8th Annual Retrovirus Conference
HIV and Hepatitis Virus
The Center of Two Epidemics
By Nancy Shulman, MD, Stanford University School of Medicine. Edited by Harvey S. Bartnof, MD, Staff Physician at the AIDS Virus Education and Research Institute (AVERI) in San Francisco
The topic of HIV co-infection with either hepatitis B or C virus is gaining increased attention by both researchers and clinicians alike. There were a substantial number of presentations addressing this issue at the 8th Annual Retrovirus Conference. During the "Opportunistic Infection/Co-Pathogen" session where the oral presentations on co-infection were delivered, there was standing room only. The following is highlights of the oral and selected poster presentations about hepatitis virus co-infection with HIV.
HIV-HCV Co-infection State-of-the-Art
Ken Sherman, MD of the University of Cincinnati in Ohio delivered a summary HIV/HCV co-infection lecture. The lecture including audio and viewing slides is available through the conference website at www.retroconference.org. There was a theme throughout his talk that there are lots of unknowns in co-infection and that new research results will become available in the next several years. The epidemiology (population studies) of HCV (hepatitis C virus) in HIV-infected patients varies somewhat by geographic location. Based on seroprevalence (positive blood test rates) studies in different risk groups an overall estimate of 20-30% of HIV positive patients are co-infected with HCV in developed countries. For individuals HIV-infected by IVDU, various studies show rates of HCV infection to be 50%-90%. Rates are very high in hemophiliacs (males with genetic blood clotting protein deficiency) and in people who have acquired HIV through injection drug use. Rates in men who have sex with men are much lower (5-10%) but are higher than the general population. Dr. Sherman recommends that all HIV positive patients be tested for HCV.
HIV increases the rate of liver fibrosis (scarring) and progression to end-stage liver disease in patients with HCV infection. The impact of HCV on the natural history of HIV disease is more controversial, with some studies showing an enhanced progression rate and others showing no effect, particularly when there is a response to HAART (highly active antiretroviral therapy). People with HIV/HCV co-infection are more susceptible to hepatotoxicity (liver toxicity) associated with HAART. However, one study at this Conference suggests that the NNRTI drugs nevirapine (Viramune) or efavirenz (Sustiva) are not associated with a higher rate of severe liver toxicity among patients with hepatitis virus co-infection and HIV. Several studies have shown an increasing proportion of morbidity (disease) and mortality (death) from liver disease in cohorts (groups of patients) where the HCV or HBV co-infection prevalence (rate) with HIV is substantial.
Information is sparse on treatment of HCV in HIV/HCV co-infected populations. Several uncontrolled studies with small numbers of subjects using interferon alfa (Intron A, Roferon) or combination therapy with interferon alfa plus ribavirin (Rebetol, combined as Rebetron) have mostly reported substantially lower sustained HCV clearance rates than is seen in studies of HCV mono-infected patients without HIV yet with similar treatment. Part of the reason for the low rates are that these studies in HIV positive patients have reported higher dropout rates due to side effects than studies of HIV negative subjects. HIV positive patients also have more diverse HCV "quasispecies" (variants) than HIV negative patients, which may be a factor in non-response to treatment. Subjects with higher CD4 counts have tended to respond better to treatment than subjects with lower CD4 counts with a cutoff somewhere between 200-400 cells per microliter. Based on this finding, Dr. Sherman recommends that patients be treated with HAART before HCV treatment to enhance the immune status. We know from several studies HAART does not lower HCV viral loads and is often associated with an increase in inflammation of the liver and an increase of HCV viral loads. A small number of HIV/HCV co-infected patients who had stable HIV disease on HAART have shown benefits when interferon alfa and ribavirin were added to treat their HCV disease. The rate of anemia from ribavirin may be higher in co-infected patients, particularly if their HAART regimen includes AZT (zidovudine, Retrovir).
