Sunday Forums: Clinical Etiology
172. Low serum adiponectin levels in a transgenic mouse model of hepatitis C infection
A. Delgado; Y. Kamegaya; S. H. Jordan; S. Agrawal; Y. Hiasa; R. T. Chung
Background and Aims:
Recent studies have demonstrated an association between HCV infection and insulin resistance. However, the underlying mechanisms for this association remain unclear. We sought to evaluate whether HCV decreases in vivo levels of adiponectin, an anti-inflammatory adipokine known to induce insulin sensitivity.
Transgenic (Tg) mouse models expressing HCV core and HCV core-E1-E2 proteins under the control of the HBV enhancer and albumin promoter, respectively, were created on a hybrid C57BL/6/FVB background as previously described (Kamegaya Y., Hepatology 41: 660-7, 2005). On this background, mice do not develop significant liver pathology. Twenty-eight Tg (18 core, 10 core-E1-E2) and 12 littermate controls were evaluated. All mice were 16 to 18 months old. Serum samples were collected following an overnight fast, and immediately frozen at –80C. Adiponectin levels were measured using Luminex ELISA technology.
Mean adiponectin levels were 16.45 μg/ml for non-Tg vs. 13.79 μg/ml for Tg mice (p = 0.2). There were no differences in adiponectin levels observed between the core and core-E1-E2 Tg animals. The greatest difference in frequencies was observed for the group of adiponectin levels <10 μg/ml (36% Tg vs. 9% non-Tg). Thus, adiponectin was dichotomized at the critical value of 10 μg/ml, and low adiponectin was defined as any value below 10 μg/ml. Logistic regression was used to compare the probability of having a low adiponectin level between the Tg and non-Tg groups. The odds of having a low adiponectin level were 30.4 (1.6 – 583.8) times greater (p=0.02) for the Tg than the non-Tg group.
HCV Tg mice were significantly more likely to have low adiponectin levels as compared with non-Tg mice. These data demonstrate an in vivo association between HCV and low adiponectin levels, and suggest that HCV structural protein expression alone may suffice to decrease adiponectin levels. These results provide a plausible mechanism for HCV-associated insulin resistance.
P. K. Pandya; S. C. Mathur; P. Callahan; P. Shah; D. Reker
In addition to traditional predictors of poor response to antiviral therapy for chronic hepatitis C such as viral load, genotype, race, and advanced fibrosis, recent evidence has implicated obesity related factors such as insulin resistance and hepatic steatosis. The prevalence of insulin resistance and related MS in veteran population with chronic hepatitis C has not been systematically studied.
We performed a cross-sectional analysis of the prevalence of the metabolic syndrome in chronic hepatitis C infection. We evaluated 38 consecutive veterans referred for evaluation of chronic hepatitis C, who were drinking less than 20 g alcohol per day (by standard questionnaire). We hypothesized that the metabolic syndrome (MS) would be associated with worse hepatic histology than HCV infection alone. We collected detailed demographic, anthropometric, serologic, histologic and metabolic parameters in order to compare metabolic, virologic and histologic status between subjects with and without the MS. Insulin resistance was defined as a HOMA-R (fasting insulin (mU/L) x fasting glucose (mmol/L)/22.5) greater than 2.5. MS was defined by the presence of 3 or more ATP III criteria. Additional data included determination of hs-CRP, TNF-α, IL-6, adiponectin and leptin. Groups were compared using t-test and/or Fisher’s exact analysis.
The average age, BMI, viral load, ALT and HOMA were 53.3±7.0, 28.0±6.3, 2.57x106, 59.4±38.3 and 4.7±7.4 respectively. All but one of the patients were male. Caucasians and African Americans comprised 71% and 18.4% of the cohort. The predominant genotype in the cohort was Type 1 (81%). Twenty of the 38 veterans studied met the ATP III criteria for the MS. Veterans with the MS had higher BMI (p=0.001), HOMA-R (p=0.012), and leptin (p=0.008), while no differences were seen (p=ns) in age, race, ALT, TNF-α, Il-6, adiponectin, hs-CRP, viral load, or genotype. Patients with MS had significantly higher prevalence of steatosis (p=0.001) and fibrosis on biopsy (p=0.001).
