Hepatitis C virus transmission

Sprinz (2003) reported that in a Brazilian cohort of 1128 HIV-infected people, the HIV/HCV co-infection rate was 28.5%. Co-infection was higher in men than women (33% vs 19%), which was attributed to the higher rate of injecting drug use in the men. HIV infection via injecting drug use was associated with a 5.55-fold increased risk of HCV infection.

Hagan (2001) found that among a group of 317 initially HCV-negative injecting drug users (IDUs) in Seattle, sharing syringes was associated with a three-fold increase in the risk of HCV infection. Among those who did not share syringes, the risk was still three times higher in individuals who shared cookers and cottons used to prepare drugs.

Haley (2003) reanalysed data on 627 people with spinal disorders who underwent HCV testing in 1991. 43 patients had HCV antibodies and 48 reported some form of acute hepatitis in the past. 40 patients reported injecting drug use and 52 had been tattooed in a commercial tattoo parlour. The investigators found that a history of injecting drug use was strong linked with having acute hepatitis (hazard ratio [HR] 5.8, 95% confidence interval [CI], 2.5-13.8) and HCV (HR 7.2, 95% CI, 3.1-16.5). Tattooing was strongly linked to chronic HCV infection (HR 6.5 95% CI, 2.9-14.4), but not to an acute episode of hepatitis (HR 1.2, 95% CI, 0.5-3.3).

Marincovich (2003) studied a cohort of cohort of 171 HIV/HCV-negative people (152 women and 19 men) whose steady heterosexual partner had both HIV and HCV antibodies. Over 529 person-years (PY) of follow-up were counted and over 5800 incidents of unprotected anal and vaginal sex were calculated to have occurred and unprotected oral sex was estimated to have happened 25,000 times. Although 31 women became pregnant and one case of HIV seroconversion occurred, not a single case of HCV transmission could be identified. The investigators calculated that unprotected anal or vaginal sex had an HCV seroconversion rate of 0 to 6.3 per 10,000 acts of unprotected intercourse, and unprotected oral sex a rate of 0 to 1.4 per 10,000 acts.

Vandelli (2004) followed 895 HCV-uninfected individuals who had monogamous sexual relationships with HCV-infected partners; 776 individuals were followed for 10 years (the rest ended their relationships, were lost to follow-up, or the infected partner spontaneously clearly HCV). On average, the couples reported sexual intercourse 1.8 times per week. They reported that they did not use condoms, did not practice anal intercourse, and did not have sex during menstruation. Just three new HCV infections in the uninfected partners were observed during follow-up, for an incidence rate of 0.37 per 1000 PY. However, in one case the newly infected partner had a different genotype than the spouse, and in the other two cases the newly infected partners had different HCV viral isolates than their spouses, indicating that sexual transmission was not the cause of infection.

Pasquier (2003) reported that 9 of 35 (26%) co-infected men had intermittent HCV RNA in their semen. No factors were found to be associated with detectable HCV in the semen.

DOliveira (2005) studied 2691 HIV-positive heterosexual and homosexual individuals in Lyon, France, between 1992-2002 who were exposed to HIV through sexual activity (controlling for injecting drug use). Amongst the heterosexual subjects, 12.9% were coinfected with HCV; amongst the gay men, 5.8% had HCV. The risk of HCV infection differed according to sexual behaviour.

Augenbraun (2003) conducted a retrospective study involving a total of 2059 HIV-positive women and 569 HIV-negative women who were enrolled in the Womens Interagency HIV Study (WIHS). Stored blood samples obtained between 1994-1999 were used for the study. At baseline, HCV antibody tests were negative for 55% of the HIV-positive women and 69% of the HIV-negative women. Over a mean 3.5 years of follow-up, a total of 22 (1.5%) women seroconverted for HCV. However, the weakness of HCV antibody response in eight women led the investigators to conclude that only 14 of the women were true incident HCV infections (10 HIV-positive, 4 HIV-negative). 86% of the women who acquired HCV reported current or prior drug use, compared to 22% of the 1517 HCV-negative women at enrolment (p < 0.01).

Njouom (2003) studied 2554 pregnant women in Yaounde, Cameroon, who were screened for antibodies to HCV and HIV. HCV prevalence was 1.9% and HIV prevalence was 8.5%. HIV-infection did not significantly affect the prevalence of HCV, suggesting little HCV sexual transmission in this population and the failure of HIV to enhance susceptibility to HCV. These results may be confounded by false-negatives among HIV-positive women.

Msellati (2003) conducted a study of blood samples taken from 1002 women (half HIV-positive) in Abidjan, Ivory Coast, between 1995-1998. 1.2% of HIV-positive women and 1.0% of HIV-negative women had antibodies to HCV. HCV (genotype 1 or 2) was found in 7 of 10 tested samples. Hepatitis B infection was much more common, with 25% of the HIV-positive women and 10% of the HIV-negative women having active hepatitis B (p = 0.06).

Researchers recently have reported numerous cases of acute hepatitis C among HIV-positive gay men. 20 cases of acute hepatitis C were identified between October 2002 and January 2003 at the Royal Free Hospital in London, with an association between HCV infection and other STIs (Aizen 2003). 8 of the men had syphilis, 2 had recently acquired HIV infection, 2 had gonorrhoea and 2 had herpes simplex virus. All the men reported unprotected anal intercourse and 17 of 20 reported fisting and group sex. 2 of 20 reported ever having shared injecting equipment for drug use, while 70% (14 of 20) had shared notes or straws to snort cocaine or ketamine. Most men were infected with HCV genotype 1. At Chelsea and Westminster Hospital in London, 36 HIV-positive gay men were identified who seroconverted for HCV between February 1997 and March 2003 (Browne 2003; Nelson 2003; Gilleece 2004). 31 of 36 men reported recent unprotected anal sex and 11 had infectious syphilis in the year preceding HCV seroconversion. Only four men in this group reported injecting drug use. In these two cohorts, rates of spontaneous clearance and clearance after HCV treatment were high.

Chaix (2004) identified 12 HIV-positive gay men presenting with acute HCV infection at Necker Hospital in Paris. The mean age was 42 years, the mean duration of HIV infection was 9 years, and the median CD4 cell count was 61 cells/mm³. The only significant shared risk factor among all the men was sexual exposure. 10 of the 12 had HCV genotype 4d (genotype 4 usually accounts for less than 15% of HCV infections in Paris) and the viral strains were found to be related in a phylogenetic analysis.

Gö´º (2005) reported on a cluster of acute HCV infections among gay men in the Netherlands. Amongst a group of 11 mostly HIV-positive gay men who had a recent bout of lymphogranuloma venereum (LGV; an uncommon STI related to chlamydia), 7 had recently contracted HCV. The men with HCV were not more likely to practice unprotected anal sex, leading the researchers to conclude that transmission by semen was unlikely. However, all the men with recent HCV seroconversion reported unprotected anal fisting (receptive or insertive), compared with just 2 of the 9 HCV-negative men (p = 0.003). Non-injecting recreational drug use was reported by all of the HCV-positive and HCV-negative men, but the HCV-positive men denied injecting drug use.

Craib (2001) investigated 662 men in the Vancouver Lymphadenopathy AIDS Study. 5.9% had HCV co-infection. 8.8% of the HIV-positive group had HCV compared with 2.6% of the HIV-negative group. The men with HCV were more likely to report the following: more than 20 sexual partners in the last year; more than 100 lifetime partners; practising insertive fisting; practicing receptive anal sex; and practising anal-oral sex. The HCV-positive group was also more likely to report a history of injecting drug use and recreational/illegal drug use. 49% of the 39 HCV-positive men reported never injecting drugs. Comparison of the non-IDU HCV-positive group compared with the non-IDU HCV-negative group found oral-anal contact and insertive fisting to be associated with HCV infection. Multivariate analysis showed 3 factors independently associated with HCV infection: injecting drug use (p < 0.001); HIV infection (p = 0.042) and large number of male partners in the last year (p = 0.066).

Williams (2003) identified 263 cases of acute hepatitis C in six US counties between 1991-2001. 13% reported sexual exposure as their only risk factor (9% HCV-positive partner, 4% multiple partners). Viral analysis of blood taken from 7 couples showed that 5 had concordant HCV genotypes with very similar structures (97.8-99.3% identical). In the other 2 pairs, the viruses were 82% and 91% identical.

Weisbord (2003) studied 687 people attending an STI clinic in Miami. 4.7% had HCV infection. Multivariate analysis showed that injecting drugs, spending a day or more in jail, sexual contact with an HCV-infected person, and older age were associated with an increased risk of being HCV-positive. When sexual contact with an HCV-infected person was controlled for spending time in prison or injecting drug use, it was still found to be significantly associated with an increased risk of becoming infected with HCV.

Mendes-Correa (2002) reported the following risk factors for hepatitis C infection among HIV-infected men and women at an STI clinic in Brazil: injecting drug use; a sexual partner with a past history of chronic hepatitis; a sexual partner who had received a blood transfusion; anal intercourse; use of inhaled illicit drugs; and history of an IDU sexual partner.

Abrescia (2002) found that 20% of 40 women who had been infected with HIV by an HIV/HCV co-infected partner were also infected with HCV. The women had no other HIV/HCV risk factors apart from heterosexual contact with the co-infected partner.

Tedaldi (2003b) studied over 2700 HIV-positive patients involved in HIV treatment trials in several US cities between 1998 and 2001, as part of the Community Programs for Clinical Research on AIDS (CPCRA). Investigators found a HIV/HCV co-infection rate at baseline of 16.6%. A history of injecting drug use, older age (baseline average of 52 years versus 39 years), being African-American or Latino/Hispanic, and not taking anti-HIV therapy were found to be positively associated with HIV/HCV co-infection. However, a history of sex with other men was not associated with co-infection, contradicting the findings of several earlier studies. Neither CD4 cell count nor HIV viral load were predictive of HIV/HCV coinfection. Indeed, the prevalence of HIV/HCV coinfection was lower among gay and bisexual IDUs compared with exclusively heterosexual IDUs.

Hammer (2003) found no cases of sexual transmission of HCV in a retrospective San Francisco study comprising 576.6 PY of observation contributed by 703 individuals. The study, conducted amongst a largely gay cohort, also failed to find any association between unprotected anal sex and new HCV infections. No new cases of hepatitis C were detected. The overall hepatitis C prevalence was 2.5%, with 2.1% of the gay men in the study infected with hepatitis C. In multivariate analysis, age 50 or above (odds ratio [OR] 8.5; 95% CI, 2.6-27.7), HIV infection (OR 5.7; 95% CI, 1.6-20.6) remained associated with hepatitis C infection. Alary (2005) reported that during an 8 month period, only one new case of HCV occurred in a cohort of more than 1000 HIV-negative gay men in Montreal. At the start of the study, 3% of the men were HCV-positive; the HCV prevalence rate among the 980 men with no reported history of injecting drug use was 0.3 percent. The HCV-positive mens sexual behaviour did not differ significantly from that of the HCV-negative men. 92% of the men reported anal sex and 63% reported unprotected anal sex with either a regular or casual partner. The single new case was in an individual who reported sharing injecting equipment in the previous six months.

Gibb (2002) studied 441 HCV-infected mothers and their babies to investigate the impact of different types of delivery on HCV transmission. 31 women had elective caesarean sections (CS), 54 had emergency CS and 339 had vaginal deliveries. No infants delivered by elective CS contracted HCV, compared with 5.9% of the emergency CS group and 7.7% of the vaginal delivery group. 59 women breast-fed for an average of 6 weeks and this rate of breast-feeding did not impact transmission.

Herring (2004) found that HCV superinfection occurred with a frequency similar to that of first-time infections in a group of 25 young (under age 30) recently infected IDUs in San Francisco over an average period of 12 months. Comparing viral genetic sequences over time, the researchers identified five individuals (20%) with evidence of HCV superinfection after their initial seroconversion. Two of these were superinfected with HCV of a different genotype, while the remaining three were superinfected with divergent strains of the same genotype. The incidence of new initial HCV infections was 25%. Superinfection with a second HCV strain did not seem to lead to higher HCV viral loads.

Natural history of hepatitis C virus

Among 50 gay men with acute HCV infection diagnosed at Londons Chelsea and Westminster Hospital, 12 (24%) spontaneously converted to HCV PCR-negative (Gilleece 2005). Factors associated with spontaneous clearance included having a CD4 count above 500 cells/mm³ and higher peak ALT levels at diagnosis, indicating a greater hepatic inflammatory response. In the Royal Free Hospital cohort, 9 of 21 men cleared HCV without treatment (Bhagani 2004).

Jauncey (2004) reported that between 24% and 42% of HIV-negative IDUs recently infected with HCV spontaneously seroreverted to HCV-negative within two years of being diagnosed with acute hepatitis C. The researchers retrospectively analysed 57 newly infected IDUs seen at a Sydney STI clinic between January 1992 and May 2002 who had at least two documented HCV viral load tests. 63% reporting sharing drug injecting equipment. 24 (42%) cleared HCV in a median 5.9 months (range 1.4-11.2 months). 42% of individuals diagnosed with a positive HCV antibody test later seroreverted, compared with just 24% of those diagnosed with a positive HCV RNA test, suggesting that some of the former group had received a false-positive diagnosis. Individuals whose ALT level normalised to less than 40 U/L were almost twice as likely to clear HCV, but this did not reach statistical significance (p = 0.16). The fact that more than half the cohort was female may help explain the high spontaneous clearance rate. The one individual with documented HIV co-infection did not clear HCV.

