In most patients, HAART reduces circulating plasma HIV RNA to levels below the detection limits of commercial assays, generally resulting in considerable immunological reconstitution . However, the duration of an initial HAART regimen may be relatively short in routine clinical care, with the most common reasons cited for discontinuation or modification being antiretroviral drug-associated intolerance or toxicity [2–6].
Given the hepatotoxic effects of hepatitis C infection (HCV) and the accelerated progression of HCV-associated cirrhosis among HIV-infected individuals [7,8], HIV/HCV-coinfected patients may experience greater antiretroviral therapy (ART)-related intolerance and adverse events. Moreover, HIV/HCV-coinfected patients, compared with HIV-infected patients, have a higher prevalence of alcohol and substance abuse , further accelerating HCV-associated cirrhosis [7,8]. Taken together, these factors may adversely affect the duration of a first HAART regimen among HIV/HCV-coinfected patients in comparison with those without HCV.
This study of an observational HIV clinical cohort contrasted HIV/HCV-coinfected patients with HIV-infected patients with respect to time to the first occurrence of either discontinuation or modification of an initial HAART regimen and examined the reasons for initial HAART discontinuation or modification.
The University of North Carolina Center for AIDS Research, HIV/AIDS Research and Clinical Cohort was established in January 2000, with on-going enrollment. HIV-infected patients aged 18 years or older, able to provide informed consent and receiving primary HIV care at UNC are eligible to participate. This analysis included ART-naive patients who initiated HAART prior to November 2003 to allow for at least 1 year of follow-up. Patients were excluded if data were missing on HCV serostatus (n = 8); HAART initiation, modification, or discontinuation dates (n = 47); or pre-HAART CD4 cell count or HIV RNA level (n = 60). The study was approved by the University of North Carolina Institutional Review Board.
Standardized abstractions of medical records and daily updated electronic information from institutional databases provided the clinical and demographic data. HAART was defined as a combination of three or more antiretroviral drugs, including at least one protease inhibitor (PI) or non-nucleoside reverse-transcriptase inhibitor (NNRTI). Additionally, given the time frame of this study, triple nucleoside(tide) reverse transcriptase inhibitor (NRTI) combinations were also considered as HAART.
Discontinuation was defined as the cessation of all antiretroviral drugs for longer than a 2 week interval; modification was defined as any one drug substitution or intensification, except for addition of ritonavir as a boosting agent. Dosage adjustments and formulation changes were not considered as modifications. Moreover, once patients discontinued initial HAART, they were no longer at risk for experiencing modification. Similarly, once patients modified initial HAART, they were no longer at risk for discontinuing initial HAART.
Reasons for HAART discontinuation and modification were grouped into three main categories: treatment failure, toxicity, and barriers to adherence. Toxicity was defined as experiencing either an ART-related side effect or an adverse event. These toxicities were grouped into seven categories: gastrointestinal (nausea, vomiting, diarrhea, and bloating), allergic (fever, rash, and hypersensitivity), neurological (headache, peripheral neuropathy, feeling drunk, nightmares, lethargy, and flush), metabolic (diabetes, hyperlipidemia, lipodystrophy, mitochondrial toxicity, and nephrolithiasis), hematological (anemia and neutropenia), organ-related (pancreatitis, abnormal liver function tests, nephritis, and myositis), and other (alopecia and development of condyloma acuminate). Cited reasons for ‘barriers to adherence’ were depression, substance abuse, alcohol abuse, financial constraints, and problems with HAART dosage. ‘Treatment failure’ was defined as either immunological or virological non-response, not secondary to barriers to adherence or toxicity. Healthcare providers did not document more than one main discontinuation or modification reason; however, within each main category, some patients had more than one reason for discontinuation or modification.
HCV serostatus was based on HCV antibody testing (Ortho-Clinical Diagnostic, Raritan, New Jersey, USA). Body mass index was calculated using the standard formula of weight in kilograms divided by height in meters squared . Baseline values for CD4 cell counts and HIV RNA levels were those closest to, but prior to, HAART initiation.
Clinical and demographic characteristics at the time of HAART initiation were contrasted by HCV serostatus, using the Pearson χ2 test for categorical variables, the Wilcoxon–Mann–Whitney test for non-normally distributed continuous variables, or the Student's t test for normally distributed continuous variables. For analyses of HAART discontinuation or modification, patients contributed person-time from the day of HAART initiation. The day when HAART was discontinued or modified corresponded to an outcome endpoint. If patients were still on their initial HAART on 31 October, 2004, their time was censored on that date. If a patient was lost to follow-up, they were censored at their last known clinic visit. Time to either initial HAART discontinuation or modification, whichever outcome occurred first, was evaluated using Kaplan–Meier survival curves. The log-rank test was used to test for differences by HCV status.