Metabolic syndrome was present in 53% of veterans referred with chronic hepatitis C. Leptin but not adiponectin was significantly associated with the metabolic syndrome. Veterans with metabolic syndrome have more steatosis and advanced fibrosis. The high prevalence of metabolic syndrome may adversely impact outcomes with antiviral therapy in veterans.
174. Immunohistochemical expression of junctional proteins and intestinal permeability are altered in Hepatitis C virus (HCV) related chronic hepatitis (HCV-CH).
L. Martorelli; S. Anna; P. Esposito; L. deMagistris; M. Russo; M. Generoso; M. Carteni'; G. Riegler; F. Ferraraccio; A. Fasano
Altered intestinal permeability values (IP) and intestinal ultra-structural abnormalities have been shown in cirrhotic patients as well as other liver diseases.
To study the IP, serum levels of the intestinal tight junction (tj) modulator zonulin (Z), and the immunohystochemical expression of junctional proteins occludin, claudin, desmoglein and intracellular actin in duodenal mucosa of HCV-CH patients.
The IP was evaluated by the Lactulose/Mannitol (LA/MA) test. Serum Z levels were determined by sandwich-ELISA. Distal duodenum biopsies were obtained by EGD and processed for immunohystochemistry. Staining was obtained with antibodies specific for occludin, claudin-4, desmoglein-3, and actin. Twenty-five consecutive naïve HCV-CH patients were enrolled (12 males, 13 females, mean age 59,9 ±10,5, range 35-72); none of them was on interferon treatment. All enrolled patients underwent IP and Z investigation; 9 of them gave informed consent to perform an EGD with duodenal biopsies.
The LA/MA mean ± SD value in the 25 patients studies was 0.025±0.020, with 8 patients (32%) having LA/MA values higher than the normal cut off (< 0.028). The mean ± SD value of Z in the 25 patients resulted 1.27±1.64 ng/mg protein, with 10 of them (40%) having Z values higher than the normal cut off (< 0.6 ng/mg protein). There was a correlation between elevated LA/MA and serum zonulin levels. Occludin immunoreactivity at the cell membrane surface resulted diminished in all cases but one (decrement ranging between 50-90% as compared to normal controls). These changes were paralleled by an increased cytoplasmatic occludin immunoreactivity (2-50%), suggesting a displacement of the tj protein from the membrane boundary to the cytoplasm. Claudin-4 expression resulted diminished by 30-60% in all cases and was associated to a decrease in actin immunoreactivity (45-85%). Desmoglein immunoreactivity also decreased by 45-70%.
Alteration in IP was detected in a one third of the HCV-CH patients studied and was associated to elevated serum Z values. These changes were associated to modification of tj proteins expression and localization. Our data suggest that a subgroup of HCV-CH patients has increased serum Z and changes in tj proteins distribution compatible with altered IP. These changes may lead to increased non-self antigens access that may influence the course of the liver disease.
A. Iwai; H. Marusawa; Y. Takada; H. Egawa; M. Nabeshima; T. Chiba
Interferon (IFN)-α2b plus ribavirin (RBV) combination therapy is widely used for recurrent hepatitis C following liver transplantation; however, the rate of sustained virological response (SVR) in transplant patients has been limited. Recurrent hepatitis C in transplant patients differs from that in immunocompetent individuals in several ways such as immunosuppressive condition and extremely high viral loads. However, it remains unclear whether viral and host factors determine the effectiveness of combination therapy in those patients.
The aim of this study was to identify the host and viral factors associated with a SVR to antiviral therapy in patients with recurrent hepatitis C after living-donor liver transplantation (LDLT).