(Editorial comments from Jules Levin: several small studies of interferon+ribivarin in coinfected persons showed response to therapy was comparable to that seen in HCV infected patients. But I think it remains uncertain whether coinfected persons will respond as well to HCV therapy as HCV-infected patients, and there is concern that coinfected persons will not in general respond as well to HCV therapy as people with HCV-infection. There are several ongoing studies in coinfection with pegylated interferon+ribivarin, which should yield data by next year. Response to therapy is individual, and so some coinfected individuals will likely respond as well as HCV-infected, while others will not. In certain circumstances it may be preferable to consider treating HCV before treating HIV. For the coinfected person, biopsy ought to be considered immediately upon diagnosis with HCV, because one can have normal ALT but still have moderate to severe liver disease, and because HCV disease can progress more quickly in coinfected persons than in individuals infected with HCV alone.)
The immune response to HCV in co-infected patients is qualitatively and quantitatively different than in HIV negative patients with HCV infection. "T-cell stimulation assays" in the laboratory reveal that co-infected patients respond less overall and preferentially less to the "non-structural" proteins of HCV, which may be important for clearing virus. Co-infected patients produce less IL-10 (interleukin 10) in response to HCV antigen (proteins) than HCV mono-infected patients. IL-10 has "anti-fibrotic" (blocking scarring in liver) activity and this may be part of the reason for the enhanced fibrogenesis (causing scarring) seen in patients with HIV/HCV co-infection.
The next standard of care for treatment for chronic hepatitis C and HCV in HIV/HCV co-infected patients will be "pegylated" interferon alfa (Peg-Intron, Pegasys; long-acting, once weekly dosing) plus ribavirin. However, Dr. Sherman said that better, less toxic drugs are needed. He gave a plea for more research about HIV/HCV co-infection, including more research into liver transplantation in HIV patients.
Organ Transplantation in HIV Positive
Two posters presented information on 9 liver transplants and 5 kidney transplants from UCSF (University of California at San Francisco) and Kings College in London, UK. The UK group performed 7 liver transplants, 4 for HCV-related liver complications and 3 for liver disease due to HBV (hepatitis B virus). The lead author was Dr. A. E. Boyd. The UCSF group transplanted 2 livers in one patient (the 2nd liver was transplanted after the first one failed) and 5 kidneys (one on the liver patient). The lead author was Michele Roland, MD. Among the group of 5 HCV liver transplant patients, all had HCV disease recur rapidly in the new liver and they died between 3 and 25 months after transplant surgery. In contrast, all 3 who were transplanted for HBV disease are alive 6 to 36 months after transplant. One person attempted suicide, but other than that all patients are well and remain on anti-HIV medication with stable CD4 counts and undetectable HIV viral loads. These are reports from only two of the sites that are doing transplantation in HIV positive patients as a part of experimental research. The take home message from these reports with only few patients is that liver and kidney transplantation is possible in HIV patients, but we need better anti-HCV therapies to contain their virus after transplant. Liver transplantation among patients with HIV/HBV co-infection may be associated with a higher long-term success rate; however, those patients were treated for HBV after transplant.
(Comments from Jules Levin: It's my understanding that liver transplants in HCV/HIV coinfected at other sites are yielding better results but follow-up is still very short term (1-2 years). More transplants need to be performed so we can better understand the challenges & potential for success.)
NNRTIs & Hepatotoxicity in
Hepatitis & HIV Coinfection
NNRTI (non-nucleoside reverse transcriptase inhibitor) drugs to treat HIV are potent, and the two most commonly used ones have conveniently dosing. One of the drugs, nevirapine (NVP, Viramune), has been associated with a 10-20% rate of any hepatotoxicity (liver toxicity).
In Study FTC-302, John Bartlett, MD reported
on the rates of liver toxicity. In this study, 908 patients were randomized
to receive d4T (stavudine, Zerit) with FTC (emtricitabine, experimental NRTI
drug) or 3TC (lamivudine, Epivir), plus either NVP or efavirenz (EFV, Sustiva,
NNRTI drug). The toxicity results showed that 17% of subjects in the NVP arm
had severe or life threatening (grade 3-4; 9% grade 4) increases in liver enzymes,
including 2 deaths. One of the women who died had chronic hepatitis B, while
the other did not; neither had acute hepatitis A or C. Whereas, no patient in
the EFV arm had grade 3-4 liver toxicity. The only factor significantly associated
with severe toxicity was being female.
