Fosamprenavir/ritonavir in patients with viral hepatitis coinfection: an observational multicohort study

Abstract

Objective: Safety and tolerability evaluation of adapted dose regimens containing fosamprenavir/ritonavir (FPV/r) in HIV-infected subjects with viral hepatitis co-infection.

Methods: A retrospective multicohort analysis was conducted. Subjects from three European cohorts who started FPV/r or lopinavir/ritonavir (LPV/r) as a comparator contributed data to a centralized database. Subjects were divided into five groups by treatment regimen and level of hepatic impairment (aspartate aminotransferase [AST] platelet ratio index [APRI] score < or ≥2). Multivariable Cox regression analyses controlling for demographic factors, baseline CD4 count, FIB-4 score, use of antiretroviral therapy, and laboratory markers (bilirubin and platelet count) were performed to identify factors independently associated with risk of developing adverse events or safety events (eg, drug discontinuation, alanine aminotransferase (ALT) elevation, hepatic decompensation/death).

Results: A total of 1096 patients contributed data to the study. Fosamprenavir/ritonavir (except in subjects with APRI ≥2 receiving standard dose) was associated with a higher two-year risk of drug discontinuation compared with LPV/r. Restricting the analysis to discontinuations due to adverse events (AEs), only subjects who received the reduced dose were more likely to discontinue ≥1 drug in the FPV/r regimen. There were no statistical differences in ALT elevation between groups. Incidence of hepatic decompensation events was similar among groups except for subjects who received non standard doses of FPV, though the number of events was small.

Conclusions: Fosamprenavir/ritonavir discontinuation rate due to AEs or ALT elevation was similar across all European-approved FPV/r doses and to that of LPV/r subjects.  Although liver tolerated antiretrovirals, such as integrase inhibitor and entry inhibitor, the use of FPV/r is acceptable in HIV infected patients with viral hepatitis.

Keywords:

• Fosamprenavir
• Hepatic impairment
• HCV
• HIV
• Lopinavir
• Viral hepatitis

Introduction

Hepatitis coinfection is prevalent among HIV-infected patients. It is estimated that 10% of HIV-positive (HIV+) persons worldwide are coinfected with chronic hepatitis B virus (HBV), and 25% are coinfected with chronic hepatitis C virus (HCV), with large geographic variability.¹ Liver disease is now the leading cause of death, after AIDS-related complications, among HIV + individuals with access to highly active antiretroviral therapy (HAART).^2,3] Although HAART use is associated with slower rates of hepatic disease progression, coinfected patients have a higher rate of drug-induced liver injury compared with HIV-monoinfected patients. This risk may be up to eightfold higher despite HAART initiation, especially under certain parameters (e.g. low CD4 + T-cell count, moderate-to-severe liver disease, HCV genotype 3).^4,5 More data on the use of antiretroviral drugs in HIV/hepatitis-coinfected patients are needed, along with adequate pharmacokinetic (PK) studies in patients with advanced liver disease.

Fosamprenavir (FPV) is the phosphate ester prodrug of the HIV-1 protease inhibitor amprenavir (APV).⁶ Amprenavir is primarily eliminated via hepatic metabolism by CYP3A4 with little of the dose (<1%) excreted as unchanged drug in urine. Similar to other protease inhibitors, APV exposures increase with coadministration of ritonavir (RTV), a potent CYP3A4 inhibitor. Fosamprenavir boosted or unboosted by RTV is commonly used in HIV/hepatitis-coinfected patients in different stages of liver disease, when therapeutic drug monitoring is available.^7,8 It is also frequently used by clinicians because it offers reductions in pill number during prolonged treatment, which is typically required in these patients. APV10017 was a PK study that evaluated the PK, safety, and tolerability of FPV boosted with RTV (FPV/r) at reduced doses (FPV 700 mg twice daily [BID]/RTV 100 mg once daily [QD] or FPV 300 mg BID/RTV 100 mg QD) over 14 days in HIV-infected subjects with mild-to-severe hepatic impairment (HI) compared with matched controls with normal hepatic function who received the standard FPV/r dose (i.e. FPV 700 mg/RTV 100 mg BID). The study demonstrated that APV clearance was significantly slower in subjects with HI compared with those with normal hepatic function, suggesting that a reduction in the standard dose of FPV/r may be necessary to achieve plasma APV exposures similar to that in subjects with normal hepatic function.⁹ On the basis of data from APV10017, new regimens for use in HIV patients were approved in Europe in 2007 (for mild/moderate HI) and 2009 (for severe HI): (1) FPV 700 mg BID/RTV 100 mg QD in patients with mild HI (Child–Pugh score 5–6); (2) FPV 450 mg BID/RTV 100 mg QD in those with moderate HI (Child–Pugh score 7–9); and (3) FPV 300 mg BID/RTV 100 mg QD in those with severe HI (Child–Pugh score 10–15). The objective of this analysis was to collect data from three established European cohorts of HIV-infected subjects with a high prevalence of hepatitis coinfection: the ICONA Foundation (Italy), the ANRS CO13-HEPAVIH (France), and the MASTER (Italy) cohorts and compare safety outcomes of subjects taking different doses of FPV/r or lopinavir/ritonavir (LPV/r) as a comparator.

Materials and methods

This was an observational, retrospective, multicohort study. Data collected in the cohorts were reflective of routine clinical practice, and no specific interventions were introduced as a consequence of being enrolled in these studies. The characteristics of the three contributing cohorts have been described in previous publications.^10–12 All subjects provided written consent to participate in the cohort studies, following procedures in accordance with the ethical standards of the responsible committee on human experimentation and the Helsinki Declaration. Subjects’ informed consent included a specification for use of samples for further HIV research, and informed consents were approved by the ethics committees of the various clinical sites. Therefore, no specific consent for inclusion in the current analysis was needed. Data were merged following a strict standard operating procedure manual and using standardized data collection and coding (HICDEP; http://www.hicdep.org/).

Subjects with viral hepatitis coinfection defined as xxx with HCV Ab positive and or HBsAg positive, who initiated FPV/r treatment on or after 1 January 2008 (baseline), or LPV/r treatment on or after 1 January 2007, were included in this analysis. People who started an FPV/r-based regimen followed by an LPV/r-based regimen (or vice versa) were allocated to the drug group that was started first and remained in this allocated group for the analysis according to an intention-to-treat principle. Subjects included in the FPV/r group or LPV/r group had to be free of FPV/r or LPV/r exposure, respectively, during the year before baseline.

Statistical analyses

Eight endpoints were identified: (1) discontinuation of FPV/r or LPV/r; (2) discontinuation of ≥1 antiretroviral (ARV) drug included in FPV/r- or LPV/r-based regimen; (3) alanine aminotransferase (ALT) elevation >200 IU/L; (4) composite clinical endpoint of treatment-emergent hepatic decompensation, defined as onset/progression of ascites, encephalopathy, jaundice, gastrointestinal (GI) bleeding, serious bacterial infection (spontaneous bacterial peritonitis, pneumonia, bacteremia), hepatorenal syndrome, or death; (5) same as endpoint 1 but restricted to discontinuations due to adverse events (AEs); (6) same as endpoint 2 but restricted to discontinuations due to AEs, (7)risk of HIV viral failure, defined as HIV_RNA >200 copies/ml, twice consecutively in patients, with detectable HIV-RNA at baseline, who started the new regimen for at least six months; and (8) risk of viral rebound, defined as HIV-RNA >200 copies/ml after switch, in patients with viral suppression (HIV-RNA <200 copies/ml) at baseline.