A number of possible confounding factors were considered, including gender, race, age, insurance status, body mass index, years from HIV diagnosis to HAART initiation, calendar year of HAART initiation, type of HAART, AIDS-defining clinical condition, baseline and nadir CD4 cell counts, baseline and peak HIV RNA levels, and history of alcohol or substance abuse or dependence. CD4 cell counts and HIV RNA levels were square root and log10 transformed, respectively.
Multivariable models were fit using proportional hazards regression. Interaction was assessed between HCV and each covariate relying on a likelihood ratio test P value < 0.2. The full model included all the candidate confounding factors and interaction terms. A backward elimination procedure based on change in point estimate was used, where potential confounding variables were removed from the model if the point estimate changed by < 10%. Log–log survival curves and Schoenfeld residuals were used to assess proportionality of hazards .
Since the primary analyses were based on a combined outcome of time to the first occurrence of either modification or discontinuation of an initial HAART regimen, further analyses additionally assessed time to discontinuation and time to modification separately. Sensitivity analyses were also performed to evaluate possible bias from differential loss to follow-up.
Patients who were lost to follow-up in the primary analysis were censored while in our sensitivity analysis these patients were considered to have discontinued or modified their HAART on their last clinic visit. Additionally, models were fitted where HCV-infected patients lost to follow-up were assumed to have had the outcome at their last visit, but HCV-uninfected patients lost to follow-up were censored 2 years after their last know visit, providing an upper bound to a possible HCV effect. To estimate the lower bound, HCV-infected patients lost to follow-up were censored 2 years after their last visit, whereas HCV-uninfected patients lost to follow-up were assumed to have had the outcome at their last visit.
Statistically significant P values were considered to be < 0.05, and all tests of significance were two sided. The SAS statistical package (version 8.2; SAS Institute, Cary, North Carolina, USA) was used for these analyses.
Of the 296 ART-naive patients initiating HAART, 33% were women and the median age was 38 years [interquartile range (IQR), 32–45; Table 1]. Sixty-four (22%) patients were infected with HCV. Compared with HIV-infected patients, HIV/HCV-coinfected patients were older (median age 44 versus 37 years; P < 0.001) but had comparable baseline CD4 cell counts [median 85 cells/μl (IQR, 15–264) and 122 cells/μl (IQR, 30–302), respectively] and HIV RNA levels [median 5.0 log10 copies/ml (IQR, 4.5–5.3) and 5.0 log10 copies/ml (IQR, 4.5–5.5), respectively].
Most patients initiated HAART after 1998, regardless of HCV serostatus (Table 1). For 40%, the regimen contained a NNRTI, for 43% a PI, for 11% both a PI and NNRTI, and for 6% only NRTI. The majority of patients initiating an NNRTI-based regimen received efavirenz (60%). Of the 128 patients who initiated a PI-based regimen, most received nelfinavir (62%), followed by indinavir (25%), and lopinavir/ritonavir (10%). Twenty-three patients received a ritonavir-containing regimen and five of these patients were HCV coinfected. Most patients took lamivudine/zidovudine (67%) or lamivudine/stavudine (23%). There were no differences in initial PI, NNRTI, or NRTI drugs by HCV serostatus.
Reasons for HAART discontinuation or modification
During follow-up, 104 patients discontinued and 91 patients modified their initial HAART regimen, where 150 (65%) HIV-infected patients discontinued or modified their initial HAART compared with 45 (70%) HIV/HCV-coinfected patients. Among patients who either discontinued or modified their initial HAART, the prevalence of discontinuing or modifying secondary to therapeutic failure, toxicity, and barriers to adherence was 12%, 41%, and 47%, respectively (Table 2). HCV serostatus did not influence reasons for discontinuation or modification (P = 0.29). The predominant causes of toxicity were neurological and hematological. Of the 16 HCV-infected patients who discontinued or modified owing to toxicity, six (38%) and four (25%) did so secondarily to hematological and neurological reasons, respectively. Among the 63 HCV-uninfected patients who discontinued or modified owing to toxicity, hematological and neurological events were responsible for 18 (29%) and 16 (25%) changes, respectively.