Piasecki (2004) analysed 203 US veterans who spontaneously cleared HCV and 293 chronically HCV-infected controls. In univariate analysis, heavy alcohol users (OR 0.52; 95% CI, 0.31-0.85; p =.006) and people of African descent (OR 0.65; 95% CI, 0.44-0.96; p =.024) were less likely to spontaneously clear HCV. In multivariate analyses adjusting for race, HIV co-infection, age, and alcohol use, alcohol remained strongly associated with reduced HCV clearance (OR 0.49; 95% CI, 0.30-0.81; p =.005), but race/ethnicity was no longer statistically significant (OR 0.72; 95% CI, 0.48-1.09; p =.125). Individuals co-infected with HIV were less likely to clear HCV (OR 0.37; 95% CI, 0.16-0.83; p =.016), whilst those co-infected with HBV were more likely to do so (unadjusted OR 5.0; 95% CI, 1.26-28.6; p =.008).

Gerlach (2003) examined 60 patients diagnosed with acute HCV infection on the basis of positive HCV antibody status, detectable HCV RNA or clinical and biochemical criteria. 51 (85%) presented with some symptoms of acute hepatitis. Among the 46 symptomatic individuals who did not receive anti-HCV treatment, 24 (52%) spontaneously cleared the virus, usually within 12 weeks of symptom onset. All 9 of the asymptomatic patients developed chronic hepatitis C. Among the small number of treated individuals, 80% achieved undetectable HCV viral loads.

Thomas (2000) reported a community-based prospective cohort study of 1667 HCV-infected individuals enrolled between 1988 and 1989. 90 of 919 cleared HCV RNA, 722 had persistent viraemia and 107 had intermittent viraemia. Duration of HCV infection was estimated at over 15 years for 75% of the cohort. After a median follow-up period of 8.8 years, the overall rate of end-stage liver disease (ESLD) was 3.1 per 1000 PY. Heavy alcohol consumption (more than 3 standard drinks daily) was associated with greater risk of ESLD, but moderate consumption was not.

Poynard (1997) recruited 2235 people with HCV from 4 cohorts who had never received treatment with interferon. All underwent biopsy to determine stage of fibrosis. Follow-up showed a fibrosis progression rate of 0.133 fibrosis units, similar to previous studies. Median estimated time from infection to cirrhosis was 30 years. Three factors were associated with faster progression: male sex, age over 40 at the time of infection, and daily alcohol consumption over 50g. HCV genotype was not associated with progression. Non-drinking women infected before the age of 40 progressed most slowly, with a median 42 years to cirrhosis. 33% of HCV-infected individuals were expected to progress to cirrhosis within 20 years, 36% to progress between 20 and 50 years, and 31% to progress after 50 years or never.

Monto (2004) reported that whilst heavy alcohol use (more than 5 drinks per day) was associated with an increased fibrosis risk (p = 0.01), light (less than 2 drinks per day) or moderate (2-5 drinks per day) alcohol consumption did not appear to worsen fibrosis in a study of 800 participants. However, a range of fibrosis severity was seen in each category of alcohol use; some individuals who drank heavily had minimal fibrosis, whilst others who drank moderate or light amounts had more severe liver damage. These data suggest that some individuals may be more susceptible to the damaging effects of alcohol on the liver. Older age, higher ALT levels, and liver inflammation were also associated with fibrosis in this study.

Kenny-Walsh (1999) studied 376 HCV-positive women with a known date of infection. Average duration of HCV infection was 17 years. 81% had symptoms, most commonly fatigue. 47% had slightly elevated liver function tests (LFTs) and 8% had significantly elevated LFTs. 98% showed liver inflammation by biopsy and 51% had fibrosis. Only 2% had cirrhosis.

Liver-related morbidity and mortality in the highly active antiretroviral therapy era

Bica (2001) retrospectively compared death rates due to ESLD among HIV-infected patients at Bostons Lemuel Shattuck Hospital in 1991 and for a 12 month period during 1998-1999. The rate of HBV/HIV co-infection rose from 69% to 73% and HCV/HIV co-infection increased from 75% to 94% during this time. In 1991, 3 of 26 deaths (11.5%) were due to liver disease. In 1998-1999, 11 of 22 deaths (50%) were due to liver disease. In two deaths, ESLD was a co-factor. The rise in liver-related deaths was statistically significant. In the 1998-1999 period, 55% of deceased patients had undetectable HIV viral load at time of death, and had had CD4 counts over 200 cells/mm^{3,/sup> within the last year of life. 32% of participants in this study had to discontinue highly active antiretroviral therapy (HAART) due to liver toxicity.}

^{Ahmad (2001) found that ESLD was the second-leading cause of non-AIDS mortality (after sepsis) among HIV-positive individuals who died at Cook County Hospital in Chicago, accounting for 35% of deaths. A retrospective chart review of 98 of the 107 total deaths between 1998-2000 revealed that the group was 72% male, 79% substance users, 80% African-American, and had a median age of 42 years. 52% had CD4 counts below 50 cells/mm3,/sup> and 62% had an opportunistic infection (OI) during hospitalisation. 34 of the 64 patients tested (53%) were co-infected with HCV and 14 (22%) were coinfected with HBV.}

^{Puoti (2001a) investigated 829 HIV-infected people attending one HIV clinic in Italy, for a total 2028 PY of follow-up. 61% were co-infected with HCV, 8% with HBV and 6% with hepatitis D virus (HDV). 58% were previous or active IDUs, 16% had been diagnosed with AIDS and the average CD4 count was 331 cells/mm3,/sup>. Of the 47 deaths in this cohort during the follow-up period, 40% were due to ESLD. Multivariate analysis showed co-infection with HBV and severe liver toxicity while on HAART were associated with risk of death. However, HCV and HDV were not associated with greater risk of death in this cohort.}

^{Martin-Carbonero (2001) reviewed all admissions between January 1996 and December 2000 to a Madrid hospital. There were 1334 admissions of 875 HIV-infected people, of whom 82% were current or former IDUs. The proportion of admissions due to chronic liver disease increased from 9.4% in 1996 to 16% in 2000 (p < 0.05). The proportion of deaths due to liver disease rose from 9.3% in 1996 to 45% in 2000 (p < 0.05). Chronic hepatitis C was the only cause of liver disease in 75% of cases.}

^{Bonnet (2001) and Rancinan (2001) reported that ESLD was a leading cause of death in the French Aquitaine Cohort. 576 (58%) of 995 HIV-infected people in the cohort were co-infected with HCV. There was a non-significant trend showing a greater risk of death in the HCV/HIV co-infected group, despite their younger ages. HCV-related ESLD was responsible for 29% of the 105 deaths recorded between 1998-1999.}

^{Cacoub (2001) conducted 2 retrospective cohort analyses on 17,487 HIV-infected people in 1995 and 26,497 HIV-infected people in 1997. In 1995, 7.5% of the cohort died of AIDS, 0.12% died of cirrhosis or hepatocellular carcinoma (HCC), and 0.56% died of other causes. In 1997, AIDS-related deaths had fallen significantly to 1.73% (p < 0.001) while the number of deaths from HCV-related causes remained steady at 0.13%. Soriano (2001b) disputed these findings for several reasons including the fact that only deaths that occurred on infectious diseases and internal medicine wards were included in the study.}

^{Ragni (2001) evaluated 157 haemophiliacs with hepatitis C, including 85 co-infected with HIV. Average duration of HCV infection was 24 years. After controlling for age at HCV infection, HBV infection and alcohol use, HIV/HCV co-infected men had a greater risk of ESLD (relative risk [RR] 3.72) and death due to ESLD (RR 3.81). Antiretroviral therapy for HIV reduced the risk of ESLD, although the result was not statistically significant (p = 0.069). Increasing number of decades with HCV infection (RR 2.26, p = 0.0013) and HIV infection (RR 2.18, p = 0013) were associated with increased risk of ESLD.}

^{Rosenthal (2003) followed 25,178 HIV-infected people during 2001. 265 deaths occurred: 129 (48.7%) due to AIDS; 38 (14.3%) due to ESLD; and 98 (36.7%) due to other causes. 36 of the 38 people who died of ESLD had chronic hepatitis C. The rate of death due to ESLD among HIV-infected people was significantly greater in 2001 than in 1995 (1.5%, p < 0.01) and 1997 (6.6%, p < 0.01). The prevalence of HCC as a cause of death also increased over time (1995, 4.7%; 1997, 11%; 2001, 25%; p < 0.05).}

^{Monga (2001) followed 263 people with HIV alone, 166 people with HIV/HCV co-infection and 60 people with HCV alone for an average of 34 months. Average CD4 counts and HIV viral loads did not differ amongst the groups. ALT levels were higher in the co-infected patients compared to the HIV patients (52 vs 35 U/L; p < 0.05) and albumin levels were lower. Liver failure occurred in 10% of co-infected patients, but in none of the people with only one viral infection (p < 0.05). 11% of co-infected people died compared to 7% of the HIV group (p < 0.02). 47% of deaths amongst the co-infected group were due to liver-related causes.}

^{Bonacini (2004) analysed data from 472 HIV-positive patients. 72 had HIV/HBV co-infection, 256 had HIV/HCV co-infection, 18 had HIV plus more than one hepatitis virus and 126 had HIV alone. During the follow-up period, 134 subjects (28.4%) died, 55 of them (about 40%) due to liver-related causes. Subjects with more than one hepatitis virus were twice as likely (28%) to die of a liver-related cause than HIV/HBV or HIV/HCV co-infected subjects (15% and 13%, respectively), who in turn were twice as likely to die as people with HIV alone (6%). However, the rates of HIV-related deaths were similar in co-infected and HIV mono-infected subjects. Amongst the co-infected subjects, using HAART and having a CD4 count above 200 cells/mm3 were significantly associated with reduced liver-related mortality.}

Salmon-Ceron (2005) looked at ESLD deaths in HIV-positive patients at French hospitals in 2000. Among the 822 patients with documented HCV and HBV serostatus test results, 29% were HIV/HCV co-infected, 8% were HIV/HBV co-infected, and 4% had both HCV and HBV. Among the HCV co-infected patients, ESLD was the leading cause of death, more frequent than AIDS (31% vs 29%), but among the HBV co-infected patients, ESLD was a less frequent cause of death than AIDS (22% vs 38%). The highest proportion of ESLD deaths (44%) were in people triple-infected with HIV, HCV, and HBV. 43% of patients who died from ESLD had CD4 cell counts above 200/mm³, compared with 27% of patients who died from other causes.

Klein (2001) showed a greater risk of morbidity and mortality amongst 78 HIV/HCV co-infected people compared to 104 with HIV alone. Baseline characteristics were similar, with average CD4 levels just above 300 cells/mm³. Fewer of the HCV/HIV co-infected patients were taking HAART at baseline (23% vs 35%). For the co-infected and HIV mono-infected groups, respectively, the OI rates were 9.77 vs 7.91 per 100 PY, the mortality rates were 6.67 vs 2.27 per PY and hospitalisations were 15.03 vs 6.79 per 100 PY. HCV co-infection was associated with faster progression to death (p = 0.04).

In another report, Klein (2003) studied 1462 HIV-positive people, comparing pre-HAART (1989-1995) and post-HAART (1996-1999) cohorts. A total of 42 co-infected patients were included in the pre-HAART sample, and 83 in the post-HAART cohort. In the pre-HAART cohort, the co-infected patients were less likely to have an AIDS diagnosis on entry (12% versus 30%), and to have higher median CD4 cell counts (228 versus 147 cells/mm³). However, the investigators stated that these differences could be explained by the shorter duration of HIV infection seen in the co-infected patients, as the majority (64%) entered the cohort in 1994-95. Only 19% of the co-infected patients had an AIDS-defining illness at baseline, compared to 28% of the patients with HIV alone. In the pre-HAART era, the HIV/HCV co-infected patients had a consistently lower risk of adverse outcome than patients who were only infected with HIV. The rates were 11.0 versus 22.9 per 100 PY of follow-up for OIs; 9.7 versus 21.00 per 100 PY of follow-up for hospitalisation and 6.9 per versus 14.9 per 100 PY of follow-up for death. In the post-HAART era there was no difference in the risk of OIs between the co-infected patients and the patients with HIV alone (12.8 versus 13.1 per 100 person years). However, co-infected patients had much greater risk of hospitalisation and death compared to the individuals with only HIV (9.5 versus 5.3 per 100 PY of follow-up), and the risk of hospitalisation was 12.8 versus 13.1 per 100 PY. In the pre-HAART era, AIDS-related causes accounted for a similar proportion of deaths amongst the co-infected and HIV mono-infected patients (50% versus 43%). However, in the period after 1996, the main causes of death in patients who also had HCV were attributable to non-HIV-related infections and complications of injecting drug use (45% versus 15% of HIV-infected patients).Backus (2005) analysed data from a cohort of 12,200 HIV-positive US veterans who first received HAART between 1997-2003. 4668 patients (38%) were co-infected with HCV. Co-infected individuals were on average older and more likely to be African-American or Hispanic; have a history of psychiatric illness; and have drug or alcohol problems. Prior to starting HAART, co-infected patients had higher HIV viral loads than HIV mono-infected participants (median 26,000 copies/ml vs 19,000 copies/ml), but baseline CD4 counts were comparable (median 257 cells/mm³ vs 248 cells/mm³). HCV-positive and HCV-negative individuals had a similar virologic response to HAART (80% in both groups achieved undetectable viral load), but CD4 cell increases were lower amongst the co-infected patients (median 199 cells/mm³ vs 239 cells/mm³; p < 0.001). The overall mortality rate was 22% in the co-infected group compared with 14% amongst HIV mono-infected patients (p < 0.001).

