HIV and Hepatitis B Coinfection


914. The Long-term Use of Tenofovir Disoproxil in HIV-HBV Co-infection Induces a Marked Decrease in Liver Fibrosis.

K. Lacombe; A. Boyd; C. Lascoux-combe; P. Bonnard; J. Molina; P. Miailhes; L. Serfaty; P. Girard



The impact of TDF on liver fibrosis dynamics has not been well characterized on the long term.



Co-infected patients from the French HIV-HBV Cohort were included in the present analysis when treated with TDF. The level of fibrosis was estimated using a biochemical score (Fibrometer«). Time-adjusted change in Fibrometer« score from baseline (DAVG) was modeled at every 12-months after treatment initiation using GEE models adjusted for serum HBV-DNA, age, CD4+ t-cell count, body mass index, alcohol consumption, YMDD and PreC mutations, and hepatitis C and/or D co-infection. Crude change in fibrosis stage was also analysed on available pairs of liver biopsies performed at baseline and during follow-up.



One hundred and forty-one co-infected patients were treated with tenofovir and followed for a median of 29.4 months (IQR 16.3). At baseline, 75 patients (mean age: 41.2 years, SD 7.3) presented with Fibrometer« stage F0-F1-F2 and 68 patients (mean age: 44.0 years, SD 7.8) with stage F3-F4. The majority of patients were treated with lamivudine at baseline (mean duration: 55 months, SD 19.6). Among patients with F3-F4 baseline fibrosis stages, there was a steep decline in fibrosis score at 12 months (adj DAVG=-0.127; 95% CI: -0.206, -0.047; p=0.002) proceeded by a slow and stable decline at 24 and 36 months of treatment (adj DAVG=-0.146; 95% CI: -0.245, -0.048; p=0.004 and adj DAVG=-0.167; 95% CI: -0.316, -0.018; p=0.03 respectively). A larger decline in Fibrometer« score was found in patients with F4 than F3 (adj DAVG at 36-months=-0.233 and -0.106, respectively). Patients with F0-F1-F2 baseline fibrosis remained consistently stable over follow-up (adj DAVG at 36-months=0.017; 95% CI: -0.102, 0.136; p=0.8). Among 38 pairs of biopsies, a 1 point-decrease in METAVIR stage was observed in 10 patients while 5 progressed and 23 remained stable.



TDF induced a significant decrease in fibrosis level after a mean treatment duration of 29.6 months. This is in favor of its extending use in HBV-infected patients.


922. De Novo Combination Therapy of Tenofovir Disoproxil Fumarate (TDF) Plus Lamividine (LAM) or TDF Plus Emtricitabine (FTC) Is Associated with Early Virologic Response in HIV/HBV Co-infected Patients

O. Lada; A. Gervais; M. Branger; G. Peytavin; G. Collin; G. Fraqueiro; R. Moucari; S. Males; M. Martinot-Peignoux; S. Matheron; P. Marcellin




Tenofovir disoproxil fumarate (TDF) has been recently approved for treatment of chronic HBV. The aim of this retrospective study was to compare the early virologic response of de novo combination therapy with TDF+LAM or TDF+FTC with add-on therapy of LAM+TDF among HIV/HBV co-infected patients followed from 2003-2006 at Bichat Claude Bernard Hospital.



Group I patients received de novo combination therapy of TDF+LAM or TDF+FTC. These patients had either never received LAM or had not received LAM in the last 4 years. Group II patients received add-on TDF+LAM (TDF added to current LAM therapy). Early virologic response (ER) to TDF was defined as HBV-DNA viral load (VL) <3 log IU/mL after 6 months (M6) of TDF. Primary non-response (PNR) was defined as VL decrease <1 log at M6 of TDF. Serum VL (Roche Cobas) and adherence to therapy (by measurement of plasma levels by validated liquid chromatographic assay) were assessed at baseline (BL) and M6. The polymerase region was cloned, sequenced and analyzed for all patients with available samples.