Twenty-three patients receiving the IFN-α2b plus RBV combination therapy for recurrent hepatitis C after LDLT were enrolled. The hepatitis C virus (HCV) genome clones circulating in sera of those recipients were cloned by reverse transcription–polymerase chain reaction.
Twenty of the 23 patients (87%) were infected with HCV genotype 1b. Before treatment, all patients had serum HCV RNA levels over 100 KIU/mL by the Amplicor HCV 2.0 assay. Moreover, HCV RNA levels were greater than 850 KIU/mL (ranging from 850 to 21,200 KIU/mL) in 17 cases (74%), which is equivalent to 8.5 × 105 to 2.12 × 107 copies/mL HCV RNA, indicating that a majority of the patients in this study had high baseline viral loads before treatment. SVR was achieved in seven of 23 patients (30%). Predictive factors for SVR included; (1) 2log10 decline in HCV RNA at two weeks after the start of therapy, (2) disappearance of HCV RNA at four or 24 weeks after the start of therapy, (3) frequency of positivity for HLA-B51, and (4) frequency of negativity for HLA-A26. Since the pretreatment high viral load showed no association with SVR, we asked whether other viral factors played roles in determining response to the combination therapy in transplant recipients. We found several novel defective HCV clones in four (33%) of twelve recipients’ sera with genotype 1b and extremely high HCV RNA loads (over 850 KIU/ml). All defective HCV clones had deletions in the envelope region of HCV genome. Interestingly, no patients with defective clones showed a 2log10 decline in HCV RNA at two weeks after the start of therapy.
Early decline in serum HCV RNA after treatment was closely associated with SVR. The circulating defective HCV clones are present and might be responsible for resistance to the combination therapy in patients with recurrent hepatitis after liver transplantation.
A. Bergk; C. Sarrazin; M. von Wagner; G. Teuber; P. Buggisch; V. Weich; B. Wiedenmann; T. Berg
Current treatment guidelines for the therapeutic management of chronic hepatitis C virus (HCV) genotype 1 infected patients include a week 12 stopping rule for nonresponders defined by a drop of viral load less than 2 log10 in patients treated with pegylated interferon and ribavirin. Nevertheless there has been no convincing prognostic tool to easily and reliably predict viral relapse after the end of 48 weeks of treatment.
Aim of the present study was to evaluate the predictive value of a minimal residual HCV viremia for a relapse in genotype 1 infected patients with early virologic response to therapy.
Patients and methods:
We retrospectively analyzed viral kinetics at week 12 and response of 773 treatment naive HCV genotype 1-infected patients using PegIFNα-2a/b and ribavirin treated at our outpatient clinics. For the detection of residual viremia at week 12 a quantitative real-time PCR with a lower limit of detection of 10IU/mL (TaqMan) was used.
Using standard quantitative HCV-RNA assays at week 12 75% (430/773) of the patients treated with PegIFNα-2a/b and ribavirin showed an early virologic response defined by week 12 viral load drop ≥2 log10. By re-analyzing 222 available serum samples of early virologic responders by TaqMan a residual viremia could be detected in 84 out of 222 patients (38%). The presence of viremia highly correlated with a viral relapse after the end of 48 weeks of treatment (p<0,001). Residual viremia was detectable by real-time PCR at week 12 in only 19 out of 126 sustained responders (15%) as compared to 69 out of the 96 relapse patients (72%) (p<0,001).
A relapse occurred in 78% of the patients who had detectable HCV viremia at week 12 (68/88) as compared to only 19% (25 out of 134 patients) if HCV RNA was undetectable at week 12. The relative risk to suffer from a relapse was 3.5 and 0.26 in patients with and without residual hepatitis C viremia at week 12, respectively (p<0,001).
Using a highly sensitive real-time PCR the detection of minimal residual hepatitis C viremia at treatment week 12 was in 78% associated with a viral relapse after the end of therapy and is therefore a valuable prognostic tool for the prediction of individual treatment outcome. Treatment of patients with detectable viremia at week 12 should be individually intensified by either prolonging treatment duration or preventing dose reductions by the application of erythropoetin or GM-CSF.