Alanine aminotransferase elevation was defined as (1) time of first ALT value >200 IU/L (single value above this threshold) for subjects with baseline ALT ≤200 IU/L; and (2) time of first value which was 50% higher than baseline value for patients with baseline ALT >200 IU/L. Notably, events were defined using 40 IU/L as the “normal” value for subjects (regardless of sex) so that a level of 200 IU/L represented a value that was fivefold greater than normal. This is a standard pragmatic definition in analysis of data from cohort studies in which it is not generally possible to establish the actual upper limit of normal used at the various local laboratories. In addition, an endpoint entailing two consecutive values >200 IU/L was also considered.

We defined drug discontinuation as due to AEs if one of the following reasons was reported for discontinuation: dyslipidemia, GI tract toxicity, central nervous system (CNS) toxicity, renal toxicity, lipodystrophy, or unspecified side effects/toxicity. Any time an ARV drug was discontinued, it was counted as an event, regardless of how many other ARV drugs were discontinued at the same time (i.e. no distinction between partial and total discontinuations). Discontinuation of the boosting dose of RTV was also counted as a drug discontinuation. The main reason for discontinuation reported was considered when there were multiple reasons.

Subjects were stratified into the following five groups: normal hepatic function, defined by receipt of FPV 700 mg BID/RTV 100 mg BID (referred to hereafter as “standard dose”) and baseline aspartate aminotransferase (AST) platelet ratio index (APRI) score <2.0 (Group A); mild HI, defined by receipt of the recommended reduced FPV/r dose, 700 mg BID/100 mg QD, (referred to as “reduced dose”; Group B); non-standard dose of FPV different from those taken by subjects in Groups A, B, and D (Group C); standard dose despite evidence of abnormal hepatic function, defined by baseline APRI score ≥2.0 (Group D); and standard dose of LPV/r (Group E). The aim was to include at least three LPV/r subjects per every FPV/r subject in the analysis (Groups A + B + C + D). For Groups A and D, baseline APRI was calculated as the mean of two APRI scores using the two most recent values among those measured during the year before starting the FPV/r-based regimen.

Subject characteristics were described across the five groups and compared using χ² or Wilcoxon test as appropriate. Incidence rates for each endpoint were calculated as number of events divided by person-years of follow-up (PYFU) with 95% confidence intervals (CIs) using a Poisson approximation. A standard time-to-event analysis was employed for all six endpoints. Patients’ follow-up in each group accrued from baseline until the date of occurrence of the endpoint in question or last clinical follow-up. Kaplan–Meier curves and Cox proportional hazards regression were used to compare incidence rates in the five groups (A–E). Lopinavir/ritonavir subjects (Group E) were chosen as the comparator group. The following factors were considered for inclusion in the multivariable analysis: age; sex; mode of HIV transmission; weight; time since first HIV + test; ART status at baseline (naive or experienced); ART before baseline (specific drugs used, time spent on ART); baseline hepatitis B surface antigen, hepatitis B e antigen, HBV DNA and HCV RNA (where available); baseline AST, ALT, total bilirubin, and platelet count; HIV-1 RNA; CD4 count; non-ARV medications and start/stop date; baseline history of hepatic decompensation and AIDS diagnosis before baseline; and APRI, FIB-4, Meld, and Child–Pugh scores.

We also aimed to identify which of these factors were independently associated with risk of developing events of interest. No data were available on dose modifications of FPV-based regimens after the dose was used as initial treatment. Child–Pugh scores were not available for all cirrhotic subjects; therefore, scores were estimated for cirrhotic subjects using information available in the database. Cirrhotic subjects were defined as those with one or more of the following conditions at baseline: decompensated liver disease, APRI >2¹³, or FIB-4 >3.25.¹⁴ Scores were calculated on the basis of the presence/severity of ascites, encephalopathy, serum bilirubin, serum albumin, and blood clotting times (prothrombin time, international normalized ratio [PT INR]).¹⁵ Prothrombin time INR was available only for a limited number of subjects, and subjects were assumed to have a normal value (<1.7) when unknown. All data manipulation and analyses were conducted using SAS^® v9.2 (Statistical Analysis Software, Cary, NC, USA).

Results

A total of 1096 subjects were included in this analysis and followed up to the end of January 2012. There were 144 patients in Group A, 66 in B, 18 in C, 50 in D, and 818 in the comparator (group E). Median clinical follow-up in the 5 groups was 258 PYFU in Group A, 115 in B, 27 in C, 77 in D, and 1941 in E. Main characteristics of the study population by treatment group are shown in Table 1. As expected, subjects with APRI ≥2 (Group D) showed significantly higher baseline ALT, AST, and bilirubin levels and APRI, FIB-4, Child–Pugh and Meld scores, while CD4 counts and platelet levels were lower compared with those in other groups. Lopinavir/ritonavir subjects, by design, initiated LPV/r on average one year earlier than subjects in some of the FPV/r groups.

Table 1 Main characteristics of study population (n = 1096) by treatment group

	Treatment groups
	FPV/r				LPV/r
Characteristics	A 700 mg BID/ RTV 100 mg BID, APRI <2.0 (N = 144)	B 700 mg BID/ RTV 100 mg QD (N = 66)	C Other dose^a (N = 18)	D 700 mg BID/ RTV 100 mg BID, APRI ≥2.0 (N = 50)	E Standard dose (N = 818)	Total (N = 1096)	P value
Cohort, n (%)
MASTER	123 (85.4%)	61 (92.4%)	6 (33.3%)	45 (90.0%)	554 (67.7%)	789 (72.0%)	<0.001^b
HEPAVIH	6 (4.2%)	5 (7.6%)	4 (22.2%)	0 (0.0%)	112 (13.7%)	127 (11.6%)
ICONA	15 (10.4%)	0 (0.0%)	8 (44.4%)	5 (10.0%)	152 (18.6%)	180 (16.4%)	<0.001^c
Age, median (IQR), y	45 (40, 48)	45 (41, 48)	45 (41, 47)	45 (41, 48)	42 (37, 46)	43 (38, 46)	0.543^b
Gender, female, n (%)	36 (25.0%)	18 (27.3%)	3 (16.7%)	8 (16.0%)	208 (25.4%)	273 (24.9%)	<0.001^b
Time on ART, median (IQR), y	5 (0, 9)	7 (1, 11)	6 (0, 10)	7 (2, 10)	4 (0, 9)	5 (0, 9)	0.015^c
AIDS before baseline, n (%)	96 (66.7%)	49 (74.2%)	6 (33.3%)	28 (56.0%)	415 (50.7%)	594 (54.2%)	<0.001^b
Died over follow-up, n (%)	8 (5.6%)	1 (1.5%)	3 (16.7%)	7 (14.0%)	39 (4.8%)	58 (5.3%)	0.006^b
HCV-Ab positive, n (%)	126 (87.5%)	63 (95.5%)	17 (94.4%)	49 (98.0%)	710 (86.8%)	965 (88.0%)	0.039^b
HBs-Ag positive, n (%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	1 (0.1%)	1 (0.1%)	0.987^b
HBe-Ag positive, n (%)	1 (0.7%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	8 (1.0%)	9 (0.8%)	0.847^b
APRI score, median (range)	0.66 (0.13, 2.00)	0.76 (0.14, 17.90)	0.74 (0.31, 10.42)	3.79 (2.03, 17.37)	0.59 (0.04, 74.81)	0.63 (0.04, 74.81)	<0.001^c
FIB score, median (IQR)	1.61 (1.20, 2.58)	1.65 (1.05, 3.49)	2.63 (1.51, 8.71)	7.36 (4.48, 11.71)	1.49 (1.07, 2.53)	1.61 (1.12, 2.93)	<0.001^c
MELD score, median (IQR)	5.66 (1.85, 9.56)	2.87 (0.85, 6.01)	3.39 (0.59, 6.68)	7.45 (6.43, 10.5)	3.10 (1.00, 6.34)	3.82 (1.00, 6.96)	<0.001^c
ALT, median (IQR), IU/L	53 (35, 92)	60 (47, 95)	47 (31, 740)	111 (67, 198)	49 (30, 81)	52 (32, 89)	<0.001^c
AST, median (IQR), IU/L	47 (33, 72)	45 (32, 84)	57 (38, 110)	140 (99, 182)	42 (29, 67)	45 (30, 74)	<0.001^c
Length of follow-up, median (IQR), y	1.97 (0.78, 2.64)	1.69 (1.01, 2.45)	1.15 (0.65, 2.47)	1.74 (0.38, 2.41)	2.53 (1.51, 3.43)	2.31 (1.26, 3.24)	<0.001^c

ALT, alanine aminotransferase; APRI, AST platelet ratio index; ART, antiretroviral therapy; AST, aspartate aminotransferase; BID, twice daily; FPV, fosamprenavir; HBe-Ag, hepatitis B e antigen; HBs-Ag, hepatitis B surface antigen; HCV-Ab, HCV antibody; IQR, interquartile ratio; LPV, lopinavir; QD, once daily; RTV, ritonavir.

