Additional Tuesday Poster Sessions,
Topic: Current Treatment - Pegasys
M. L. Shiffman; D. R. Nelson; G. Hooper; D. Messinger;
S. Zeuze
Background:
Patients with HCV genotypes (GT) 2 or 3 have
traditionally been referred to as “easy to treat”. However, data from the
ACCELERATE trial have demonstrated that this is not true for all of these
patients.
In this study 1463 treatment-naive patients with HCV
genotype 2 or 3 were randomized to receive either 16 or 24 weeks of
peginterferon alfa-2a (40KD) 180 µg/week and ribavirin 800 mg/d. Overall 66% of
these patients were HCV RNA undetectable at treatment week 4; defined as a
rapid virologic response (RVR). These patients had an SVR of 79% and 85% with
16 and 24 weeks of treatment respectively. In contrast, patients with HCV
genotype 2 or 3 who had detectable HCV RNA at week 4 (non-RVR) had an SVR rate
of only 45% with 24 weeks of treatment. This analysis explores baseline (BL)
factors that may identify those non-RVR patients who are more likely to achieve
an SVR with either 16 or 24 weeks of therapy.
Methods:
All patients who received at least one dose of study
drug and who did not achieve an RVR were included in this analysis (ITT
population). Patients were dichotomised by BL factors and rates of SVR
calculated as a function of treatment duration.
Results:
466 non-RVR patients were included in this analysis
(Table). Patients with a BL HCV RNA <400,000 IU/mL had similar rates of SVR
with 16 and 24 weeks of treatment. For all other BL characteristics non-RVR
patients treated for 24 weeks had a higher SVR rate compared to patients
treated for 16 weeks (overall 44.7 vs 26.4%). In patients treated for 24 weeks
GT 2 (vs GT 3), absence of cirrhosis and younger age were associated with at
least 10% higher rates of SVR.
Conclusions:
Patients with HCV GT 2 or 3 who do not achieve an RVR are not “easy to treat” and have low rates of SVR. The SVR in those patients treated for 24 weeks with GT 2, absence of cirrhosis and younger age is consistently above 50%, but this is still significantly inferior to that observed in patients with an RVR. RVR remains the strongest predictor of SVR and studies to evaluate the impact of prolonging treatment to 48 weeks in patients with HCV GT 2 or 3 who do not achieve an RVR are warranted.
Rate of SVR in non-RVR pts (ITT)
|
Characteristic |
16wks
(n=220) |
24wks (n=246) |
|
Geographical region |
18/62 (29.0%) 40/158 (25.3%) |
29/66 (43.9%) 81/180 (45.0%) |
|
Genotype 2 Genotype 3 |
27/103 (26.2%) 31/117 (26.5%) |
53/99 (53.3%) 57/145 (39.3%) |
|
HCV RNA (IU/mL) ≤400 000 >400 000 |
14/28 (50.0%) 44/192 (22.9%) |
17/32 (53.1%) 93/214 (43.5%) |
|
HCV RNA (IU/mL) ≤800 000 >800 000 |
19/46 (41.3%) 39/174 (22.4%) |
27/57 (47.4%) 83/189 (43.9%) |
|
Cirrhosis No Yes |
47/146 (32.2%) 11/74 (14.9%) |
90/174 (51.7%) 20/72 (27.8%) |
|
Age (years) <45 ≥45 |
30/85 (35.3%) 28/135 (20.7%) |
44/83 (53.0%) 66/163 (40.5%) |
|
BMI (kg/m2)
≤30 >30 |
45/139 (32.4%) 13/81 (16.0%) |
75/157 (47.8%) 35/89 (39.3%) |
Topic: Experimental Therapies - R1626
P. J. Pockros; D. Nelson; E. Godofsky; M.
Rodriguez-Torres; G. Everson; M. W. Fried; R. H. Ghalib; S. A. Harrison; L. M.
Nyberg; M. L. Shiffman; G. Z. Hill; A. Chan
Introduction:
The novel nucleoside, R1626, is a potent inhibitor of
the HCV RNA polymerase. When administered as monotherapy to patients with
chronic hepatitis C, R1626 demonstrates dose-dependent decreases in HCV RNA
blood levels.
Method:
This Phase 2a study evaluated the safety and efficacy
of R1626 administered for 4 weeks in combination with 180 µg peginterferon
alfa-2a (PEG-IFNα-2a) ± 1000–1200 mg ribavirin (RBV) in HCV genotype 1
treatment-naive patients. 104 patients were randomized to: Dual Low: 1500 mg
bid + PEG-IFNα-2a (n=21); Dual High: 3000 mg bid + PEG-IFNα-2a
(n=32); Triple Low: 1500 mg bid + PEG-IFNα-2a + RBV (n=31); SOC:
PEG-IFNα-2a + RBV (n=20). Virological response was measured by Roche COBAS
TaqMan (undetectable <15 IU/mL).
