Treatment of Hepatitis C Virus (HCV) Infection in Patients Coinfected with HIV in the HIV Outpatient Study (HOPS), 1999–2007

[Guest guest]

Treatment of Hepatitis C Virus (HCV) Infection in Patients Coinfected with HIV in the HIV Outpatient Study (HOPS), 1999–2007

Abstract

Liver disease due to hepatitis C virus (HCV) infection is a leading cause of non-AIDS-related morbidity and mortality in patients infected with HIV. We assessed the frequency of and predictors for initiation of treatment for HCV infection among patients coinfected with HCV/HIV enrolled in the HIV Outpatient Study (HOPS) during 1999–2007. We included patients with confirmed HCV infection, at least 1 year of subsequent follow-up, and no evidence of prior HCV treatment. We assessed predictors of HCV treatment initiation using proportional hazards analyses. During 1999–2007, 103 (20%) HOPS patients coinfected with HCV/HIV initiated HCV treatment during a median of 4.3 years of follow-up (interquartile range: 2.7, 6.7). In multivariable analysis, non-Hispanic black race/ethnicity (hazard ratio HR] 0.3; 95% confidence interval [CI] = 0.2, 0.6) was independently associated with a lower likelihood of HCV treatment. Elevated alanine aminotransferase

(ALT; HR 3.5; 95% CI = 2.2, 5.6) and CD4+ cell count ≥500 cells/mm3 (HR 1.8; 95% CI = 1.2, 2.8) at the start of observation were independently associated with higher likelihood of HCV treatment. For patients starting observation in 1999–2001, 2002–2004 and 2005–2007, 5%, 11% and 21% of patients initiated treatment during the first year of follow-up, respectively. Between 1999 and 2007, despite a stable low fraction of patients coinfected with HCV/HIV initiating treatment for HCV infection, an increasing proportion initiated treatment within the first year after the infection was confirmed. Treatment of HCV infection in patients coinfected with HCV/HIV should be considered a priority, given the increased risk of accelerated end-stage liver disease.

Introduction

With the advent of highly active combined antiretroviral therapy (cART), HIV-associated morbidity and mortality has declined dramatically in the United States (US).[1,2] With longer survival and fewer opportunistic infections, liver disease because of hepatitis C (HCV) infection has emerged as a leading cause of non-AIDS-related death.[3–5] Up to 30% of HIV-infected individuals are coinfected with HCV.[6–8] Among injection drug users, prevalences of coinfection range from 50% to 90%.[6–8] Coinfected individuals have an increased risk of, and

accelerated progression to, end-stage liver disease and hepatocellular carcinoma compared with HCV- or HIV-monoinfected individuals.[9–14] It is therefore recommended that all HIV-infected individuals be screened for HCV infection and all coinfected individuals evaluated for HCV treatment.[10,15–17]

In 1999, the US Public Health Service (USPHS) and Infectious Diseases Society of America (IDSA) Prevention of Opportunistic Infections Working Group recommended routine screening of all patients infected with HIV for HCV infection.[18] However, the safety and efficacy of combination pegylated interferon (IFN) and ribavirin (used and approved for the treatment of HCV monoinfection[19]) for the treatment of coinfected individuals was not demonstrated until 2004.[20–24] In the same year, guidelines on the treatment of opportunistic infections among patients infected with HIV for the first time provided discrete guidance for treating HCV coinfection

and recommended all patients with coinfection be considered for treatment.[25] In 2005, the US Food and Drug Administration (FDA) approved anti-HCV combination therapy for treatment of patients coinfected with HCV/HIV, and a European consensus statement reiterated that treatment of HCV and hepatitis B virus coinfections be prioritized.[10] Despite these broadly endorsed guidelines, recent data suggest that barriers remain and few eligible HCV/HIV-coinfected individuals are treated for HCV infection. In two recent retrospective analyses using a national cohort of US veterans infected with HCV, only 10–15% of treatment-eligible patients with coinfection received anti-HCV therapy.[26,27]

We evaluated data from the HIV Outpatient Study (HOPS)[2] to assess HCV treatment practices for HCV/HIV-coinfected individuals receiving care in typical HIV care settings at ten HIV-specialty clinics throughout the United States.

The HIV Outpatient Study (HOPS)

The HOPS is an ongoing, open prospective cohort study conducted at HIV-specialty clinics in the United States which has continuously recruited and followed patients infected with HIV since 1993.[9] The study protocol is approved and renewed annually by the participating institutions' ethical review board. All study participants provide written, informed consent.