^a Either FPV 700 mg QD + RTV 100 mg QD (n = 5) or FPV 1400 mg QD (n = 14).

^b χ² or Fisher’s exact test as appropriate.

^c Kruskal–Wallis test.

Discontinuation of FPV/r or LPV/r alone

A total of 517 discontinuations of FPV/r or LPV/r were observed over 1636 PYFU for an overall incidence rate of 32 per 100 PYFU (95% CI, 29–34). The most common reason for FPV/r discontinuation (where reason was known) was subject decision (23.1% across all FPV groups compared with 18.9% in the LPV/r group). More FPV/r subjects discontinued because of structured treatment interruptions (STIs) than LPV/r subjects (10.0% and 15.2% among those in Groups D and B, respectively compared with 3.6% of LPV/r subjects). Approximately one-third of subjects in both the FPV/r and LPV/r groups discontinued for unknown causes.

All FPV/r subjects, except those with APRI ≥2 taking the standard dose, showed an increased risk of discontinuation compared with LPV/r subjects (Table 2) In addition, baseline CD4 count was also associated with risk of discontinuing FPV/r or LPV/r (the higher the count, the lower the risk of discontinuation). These associations remained or became statistically significant after adjustment for potential confounders. The analysis was also performed using FIB-4 score and gave similar results. As per analysis plan, we also performed this analysis after restricting to discontinuations due to AEs alone and using a competing risk approach to analysis.

Table 2 RH of risk of discontinuing FPV/r by dose compared with risk of discontinuing LPV/r from fitting a multivariable Cox regression model (CD4 and platelet counts fitted as time-fixed covariates at baseline)

	Crude and adjusted RH of stopping FPV/r or LPV/r
Characteristic	Crude RH (95% CI)	P value	Adjusted^a RH (95% CI)	P value
Treatment group (total events, N = 517)
(A) FPV 700 mg BID/RTV100 mg BID and APRI <2 (N = 70)	1.48 (1.14, 1.92)	0.003	1.56 (1.18, 2.06)	0.002
(B) FPV 700 mg BID/RTV 100 mg QD (N = 33)	1.45 (1.01, 2.09)	0.043	1.53 (1.05, 2.23)	0.025
(C) FPV other dose (N = 8)	1.49 (0.74, 3.03)	0.265	2.05 (0.99, 4.22)	0.052
(D) FPV 700 mg BID/RTV100 mg BID and APRI ≥2 (N = 19)	1.09 (0.69, 1.74)	0.712	0.97 (0.59, 1.60)	0.913
(E) LPV/r (N = 387)	1.0		1.0
Sex
Male	1.0		1.0
Female	1.27 (1.05, 1.54)	0.015	1.25 (1.00, 1.56)	0.053
Mode of HIV transmission
IDU	1.0		1.0
Homosexual contacts	1.04 (0.75, 1.44)	0.830	1.22 (0.84, 1.77)	0.300
Heterosexual contacts	1.17 (0.95, 1.44)	0.144	1.18 (0.94, 1.50)	0.157
Other/unknown	1.24 (0.87, 1.78)	0.241	1.54 (0.99, 2.39)	0.056
Baseline CD4 count
Per 100 cells/mm^3 higher	0.96 (0.92, 1.00)	0.035	0.95 (0.92, 1.00)	0.030
Baseline APRI score
Per 1 unit higher	1.01 (0.99, 1.03)	0.250	1.00 (0.98, 1.02)	0.851
ART naive
No	1.0		1.0
Yes	1.00 (0.80, 1.24)	0.973	0.81 (0.62, 1.07)	0.136
Baseline platelets
Per 100 109/L higher	0.90 (0.80, 1.00)	0.058	0.92 (0.82, 1.04)	0.168
Use of non-ARV drugs at baseline
No	1.0		1.0
Yes	1.17 (0.83, 1.66)	0.369	1.03 (0.68, 1.57)	0.885
Baseline bilirubin
Per mg/dL higher	0.98 (0.95, 1.02)	0.300	0.98 (0.94, 1.02)	0.235
Baseline FIB-4 score
Per 1 unit higher	1.00 (0.99, 1.01)	0.548
Baseline hepatic decompensation
No	1.0		1.0
Yes	1.85 (1.12, 3.05)	0.016	1.45 (0.84, 2.52)	0.185

APRI, AST platelet ratio index; ART, antiretroviral therapy; ARV, antiretroviral; AST, aspartate aminotransferase; BID, twice daily; CI, confidence interval; FPV, fosamprenavir; IDU, intravenous drug use; LPV, lopinavir; QD, once daily; RH, relative hazard; RTV, ritonavir.

All models are stratified by cohort study.

^a Adjusted for the factors shown in table.

In summary, there were 112 discontinuations of FPV/r or LPV/r due to AEs: 4 lipodystrophy, 7 dyslipidemia, 1 hypersensitivity reaction, 28 abdomen and GI tract toxicities, 1 CNS toxicity, 1 renal toxicity, 7 other unspecified side effects, and 63 other unspecified toxicities. When details were given for the unspecified toxicities, the extra information collected after data merging was generic: “toxicity grade II” for 53 subjects (84%), “toxicity grade IV” for 3 subjects (5%), and “concomitant pathology” for the remaining 3 subjects (5%). There was no difference in rate of discontinuation due to AEs by group (adjusted relative hazard [RH] and 95% CI in different groups: A = 0.78 [0.37–1.64], p = 0.514; B = 1.36 [0.61–3.01], p = 0.448; C = 0, p = 0.979; D = 1.87 [0.77–4.51], p = 0.164 compared with the LPV/r group). None of the other factors considered showed an association with risk of discontinuing because of AEs.

Discontinuation of ≥1 drug included in the FPV/r- or LPV/r-based regimens

A total of 634 subjects discontinued ≥1 drug of those used at baseline over 1459 PYFU for an overall incidence rate of 43 per 100 PYFU (95% CI, 41–46). Besides FPV and LPV, RTV (including both full dose and use as booster), tenofovir (22%), emtricitabine (22%), and lamivudine (19%) were the most frequently discontinued drugs in all groups. The most common reasons for discontinuation of any drug in FPV-based regimens were patient decision (n = 34, 22%), AEs (n = 33, 20.9%), and STIs (n = 24, 15%). Results were similar to those for discontinuation of FPV/r or LPV/r alone. Subjects taking the reduced dose of FPV were 64% more likely to discontinue any drug in the FPV/r-based regimen compared with those taking the LPV/r-based regimen. Notably, subjects taking the FPV/r standard dose also had a 55% increased risk of discontinuation compared with those on LPV/r and, although this was not formally tested, it is unlikely that a large difference between the 2 doses of FPV/r exist (Table 3). In addition, the analysis was performed using FIB-4 score instead of APRI and gave similar results. There were 132 discontinuations of ≥1 drug due to AEs (5 lipodystrophy, 5 dyslipidemia, 1 hypersensitivity reaction, 23 abdomen and GI tract toxicities, 1 CNS toxicity, 2 renal toxicities, 11 unspecified side effects, 3 hematological, and 81 other unspecified toxicities). For the 81 unspecified toxicities, again, only generic information was available: “toxicity grade II” for 67 subjects (83%), “toxicity grade IV” for 7 subjects (8%), “concomitant pathology” for 7 patients (8%), and no information for the remaining 6 subjects. When restricting the analysis to these discontinuations for AEs alone, only Group B showed an increased risk of discontinuing any drug in their regimen (adjusted RH = 2.05; 95% CI, 1.10–3.84; p = 0.024) compared with LPV/r subjects. All other FPV/r groups showed a non-statistically significant increased risk (adjusted RH and 95% CI in A: 1.11 [0.61–2.02], p = 0.730; D: 1.30 [0.51–3.34], p = 0.583). No AEs were described in Group C.