Results:
At 4 weeks HCV RNA was undetectable in 81% of patients
treated with Triple Low (mean reduction of 5.2 log10 IU/mL), 69% treated with
Dual High (mean reduction of 4.5 log10 IU/mL), and 33% treated with Dual Low
(mean reduction of 3.6 log10 IU/mL), compared to only 5% of patients treated
with SOC (mean reduction of 2.4 log10 IU/mL). Synergy was observed between
R1626 and SOC (additional 2.8 log10 IU/mL, Triple Low vs. SOC), and between
R1626 and RBV (additional 1.6 log10 IU/mL, Triple Low vs. Dual Low) (Figure).
ALT normalized in approximately 50% of patients in R1626 treatment groups.
Most adverse events (AEs) were mild or moderate. Six
patients had 8 serious AEs during the 4-week period: 4 in Dual High, 1 in
Triple Low and 1 in SOC. The incidence of Grade 4 neutropenia was 48%, 78%, 39%
and 10% in Dual Low, Dual High, Triple Low and SOC, respectively and was the
main reason for dose reductions.
Conclusion:
In conclusion, a robust synergistic antiviral effect
was observed when R1626 is combined with PEG-IFNα-2a ± RBV, with up to 81%
of patients undetectable by week 4. Dosing of R1626 may be limited mainly by
neutropenia; additional studies of different dosages of R1626 in combination
with PEG-IFNα-2a and RBV are underway.
Topic:
Current Treatment – Consensus Interferon
B. Bacon ; A. Regev; R. H. Ghalib; G. Morelli; K. D.
Rothstein; M. L. Shiffman; T. Hassanein; J. Hammond
Background:
Approximately 50% of HCV treatment-naďve patients
treated with pegIFN/RBV fail therapy. A controlled trial of re-treatment of
pegIFN/RBV non-responders with a subsequent course of pegIFN/RBV resulted in
only a 3% sustained response. Daily use of consensus interferon (CIFN), which
in vitro has been shown to be more effective than IFN alfa-2a and IFN alfa-2b,
allows constant pharmacologic pressure on HCV and, when combined with RBV, may
be an effective treatment option for previous pegIFN/RBV non-responders.
Methods:
The DIRECT clinical trial was a Phase 3, open-label,
multi-center US-based study that enrolled 343 patients who were previous
non-responders to pegIFN/RBV therapy. Patients were randomized to receive 48
weeks of CIFN 9μg/day or CIFN 15μg/day both with RBV (1.0-1.2 g/day).
Prior to study entry, patients had to have <2log10 decline in viral load
(VL) while undergoing at least 12 weeks of previous pegIFN/RBV therapy or have
detectable VL at end of treatment after completing at least 24 weeks of therapy
with ≥80% adherence. Patients had a negative VL if virus was undetectable
by both bDNA and TMA assays. No adjunctive growth factors were used.
Results:
Patient demographics included mean age 50 yr, 71%
male, 68% high viral load, 94% genotype 1, mean weight 89 kg, 17% African
American, and 67% Caucasian. Other important patient characteristics included
59% with evidence of bridging fibrosis or cirrhosis on biopsy and 79% null
responders (<2log10 drop in VL during their previous pegIFN/RBV). In the intent-to-treat
analysis, SVR rates were 5% and 10% for CIFN 9μg/day and CIFN
15μg/day, respectively. SVR rates by previous response to therapy and
fibrosis score for available patients are listed in Table 1. Discontinuations
due to adverse events occurred in 11% and 17% in the CIFN 9μg/day and CIFN
15μg/day arms, respectively.
Conclusions:
Use of daily CIFN plus RBV in patients with advanced
liver disease who were primarily null responders to previous pegIFN/RBV therapy
resulted in viral clearance in a dose-dependent manner. Non-cirrhotic patients
who achieved a partial response on their previous pegIFN/RBV treatment achieved
the best outcomes, which approached 30% SVR. Daily CIFN up to 15μg/day and
RBV was well tolerated in this population of
pegIFN/RBV non-responders.
Table 1: Sustained Virologic Response (%) By Previous Response to Therapy and Fibrosis Score
|
|||||||||||||||||||||||||||||
Topic:
Experimental Therapies - Telaprevir
175.
Final Results of Patients Treated with Peg-Interferon-Alfa-2a (Peg-IFN) and
Ribavirin (RBV) Follow-on Therapy After
28-Day Treatment with the Hepatitis C Protease Inhibitor Telaprevir (VX-950),
Peg-IFN and RBV
M. Rodriguez-Torres; E. J. Lawitz; J. G. McHutchison
Purpose:
Telaprevir (TVR, VX-950) is an orally administered,
highly selective peptidomimetic inhibitor of the Hepatitis C virus (HCV) NS3-4A
protease. The VX05-950-102 study was designed to assess the safety of
telaprevir when given in combination with Peginterferon alfa-2a (Peg-IFN) and
ribavirin (RBV) and to evaluate the antiviral response during 28 days of
dosing. After completion of the 28-day study, all subjects received off-study
therapy with Peg-IFN/RBV under the clinical care of their physicians. Here we
report the outcome of treatment after this post-study therapy.