We treated 141 HIV-HBV co-infected patients (131 LAM experienced) with TDF in our center. At initiation of TDF, 57/141 subjects had VL under 3 log IU/mL and 61 had VL >3 log UI/mL. Of these, 15/61 received de novo combination therapy (Group I), and 46 received add-on combination (Group II). All Group I patients (100%) achieved ER at M6 versus 75% in group II (p=0.035). Fourteen Group II patients met criteria for PNR, but 7 were non-adherent by drug level measurement. The 7 remaining patients achieved a delayed response (DR) to TDF (VL <3 log IU/mL) after a median of 20 (range 17-24) month of TDF. BL patient characteristics from both group were comparable: median age 41 (range 29-64) vs 42 (range 33-60) years, 88% vs 87% male, median VL 6.1 vs 6.5 log IU/mL, median HIV RNA <50 vs 109 cop/mL, median CD4 counts 381 vs 290 cells/mm3. As expected, Group II patients had a higher incidence of LAM resistance-associated mutations at BL (46%) than Group I patients (11%) (p=0.026). In DR, the rtA194T mutation was not observed and no new mutations were detected at M6.



In our cohort of HIV/HBV co-infected patients de novo combination therapy appears to lead to an earlier virologic response than add-on therapy. De novo combination therapy should be considered in HIV/HBV co-infected patients. We observed a small subgroup of patients with delayed response to TDF. The reason is unclear and may possibly be related to the multiple LAM resistance mutations. No new mutations were identified among the delayed responders. Ongoing viral phenotyping analyses will further clarify this observation.


967. Prior Lamivudine (LAM) Failure May Delay Time to Complete HBV-DNA Suppression in HIV Patients Treated with Tenofovir plus LAM

P. Tuma; M. Bottecchia; J. Sheldon; J. Medrano; E. Vispo; A. Madejˇn; L. Martin-Carbonero; P. Barreiro; V. Soriano



Tenofovir (TDF) has recently been approved in Europe for treatment of chronic hepatitis B. However, this drug has been used for longer than 7 years in HIV-HBV coinfected patients. HBV resistance to TDF has only sporadically been reported in subjects with LAM mutations plus rtA194T. Although most patients failing LAM respond well to TDF, it is unclear if response to TDF is somewhat impaired compared to drug-na´ve patients.
Methods: Retrospective analysis of serum HBV-
DNA in all HIV-HBV coinfected patients who initiated TDF+LAM (plus a third antiretroviral agent) at one referral center. Drug-na´ve subjects and patients with prior LAM experience were split for comparisons.



A total of 56 HIV/HBsAg+ patients (mean age 45.1 years, 89% males) were identified. Their mean follow-up was 826 patients-months. In 9 patients TDF+LAM was part of their first antiretroviral regimen, while 47 had prior LAM experience. In this last group, TDF was added even when serum HBV-DNA was <10 IU/ml in 18 (38%), and with ongoing HBV replication under LAM in 29 (62%). At the time of TDF+LAM initiation, median serum HBV-DNA was 6.21▒1.91 in patients failing LAM vs 7.23▒0.27 log IU/mL in drug-na´ve patients (p=0.01).


A total of 49 patients (88%) showed undetectable serum HBV-DNA at week 48 of TDF+LAM; in this respect there were no significant differences comparing patients with prior LAM exposure vs drug-na´ve patients (87% vs 100%; p=0.35). However, mean time to reach undetectable serum HBV-DNA was shorter in drug-na´ve than in LAM-failing patients (34.7▒24.5 vs 73.5▒67.2 weeks, respectively; p=0.01).



Dual therapy with TDF+LAM provides serum HBV-DNA suppression in a large proportion of HIV-HBV coinfected patients. However, time to achievement of undetectable serum HBV-DNA levels is longer in prior LAM-failing patients than in drug-na´ve subjects, what makes the former at greater risk for impairment in TDF susceptibility. Further studies are needed to confirm these findings.