Table 3 RH of risk of discontinuing ≥1 drug in FPV/r-containing regimens compared with risk of discontinuing ≥1 drug in the LPV/r-containing regimen from fitting a multivariable Cox regression model (CD4 and platelet counts fitted as time-fixed covariates at baseline)

	Crude and adjusted RH of stopping ≥1 drug in regimen
Characteristic	Crude RH (95% CI)	P value	Adjusted^a RH (95% CI)	P value
Treatment group (total events, N = 634)
(A) FPV 700 mg BID/RTV 100 mg BID and APRI <2 (N = 85)	1.54 (1.21, 1.95)	<0.001	1.55 (1.20, 2.00)	<0.001
(B) FPV 700 mg BID/RTV 100 mg QD (N = 42)	1.56 (1.14, 2.16)	0.006	1.64 (1.18, 2.28)	0.003
(C) FPV other dose (N = 9)	1.44 (0.74, 2.81)	0.280	1.48 (0.72, 3.02)	0.286
(D) FPV 700 mg BID/RTV100 mg BID and APRI ≥2 (N = 22)	1.10 (0.71, 1.69)	0.679	1.04 (0.66, 1.64)	0.866
(E) LPV/r (N = 476)	1.0		1.0
Sex
Male	1.0		1.0
Female	1.27 (1.07, 1.52)	0.006	1.26 (1.03, 1.54)	0.026
Mode of HIV transmission
IDU	1.0		1.0
Homosexual contacts	1.13 (0.85, 1.52)	0.398	1.33 (0.95, 1.85)	0.100
Heterosexual contacts	1.15 (0.95, 1.39)	0.152	1.19 (0.97, 1.47)	0.099
Other/unknown	1.17 (0.84, 1.63)	0.349	1.44 (0.96, 2.15)	0.078
Baseline CD4 count
Per 100 cells/mm^3 higher	1.00 (0.97, 1.03)	0.939	0.99 (0.96, 1.03)	0.754
Baseline APRI score
Per 1 unit higher	1.00 (0.99, 1.02)	0.695	1.00 (0.98, 1.02)	0.799
ART naive
No	1.0		1.0
Yes	0.91 (0.74, 1.11)	0.362	0.80 (0.62, 1.02)	0.074
Baseline platelets
Per 100 109/L higher	0.95 (0.87, 1.04)	0.269	0.95 (0.86, 1.05)	0.279
Use of non-ARV drugs at baseline
No	1.0		1.0
Yes	1.12 (0.81, 1.54)	0.487	1.03 (0.70, 1.52)	0.875
Baseline bilirubin
Per mg/dL higher	0.98 (0.95, 1.01)	0.185	0.97 (0.93, 1.01)	0.134
Baseline FIB-4 score
Per 1 unit higher	1.00 (0.99, 1.01)	0.942
Baseline hepatic decompensation
No	1.0		1.0
Yes	1.62 (1.01, 2.60)	0.044	1.38 (0.83, 2.30)	0.220

APRI, AST platelet ratio index; ART, antiretroviral therapy; ARV, antiretroviral; AST, aspartate aminotransferase; BID, twice daily; FPV, fosamprenavir; IDU, intravenous drug use; LPV, lopinavir; QD, once daily; RH, relative hazard; RTV, ritonavir.

All models are stratified by cohort study.

^a Adjusted for the factors shown in table.

ALT elevation

A total of 124 patients experienced an ALT elevation >200 IU/L (defined as 1 result after baseline) over 2212 PYFU for an overall incidence rate of 5.6 per 100 PYFU (95% CI, 4.7–6.6). Using the 2 consecutive values definition, only 40 subjects would have been declared to have an ALT elevation event; hence, this analysis was not conducted as it would have been underpowered. Results are shown in Table 4. The only tendency observed was that subjects receiving the FPV/r standard dose had an increased risk of ALT elevation compared with subjects receiving LPV/r. This association was only borderline statistically significant (p = 0.065), and the CIs around the point estimates overlapped with those of the other FPV/r groups. Again, a further analysis adjusting for FIB-4 score instead of APRI score produced similar results.

Table 4 RH of risk of ALT elevation (single value >200 IU/L) from fitting a multivariable Cox regression model (CD4 and platelet counts fitted as time-fixed covariates at baseline)

	Crude and adjusted RH of ALT elevation >200 IU/L
Characteristic	Crude RH (95% CI)	P value	Adjusteda RH (95% CI)	P value
Treatment group ((total events, N = 124)
(A) FPV 700 mg BID/RTV 100 mg BID and APRI <2  N=18)	1.37 (0.82, 2.28)	0.230	1.65 (0.97, 2.80)	0.065
(B) FPV 700 mg BID/RTV 100 mg QD (N = 8)	1.24 (0.60, 2.57)	0.568	1.21 (0.57, 2.54)	0.619
(C) FPV other dose (N = 1)	0.87 (0.12, 6.29)	0.888	1.03 (0.14, 7.72)	0.977
(D) FPV 700 mg BID/RTV 100 mg BID and APRI ≥2 (N = 91)	1.39 (0.61, 3.19)	0.438	1.84 (0.76, 4.42)	0.175
(E) LPV/r	1.0		1.0
Sex
Male	1.0		1.0
Female	0.75 (0.48, 1.17)	0.203	0.70 (0.42, 1.15)	0.160
Mode of HIV transmission
IDU	1.0		1.0
Homosexual contacts	0.75 (0.36, 1.56)	0.441	0.45 (0.18, 1.14)	0.092
Heterosexual contacts	0.83 (0.53, 1.30)	0.421	0.89 (0.55, 1.46)	0.655
Baseline CD4 count
Per 100 cells/mm^3 higher	0.96 (0.89, 1.04)	0.352	0.98 (0.90, 1.07)	0.657
Baseline APRI score
Per 1 unit higher	1.01 (0.98, 1.05)	0.403	1.01 (0.98, 1.05)	0.460
ART naive
No	1.0		1.0
Yes	1.74 (1.17, 2.60)	0.006	2.04 (1.28, 3.27)	0.003
Baseline platelets
Per 100 10^9/L higher	0.94 (0.75, 1.16)	0.558	0.99 (0.80, 1.22)	0.926
Use of non-ARV drugs at baseline
No	1.0		1.0
Yes	2.07 (0.90, 4.73)	0.087	1.66 (0.61, 4.51)	0.322
Baseline bilirubin
Per mg/dL higher	0.95 (0.82, 1.09)	0.442	0.96 (0.85, 1.09)	0.536
Baseline FIB-4 score
Per 1 unit higher	0.99 (0.96, 1.03)	0.715
Baseline hepatic decompensation
No	1.0		1.0
Yes	0.83 (0.20, 3.36)	0.794	0.70 (0.16, 2.98)	0.629

ALT, alanine aminotransferase; APRI, AST platelet ratio index; ART; antiretroviral therapy; ARV, antiretroviral; AST, aspartate aminotransferase; BID, twice daily; FPV, fosamprenavir; IDU, intravenous drug use; LPV, lopinavir; QD, once daily; RH, relative hazard; RTV, ritonavir.