Methods:
This study included 12 treatment-naďve, mostly Latino
(10/12),patients infected with genotype 1. All
subjects received TVR (750 mg q8h), Peg-IFN alfa-2a (180 µg weekly), and RBV
(1000 or 1200 mg daily). At the completion of the 28 days, patients began
off-study follow-on therapy with Peg-IFN alfa -2a/RBV.
Results:
TVR/Peg-IFN/RBV was well tolerated in the 28-day
study, with no serious adverse events. The adverse event profile was consistent
with the profile commonly seen with Peg-IFN/RBV therapy. All subjects
demonstrated a response to the study drug regimen, with 2 subjects reaching
undetectable (< 10 IU/mL, Roche Taqman® Assay) levels of plasma HCV RNA
within 8 days of the start of dosing, and all subjects had undetectable HCV RNA
at the end of the 28-day study dosing period. At 12 weeks of follow-on therapy
after completing the 28-day study dosing, 11 subjects had undetectable HCV RNA.
All subjects continued on Peg-IFN/RBV therapy, and were followed for response
in accordance with standard practice. Seven patients received a total of 48
weeks of treatment and achieved SVR. One patient received Peg-IFN/RBV for only
18 weeks (total treatment 22 weeks) before discontinuing, but also achieved
SVR. Two patients had viral breakthroughs at 12 weeks and 24 weeks of
treatment.Two patients were lost to follow up and were undetectable at last
assessment. In total, 8/10 patients for whom results are available, achieved
SVR. The side effect profile observed during the post-study dosing was
consistent with the expected profile of Peg-IFN/RBV therapy.
Conclusions:
The rapid and substantial initial antiviral effect of
telaprevir was maintained by the majority of patients during post-study therapy
with PegIFN/RBV. The observation that SVR was achieved in eight patients,
including 1 who completed only 22 weeks of treatment, suggests that
telaprevir-based regimens may allow increased SVR rates as compared to current
therapies.
Topic:
Current Treatment - General
J. Klarquist; L. Golden-Mason; A. S. Wahed; H. R.
Rosen
Introduction:
African Americans (AA) with chronic HCV are less
likely to have a sustained virological response (SVR) to IFN-based antiviral
therapy than Caucasian Americans (CA), but mechanisms for these differences are
not fully elucidated. Host immune responses are postulated to affect viral
kinetics and ultimate outcome of treatment.
Aim:
To asses the association of PD-1, recently shown to be
an important marker for CTL exhaustion, with early viral decline and ultimate
response to treatment in AA and CA with chronic HCV. Methods: 72 genotype 1
patients (30 AA, 42 CA) with chronic HCV were selected from the Virahep-C study
based upon presence of relevant HLA alleles. All patients received
peginterferon and ribavirin for 24-48 weeks; 31 (43%) experienced SVR. Poor
viral decline was defined as less than 1.4 log10 drop in HCV RNA levels between
baseline and day 28 (n=28); intermediate, 1.4-3.5 log10 drop (n=24); and
marked, >3.5 log10 drop (n=18). Pentamers incorporating 10 HCV epitopes were
synthesized to stain PD-1 on a total of 168 pre-treatment, individual,
HCV-specific CTL responses.
Results:
AA with poor kinetics demonstrated significantly
higher pretreatment PD-1 on HCV-specific CTLs than those with marked or intermediate
responses (p=.0005). Remarkably, PD-1 on HCV-specific CTLs was 15% higher among
AA poor responders compared to the intermediate group. Among CA, pretreatment
PD-1 was not significantly associated with early viral kinetics. PD-1 as a
predictor of SVR was examined using Poisson regression analysis, with
adjustment for race. Among AA, PD-1 expression on HCV-specific CTLs was
significantly, inversely associated with SVR (RR=0.95; 95% CI 0.93-0.97;
p<.0001). This association, however, was not significant in CA (Figure).
Conclusions:
In this large cohort of prospectively characterized patients, pre-treatment PD-1 expression on HCV-specific CTLs was associated with early and long-term virologic response to combination antiviral therapy in AA but not CA patients. These findings suggest that immune factors may contribute to the poor antiviral responses found among AA patients with hepatitis C and provide potential immune targets for manipulation.
Topic:
Experimental Therapies - Telaprevir
I. M. Jacobson; G. T. Everson; S. C. Gordon; R.
Kauffman; L. McNair; A. Muir; J. G. McHutchison
Background:
The VX05-950-104 study (PROVE1) is a randomized, placebo-controlled
Phase 2 study of telaprevir (TVR, VX-950), an HCV protease inhibitor, with
Peg-IFN-2a and RBV, in naive subjects with genotype 1 hepatitis C. We report
the results of a planned interim analysis.