HCV Treatment Forum:
195. Generation and Characterization of HCV Replicons with Reduced Sensitivity to ITMN 191, a Macrocyclic Inhibitor of NS3/4A
S. Seiwert; S. W. Andrews; H. Tan; K. R. Condroski; J. A. Ballard; B. A. Bernat; J. A. Josey; L. M. Blatt
The standard of care for chronic HCV infection affords a sustained virologic response in ~50% of chronic HCV patients—thus the need for novel therapeutic approaches. We therefore embarked on a rational drug design campaign to produce inhibitors of the HCV NS3/4A protease. ITMN 191 (an active site inhibitor with a macrocyclic structure) emerged from this discovery effort and was nominated as a preclinical candidate. In a genotype-1b replicon system, ITMN 191 displays an EC50 of 2.1 nM and an EC90 of 5.6 nM. It has liver levels in multiple species predictive of efficacious exposure in humans at moderate doses, and displays an acceptable tolerability profile.
To understand the resistance profile of ITMN 191, we incubated it with a genotype-1b HCV replicon in increasing concentrations ranging from 20 nM to 320 nM (10-fold to 160-fold the original EC50 value). As measured by its EC50 value, the potency of ITMN 191 was reduced by 149-fold against replicons that persist in the presence of 320 nM ITMN 191. Sequence analysis demonstrated that all such replicons carried the previously identified mutation D168A; this mutation disrupts a salt bridge to R155. Structural modeling suggests that loss of this salt bridge would partially occlude binding of ITMN 191’s fluoro-isoindolene P2 group. In a biochemical activity assay, D168A shifted the EC50 of ITMN 191 from <300 pM to 4.5 nM. The majority of replicons that persist in the presence of 320 nM ITMN 191 additionally carried both V116M and E30A mutations. Biochemical characterization of the potential contribution of these mutations is underway.
In conclusion, HCV replicons with reduced sensitivity to ITMN 191 harbor a D168A mutation. The additional presence of V116M and E30A mutations in ITMN 191–resistant replicons, residues that are far from the binding site of ITMN 191, suggests that if these sequence changes contribute to replicon fitness or ITMN 191 resistance, it is through an indirect mechanism. Interestingly, the D168V mutation that confers resistance to the macrocyclic inhibitor BILN-2061 was not observed in this study. Biochemical analysis has previously demonstrated that ITMN 191 retains subnanomolar potency against D168V. Additionally, it is noteworthy that mutations that confer reduced sensitivity to both BILN-2061 and linear tetrapeptide inhibitors such as VX-950 (A156T/V) were not identified in this study. Again, previous biochemical analysis has confirmed that ITMN 191 retains subnanomolar potency against A156T. Thus, ITMN 191 has a favorable cross-resistance profile with VX-950 and related linear tetrapeptides, and potentially with other macrocyclic inhibitors as well.
D. Nelson; V. Rustgi; V. Balan; M. Sulkowski; L. Lambiase; R. Dickson; R. Weisner; G. Davis; A. Muir; L. Novello; R. Yu; W. Freimuth; A. Neumann; J. McHutchison; M. Subramanian
Background and Aims:
Albuferon (alb-IFN) is a novel recombinant protein consisting of IFNa genetically fused to human albumin. This ongoing Phase 2, dose-ranging study evaluates the safety and efficacy of alb-IFN in chronic HCV patients who were non-responders (failed to achieve EVR12 or clear HCV RNA on therapy) to previous IFNa based regimens.
Subjects were randomized into 3 alb-IFN SC treatment cohorts (900 mcg Q2w, 1200 mcg Q2w or 1200 mcg Q4w) in combination with ribavirin (RBV) 1000-1200 mg/d. After evaluating safety data, 2 higher dose cohorts of alb-IFN 1500 mcg Q2w and 1800 mcg Q2w were enrolled. The treatment duration is 48w with 24w follow-up. The primary efficacy end-point is SVR. The initial 3 cohorts have completed 48w treatment and 12w data for the highest dose cohort is currently available.