954. Baseline CD4+ T Cell Count Predicts HBV Decline in HIV Negative and Positive and Patients Treated with Adefovir Dipivoxil

E. Formentini; A. U. Neumann; M. G. Ghany; A. C. Frank; R. T. Davey; S. Kottilil



Coinfection with HIV and hepatitis B virus (HBV) substantially alters the natural course of HBV infection as well as its management. Use of directly acting anti-HBV agents have been successful in suppressing HBV viral levels, but the kinetics of HBV viral decline and the factors that predict antiviral effect of nucleoside analogs have not yet been clearly established. In this study, we evaluated the HBV viral kinetics to adefovir therapy among lamivudine resistant HBV-infected HIV positive and negative patients



A double blind, randomized placebo-controlled study of HIV infected (N=12) and uninfected (N=5) chronic HBV patients treated with adefovir was conducted. All five HIV uninfected patients received adefovir, whereas co-infected patients were randomized to receive adefovir (n=8) or placebo (n=4) for a total of 48 weeks. At the end of 48 weeks, all patients received open label adefovir for an additional 48 weeks. HBV and HIV viral loads were performed on days 0, 1, 3, 5, 7, 10, 14, 28, and then every 4 weeks by Bayer bDNA 3.0 assay. Immune profile, liver chemistry and safety labs were performed at baseline and then serially after initiating treatment.



A significantly lower HBV viral decline was observed among co-infected patients receiving active drug when compared to that seen in monoinfected patients at week 4, (Median log -2.54 vs. -3.52, p<0.03) and at the end of treatment (Median log -5.37 vs. -7.00, p<0.03). Adefovir pharmacokinetics were similar in both groups (p>0.5). Baseline CD4+ T cell counts correlated positively with second slope decline of HBV in all patients (r=0.8; p<0.001), particularly in co-infected patients (r=0.65; p<0.05). Co-infected patients with CD4+ T cell counts less than 600 cells/mm3 had a much lower HBV viral load decline than those with CD4+T cell counts > 600 cells/mm3 (p <0.05).



HIV co-infection status is associated with lower HBV viral response rates to adefovir. Baseline CD4+ T cell count is an independent predictor of HBV decline in both HIV positive and HIV negative subjects, emphasizing the role of immune status on clearance of HBV-infected hepatocytes. The pharmacokinetics of adefovir was not affected by HIV serostatus and did not predict HBV viral response in both HIV negative and HIV positive subjects. Future studies will focus on evaluating the relationship between baseline CD4+ T cell counts and long-term suppression of HBV.


938. Entecavir Monotherapy Selects M184V Mutation in Lamivudine-na´ve and Lamivudine-experienced HIV-HBV Co-infected Patients

H. Castel; L. Bocket; C. Tamalet; M. BourliŔre; Y. Yazdanpanah



When approved by drug regulatory authorities, entecavir (ETV) was considered as the treatment of choice for HBV chronic infection (CHB) in HIV-infected patients (pts) who do not require anti-HIV treatment. But soon after, reports raised concern about the anti-HIV effect of ETV and its ability to select for M184V, a resistance mutation to HIV.



1) to assess the effect of ETV on HIV viral load (VL) in lamivudine (LAM)-experienced and LAM-na´ve HIV-infected pts with CHB who do not receive anti-HIV treatment; 2) to screen for M184V mutation emergence during ETV therapy.



HIV-HBV co-infected pts receiving ETV monotherapy for CHB were included. CD4 cell count, HIV VL, HBV DNA, and HBeAg status were performed at baseline and every 3 months. After the anti-HIV effect of ETV was reported, genotype resistance testing was done every 3 months using the Trugene« HIV-1 Genotyping Kit and the OpenGeneÖ DNA Sequencing System. Clonal analysis was performed to assess for HIV-resistance mutations at the last visit.