(Editorial comment: in a previous retrospective chart review study in the USA from Judith Aberg & colleagues higher incidence of rash was seen, but not the higher incidence of hepatotoxicity in women. In trying to explain the South African results it's been suggested that possibly elevated ALTs were not noticed & patients continued on therapy anyway. It's also been suggested that different diet resulting in interaction, lifestyle, or genetics could be a factor. Thirdly, if rash had occurred it's possible it may not have been noticeable.)
In another presentation on this issue, Mark Sulkowski, MD reviewed the large Hopkin's HIV patient database where 45% of patients are co-infected with HCV. Between 1998 and mid-2000, 203 patients were prescribed NVP-based HAART, while 96 were prescribed an EFV-based HAART. The results showed that 17% of patients taking NVP developed grade 3 or 4 liver toxicity, compared to 8% those taking EFV. However, when analyzing the rate by total patient-months of exposure, the rates for the two regimens were essentially the same: 1.3 cases per 100 patient-months for NVP versus 1.2 cases per 100 patient-months for EFV. Previously, studies have shown the NVP liver toxicity may be more likely to occur within the first few months after starting the drug. (Editorial comment: I don't think the cases per 100 patient months method captures the tendency for NVP hepatotoxicity to occur within the first 4 or 12 weeks. A second limitation of this study is it's not prospective & randomized.) When analyzed by the presence or absence of chronic hepatitis B or C, patients with chronic viral hepatitis were at no greater risk for severe liver toxicity than those without hepatitis co-infection. The only independent predictor of hepatotoxicity in a "multivariate" statistical model was a CD4 cell increase greater than 50 cells per microliter. This also might have been a marker for higher adherence. Liver biopsy sampling was not performed. Alcohol consumption as a co-factor was not reported. Dr. Sulkowski concluded that NNRTI drug combination HAART should not be withheld in HIV patients who are co-infected with chronic hepatitis B or C.
Low Response to Interferon Alfa plus
Ribavirin In HIV/HCV Co-infected Patients
Dr. M. Bochet of Hospital Pitie-Salpetriere in Paris, France presented interim results of 56 HIV/HCV (25% women) co-infected patients treated with combination therapy for their HCV disease. She used interferon alfa (3 million units injected 3 times weekly) plus ribavirin (800-1200 mg daily) for 6-12 months. All patients had been taking stable HAART for their HIV disease for at least 3 months, with reasonably suppressed HIV viral loads. A previous course of monotherapy with interferon alfa was not successful for 46% of the enrollees; the remaining 54% were treatment-naïve (no previous treatment) for their HCV disease.
The mean baseline CD4 count was 377 cells per microliter, HIV RNA was 3.3 log (2,000) copies per milliliter, HCV RNA was 6.9 log (7.9 million) copies (20 "milliequivalents") per milliliter, and 63% had HCV genotype 1 (the most difficult to treat). The mean baseline liver fibrosis score was 2.8 ("Metavir" scale, worst is 4), and 23% had cirrhosis (fibrosis score 4).
Early withdrawal occurred among 29%, due to known side effects of the two anti-HCV drugs, mainly anemia (low red cells), asthenia (weakness), "flu"-like symptoms and/or psychiatric symptoms. One patient died after 3 months of treatment due to "liver decompensation" associated with cirrhosis (scarring). Of the remaining 71% of patients, treatment took place for 6-12 months. Completing 12 months of treatment was 47% of the entire group.
The results showed that the sustained HCV viral clearance rate 6 months after treatment for HCV was stopped was a mere 19.6% in a strict "intent-to-treat" analysis (all enrolled patients considered). The response rate was nearly identical when comparing pre-treated with treatment-naïve patients. This is compared to about 40% in HIV negative, HCV positive patients receiving the same type of treatment for chronic hepatitis C. Part of the poor response was due to a high 29% discontinuation rate. Most patients discontinued for intolerance of flu-like symptoms or psychiatric symptoms, mainly depression. Of the patients who remained on therapy ("on-treatment" analysis), approximately 40% achieved a sustained virologic response (SVR). SVR was significantly associated with HCV genotype non-1, a baseline HCV RNA less than 3 milliequivalents per milliliter, and a normalized ALT (liver enzyme, alanine aminotransferase). CD4 cell counts and HIV RNA levels did not change significantly during the study. Paired liver biopsies performed pre-HCV treatment and 6 months after treatment on 13 patients revealed a trend towards improvement to no change. Dr. Bochet did not have the breakdown of response by genotype, but stated it was not significant in the multivariate model.