Subject demographics and antiviral response for the 115 subjects enrolled are summarized in the table. At w12, the overall antiviral response in the 1800 mcg Q2w cohort was the highest despite having a greater proportion of prior PEG-IFN+RBV non-responders. The slope of HCV RNA decline at w4-12 was significantly more rapid (P<0.01) for the 1800 mcg cohort compared to the 900-1500 mcg cohorts in genotype 1, prior PEG-IFN+RBV non-responders. The w24 antiviral response was comparable across the 900-1500 mcg cohorts. Most patients who were RNA negative at w24 remained negative at w48. Antiviral response at w12 and w24 was predictive of w48 end-of-treatment response for the 900-1200 mcg cohorts. Overall, alb-IFN in combination with RBV was well tolerated and the safety profile in the 1500 mcg and 1800 mcg cohorts was comparable to the 900-1200 mcg cohorts in type, incidence and severity of AEs.
Alb-IFN at doses up to 1800 mcg Q2w in combination with RBV is safe and demonstrates significant antiviral activity in prior IFNa non-responder patients. Further studies are warranted and are ongoing.
M. R. Kraus; O. Al-Taie; A. Schaefer; M. Pfersdorff; M. Scheurlen
Background and aims:
Interferon-induced depression is still a major problem in antiviral therapy of chronic hepatitis C. By the use of SSRIs, this therapy-induced side effect can be successfully managed in most cases. However, so far little is known about underlying mechanisms and reliable predictive factors associated with cytokine-induced depressive symptoms.
The aim of the present study was to investigate the putative role of the 5-HT1A C(-1019)G serotonin receptor polymorphism in a prediction model for IFN-induced depression.
In a longitudinal single-center study, we included a volunteer sample of 139 HCV outpatients. These were enrolled between 1997 and 2004 and received (combination) treatment with (pegylated) interferon alfa-2b (and ribavirin).
Before and during interferon therapy, depression was monitored using the Hospital Anxiety and Depression Scale (HADS) and interview data (DSM-IV criteria for major depression / major depressive disorder).
EDTA blood samples were drawn from all patients, and subsequent genotyping of the serotonin receptor polymorphism (HTR1A-1019) was realized by DNA isolation extraction and polymerase chain reaction in an ABI PRISMTM 310 Genetic Analyzer (according to standard protocols).
According to our data, there was a significant effect of the HTR1A receptor polymorphism on the occurrence of interferon-induced depression during antiviral therapy. Both higher incidence and extent of depressive symptoms were significantly associated with carriers of the G allele. ANOVA analyses showed that maximum (P=0.011) and mean (P=0.024) increases in HADS depression scores on antiviral therapy were significantly affected by the HTR1A-1019 polymorphism. Also, the occurrence of clinically relevant HADS depression scores (cutoff criterion ≥ 9) was significantly associated with the allelic variation (Chi-square test, P=0.017).
Our findings suggest allelic variation in 5HT1A expression to play a significant role in the development of interferon-induced depression in the antiviral treatment of chronic hepatitis C. Prediction models of interferon-induced depressive symptoms that are based on the HTR1A-1019 polymorphism should be further investigated.