All models are stratified by cohort study.

^a Adjusted for the factors shown in table.

Severe hepatic disease and death

Twenty-eight events of hepatic decompensation were reported at baseline (6 in Group A, 2 in B, 3 in C, 7 in D, and 10 in subjects on LPV/r). Twenty events of hepatic decompensation were observed after baseline, all leading to death, over a total clinical follow-up of 2418 person-years (rate = 0.83 per 100 PYFU; 95% CI, 0.51–1.27). There were 38 additional deaths due to other reasons, 8 of which occurred in the FPV/r groups (1 non-AIDS-defining cancer in the reduced dose group), and the remaining 30 occurred in the LVP/r group. Overall, there were 58 events of hepatic decompensation or death over follow-up, only 1 of which occurred in the FPV/r-reduced dose group. Table 5 shows the breakdown with all causes of death by treatment group. Because of the limited number of clinical events, the analysis was underpowered to perform a formal comparison of the rate of disease progression between groups.

Table 5 Causes of death by treatment group

Cause of death, n (%)	Treatment groups
	FPV/r				LPV/r
	A	B	C	D	E
	700 mg BID/RTV 100 mg BID, APRI <2.0	700 mg BID/RTV 100 mg QD	Other dose	700 mg BID/RTV 100 mg BID, APRI ≥2.0	Standard dose	Total number of deaths
Total deaths	N = 8	N = 1	N = 3	N = 7	N = 39	58
Deaths due to hepatic decompensation	4 (50%)	0 (0%)	2 (67%)	5 (71%)	9 (23%)	20
Deaths for other reasons	N = 4	N = 1	N = 1	N = 2	N = 30	38
Cardiovascular	1 (13%)	0 (0%)	0 (0%)	0 (0%)	1 (3%)	2
HIV related	1 (13%)	0 (0%)	0 (0%)	0 (0%)	13 (33%)	14
Non-AIDS malignancies	0 (0%)	1 (100%)	0 (0%)	2 (29%)	2 (5%)	5
Accidental	0 (0%)	0 (0%)	0 (0%)	0 (0%)	3 (8%)	3
Unknown	2 (26%)	0 (0%)	1 (33%)	0 (0%)	11 (28%)	14

APRI, AST platelet ratio index; AST, aspartate aminotransferase; BID, twice daily; FPV, fosamprenavir; LPV, lopinavir; QD, once daily; RTV, ritonavir.

HIV Viral failure and rebound

Both HIV viral failure and rebound were investigated among groups (Tables 6 and 7). In total, 172 subjects experienced viral failure over 1,906 PYFU for an overall incidence rate of 9.0 per 100 PYFU (95% CI: 7.7–10.5). Only FPV/r subjects receiving the standard dose with APRI ≥2 showed a twofold risk of viral failure compared with LPV/r subjects (APRI-adjusted RH = 2.11; 95% CI, 1.08–4.13; p = 0.03). ART-naive subjects were approximately 50% less likely to experience viral failure (adjusted RH = 0.54; 95% CI, 0.32–0.89; p = 0.015, Table 6). Overall, 54 subjects of the total 172 who had viral failure started treatment with a VL ≤200 copies/mL and experienced viral rebound over 838 PYFU for an overall incidence rate of 6.4 per 100 PYFU (95% CI:4.8–8.4). There was no difference in viral rebound among FPV patients on the standard or reduced FPV dosage, compared to LPV/r users; however, a fivefold elevated risk was seen among those on other FPV doses (adjusted RH = 5.68; 95% CI, 0.1.21–26.65; p = 0.028, Table 7). For both viral failure and rebound, analyses were repeated adjusting for FIB-4 score instead of APRI score and similar results were seen.

Table 6 Variables related with risk of HIV viral failure after starting the new regimen.

	Crude and adjusted relative hazards of viral failure >200 copies/mL
Characteristic	Crude RH (95% CI)	p-value	Adjusted* RH (95% CI)	p-value	Adjusted** RH (95% CI)	p-value
Treatment group
LPV/r	1.0		1.0		1.0
FPV 700 mg BID/RTV100 mg BID and APRI <2	1.44 (0.94, 2.21)	0.092	1.49 (0.96, 2.31)	0.074	1.49 (0.96, 2.30)	0.075
FPV 700 mg BID/RTV 100 mg QD	1.48 (0.81, 2.70)	0.202	1.50 (0.79, 2.82)	0.212	1.50 (0.79, 2.83)	0.211
FPV 700 mg BID/RTV100 mg BID and APRI>=2	2.04 (1.09, 3.79)	0.025	2.10 (1.08, 4.12)	0.030	2.12 (1.09, 4.15)	0.028
FPV other dose	1.82 (0.44, 7.47)	0.408	2.56 (0.60, 10.80)	0.202	2.55 (0.60, 10.77)	0.203
Gender
Male	1.0		1.0		1.0
Female	1.58 (1.15, 2.18)	0.005	1.47 (1.03, 2.08)	0.032	1.47 (1.03, 2.09)	0.032
Mode of HIV transmission
IDU	1.0		1.0		1.0
Homosexual contacts	0.90 (0.50, 1.65)	0.743	1.27 (0.69, 2.36)	0.444	1.27 (0.69, 2.35)	0.448
Heterosexual contacts	0.93 (0.65, 1.34)	0.699	0.88 (0.60, 1.30)	0.535	0.89 (0.60, 1.30)	0.538
Other/unknown	0.61 (0.25, 1.49)	0.277	0.95 (0.37, 2.39)	0.906	0.94 (0.37, 2.39)	0.903
Baseline CD4 count
per 100 cells/mm3 higher	0.96 (0.90, 1.02)	0.195	0.93 (0.87, 1.00)	0.054	0.93 (0.87, 1.00)	0.054
Baseline APRI score
per 1 unit higher	1.01 (0.99, 1.04)	0.331	1.01 (0.98, 1.04)	0.599
ART-naive
No	1.0		1.0		1.0
Yes	0.46 (0.29, 0.74)	0.001	0.54 (0.32, 0.89)	0.015	0.54 (0.32, 0.89)	0.015
Baseline platelets
Per 100 109/l higher	0.95 (0.79, 1.13)	0.534	0.99 (0.85, 1.16)	0.925	0.99 (0.85, 1.16)	0.913
Use of non-ARV drugs at baseline
No	1.0		1.0		1.0
Yes	0.84 (0.39, 1.82)	0.667	0.95 (0.43, 2.09)	0.897	0.95 (0.43, 2.08)	0.891
Baseline bilirubin
Per mg/dl higher	0.90 (0.77, 1.05)	0.190	0.87 (0.73, 1.03)	0.111	0.87 (0.73, 1.03)	0.111
Baseline FIB score
per 1 unit higher	1.01 (0.99, 1.02)	0.411			1.00 (0.99, 1.02)	0.689
Baseline liver disease
No	1.0		1.0		1.0
Yes	1.39 (0.52, 3.75)	0.514	1.09 (0.39, 3.08)	0.868	1.09 (0.39, 3.07)	0.871

All models are stratified by cohort study.

^* Adjusted for factors shown in table.

^** Adjusted for FIB score instead of APRI score.