Methods:
Subjects were randomized into 4 groups; 3 groups
received TVR 750 mg q8h, Peg-IFN-2a 180 μg/week, and RBV 1000-1200 mg/day
for 12 weeks followed by 0 (n=20), 12 (n=80) or 36 (n=82) weeks of Peg-IFN-2a
and RBV (TVR/PR groups). The control group (n=81) received up to 48 weeks of Peg-IFN-2a/RBV
(PR). This analysis was performed when all treated subjects had completed 12
weeks of dosing. Follow-up information was also obtained for the subjects who
completed the 12-week dosing arm. Plasma HCV RNA was isolated for viral
sequencing at baseline and at each HCV RNA assessment in samples that contained
greater than 1,000 IU/mL.
Results:
The proportion of subjects with undetectable HCV RNA
(LOD 10 IU/mL) at Week 4 was 79% (TVR/PR group), and 11% (control group)
(p<0.001), and at Week 12 was 70% (TVR/PR group), and 39% (control group)
(p<0.001). Viral breakthroughs (increase of > 1-log above HCV RNA nadir)
with previously described TVR resistant variants were observed in 12 of 175 (11 genotype 1a and 1 genotype 1b) subjects in the TVR/PR
groups. In the group assigned to 12 weeks of treatment, 6 of 9 subjects with
RVR and completing 12 weeks of treatment achieved SVR. The remaining 3 subjects
relapsed with TVR resistant variants (genotype 1a: R155T/I). Discontinuations
due to adverse events were more frequent in the TVR/PR group (11% vs. 3%).
Rashes, gastrointestinal events and anemia were more common, and rashes more
severe in the TVR/PR groups. An additional interim analysis will be conducted
when all subjects have completed 36 weeks on-study, including 12-week
post-treatment follow-up for the 24-week study arm subjects.
Conclusions:
Compared with standard therapy, significantly more
subjects receiving a TVR/PR regimen achieved undetectable HCV RNA at Weeks 4
and 12, and some subjects achieved an SVR after 12 weeks of TVR/PR. Other
patients relapsed with TVR-resistant variants, suggesting that 12 weeks of
TVR/PR may be adequate to eliminate wild-type virus. The post-treatment
follow-up results in the 24-week group will evaluate whether the additional 12
weeks of PR is sufficient to eliminate the remaining TVR-resistant variants.
* for the PROVE1 study team
Topic:
Experimental Therapies - NTZ
178. Interim Data from a Randomized Controlled Trial of Nitazoxanide-Peginterferon-Ribavirin,
Nitazoxanide-Peginterferon and Peginterferon-Ribavirin in
the Treatment of Patients with Chronic Hepatitis C Genotype 4
J. Rossignol1; A. Elfert; Y. El-Gohary; E. B. Keeffe;
J. Glenn
Introduction:
Nitazoxanide (NTZ), a thiazolide anti-infective agent
approved in the United States for treating emerging intracellular protozoan
infections, blocks viral protein synthesis at a cellular level by selectively
inhibiting dephosphorylation of eukaryotic initiation factor 2α (eIF2α).
A randomized controlled clinical trial was conducted to evaluate the efficacy
and safety of NTZ added to peginterferon α-2a (PegIFN) plus ribavirin
(RBV) or PegIFN without RBV in treating chronic hepatitis C genotype 4.
Methods:
120 patients with chronic hepatitis C genotype 4 were
enrolled in the trial at two centers in
Results:
Baseline viral loads, BMI, fibrosis, age, sex and
other demographic and disease-related characteristics were similar across the
three groups. RVR, EVR and ETR rates are presented in the table below. Adverse
events were similar across the three treatment groups except that the rate of
anemia was significantly reduced in the group that did not receive RBV.
Conclusions:
The addition of NTZ to PegIFN or PegIFN-RBV improved
RVR, EVR and ETR rates compared with PegIFN-RBV therapy without increase in
adverse events. Patients are being followed to evaluate SVR rates. Further
trials are being conducted in the
|
|
Treatment-naive |
Treatment-experienced |
||||
|
|
RVR |
EVR |
ETR |
RVR |
EVR |
ETR |
|
PegIFN-RBV |
14/40 (35%) |
26/40 (65%) |
20/40 (50%) |
- |
- |
- |
|
NTZ-PegIFN |
18/29 (62%) |
22/29 (76%) |
20/29 (69%) |
5/11 (45%) |
5/11 (45%) |
5/11 (45%) |
|
NTZ-PegIFN-RBV |
21/29 (72%) |
26/29 (90%) |
27/29 (93%) |
3/11 (27%) |
4/11 (36%) |
4/11 (36%) |
Topic:
Current Treatment - General
T. Berg; V. Weich; G. Teuber; H. Klinker; B. Möller;
J. Rasenack; H. Hinrichsen; T. Gerlach; U. Spengler; P. Buggisch; H. Balk; M. Zankel; C. Sarrazin; S. Zeuzem
Introduction:
The exact estimation of early virologic response rates
in the course of antiviral therapy is an important goal in order to improve
individualized therapeutic strategies in HCV infection. The sensitive TMA test
could provide better advantage to distinguish at early stages sustained from
non-sustained responders.
