Hepatitis B and C Coinfection Report
Written by Nancy Shulman, MD, Stanford University
Hepatitis B and C co-infection have now made it to prime time. There were over 40 abstracts and a symposium on hepatitis coinfection at this year's retrovirus conference and not only are the numbers of abstracts up but the quality of the studies are improving. The "big boys" like the ACTG and the pharma companies have gotten into the action. Below I have summarized some of the abstracts by topic.
--Response to treatment in HCV coinfected patients
--Response to treatment in HBV coinfected patients with 3TC resistance
--Liver transplantation in HIV coinfected
--Does HCV blunt CD4 response after HAART
--Does Hepatitis coinfection impact survival?
--HIV+ persons not receiving HBV vaccine
Treatment of HCV in HIV:
We heard interim results of ACTG study 5071 that randomized 133 HCV-HIV coinfected patients to receive either Pegylated interferon-alfa 2a (Pegasys, Roche) at 180mg per week for 48 weeks or Interferon-alfa 2a 6 million units three times per week for 12 weeks followed by 3 MIU 3 times per week. Both groups received ribavirin 600mg per day escalated to 1 gm per day as tolerated. All patients had CD4 >100/ml and HIV RNA of <10,000 if they were on antiretroviral therapy. A small subset of patients with CD4 >300 were not on antiretroviral therapy.
Results - At the end of 24 weeks of therapy, 44% were HCV RNA <50 copies/ml in the pegylated arm vs. only 15% in the standard interferon arm by intent-to-treat analysis where dropouts are considered failures. In Genotype I patients who are generally less responsive to treatment than other genotypes, the numbers were less at 33% and 7%. Genotypes 2 and 3 combined were had much higher rates of virologic response at 80% and 40% in the Peg and non-peg arms respectively. This is less than would be expected in HCV mono-infected patients. Overall about 14% dropped out of the study due to side affects of the medications, similar to studies of HCV mono-infected patients. We await the sustained virologic response rates (HCV RNA negative 6 months after completing therapy).
Another large study by the HRN (Hepatitis Resource Network) randomized co-infected patients to receive either daily or three times per week standard interferon with ribavirin 800mg/day. 180 patients were enrolled, 90 in each arm. CD4 counts were all >100. This study enrolled about 50% non-whites (higher than other studies). By 12 weeks, 23% had discontinued equally distributed between the two groups, mainly for side affects (flu-like illness and neuro-psychiatric complaints were the main ones). 25% enrolled in the daily interferon group vs. 10% in the three time a week group had undetectable HCV RNA at 12 weeks (both less than the ACTG results with pegylated at 24 weeks).
In a follow-up from their IAS presentation, the Madrid group presented their study using Peg-Intron (Schering) plus ribavirin in a single armed open-label study. 65 coinfected patients with CD4>300 cells/ml and HIV RNA <5000 copies, received Peg-Intron 150mg every week for 3 months, followed by 100mg for 3 more months in genotype 2 or 3 and 9 months in genotype 1 or 4 (a genotype in North Africa that is less responsive to interferon treatment). The ribavirin dose was 800mg/day. Overall 54% were HCV RNA negative at the end of treatment, 37% in genotypes 1 and 4 and 63% in 2 and 3. This is similar to the ACTG results at 24 weeks. The current sustained response rate was 33%. Discontinuation of therapy occurred in 14%, mainly due to flu-like symptoms and neuro-psychiatric complications.
HCV Viral Kinetics
Written by Nancy Shulman MD, and Jules Levin
Two studies looked at how fast the viral load drops after pegylated
interferon or standard interferon in HCV/HIV patients. F. Torriani (UCSD)
reported the data from the Apricot study. K. Sherman (Hepatologist at
University of Cincinnati) reported the data from ACTG 5071. 10 patients, all
genotype 1, from the large Roche study (Apricot) were randomized patients to
full dose Pegasys + ribavirin (800mg/day), full dose Pegasys + ribavirin
placebo, vs. interferon (3 MIU/3x/day) + ribavirin (800mg/day). In this
subanalysis 5 patients received Pegasys+ribavirin and 5 received
interferon+RBV. 10 patients, 8 who were genotype 1, in the ACTG 5071 (5 Peg
and 5 interferon) were also studied and the data presented. Patients were
randomized to receive Pegasys (full dose, 180 mcg once weekly + ribavirin
600-1000mg/day (dose escalated), or interferon (6 MIU 3x/week initially and
then 3 MIU 3x/week) + ribavirin (600-1000mg, also dose escalated) All
patients had frequent blood draws in the first few days to determine the rate
HCV levels decline after initiating treatment.
