P. Kennedy

P0485 : Functional immune restoration correlates with HBsAg decline and may predict treatment response on sequential NUC therapy in chronic hepatitis B

Results: Frequencies of CD3(+)CD56(+) NK-like T cells did not differ significantly between the study groups. Phenotypic analysis revealed that acutely HCV infected patients displayed a significantly lower frequency of CD161(+)CD3(+)CD56(+) NKlike T cells compared to controls. Of note, frequency of CD161expressing cells was positively correlated with IFN-g production. We found, IL-12/IL-15 stimulated CD3(+)CD56(+) NK-like T cells from healthy donors effectively block HCV replication in vitro. Supernatants of IL-12/IL-15 stimulated CD3(+)CD56(+) NK-like T cells also significantly inhibited HCV replication in vitro, suggesting that non-cytolytic mechanisms may play a major role. In addition, we only observed minimal killing of HuH7A2HCVreplicon cells by CD3(+)CD56(+) NK-like T cells and could confirm that blocking of IFN-g with a specific antibody significantly reduced the antiviral activity of these cells. However, when CD3(+)CD56(+) NKlike T cells from HIV(+) patients were studied we found HIV infection to be associated with a significantly impaired IFN-g production, irrespective of HCV co-infection. In line with this observation, CD3(+)CD56(+) NK-like T cells from HIV(+) patients were significantly less effective in blocking HCV replication in vitro than cells from healthy individuals. Conclusions: Taken together, our data indicate that CD3(+)CD56(+) NK-like T cells have the potential to block HCV replication but are functionally impaired in HIV(+) patients. This might represent a novel mechanism of dysregulated immune response in HIV/HCV co-infected patients.

PTU-117 Sequential NUC therapy following pegylated interferon provides a greater decline in HBSAG and potentially offers a treatment advantage over current therapies

Introduction Tenofovir and Entecavir are potent nucleots (t)ide analogues (NUCs) and are the standard of care in the treatment of Chronic Hepatitis B (CHB). Despite this, they have limited ability to reduce HBsAg, thus indefinite or life-long therapy is mandated in the majority of patients. Pegylated-Interferon-Alpha (PegIFNa) is associated with better rates of HBsAg decline, but its effect is limited to a small proportion of patients. In isolation, these therapies are inadequate. By optimising the use of these therapies, delivered in combination or sequentially, better treatment outcomes may be achievable. We report on the treatment response in a cohort of patients treated with sequential NUC therapy. Method 31 patients (male = 25); median age 32 (range 18–55), HBeAg positive (n = 20), were treated with PegIFNa; deemed non-responders and treated sequentially with a NUC analogue. Treatment response in this cohort was compared with 58 patients (male = 50); median age 45 (range 21–72), HBeAg positive (n = 25) who were treated with do-novo NUC therapy without previous PegIFNa exposure. Patients in both cohorts were virally suppressed at the time of analysis. Serum ALT, HBV DNA and HBsAg were quantified at baseline and longitudinally in the cohorts. Results In the sequential NUC therapy group, baseline median ALT was 102 IU/L (range 24–420) and median HBV DNA 6.93 logIU/ml compared with 43 IU/L and 4.48 logIU/ml respectively for the de-novo NUC group. ALT normalisation and reduction in HBV DNA to undetectable levels was similar in both groups over follow-up (p = n.s). At the time of viral suppression (9–12 months in both cohorts) the decline in HBsAg in the de-novo NUC group overall was 0.02 logIU/ml compared to baseline (p = n.s). In patients receiving sequential NUC therapy there was a significant decline in HBsAg over follow-up compared to baseline (0.70 log IU/ml, p = 0.0001). Conclusion Sequential NUC therapy following treatment failure with PegIFNa is associated with greater reductions in HBsAg than PegIFNa alone or NUC monotherapy. This suggests PegIFNa may prime the immune response, even in the context of treatment failure, leading to better outcomes with sequential NUC therapy. We have identified an immune phenotype associated with HBsAg decline in sequential NUC therapy. Further studies are required to confirm this finding and the applicability of an immune marker to predict treatment response in clinical practice. Disclosure of interest None Declared.