Method:
We evaluated HCV type 1 patients who took part in a
prospective study asking whether the application of TMA in bDNA-negative
patients may be a better indicator to predict long-term outcome of the HCV
infection.
433 patients were randomized to receive either
1.5ug/kg PEG-INFa-2b plus 800-1400 mg RBV for 48 weeks (n=225, group A) or an
individualized tailored treatment duration (n=208, group B). In the latter
group treatment duration was calculated by the time required to become for the
first time HCV RNA negative as defined by bDNA assay (detection limit 615
IU/mL) multiplied by the factor 6. HCV RNA levels were quantified weekly until
week 8, at week 12 and 24. For all those patients who were bDNA negative the
more sensitive TMA test (detection limit 5.3 IU/mL) was also prospectively
assessed. The different response groups were classified according to the HCV
RNA levels at week 4 and 12 (Table).
Results:
Table shows the relevant data and refers to the
relative relapse rates in group A and B at week 4 and 12 in relation to the
treatment schedule. There is clear evidence for a high relapse rate in patients
with a positive TMA within the first 12 weeks of therapy being more pronounced
when treatment duration shortened in the individualized treatment group. In
contrast patients shown to respond as early as week 4 evidenced by a negative
TMA test had relapse rates below 10% irrespectively from treatment group.
Conclusion:
The application of the highly sensitive HCV RNA tests must be considered to be now mandatory because this new test helps to evaluate in a much more refined way treatment response as well as relapse rates and provides better clues for an individualized tailored treatment strategy. Thus our study clearly indicates that patients even with a minimal amount of HCV RNA detected at week 12 can suffer from relapse rates greater than 50%. These patients may indeed benefit when their treatment duration is adapted to their individual needs.
|
Response groups |
Relative relapse rate (%) |
|||
|
|
|
|
||
|
week 4 response |
>= log decline, bDNA positive |
36% (all patients) |
19% (group A) |
63% (group B) |
|
bDNA negative, TMA positive |
38% (all patients) |
22% (group A) |
49% (group B) |
|
|
bDNA negative, TMA negative |
4% (all patients) |
0% (group A) |
8% (group B) |
|
|
week 12 response |
>= 2log decline, bDNA positive |
77% (all patients) |
78% (group A) |
75% (group B) |
|
bDNA negative, TMA positive |
64% (all patients) |
56% (group A) |
69% (group B) |
|
|
bDNA negative, TMA negative |
20% (all patients) |
9% (group A) |
32% (group B) |
|
Topic:
Current Treatment - Pegasys
180. Sustained
Virologic Response rates with albinterferon alfa-2b plus ribavirin treatment in
IFN-naďve, chronic hepatitis C genotype 1 patients
S. Zeuzem; E. M. Yoshida; Y. Benhamou; V. G. Bain; D.
Shouval; S. Pianko; R. Flisiak; M. Grigorescu; V. Rehak; K. D. Kaita; P.
Cronin; E. Pulkstenis; G. M. Subramanian; J. G. McHutchison
Aim/Background:
This Phase 2b study evaluates the efficacy and safety
of albinterferon alfa-2b (alb-IFN), a novel recombinant protein consisting of
IFNa-2b genetically fused to human albumin, in combination with oral ribavirin (RBV)
in IFN-naďve, chronic hepatitis C genotype 1 patients.
Methods:
458 patients were randomized to 4 treatment arms:
PEG-IFNa-2a 180mcg Q1w (PEG-IFN) or one of 3 alb-IFN arms (900mcg Q2w, 1200mcg
Q2w or 1200mcg Q4w), all in combination with weight-based oral RBV1000-1200
mg/d (Table). HCV RNA was measured by real-time PCR (limit of detection [LOD]:
10 IU/mL). The primary efficacy endpoint was sustained virologic response (SVR)
defined as HCV RNA
Results:
By ITT analysis, SVR rates were 58.5% for the 900Q2,
55.5% for 1200Q2, 50.9% for 1200Q4 and 57.9% for PEG-IFN (Table). In patients
with ≥80% adherence to therapy, SVR rates were 71.6% for the Q2w alb-IFN
arms compared to 66.7% for PEG-IFN. Of note, in heavier patients (≥75 kg)
adherent to therapy, SVR rates were maintained for the alb-IFN arms and
declined for PEG-IFN. Discontinuations due to adverse events in the alb-IFN
arms were 9.3% in the 900Q2, 18.2% in the 1200Q2 and 12.1% in the 1200Q4
compared with 6.1% for PEG-IFN. Hematologic reductions were lowest in the Q4w
arm and comparable across other arms. Quality-of-life as measured by SF-36v2
and missed work was significantly more favorable for the 900Q2w arm.
