the study was performed during November 1989. The liver biopsies of 109 of 152 patients included in the study were obtained since 1996.
Antiretroviral therapy
Patients were offered antiretroviral therapy according to the availability of drugs, evidence from clinical trials and international guidelines. Thus, PI were extensively introduced in our practice since 1997 and NNRTI in 1999. HAART was defined as the combination of, at least, two nucleoside analogues plus one PI or one NNRTI. The different antiretroviral drugs prescribed to each patient during the follow-up, and the dates of initiation and discontinuation of each agent were recorded.
Histologic evaluation
All of the patients gave their written informed consent for liver biopsy. The specimens were immediately placed in buffered formalin. After a 24-h fixation, they were embedded in paraffin using routine methods. For the present study, 4-mm-thick sections were deparaffinized and rehydrated. Histological evaluations were made on sections stained with haematoxylin and eosin, and Masson's trichrome. A single pathologist, who was not aware of the clinical status of the patients, evaluated all the stained sections. Liver fibrosis was scored following the Knodell histological activity index modified by Scheuer.
Calculation of the fibrosis progression rate
The estimated fibrosis progression rate is defined as the ratio between the fibrosis stage and the estimated duration of HCV infection in years. In this model it is assumed that the patient has no liver fibrosis the day of infection and that the fibrosis progression rate is constant. The fibrosis stage F4 (cirrhosis) was excluded from the calculation, as cirrhosis may have been present for a long period of time before the liver biopsy. For example, for a patient with fibrosis stage 2 (F2) and a 10-year duration of HCV infection, the fibrosis progression rate is 0.2.
Variables that could be associated, due to previous evidence or because of plausibility, with liver fibrosis were considered in the statistical analysis: age at infection, age at liver biopsy, duration of HCV infection, sex, alcohol intake, route of HCV transmission, CD4 cell count at time of liver biopsy, nadir CD4 cell counts, clinical AIDS, alanine aminotransferase (ALT) levels at liver biopsy, HCV genotype and serum RNA levels, use of individual antiretroviral drugs at any time before liver biopsy and the duration of exposure to each of them.
Liver fibrosis was categorized into advanced, which included stage F3 (numerous septa with architectural distortion without cirrhosis) and stage F4 (definitive or probable cirrhosis), and non-advanced, which included stages F0 to F2. We chose the cut-off level of 0.2 to categorize the fibrosis progression rates. This level is clinically meaningful as patients with rates > 0.2 could progress at least 1 fibrosis stage after 5 years of infection.
Statistical methods
Continuous variables were expressed as median (interquartile range) and the categorical variables as numbers (percentage). Continuous variables were compared using Student's t test or the Mann-Whitney U test when appropriate. Categorical variables were compared using the Chi-square test with the Yate's correction or the Fishers' test when appropriate. The variables that showed a P < 0.2 in the univariate analysis were entered in the multivariate analysis.
The variables associated with the stage of fibrosis and with faster fibrosis progression were assessed by stepwise logistic regression analysis. The adjusted odds ratio (AOR) for advanced fibrosis and for fibrosis progression rate > 0.2 and the respective 95% confidence intervals (CI) were calculated. The goodness-of-fit of the final models was assessed by the Hosmer-Lemeshow test.
The statistical analysis was carried out using the SPSS 11 statistical software package (SPSS, Chicago, Illinois, USA).
Results
Characteristics of the study population
The characteristics of 152 HIV/HCV-coinfected patients according to the antiretroviral therapy regimen that they had received before liver biopsy are shown. Sixty-two (40.8%) patients had never received HAART; 47 (75.8%) of them were antiretroviral naive; 90 (59.2%) patients had been prescribed HAART before liver biopsy; 78 (86.7%) of them were prescribed PI at any time before liver biopsy.
Charactertistics of patients according to ART received before liver biopsy: no differences except—median duration of HCV infection -- 15 yrs for HAART group vs 9 yrs for non-HAART group (<0.001); median CD4 count nadir- 231 (HAART group) vs 390 (non-HAART group). Median age at liver biopsy- 38 yrs (HAART) vs 30 yrs (non-HAART), (<0.001).
Factors associated with advanced fibrosis (fibrosis stages 3 and 4)
Among the variables considered in the study age at HCV infection was associated with advanced liver fibrosis. Advanced liver fibrosis was observed in 21 (29%) and 38 (48%) patients with ALT levels below and above 100 IU/ml, respectively (P = 0.02).
The estimated duration of HCV infection of the group with F0-F2 stages was median 13 (range, 9-17) years and of the group with F3 and F4 stages median 13 (range, 9-16) years (P = 0.5). Twenty-two (37.3%) patients with advanced liver fibrosis and 41 (44%) patients with non-advanced liver fibrosis showed a CD4 cell count nadir of <= 250 x 106 cells/l (P = 0.4). The proportion of women with advanced fibrosis tended to be higher than in men, but the differences did not reach statistical significance. Sixteen (76%) of 21 women and 77 (59%) of 131 men became infected by HCV after the age of 20 years (P = 0.1). Six (29%) women and 37 (28%) men reported daily alcohol intake > 50 g/day (P = 0.9).