PTU-116 Functional immune restoration correlates with HBSAG decline and may predict treatment response on sequential NUC therapy in CHB

Introduction Sequential/combined therapeutic approaches comprising Pegylated-Interferon (Peg-IFNα) and nucleot (s)ide analogues (NUC) are being given greater consideration as treatment strategies for chronic hepatitis B (CHB) to achieve HBsAg loss. We demonstrated boosting of NK cells in eAg- patients treated with Peg-IFNα (Micco et al, J. Hepatol, 2013),and postulated this effect could be maintained with sequential NUC therapy. Differential NK cell responses in patients receiving sequential NUCs were analysed and correlated with HBsAg response, to elucidate a possible treatment advantage with Peg-IFNα exposure. Method PBMC from 18eAg+ patients during Peg-IFNα therapy were utlised. 9/18 patients considered Peg-IFNα non-responders after 48-weeks therapy progressed to sequential NUCs and were followed until virally suppressed. Phenotypic and functional analysis of NK cell subsets was performed by multicolour flow-cytometry and findings correlated with on-therapy HBsAg changes. Results Peg-IFNα expanded CD56bright NK cells by 3-fold (p = 0.0001) which was maintained on sequential therapy. NK cell receptor expression was analysed. All receptors, except NKG2C, were maintained on sequential NUCs, with marked augmentation in the expression of NKp30 and NKp46 on CD56brightNK cells (p = <0.05). These NK cells maintained their ability to degranulate and produce IFNγ during sequential therapy, functional restorations not seen on NUCs alone (p = 0.0001&0.002 respectively). TRAIL expression was analysed; this decreased on sequential NUCs, but remained higher than baseline. 6/9 patients had significant declines in HBsAg (>0.5 log10IU/ml) on sequential NUCs. We noted that only these patients showed the ability to increase the proportion of functional NK cells (IFNγ+&CD107+) on sequential NUCs. Additionally only these responders demonstrated a reduction in TRAIL expression on sequential NUCs, compared to those without HBsAg decline, who showed the reverse. Conclusion Restoration of NK cell cytotoxic/effector functions is seen on sequential therapy, but only in those patients with HBsAg decline. Lower expression of TRAIL also correlates with treatment response, in line with our finding that TRAIL+ NK cells can delete antiviral T-cells (Peppa et al, JEM2013). IFNγ, CD107 and TRAIL expression on NK cells may predict those patients who are likely to demonstrate HBsAg decline on sequential therapy. Thus we have identified a phenotype of TRAILlowIFNγ/CD107hiNK cells which may act as a biomarker in predicting treatment response in this setting. Given these findings, the TRAIL pathway may be a potential future target in order to improve treatment outcomes in CHB. Disclosure of interest None Declared.

PTU-106 Identifying true immune-tolerant disease in children and young adults with chb using quantified hepatitis b surface antigen levels