Conclusions:
These data suggest that the Q2w alb-IFN may offer
comparable efficacy, with an improved dosing schedule and quality-of-life
compared with current standard of care with PEG-IFN.
Topic: Diagnostic Tools
D. Lucidarme; J. Foucher; B. Le Bail; L. Castera; S.
Villars; G. Forzy; B. Filoche; P. Couzigou; V. de Lédinghen
Background:
The result of transient elastography using FibroScan
is expressed as a median value of 10 liver stiffness measurements (LSM). To our
knowledge, the role of the interquartile range (IQR) and the ratio IQR/LSM for
the performance of FibroScan has not been studied. The aim of this study was to
evaluate the influence of these 2 parameters on the accuracy of transient
elastography for the diagnosis of liver fibrosis and cirrhosis.
Patients and
methods:
A total of 254 consecutive HCV patients (mean age 52
y; males 60%, BMI 25.0 Kg/m2) had LSM using FibroScan and liver biopsy (LB)
> 15mm (median 22 mm, range 15-50 mm) within one month. Liver fibrosis was
staged according to the METAVIR scoring system and cut-off values of LSM
according to those previously reported (1). Discordances of at least 2 stages
between LSM and histological assessment were observed in 28 cases (11 %). The
effect of age, gender, BMI, LB length, activity grade, LSM, success rate of
LSM, IQR and IQR/LSM on discordances was assessed by comparing the 28
misclassified cases to the 226 others. The influence of one or several possible
predictive factors of discordances was assessed by using the receiver operating
characteristics (ROC) curves after assessment of the most significantly
discriminant cut-off values.
Results:
By univariate analysis, LSM and IQR/LSM were
associated with discordances (P< 0.10). By multivariate analysis, IQR/LSM
was the only factor independently associated with discordances (P < 0.01).
The most significantly discriminant cut-off value was IQR/LSM = 0.2. For
IQR/LSM < 0.2 versus > 0.2, discordances of at least two stages of
fibrosis were observed in 8/127 cases (6.3 %) versus 20/127 cases (15.7 %), P
< 0.02, respectively. In patients with IQR/LSM > 0.2 versus < 0.2, for
the diagnosis of liver fibrosis F > 2, F > 3 and cirrhosis, AUROCs were
0.77 versus 0.83 (P = 0.18), 0.79 versus 0.91 for (P < 0.01) and 0.86 versus
0.95 (P = 0.11), respectively.
Conclusion:
IQR/LSM is a key factor of accuracy of FibroScan for
the diagnosis of liver fibrosis. This ratio significantly improves the
performance of FibroScan for the diagnosis of extensive fibrosis (F3). In case
of IQR/LSM > 0.2, a new measurement of LS or a LB should be proposed.
1) Castera et al.
Gastroenterology 2005;128:343-50.
Topic:
Diagnostic Tools
S. M. Martinez; M. Domínguez; G. Fernandez-Varo; P.
Gonzalez; R. Bataller; E. Sampson; W. Jimenez; X. Forns; J. M. Sánchez-Tapias
Background:
Prediction of liver fibrosis by means of noninvasive
tests is intended to replace liver biopsy in the assessment of disease severity
in chronic hepatitis C (CHC).
Aim:
The aim of this study was to evaluate the diagnostic
performance of several noninvasive methods, alone or in combination, in the
detection of significant liver fibrosis or cirrhosis in patients with chronic
hepatitis C.
Methods:
The performance to predict liver fibrosis was
investigated in a cohort of 373 patients with CHC who were candidates to
antiviral therapy by the following non-invasive methods: Forns score, AST to
platelet ratio index (APRI), liver stiffness measurement (LSM) by transient
elastography, and the extracellular matrix markers hyaluronic acid (HA),
collagen IV(Col4), amino-terminal propeptide of type
III collagen (PIIINP), and tissue inhibitor of matrix metalloproteinase
1(TIMP-1). The latter markers were combined in two models: ELF Scheuer (age,
PIIINP, TIMP-1 and HA) and 4-marker (PIIINP, TIMP-1, Col4 and HA) (Rosenberg et
al. Gastroenterol 2004). Liver fibrosis stage was assessed according to Scheuer
classification. Diagnostic accuracies of each method were determined by the
area under the ROC curve (AUROC). The predictive value of combinations of these
methods was assessed by multiple regression.
Results:
Mean age was 47,26 (17-69)
years, and 251 (62,6%) were male. The distribution of liver fibrosis stages
was: F0: 11,5% , F1: 25,9%, F2: 25,7%, F3: 9,5% and
F4: 27,4%. The AUROCs of the different methods for F ≥2 and F4 are shown
in table 1. The best diagnostic performance was obtained with combination of
APRI index and FibroScan for F ≥ 2 (AUROC = 0,88),
and Forns Score and FibroScan (AUROC= 0,90) for F4.