Wong (2003) used fibrosis rates to estimate the risk of cirrhosis in HIV/HCV co-infected people in the age of HAART. Predicted outcomes were for people 35 years old with CD4 counts between 200-349 cells/mm³ and stage F2 fibrosis. He calculated that at 10 years, co-infected people have a 6% risk of cirrhosis or death due to liver disease and at 20 years, there is a 21% risk of cirrhosis and a 16% risk of death due to liver disease. Life expectancy for HIV mono-infected people with access to HAART is 16.5 quality years, and for those with HIV/HCV co-infection with access to HAART, life expectancy is 14.2 years. These data suggest that prolonged survival of co-infected people may lead to increased rates of liver-related illness and death.

Does HIV affect hepatitis C?

Darby (1999) examined mortality due to liver disease and liver cancer in 4865 haemophilic men and boys in the UK who had been treated between 1969-1985 with potentially HCV-tainted blood products. Based on death-certificate information, these men were 16.7 times more likely than men in the general population to die of liver disease (95% CI 12.5-22.0) and 5.6 times more likely to die of liver cancer (CI 1.8-13.0). For those not infected with HIV, the cumulative rates of death from liver-related causes was 1.4% (but increasing from 0.10% to 14.3% for boys and men first exposed to tainted blood products at increasingly older ages), as compared with 6.5% (increasing from 3.8% to 18.7%) for HIV/HCV co-infected individuals.

Lesens (1999) compared 81 co-infected haemophiliacs with 53 haemophiliacs infected with HCV alone. 22 co-infected haemophiliacs developed progressive liver disease (PLD) of whom 20 died. Mean survival period with PLD was 3.2 years. Co-infection increased the risk of PLD by 7.4 fold. HIV disease progression was significantly faster among the co-infected haemophiliacs with PLD, regardless of age at time of HCV infection.

Soto (1997) conducted a multi-centre, cross-sectional study of 547 people with HCV (of whom 116 were HIV co-infected). Time of HCV infection was estimated. Liver biopsies revealed 13 of 87 (15%) among the co-infected individuals and 7 of 272 (2.6%) among the HCV mono-infected people developed cirrhosis within 10 years of HCV infection. Time to cirrhosis was 23.2 years in the HCV individuals compared with 6.9 years in the co-infected individuals. This was a significant difference.

Sulkowski (2005a) looked at sequential liver biopsies from 61 HIV/HCV co-infected individuals. Median time between biopsies was 2.84 years; patients with advanced fibrosis (stage F5-F6) on first biopsy were excluded from analysis. At the second biopsy, 73% experienced one stage or less fibrosis progression, whilst 17% experienced 2- or 3-stage progression. Amongst those with an initial score of F0-F1, 28% had at least a 2-stage increase in fibrosis. Rapid progression was slightly but not significantly associated with higher HIV viral load (p = .03). No relationship was detected between fibrosis progression and age, sex, alcohol consumption, use of HAART, or CD4 cell count. The researchers suggested co-infected patients - even those with little or no fibrosis at baseline - may need a liver biopsy at least every 3 years.

Benhamou (1999) reported that liver fibrosis progresses more rapidly in HIV/HCV co-infected individuals. Extensive fibrosis (stages F2 through F4) and severe inflammatory activity were detected in 54% of 122 HIV/HCV co-infected patients, compared with 30% of 122 individuals with HCV alone. The overall median time to develop cirrhosis in the co-infected individuals was 26 years compared with 34 years in those with HCV alone. However, other factors including age at the time of HCV infection, alcohol consumption, and immunosuppression (CD4 count below 200 cells/mm³) also affected the rate of liver damage. HIV/HCV-co-infected participants with over 200 cells/mm³ had progression rates similar to those of people with HCV alone. As examples, a co-infected individual with a CD4 count of 200 cells/mm³ and alcohol intake of more than 50 grams per day would be expected to take a median 16 years to develop cirrhosis, whilst a co-infected person with a high CD4 count and alcohol intake of less that 50 grams per day would be expected to take a median of 36 years.

Mohsen (2003) compared the progression rate of liver fibrosis in 55 HIV/HCV co-infected people and 153 people with HCV alone at two London hospitals. The median progression rate in coinfected people was 0.17 fibrosis units per year and 0.13 in HCV mono-infected people. The average time from HCV infection to cirrhosis was estimated to be 23 years in co-infected people and 32 years in those with HCV alone. In addition to HIV serostatus, older age at the time of HCV infection, higher ALT levels and higher inflammatory activity were also linked to more rapid progression.Pineda (2005) retrospectively analysed data from 1037 HCV mono-infected and 180 HCV/HIV co-infected patients with decompensated cirrhosis. Survival with ESLD was significantly shorter in the co-infected group. During follow-up, 100 (56%) co-infected and 386 (37%) HCV mono-infected patients died, yielding median survival times of 16 and 48 months, respectively (p < .001). The researchers suggested that liver transplantation should be considered sooner for co-infected individuals

Sulkowski (2005b) reported that hepatic steatosis was present on biopsy in 40% of 112 co-infected patients (99% with HCV genotype 1), most of whom were taking HAART. 22% had grade 1 steatosis, 13% had grade 2, and 5% had grade 3. Steatosis was more common amongst Caucasian patients; those with higher fibrosis scores; those with hyperglycaemia; obese individuals; and those with prior exposure to d4T (Zerit); only 4 subjects had never taken d4T, none of whom had steatosis. However, steatosis risk was not associated with history of alcohol use.

Garcia-Samaniego (2001) reported on 7 HIV/HCV co-infected people with ESLD. Mean time between HCV infection and development of HCC was 17.8 years. Only one person had been diagnosed with AIDS prior to HCC diagnosis but all had CD4 counts below 500 cells/mm³. 5 of 7 died within the first 4 months of follow-up. Compared with a control group of HCV mono-infected people with HCC, the co-infected group were significantly younger (42.2 vs 68.9 years; p < 0.001) and had been HCV-infected for less time (17.8 vs 28.1 years; p < 0.05).

In a retrospective chart review including 40 HIV/HCV co-infected and 50 HCV mono-infected US and Canadian veterans with HCC, Bräµ (2005) found that the co-infected individuals developed HCC at younger ages and that time from initial HCV infection to HCC was 7.5 years less. However, survival times were similar in both groups.

Vincent di Martino (2002) found that co-infected people had higher HCV viral loads, higher inflammatory activity scores and were less likely to respond to treatment with conventional interferon plus ribavirin than HCV mono-infected individuals. However, whilst sustained virological response (SVR) rates in co-infected people were lower than those in people with HCV alone, histological response rates were similar. Kramer (2005) conducted a retrospective analysis of data from more than 26,641 US veterans with HCV alone and 4761 with HIV/HCV co-infection. After controlling for demographic and confounding factors, co-infection was not significantly associated with cirrhosis overall, although this association was present before the advent of HAART (October 1, 1996) (HR=1.48; p = 0.02). Co-infection was not associated with increased risk of HCC either before or during the HAART era.

Fultz (2004) examined data from the US Veterans Administration (VA) and two cohorts of HIV-positive patients (VACS 3 and CHORUS) to determine the effect of HCV co-infection on survival. In the two clinical cohorts, HCV co-infection was associated with increased mortality (HR 1.38; p = 0.01). When analysing the data by CD4 count, however, the death rate increase was primarily seen in co-infected people with fewer than 100 cells/mm³.

Puoti (2001b) conducted liver biopsies on 84 HIV/HCV co-infected individuals and 120 individuals with HCV alone. Fibrosis was associated with a CD4 count below 500 cells/mm³ (OR 3.2; p = 0.037) regardless of HIV serostatus, duration of HIV infection, sex or alcohol intake.

Tor (2001) compared 114 HCV/HIV co-infected individuals with 57 matched controls with HCV alone. Degree of liver inflammation and fibrosis were established by biopsy. There was no significant difference in inflammation or fibrosis between the two groups. Factors associated with inflammation were high ALT and AST levels, high HCV viral load and non-use of HAART. CD4 count and HIV viral load did not impact liver inflammation or fibrosis, suggesting that HIV does not accelerate HCV disease.

Sterling (2003) compared a range of disease markers in 66 consecutive HIV/HCV co-infected patients and 119 HCV mono-infected controls. Virological, histological and biochemical markers were not significantly different between the co-infected and mono-infected patients. In the co-infected group, 86% had CD4 counts above 200 cells/mm³, 32% had advanced fibrosis, 51% had normal ALT levels, 92% had HCV genotype 1, the mean HCV RNA was 5.7 log IU/ml and the mean histologic activity index score was 6.86. As previously reported, markers and symptoms did not provide a reliable assessment of degree of fibrosis or cirrhosis.

Monto (2005) studied 92 co-infected and 372 HCV mono-infected patients undergoing liver biopsy. While liver inflammation and degree of fibrosis were similar in both groups, steatosis was less in the co-infected patients. In a multivariate analysis, amongst the mono-infected patients, steatosis was associated with elevated body mass index, male sex, older age, and genotype 3a HCV infection, while amongst the co-infected patients age was the only predictor of steatosis.

Does hepatitis C virus affect HIV disease?

Sulkowski (2002b) studied HIV/HCV co-infection in 1955 HIV mono-infected patients and 873 HIV/HCV co-infected people over 2 years. The rate of death over 2 years was 17.5% in the co-infected group compared to 15.5% in the group with HIV alone. A similar proportion of co-infected and HIV mono-infected individuals progressed to AIDS (26.4% vs 24.4%). Only 54% of co-infected people received antiretroviral therapy compared to 67% of the HIV mono-infected group. Although the untreated HIV/HCV co-infected patients had a relatively high incidence of AIDS and death, this study suggests that HCV does not accelerate HIV disease progression.

Amin (2004) retrospectively analysed stored blood samples of 1649 people enrolled in the CAESAR trial, a pre-HAART (early 1990s) international study of 3TC. Participants had relatively advanced immune suppression (CD4 count range 25-250 cells/mm³). Overall, 265 people (16.1%) were found to be HIV/HCV co-infected. However, co-infection rates varied considerably by region, from 1.6% in South Africa to 9.2% in the UK to 48.6% in Italy. Co-infection had no significant impact on HIV disease progression. Median CD4 count increases were similar in the co-infected and HIV mono-infected individuals (4 vs 6 cells/mm³), as were the rates of new OIs (11% vs 13%).

Tedaldi (2003), with the US HIV Outpatient Study (HOPS), analysed data from 823 HIV-positive patients from three large urban clinics. HIV/HCV co-infection was present in 267 (32.4%). Similar numbers of patients in the HIV/HCV co-infected and HIV mono-infected groups were treated with HAART (approximately 85%), for an average of 1.9 years. Co-infected patients, however, started HIV therapy later than HIV mono-infected patients, and with a lower CD4 count (164 vs 231 cells/mm³). 7% of the co-infected patients received anti-HCV therapy, normally late in the follow-up period. Co-infected patients were more likely to be unwell, and more likely to die during follow-up (10.9% vs 6.7%). The researchers, however, attributed this to differences in CD4 count, duration of HAART and age, not to co-infection with HCV. Of all co-morbidities, only liver abnormalities continued to be seen at increased rates amongst the co-infected patients.Hershow (2005) reported on 652 HIV-positive participants in the US Women and Infants Transmission Study enrolled between 1989-2003, for a total of 1761 PY of follow-up. 190 (29%) were co-infected with HCV and 19% received HAART during follow-up (21% of the HIV mono-infected and 13% of the co-infected women). Co-infected women were significantly older at the time of delivery (29 vs 26 years) and more likely to have a history of hard drug or alcohol use. HIV viral load was similar in both groups, whilst average CD4 cell percentage was 1.9% higher amongst the co-infected women (p < 0.001). The co-infected women did not progress to an AIDS-defining illness or death faster than those with HIV alone. The researchers concluded that co-infected women did not experience faster progression to AIDS or death.