Table 7 Variables related with risk of HIV viral rebound after starting the new regimen in patient with HIV-RNA <200 copies/ml at baseline

	Crude and adjusted relative hazards of viral rebound >200 copies/mL
Characteristic	Crude RH (95% CI)	p-value	Adjusted* RH (95% CI)	p-value	Adjusted** RH (95% CI)	p-value
Treatment group
LPV/r	1.0		1.0		1.0
FPV 700 mg BID/RTV100 mg BID and APRI <2	1.55 (0.73, 3.27)	0.254	1.75 (0.81, 3.82)	0.157	1.73 (0.80, 3.78)	0.166
FPV 700 mg BID/RTV 100 mg QD	1.60 (0.61, 4.15)	0.338	1.74 (0.64, 4.72)	0.279	1.78 (0.66, 4.77)	0.253
FPV 700 mg BID/RTV100 mg BID and APRI> = 2	2.17 (0.84, 5.60)	0.110	2.79 (0.78, 9.94)	0.114	3.06 (0.85, 11.08)	0.088
FPV other dose	4.49 (1.04, 19.40)	0.044	5.68 (1.21, 26.65)	0.028	5.71 (1.22, 26.74)	0.027
Gender
Male	1.0		1.0		1.0
Female	1.42 (0.78, 2.58)	0.251	1.61 (0.81, 3.20)	0.176	1.65 (0.84, 3.25)	0.144
Mode of HIV transmission
IDU	1.0		1.0		1.0
Homosexual contacts	1.44 (0.56, 3.71)	0.453	2.23 (0.81, 6.15)	0.122	2.22 (0.81, 6.13)	0.123
Heterosexual contacts	0.74 (0.36, 1.54)	0.424	0.78 (0.36, 1.72)	0.543	0.77 (0.35, 1.70)	0.521
Other/unknown	0.47 (0.06, 3.44)	0.457	0.58 (0.07, 4.73)	0.611	0.56 (0.07, 4.60)	0.592
Baseline CD4 count
per 100 cells/mm^3 higher	1.00 (0.90, 1.10)	0.963	0.99 (0.89, 1.11)	0.881	0.99 (0.89, 1.11)	0.867
Baseline APRI score
per 1 unit higher	1.08 (0.96, 1.22)	0.213	1.02 (0.86, 1.21)	0.852
ART-naive
No	1.0		1.0		1.0
Yes	0.53 (0.07, 3.87)	0.534	0.76 (0.09, 6.05)	0.793	0.76 (0.10, 6.08)	0.799
Baseline platelets
Per 100 10^9/l higher	0.88 (0.62, 1.25)	0.481	0.93 (0.62, 1.40)	0.727	0.91 (0.59, 1.41)	0.673
Use of non-ARV drugs at baseline
No	1.0		1.0		1.0
Yes	5.69 (1.45, 22.29)	0.013	7.75 (1.91, 31.38)	0.004	7.77 (1.92, 31.50)	0.004
Baseline bilirubin
Per mg/dl higher	0.85 (0.65, 1.11)	0.231	0.80 (0.59, 1.08)	0.143	0.80 (0.59, 1.09)	0.152
Baseline FIB score
per 1 unit higher	1.04 (0.96, 1.12)	0.394			0.99 (0.87, 1.13)	0.907
Baseline liver disease
No	1.0		1.0		1.0
Yes	1.33 (0.32, 5.46)	0.697	1.20 (0.27, 5.36)	0.809	1.22 (0.27, 5.54)	0.796

All models are stratified by cohort study.

* Adjusted for factors shown in table.

** Adjusted for FIB score instead of APRI score.

Discussion

The main objective of the current analysis was to gather further real-world safety data on FPV/r regimens in HIV/hepatitis-coinfected patients, given that the most recent approved doses in this HI group had been approved largely on the basis of clinical pharmacology studies.

Safety clinical endpoints

Although severe hepatic events and death were infrequent in our study population, and therefore formal comparison of rates in these groups was not possible, it is noted that we recorded very few severe events in subjects receiving FPV/r. Only one death was observed in the reduced dose group; the fatality was not considered liver associated. Regarding the risk of ALT elevation, there was some evidence that subjects with APRI <2 in the FPV/r standard dose group were at increased risk of ALT elevation >200 IU/L compared with subjects in the LPV/r group (65% increase, close to statistical significance). Overall, there was little evidence for a difference between the groups, demonstrated by the wide CIs around the estimates of the relative hazards. Despite a mean follow-up of approximately two years per person, the power of this analysis was low. A limitation of this analysis is that ALT elevation was assessed using a suboptimal definition of this outcome based on a single value, >200 IU/L, which is likely to have lower specificity than more robust outcomes based on repeated values. However, the power would have been even lower if the 2 consecutive values definition had been used (40 events instead of the 124 used in this analysis).

Drug discontinuation endpoints

Rates of FPV/r discontinuation (all doses except perhaps the standard dose when used with APRI ≥2) were significantly higher than the rates of LPV/r discontinuation (Table 2). These results were confirmed in the Cox regression analysis after controlling for potential confounders including sex, modality of HIV transmission, baseline APRI score (FIB-4 score in a sensitivity analysis), CD4 count, platelet count, bilirubin levels, and previous history of ART use. Notably, the higher discontinuation rate for FPV/r-reduced dose subjects compared with LPV/r subjects was not due to a higher risk of discontinuing because of AEs. The percentage of subjects discontinuing because of AEs was similar in these 2 groups (21 vs. 23%).

Subjects on the reduced dose of FPV/r seemed to have discontinued more frequently than those on LPV/r because of treatment failure (12 vs. 6%).

Regarding the analysis of discontinuation of any drug included in the FPV/r- and LPV/r-based regimens, the most frequently discontinued drugs (besides FPV/LPV) were RTV, followed by the nucleosides tenofovir, emtricitabine, and lamivudine. In the FPV/r-reduced dose group, the most common reason for discontinuing ≥1 drug was structured treatment interruption, and reasons for this are unclear. In contrast to the analysis of reasons for discontinuing FPV/r alone, when we analyzed the rate of discontinuation of any drug in the regimens due to AEs, subjects receiving the FPV/r-reduced dose were at significantly higher risk of discontinuing other drugs compared with subjects receiving LPV/r. This seems to indicate that when the reduced dose of FPV/r was used in the regimen, it led to a more significant change in regimen than that observed for subjects discontinuing LPV/r and could be the result of confounding by indication or could reflect that this group is more challenging to manage initially. The strong limitation of the study is that there were sizable discontinuation rates with limited information on reason for discontinuation in the study. The high rate of gastrointestinal disturbances which characterized the regimen based on fosamprenavir could be a reason even if not specifically reported .Unfortunately, no further information on reasons could be provided upon subsequent investigation with the original reporting cohort. It is however noted that if reason was captured under the broad ‘subject decision’ category, that this almost certainly excludes toxicity as a reason for discontinuation as this is specifically detailed in the Case Report Form (CRF). The subject decision category may be reflective of issues with patient compliance due to dose changes following liver function tests.

The major strength of this analysis is the use of data collected in the observational setting, which allowed for the evaluation of response to a variety of FPV-based regimens in a larger number of subjects followed up over a longer period of time and better reflecting routine clinical practice than that of patients enrolled in clinical trials. One advantage of the longer follow-up is the possibility of collecting long-term clinical progression events. However, ALT elevation, severe hepatic disease, and death were rare, and a small number of events were observed over our average follow-up of two years.

There are some limitations to the results of this study. The prevalence of use of the reduced FPV/r dose as initial dose was lower than expected. Only 66 subjects (23% of total FPV/r subjects) used FPV 700 mg BID/RTV 100 mg QD, and none of the subjects appeared to have started the FPV 450/300 mg BID/RTV 100 mg QD dose. The dose for severe HI patients was only approved in March 2009; therefore, it is likely that full implementation of this recommendation has not yet been integrated into clinical practice, and for this reason, it is not reflected in our data-set.

In addition, the prevalence of subjects with estimated severe HI was less than 4% in our study population. The calculated Child–Pugh scores in the different treatment groups suggest that dosing was not driven by Child–Pugh scores alone, indicating that the dosing recommendations in the summary of product characteristics (SPC) were not always followed. Because of the introduction of the new recommendations, it is possible that the full dose of FPV/r was initially given, regardless of Child–Pugh score, until there was evidence from drug levels or clinical assessment that a dose adjustment was needed. As a result of the modality of data collection (e.g. discontinuation due to PK analysis [guided dose adjustment is not an explicit reason for stopping in the standardized HICDEP coding and only the initial dose was collected]), we were not able to capture whether dose modifications from the standard dose to these reduced doses took place and the impact they may have had on our observed results.