It is likely that a higher rate of sustained virologic response in HIV/HCV co-infected patients will occur when the experimental pegylated version of interferon alfa is combined with ribavirin, on top of stable HAART. Many physicians are using weekly erythropoietin (Procrit) to prevent or treat the anemia that appears to be more common among this patient group. Also, many physicians are also using a lower dose of ribavirin 800-1,000 mg to help avoid anemia, particularly for patients who are not overweight based on BMI (body mass index, weight divided by height squared). Recent evidence suggests a weight-based dosing for ribavirin among HCV-mono-infected patients. Whether this will be the case for HIV/HCV co-infected patients remains to be seen.
HIV/HCV Co-Infection in Hemophiliacs
Eric Daar, MD of the University of California at Los Angeles presented follow-up data about HIV/HCV co-infected hemophiliacs (males with genetic deficiency of normal blood clotting protein) that was originally presented at last year's 7th Annual Retrovirus Conference and recently published in the February 15th Journal of Infectious Diseases. Patients were enrolled in the Hemophilia Growth and Development Study. This was a multicenter study that began between 1989-1990 and followed 207 hemophiliac children ages 7-19 who were co-infected with HIV and HCV and 121 infected with only HCV. They were evaluated annually with T-cell counts and viral load measurements of HCV and HIV.
In considering finding's from this study in hemophiliacs, one should bear in mind that we don't know if HIV and hepatitis behaves differently in hemophiliacs. The mode of transmission is different, the viruses may be different in their virulence or in some other way, and hemophiliacs were generally younger when infected so they have more thymus.
During the 7-year follow-up period that was predominantly before the era of HAART, 31% of the co-infected patients progressed to AIDS and 35% to AIDS-related death. Baseline HCV RNA levels were associated significantly with progression to clinical AIDS and AIDS-related death, even after controlling for HIV-1 RNA levels and CD4 cell counts. For every log or 10-fold increase in HCV RNA level, there was a 1.7 increase in the relative risk ("RR") for progression to AIDS, and a 1.5 increase in the RR for AIDS-related death.
In an additional analysis, the patients were divided into two groups based on their average HCV RNA levels over the study. The patients who had the higher HCV RNA levels (above the median) had a higher risk of progressing to AIDS or progressing to an AIDS-related death than those with lower HCV RNA levels (below the median). In a "stratified" statistical analysis looking at HIV viral loads (above or below the median) and HCV viral loads (above or below the median), when controlling for CD4 cell count, those with both HIV and HCV viral loads in the upper halves had the greatest risk of developing AIDS or death. Having a high HCV viral load with a low HIV viral load was associated with a somewhat slower progression to AIDS than having a low HCV with a high HIV viral load. The slowest rates of progression to AIDS or death were in the group with low HIV and low HCV viral loads. This is one of the first studies to find that HCV viral loads had a significant impact on anything other than response to HCV treatment. Unlike high HIV viral loads in HIV disease, there has not been any good evidence to date that HCV viral loads even impact the natural history of HCV liver disease. (HCV viral loads do not correlate well with liver pathology [disease] as seen under the microscope.) The author theorizes that control of HCV viremia may be a marker for better cell-mediated immunity that helps protect from HIV progression and AIDS opportunistic conditions in HIV/HCV co-infected patients.
Dr. Daar also looked at the relationship between HCV viral load, HIV viral load, CD4 cell counts and blood serum (no cells) ALT (liver enzyme) in the co-infected group. Baseline HIV viral loads were directly related to HCV viral loads, supporting the hypothesis if one is controlled, the other is also likely to be controlled due to a good cellular immune response. As has been shown in other studies, HCV viral loads were inversely related to CD4 counts, meaning a high HCV viral load is associated with a low CD4 count and that a low HCV viral load is associated with a high CD4 count. Also, blood serum ALT levels did correlate (although weekly) with HCV viral loads. Co-infected subjects had higher HCV RNA levels than the HCV mono-infected, as has been observed in other studies.