Results: The Roche data was difficult to interpret due to the great degree
of variability in the patient responses. They did not see the characteristic
biphasic reduction of viral load (a steep initial reduction, followed by a
slower second phase). In the presentation on ACTG 5071, Sherman described
phase 1 and 2. It is suggested that there are two phases of HCV decline. Itıs
suggested that the initiation of HCV therapy results in inhibition of viral pr
oduction and release from hepatocyte cells during phase 1, which appears to
occur in the first 10 days. Phase 1 is associated with a certain level of
efficiency of HCV clearance. When the efficiency is 100%, the virus will be
totally cleared. If this fails to result in 100% efficiency in clearing HCV,
the ultimate outcome depends on the decline of HCV in phase 2, which can
extend as far out as 80 days or longer. Decline in phase 2 is presumably
related to the clearance of infected hepatocyte cells and the continued
inhibition of production at some level of efficiency. The estimation of the
time to clearance of HCV, when phase 1 is not 100% efficient in clearing HCV,
is based on how long it will take to clear HCV through phase 1 and 2. The
model Sherman used to make such estimates is based on this and referred to
below. Bear in mind that clearance of HCV does not necessarily result in
end-of-treatment response or in sustained viral response. Non-adherence and
perhaps other unidentified factors may play roles in achieving sustained
responses. The results from the analysis by Torriani suggest that having HIV
may impair the ability for initial HCV clearance. She reported that in the 5
subjects with HCV RNA decline, the HCV virion half-life was 8.7 hours which
compares to 2.7 hours in HCV monotherapy using 5-15 MIU of daily interferon.
Again, she did not detect a phase 2 decline. Is this may related to the
virion half-life, I donıt know the answer to that.
In the Apricot/Torriani study, four patients responded by 24 weeks, 2 of
whom became undetectable within 5 days and one who became undetectable at 9
days, and the 4th reached undetectable at week 21. These 4 patients received
Pegasys+ribavirin. The 5th patient received interferon+ribavirin and reached
undetectable at 28 days but relapsed. The virologic responders had a lower
pre-treatment HCV RNA, they appeared to be below 1 million. The remaining 5
patients had transient or minimal reduction in HCV RNA; 4/5 received
interferon+ribavirin while 1 patient received Pegasys+ribavirin. Torriani
speculated that the reason she did not observe a phase 2 may be due to the
presence of a second phase below the detection of the limit of the assay. Or,
the interaction with HIV, which might require a modification of the current
model for HCV dynamics. She said the first hypothesis regarding the limits of
the assay makes more sense to her. Her explanation about this did not make
sense to me. In all 4 subjects below the detection limit by 12 weeks, decline
in ALT closely paralled HCV RNA decline, leading her to suggest that frequent
ALT measures may prove useful in monitoring treatment response. She
suggested, and this makes sense to me, that perhaps coinfected patients may
require higher interferon dosing or longer treatment to achieve sustained
viral suppression. Based on statistical modeling, Torriani estimated the efficacy of the initial
viral response was 60% for patients receiving the standard interferon regimen
vs. 99% for patients receiving the Pegasys regimen. But this efficacy data is
based on preliminary initial response data; final study results on patient
response to therapy is not available yet. Clearly, HCV in HIV-infected patients needs more in-depth
study as it appears there may be important and identifiable differences in
response to therapy from HCV monoinfected patients.
The ACTG data was more interpretable. They saw first and second phase HCV
declines, unlike in the Apricot study. The slope of the first phase slope and
the second phase of the Pegasys vs. the interferon were both much steeper
(which generally is accepted to correlate to potency). The lag time to
response (first few hours before viral load starts to decline) was about the
same for Pegasys vs IFN (7.7-9 hours). The overall efficacy of phase 1
response was 90% in the Pegasys patients vs 65% in the interferon patients
(p<0.02). Based on these data and Shermanıs modeling technique, the estimated
time to clear the virus from the body in the Pegasys group was a mere 194
days, where the estimated time clearance for the interferon group was
calculated to be 2400 days, because some did not reduce viral load at all
making the number very high. Sherman said that if the model is correct you
ought to be able to predict response to therapy, based on the initial
response in the early days after starting therapy. In applying the model,
they were able to predict response (n=4) and non-response (n=6) at week 24,
with 100% accuracy, suggesting the validity of their model although the
number of patients is small.