Duration of initial HAART
The median time to first occurrence of either discontinuation or modification was 473 days (IQR, 167–940), and this did not differ by HCV serostatus, with a median time of 401 days (IQR, 128–821) observed among HCV-infected patients and 493 days (IQR, 204–952) among HCV-uninfected patients (P = 0.22). Overall, just under two-thirds of patients remained on their initial HAART regimen after 12 months of therapy. When considered separately, modification of therapy occurred at earlier times than discontinuations (Fig. 1).
Duration of initial HAART was associated with reasons for discontinuation (P = 0.02). Among the 27 patients discontinuing therapy in the first 3 months, the reasons were barriers to adherence (48%), toxicity (52%), and treatment failure (0%); these percentages were 72%, 22%, and 6% for the 36 patients discontinuing at 3–12 months, and 78%, 22%, and 0% for the 41 patients discontinuing after more than 12 months. Reasons for modification were also different by duration of therapy (P < 0.001). Among the 19 patients who modified their therapy in the first three months, the reasons were barriers to adherence (16%), toxicity (79%), and treatment failure (5%). These percentages were 7%, 70%, and 22% for the 27 patients modifying at 3–12 months, and 36%, 31%, and 33% for the 45 patients modifying after more than 12 months. However, reasons for discontinuation or modification across time did not differ by HCV coinfection status (all P > 0.28), or any other clinical or demographic characteristics considered (all P > 0.10).
Effect of hepatitis C infection on duration of initial HAART
At 6 and 12 months from HAART initiation, 70% and 56% of HIV/HCV-coinfected patients remained on their initial regimen, respectively, which was slightly lower than the 79% and 64%, respectively, for HIV-infected patients (Fig. 2; log rank P = 0.20). The unadjusted hazard ratio (HR), contrasting the time to either discontinuation or modification of the initial HAART regimen by HCV serostatus, was 1.24 [95% confidence interval (CI), 0.89–1.73]. There was a similar effect of HCV serostatus on the HR of either discontinuation or modification owing to toxicity (HR, 1.24; 95% CI, 0.90–1.71) and treatment failure (HR, 1.24; 95% CI, 0.93–1.64), but a smaller effect owing to barriers to adherence (HR, 1.02; 95% CI, 0.72–1.45).
The full model included gender, race, age, insurance, body mass index, years from HIV diagnosis, prior AIDS-defining clinical condition, baseline and nadir CD4 cell count, baseline and peak HIV RNA level, calendar year of HAART initiation, type of HAART, and history of alcohol and substance abuse or dependence. The final reduced model included gender, race, age, baseline and nadir CD4 cell counts, and baseline and peak HIV RNA levels. The full and final reduced model results were comparable to those observed prior to adjustment for any covariates: HR for the full model 1.39 (95% CI, 0.95–2.03) and HR for the final model 1.39 (95% CI, 0.97–1.99). Comparable results were also observed by HCV serostatus when time to modification and time to discontinuation were considered in separate models. The unadjusted and adjusted HRs for the effect of HCV on time to modification were 1.11 (95% CI, 0.67–1.85) and 1.22 (95% CI, 0.70–2.11), respectively; and for time to discontinuation were 1.36 (95% CI, 0.87–2.11) and 1.58 (95% CI, 0.97–2.56), respectively.
Initial HAART outcomes
Only 12% of patients who discontinued or modified their initial HAART did so as a consequence of therapeutic failure. However, the category of therapeutic failure was not based on immunological or virological response to HAART but rather on the healthcare providers' documented reasons for discontinuation or modification. The most proximal CD4 cell count prior to the date of discontinuation, modification, or censoring did not differ by HCV serostatus, (P = 0.21, 0.42, and 0.86, respectively). HIV RNA levels at these time points were also similar by HCV serostatus (P = 0.85, 0.80, and 0.05, respectively). However, median CD4 cell counts were different at the time of discontinuation, modification, or censoring (P < 0.001), as were HIV RNA levels (P < 0.001) (Table 3).
Among the 104 patients who discontinued therapy, the median time to restarting therapy was 143 days (IQR, 61–324). To the best of our knowledge, no patient was placed on a structured treatment interruption. A greater proportion of patients who remained off therapy for over 3 months stopped their first regimen because of barriers to adherence (P = 0.02); however, time off therapy was similar by HCV serostatus (P = 0.43) (Table 3).