Conclusions:
These results show that non-invasive methods are
accurate in the prediction of significant fibrosis and cirrhosis. Combinations
of these methods improve the diagnostic accuracy, with best diagnostic
performance obtained with combination of transient elastography and indirect
markers of fibrosis.
Topic: Diagnostic Tools
L. Huwart; C. Sempoux; N. Salameh; L. Annet; Y. Horsmans; B. Van Beers
Introduction:
Recently,
magnetic resonance (MR) elastography has emerged as a non-invasive method to
assess the fibrosis stage. The purpose of our study was to prospectively
compare the diagnostic accuracy of MR elastography,
ultrasound elastography, and aspartate aminotransferase to platelets ratio
index (APRI) for the staging of fibrosis in patients with chronic liver
disease.
Materials and methods:
Ninety-six
patients who had liver biopsy for suspicion of chronic liver disease had MR elastography, ultrasound elastography and APRI within 2
days. Fibrosis stage was assessed by two pathologists according to the METAVIR
score (from F0 to F4) that served as reference. MR elastography
was performed with a 1.5 Tesla scanner using a motion-sensitized sequence and a
transducer that applied mechanical waves into the liver. Ultrasound
elastography was performed with the commercially available FibroScan. The
measurements of MR elasticity, ultrasound elasticity and APRI were analyzed
with receiver operating characteristic curves.
Results:
The
distribution of fibrosis stage was F0 in 22/96 patients (23%), F1 in 22 (23%),
F2 in 19 (20%), F3 in 15 (15%) and F4 in 18 (19%). The areas under the receiver
operating characteristic curves of MR elastography (0.994 for F ≥ 2,
0.985 for F ≥ 3 and 0.998 for F = 4) were significantly (P < 0.05)
larger than those of ultrasound elastography, APRI and the combination of
ultrasound elastography and APRI (respectively 0.837, 0.709 and 0.849 for F ≥
2; 0.906, 0.816 and 0.936 for F ≥ 3; 0.930, 0.820 and 0.944 for F = 4).
The most discriminant cut-off values of MR elasticity were respectively 2.5 kPa
for F ≥ 2, 3.1 kPa for F ≥ 3, and 4.1 kPa for F = 4.
Conclusion:
The results of
this study show that MR elastography is more accurate
than ultrasound elastography and APRI to stage liver fibrosis and suggest that
MR elastography can be used for the assessment and follow-up of liver fibrosis.
|
|
F ≥ 2 |
F ≥ 3 |
F = 4 |
|
Cut-off (kPa) |
2.49 |
3.13 |
4.13 |
|
Sensitivity |
1.0 (0.93-1.0) |
0.91 (0.76-0.98) |
1.0 (0.81-1.0) |
|
Specificity |
0.91 (0.78-0.97) |
0.97 (0.89-1.0) |
0.96 (0.89-0.99) |
|
Positive predictive value |
0.93 (0.83-0.98) |
0.94 (0.79-0.99) |
0.86 (0.64-0.97) |
|
Negative predictive value |
1.0 (0.91-1.0) |
0.95 (0.87-0.99) |
1.0 (0.95-1.0) |
Table.
Most discriminant MR elasticity cut-off values (kPa) and corresponding sensitivities,
specificities, positive and negative predictive values with 95% confidence
intervals in parentheses for METAVIR scores F ≥ 2, F ≥ 3 and F = 4.
Topic: Diagnostic Tools
185.
Fibrotest versus liver biopsy: An independent multicenter evaluation of
performance
M. Guido; A. Alberti; G. Bellati; G. Colloredo; A.
Craxě; V. Di Marco; S. Fagiuoli; M. Fassan; L. Giacomelli; A. Mangia; G.
Sebastiani; M. Rugge
BACKGROUND
AND AIMS:
Algorithms based on indirect biochemical markers of
fibrosis are becoming increasingly popular. Though the Fibrotest (FT) is
accredited for the non-invasive assessment of liver fibrosis, its validation
against liver biopsy comes mainly from data generated by the center where it
was originally developed. In an independent nationwide multicenter study, we
aimed to assess the diagnostic accuracy of FT in a large series of patients
with chronic hepatitis B and C undergoing liver biopsy (LB) for disease grading
and staging. This is the first multicenter study in which all liver biopsies
were reviewed at a single center by the same experienced liver pathologist.
METHODS:
In all, 496 consecutive patients with chronic HBV (55
cases) or HCV hepatitis were enrolled at five centers where liver biopsies and
biochemical markers were collected on the same day. All subjects were HIV
negative. The diagnostic value of the FT was assessed, applying logistic
regression as a discriminant function, in terms of area under the receiver
operating characteristic (ROC) curves (AUCs), sensitivity, specificity, PPV,
NPV, general agreement, and kappa score.