Introduction Immune tolerant (IT) chronic hepatitis B (CHB) is a clinical definition based on normal serum ALT and high HBV DNA. NICE guidelines identify a need to improve knowledge of the natural history of eAg+ IT disease. Importantly, recent studies have reported an inverse relationship between quantitative HBsAg (qHBsAg) and the degree of liver fibrosis in eAg+ patients.1We aimed to explore the utility of qHBsAg in combination with clinical variables to more accurately define the IT disease phase in young patients. Method Children and young adults were followed in dedicated clinics. 155 consecutive treatment naïve, eAg+ patients were included for analysis; female = 79, median age 22 (range 6–30). Data on longitudinal ALT, HBV DNA and Ishak fibrosis stage (FS) were recorded. HBsAg levels were quantified to elucidate its relationship with IT disease. Results There was no correlation between qHBsAg and age, HBV genotype or ALT levels (p = ns). High HBV DNA correlated with high qHBsAg; HBV DNA >8 log, (p = 0.016); HBV DNA >9 log (p = <0.001). 133/155 patients underwent liver biopsy. Patients with FS 0–2, indicating no or minimal liver damage, demonstrated higher qHBsAg compared to patients with a higher FS (≥3) p = 0.01. Combining clinical parameters (ALT <40, HBV DNA >8 log, FS 0–2), to denote the IT disease phase; we noted significantly higher qHBsAg levels in IT vs. non-IT eAg+ patients (p = <0.001). Mean qHBsAg was 139,471 IU/ml (5.14 log) in patients considered IT. Interestingly, qHBsAg >1,000,000 IU/ml (6 log) was only seen in patients who had the combined clinical parameters of ALT, HBV DNA and FS consistent with IT disease; thus representing a theoretical qHBsAg threshold to identify genuine IT patients. Conclusion Serum ALT and HBV DNA alone do not accurately define IT CHB. We demonstrate that high qHBsAg levels combined with these parameters can more accurately characterise IT disease, confirmed histologically with minimal or no fibrosis. We propose that qHBsAg used in combination with serum ALT and HBV DNA could obviate the need for liver biopsy in young patients. Disclosure of interest None Declared. Reference Marcellin P, Martinot-Peignoux M, Asselah T, et al. Serum levels of hepatitis B surface antigen predict severity of fibrosis in patients with E antigen-positive chronic hepatitis B. Clin Gastroenterol Hepatol. 2014

PTU-103 Characterisation of the immune profile in chronic hepatitis b patients following nucleos(t)ide discontinuation by cytof mass cytometry

Introduction Identifying chronic Hepatitis B virus (HBV) patients in whom nucleos (t)ide analogues (NUCs) can be safely discontinued with sustained immune control is a major unmet need of current treatment strategies. Predicting the onset of biochemical relapse or hepatic flares (HF) following NUC discontinuation is currently not possible. Through the utilisation of cytometry time of flight (CyTOF), a novel mass cytometry technology, in conjunction with traditional flow cytometry tools, the aim of this study was to establish an immune profile associated with the onset of HF upon withdrawal of NUC therapy. Method Cytokine and chemokine serum levels (IL-1β, IL-6, TNF-α, IL-10, CXCL-8, CXCL-10) were measured longitudinally using Luminex in chronic HBV patients who developed HF (n = 6) and in those who did not (n = 2) upon the discontinuation of NUCs. HBV-specific T cell responses were assessed after in vitro expansion of patient peripheral blood mononuclear cells with overlapping HBV peptide pools. An in-depth longitudinal analysis of the expression of 35 markers involved in the activation, differentiation, exhaustion and anti-viral effector function of T and NK cells was performed by CyTOF mass cytometry. Results The analysis of HBV-specific T cells using overlapping HBV peptides, shows that in patients with biochemical relapse after therapy withdrawal, there was no detection of functional antigen-specific T cells at all time points. The onset of HFs was, however, associated with a significant increase in the serum levels of CXCL-10 and IL-10. In contrast, a chronic HBV patient that controlled viral relapse without experiencing HF displayed functional circulating HBV-specific T cells. Preliminary analyses by CyTOF also revealed significant differences in terms of the frequency and functionality of the bulk of the T and NK cell populations both longitudinally (during and after discontinuation of therapy), as well as amongst patients characterised by relapse or immune control. Conclusion Consistent with previous studies, HFs are associated with an increase in the serum levels of CXCL10 and IL-10 while HBV-specific T cells remain undetectable in the periphery. The absence of circulating virus-specific CD8+T cells during HFs suggests that hepatic inflammation after NUC withdrawal in chronic HBV patients is induced independent of virus-specific CD8+T cells. CyTOF technology represents a major advance, allowing us to investigate multiple parameters, providing a unique insight into potential biomarkers associated with sustained immune control of HBV. Disclosure of interest None Declared.

PTU-107 Defining the low-risk inactive carrier in chronic hepatitis b with qhbsag: do the same rules apply in children and young adults?