Rockstroh (2005) reported data on 5957 individuals in the EuroSIDA cohort; 1960 (33%) were HCV co-infected. On study entry, co-infected patients had higher median viral loads (1600 copies/ml vs 700 copies/ml; p = 0.002) and were more likely to have an AIDS diagnosis (27% vs 21%; p < 0.001), abnormal liver function and HBV co-infection (9% vs 7%; p < 0.001) compared with HIV mono-infected people. Significantly more AIDS-defining events (p = 0.024), mortality due to any cause (p < 0.001) and liver-related deaths (p < 0.001) occurred in the co-infected group. Co-infected patients responded to HAART as well as HIV mono-infected participants, with 91% vs 89%, respectively, achieving viral loads below 500 copies/ml, and 80% and 83%, respectively, experiencing CD4 increases of at least 50 cells/mm³. After adjusting for baseline characteristics, initiation of HAART, and CD4 cell changes, the investigators concluded that HCV co-infection did not increase the risk of HIV disease progression or death due to AIDS.

Greub (2000) conducted a prospective study of 1157 HIV/HCV co-infected people and 1954 people infected with HIV alone in the Swiss HIV Cohort who commenced HAART between June 1996 and May 1999. At baseline, the HIV/HCV co-infected group had an median CD4 count of 172 cells/mm³ and median viral load of 39,800 copies/ml, whilst the HIV mono-infected group had a CD4 count of 222 cells/mm³ (p < 0.001) and viral load of 31,600 copies/ml (p = 0.029). In addition to the lower baseline CD4 count, the co-infected group were more likely to have taken anti-HIV drugs previously (52% vs 59%) and were more likely to have been diagnosed with AIDS (23.5% vs 28%). About 88% of the co-infected group had a history of injecting drug use. Median follow-up was 28 months. The probability of disease progression at 2 years was 6.6% for HCV-negative non-IDUs; 9.7% for HCV-positive non-IDUs; 15% for HCV-positive IDUs. The co-infected subjects (37% of the total) were more likely to develop new AIDS-defining conditions (7.5% vs 4.7%) and more than twice as likely to die of any cause (8.8% vs 4%) than those with HIV alone (combined HR 1.7). Among those with HIV viral loads below 400 copies/ml, disease progression was more strongly associated with HCV co-infection (HR 3.5).

Soriano (2001c), in a letter in The Lancet responding to Greub, reported further evidence of poorer immunological and virological outcomes in HIV/HCV co-infected patients despite similar rates of anti-HIV treatment and comparable adherence. Among 902 HIV-infected patients followed since January 1998, HCV-negative people had higher mean CD4 counts (620 vs 518 cells/mm^{3,/sup>; p < 0.001) and lower HIV viral load (6351 vs 11188 copies/ml; p < 0.001) than co-infected people at 2 years. Soriano stated that chronic immune activation in co-infected people may be the cause of poorer clinical outcomes. Anderson (2004) conducted a retrospective review of medical records from 970 individuals in the HIV Atlanta VA Cohort Study (HAVACS). 76% of patients received HAART during the study period and 67% had been diagnosed with AIDS. A majority of the cohort were black (73%). The HIV/HCV co-infection rate was 32%, and those with HCV were significantly more likely to be older, have a history of injecting drug use, and to have never taken HAART. Co-infected patients had shorter survival times after HIV diagnosis than did HIV mono-infected patients (p = 0.009). The time from AIDS diagnosis to death was also shorter for co-infected individuals (p = 0.022). After starting HAART, both co-infected and HIV mono-infected patients experienced similar short-term and long-tern CD4 cell recovery. However, injecting drug use was independently associated with decreased CD4 T-cell gain.}

^{Dorruci (2004) analysed data from 1052 recent HIV seroconverters in the Italian HIV Seroconversion Study (924 enrolled pre-HAART and 795 post-HAART); about 57% were co-infected with HCV. During the pre-HAART period, fewer co-infected participants developed AIDS compared with HIV mono-infected patients, but the difference did not reach statistical significance (21% vs 24%; p = 0.180). During the post-HAART period, however, the co-infected group was significantly more likely to progress to AIDS (15% vs 9%; p = 0.009). During the post-HAART period, co-infected individuals had a shorter duration of HAART use than the HIV mono-infected patients (31% vs 41% of their total time at risk for HIV disease progression; p = 0.001).}

^{Stebbing (2005) analysed data from about 1500 HIV-positive patients tested for HCV at Londons Chelsea and Westminster Hospital since HAART became available (January 1996). 85 patients (about 6%) were co-infected with HCV. In a multivariate analysis, co-infected patients were 52% more likely to develop an AIDS-defining illness or progress to a CD4 cell count below 200 cells/mm3, although the rate of CD4 cell decline was similar in both groups. The researchers said their data suggest that HCV infection is having an effect on HIV-1 disease progression that is not reflected in CD4 cell count.}

Van Asten (2004) analysed data from 126 HIV/HCV co-infected injecting drug users in the European Seroconverter Study and the Italian Seroconverter Study. HCV genotype 1 was most common (48%), followed by genotype 3 (34%), genotype 4 (13%), and genotype 2 (1%); 5% were infected with more than one genotype. After a median follow-up of about 7 years, subjects with genotype 1a and those with multiple HCV genotypes experienced more rapid progression to 200 or fewer CD4 cells/mm³ compared to those with genotype 3 (adjusted HR 2.02, 95% CI 1.04-3.92 for genotype 1; HR 2.74, 95% CI 0.95-7.90 for multiple genotypes). Those with multiple HCV genotypes experienced slightly faster clinical progression to AIDS overall, but significantly faster progression in the pre-HAART era, leading the researchers to suggest that effective anti-HIV therapy may diminish the effect of HCV genotype on HIV disease progression.

In a meta-analysis of data from 8 trials involving 6216 participants, Farmer Miller (2005) found that HIV/HCV co-infected patients had mean CD4 cell increase 33.4 cells/mm³ less than that seen in HIV mono-infected patients. The results were not influenced by the year in which patients started HAART. The researchers concluded that their analysis shows that patients with HIV/HCV co-infection do, in fact, have less immune reconstitution, as determined by CD4 cell count after 48 weeks of HAART, than do patients with HCV infection alone.

Braitstein (2003) presented data from the HOMER cohort demonstrating that antiretroviral-naive HIV/HCV co-infected patients gained far fewer CD4 cells after starting HAART than HIV-monoinfected individuals. Between 1996 and 2000, 1416 HIV-positive patients started HAART; of these 552 (39%) were also infected with HCV. 18 months after starting HAART, the average increase in CD4 count was 190 cells/mm^{3,/sup> in HIV mono-infected people compared with less than 50 cells/mm}3,/sup> in the co-infected patients. Amongst patients with 95% adherence, the average CD4 cell count in the HIV mono-infected patients was 230 cells/mm^{3,/sup> compared with 120 cells/mm}3,/sup> in co-infected individuals. Looking at immunological recovery in patients with baseline CD4 counts below 200 cells/mm^{3,/sup>, HIV mono-infected people gained an average of 78 cells/mm}3,/sup> after 18 months on HAART, compared with just 30 cells/mm^{3,/sup> in co-infected individuals (p < 0.001). Controlling for all possible confounders, the investigators found that HIV mono-infected patients gained an average of 33 CD4 cells per day, compared to an average loss of 8 cells per day in co-infected individuals (p < 0.001).}

^{Martin (2001) also found co-infection to be linked to impaired CD4 cell recovery in a study of 902 HIV-positive individuals (72% co-infected with HCV). During follow-up between January 1998 and April 2000, the co-infected people gained an average 53 cells/mm3,/sup> whilst the HIV mono-infected individuals saw an average 111 cells/mm3,/sup> increase. Furthermore, HIV RNA declined by an average 606 copies/ml in the co-infected group, compared with a decrease of 5788 copies/ml in the HIV mono-infected group.}

Lincoln (2003) found that among 2086 HIV-positive people in the Australia HIV Observational Database, both HIV/HCV and HIV/HBV co-infected individuals had a virological response similar to that of subjects with HIV alone after a year on HAART, but those with HCV had a slightly poorer immunological response, gaining 32 fewer CD4 cells/mm^{3,/sup> than HIV-monoinfected subjects.}

^{In Kleins study (also discussed above, 2003), similar proportions of co-infected and HIV mono-infected individuals achieved HIV viral loads below 500 copies/ml (82% vs 81%) and there was no significant difference in the time to virological failure. However, the investigators did find that HCV-positive individuals were significantly less likely to achieve a CD4 cell count increase than HCV-negative patients (HR=0.48; p = 0.04).}

^{Cowling (2004) studied 6401 HIV-positive patients in the international ATHENA cohort, of whom 947 (15%) were co-infected with HCV. After adjusting for baseline HIV disease stage and year of HAART initiation, HCV co-infection was significantly associated with lower mean CD4 counts, but the time to HIV viral suppression did not differ significantly in the co-infected and HIV mono-infected groups.}

^{In Andersons study (2004), both co-infected and HIV mono-infected patients experienced a similar CD4 cell recovery by month 6 after starting HAART. However, injecting drug use was independently associated with decreased CD4 T-cell gain.}

^{Duncombe (2004) studied a cohort of 692 HIV-positive individuals in Thailand. Unlike US and European studies, about half were male, the mean age was 32 years, and only 22% had previously received HAART. 7.2% were co-infected with HCV, 8.7% with HBV, and 0.4% with both hepatitis viruses. After starting HAART, median reductions in HIV RNA were similar in the HIV mono-infected, HIV/HCV co-infected and HIV/HBV co-infected groups. CD4 cell increases were significantly lower in the HIV/HCV and HIV/HBV groups at week 4 (62, 33, and 29 cells/mm3, respectively), but were similar by week 48 (115, 97 and 113 cells/mm3).}

Antonucci (2005) analysed data from 284 HCV-positive and 1219 HCV-negative patients in the Italian Cohort Naive for Antiretrovirals who were starting HAART for the first time. The HCV-negative people were more likely to achieve a CD4 count increase of at least 100 cells/mm³ - and to do so faster - compared with HCV-positive individuals (23 vs 29 weeks; p = 0.001). HCV viraemia was only slightly associated with reduced likelihood of achieving a CD4 count increase of 100 or more cells/mm³ (p = 0.06), but significantly increased the time it took to achieve a CD4 count of at least 300 cells/mm³ (p = 0.01). The likelihood of achieving a CD4 cell increase of 100 or more cells/mm³ did not differ by HCV genotype, but those with genotype 3 were less likely to achieve a CD4 count of at least 300 cells/mm³ compared to patients with genotype 1 (p = 0.02). The researchers said their data support a direct role of HCV viremia in the CD4 cell count response to HAART, speculating that HCV may actively replicate in the same T-cells as HIV.

Yang (2004) reported that HCV/HIV co-infection adversely affected neuropsychological performance and may also be associated with depressed mood. In the US ACTG A5095 study, neuropsychological assessment was performed on 25 HIV/HCV co-infected people and 210 who had HIV infection alone. The co-infected participants had significantly lower neuropsychological performance scores overall (p = 0.012); in particular, they did less well on the Digit Symbol task used to measure attention, visual-motor coordination, mental flexibility and information processing speed (p < 0.001). In terms of depression, 52% of the co-infected subjects had clinical depression, compared with 33% of those with HIV alone. A multivariate analysis suggested that HCV infection was significantly associated with reduced performance on the Digit Symbol task, even after controlling for depression and other potentially confounding variables such as education level, injection drug use, alcohol use, hepatitis B status, CD4 count, and HIV viral load. Although the co-infected subjects had higher ALT and AST, levels were not exceedingly high, and the types of cognitive impairment seen did not resemble the profile of hepatic encephalopathy.

Cherner (2005) looked at the interaction of HCV, HIV, and methamphetamine use as risk factors for neurocognitive impairment. The study included 40 HIV/HCV co-infected methamphetamine users, 110 participants with two risk factors, 190 with one risk factor, and 90 uninfected non-users. 37% of HIV-positive methamphetamine users had HCV, compared with 2% of HIV-negative non-users. The researchers used various neuropsychiatric tests to assess learning, recall, attention, working memory, information processing speed, verbal fluency, abstraction, problem solving, and motor ability. After adjusting for demographic factors, they found that HCV serostatus was a significant predictor of decreased neuropsychological performance, both overall (p < 0.02) and in the areas of learning (p < 0.02), abstraction (p < 0.02), and motor skills (p < 0.04). However, HCV serostatus did not predict attention and working memory or verbal fluency. HCV, HIV, and methamphetamine use all were independently associated with cognitive impairment, and also had an additive effect.

Diagnosis, genotypes and normal alanine aminotransferase

Berggren (2001) compared 10 HIV-infected people who had detectable HCV RNA but no HCV antibodies. All had CD4 counts below 150 cells/mm^{3,/sup>, and a CD4 count below 100 cells/mm}3,/sup> was strongly associated with a false negative HCV antibody result. 3 became HCV antibody positive as CD4 count increased. From a control group of HIV-positive, HCV RNA-negative, HCV antibody-positive individuals, 2 people became HCV antibody-negative as their CD4 counts declined. Some patients were HCV antibody-negative despite having antibodies to cytomegalovirus (CMV), toxoplasmosis, and hepatitis A and B.