Who Clears HCV Viremia?
In Dr. Daar's hemophilia cohort, an analysis was performed looking at who had cleared HCV infection. This was accomplished by measuring who was HCV antibody positive but repeatedly HCV RNA negative by 2 viral load tests (RT-PCR [reverse transcriptase-polymerase chain reaction] and bDNA [branched chain DNA]). Interestingly, only 3% of the HIV/HCV co-infected cleared HCV RNA, compared with 14% of those HCV mono-infected, without HIV. Co-infected patients had nearly 5-times the odds (significant "odds ratio") of having chronically detectable HCV RNA than those without HIV infection. Two other factors also significantly associated with chronically detectable HCV RNA were older age ("odds ratio" of 1.3) and chronic hepatitis B infection ("odds ratio" of 3.2). Dr. Daar also found that the blood levels of HCV RNA were significantly higher at baseline among HIV/HCV co-infected patients than among HCV mono-infected patients by approximately 0.5 log (3-fold) copies per milliliter. This was maintained throughout the 7 years of observation.
Somewhat different results were reported in a much smaller sub-study of HIV positive patients followed at the University of Texas Southwestern, as reported by Dr. M. Jain. In that case-control study, 15% (n=17) of the HCV antibody positive patients with HIV had cleared HCV RNA (85% chronic infection), as determined by a negative HCV PCR test. These patients were matched to 34 randomly selected HIV patients who were HCV positive for antibodies and RNA (chronic HCV infection). The 2 groups were similar with regard to race, gender (sex), CD4 count, HIV viral load, risk factor for HIV, and chronic hepatitis B. Duration of HCV infection or sequence of acquisition of HIV and HCV infection was unknown. In this study, no significant predictors of HCV RNA clearance were found by "multivariate" analysis. One possibility is that the subjects who cleared HCV were exposed to HCV well before HIV. One main difference in this study compared to the Daar study above is that all patients were HIV positive and none had only HCV infection.
(Editorial comments: Previously at AASLD Sulkowski presented data suggesting that HIV impairs HCV clearance in coinfection.)
Adefovir for HIV/HBV Co-infected Patients
With HBV-Resistant to 3TC
HBV (hepatitis B virus) co-infection occurs in approximately 10% of HIV-infected patients. 3TC (lamivudine, Epivir-HBV) is the drug of choice for most patients with chronic HBV infection, but after 4 years, resistance has developed in 90% of HIV/HBV co-infected patients and in 67% of HBV mono-infected patients. (Interferon alfa also is FDA-approved to treat chronic hepatitis B.) Adefovir dipivoxil is a nucleotide reverse transcriptase inhibitor (NtRTI) drug that was initially developed to treat HIV. However, even at a much lower dose than that used for HIV, it has good activity against HBV, including 3TC-resistant strains. Adefovir was denied FDA approval for HIV treatment due to the high rates of kidney toxicity at the doses given (60-120 mg per day). Much lower doses are used to treat hepatitis B effectively (10 mg daily).
Yves Benhamou, MD of Groupe Hospital Pitie-Salpetriere in Paris presented the interim results of an open-label, pilot study of adefovir treatment in 35 HIV/HBV co-infected patients (3% women) who had 3TC-resistant HBV for a mean of 24 months. The study will last 12 months; the interim results are after 28-36 weeks. 3TC had been taken as a part of HAART for a mean of 45 months. Drug resistance in the HBV "DNA polymerase" gene included mutations "M550V" or "M550I," almost always with "L526M". During the 6 months on HAART prior to adding adefovir, patients were required to have an undetectable HIV viral load (limit 400 copies per milliliter) and detectable HBV DNA in blood serum.
