Skip directly to content

Diabolic Duo

tagline May 2001

End Stage Liver Disease Emerges As Leading Cause of Death in PWAs with Hepatitis C


“While many people with HIV are living longer, more productive lives,” Michael Marco explains,” a good many who are also co-infected with the hepatitis C virus are winding up in the hospital and dying of end-stage liver disease. As an activist who eschews hepatitis C media hype (e.g., New York magazine’s outrageous 2000 cover story, “Hepatitis C, The Silent Killer”) and alarmist co-infection data from misleading, outdated and small retrospective case studies, I am here to say that we have a big problem on our hands.”

End stage liver disease (ESLD) (characterized as conditions, including encephalopathy, ascites, jaundice, gastrointestinal bleeding, hepatorenal syndrome or peritonitis) in PWAs with HCV appears to be a leading cause of hospital admissions and death according to a number of U.S. and European data presentations at the 8th Conference on Retroviruses and Opportunistic Infections. Conflicting data from the conference (and recent journal articles) have also reinvigorated the old debate of whether HCV infection independently leads to accelerated disease progression and death in PWAs.

End stage liver disease as the leading cause of death in HIV clinical cohorts

Epidemiology studies from various HIV cohorts in the U.S. report coinfection in 25-75% of individuals. The higher incidence rates are often found in major metropolitan inner-city clinics treating persons with a history of intravenous drug use. Many HIV clinics in southern Europe have a co-infection exceeding 50%. Co-infected HAART-treated individuals who are living longer and entering their late 40s and 50s are having their HCV disease go from asymptomatic to symptomatic and, worse yet, from symptomatic to ESLD. For 20% of HCV-infected individuals—especially those co-infected with a nadir CD4 cell count of <200 cells/mm3—progression to ESLD is only a matter of time.

Researchers in Madrid documented a steep rise in the number of hospital admissions and deaths caused by HCV- and HBV-related ESLD between 1995 and 2000. Of the 843 HIV-positive individuals seen at the institution over the five year period, 46% had viral hepatitis; about 30% of whom had HCV. Hospitalization due to ESLD rose from 5.2% in 1996 to 8.4% in 2000. In fact, 43% of all deaths among the HIV-positive individuals were caused by ESLD.

HCV ESLD was again the leading cause of death in a French HIV cohort. French researchers reported that there were 105 deaths recorded between 1998-1999 in their 2,200 person cohort, and HCV ESLD was the cause of 29% of them. Similarly, investigators from northern Italy found that HCV ESLD was the leading cause of death in their HIV-positive individuals in 1998 and 1999.

This surge in deaths due to ESLD in co-infected individuals is not confined to southern Europe. At Cook County Hospital in Chicago where over 50% of the HIV-positive individuals have HCV, researchers there found ESLD to be the cause of death in 35% of the individuals between January 1998 and September 2000. Overall, ESLD was the second leading cause of death after sepsis (38%).

Does HCV infection negatively affect HIV disease?

A number of studies presented at the Retrovirus meeting attempted to answer the controversial co-infection question: Does HCV have a negative impact on HIV disease progression and survival? The best data came from the well-established European HIV cohorts and the Johns Hopkins group.

Studies concluding “Yes”

Researchers from Montreal’s McGill University presented results from a 182 HIV-positive patient retrospective chart review suggesting that HCV co-infection was associated with faster progression to death and increased hospitalization. 78 HIV/HCV co-infected individuals were compared with 104 HIV-monoinfected individuals seen at the McGill clinic between January 1996 and June 1999. While both groups had similar baseline demographics (38 years old; 70% male; 310 CD4 cells/mm3; HIV RNA of 10,000 copies/mL), 23% of the co-infected individuals were on HAART compared to 35% of those infected with HIV alone.

The rate of opportunistic infections was 9.77 vs. 7.91 per 100 person years; rate of death: 6.67 vs. 2.27/100 person years; and hospitalization rate: 15.03 vs. 6.79/100 person years in co-infected and mono-infected individuals, respectively. Co-infection was therefore associated with a faster progression to death. After adjusting for differences in baseline and follow-up measures of CD4 cell count, HIV viral load, duration of HIV infection, and use of HAART, the relative risk of death for HCV/HIV co-infected individuals was 11.7; the relative risk of hospitalization was 2.5.

Even though HAART use was controlled for, the authors speculated that the differences in HIV clinical progression “may be explained in part by the lower use of HAART.” The data here are important, but it is difficult to garner definitive conclusions from such a small, retrospective single-institution case study.

The McGill study corroborates results of a much larger analysis from the Swiss HIV Cohort—which was recently published in the Lancet. 3,111 individuals initiating HAART between June 1996 and May 1999 were prospectively followed for survival, clinical progression, HIV RNA suppression, CD4 cell recovery, and HAART change according to HCV status. Of the 3,111 HIV individuals who were followed for a median of 28 months, 1,157 (37.2%) were co-infected with HCV, 1015 (87.7%) with a history of IVDU. There were significant differences in baseline HIV characteristics of the HCV-infected and HCVuninfected individuals: 27.7% vs. 23.5% had AIDS; 58.9% vs. 52.3% were antiretroviral treatment naive; and median CD4 cell count was 172 vs. 222 cells/mm3.