John (1998) reported 3 cases of acute hepatitis in 133 individuals with HIV who started HAART. All had HCV, although 2 had not previously been diagnosed. Stored blood samples from these two individuals showed HCV RNA but undetectable antibodies to HCV. HCV antibodies became detectable after immune recovery due to HAART.

George (2002) collected blood samples from 131 HIV-positive patients from a US VA clinic and 102 HIV-negative diabetics. Participants were asked to complete a questionnaire about hepatitis C risk factors and symptoms and alcohol consumption. All the HIV-infected patients were receiving HAART and none of the study population was being treated for hepatitis C. Tests were performed for the presence of HCV antibodies and HCV RNA. Of the HIV-positive individuals, 31 (23.7%) tested positive for HCV antibodies, as did 2% of the HIV-negative control group. HCV RNA testing of the HCV antibody-negative blood samples found that an additional 20 HIV-positive patients, but none of the HIV-negative individuals, had detectable HCV RNA. Among the HCV-infected people, the absence of HCV antibodies was associated with a lower CD4 count and acquisition of HCV via sexual transmission.

Quereda (2003) studied the impact of liver biopsy in 112 HIV/HCV co-infected people. 42% had no or mild fibrosis (F0 or F1) and 58% had significant or severe fibrosis, including 39.2% with cirrhosis. In 26% of cases, the physician predicted the finding of cirrhosis; pre-biopsy testing did not accurately predict cirrhosis.19 co-infected people reversed the decision to treat following a biopsy finding of little or no fibrosis. In another study (2002), Quereda reported that gamma-glutamyltranspeptidase (GGT) level along with alcohol abuse, duration of HCV infection, age at HCV infection, HCV genotype, HCV viral load, and having an AIDS diagnosis were all predictive of fibrosis in 99 HIV/HCV co-infected patients with a median duration of 17 years HCV infection and median CD4 count of 494 cells/mm³ at biopsy. GGT level above 100 IU/ml was found to correlate with severe histological activity index (OR 6.6) after controlling for alcohol use, duration of HCV infection and age at HCV acquisition, HCV genotype, HCV RNA level and AIDS diagnosis.

Myers (2003) reported on the utility of an index of five biochemical and virological markers at predicting HCV-related liver fibrosis in 130 HIV/HCV patients: age, sex, total bilirubin, GGT, alpha2-macroglobulin, apolipoprotein A1 and haptoglobin.Investigators found a significant relationship between the stage of fibrosis and alpha2-macroglobin and apolipoprotein. Predictors of septal fibrosis were alpha2-macroglobulin, apolipoprotein A1 and male sex, with GGT of borderline significance. Investigators argued that it would have been possible to restrict biopsies to patients with scores of borderline significance, therefore avoiding liver biopsies in 55% of patients while maintaining 89% accuracy.

Kelleher (2005) evaluated non-invasive fibrosis prediction in 95 HIV/HCV coinfected patients using the SHASTA index, measuring ALT, AST levels, AST-to-platelet ratio index (APRI), albumin, total bilirubin, hyaluronic acid (HA), and YKL-40 (a glycoprotein thought to be associated with tissue regeneration). Fibrosis scores of F3 or higher were 27 times more common in subjects with high HA levels (above 86 ng/mL). Advanced fibrosis was less strongly associated with decreased albumin and elevated AST. All 35 subjects who had normal or near-normal HA, albumin, and AST levels also had minimal fibrosis. The authors concluded the SHASTA index could accurately stage mild and advanced fibrosis.

Lyra (2004) reported on the epidemiology of genotype 4 HCV in the US, asking hepatologists for information about genotype 4 patients they had treated and searching the St. Louis University Hospital database for cases seen between 1999 and 2002. Medical charts were reviewed for patient demographics, HCV risk factors, and response to therapy. 20 individuals with genotype 4 were identified; most had a history of injection drug use. Of the 17 patients treated with interferon plus ribavirin (14 of whom completed therapy), 10 (59%) achieved SVR.

Uberti-Foppa (2004) conducted a retrospective analysis of 354 liver biopsies from HIV/HCV co-infected patients. At the time of biopsy, 26 subjects had normal and 328 had abnormal ALT levels. HIV RNA and CD4 counts were similar in the two groups, as was the range of antiretroviral regimens. Mild liver fibrosis (stage 0-1) was detected in 75% of patients with normal and in 42.9% with abnormal ALT. Moderate fibrosis (stage 2-3) was present in 12.5% and 33.8%, respectively. Severe liver damage (stage 4-6) was seen in 12.5% and 23.2% (p = 0.039).

Fonquernie (2004) found that among 155 HIV/HCV co-infected individuals, 28.5% had persistently normal ALT levels and 71.5% had high ALT levels. Those with normal ALT had significantly lower fibrosis scores and a lower fibrosis progression rate. Three factors were associated with consistently normal ALT: being female (p = 0.05), being negative for HBsAg (HBV surface antigen) (p = 0.003), and having HCV genotype 4 (p = 0.01).

Peginterferon / ribavirin in HCV-monoinfected people

In pivotal study C/198-589 (Manns 2001), 1530 individuals with chronic hepatitis C diagnosed by HCV PCR were randomised to receive peginterferon 1.5µg/kg/week plus ribavirin 800mg or peginterferon 1.5µg/kg for one month plus ribavirin 1000mg or 1200mg by body weight, then 1.0µg/kg plus ribavirin for eleven months, versus conventional interferon alfa 2b (3 million IU three times weekly) plus ribavirin 1000mg or 1200mg by body weight for 12 months. SVR 24 weeks after completing treatment was significantly greater in the peginterferon 1.5µg/kg arm compared to the interferon alfa group (54% vs 47%, p = 0.0143).

Fried (2002) randomised 1121 people with hepatitis C to one of three treatment arms: peginterferon alfa 2a (Pegasys) once weekly plus daily ribavirin (1000mg or 1200mg, depending on body weight); weekly Pegasys plus daily placebo, or 3 million units of conventional interferon alfa 2b thrice weekly plus daily ribavirin for 48 weeks. 56% of the Pegasys/ribavirin group had undetectable HCV 24 week after the end of treatment compared to 44% of those on interferon alfa 2b/ribavirin, and 29% of those taking Pegasys alone (p < 0.001). Among those with HCV genotype 1, a sustained response occurred in 46%, 36%, and 21%, respectively. Among people with HCV genotype 1 and high baseline HCV viral load, 41%, 33%, and 13%, respectively, had a sustained response. Side-effects were similar for the 3 groups. Flu-like symptoms and depression were less common among people receiving Pegasys than in the group receiving interferon alfa 2b plus ribavirin.

Wirth (2005) conducted an open-label study treating 62 children and adolescents (age 2-17 years) with Peg-Intron (1.5µg/kg body weight once weekly) plus ribavirin (15 mg/kg daily) for 48 weeks; 23 subjects who did not show an early virological response (EVR) discontinued therapy after 6 months, and 1 dropped out early. 22 of 46 (48%) children with genotype 1 HCV achieved SVR, as did all 13 (100%) with genotypes 2 or 3, and 1 of 2 (50%) with genotype 4. By transmission route, 19 of 27 (70%) with parenteral (blood-to-blood) infection, 12 of 25 (48%) with mother-to-child infection, and 5 of 9 (56%) with unknown transmission route achieved SVR.

Jeffers (2004) conducted an open-label study of 78 HIV-negative participants of African descent and 28 non-Hispanic white patients, all with genotype 1, receiving Pegasys plus ribavirin for 48 weeks. Black participants were more likely to be male, older, somewhat heavier, and have higher HCV viral load. 26% of blacks and 39% of whites achieved SVR in an intent-to-treat analysis. About twice as many blacks had their Pegasys dose reduced due to neutropenia.

Muir (2004) treated 100 black and 100 white patients (98% in both groups with genotype 1) with Peg-Intron plus ribavirin for 48 weeks. The black patients were more likely to be male, older, heavier, and infected with HCV slightly longer. 19% of blacks achieved SVR, compared with 52% of whites. In a multivariate analysis, only black race/ethnicity was significantly associated with lower response rates.

Peginterferon monotherapy in hepatitis C virus monoinfected people

Heathcote (2000) compared Pegasys with conventional interferon alfa in 531 individuals with chronic hepatitis C. The study lasted 48 weeks, followed by 24 weeks follow-up after treatment was completed. All patients were HIV-negative. Patients received either 180µg of subcutaneous Pegasys once per week for 48 weeks, or 6 million units of interferon alfa three times weekly for 12 weeks, followed by 3 million units three times weekly for a further 36 weeks. Nineteen individuals discontinued Pegasys due to adverse events compared with 27 discontinuations for adverse events in the conventional interferon group, whilst discontinuations due to lack of treatment responses numbered 13 in the Pegasys arm compared to 53 in the conventional interferon group. By intent-to-treat analysis, 69% of the Pegasys group had undetectable HCV RNA at week 72, compared with 19% of the conventional interferon group. The Pegasys group was also significantly more likely to experience a decline in liver enzyme levels, and to experience improvements in liver tissue inflammation. Dose modification due to neutropenia was required in 11% of the Pegasys group and 7% of the conventional interferon group. Responses in the Pegasys monotherapy group were similar to those seen in the studies of conventional interferon plus ribavirin in comparable populations and with comparable follow-up periods.

Zeuzem (2000) randomised 271 patients with HIV to receive either 90µg or 180µg of peginterferon once per week or 3 million units of conventional interferon alfa three times per week. Participants had already been diagnosed with liver cirrhosis or bridging fibrosis by liver biopsy. Of the 271 patients enrolled, 8% of the conventional interferon group, 15% of the 90µg peginterferon group and 30% of the 180µg peginterferon group (p = 0.001) had undetectable HCV RNA at week 72 by intent-to-treat analysis. Liver enzyme levels normalised in 15% of the conventional interferon group, 20% of the 90µg peginterferon group and 34% of the 180µg peginterferon group (p = 0.004). Improvements in liver inflammation were also significantly more common in the higher dose peginterferon group (54% vs 44% and 31%, p = 0.02) among 184 individuals who received a follow-up liver biopsy at week 72. There were no discontinuations due to neutropenia, and only two due to thrombocytopenia (in peginterferon 180µg recipients). The peginterferon 180µg group had the highest rate of discontinuations due to adverse events (13%), largely due to the flu-like symptoms commonly seen with interferon treatment. Histological (liver tissue) improvements occurred in some patients who had detectable HCV RNA at week 72, suggesting that interferon may still have benefit to the liver even if it does not control viraemia.

Reddy (2001) reported a randomised dosing study of peginterferon (45, 90, 180 and 270µg once weekly) vs conventional interferon alfa 2a (3 MIU 3 times weekly) in 159 HCV-infected people without HIV infection. Undetectable HCV viral load occurred in 10% (non-significant), 30% (p = 0.009), 36% (p = 0.0006), 29% (p = 0.004) of each group, respectively, at week 24, compared with a response rate of 3% in the conventional interferon group.

Conventional interferon / ribavirin in hepatitis C virus monoinfected people

Poynard (1998) randomised 832 individuals with HCV, who had not previously received interferon treatment, to one of the following treatment arms: conventional interferon alfa plus 1000-1200mg ribavirin per day for 48 weeks; interferon plus 1000-1200mg ribavirin per day for 24 weeks; or interferon plus placebo for 48 weeks. Undetectable HCV viral load at 24 weeks occurred in 43%, 35% and 19% of participants in the three treatment groups. Factors associated with response were: genotype 2 or 3; age under 40 years; minimal fibrosis; viral load under 2 million copies and female sex. The discontinuation rate at 48 weeks was 19% in the combination arm and 13% in the interferon arm.

McHutchison (1998) reported results of an international study of conventional interferon alfa plus ribavirin as initial treatment for HCV. 912 individuals were randomised to receive either interferon plus ribavirin or interferon alone for 24 or 48 weeks. After 24 and 48 weeks, undetectable HCV viral load was more likely among people taking dual therapy vs interferon alone (31% and 38% vs 6% and 13%). Liver function tests showed greater improvement at 6 months after 24 and 48 weeks in the dual therapy arm (57% and 61% vs 44% and 41%). Sub-group analyses showed that amongst those with HCV genotype 1, the response rate was higher if treated for 48 weeks rather than 24 weeks (28% vs 16%). Similarly, response rates in those with viral load above 2 million copies or cirrhosis were greater in those treated for 48 weeks. Up to 50% of responders in each treatment group did not achieve viral clearance until they had received more than 12 weeks of treatment. 21% of the dual therapy group discontinued treatment due to adverse events.

Reichard (1998) treated 100 treatment-naive HCV-infected people (HIV-negative) with conventional interferon plus ribavirin or interferon monotherapy for 24 weeks, with follow-up at 48 and 60 weeks. Intent-to-treat analysis found that 36% of those on the combination had undetectable HCV viral load at week 48. In comparison, 18% of the interferon arm had a sustained response. 1 year follow-up showed the superiority of the combination was sustained (42% vs 20%). However, those with HCV viral load below 3 million copies had similar rates of response in both treatment arms. Combination therapy had a superior effect for people with higher viral loads.