Introduction A substantial proportion of e-Antigen negative chronic hepatitis B (CHB) under specialist follow-up are inactive carriers (IC) defined as those with low serum ALT and HBV DNA. Recent studies have described the potential utility of quantitative Hepatitis B surface antigen (qHBsAg) (<1,000 IU/ml) to represent a low-risk IC state, which is associated with reduced risk for disease progression and the development of HCC. Conversely higher qHBsAg levels (>1,000 IU/ml) are associated with an increase in risk of disease progression.1Accurate identification of low-risk IC’s would reduce the frequency of follow-up and the need for HCC screening in selected patients. We investigated whether a qHBsAg threshold (<1,000 IU/ml) representing a low-risk IC profile can be applied to children and young adults with CHB. Method Fifty-six consecutive treatment naïve eAg negative young patients (<30 years) considered IC’s (ALT <40, HBV DNA <2,000 IU/ml) over longitudinal follow-up were analysed; female = 38, median age = 26 years (range 12–30 years). A cohort of older patients (>30 years) based on the same parameters were included for comparison; female = 33%, median age = 42 years (range 31–61 years). HBsAg levels were quantified in all patients to determine any correlation between the qHBsAg threshold (<1,000 IU/ml) and age. Results In keeping with an IC disease profile, serum ALT and HBV DNA levels in both cohorts were similar (p = ns), however, qHBsAg levels were significantly higher in the <30s vs. >30s; mean qHBsAg 10,545 IU/ml vs. 5,278 IU/ml (p = 0.006). In accordance with this we detected a significant negative correlation with age and qHBsAg (Spearman Rho rs = -0.35, p = 0.0004). A significantly higher proportion of older patients had qHBsAg <1,000 IU/ml, (36% vs. 13%, p = <0.001), consistent with low-risk IC state. Conclusion The addition of qHBsAg to serum ALT and HBV DNA can enhance risk stratification. However, a threshold level (<1,000 IU/ml) cannot be safely utilised in young patients to define a low-risk IC state. The natural decline in qHBsAg with advancing age underlines the need for further studies to define an age limit above which this qHBsAg level (<1,000 IU/ml) can be used to identify low-risk ICs and streamline clinical monitoring. Disclosure of interest None Declared. Reference Tseng TC, Liu CJ, Yang HC, et al. High levels of hepatitis B surface antigen increase risk of hepatocellular carcinoma in patients with low HBV load. Gastroenterology 2012;142:1140–1149

PWE-144 Frax Score in the Assessment of Bone Mineral Density Changes in Tenofovir Treated Chronic Hepatitis B Patients

Introduction Tenofovir Disoproxil Fumarate (TDF) is an established oral antiviral (OAV) in the treatment of Chronic Hepatitis B (CHB). Bone Mineral Density (BMD) loss has been described in TDF treated HIV patients, but limited data exist in CHB. We have used DEXA scanning to determine BMD changes in TDF treated patients and have reported the possibility of BMD loss. DEXA scanning, however, is costly and requires longitudinal follow-up. We assessed the value of the FRAX® score and of bone biochemistry to evaluate their utility in TDF treated patients. Methods The FRAX score is a WHO web-based tool, used to calculate 10-year fracture risk and the need for lifestyle modification, DEXA scanning or preventative treatment. CHB patients treated with TDF for a minimum of 12-months and a control group not exposed to TDF were studied. 122 TDF exposed patients (male = 89), median age 45 (range = 26–64) and 48 patients (male = 31), median age 36 (range = 20–62) not exposed to TDF were DEXA scanned and included in the study. We calculated FRAX scores and recorded bone biochemical markers, comprising serum Alkaline Phosphatase (sALP), Calcium (sCa) and Phosphate (sPO). Results TDF treated patients had lower hip T-scores compared to controls (p = 0.02). On univariate analysis factors associated with a hip T-score < 1 included older age, lower BMI, smoking and TDF exposure (p = < 0.05). On multivariate analysis the same factors were associated with a hip T-score < 1, but TDF lost significance. For the development of a major osteoporotic fracture the pre-DEXA FRAX score was 4.77% compared to 4.33% (post-DEXA FRAX) (p = 0.9) and for a hip fracture this was 0.54% (pre-DEXA FRAX) and 0.77% (post-DEXA FRAX) (p = 0.5). The pre-DEXA FRAX score was a significant predictor of the post-DEXA FRAX treatment recommendation (p = < 0.001). TDF therapy was associated with increased sALP after 12-months, but this was not significant. No change was observed in pre-treatment sCa and sPO levels compared to those after 12-months exposure (p = 0.5 & 0.09 respectively). Conclusion Our results demonstrate the FRAX score alone can accurately predict the risk of developing an osteoporotic fracture in TDF treated CHB patients. This potentially obviates the need for DEXA scanning and the associated costs. The relationship between sALP and TDF is noteworthy, but bone parameters are of limited use in predicting BMD changes. Although BMD loss in TDF treated CHB patients remains unproven, we demonstrate the use of the FRAX score may determine those at risk of osteoporotic fractures in CHB. Disclosure of Interest None Declared.