198. Treatment of Chronic Hepatitis C in Thalassemic and Sickle Cell Disease Patients with Interferon Alfa2b (IFN) and Ribavirin(RBV).
d. b. ancel; D. Chaslin-Ferbus; x. J. Amiot; I. Hagege; J. Schaeffer Plumet; S. Pol; R. Girot; j. b. grange
Prognosis of patients with congenital haemolytic anaemia has greatly improved in recent years, particularly among those suffering from thalassemic syndrome. In this population, hepatic complications due to post-transfusion viral hepatitis and iron overload become one of the main causes of death. The 1999 Consensus statement of the EASL listed anaemia as an absolute contra-indication to RBV in hepatitis C. Aim: to assess the efficacy and safety of IFN and RBV treatment among patients with congenital haemolytic anaemia
Ten patients (mean age: 27.4 ± 6.3 yr) were included: 5 sickle cell disease (4 treatment-naïve, 1 non responder) and 5 thalassemic major patients (1 naïve, 4 non responders). Eight patients (5 sickle cell, 3 thalassemic patients) were treated with PEGIFN and RBV. Two thalassemic patients were treated with IFN monotherapy. PEGIFN was given at full dose. RBV was started with an undervalued dose (400mg/d), then progressively increased according to weight based recommended dose. Duration of treatment was 24 or 48 wk.
HCV genotypes were 1 and 4 in 8 cases. All sickle cell disease patients had a fibrosis METAVIR score ≦2, whereas 4/5 thalassemic patients were F4. A virological response at the end of treatment was achieved in 9/10 patients, and a sustained virological response (SVR) in 6/10 patients (3 thalassemic patients including 2 with cirrhosis). Five of the six SVR were obtained with combination therapy. The 6 sustained responders did not present any hepatic complication during the follow-up. A non responding thalassemic patient died of hepatocellular carcinoma recurrence after liver transplantation. The mean increase in transfusion requirements during treatment compared with the same period prior to treatment was 22% in thalassemic patient group (growth factors were not administered). No sickle cell patient needed transfusion during or after treatment periods. Surprisingly, the haemoglobin level during and at the end of treatment was higher than the pretreatment value among 4/5 sickle cell patients (none of them being treated beforehand by hydroxyurea).
In this pilot study, overall SVR was 60 %, despite the unfavourable genotypes of most patients. Combination therapy with IFN and RBV achieved a high SVR rate (62.5%, 5/8 patients). The increased transfusion requirements during treatment in thalassemic patients are acceptable. Among sickle cell patients, no previous case of excellent haematological tolerance under RBV and without hydroxyurea pre-treatment has been reported in literature: our data suggest the use of a full dose RBV and PEGINF from the start of treatment in these patients.
199. Relationship between Serum Peginterferon and 2,5-OAS Kinetics and Virologic Responses in African Americans and Caucasians Patients with Chronic Hepatitis C, genotype 1, during Peginterferon Combination Therapy
C. Howell; T. Dowling; M. Haritos; N. Terrault; W. Thelma; M. Taylor
African Americans (AA) infected with hepatitis C virus (HCV) genotype 1 have significantly lower rates of HCV clearance than Caucasian Americans (CA) following treatment with peginterferon alfa (PEGIFN) and ribavirin. In the Virahep-C study, this racial difference in virologic response was observed during the first 4 weeks of therapy.
To compare PEGIFN pharmacokinetics and pharmacodynamics (i.e. serum 2,5 OAS kinetics) in relationship to virologic responses during the first 28 days of PEGIFN-2a and ribavirin treatment in AA and CA HCV genotype 1 patients.
401 treatment-naïve, HCV genotype 1 patients (196 AA and 205 CA) were treated with PEGIFN-2a (180 mcg/wk) and ribavirin (1000-1200 mg/day) for up to 48 weeks. Ninety-six patients who received PEGIFN without dose alteration during the first 4 weeks of treatment and who had PEGIFN and 2,5-OAS serum levels measured before treatment and on treatment-days 1, 2, 3, 7, 14 and 28 were selected for this study. PEGIFN and serum 2,5-OAS (pharmacodynamic marker) levels were measured by ELISA and by RIA (Roche Molecular Diagnostics, Alameda, CA). The maximum PEGIFN concentration (Cmax) and area under the serum concentration-time curve (AUC) values from day 0 to 7 were determined using non-compartment PK analysis. Serum HCV RNA levels were measured using the Amplicor HCV Monitor Test, version 2.0 assay.