Marcellin (1997) reported long-term follow-up of 80 patients with HCV who had undetectable HCV RNA after 6 months of treatment with conventional interferon alfa. Follow-up was between 1-7.6 years (mean 4 years). During follow-up, HCV RNA remained undetectable in 96% of participants and 93% had normal ALT. No HCV RNA was detected from 27 liver tissue samples taken 1-5 years after treatment.

Interferon maintenance therapy and re-treatment of non-responders

Abergel (2004) studied 96 patients treated with conventional interferon alfa plus ribavirin and 64 treated with interferon alone. In sustained responders, the fibrosis progression rate decreased from 0.26 to 0.67 Metavir fibrosis units. Even amongst virological non-responders, the fibrosis rate decreased from 0.25 units before treatment to 0 units during therapy. 6 of 18 (33%) and 4 of 43 (9%) non-responders experienced reversion from cirrhosis (stage F4) to severe fibrosis (stage F3); however, no patients with cirrhosis had a two-point decrease in fibrosis score.

Research into hepatitis C virus treatment in co-infected individuals

Soriano (2004a) reported that reduction in HCV viral load of at least 2 log₁₀ by 12 weeks of treatment with peginterferon plus ribavirin predicts SVR in HIV/HCV co-infected patients, as it does in people with HCV alone. However, whilst no co-infected patients in this study who failed to achieve an early virological response at week 12 went on to achieve SVR, about 40% who did achieve an end-of-treatment response (ETR) later relapsed, a higher relapse rate than usually seen in HCV mono-infected individuals when using the currently recommended treatment durations (24 weeks for genotypes 2 or 3; 48 weeks for genotype 1 or 4). Soriano also found that it may be beneficial to extend therapy to 48 weeks for genotypes 2 or 3 and 72 weeks for genotypes 1 or 4.

Rodriguez-Torres (2004a) analysed data from a subset of nearly 300 subjects in the APRICOT study (described below), about 60% with genotype 1. Only two co-infected patients who failed to achieve an early virological response (at least a 2 log₁₀ reduction in HCV RNA by week 12) went on to achieve a sustained response. Thus, while a good response at week 4 or week 12 did not necessarily predict SVR (positive predictive value of 66% and 56%, respectively), a poor response at these time points did predict lack of SVR (negative predictive value of 88% at week 4 and 98% at week 12). The researchers concluded that 12 weeks was the best time to assess progress and make a decision about continuing therapy; 4 weeks seemed too early and could exclude some patients who would benefit from longer therapy.

Camino (2004) and Crespo (2004) presented early response data from the PRESCO study (described below). This study also validates a 2-log₁₀ HCV RNA decrease at week 12 as a predictor of sustained response.

Uberti-Foppa (2003) found that undergoing HCV treatment for six months before starting anti-HIV therapy reduced the proportion of HIV/HCV co-infected individuals who discontinued HAART due to drug-related liver toxicity.

Perez-Olmeda (2003) treated 58 HCV/HIV co-infected people (CD4 counts above 300 cells/mm³) with conventional interferon alfa plus ribavirn (n=27) or peginterferon plus ribavirn (n=31). Duration of treatment was 24 weeks for HCV genotypes 2 or 3 (n=36), and 48 weeks for HCV genotypes 1 or 4 (n=22). 19 (32.8%) of those treated relapsed (HCV became detectable again) after therapy was stopped. Rate of relapse was not affected by genotype, use of conventional interferon or peginterferon, CD4 count, HIV viral load or concurrent anti-HIV therapy.

Rodriguez-Torres (2004b) randomised 76 patients to receive Pegasys monotherapy or Pegasys plus ribavirin for 24 weeks. All responders continued on treatment for an additional 24 weeks. Paired biopsies were obtained to assess liver disease progression. Patients that achieved SVR had improvement in fibrosis grade, but patients who did not achieve a sustained response also experienced histological improvement. All non-responders who received at least 24 weeks of treatment in either arm experienced a decrease in fibrosis grade from 6.8 to 4.8, as well as a decrease in fibrosis progression rate.

Looking at histological response in the APRICOT trial, Lissen (2004) also found that a substantial proportion of patients experienced histological improvement even in the absence of SVR. Decreased fibrosis was seen in 69% of those who achieved SVR and in 43% of those who did not achieve a sustained response.

Cargnel (2005) studied 135 HIV/HCV co-infected patients in Italy, analysing factors that predicted treatment response. Patients received Peg-Intron plus ribavirin or Peg-Intron monotherapy for 48 weeks; 55 completed therapy. The best predictors of SVR were use of combination therapy, infection with HCV genotype 3, and early virological response at week 8. The positive predictive value of week 8 EVR was 65%, while the negative predictive value was 86%.

Peginterferon / ribavirin in co-infected people

Torriani (2004) reported that Pegasys plus ribavirin produced higher SVR rates among HIV/HCV co-infected people than conventional interferon plus ribavirin or Pegasys monotherapy. The APRICOT study included 868 co-infected participants in 19 countries, all with stable HIV disease with or without antiretroviral therapy. Participants were randomly assigned to one of three arms: 3 million IU conventional interferon alfa 2a three times weekly plus 800 µg ribavirin daily (285 subjects); 180µg Pegasys once weekly plus placebo (286 subjects); or the same doses of Pegasys plus ribavirin (289 subjects) for 48 weeks. Baseline characteristics were similar in the three arms. About 60% had genotype 1;, 5% had genotype 2; 27% had genotype C; and 7% had genotype 4 (40% total non-1). Mean CD4 count was above 500 cells/mm³. About 16% had bridging fibrosis or cirrhosis. SVR was defined as HCV RNA viral load less than 50 IU/ml at the end of a 24 week post-treatment follow-up period (72 weeks total). Overall, 40% of co-infected patients treated with Pegasys/ribavirin achieved SVR, compared with 20% of those receiving Pegasys monotherapy and 12% of those receiving conventional interferon/ribavirin; corresponding ETR rates were 49%, 33%, and 14%. Among subjects with genotypes 2 or 3, the respective SVR rates were 62%, 36%, and 20%, while in genotype 1 patients SVR was seen in 29%, 14%, and 7%, respectively. The higher SVR rates in the Pegasys/ribavirin arm compared with the Pegasys monotherapy arm suggest that ribavirin may play an important role in preventing HCV relapse. HCV therapy had no negative impact on HIV disease progression, and may even have had a small positive effect. In terms of safety and tolerability, 39% in the conventional interferon/ribavirin arm, 31% in the Pegasys monotherapy group, and 25% in the Pegasys/ribavirin arm discontinued treatment. Adverse side-effects or laboratory abnormalities were seen in about 15% in all three arms. Serious adverse event rates were 5%, 10%, and 8%, respectively; neutropenia was more common in the two Pegasys arms (about 12%).

Chung (2004) presented final 72 week results of ACTG A5071, a randomised study looking at the safety and efficacy of conventional interferon/ribavirin vs Pegasys/ribavirin in patients with HIV/HCV co-infection. Baseline characteristics were similar in both arms. About 80% were men, about half were African-American, mean CD4 count was above 400 cells/mm³, and 75% had HCV genotype 1. The median fibrosis score was 2.0 and the median HAI score was 5.0. Overall, at week 24, virological response (HCV RNA viral load less than 60 IU/ml) was seen in 44% of patients in the Pegasys/ribavirin arm, compared with 15% in the conventional interferon/ribavirin arm. After 48 weeks of therapy, 41% and 12%, respectively, showed an end-of-treatment response. Within the Pegasys/ribavirin arm, 80% with genotypes other than 1 showed an end-of-treatment response, compared with 29% of genotype 1 patients. By 72 weeks, overall SVR rates were 27% in the Pegasys/ribavirin arm and 12% in the conventional interferon/ribavirin arm. Notably, while the ETR and SVR rates were the same in the conventional interferon/ribavirin arm, the response rate declined dramatically from week 48 to week 72 in the Pegasys/ribavirin arm. By genotype, in the Pegasys/ribavirin arm, SVR was 73% in non-1 patients and just 14% in genotype 1 patients. In the conventional interferon/ribavirin arm, 33% of the non-1 genotype patients achieved SVR compared with just 6% of those with genotype 1. No patient who failed to achieve at least a 2-log₁₀ reduction in HCV RNA by week 12 went on to achieve SVR (negative predictive value of 100%). Besides non-1 genotype, SVR was also associated with absence of current or past injecting drug use and with detectable HIV viral load at study entry. Histological response was seen in 52% of virological responders and 36% of virological non-responders who underwent biopsy. Incidence of side-effects (flu-like symptoms, depression) was similar in both arms, and 12% in both arms prematurely discontinued therapy.

In the 412-person multi-centre French ANRS HC02-RIBAVIC study, Carrat (2004) compared conventional interferon and Peg-Intron. Participants had abnormal liver histology and stable HIV disease with or without antiretroviral therapy. Participants were randomly assigned to receive 3 million IU conventional interferon three times weekly (207 subjects) or 1.5 µg/kg Peg-Intron once weekly (205 subjects) plus 800mg ribavirin daily. Baseline characteristics were similar in both arms. More than half (58%) had HCV genotypes 1 or 4; 34% had genotype 3; and 8% had other genotypes. Stage F3-F4 fibrosis was detected in about 40%, of whom 17% had persistently normal ALT. At 72 weeks, 19% of subjects receiving conventional interferon versus 27% of those on Peg-Intron achieved SVR using an intent-to-treat analysis. Considering only those who stayed on therapy (as-treated analysis), the respective response rates were 12 % vs 20% at week 4; 34% vs 41% at week 12; 41% vs 54% at week 24; 34% vs 52% at week 48; and 26% vs 35% at week 72. Among patients with genotype 1, SVR was seen in 5% of the conventional interferon arm and 15% of the Peg-Intron arm; among non-1 patients, the corresponding rates were 40% and 45%. Early virological response at week 4 strongly predicted later SVR, while lack of response at week 12 predicted failure to achieve SVR. Some improvement in liver histology was seen in responders. In terms of tolerability, about 40% discontinued therapy in both arms and about 30% experienced severe adverse events (including six cases of symptomatic hyperlactatemia). Significant decreases were seen in haemoglobin (a marker for red blood cells) and platelets, but not white blood cells.

Laguno (2004) studied 95 HIV/HCV co-infected patients (68% male; mean age 40; 80% IDU history) receiving either conventional interferon (n=43) or Peg-Intron (n=52), both plus daily ribavirin. At baseline all had well-controlled HIV (94% on HAART; mean CD4 count 570 cells/mm³), detectable HCV RNA, elevated ALT and biopsy-proven liver fibrosis. Subjects with genotypes 1 or 4 were treated for 48 weeks, and those with genotypes 2 or 3 for 24 weeks. The overall SVR rates were 44% for a Peg-Intron and 21% for conventional interferon (p = 0.017). Among those with genotypes 2 or 3, the SVR rates were 53% and 47%, respectively (p = 0.730). The corresponding rates for genotypes 1 or 4 were 38% and 7% (p = 0.007). In addition, individuals with a low HCV RNA (below 80,000 IU/ml) were more likely to achieve SVR (49% versus 22%; p = 0.007). Side-effect rates were high, but not significantly different in the two arms. 37% in the Peg-Intron arm and 51% in the interferon arm were diagnosed with depression; corresponding rates for anaemia were 31% and 23%. 9 cases of mitochondrial toxicity (all but one asymptomatic) were detected. 9 subjects in the Peg-Intron arm and six in the conventional interferon arm discontinued due to side-effects. CD4 cell counts, but not percentages, dropped during treatment (a consequence of the general white blood cell loss due to interferon), and HIV viral load did not change from baseline.

The ongoing, non-randomised Spanish PRESCO trial includes about 350 HIV/HCV co-infected participants (52% genotype 1, 1% genotype 2, 37% genotype 3, 10 genotype 4). The trial is testing Pegasys plus relatively high, weight-adjusted doses of ribavirin in an attempt to reduce the rate of HCV relapse. All participants have well-controlled HIV (CD4 count at least 300 cells/mm³) and elevated ALT. Interim data (Nunez 2004) showed that after 24 weeks of treatment, 63% had undetectable HCV RNA (50% for genotype 1; 85% for genotypes 2 or 3; 44% for genotype 4). As-treated response rates were higher (71% overall and 61% for genotype 1), since this analysis excluded the approximately 20% of subjects who discontinued therapy early due to adverse events or other causes.

Peginterferon monotherapy in co-infected people

Hoffman-Terry (2002) treated 154 HIV/HCV co-infected people with Pegasys for 48 weeks. Those who failed to achieve at least a 2 log₁₀ decline in HCV viral load by week 12 were then randomised to receive ribavirin or placebo. Preliminary data on 119 people who completed 24 weeks on treatment showed that 39% had achieved a virological response by week 12. However, by week 24, 59% of these 'responders' had HCV viral rebound.