PMO-166 Early experience with telaprevir for patients with advanced fibrosis or cirrhosis

Introduction The direct-acting HCV protease inhibitor telaprevir has recently been licensed for treatment of chronic genotype 1 HCV infection, and promises significant improvements in sustained virological response for these patients. However the patients who may benefit most from novel HCV therapies, namely those with advanced fibrosis or cirrhosis who have previously failed to respond to pegylated interferon (pegIFN) and ribavirin treatment, are relatively poorly represented in the telaprevir clinical trials. Efficacy, safety and tolerability were assessed in patients with genotype 1 HCV and advanced fibrosis/cirrhosis who have received telaprevir-containing treatment at the Royal London Hospital. Methods Laboratory results and case notes were reviewed for all patients treated with pegIFN, ribavirin and telaprevir at the Royal London Hospital between September 2011 and January 2012. Results Eight patients with genotype 1 HCV had commenced telaprevir-containing treatment. All had advanced fibrosis/cirrhosis (median Ishak score 5, range 4–6). One was treatment-naïve, three had previously failed to respond to pegIFN/ribavirin and four had relapsed after therapy. All patients had completed at least 4 weeks of telaprevir-containing therapy. With one exception, all patients achieved undetectable HCV RNA at week 4 of treatment; the patient who did not had a viral load of 168 IU/ml at week 4 and undetectable HCV RNA at week 8. One patient had completed 12 weeks of therapy, with undetectable HCV RNA. The most common side effects were fatigue (8/8), pruritis (4/8), rash (3/8), anal pain (3/8), depression (3/8), nausea (3/8), gastrointestinal disturbance (2/8) and oral candidiasis (2/8). Most side effects were successfully managed, although telaprevir was stopped in two patients at week 8 due to worsening rash and one patient withdrew from all therapy at week 4 due to tolerability. The most common laboratory abnormality was an early, transient rise in bilirubin (3/8). Significant anaemia (Hb). Conclusion Telaprevir in combination with pegIFN and ribavirin appears efficacious in patients with advanced fibrosis or cirrhosis, who have previously failed treatment with pegIFN and ribavirin alone. However, the incidence of significant side effects in this subgroup of patients is high and necessitates frequent follow-up with medical support. Side effects, particularly rash, may limit duration of telaprevir treatment. Whether this impacts on sustained virological response remains to be seen. Competing interests M Cunningham: None declared, J Schulz: None declared, L Payaniandy: None declared, Y Kallis: None declared, P Kennedy: None declared, P Kooner: None declared, R Marley: None declared, G Foster grant/research Support from: Roche, Janssen, Tibotec, Novartis, Consultant for: Abbott, BI, BMS, Chughai, Janssen, Merck, Novartis, Roche, Tibotec.

PMO-176 Induction maintenance treatment in chronic hepatitis B; step-down from tenofovir and lamivudine to lamivudine monotherapy is effective