Liver biopsies more than 2 cm long & with ≥
11 complete portal tracts (= 277 cases) were considered the “platinum”
standard: all analyses were repeated in these biopsies to test the hypothesis
that FT might significantly improve its diagnostic accuracy with a better gold
standard, assuming that the discrepancy between FT and LB is due mainly to poor
biopsy quality.
RESULTS:
73% of the biopsy samples were ≥ 2 cm long
(median 2 cm; range: 0.5–6.5). According to the METAVIR score, histological
fibrosis was F0 in 8% (39/496) of cases, F1 in 29% (142/496), F2 in 35%
(176/496), F3 in 13% (63/496), and F4 in 15% (76/496). FT proved scarcely able
to discriminate between single fibrosis stages and 34% of cases had
intermediated values and cannot be classified. Table 1 summarizes the
diagnostic value of FT in discriminating between absent/mild fibrosis and
severe fibrosis/cirrhosis. The results did not change when the analysis was
restricted to the “platinum” standard biopsies.
CONCLUSIONS:
FT performs best in discriminating severe
fibrosis/cirrhosis from absent/mild fibrosis, but it does not accurately
predict the various intermediate stages of fibrosis. Its performance does not
improve significantly with better-quality liver biopsies.
Table 1
|
|
AUC-ROC |
SENS |
SPEC |
NPV |
PPV |
Agreement |
k |
|
F0/F1 vs. F2/F4 |
0.77 |
0.71 |
0.72 |
0.59 |
0.82 |
71.98 |
0.42 |
|
F0/F2 vs. F3/F4 |
0.82 |
0.64 |
0.83 |
0.85 |
0.64 |
78.43 |
0.47 |
Topic:
Diagnostic Tools
L. Castera; G. Sebastiani; B. Le Bail; V. de Ledinghen; P. Couzigou; A. Alberti
Background:
Non invasive assessment of liver fibrosis is a
challenging area. Numerous markers are available but their performance may be
better when combined. Two algorithms have been proposed recently in patients
with chronic hepatitis C, combining either transient elastography (Fibroscan)
and Fibrotest (Castera et al. Gastroenterology 2005) or Fibrotest and AST to
Platelet Ratio (APRI) (Sebastiani et al. J Hepatol 2006).
Aim:
to prospectively compare the diagnostic
performance for significant fibrosis and cirrhosis of these 2 algorithms in the
same population.
Methods:
302 consecutive HCV patients (58% male, mean age 52±12
yrs) who undergone liver biopsy (>10mm and 6 portal triads) were studied.
Fibroscan, Fibrotest and APRI were performed in the same lab the day of liver
biopsy taken as reference. Fibrosis was scored according to the Metavir scoring
system. Diagnostic performances were assessed by measuring area under the ROC
curve (AUROC).
Results:
Significant fibrosis (F2-F3-F4) was present in 76% of
patients and cirrhosis (F4) in 24%. The mean liver biopsy length was: 20±8 mm.
Fibroscan failure was observed in 8 cases (3%). Performance of the 2 algorithms
using the cutoffs for APRI and fibrotest of the original studies are shown in
the table.
For significant fibrosis, Castera algorithm saved 23%
more liver biopsies than Sebastiani algorithm but its accuracy was lower (AUROC
0.91 vs. 0.94; accuracy 87.1% vs. 97.0%, respectively). Regarding cirrhosis,
Castera algorithm was slightly more accurate (AUROC 0.93 vs. 0.87; accuracy
95.7% vs. 88.7%, respectively) and saved 4% more liver biopsies. However,
Castera algorithm was less cost-effective (using both 100% Fibroscan and
Fibrotest) than Sebastiani algorithm (using 100% APRI (virtually no cost) and
around 50% Fibrotest).
Conclusion:
Both algorithms are effective for the non invasive
staging of liver fibrosis in chronic hepatitis C. Their use in clinical
practice could avoid the need for liver biopsy in 48 to 71% of cases for the
diagnosis of significant fibrosis and in 74 to 78% of cases for cirrhosis.
|
|
Castera algorithm |
Sebastiani algorithm |
||
|
End Point |
F2-F3-F4 |
F4 |
F2-F3-F4 |
F4 |
|
APRI performed |
none |
none |
100 |
100 |
|
Fibrotest performed |
100 |
100 |
43.7 |
61.9 |
|
Fibroscan performed |
100 |
100 |
none |
none |
|
Sensitivity (%) |
85.1 |
89.4 |
100 |
86.4 |
|
Specificity (%) |
97.2 |
98.2 |
87.3 |
89.7 |
|
PPV (%) |
98.9 |
95.0 |
96.3 |
77.6 |
|
NPV (%) |
66.9 |
95.9 |
100 |
94.1 |
|
Accuracy (%) |
87.1 |
95.7 |
97.0 |
88.7 |
|
AUROC |
0.91 |
0.93 |
0.94 |
0.87 |
|
Saved biopsies (%) |
71.9 |
78.8 |
48.3 |
74.8 |