Finally, Child–Pugh values were not available for a large proportion of subjects and were estimated using incomplete information (prothrombin ratio and INR were seldom recorded in both subjects with and without cirrhosis) and therefore, it was not possible to address the issue of whether SPC recommendations were strictly followed. Because FVP/r subjects were not matched by design to those starting LPV/r-based regimens, there were imbalances at baseline between the considered groups even in measured factors. However, all associations between treatment groups and the studied outcomes that were significant in univariable analyses remained significant after adjusting for most factors in the multivariable analyses both when fitted as time-fixed or time-dependent factors. Results also remained unchanged after controlling for CD4 and platelet counts as time-dependent covariates. Nevertheless, since the study was not randomized, we cannot rule out bias introduced by unmeasured confounding, which is the main limitation of our study. In conclusion, our results show that in patients, receiving the FPV/r standard dose, with viral hepatitis coinfection, there is a borderline statistically significant (p = 0.065) increased risk of ALT elevation compared with subjects receiving LPV/r.

The analysis was underpowered for other severe clinical endpoints; nonetheless, very few events were observed in FPV/r subjects, regardless of dose.

Use of FPV/r (except for subjects with APRI ≥2 receiving the standard dose) was associated with a higher two-year risk of drug discontinuation compared with use of LPV/r. However, it is important to note that the reason for this difference was not due to a higher risk of discontinuing because of AEs. Rates of discontinuation of FPV/r due to AEs were similar across all FPV/r dose groups and to those of the LPV/r group, indicating an acceptable safety profile for both the standard and reduced dosage of FPV/r. Despite the recent introduction of new generation PIs, integrase inhibitors, and entry inhibitors, FPV/r provides another treatment option for individuals in this growing subset of HIV-infected patients.

Abbreviations
(AE)	=	adverse event
(ALT)	=	alanine aminotransferase
(APRI)	=	AST platelet ratio index
(ART)	=	antiretroviral therapy
(ARV)	=	antiretroviral
(AST)	=	aspartate aminotransferase
(BID)	=	twice daily
(CI)	=	confidence interval
(CNS)	=	central nervous system
(FPV)	=	fosamprenavir
(FPV/r)	=	fosamprenavir/ritonavir
(GI)	=	gastrointestinal
(HAART)	=	highly active antiretroviral therapy
(HBV)	=	hepatitis B virus
(HCV)	=	hepatitis C virus
(HI)	=	hepatic impairment
(HIV)	=	human immunodeficiency virus
(LPV)	=	lopinavir
(LPV/r)	=	lopinavir/ritonavir
(PK)	=	pharmacokinetic
(PT INR)	=	prothrombin time, international normalized ratio
(PYFU)	=	person-years of follow-up
(QD)	=	once daily
(RH)	=	relative hazard
(RTV)	=	ritonavir
(SPC)	=	summary of product characteristics
(STI)	=	structured treatment interruption

Funding

This work was supported by ViiV Heathcare.

Acknowledgements and Disclosures

Funding for this work was provided by ViiV Healthcare. All listed authors meet the criteria for authorship set forth by the International Committee for Medical Journal Editors. The authors wish to acknowledge the following individuals for their contributions:

From GlaxoSmithKline: We would like to thank Katharine Cheng, Teodora Perger and Jorg Sievers for their critical review of study reports throughout the study and comments on the manuscript.

From the MASTER Cohort: All patients participating in the cohort and Board of Directors: G Angarano, F Maggiolo, P Viale, F Castelli, A Scalzini, A Pan, L Sighinolfi, F Mazzotta, A Gori, L Minoli, R Cauda Scientific Secretary: C Torti, S Digiambenedetto, E Quiros, M Di Pietro.

Project Management: Nicola Mazzini.

Steering Committee: P Nasta, F Castelnuovo, E Quiros Roldan, S Locaputo, L Sighinolfi, N Ladisa, A Pan, G Lapadula, M Borderi, V Colangeli, S Novati, S Digiambenedetto, M Fabbiani.

Participating Physicians and Centers: F Castelli, G Paraninfo, S Casari, E Focà, E Quiros, P Nasta (Brescia), A Scalzini, I El-Hamad, F Castelnuovo (Brescia), R Cauda, S Digiambenedetto, M Colafigli, M Fabbiani (Roma), F Maggiolo (Bergamo), F Mazzotta, S Locaputo, P Pierotti, C Blè, M Di Pietro (Firenze), L Sighinolfi (Ferrara), G Angarano, N Ladisa (Bari), A Pan, F Zacchi (Cremona), A Gori, G Lapadula (Monza), PG Viale, M Borderi, V Colangeli (Bologna), L Minoli, S Novati (Pavia).

From the ANRS CO13-HEPAVIH Cohort: All patients participating in the cohort and

Scientific Committee of the ANRS CO13 HEPAVIH Study Group: D Salmon (principal investigator), F Dabis (principal investigator), M Winnock, MA Loko, P Sogni, Y Benhamou, P Trimoulet, J Izopet, V Paradis, B Spire, P Carrieri, C Katlama, G Pialoux, MA Valantin, P Bonnard, I Poizot-Martin, B Marchou, E Rosenthal, D Garipuy, O Bouchaud, A Gervais, C Lascoux-Combe, C Goujard, K Lacombe, C Duvivier, D Vittecoq, D Neau, P Morlat, F BaniSadr, L Meyer, F Boufassa, S Dominguez, B Autran, AM Roque, C Solas, H Fontaine, L Serfaty, G Chêne, D Costagliola, D Zucman, A Simon, S Dominguez, E Billaud, P Miailhes, J Polo Devoto, S Couffin-Cadiergues (ANRS).