The mean baseline HBV viral load was 8.64 log (436 million) copies per milliliter (approximately 1,800 picograms per milliliter). Hepatitis "e" antigen (chronic infection) was present in 94% at baseline, with the remaining 6% being "e" antibody positive (greater control of infection). The mean baseline HIV RNA was 2.88 log (759) copies per milliliter with a mean CD4 count of 423 cells per microliter. The baseline ALT (liver enzyme alanine aminotransferase) was approximately 100 IU (international units) per milliliter. Baseline liver histology within 18 months before starting adefovir for 66% of patients who had one showed a mean fibrosis score of 2 (worst is 4), with 23% having cirrhosis (F4 fibrosis score, scarring). All patients added adefovir 10 mg daily to their existing HAART regimen that included 3TC.
The interim results after 36 weeks revealed a mean HBV DNA decrease at week 36 of -3.8 log (6,309-fold reduction) copies per milliliter to 4.84 log (69,183) copies per milliliter. Each time point measurement revealed an additional mean decrease in HBV DNA. Two patients (6%) achieved an undetectable HBV viral load less than 3 log (1,000) copies per milliliter. Three patients (9%) lost "e" antigen, while two (6%) had "e" antigen conversion (developed antibodies to "e" antigen). Surprisingly, the decrease in HBV viral load was not associated with a decrease in ALT that remained essentially unchanged from baseline. Dr. Benhamou had no explanation for this.
When 11 patients were tested for adefovir genotypic resistance at weeks 12 and 24, none had developed RT (reverse transcriptase) mutations at "codons" 65 or 70 that have occurred previously with this drug. The mean CD4 count increased insignificantly by approximately 15 cells per microliter, while the HIV viral load remained unchanged. Adefovir was well tolerated. No kidney toxicity occurred in terms of significant serum creatinine increase or phosphate decrease. No other biochemical, organ or cellular toxicities were noted, with the exception of one patient who developed insulin-dependent diabetes at week 7. (This patient had a family history of diabetes and adefovir was not believed to be related.) One patient developed mild (grade 1) insomnia (difficulty sleeping) after 7 weeks. Two patients discontinued the study for reasons unrelated to adefovir toxicity.
Dr. Benhamou concluded that in HIV/HBV co-infected patients, adding adefovir 10 mg once daily for 36 weeks to HAART led to a significant decrease in HBV DNA in spite of baseline 3TC resistance, without any detriment to the CD4 count or HIV viral load. The drug was tolerated well. This study is ongoing for a full year (and possibly longer) to assess the "durability of HBV DNA suppression," additional potential conversion of "e" antigen, longer-term tolerance and to monitor for the "potential emergence of adefovir resistance." It would be helpful if repeat liver biopsies could be performed after 12 months on those patients with baseline biopsies.
Rising Rates of
Liver-Related Deaths from Hepatitis Virus Co-infection in the HAART Era
In the era of HAART (highly active antiretroviral therapy) to treat HIV disease, death from AIDS has been decreasing, and some studies have shown rising rates of non-AIDS deaths including liver disease related to HBV (hepatitis B virus) or HCV (hepatitis C virus) co-infection. In one study by Dr. M. Puoti of the University of Brescia in Italy, approximately two-thirds of HIV positive patients have co-infection with HCV or HBV. Up to September 2000, Dr. Puoti reported that 40% of the deaths among HIV patients in the era of HAART have been related to liver disease. (This was among 2,028 "person-years" of follow-up.) This was statistically associated with HBV infection and HAART-related toxicity. Interestingly, HCV and HDV (hepatitis D or delta infection, a complication or "super-infection of chronic hepatitis B) were not statistically associated after controlling for various factors.
In another study by Dr. L. Martin-Carbonero of the Institute de Salud Carlos III in Madrid, Spain, 46% of their 843 HIV positive patients have chronic viral hepatitis, including HCV in 78% of that 46%. During the last 4 years, an increasing proportion of in-hospital deaths have been due to liver disease. In the first 9 months of 2000, 43% of their hospital deaths in HIV patients have been due to liver disease. HCV was the specific cause in 2/3 of those 43%.
Liver disease due to chronic hepatitis C has become the leading cause of death among HIV positive patients during 1998-1999 in the two major referral centers in Northern Italy (University of Torino and University of Verona). The presenting author was Dr. G. Perri.
Researchers at the US CDC (Centers for Disease Control and Prevention) reported last year that liver disease also has become an increasing cause of death among HIV patients in the US, although not to the same degree as the European reports above.