CD4 cell counts <= 250 x 106 cells/l at liver biopsy were non-significantly more frequent in the advanced fibrosis group. The increases from baseline CD4 cell counts among patients who started HAART were similar in the non-advanced fibrosis group and in the advanced fibrosis group, 197 (4 x 106 to 371 x 106) cells/l and 141 (90 x 106 to 385 x 106) cells/l, respectively (P = 0.6).
A significantly greater proportion of patients who had received PI-based therapy showed liver fibrosis below F3 stage (49% vs 31%, multivariate: 0.008, AOR 0.39). Thirty (71.4%), 33 (67.3%) and 26 (65%) patients ever on saquinavir, indinavir and nelfinavir, respectively, showed liver fibrosis below F3 stage. Similarly, 15 (75%) patients at any time on efavirenz-based regimens showed non-advanced liver fibrosis. However, patients who had ever received regimens containing nevirapine demonstrated advanced liver fibrosis more frequently (56% vs 36%, multivariate- 0.04, AOR 2.56). The time of exposure to the different antiretroviral drugs was not associated with the stage of liver fibrosis. Eleven (64.7%) of 17 patients who had used nevirapine for more than 1 year showed advanced liver fibrosis (AOR, 3.5; 95% CI, 1.2-10.6; P = 0.03).
The variables independently associated with advanced fibrosis were age at infection, PI-based HAART and nevirapine-based HAART.
Factors associated with fibrosis progression rate
The median fibrosis progression rate of the study population was 0.143 (range, 0.059-0.250). The duration of exposure to different antiretroviral drugs was not associated with the rate of fibrosis progression. Eight (47%) of 17 subjects who used nevirapine for more than 1 year showed a fibrosis progression rate > 0.2 (AOR, 3.4; 95% CI, 1.8-6.5; P < 0.001). Age at infection, CD4 cell counts at liver biopsy, exposure to PI at any time of the follow-up and having ever used nevirapine were independently associated with faster fibrosis progression.
Changes in ALT levels and antiretroviral drugs
The median (interquartile range) levels of ALT before starting therapy among patients who used PI were 72 (42-126) IU/ml and among those who used nevirapine were 59 (45-104) IU/ml. The levels of ALT increased 2.5-fold from the baseline levels in 21 (27%) of 78 patients during their follow-up on PI and in 19 (76%) of 25 patients during their follow-up on nevirapine (P < 0.001). The median levels of ALT at the time of these increases were 151 (92-234) and 174 (113-369) IU/ml among the patients on PI and nevirapine, respectively.
Discussion
In this study, we found that the use of HAART regimens including nevirapine is associated with an increased degree of liver fibrosis in HIV-infected patients with chronic hepatitis C. On the contrary, patients on PI as the backbone of potent antiretroviral therapy are more likely to show less liver fibrosis.The outcome variable in this study was the development of liver fibrosis, which we have defined by following a double approach: as the stage of liver fibrosis and as the liver fibrosis progression rate. The stage of liver fibrosis is not affected by some artefacts that could influence the calculation of the fibrosis progression rate, such as errors in the estimation of the duration of HCV infection nor non-constant progression of fibrosis. On the other hand, the liver fibrosis progression rate would give a dynamic view of the development of fibrosis. Both approaches yielded similar results. This reinforces the conclusions of the study.
The present study has a few limitations. On the one hand, the development of fibrosis does not follow a continuous linear progression. Thus, the stage of liver fibrosis may have been reached a long time before, for example, due to a severe acute liver damage followed by a prolonged period of inactivity. This fact limits the worth of the estimated liver fibrosis progression rate. The only alternative design to achieve a more reliable assessment of the rate of fibrosis progression would be a prospective follow-up with sequential liver biopsy withholding anti-HCV therapy. This study would be very difficult to carry out, mainly due to lack of patient acceptance and ethical concerns. Because of this, most surveys where factors associated with liver fibrosis progression have been investigated were cross-sectional studies in patients with estimated or known date of infection. On the other hand, the patients included in the present study may not be representative of the general population of HIV/HCV-coinfected patients. The study patients were selected as candidates for liver biopsy because they were adherent and had a better control of HIV infection. On the other side, most patients with clinically overt liver failure are not biopsied at our unit. Because of this, the patients with most advanced liver damage have not been included in this study.
We did not find a significant association between sex, alcohol intake and estimated duration of HCV infection and liver fibrosis. The duration of HCV infection, as estimated in this study, has not correlated with fibrosis in other studies. This is difficult to explain, but it is probably the result of inaccurate estimation of duration and the limited duration of most surveys and lack of linearity in fibrosis progression over time.