Treatment of acute HCV infection

Jaeckel (2001) treated 44 patients presumed to have acute hepatitis (known or suspected HCV exposure, documented seroconversion, and/or sudden ALT increases). All had detectable HCV viral load and elevated ALT; 68% had jaundice. All received 5 million units of conventional interferon daily for four weeks, then three times weekly for an additional 20 weeks. The average time from presumed infection to the initiation of treatment was 89 days. At the end of a 24 week post-treatment follow-up period, 43 of the 44 (98%) had undetectable HCV viral load and normal ALT (including one patient who discontinued after 12 weeks). This trial did not include a control group, and some unknown proportion of subjects would have spontaneously cleared HCV without treatment.

As discussed above, numerous HIV-positive gay men in London have recently been identified soon after HCV seroconversion. 20 men were identified at the Royal Free Hospital, half through routine liver function tests and one-third through STI screening (Aizen 2003). Over half of the men (55%) were on HAART at the time of their hepatitis C diagnosis. All were offered treatment within six months of documented seroconversion. Preliminary results suggested a high rate of spontaneous HCV clearance, with treatment in acute infection being a viable option (Lascar 2003). Ten patients received treatment with peginterferon with the addition of ribavirin for 12 weeks if they were still HCV RNA positive. Only two of the first eight men to complete the treatment course responded to therapy. Six of those for whom treatment failed had genotype 1 infection. Bhagani (2004) offered treatment with peginterferon plus ribavirin to acutely infected men after 12 weeks of persistently testing HCV RNA positive. Seventeen agreed to start treatment, thirteen of whom were infected with genotype 1, three with genotype 3, and one with genotype 4. After 12 weeks of therapy, 11 of 15 (73%) achieved an early virological response (either undetectable HCV RNA or at least a 2-log₁₀ HCV viral load decrease). After 24 weeks, 6 of 9 men (66%) achieved an undetectable HCV viral load. After 48 weeks, 5 of 7 men (71%) had undetectable HCV RNA. 30% withdrew from therapy, reflecting the difficulty of tolerating interferon due to its major side-effect, severe depression.

Gilleece (2005) reported on treatment of 50 gay HIV-positive men diagnosed with acute HCV at Chelsea and Westminster Hospital in London between 1997-2003. 12 patients (24%)spontaneously converted to HCV PCR negative by week 12. Spontaneous clearance was associated with a CD4 cell count above 500 cells/mm³ and lower HCV viral load (p = 0.04). 27 individuals who still had detectable HCV RNA at 12 weeks were treated with peginterferon/ribavirin for 24 weeks. 16 of 27 (59%) achieved SVR, which was associated with lower mean peak ALT (p < 0.001).

Santantonio (2005) studied 16 patients with acute HCV who still had detectable HCV viral load 12 weeks after initial presentation. 15 of 16 (94%) patients achieved undetectable HCV RNA at the end of a 24 week course of Peg-Intron monotherapy (1.5µg/kg once weekly). ALT normalised in 14 of 16 (88%). Virologic and biochemical response was sustained after one year of follow-up. The researchers said their data support waiting 12 weeks to check for spontaneous HCV clearance before beginning therapy.

Highly active antiretroviral therapy in co-infected people

Flexman (1999) retrospectively studied HCV-related hepatitis in 133 people treated with HAART. Three (2%) reported acute hepatitis related to immune restoration following treatment with anti-HIV drugs.

Puoti (2003) studied liver function among 755 HIV-infected individuals who commenced HAART between January 1997-June 1998. 26 people developed life-threatening hepatotoxicity - defined as liver failure, or an increase in liver enzymes 10 times baseline levels, or 5 times baseline if baseline elevations already existed. The only variables found to be significantly associated with hepatotoxicity were being HCV-positive and having elevated ALT levels. Among the 16 patients who received biopsies, all showed exacerbation of liver disease. Seven individuals died as a result of hepatotoxicity. CD4 count was below 200 cells/mm³ in all patients who died, and in 42% of those who survived. HAART was restarted in 19 individuals following interferon treatment or a break from treatment, with no further cases of hepatotoxicity.

Mendel (1998) studied HIV and HCV viral load in 43 people who started antiretroviral therapy. Although there was an increase in HCV viral load at 3 months, this had declined by 6 months. No correlation was found between HIV and HCV viral load, and HAART was found to have no effect on HCV.

Ragni (1999) studied 21 men co-infected with HIV and HCV. HCV RNA was measured before participants commenced HAART, and at weeks 24, 48 and 96. A significant HCV viral load increase was detected at week 48 (p = 0.03) and at week 96 (p < 0.001). HCV RNA increases were associated with CD4 cell increases, HIV viral load decreases and unchanged liver enzyme test results.

Kottilil (2002) reported on 17 HIV/HCV co-infected patients who suppressed HIV viral load to below 200 copies/ml after starting an anti-HIV regimen. The median baseline CD4 count was 241 cells/mm³, mean HIV viral load was 4.75 log copies/ml and mean HCV viral load was 6.36 log copies/ml. 14 of 17 patients had increases in their HCV viral load at the time of HIV suppression. The average rise in HCV viral load was 5.61 log₁₀.

Chung (2002) analysed HCV RNA changes and liver toxicity in co-infected patients enrolling in AIDS Clinical Trials Group (ACTG) studies. Of 60 co-infected patients who took HAART for at least 16 weeks, HCV viral load increased by 0.35 log₁₀ at week 16 and 0.43 log₁₀ at week 48. For those with a baseline CD4 count below 350 cells/mm³, the increases at weeks 16 and 48 were 0.43 and 0.59 log₁₀, respectively; for those with a baseline CD4 count above 350 cells/mm³, average viral load increases were 0.26 log₁₀ and 0.1 log₁₀. Extreme ALT elevations occurred in 3.3% of co-infected HAART recipients. Comparing co-infected patients and those with HIV mono-infection, CD4 count increases were not affected by HCV infection.

Puoti (2000) reported a study of 12 HIV/HCV co-infected people commencing combination antiretroviral therapy. HIV RNA declined rapidly, CD4 count increased slowly, and CD8 count was stable, as was liver function. HCV RNA underwent a transient increase at days 14 and 21, and 4 people had an increase in their ALT levels. The authors argued that these results demonstrated virus-virus interference.

Rutschmann (1998) studied 28 co-infected people who commenced antiretroviral therapy. HCV viral load rose during the first 6 weeks; however, after 2 years on the study, anti-HIV therapy had no effect on HCV viral load or liver function tests. The researchers concluded that protease inhibitors (PIs) are safe and effective for people with both HIV and HCV.

Perez-Cano (1998) studied HIV and HCV viral load in 30 co-infected people treated with antiretroviral therapy for HIV. HIV viral load fell and CD4 levels increased as expected. At 3 months, HCV viral load had risen. However it later fell, and by 12 months was below baseline. The transient increase in HCV viral load was attributed to immune reconstitution.

Benhamou (2001) conducted a retrospective analysis of 182 HIV/HCV co-infected people treated with PI-based therapy. Comparing those taking PIs with those who never took this class of drugs, the rates of cirrhosis at 5, 15 and 25 years were: 2 vs 5%; 5 vs 18%, and 9 vs 27% (p = 0.0006), respectively. Independent predictors of cirrhosis were: absence of PI therapy, heavy alcohol consumption, a CD4 count below 200 cells/mm³ and more than 20 years with HCV infection.

Mariné­‚arjoan (2004) concluded that early HAART may slow liver disease progression in co-infected individuals, based on data from a case-control study that included 116 HIV/HCV co-infected patients and 235 patients infected with HCV alone matched by sex, age at infection and duration of infection. More co-infected patients had a history of injecting drug use and/or consumed more than 50g of alcohol a day. 91 (78%) co-infected patients receiving HAART at the time of liver biopsy were stratified by fibrosis severity: none-to-moderate (F0-F2) and severe (F3-F4). Among co-infected patients, F3-F4 fibrosis scores were significantly more frequent (26% vs. 7%; p < 0.001) and the median rate of fibrosis progression was significantly higher (0.106 vs. 0.071 fibrosis unit per year; p < 0.001), compared with the HCV mono-infected patients. Patients with moderate fibrosis had a significantly shorter interval between presumed date of HCV infection to initiation of HAART (adjusted OR 1.16; p = 0.03) and longer duration of HAART use (adjusted OR 0.96; p = 0.06) than patients with severe fibrosis. After adjusting for duration of disease, age at HCV infection and CD4 cell count, fibrosis progressed on average significantly more slowly in patients who had been on HAART longer (p = 0.02) and in those who had a shorter interval between presumed HCV infection and HAART initiation (p = 0.02).

Martin-Carbonero (2004) conducted a retrospective multi-centre study of 914 HIV/HCV co-infected people (56% with genotype 1; 32% with genotype 3). 55% of patients were on HAART and 54% had undetectable HIV viral load (below 200 copies/ml). The median CD4 cell count was 480 cells/mm³, and only 9% had CD4 cell counts below 200 cells/mm³. 47% had severe fibrosis (stages F3-F4). In a multivariate analysis, factors associated with severe fibrosis were age over 35 years (p < 0.001), heavy alcohol consumption (p = 0.01), and CD4 count below 500 cells/mm³ (p = 0.02) Male sex, duration of HCV infection, HCV genotype and HCV viral load did not predict severe fibrosis. HAART was associated with neither improved nor worsened fibrosis.

Qurishi (2003) examined liver-related deaths in a cohort of 285 HIV/HC co-infected patients as part of an observational study that ran from 1990 to 2002. In total, 93 patients were treated with HAART, 55 individuals received dual therapy or monotherapy, and 137 patients took no anti-HIV drugs. None of the patients received anti-HCV therapy. The rate of liver-related mortality was significantly lower in individuals who received HAART (2 patient deaths, 0.45 per 100 PY; p < 0.001) than in patients who took one or two anti-HIV drugs (5 patient deaths, 0.69 per 100 PY; p < 0.001), or in individuals who received no anti-HIV treatment (18 patient deaths, 1.70 per 100 PY). CD4 cell gain after the initiation of HAART was associated with a significant decrease in the risk of dying from liver-related causes (p < 0.001), whilst each one year increase in age (p = 0.001) and an increase in bilirubin (p < 0.001) were significantly associated with an increased risk of liver-related mortality. The investigators noted that HCV viral loads increased significantly in all three groups of patients, regardless of their anti-HIV treatment histories (p < 0.001), but that the increase was particularly marked in patients who received HAART. Severe drug-related liver toxicity occurred in 13.8% of patients taking HAART, but resulted in no patient deaths.

Sulkowski (2001) reported a poor response to anti-HIV treatment amongst a largely African American HIV/HCV co-infected cohort. However, multivariate analysis suggested this was due to low CD4 counts, poor access to healthcare or ineffective treatment, rather than HCV co-infection per se. However, timing of HCV infection may also explain differences in HCV disease progression in haemophiliacs and IDU co-infected cohorts.

Liver toxicity, liver disease and highly active antiretroviral therapy

Monforte (2001) conducted a prospective study of 1255 HIV-infected individuals starting HAART. 7.2% had active hepatitis B (HBsAg-positive) and 46.5% had hepatitis C antibodies. After initiation of anti-HIV treatment, 61 people developed liver toxicity (ALT above 200 IU/L) with a probability of 7.9% at 2 years. Factors that predicted liver toxicity were: higher baseline ALT levels; hepatitis C infection or infection with both hepatitis B and C; and previous anti-HIV treatment. Use of d4T was associated with a lower risk of toxicity compared with AZT.

Aceti (2002) conducted a retrospective study of 1,325 HIV-infected people treated with anti-HIV drugs for at least 6 months. Severe liver toxicity at 6 months was associated with viral hepatitis co-infection and use of ritonavir (Norvir). At 12 months, HCV-infected people who did not take anti-HIV treatment had a higher rate of liver toxicity compared with those who responded to treatment and showed decreased ALT levels.

Nunez (2001) conducted a retrospective review of liver enzyme elevations in 222 HIV-infected people commencing HAART at one clinical centre. 38% had hepatitis C, 5% had hepatitis B and 2% had hepatitis D. 21 of 222 developed severe liver enzyme elevations (5-fold above normal or 3.5-fold above already elevated baseline levels). Multivariate analysis identified heavy alcohol use (RR 5.87), co-infection with HCV (RR 3.99) and older age (RR 1.11) as independent risk factors.

Sulkowski (2000) compared hepatotoxicity associated with antiretroviral therapy among 298 HIV-infected individuals, of whom 52% had HCV and 2.7% had HBV. Combinations containing ritonavir were associated with increased liver toxicity. Chronic viral hepatitis was also associated with an increased risk of toxicity, with 12% of those with HBV or HCV reporting toxicity.

Reisler (2003) found that among a cohort of 2947 HIV-positive individuals receiving antiretroviral therapy from 1996 to 2001, co-infection with HCV significantly increased the risk of severe (grade 4) liver-related side-effects compared with the rate in people with HIV alone. Nevertheless, severe hepatotoxicity was relatively uncommon even among the co-infected individuals.