--Bartlett J and others. Severe liver toxicity in patients receiving two nucleoside analogues and a non-nucleoside reverse transcriptase inhibitor. Abstract and oral presentation 19.
--Benhamou Y and others. An open-label pilot study of the safety and efficacy of adefovir dipivoxil in HIV/HBV co-infected patients with lamivudine resistance HBV. Abstract and oral presentation 36.
--Bochet M and others. Efficacy and tolerance of IFN alpha plus ribavirin for chronic hepatitis C in HIV-infected patients. Abstract and poster presentation 574.
--Boyd AE and others. Liver transplantation and HIV-a case series of 7 patients. Abstract 578.
--Darr ES and others. Hepatitis C virus clearance, viral load and alanine aminotransferase (ALT) levels in HIV-infected and uninfected hemophiliacs. Abstract and oral presentation 35.
--Di Perri G and others. Liver failure from HCV as the current leading cause of death in HIV-infected patients in Northern Italy. Abstract 573.
--Jain M and others. HIV does not diminish eradication of hepatitis C viremia: analysis of viral hepatitis co-infection in a Dallas HIV clinic. Abstract 568.
--Martin J and others. Impact of hepatitis C in HIV-infected individuals in an urban center in Madrid, Spain. Abstract 572.
--Martin-Carbonero L and others. Impact of chronic viral hepatitis on hospital admission and mortality in HIV-infected patients. Abstract 297.
--Puoti M and others. Hepatitis viruses co-infections, antiretrovirals hepatotoxicity and risk of death in HIV-infected patients: prospective cohort study. Abstract 576.
--Roland M and others. Solid organ transplantation in HIV disease. Abstract 579.
--Sherman K. State-of-the-Art Lecture on HCV Co-Infection [with HIV] and Summary at Session 9, "Co-Pathogens and Opportunistic Infections"
--Sulkowski M and others. Hepatotoxicity associated with NNRTI use: role of drugs and chronic viral hepatitis. Abstract and poster presentation 618.
HIV and Hepatitis Virus
The Center of Two Epidemics
By Jules Levin, Edited by Harvey S. Bartnof, MD
HCV Does Not Increase HIV Progression
If HAART Is Used
Many studies have shown that HIV can accelerate HCV disease progression in patients who are co-infected. However, studies have not shown definitively whether HCV can effect HIV disease progression. The goal of Dr. Mark Sulkowski's study was to examine the effect of HCV co-infection on CD4 cell recovery with HAART (highly active antiretroviral therapy). He also measured the effect of co-infection on HIV disease progression and survival after controlling for baseline CD4 counts, HIV viral load, and the use and effectiveness of HAART. These data came from patients followed at the Johns Hopkins HIV Clinic from 1996-2000. HCV antibody screening using the 2nd generation EIA (enzyme immunoassay) test has been ongoing since '96.
The primary outcome of the study was survival, CD4 cell recovery, progression to OI (opportunistic infections) and to a CD4 cell count less than 200 CD4 cells per microliter using "Cox proportional hazards" (statistical) models. A total of I,742 patients were in this analysis, 786 (45%) HIV+/HCV+, 958 (55%) HIV+/HCV-.
HCV+ patients were significantly older (39 vs. 37 years), significantly more likely to be African American (85 vs. 68%), and had significantly more "active or past" use of injection drug use (85 vs. 13%). Comparable factors at baseline for HCV+ and HCV- patients were CD4 counts (approximately 250 cells per microliter), HIV viral load (approximately 4.4 log or 25,000 copies per milliliter), and % with AIDS (38%). But, 76% of HCV+ and 71% of HCV- were treatment-naïve (no previous treatment), a significant difference. Follow-up for this study was a median of 549 days (range 365-900+), with no significant difference between arms.
HCV Blunts CD4 Increase to HAART
After one year on HAART, there was a significant difference in the CD4 count increase between HCV+ (n=272 patients) and HCV- (n=406) (60 vs. 96 cells, p=0.04). A similar significant difference occurred after the 2nd year of HAART (77 vs. 125 CD4 cell increase, p=0.05), among those who were HCV+ (n=161), or HCV- (n=240), respectively. In the 3rd year, the number of patients observed was smaller (81 & 46, respectively), and a non-significant but similar trend was observed (increase of approximately 140 vs. 180 CD4 cells, respectively). When looking at a small group with complete viral suppression similar trends were observed. Several other studies at Retrovirus reported similar findings in seeing CD4 increases blunted when HCV was present.