Clinical Centres (ward/participating physicians): CHU Cochin, Paris (Médecine Interne et Maladies Infectieuses / D Salmon, H Mehawej; Hépato-gastro-entérologie / P Sogni; Anatomo-pathologie / B Terris, Z Makhlouf, G Dubost, F Tessier, L Gibault, F Beuvon, E Chambon, T Lazure; Virologie / A Krivine); CHU Pitié-Salpétrière, Paris (Maladies Infectieuses et Tropicales / C Katlama, MA Valantin, H Stitou; Hépato-gastro-entérologie / Y Benhamou; Anatomo-pathologie / F Charlotte; Virologie / S Fourati); CHU Pitié-Salpétrière, Paris (Médecine Interne / A Simon, P Cacoub, S Nafissa; Anatomo-pathologie / F Charlotte; Virologie / S Fourati), CHU Sainte-Marguerite, Marseille (Service d'Immuno-Hématologie Clinique - CISIH/ I Poizot-Martin, O Zaegel, A Ménard; P Geneau, Virologie / C Tamalet); CHU Tenon, Paris (Maladies Infectieuses et Tropicales / G Pialoux, P Bonnard, F Bani-Sadr, L Slama, T Lyavanc; Anatomo-pathologie / P Callard, F Bendjaballah; Virologie / C Le-Pendeven); CHU Purpan, Toulouse (Maladies Infectieuses et Tropicales / B Marchou; Hépato-gastro-entérologie / L Alric, K Barange, S Metivier; A Fooladi, Anatomo-pathologie / J Selves; Virologie / F Nicot); CHU Archet, Nice (Médecine Interne / E Rosenthal; Infectiologie / J Durant; Anatomo-pathologie / J Haudebourg, MC Saint-Paul); CHU Avicenne, Bobigny (Médecine Interne – Unité VIH / O Bouchaud; Anatomo-pathologie / M Ziol; Virologie / Y Baazia); Hôpital Joseph-Ducuing, Toulouse (Médecine Interne / M Uzan, A Bicart-See, D Garipuy, MJ Ferro-Collados; Anatomo-pathologie / J Selves; Virologie / F Nicot); CHU Bichat – Claude-Bernard, Paris (Maladies Infectieuses / P Yéni, A Gervais; Anatomo-pathologie / H Adle-Biassette); CHU Saint-Louis, Paris (Maladies infectieuses / JM Molina, C Lascoux Combe; Anatomo-pathologie / P Bertheau, J Duclos; Virologie / P Palmer); CHU Saint Antoine (Maladies Infectieuses et Tropicales / PM Girard, K Lacombe, P Campa; Anatomo-pathologie / D Wendum, P Cervera, J Adam; Virologie / N Harchi); CHU Bicêtre, Paris (Médecine Interne / JF Delfraissy, C Goujard, Y Quertainmont; Virologie / C Pallier); CHU Paul-Brousse, Paris (Maladies Infectieuses / D Vittecoq); CHU Necker, Paris (Maladies Infectieuses et Tropicales / O Lortholary, C Duvivier, M Shoai-Tehrani), CHU Pellegrin, Bordeaux (Maladies Infectieuses et Tropicales / D Neau, A Ochoa, E Blanchard, S Castet-Lafarie, C Cazanave, D Malvy, M Dupon, H Dutronc, F Dauchy, L Lacaze-Buzy; Anatomo-pathologie / P Bioulac-Sage; Virologie / P Trimoulet, S Reigadas), Hôpital Saint-André, Bordeaux (Médecine Interne et Maladies Infectieuses / P Morlat, D Lacoste, F Bonnet, N Bernard, M Bonarek Hessamfar, J Roger-Schmeltz, P Gellie, P Thibaut, F Paccalin, C Martell, M Carmen Pertusa, M Vandenhende, P Mercier, D Malvy, T Pistone, M Catherine Receveur, S Caldato; Anatomo-pathologie / P Bioulac-Sage; Virologie / P Trimoulet, S Reigadas); Hôpital du Haut-Levêque, Bordeaux (Médecine Interne / JL Pellegrin, JF Viallard, E Lazzaro, C Greib; Anatomo-pathologie / P Bioulac-Sage; Virologie / P Trimoulet, S Reigadas), Hôpital FOCH, Suresnes (Médecine Interne / D Zucman, C Majerholc; Virologie / F Guitard), CHU Antoine Béclère, Clamart (Médecine Interne / F Boue, J Polo Devoto, I Kansau, V Chambrin, C Pignon, L Berroukeche, R Fior, V Martinez; Virologie / C Deback), CHU Henri Mondor, Créteil (Immunologie Clinique / Y Lévy, S Dominguez, JD Lelièvre, AS Lascaux, G Melica), CHU Hôtel Dieu, Nantes (Maladies Infectieuses et Tropicales / F Raffi, E Billaud, C Alavena; Virologie / A Rodallec), Hôpital de la Croix Rousse, Lyon (Maladies Infectieuses et Tropicales / D Peyramond, C Chidiac, P Miailhes, F Ader, F Biron, A Boibieux, L Cotte, T Ferry, T Perpoint, J Koffi, F Zoulim, F Bailly, P Lack, M Maynard, S Radenne, M Amiri; Virologie / Le-Thi Than-Thuy); CHU Dijon, Dijon (Département d'Infectiologie / P Chavanet, L Piroth, M Duong Van Huyen, M Buisson, A Waldner Combernoux, S Mahy, R Binois, A Laure Simonet Lann, D Croisier-Bertin).

From the ICONA Cohort: All patients participating in the cohort and Board of Directors: M Moroni (Chair), G Angarano, A Antinori, O Armignacco, A d’Arminio Monforte, F Castelli, R Cauda, G Di Perri, M Galli, R Iardino, G Ippolito, A Lazzarin, CF Perno, F von Schlösser, PL Viale.

Presidential Committee: A Antinori, A d’Arminio Monforte, A Lazzarin, M Moroni, G Ippolito, CF Perno.

Scientific Secretariat: A d’Arminio Monforte (Coordinator), A Antinori, A Castagna, F Ceccherini-Silberstein, A Cozzi-Lepri, E Girardi, S Lo Caputo, C Mussini, M Puoti.

Scientific Committee: A Ammassari, M Andreoni, A Antinori, A d’Arminio Monforte, C Balotta, P Bonfanti, S Bonora, M Borderi, MR Capobianchi, A Castagna, F Ceccherini-Silberstein, A Cingolani, P Cinque, A Cozzi-Lepri, A De Luca, A Di Biagio, N Gianotti, E Girardi, A Gori, G Guaraldi, G Lapadula, M Lichtner, S Lo Caputo, G Madeddu, F Maggiolo, G Marchetti, S Marcotullio, L Monno, C Mussini, M Puoti, E Quiros Roldan, S Rusconi.

Statistical and Monitoring Team: A Cozzi-Lepri, P Cicconi, I Fanti, T Formenti, L Galli, P Lorenzini.

Biological Bank INMI: F Carletti, S Carrara, A Castrogiovanni, A Di Caro, F Petrone, G Prota, S Quartui.

Participating Physicians and Centers: Ancona - A Costantini, O Cirioni, A Giacometti; Bari - G Angarano, L Monno, C Santoro; Bergamo - F Maggiolo, C Suardi; Bologna - M Borderi, V Donati, E Vanino, G Verucchi, P Viale; Brescia - F Castelli, C Minardi, C Torti; Busto Arsizio - C Abeli, T Quirino; Chieti - K Falasca, J Vecchiet; Ferrara - D Segala, L Sighinolfi; Firenze - F Mazzotta, S Lo Caputo; Genova - AI Alessandrini, N Bobbio, G Cassola, G Mazzarello, R Piscopo, C Viscoli; Lecco - P Bonfanti, I Caramma, Macerata - P Castelli, A Chiodera; Milano - L Carenzi, A Castagna, P Cicconi, A d'Arminio Monforte, M Galli, A Lazzarin, G Marchetti, MC Moioli, R Piolini, M Puoti, A Ridolfo, G Rizzardini, S Salpietro; Modena - C Mussini, C Puzzolante; Monza - A Gori, G Lapadula; Napoli - N Abrescia, A Chirianni, M Gargiulo, MG Guida; Perugia - F Baldelli, D Francisci; Pescara - G Parruti, T Ursini; Reggio Emilia - G Magnani, M Ursitti; Roma - R Acinapura, A Ammassari, M Andreoni, A Antinori, M Capozzi, R Cauda, A Cingolani, A d'Avino, L Gallo, R Libertone, C Mastroianni, E Nicastri, G Tebano, V Vullo; Rovigo - AM Cattelan; Sassari - G Madeddu, PE Manconi, MS Mura, P Piano; Torino - S Bonora, P Caramello, G Di Perri, G Orofino, M Sciandra; Udine - M Bassetti, A Londero; Vicenza - V Manfrin, G Pellizzer.

The authors also wish to acknowledge the following individual for editorial assistance during the development of this manuscript: Jennifer Rossi.

1.

Authors’ declarations of personal interests: (i) Carlo Cerini has served as an advisory board member and participated in scientific conferences for Gilead Sciences and has received grants for such. (ii) Paola Nasta has served as an advisory board member and participated in scientific conferences for ViiV Healthcare, Bristol-Myers Squibb and Merck and has received grants for such. (iii) Jeanne M. Pimenta is a full-time employee of GlaxoSmithKline and holds stock and shares in the company.

Dominique Salmon has served as an advisory board member and participated in scientific conferences for ViiV Healthcare and has received grants for such.

(i)	Alessandro Cozzi-Lepri, Erika Chiari, Antonella d’Arminio Monforte and Maria Winnock have no conflicts of interest to declare.
2.	Declaration of funding interests: (i) This study was funded by ViiV Healthcare.