HIV Outpatient Study clinicians have extensive experience treating patients infected with HIV. Information is abstracted from outpatient charts at each visit, entered electronically by trained staff, compiled centrally, and reviewed and edited before analysis. Abstracted information includes: demographic characteristics and risk factors for HIV infection; diagnoses; symptoms; prescribed medications, including dosage and duration; laboratory values, including CD4+ cell counts and plasma HIV viral loads; mortality and hospitalizations.

Study Population

We studied HOPS participants with confirmed HCV infection followed during 1999–2007. We chose to initiate our observation in 1999 to coincide with the year recommendations were first issued to screen all patients infected with HIV for chronic HCV infection.[18] We used HOPS data updated as of 30 June 2008. Patients were eligible for analysis if they had a confirmed HCV infection (positive qualitative or quantitative HCV RNA, HCV b-DNA or HCV RNA genotype), had at least 1 year of follow-up in the HOPS after the date of confirmed HCV infection and no evidence of prior HCV treatment (i.e. IFN, pegylated-IFN [peg-IFN], ribavirin). Start of observation (baseline) was the first date on or after 1 January 1999 when the patient was under active HOPS follow-up and had a confirmed HCV infection. We compared individuals during three 3-year time periods based

on the year they initiated observation: 1999–2001, 2002–2004 and 2005–2007; annualized data were considered too sparse for robust analyses. Initiation of HCV treatment was defined as receipt of at least one day of HCV therapy: IFN, peg-IFN and/or ribavirin.

We defined a person as having chronic liver disease if they had any of the following liver-related diagnoses (identified using diagnostic codes): cirrhosis, hepatic steatosis, hepatomegaly, hepatosplenomegaly or liver failure. The presence of anaemia, neutropenia and thrombocytopenia were determined by either the presence of the corresponding diagnostic code in the abstracted medical record or by a qualifying laboratory value: haemoglobin <11 grams/dL (anaemia), absolute neutrophil count <1500 cells/μL (neutropenia) and thrombocyte count <100 000 cells/μL (thrombocytopenia). For individuals with a known cause of death, we categorized the death to be liver related if any of the following were present as the primary or secondary cause of death: ascites; chronic hepatitis C infection; cirrhosis of the liver; gastrointestinal or oesophageal varices; hepatic encephalopathy; chronic hepatitis B infection; hepatomegaly; liver cancer (including

hepatocellular carcinoma) or liver failure.

Statistical Analyses

We assessed the proportion of patients who initiated HCV treatment after baseline. We compared the characteristics of patients who did vs did not initiate HCV treatment during follow-up. Characteristics included conditions that might be considered a contraindication to or cause to defer treatment of HCV infection, such as depression, chronic hepatitis B virus coinfection, chronic liver disease, CD4+ cell count <200 cells/mm3, hypertension, diabetes or alcohol abuse. To explore potential effects of restricting our analysis to patients with a confirmed HCV infection, we also compared the characteristics of patients who were HCV-seropositive without a documented confirmatory test to patients who had confirmed HCV infections. We used the chi-square or Fisher's exact test for categorical independent variables, the Wilcoxon rank sum test for continuous independent variables and the Cochran-Armitage test for trend in proportions.

Demographic and clinical factors associated with initiation of HCV treatment in bivariate analyses at P < 0.10 were included in a multivariable proportional hazards model.

We examined differences in time to initiating HCV treatment among patients by baseline calendar period (start of observation) using the Kaplan–Meier method and log-rank test. We also assessed the proportion of patients initiating HCV treatment within the first year of follow-up, according to baseline period, using the Cochran-Armitage test. Finally, we described the proportion of patients coinfected with HCV/HIV who died from a liver-related condition and who had no evidence of having received HCV treatment.

During 1999–2007, among 5203 active HOPS participants, 941 had some evidence of HCV infection (either a positive HCV-antibody or a positive confirmatory HCV test, or both). We focused our study on 639 patients who had a positive confirmatory HCV test. We further restricted analyses to the 507 (9.7%) patients who had at least 1 year of HOPS follow-up after HCV infection was confirmed (n = 104 excluded) and no evidence of prior HCV treatment (n = 28 excluded). These 507 patients were followed in the HOPS for a median of 4.3 years (interquartile range [iQR]: 2.7, 6.7); HCV confirmatory testing was completed for 115 patients prior to entry into observation (baseline) of whom 45 patients had confirmatory testing a median of 5.3 months (IQR: 2.4, 15.5) before entry into HOPS (the remaining 70 patients had confirmatory testing after entry into HOPS but before entry into observation).