Sherman summarized the following from ACTG 5071. Pegasys appears to be
superior to standard IFN alfa-2a in terms of phase 1 efficacy in coinfected
patients, and in phase 2 decline in viral load. Using this combination of
measures and the model to evaluate the initial response (phase 1 and 2 res
ponses) appears to help estimate initial response, and initial response
appears to correlate with week 24 response and to be superior for Pegasys.
Take home message on HCV treatment: Pegylated interferons are superior to
interferon in HIV+ as it is in HIV- and are the current standard of care. We
donıt have any data to suggest that one Pegylated product is superior to
another in HIV+ patients. HCV+/HIV+ have lower response rates than HCV
1. Abstract LB-15. Chung, et al.
2. Abstract 651. Sulkowski, et al.
3. Abstract 652. Perez-Olmeda, et al.
4. Abstract 121. Torriani, et al.
5. Abstract 122. Sherman, et al.
Treatment of HBV in HIV+:
Adefovir, although it failed as an HIV drug, still survives as a good HBV drug that is active against 3TC resistant HBV. It has not had the kidney toxicity seen at the HIV doses of 60 and 120 mg per day. The dose used in HBV is 10mg per day. A study of 35 HBV/HIV coinfected patients who were on 3TC as a part of their HAART regimen, had detectable HBV DNA, and detectable mutations in their HBV associated with 3TC resistance, and had adefovir 10mg added to their regimen. All had <400 copies HIV RNA. The median HBV DNA at baseline was 8.64 logs/ml. At week 72, the median change in DNA was -4.77 logs/ml including 4 who were undetectable. No one experienced HBV DNA rebound during the study period and no new mutation in the HBV were evident. Two cases of creatinine elevation (kidney toxicity) required stopping all meds, but adefovir was than restarted in both with no recurrence of the creatinine elevations.
Tenofovir, Gilead's successful sequel in the HIV world to adefovir, also has substantial activity against HBV. An analysis was presented from two of their registration trials where tenofovir or placebo was added to an existing antiretroviral regimen that was not fully suppressive. The placebo arm received tenofovir after 24 weeks. 550 patients were enrolled. They wanted to determine the maximum anti-HBV activity of tenofovir, so they selected all patients with HBV DNA levels of 6 million or more. 14 patients met the criteria, 12 in the tenofovir arm and 2 in the placebo arm. The mean change in HBV DNA at 48 weeks was about -5 logs. The placebo group initially had no reduction but had a sharp reduction in HBV DNA upon switching to tenofovir. The 4 who had no 3TC mutations had a greater response (-5.39) vs. the 8 who had mutations (-4.58). Tenofovir looks as good as adefovir for HBV. There is an ACTG study that will compare the two head to head for HIV/HBV coinfection.
1. Abstract 123. Benhamou, et al.
2. Abstract 124. Cooper, et al.
Transplantation in HIV+
Early small reports of liver transplantation of HIV+ at last year's retrovirus conference, particularly those with HCV looked bleak. This year reports are looking much better.
The outcomes of 26 liver transplants were reported that have been performed on HIV+ patients beginning from 9/97 through 12/01 at the University of Pittsburgh, University of Miami, Kings' College in London, University of California San Francisco, and the University of Minnesota. One had a living related donor and the rest were from cadavers. The living related one required re-tranplantation with a cadaveric liver due to too small a donor graft size. 17 were HCV+, 6 were HBV+, and 3 were for fulminant hepatic failure. Median survival time so far is 15 months (range 1-53). 7 patients died: 5 from recurrent HCV, 1 stopped his HAART and his immunosuppressant levels were low and he rejected the liver, and one had severe pancreatitis related to the operation. The median T-cell count at transplant was 192 (range 76-506), which increased to 295 currently. Tolerating HAART after transplant was a significant predictor of survival. One year post-transplant survival rates are similar to HIV- patients. This is progress. Of note, all patients with HCV are maintained on interferon and ribavirin therapy after transplant at this point indefinitely.