After the initiation of HAART, there was no association between HCV infection and the probability of reaching an HIV RNA <400 copies/mL. There were, however, differences with regard to CD4 cell recovery. After one year on HAART, the probability of fail-ing to increase CD4 counts by 50 cells/mm3 was 25.1% for HCV-infected compared to 16% for HCV-uninfected individuals (p<0.05).

Throughout the follow-up period, 7.5% of the HCV-infected individuals developed an OI compared to 4.7% of the HCV-uninfected ones (p= 0.001). Death was also more common in the HCV co-infected group: 8.8% vs. 4% (p<0.001). Interestingly, there were significant differences in the probability of clinical progression to AIDS and death when data were stratified by active IVDU and HCV. The estimated probability of clinical progression at 2 years was: 6.6% for HCV-/no active IVDU; 9.7% for HCV+/no active IVDU; and 15% for HCV+/active IVDU.

The Swiss study is one of the first (and largest) to detect an increase in OIs and death due to co-infection with HCV. One explanation for this could be the impaired CD4 cell recovery in HCV-infected individuals on HAART. These results will need to be confirmed in another study of equal size in a different country. (The fact that >85% of these 3,000 Swiss individuals had an HIV RNA <400 copies/mL makes this author think that Switzerland is not “real world” when it comes to HIV and its medical management.)

Impairment of CD4 cell recovery in HIV/HCV co-infected individuals on HAART was also seen in a Spanish study. Retrospective data were collected on 902 HIV-positive individuals (72% of whom were co-infected with HCV) seen between January 1998 and April 2000 in order to determine the immunologic and virologic impact of HCV infection on those initiating HAART. There were significant differences in baseline HIV characteristics between the HCV-infected and HCVuninfected individuals: mean CD4 count was 518 vs. 620 cell/mm3 and HIV RNA was 11,000 vs. 6,000 copies/mL, respectively. Similar proportions of individuals in each group were on HAART (92%), and there was no difference in drug adherence (83% taking >90% of pills).

After two years, there were striking immunologic and virologic differences between the two groups. HIV RNA on average declined only 606 copies/mL (5%) in the HCV-infected group compared to 5,788 copies/mL (53%) in the HCV-uninfected group. Likewise, CD4 counts on average increased 53 cells/mm3 (11%) compared to 111 cells/mm3 (22%) in the HCV-infected and HCV-uninfected group, respectively.

While this study does not provide actual data on HIV clinical disease progression, it is interesting to note such striking differences in HIV surrogate marker changes. Accordingly, the investigators raise an interesting question: Would treating a co-infected person’s HCV (regardless of liver fibrosis state) indirectly help combat the underlying HIV disease?

Studies concluding “No”

Data from the Johns Hopkins 1,742 person HIV cohort suggest that HCV does not affect HIV disease progression or survival. Hopkins researchers followed this cohort, 45% of whom were HCV-infected, from January 1996 to June 2000 in order to monitor endpoints such as CD4 cell decline to <200 cells/mm3, development of OIs, and death. HCV-infected individuals were older, more likely to be black (85% HCV+ vs. 65% HCV-) and had past or present IVDU (85% HCV+, 15% HCV-), yet no differences were observed in baseline CD4 or HIV RNA.

At the end of follow-up, no difference in HIV clinical progression was observed in the HCV co-infected individuals. Initially, an increased risk of progression to CD4 <200 cells/mm3 and death was documented in HCV-infected individuals with baseline CD4 counts between 50 and 200/mm3. When HAART use and lack of HIV RNA suppression to <400 copies/mL were controlled for, however, HCV infection was no longer significantly associated with CD4 decline or survival. Of interest, impairment of CD4 cell recovery on HAART was also noticed in the coinfected individuals at Hopkins. With three studies documenting this fact, TAG believes that intensified research in the immunology of HCV co-infection is warranted.

Two additional studies presented at the meeting failed to document accelerated HIV clinical progression or increased mortality in co-infected individuals. French researchers followed 995 HIV-positive individuals (58% of whom were co-infected with HCV) for 3 years. No significantly increased risk of AIDS or death was noted in the HCVinfected individuals compared to the HCV-uninfected individuals. And in a 504 person Seville HIV cohort, deaths due to liver failure were shown to have increased since 1997, but no significant difference in survival was observed between HCV-infected and HCV-uninfected individuals.

Opening up the debate of whether HCV alters HIV clinical progression adds to the never-ending list of questions in HIV/HCV co-infection clinical and basic research. In this fledgling field we desperately need large, collaborative, multicentered natural history co-infection studies (as well as treatment trials) in the U.S. and abroad—and the financial resources to implement, follow, and properly carry them out.