Four men with HBeAg-positive CHB were included. Median age was 41 [range: 35-56]. At baseline, median HBV DNA was at 7.2 log10 UI/ml [6.2-8.4], median HIV VL at 3.6 log10 copies/ml [3.0-4.2] and median CD4 cell count at 644 cells/mm3 [547-1033]. Metavir Fibrosis score was at F1 in 1 pt, F3 in 2 pts and F4 in 1 pt. Two pts were LAM-na´ve, and 2 pts LAM-experienced. Median duration on ETV was 12 months [9-16]. At month 9, when compared to baseline, the median reduction in HBV DNA compared to baseline was -4.4 log10 UI/ml [-3.5 to -4.7], and the median reduction in HIV VL -0.33 log10 copies/ml [-0.89 to +0.23]. Two pts acquired M184V mutation (1 LAM-experienced and 1 LAM-na´ve) at month 9 and at month 10 (reduction in HIV VL compared to baseline= - 0.89 and - 0.51 log10 copies/ml respectively). In the 2 others, no M184V mutation was found at month 14 and 16 of ETV therapy despite clonal analysis. In these 2 pts with consequent reductions in HBV DNA from baseline at the last visit (-4.9 and -5.5 log10 UI/ml respectively), variations in HIV VL were negligible (+0.24 and +0.38 log10 copies/ml, respectively).



Entecavir monotherapy has a significant anti-HIV activity and can select for M184V mutation in a high proportion of HIV-HBV pts, including those, as demonstrated here, who are LAM-na´ve pts. Thus, ETV should not be used without concomitant anti-HIV therapy in this population. The best strategy for HBV treatment in HIV co-infected pts needs to be defined, particularly in those with CD4 cell counts>500/mm3 who do not require anti-HIV therapy but have severe liver fibrosis stage.


852. Inter- and Intra-genotypic Recombinations of HBV DNA in Subjects Coinfected by HIV

G. Fallot; T. Durand; A. Roque-Afonso; M. Gassin; E. Billaud; A. Gervais; S. Matheron; D. Samuel; C. Feray



HBV as well as HIV can be the object of genomic recombinations due to the jump of the reverse transcriptase between different genomes. Intergenotypic HBV recombinants are described in different populations but their emergence in an individual has never been described. The aim of this study was to longitudinally analyze the variability of the whole HBV genome in patients submitted to antiretroviral therapies.


Patients and Methods

Twenty-three HIV and HBV coinfected and 9 HBV monoinfected subjects were included. All plasma samples were obtained in patient with marked HBV viremia (>log5) at least twice (mean interval= 51 months ▒ 24). They received 3TC, ADV or TDV during follow-up. The almost whole HBV genome was amplified and submitted to multiple restrictions and single strand conformational polymorphism (RFLP-SSCP). In subjects in whom RFLP-SSCP of HBV indicate changes, HBV genome was almost entirely subcloned in one or two fragments. Mutations and recombinations were searched by phylogenetic analysis (MEGA and Simplot).



RFLPľSSCP showed a marked variability of HBV DNA during the follow-up of 14/23 coinfected and in 5/9 monoinfected. Four of them, all coinfected, were infected by 2 different genotypes during the follow-up. In patient #1, initially infected by a genotype G, was infected 3 years later by a genotype F strain in which a genotype G fragment (250 bases) was recombined (within the C-terminal part of reverse transcriptase). In patient #2, initially infected by a genotype A, a recombinant HBV having of a 400 bp segment of genotype G within the N terminal protein of the polymerase appeared 12 months after. Surprisingly, these two strains gave different recombinant variants 12 and 36 months later. In patient #4 who was initially infected by the genotype A and G became infected only by genotype G without any recombination. In all cases, viral sequences (from the same genotype) which were detected in the same patient were closer each others than with all the other available sequences. Recombinant HBV strains lead to the appearance of profound changes in envelope proteins and resistance to antiretroviral therapies.



This is the first demonstration of the occurrence of inter- as well intra-genotypic recombinations of HBV during the follow-up of chronically infected patients. These recombinant events deeply modified the HBV quasispecies. These phenomenon were frequent, at least among HIV coinfected subjects with high HBV viremia, and increased the possibility of immune and drug escape strains.