Of 507 patients, 70% were men, 44% were black people and the median age at baseline was 44 years (Table 1). Nearly half of patients (49%) had injection drug use (IDU) as their HIV transmission risk. Median baseline CD4+ cell count was 365 cells/mm3 (IQR: 208, 538), and 87% of patients were antiretroviral therapy (ART) experienced at the time HCV infection was confirmed. The baseline prevalences of chronic liver disease and depression were 7% and 28%, respectively (Table 1). HCV genotype results were available for 178 (35%) patients: 84% were infected with HCV genotype 1, and 6%, 8% and 2% had types 2, 3 and 4, respectively. Only 36 (7%) patients had a documented liver biopsy during follow-up.

One hundred and three patients (20%) initiated HCV treatment, of whom 8% (n = 8) initiated treatment within the first year after confirmation of their HCV infection. After limiting the analyses to 116 patients without potential contraindications to, or conditions that might have led to deferral of, treatment (n = 391 excluded), the proportions of patients initiating HCV treatment during entire follow-up (23%) and within the first year (9%) remained nearly the same as in the overall cohort.

In multivariable analysis, non-Hispanic black race/ethnicity (hazard ratio

---

0.3; 95% confidence interval [CI] = 0.2, 0.6) was independently associated with a lower likelihood of HCV treatment, whereas elevated alanine aminotransferase (ALT; HR 3.5; 95% CI= 2.2, 5.6) and CD4+ cell count ≥500 cells/mm3 (HR 1.8; 95% CI = 1.2, 2.8) were independently associated with higher likelihood of HCV treatment (Table 2). In a separate multivariable analysis limited to the 167 patients with available HCV genotype results using a model that controlled only for race/ethnicity, baseline CD4+ cell count and ALT levels, we found no association between HCV genotype and the likelihood of initiating HCV treatment (genotypes 1 and 4 [genotypes 2 and 3 referent]; HR 1.2, 95% CI = 0.59, 2.4).

For each of the 3-year baseline periods 1999–2001, 2002–2004 and 2005–2007, although the proportions of patients eligible for HCV infection treatment remained stable, the proportions of patients who initiated HCV treatment during the first year after confirmation of their HCV infection increased: 5%, 11% and 21%, respectively (test for trend P-value <0.001). The prevalences of depression, low CD4+ cell count, anaemia or chronic HBV infection were similar during each observation period (data not shown).In Kaplan–Meier analyses, the time to initiation of HCV treatment decreased significantly (log-rank test P < 0.001) from the earliest to the most recent of these periods (Fig. 1). Further analysis showed that the difference was significant comparing patients who started observation during 1999–2001 vs 2005–2007 (P < 0.001), but statistical significance was not demonstrated when comparing

1999–2001 vs 2002–2004 (P = 0.06) or 2002–2004 vs 2005–2007 (P = 0.08).

(Enlarge Image)

Figure 1.

Time to initiation of treatment for hepatitis C virus (HCV) infection, by baseline period (beginning of observation), the HIV Outpatient Study (HOPS), 1999–2007.

[ CLOSE WINDOW ]

Figure 1.

Time to initiation of treatment for hepatitis C virus (HCV) infection, by baseline period (beginning of observation), the HIV Outpatient Study (HOPS), 1999–2007.

We identified 159 patients who had positive HCV serology and no documented confirmatory HCV test. Compared to patients with confirmed HCV infections, the HCV-seropositive patients who lacked a confirmatory test had lower median CD4+ cell counts (365 vs 299 cells/mm3, P = 0.02) and were more likely to have a baseline HIV plasma viral load ≥10 000 copies/mL (24%vs 41%, P < 0.01) and an unknown type of health care coverage (5%vs 10%, P = 0.05). No significant differences between these groups were observed with regard to gender, race/ethnicity, HIV transmission risk group and nadir CD4+ cell count or history of AIDS. Three (2%) patients who were HCV seropositive without confirmatory testing initiated treatment for HCV infection during follow-up.

Eighty-nine (18%) patients with confirmed HCV infection died; cause of death was known for 53 and of these, 21 (24%) deaths were liver related: cirrhosis (n = 10), liver failure (n = 6), encephalopathy (n = 4) and hepatocellular carcinoma (n = 1). Six (29%) of these patients had initiated HCV treatment.