Immunological and virological outcomes
Based on an intention-to-treat analysis, HIV/HCV-coinfected patients had a slightly reduced CD4 cell count response by 24 weeks of therapy compared with HIV-infected patients (median increase 88 cells/μl (IQR, 34–159) versus 110 cells/μl (IQR, 51–197); P = 0.18). Among HIV/HCV-coinfected patients, 65% had at least a 50 cells/μl increase in CD4 cell count by 24 weeks of therapy, compared with 73% of HIV-infected patients (odds ratio, 1.52; 95% CI, 0.87–2.65). This difference remained after adjustment for pretherapy CD4 cell count, HIV RNA level, gender, and age. By 48 weeks of therapy, HIV/HCV-coinfected patients had an average increase of 130 cells/μl (IQR, 77–207) in contrast to 170 cells/μl (IQR, 79–293) among HIV-infected patients (P = 0.07). Similar results were observed in the as-treated analysis.
Differences in HIV RNA levels were not observed at 24 weeks of therapy in either the intention-to-treat or as-treated analyses. Using an intention-to-treat analysis, 73% of HIV/HCV-coinfected patients and 72% of HIV-infected patients had undetectable HIV RNA levels (P = 0.88). However, 54% of HIV/HCV-coinfected patients versus 71% of HIV-infected patients remained virologically suppressed at 48 weeks of therapy in an intention-to-treat analysis (P = 0.04), which was comparable to the results in the as-treated approach (62% of HCV/HIV-coinfected patients and 78% of HIV-infected patients; P = 0.09).
Six (9%) HCV-infected and 35 (15%) HCV-uninfected patients were lost to follow-up. Analogous results to those presented above were obtained for HR values of time to either discontinuation or modification of initial HAART in the sensitivity analyses where all patients lost to follow-up were considered to have the outcome of interest (HR, 1.20; 95% CI, 0.88–1.64). When HCV-uninfected patients who were lost to follow-up were assumed to have the outcome of interest while HCV-infected patients who were lost to follow-up were assumed to have remained on their therapy for an additional 2 years of follow-up, a lower bound of 0.91 was estimated for the effect of HCV on time to discontinuation or modification. Conversely, an upper bound of 1.55 was estimated when HCV-infected patients were considered to have the outcome and HCV-uninfected patients assumed to remain on therapy.
Given the limited number of antiretroviral agents with differing mechanisms of action, understanding the time on an initial HAART regimen and factors associated with a shorter duration is clinically relevant. In line with prior research [2–6], we also found that an initial HAART regimen was either modified or discontinued in over one-third of patients within the first year of therapy, and that ART-associated intolerance and psychosocial issues were the primary reasons for altering the first HAART regimen.
Patients with HCV/HIV-coinfection are at high risk for experiencing psychosocial problems  and hepatotoxic effects [7,8], possibly leading to greater HAART intolerance [12,13]. Therefore, in this study, we assessed whether HCV/HIV-coinfection predicted a shorter time on an initial HAART regimen. Overall, HCV/HIV-coinfected patients had a slightly higher hazard of discontinuing or modifying their initial HAART regimen (HR, 1.39; 95% CI, 0.95–2.03). Our results are comparable to a prior study of ART-naive HIV-infected patients in Italy, where 80% received PI-based HAART and 46% were coinfected with HCV/HIV .
The relatively small effect of HCV on duration of first HAART may be related to the modest clinical relevance of HCV-associated hepatotoxicity. For example, Sulkowski et al.  observed that HCV/HIV-coinfected patients were at increased risk of HAART-related severe hepatotoxicity when prescribed non-ritonavir combinations; however, the majority of HCV/HIV-coinfected patients did not experience any significant HAART intolerance. In another study, hepatotoxicity leading to ART discontinuation was observed in 13% of HCV-infected and 4% of HCV-uninfected patients over a 2 year follow-up . In our study, only three cases of hepatotoxicity occurred. They were all in HCV-uninfected patients, and all resulted in HAART discontinuation.
The considerable proportion of patients with psychosocial issues in the overall study population may also explain why any additional psychiatric or substance abuse comorbidity among HCV/HIV-coinfected patients did not result in a substantial decrease in the time to discontinuation or modification of an initial HAART regimen. As we have previously documented, HIV/HCV-coinfected patients are more likely to report heavy alcohol use and symptoms of major depression than HIV-infected patients ; however, overall psychiatric and substance abuse comorbidity is substantial among our patients .