Aranzabal (2005) reported on a study of 107 HIV/HCV co-infected patients who underwent liver biopsy. Overall, 25% of patients experienced liver toxicity (defined by elevated ALT), but rates differed based on liver disease severity. 38% of patients with severe fibrosis (stage F3 or F4) experienced hepatotoxicity, compared with 15% of subjects with mild-to-moderate liver damage (F1 or F2). Toxicity risk was not associated with duration of HCV infection, HCV viral load, or HCV genotype. Amongst people with severe fibrosis, nevirapine (Viramune) or efavirenz (Sustiva / Stocrin) were more likely than non-NNRTIs to cause liver problems, but this difference was not apparent in the mild-to-moderate fibrosis group.

den Brinker (2000) evaluated 394 HIV-positive patients who started a PI-containing regimen between July 1996-February 1998. 18% developed liver enzyme elevations to more than five times the upper limit of normal. Cases occurred after a median of 25 weeks on the HAART regimen, and people with either hepatitis C (OR 2.46) and hepatitis B co-infection (OR 2.78) were significantly more likely to experience liver enzyme elevations. The co-infected patients also experienced the most rapid onset of liver enzyme elevation (90% of all cases occurred between 12 and 42 weeks after starting HAART). Modification of treatment did not affect the likelihood that liver enzyme levels would subsequently decline in those who experienced elevations, but those who did modify treatment generally experienced greater declines in ALT and AST levels.

Dieterich (2003) retrospectively reviewed the cases of 1094 HIV/HCV co-infected patients to assess hepatotoxicity related to PIs. 77% were male, median baseline CD4 cell count was 240 cells/mm³ and median HIV viral load was 4.1 log copies. 448 patients used nelfinavir (Viracept) whilst 377 used indinavir (Crixivan), 185 used saquinavir (Invirase / Fortovase), 151 used ritonavir and 8 used amprenavir (Agenerase). 110 were taking more than one PI. Grade 3 (5 times the upper limit of normal) or grade 4 (10 times the upper limit of normal) ALT elevations were observed in 4% of patients who took nelfinavir, 7% on indinavir, 8% on ritonavir, and 11% on saquinavir. Corresponding rates of grade 4 AST elevation were 3% 7%, 3%, and 5%. The researchers concluded that nelfinavir is safe for use in co-infected individuals.

Chihrin (2004) conducted a retrospective chart review of 202 co-infected individuals starting HAART. Most were men (93%), the median age was 43, 54% were HIV/HCV co-infected, 40% had HIV/HBV and 6% had HIVHCV/HBV. 33% (n=66) started on a PI-containing regimen and 13% (n=26) started on a NNRTI-containing regimen. 42 individuals (21%) had a grade 3 or 4 ALT elevation at least once in follow-up. In a multivariate analysis, only older age and use of lopinavir/ritonavir (Kaletra) predicted severe ALT elevations. Meravaglia (2004) studied 782 HIV-positive patients to assess the association between Kaletra and liver enzyme elevations. 34% had elevated baseline ALT/AST, and this was associated with HCV or HBV co-infection. During follow-up (median 349 days), 9% overall developed elevated liver enzymes while taking Kaletra. By co-infection status, 16% of HCV or HBV co-infected individuals experience liver enzyme increases compared with 3% of HIV mono-infected patients. Three-quarters of the elevations were moderate (grade 2), but 20% developed grade 3 and 5% experienced grade 4 toxicity. 13 patients stopped Kaletra due to elevated ALT/AST. In a multivariate analysis, HCV or HBV co-infection (p < 0.001), younger age (p = 0.005), elevated baseline liver enzymes p = 0.005), and use of efavirenz (p = 0.008) predicted enzyme elevations while using Kaletra. The researchers recommended that HIV-positive individuals should be tested for hepatitis B and C before starting HAART.

Bonfanti (2005) analysed data from the SCOLTA project, an online reporting system for antiretroviral side-effects, to determine the incidence of severe (grade 3 and 4) liver-related events in 755 patients using Kaletra; 44% were co-infected with HBV or HCV. After a mean follow-up period of 17 months, the overall combined incidence of all severe adverse events was 11 per 100 PY, with a lower rate amongst treatment-naive individuals (7 per 100 PY) than treatment-experienced patients (12 per 100 PY). Looking only at liver-related side-effects, the incidence was just 0.59 per 100 PY. One treatment-naive and four treatment-experienced patients experienced severe liver toxicity, four of whom were HCV co-infected. Although two cases developed shortly after starting the drug, the other three occurred after a year on therapy.

Mací¡³ (2004) reported that nevirapine appears to be associated with more rapid fibrosis progression. The study included 152 HIV/HCV co-infected individuals with a known or estimated date of HCV infection and available liver biopsy results. 14 of 46 patients (30%) taking nevirapine had evidence of advanced fibrosis (stages F3-F4), leading the researchers to extrapolate a progression rate greater than 0.2. However, it is not certain that the more severe fibrosis in this group was attributable to the drug.

Sulkowski (2002a) prospectively assessed liver function among 256 people commencing nevirapine and 312 starting efavirenz. 43% had HCV and 7.7% had HBV co-infection. Severe liver toxicity, defined as a grade 3 or 4 change in ALT or AST levels, occurred in 15.6% of nevirapine recipients and 8% of the efavirenz recipients. Only one-third and one-half of the nevirapine and efavirenz-related adverse events, respectively, occurred within 3 months of starting treatment. 69% of liver toxicity occurred among people co-infected with viral hepatitis and 82% were taking concurrent PIs. However, 84% of co-infected people did not develop liver toxicity. According to Sulkowski, HIV/HCV-coinfected people can do just as well on antiretroviral therapy as those with HCV alone.

Dieterich (2002) detected severe (grade 3 or 4) ALT or AST elevations in just 1.1% of a cohort of 272 HIV positive subjects (mostly white men) taking NNRTIs, typically nevirapine or efavirenz. None of these severe events occurred in individuals coinfected with HCV or HBV. Hoffman-Terry (2000) conducted a retrospective chart review of 30 individuals on PI-containing HAART and 22 on NNRTI-containing HAART. AST levels, ALT levels and total bilirubin were similar for the two groups at baseline, 3 and 6 months, and no significant changes occurred in either group. One individual in each group ceased treatment due to hepatotoxicity.

Cooper (2002) conducted a retrospective review of 66 HIV/HCV co-infected people to determine factors associated with cessation of anti-HIV treatment due to liver toxicity. Dual versus single PI treatment did not impact significantly on rates of high ALT levels (19 vs 26%, respectively), hyperbilirubinaemia (30 vs 38%, respectively) or discontinuation. Ritonavir also had no impact on liver toxicity in this study.

Metabolic disorders in co-infected people

Mehta (2003) studied 1230 HIV-positive individuals who started their first HAART regimen between January 1996 and May 2002; 53% were co-infected with HCV. At baseline, co-infected patients tended to be older, African-American, have a history of injecting drug use, and have abnormal liver function test. At baseline, glucose levels were lower in patients treated with PIs than those taking other regimens. In 1128 HAART-treated patients, new-onset hyperglycaemia was observed in a total of 47 individuals, an incidence of 3.5 per 100 PY of follow-up (95% CI, 2.6-4.6). Diabetes occurred in 37 patients, and 37 of the people with new-onset hyperglycaemia were co-infected with hepatitis C. The incidence of hyperglycaemia was 4.9 per 100 PY of follow-up (95% CI, 3.4-7.1) in the HIV/HCV co- infected group compared to 2.3 per 100 PY in the HIV only group (95% CI, 1.4-3.7). Individuals taking a PI were significantly more likely to be diagnosed with new-onset hyperglycaemia, with an incidence rate of 4.1 per 100 PY (95% CI, 3.0-5.6) compared to 1.5 per 100 PAY (95% CI, 0.6-4.1) for patients on NNRTIs and 2.8 per 100 PY (95% CI, 1.0-7.4) for those taking both a PI and a NNRTI.

Hoffman-Terry (2001) conducted a retrospective study of 40 co-infected people and 40 matched controls infected with HIV alone. 31 co-infected and 27 HIV patients were taking PI treatment. At baseline, average CD4 count was higher in the co-infected group (279 vs 394 cells/mm³) and average blood sugar and ALT levels were higher (93.4 vs 86.1 mg/dl and 85.4 vs 40.2 U/L). There was no difference in viral load between the two groups. At 9 months, 5 of the co-infected and none of the HIV mono-infected patients developed hyperglycemia. 4 of 5 were also diagnosed with diabetes. All who did so had body mass indexes above 25 and developed high sugar levels with 6 months of starting PI treatment.

Duong (2001) reported that co-infection with HCV increases the risk of metabolic abnormalities such as insulin resistance associated with antiretroviral therapy, based on a cross-sectional study of 28 co-infected people on HAART, 76 HIV-infected people and 121 controls with HCV alone. Both HCV-positive groups had significantly higher insulin resistance than the HIV mono-infected group, while body mass index was lower in co-infected group. Lipoatrophy was greater in the co-infected group (48% vs 19.5%), but cholesterol levels were lower in the co-infected group.

Patroni (2002) retrospectively analysed 282 HIV-infected patients who had been taking their first anti-HIV regimen for at least 4 months. Average follow-up time was 18.5 months. During this time, the median cholesterol level rose significantly, independent of exposure to a PI vs a NNRTI, and independent of d4T exposure. Univariate analysis showed HCV infection protected against elevated cholesterol, while high baseline lipids and the size of the CD4 cell increase were linked to the cholesterol increase during HAART. Multivariate analysis showed cholesterol increase was associated with baseline lipid levels, while HCV co-infection was associated with a smaller cholesterol increase.

Polgreen (2004) studied 881 HIV-positive individuals in the US VACS cohort. 700 had had an HCV antibody test performed and 300 (42.8%) were HCV co-infected. The HCV-positive and HCV-negative patients had similar ages (about 48 years), but HCV-positive patients were more likely to be non-white (80% versus 20%) and to have a history of injecting drug use (62% versus 13%). In a multivariate analysis, being HCV co-infected was independently associated with lower total cholesterol (p = 0.001) and lower LDL (bad) cholesterol (p < 0.001), but not with lower HDL (good) cholesterol or triglycerides. The researchers suggested that HCV might enhance lipid uptake by cells or might impair cholesterol synthesis in the liver.

Side-effects of hepatitis C virus treatments

Schaefer (2005) reported on a study of the effectiveness of pre-emptive antidepressant therapy in preventing depression during HCV treatment. Before starting peginterferon plus ribavirin, 14 HCV-positive patients with pre-existing psychiatric conditions received 20mg daily of the SSRI antidepressant citalopram (Celexa), which they continued to take during anti-HCV therapy. Subjects who received citalopram were significantly less likely to develop major depression during the first six months of anti-HCV treatment, compared with 11 HCV-positive patients with psychiatric disorders who did not receive the antidepressant and 11 HCV-positive patients without psychiatric risk factors (14%, 64%, and 55%, respectively). Among those who did develop depression while taking interferon, citalopram was associated with improvement.

Kraus (2005) studied patients retreated with peginterferon plus ribavirin after previous unsuccessful therapy attempts. 8 patients who had developed major depression during their prior treatment were given prophylactic SSRIs, while 9 were retreated without preventive antidepressants. The Hospital Anxiety and Depression Scale was used to assess mood before, during, and after therapy. All patients who received prophylactic SSRIs completed HCV therapy and depression scores were significantly lower compared with non-SSRI patients, who experienced depression levels similar to those during their first course of HCV treatment.

Farel (2004) found that peginterferon/ribavirin can lead to serious eye problems in some HIV/HCV co-infected individuals. The 23 patients in the analysis had received Peg-Intron plus ribavirin for at least 12 weeks. 35% developed some type of ophthalmological pathology during anti-HCV therapy, including six patients who developed cotton wool spots (indicative of retinal haemorrhage) and two who developed cataracts. Two patients experienced decreases in red-green colour vision. One patient discontinued hepatitis C treatment and within ten weeks normal colour vision had returned; in the other patient colour vision returned to normal while still on therapy. There were no differences in CD4 cell count, HIV viral load, or hepatitis C viral load between the patients who did and did not develop vision problems.

Looking at data from the APRICOT study, Brau (2004) found that HIV/HCV co-infected and HCV mono-infected patients receiving peginterferon plus ribavirin typically had similar rates of side-effects leading to early treatment discontinuation, but adverse events tended to be more severe adverse in the co-infected individuals.

Moreno (2004) conducted a prospective study of 56 HIV/HCV co-infected subjects (80% male, 95% history of IDU). 88% were on HAART and most had well-controlled HIV disease (although 18% had an AIDS diagnosis). 59% had genotype 1 HCV, median duration of HCV infection was 20 years, and 43% had evidence of cirrhosis. At the end of HCV treatment (mostly with peginterferon plus ribavirin), 29% of patients experienced an end0of-treatment response, but only 14% achieved SVR. 29% discontinued treatment due to side-effects; one-fifth of the subjects were hospitalized and two (4%) died during therapy. Blood cell deficiencies were particularly common: neutropenia, 73%; thrombocytopenia, 61%; anemia, 38%. The poor results seen in this study may be related to the fact that a substantial number of the subjects had a history o