Probability of CD4 Decline to Less Than
200 Cells Not Associated With HCV
959 patients had a CD4 cell count greater than 200 cells at baseline, and 173 of them progressed to a CD4 count less than 200 during the observation period of up to almost 5 years. The "hazard ratio" (risk) was 1.28 for those who were HCV+ to have a decrease to less than 200 cells, when compared to those who were HCV-. But, in a "multivariate" analysis (that considers multiple factors) HCV+ was not significantly associated with a decline of the CD4 count to less than 200cells. In this analysis, Sulkowski considered HCV-infection, total exposure time on HAART, % of HIV viral load measurements that were greater than 400 copies per milliliter, and CD4 count at baseline. Those results suggested that the last three factors, but not HCV infection per se, were significantly associated with progression to a CD4 count less than 200 cells per microliter. (However, Sulkowski did not report the "p" values for that analysis.)
In a similar analysis to the one above, using "Kaplan-Meier" (statistical) survival estimates, among patients with baseline CD4 counts between 50-200 cells, there was a significant increased risk of death for HCV+ (HR "risk" 1.74) during the same observation period, when compared to HCV-. This might suggest that if someone started HAART when his/her CD4 count was between 50 and 200 cells, that there would be an increased risk for HIV disease progression & death if there also were HCV infection, when compared to someone with HIV but no HCV infection.
However, using a similar "multivariate Cox proportional hazards model," being HCV positive was no longer significantly associated with increased risk for death (based upon 95% "confidence intervals," BUT:
However, Sulkowski did not yet report statistical "p" values for the above analysis.
HIV patients with HCV co-infection and a CD4 count greater than 50 cells had significantly less exposure (total time) to HAART compared to HIV+/HCV- within the same CD4 range. Coinfected patients had detectable HIV viral load more often (viral load test results) compared to HIV+/HCV- patients (89% vs 20%). This could be related to adherence.
No Added Risk to HIV Progression
Using "Kaplan-Meier" statistical estimates, there was no difference in the probability of surviving without an OI between people with HCV or without HCV infection (during the period of up to five years).
No Significant Added Risk of Death
Having HCV did not significantly increase risk of death, again using Kaplan-Meier estimates, during the observation period of up to five years.
Sulkowski concluded HCV co-infection might be associated with impaired CD4 recovery from HAART for HIV at least during the first two years. But, he said his data appears to show that co-infection does not appear to adversely affect HIV disease progression after controlling for the use & effectiveness of HAART and the baseline CD4 count. He suggested adherence, active IVDU (intravenous drug usage), and HAART-associated liver toxicity might be related to immune recovery and the use & effectiveness of HAART in coinfected patients.
(Comments from Dr. Nancy Shulman: One piece of data that Dr. Sulkowski did not have available is the cause of death in the groups. If many of the HCV+ group died of liver-related causes vs. none of the HCV- subjects, particularly if they died with higher CD4 counts, then HCV might be impacting their survival.)
(Comments from Dr. Harvey Bartnof: There might be other co-factors that could affect the results of Dr. Sulkowski's study. They include: the types of HAART used (PI- vs. NNRTI-based HAART, HAART regimens changed or maintained), baseline and post-HAART levels of HCV viral load, HCV genotype (however, most have been reported elsewhere to be genotype 1), HBV (hepatitis B virus) co-infection and the inclusion or not of 3TC (lamivudine, Epivir, Epivir-HBV) that has anti-HBV activity as a part of HAART, other non-injection illicit drug usage including intranasal cocaine and oral alcohol, and nutrition. Nonetheless, Dr. Sulkowski's report certainly adds to our knowledge about HIV/HCV co-infection and response to HAART in this population.)
Sulkowski M and others. Effect of HCV co-infection on HIV disease progression and survival in HIV-infected adults. Abstract and oral presentation 34.