The first author (D.H., an infectious diseases physician) reviewed all reasons and dates for discontinuation or modification to increase validity and reliability; however, misclassification of time to event may still exist. We also considered a relatively short 2 week interval of not receiving any antiretroviral drugs as discontinuation, but the results were unaltered if this was increased to 4 weeks. We excluded 60 patients because there were no CD4 cell count or HIV RNA data available before therapy. Clinicians may have had this information from outside providers or screening for clinical research studies, but documentation was not available in the patient's medical record. Patients with available and unavailable values did not differ by HCV/HIV-coinfection status (P = 0.77), time to modification or discontinuation (log rank P = 0.46), or any other demographic or clinical characteristic (all P > 0.10). Additionally, as in any observational study, we were unable to adjust for possible confounding by unmeasured confounders. Therefore, confirmation of our findings will require using prospective, standardized data collection with more thorough evaluation of factors that are generally not routinely available in clinical care documentation.
HCV antibody testing was relied on to classify patients as infected with HCV as HCV viral loads or genotypes were generally unavailable, but given that this test was preformed in a high-risk population its positive predictive value is estimated to be between 90 and 95% . HCV antibody was also used as a proxy for hepatic pathology since liver biopsies were not routinely available. Since HCV antibody is a sensitive but not a specific marker of hepatic damage, we may be underestimating the effect in a subgroup of HCV-infected patients with more advanced liver disease.
Our study only included patients initiating a first HAART regimen with no prior ART exposure. Therefore, we cannot generalize our findings to subsequent HAART regimens, which may be less tolerable. Additionally, larger studies with longer follow-up are needed to assess the effect of HCV on HAART-related adverse events that are rare or have a delayed onset, as well as the impact of HCV on HIV disease progression and survival [19,20].
Prior work suggested that, despite similar virological response, immunological reconstitution may be blunted in HCV-infected compared with HCV-uninfected patients [20–22]. In this cohort, patients with HCV/HIV coinfection experienced inferior immunological reconstitution, as measured by CD4 cell counts at 24 and 48 weeks post HAART initiation, as compared to patients without HCV/HIV-coinfection. These findings are comparable to the results of a recent meta-analysis .
In summary, although considerable concern exists that HCV/HIV-coinfected patients may experience inferior HIV therapeutic and clinical outcomes, we did not observe a substantial difference in time to discontinuation or modification of a first HAART regimen by HCV serostatus. Further work evaluating the effect of HCV would benefit from the inclusion of additional HCV assessment markers, including HCV viral load, genotype, and HCV-associated cirrhosis progression. Additionally, in our study population, < 3% of HCV/HIV-coinfected patients had received HCV therapy. As more patients are treated for their HCV, the impact on HIV therapeutic management will need to be reassessed. Finally, as HIV management improves, the effect of comorbidities such as HCV/HIV-coinfection on HIV therapeutic and clinical outcomes may increase and deserves further close monitoring.
We greatly appreciate the support of all study staff members, HIV care providers, the staff of the infectious disease clinic, and particularly the patients who participated in this study.
Sponsorship: This study was supported by the University of North Carolina at Chapel Hill, Center for AIDS Research, National Institutes of Health funded program P30 AI 50410; The US Department of Health and Human Services, HRSA, HAB, Office of Science and Epidemiology; the Epidemiology Department at GlaxoSmithKline; SAS Institute; the Medical Foundation of North Carolina, Inc.
1. US Public Health Service. Guidelines for the Use of Antiretroviral Agents in HIV-infected Adults and Adolescents
. Washington, DC: US Department of Health and Human Services; updated 6 October 2005. Accessed 9 October 2005: http://aidsinfo.nih.gov/guidelines
2. d'Arminio Monforte A, Lepri AC, Rezza G, Pezzotti P, Antinori A, Phillips AN, et al
. Insights into the reasons for discontinuation of the first highly active antiretroviral therapy (HAART) regimen in a cohort of antiretroviral naive patients. AIDS 2000; 14:499–507.
3. Mocroft A, Youle M, Moore A, Sabin CA, Madge S, Lepri AC, et al
. Reasons for modification and discontinuation of antiretrovirals: results from a single treatment centre. AIDS 2001; 15:185–194.
4. Palella FJ Jr, Chmiel JS, Moorman AC, Holmberg SD. Durability and predictors of success of highly active antiretroviral therapy for ambulatory HIV-infected patients. AIDS 2002; 16:1617–1626.