Male sex has been associated with liver fibrosis in some studies. However, other reports failed to find such association in HCV-monoinfected and HIV/HCV-coinfected patients. In this study, women and men showed similar frequency of advanced liver fibrosis and a close fibrosis progression rate. This finding is most probably due to several confounders. Thus, women tended to become infected with HCV at an older age and drank as much alcohol as men. Women are more susceptible to alcohol-mediated liver damage. In fact, the risk of cirrhosis is twice as high in women as in men with the same amount of alcohol intake. Perhaps these characteristics of the women included in this study and the impact of antiretroviral drugs may explain the lack of differences.
Alcohol was surprisingly not related to fibrosis. However, this finding is in agreement with surveys that assessed liver fibrosis in HIV/HCV-coinfected patients and HCV-monoinfected injecting drug users. In one study of follow-up biopsies on HCV-monoinfected patients, a detailed assessment of alcohol exposition was done by several approaches. However, alcohol consumption was associated neither with observed nor estimated liver fibrosis progression in that study. Moreover, in HIV infected patients the effect of alcohol could be overshadowed by the impact of antiretroviral therapy.
We report, for the first time to our knowledge, more advanced stage of liver fibrosis and faster liver fibrosis progression among coinfected patients exposed to nevirapine. A previous report on this issue included too few patients on NNRTI to draw any conclusion. Nevirapine is a well-characterized hepatotoxic agent. Nevirapine-associated severe hepatotoxicity is more common in patients with hepatitis C infection. There are two possible presentations of toxicity related to nevirapine: an early idiosyncratic reaction and a late onset cumulative toxicity both of which might be implicated in the worsening liver fibrosis among HIV-infected patients with chronic hepatitis C. We found that patients with exposure to nevirapine over 1 year were more likely to show advanced liver fibrosis and faster liver fibrosis progression. These data support that the cumulative nevirapine hepatotoxicity is involved in the development of liver fibrosis in HIV/HCV-coinfected patients. On the other hand, efavirenz was not associated with liver fibrosis progression. Thus, the effect of nevirapine is not related to the NNRTI drug class, but rather to individual characteristics of this drug.
We found that PI-based antiretroviral regimens were associated with non-advanced liver fibrosis. We also found a lower rate of fibrosis progression among patients exposed to PI, in agreement with a previous report. HAART including efavirenz showed a similar tendency that might have not reached the significance level due to lack of statistical power. There is no clear explanation for the protective effect of regimens including PI as backbone drugs for liver fibrosis in HIV/HCV coinfection. Indeed, the altered natural history of chronic hepatitis C in HIV infection is incompletely understood. HIV infection is associated with loss of CD4 lymphocytes. In this regard, low CD4 cell counts have been linked with the progression of chronic hepatitis C in HIV infection. Additionally, HIV infection seems to alter the pattern of cytokine response to HCV antigens. Therefore, HIV infection can induce quantitative and qualitative changes in the immune response to HCV infection. The immune reconstitution associated with HAART may partially reverse these alterations in the immune response to HCV. In the present study an absolute level of CD4 cells > 250 ¥ 106 cells/ml at liver biopsy was associated with slower liver fibrosis progression, which supports that hypothesis.
The time of exposure to antiretroviral drugs was not associated with liver fibrosis progression, except for nevirapine. Again, this finding should be a consequence in part of the non-linear evolution of fibrosis; likewise, the different degree of liver injury in patients at starting HAART could also partly explain this observation.
Chronic hepatitis C has emerged as a serious concern in the HAART era. Coinfected patients can show more unwanted drug reactions. This may complicate the choice of a safe HAART regimen in coinfected patients. The results of this study suggest that HAART including PI could be more advantageous in terms liver fibrosis progression than nevirapine-based regimens in this setting. Whether the associations found in this study have a clinical impact need to be confirmed by randomized prospective studies with hard clinical end-points, such as development of decompensated cirrhosis or death attributable to liver failure.
SOME REFERENCES
9. Benhamou Y, Martino V, Bochet M, Colombet G, Thibault V, Liou A, et al. Factors affecting liver fibrosis in human immunodefiency virus- and hepatitis C virus-coinfected patients: impact of protease inhibitor therapy. Hepatology 2001, 34:283-287.
10. Tural C, Fuster D, Tor J, Ojanguren I, Sirera G, Ballesteros A, et al. Time on antiretroviral therapy is a protective factor for liver fibrosis in HIV and hepatitis C virus (HCV) co-infected patients. J Viral Hep 2003, 10:118-125.
11. Martinez-Sierra C, Arizcorreta A, Diaz F, Roldán R, Martin-Herrera L, Pérez-Guzmán E, et al. Progression of chronic hepatitis C to liver fibrosis and cirrhosis in patients cinfected with hepatitis C virus and human immunodeficiency virus. Clin Infect Dis 2003, 36:491-498.
12. Zylberberg H, Barennes C, Carnot F, Chaix ML, Viard JP, Savès M, et al. Progression of liver histological status in HIV hepatitis C virus co-infected patients started on HAART: a prospective study. Tenth Conference on Retroviruses and Opportunistic Infections. Boston, February 2003 [abstract 831].
|
|
| |
| |
|
|
|