We found that only one-fifth of patients confirmed with HCV/HIV coinfection observed in the HOPS during 1999–2007 initiated treatment for HCV infection during a median of 4.3 years of observation. Non-Hispanic black patients were less likely to initiate treatment, whereas patients with elevated ALT or higher CD4+ cell counts were more likely to initiate treatment. Although the fraction of patients treated in each time period remained stable, the fraction of persons initiating anti-HCV treatment early (within 1 year after confirmation of HCV infection) steadily increased; this trend might reflect increasing appreciation of the value of early HCV treatment. We are aware of only one other study demonstrating an improvement in HCV care and treatment practices by HIV clinicians[28] since the European consensus statement was published in 2005.[10]

Other studies have reported lower fractions of patients initiating treatment for HCV infection, ranging from 3% to 13%.[28–32] We followed patients a median of approximately 4 years and cannot be certain that the duration of observation and subsequent opportunity to initiate anti-HCV therapy in these other studies was comparable. Four US studies during 1999–2003 demonstrated that <4–15% of patients coinfected with HCV/HIV were prescribed HCV treatment.[26,27,29,33] Similar low rates of HCV treatment have been documented in Europe.[31,32] Because we included data through 2007 and possibly observed patients for a longer duration, we might

have found a higher proportion of patients with coinfection initiating HCV treatment than has been reported in previous studies.

In our multivariable analysis, we found no association between the absence of treatment and prevalent anaemia, depression, alcohol use and chronic liver disease. This finding suggests that there might be other barriers to initiate HCV treatment. The findings that only a relatively low proportion of patients coinfected with HCV/HIV were treated in this and other studies, and that only one-third of HOPS patients whose death was caused by or related to a perimortal liver-related condition initiated HCV treatment, underscore the need to identify and overcome barriers to HCV treatment.

Consistent with findings from other studies of HCV treatment practices,[27,28,34,35] we found non-Hispanic black people were less likely than non-Hispanic whites to initiate HCV treatment. Although studies have also found poorer response to HCV treatment among black patients,[36–38] current guidelines recommend HCV treatment for patients coinfected with HCV/HIV regardless of race.

Our findings that elevated serum ALT level and a higher CD4 cell count were independently associated with a greater likelihood of being treated for HCV infection are consistent with data reported by others[39] and with guidelines available during the period under observation (1999–2007).[40] However, as it has been recognized that liver fibrosis can develop in patients with normal serum ALT values,[41] and the prevalence of fibrosis is higher in patients with coinfection compared to patients with monoinfection (up to 40%vs 30%),[42,43]

subsequent international[16] and US guidelines[44] have recommended that patients with coinfection should be offered treatment regardless of their serum ALT level.

There are several important limitations to our study. First, some patients might have received treatment prior to enrolling in HOPS and were included erroneously in this analysis, especially the 45 patients with HCV infection confirmed prior to HOPS enrollment. Second, it is possible that data were unavailable for patients referred to a gastroenterologist or hepatologist for HCV-related treatment (e.g. no referral note was present) and thereby decreased artificially the number of patients observed as treated. However, HOPS providers record all medications patients are taking, including medications not necessarily prescribed by the HOPS clinician. Third, although HCV genotype and the degree of liver damage have been shown to be predictors for initiating HCV treatment,[27,28,35,45] we were unable to demonstrate these associations in our study, perhaps

because of the small numbers of patients for whom this information was available. Fourth, we included only patients with confirmed HCV infection, however, including patients who are HCV seropositive who lacked confirmatory testing would have lowered the fraction of patients treated as only 2% of such persons initiated anti-HCV therapy. Finally, we cannot infer patients' and providers' intentions regarding decisions to obtain confirmatory testing of HCV infection or to initiate treatment. Our ability to identify actionable barriers to treatment (e.g. number of patients who refused treatment and their reasons, providers' reasons for not offering treatment) was limited.

In summary, our data suggest that although the fraction of patients treated for HCV infection in the HOPS is higher than reported elsewhere, the fraction treated remains low. The trend towards earlier initiation of therapy is encouraging, and every effort should be undertaken to provide HCV therapy to every eligible patient. Relative to the past, patients infected with HIV now are generally healthier and visit their primary providers frequently so that opportunities to address HCV infection and its treatment are numerous. Newer technologies to assess liver fibrosis (e.g. biomarkers, non-invasive imaging) will likely diminish the need for liver biopsies, a frequently cited deterrent to HCV treatment.[46,47,48] Where providers have aggressively increased efforts to treat HCV infection in eligible patients coinfected with HCV/HIV, hospitalizations and

deaths related to liver disease have stabilized.[49,50] Further research to understand and eliminate barriers to HCV treatment in coinfected persons warrants increased attention.

http://www.medscape.com/viewarticle/741470?src=rss