Introduction Tenofovir Disoproxil Fumarate (TDF) is a potent and effective oral antiviral used to treat Chronic Hepatitis B (CHB), but concerns remain about possible long-term toxicity and the costs of indefinite use. An induction-maintenance treatment strategy may allow the use of combination Lamivudine (LAM) and TDF, to avert the development of resistance, followed by maintenance of viral suppression with LAM. To date, there are no data on such a step-down strategy in HBeAg negative CHB. Here we report on patients in whom we safely discontinued TDF, while maintaining viral suppression and normal liver biochemistry. Methods We selected patients who were had received combination therapy for a minimum of 18 months. Selection criteria included HBeAg negative disease, fibrosis score of <4/6 on biopsy, undetectable HBV DNA and normal serum ALT for a minimum of 12 consecutive months. Patients meeting these criteria were invited to stop TDF and step-down to maintenance LAM monotherapy. Patients were followed at monthly intervals to determine whether viral suppression and ALT normalisation was maintained in the absence of TDF. Results 21 patients (13 male), median age 47, (range 39–62) discontinued TDF. Median follow-up was 3 months (range 1–10 months). During monthly follow-up biochemical and serological data have been measured. All patients had undetectable HBV DNA prior to step-down therapy to LAM and this remained undetectable during follow-up. Pre-discontinuation of TDF the median ALT was 27 (range 15–38) and during follow-up, on LAM monotherapy, was 22 (range 15–45), (p=NS). Median HBsAg level pre-discontinuation of TDF was log 3.48 (range 1.55–4.49) and 3.49 (range 1.55–4.55), (p=NS) on LAM monotherapy. Conclusion We demonstrate no viral breakthroughs or biochemical flares on discontinuing TDF. These data suggest that an induction-maintenance strategy may be pursued in selected CHB patients to avoid long-term exposure to TDF and reduce the burden on healthcare budgets in the context of lifelong oral antiviral therapy. Competing interests None declared.

PMO-175 Can a 3-month “stopping rule” for pegylated-interferon-α be applied to a UK population of chronic hepatitis B infected patients of mixed genotype?

Introduction Stopping rules have been proposed for the early discontinuation of Pegylated-Interferon-α (PEG-IFN-α) therapy in those patients who are considered unlikely to respond. Recent studies have shown that no reduction in quantitative HBsAg and the absence of >2 log decline in HBV DNA at 12 weeks therapy can predict non-response. However, these data are almost exclusively from genotype A and D cohorts. Here we test how robust this strategy would be in clinical practice and whether this rule could be applied to a UK population of diverse HBV genotypes. Methods 49 patients (male=35) were treated with PEG-IFNα for CHB over the course of the study. Ten patients remain on therapy and eight patients discontinued due to poor compliance or intolerance. 31 patients (male=20), HBeAg positive (n=24), median age 31 (range 18–55) completed 48 weeks PEG-IFNα and were included in the analysis. HBV genotype was recorded for all patients (A=6, B=5, C=10, D=9, E=1). ALT, HBV DNA and HBsAg was quantified at baseline and longitudinally at 12-week intervals. Results Of the 31 patients, 10 were considered responders; seven were HBeAg positive and seroconverted on therapy and three were HBeAg negative pre-therapy and considered responders with sustained immune control off treatment. The decline in HBV DNA and qHBsAg by 12 weeks was 3.99 log, 0.17 log (HBeAg positive group) and 2.9 log, 0.5 log (HBeAg negative group) respectively. 16/31 patients were non-genotype A or D. Of the responders from this group there was a decline in HBV DNA and qHBsAg of 4.10 log and 0.58 log respectively by 12 weeks. On sub-group analysis by genotype, there was no statistically significant difference in HBV DNA and qHBsAg decline at 12 weeks across all genotypes, when comparing HBV DNA and qHBsAg between genotype A and D and non A and D patients (p=0.40 and 1.0 respectively). More over adopting the rule of >2 log decline in HBV DNA and no decline in qHBsAg by 12 weeks, reveals we would not exclude those likely to respond; as all responders achieve the outlined viral response by 12 weeks therapy. Conclusion These data highlight the utility of this stopping rule for PEG-IFNα across all genotypes. The absence of >2 log decline in HBV DNA and reduction in qHBsAg at 12 weeks therapy makes a favourable response unlikely. This rule should be adopted in clinical practice to avoid poorly tolerated side effects and the cost of completing 48 weeks therapy. Furthermore, this 12-week milestone would allow the early switch to an oral antiviral in PEG-IFNα non-responders. Competing interests None declared.