At baseline, AA and CA patients were similar in regards to age, gender distribution, body weight, BMI, alcohol or tobacco use, liver histology, creatinine clearance, baseline HCV RNA levels, and HCV 1 subtype distribution, but AA had a lower mean serum ALT level (p = 0.003). There were no differences between AA and CA in either Cmax or AUC from day 0-7 after first PEGIFN dose, but AA were more likely to have a Tmax by day 3 (p = 0.03) and had a greater PEGIFN level on day 28 (18.0 ± 8 vs. 15.5 ± 7 ng/mL, p = 0.03) and higher 2,5-OAS levels on day 2, 3, 14, and 28 (p < 0.05). In contrast, AA had a significantly smaller decline in serum HCV RNA from pretreatment to day 28 (p =0.02 day 14 and 28). There was a negative correlation between serum PEGIFN-2a (p < 0.05) and HCV RNA levels on days 7 (r = -0.23, p = 0.09), 14 (r = -0.30, p = 0.03), and 28 (r = -0.32, p = 0.02) in CA but not in AA. Likewise, there was a negative correlation between serum 2,5-OAS levels on day 2-14, and HCV RNA levels from day 2-28 in CA (p < 0.05) but not in AA.
The weaker virologic response in AA HCV genotype 1 patients compared to CA patients during the first 4 weeks of PEGIFN-2a and ribavirin treatment is not explained by racial differences in serum PEGIFN and serum 2,5-OAS kinetics.
M. Hussein; J. S. Benner; D. Lee; A. Sesti; D. S. Battleman
For patients with hepatitis C virus (HCV), poor adherence to antiviral therapy is a common barrier to treatment success. Challenges associated with administration and tolerability make this issue especially true of pegylated interferons + ribavirin. The Be In Charge® program (BIC) is a comprehensive patient support program for HCV patients; it encourages adherence by providing 24-hour inbound and proactive outbound on-call nursing support and mailings of HCV educational materials throughout therapy.
The purpose was to determine the effect of BIC on patient adherence to peginterferon alfa-2b combination therapy (peg-2b).
A retrospective cohort analysis comparing BIC enrollees to a matched control group was conducted. Subjects were included if they were ≥18 years of age; started peg-2b on or after 1/1/2004; and could be followed for at least 12 weeks after treatment initiation. To reduce potential selection bias from self-enrollment in the program, BIC enrollees were propensity score-matched to peg-2b starters not enrolled in BIC (controls), based on clinical and demographic characteristics. Filled prescription records were used to measure adherence based on the number of injections dispensed and proportion of patients who received an average of ≥1 injection per week during follow-up. Adherence was compared using paired chi-square and t-test.
The BIC group consisted of 780 eligible subjects who were observable for ≥12 weeks, including 638 and 333 subjects observable for 24 and 48 weeks, respectively. Of control subjects, 8572, 7014, and 4071 subjects were observable for 12, 24, and 48 weeks, respectively. Compared to non-BIC subjects, BIC enrollees were more likely to be female (53% vs. 41%, P<0.0001) and to use injection delivery devices vs. vials (74.0% vs. 56.7%, P<0.0001). In the matched analysis, BIC subjects refilled 1.2 more injections (95% confidence interval [CI] 0.52, 1.83; P<0.0001) than the control cohort within 12 weeks, 2.7 more (95% CI 1.5, 3.8; P<0.0001) within 24 weeks, and 6.7 more (95% CI 4.3, 9.1; P<0.0001) within 48 weeks. Compared to matched controls, BIC enrollees were more likely to refill ≥12 injections within 12 weeks of initiation (72% vs. 64%, P=0.0005), ≥24 injections within 24 weeks (52% vs. 41%, P<0.0001), and ≥48 injections within 48 weeks (22% vs. 13%, P=0.0020).
These findings suggest the BIC program significantly improves adherence to peg-2b. Additional research is needed to ascertain which aspects of the program are most effective and which patients are most likely to benefit from this intervention.