5. O'Brien ME, Clark RA, Besch CL, Myers L, Kissinger P. Patterns and correlates of discontinuation of the initial HAART regimen in an urban outpatient cohort. J Acquir Immune Defic Syndr 2003; 34:407–414.
6. Chen RY, Westfall AO, Mugavero MJ, Cloud GA, Raper JL, Chatham AG, et al
. Duration of highly active antiretroviral therapy regimens. Clin Infect Dis 2003; 37:714–722.
7. Sulkowski MS, Thomas DL. Hepatitis C in the HIV-infected person. Ann Intern Med 2003; 138:197–207.
8. Monga HK, Rodriguez-Barradas MC, Breaux K, Khattak K, Troisi CL, Velez M, et al
. Hepatitis C virus infection-related morbidity and mortality among patients with human immunodeficiency virus infection. Clin Infect Dis 2001; 33:240–247.
9. Hooshyar D, Theodore D, Napravnik S, Robertson K, Fried M, Eron JJ. Higher prevalence of active psychiatric diagnoses and substance abuse among patients with HIV and HCV co-infection.41st Annual Meeting of the Infectious Diseases Society of America
. San Diego, October 2003 [abstract 366-M].
10. Centers for Disease Control and Prevention. Prevalence of overweight and obesity among adults with diagnosed diabetes: United States, 1988–1994 and 1999–2002.MMWR
11. Hess KR. Graphical methods for assessing violations of the proportional hazards assumption in Cox regression. Stat Med 1995; 14:1707–1723.
12. Sherman KE, Rouster SD, Chung RT, Rajicic N. Hepatitis C virus prevalence among patients infected with human immunodeficiency virus: a cross-sectional analysis of the US adult AIDS clinical trials group. Clin Infect Dis 2002; 34:831–837.
13. Mocroft A, Phillips AN, Soriano V, Rockstroh J, Blaxhult A, Katlama C, et al
. Reasons for stopping antiretrovirals used in an initial highly active antiretroviral regimen: increased incidence of stopping due to toxicity or patient/physician choice in patients with hepatitis C coinfection. AIDS Res Hum Retroviruses 2005; 21:527–536.
14. Ripamonti D, Arici C, Pezzotti P, Maggiolo F, Ravasio L, Suter F. Hepatitis C infection increases the risk of the modification of first highly active antiretroviral therapy in HIV-infected patients. AIDS 2004; 18:334–337.
15. Sulkowski MS, Thomas DL, Chaisson RE, Moore RD. Hepatotoxicity associated with antiretroviral therapy in adults infected with human immunodeficiency virus and the role of hepatitis C or B virus infection. JAMA 2000; 283:74–80.
16. Melvin DC, Lee JK, Belsey E, Arnold J, Murphy RL. The impact of co-infection with hepatitis C virus and HIV on the tolerability of antiretroviral therapy. AIDS 2000; 14:463–465.
17. Whetten K, Reif SS, Napravnik S, Swartz MS, Thielman NM, Eron JJ Jr, et al
. Substance abuse and symptoms of mental illness among HIV-positive persons in the Southeast. South Med J 2005; 98:9–14.
18. Fried MW. Diagnostic testing for hepatitis C: practical considerations. Am J Med 1999; 107(Suppl 6B):S31–S35.
19. Sulkowski M, Moore R, Mehta S, Chaisson R, Thomas D. Hepatitis C and progression of HIV disease. JAMA 2002; 288:199–206.
20. Greub G, Ledergerber B, Battegay M, Grob P, Perrin L, Furrer H, et al
. Clinical progression, survival, and immune recovery during antiretroviral therapy in patients with HIV-1 and hepatitis C virus coinfection: the Swiss HIV Cohort Study. Lancet 2000; 356:1800–1805.
21. De Luca A, Bugarini R, Lepri AC, Puoti M, Girardi E, Antinori A, et al
. Coinfection with hepatitis viruses and outcome of initial antiretroviral regimens in previously naive HIV-infected subjects. Arch Intern Med 2002; 162:2125–2132.
22. Braitstein P, Palepu A, Dieterich D, Benhamou Y, Montaner SG. Special considerations in the initiation and management of antiretroviral therapy in individuals coinfected with HIV and hepatitis C. AIDS 2004; 18:2221–2234.
23. Miller MF, Haley C, Koziel MJ, Rowley CF. Impact of hepatitis C virus on immune restoration in HIV-infected patients who start highly active antiretroviral therapy: a meta-analysis. Clin Infect Dis 2005; 41:713–720.