1. Abstract 125. Ragni, et al.
Does HCV affect Immune Reconstitution after HAART:
The Swiss cohort study recently published their findings of a diminished CD4 increase in patients who were coinfected with HCV. Several studies addressed this question at the conference.
The ACTG analyzed 60 HCV coinfected patients enrolled in 6 different ACTG studies who had at least 16 weeks of therapy on the trials. They looked at CD4 increase compared to 133 HCV- cases in a case control study. No difference was seen in median CD4 counts, median CD4 increases, or HIV RNA levels in the two groups. They also looked at viral load after therapy initiation and found that HCV RNA increases about 0.5 logs at 16 and 48 weeks in patients with <350 CD4, while in those whose T-cells were >350, HCV RNA increased by a median of only 0.2 logs. ALT also increased in most, but was mild (grade 1 or 2) in nearly all. Dr. Chung was not sure of the reason for the differences in the two, but hypothesized that it could be that they HIV and HCV compete for certain cells like PBMCS (in the blood) and the balance shifts with treatment of one. Another hypothesis is that the immune reconstitution stimulates the replication and evolution of quasispecies. These are educated guesses at present with no data to back them.
The Madrid group analyzed their highly coinfected cohort of patients for the inpact of HCV+ on immune reconstitution. Of their 503 patients who initiated HAART during the study period of 1997 through 2001, 58% had undetectable HIV RNA levels for at least 24 months. Two-third of those patients are HCV-coinfected. Age/sex were similar in the HCV- vs. HCV+. 90% received protease inhibitors in each group. The HCV+ had a slightly lower T-cell count to begin with (median 213 vs. 265) which was not statistically significant. Both groups had good increases in T-cells from baseline at 12 months, (220 for HCV+ vs 194 for HCV-), but the HCV+ group dropped off slightly after 18 months (233 vs. 301 in the HCV-) and at 24 months (272 vs. 325). These numbers, although different are very respectable increases in CD4 counts.
A study of 576 patients initiating HAART in Frankfurt looked at immunologic response (defined as an increase of100 CD4 cells) and factors that are predictive. 120 were HCV+. When controlling for viral load response, age, baseline CD4, baseline viral load, and previous therapy, being HCV+ was independently associated with not achieving an immunologic response.
Editorial comment: These studies don't definitively answer the question, but one thing to take home if you are coinfected, almost all patients' CD4 cells increase with good control of the HIV. If they increase 150 vs. 250, I am not convinced that that is clinically going to make a difference.
1. Abstract 637. Chung, et al.
2. Abstract 638. Moreno, et al.
3. Abstract 639. Sabin, et al.
Does Hepatitis coinfection impact survival?
A study from the MACS cohort asked the question of the impact of HBV in HIV+ and HIV- mortality. Of the 5293 gay men followed from 1/84-3/00, 6% were HBV+ as determined by having a positive HBSurface antigen in the blood. About half of the cohort either were or became HIV+ during the study period. The liver related deaths were 10 times higher in those who were HIV/HBV coinfected vs. those who were HBV monoinfected. The mortality from liver disease has doubled in HIV/HBV coinfected patients in the post-HAART era (like what we've seen with HCV). Overall mortality was not impacted by HBV coinfection.
In contrast to their earlier study/publication a couple of years ago, the group at the Atlanta VA found that now they are seeing a decreased survival in their HCV+ coinfected patients. Observation period is 1992 to 2001. 31% were HCV+. The time from an AIDS diagnosis to death is shorter in co-infected patients vs. HCV- patients even after controlling for CD4 count at the time of diagnosis, antiretroviral use, and having an opportunistic infection. They suspect that this could be due to longer follow up times or due to the impact of HAART.
In the HOPS (HIV outpatient study) cohort, a similar analysis of HCV impact on morbidity, mortality and taking HAART therapy was performed. 852 patients had known HCV status and were included. The observation period was from 1996 - 2001. 38% of patients were HCV+. Co-infected overall had higher mortality rates, rates of AIDS diagnoses, renal and cardiovascular diseases than HCV-, but when controlling for antiviral therapy usage and baseline CD4 at presentation, HCV status did not significantly affect survival. Taking HAART was the biggest factor in survival and HCV+ patients took less HAART.
1. Abstract 656. Thio, et al.
2. Abstract 658. Rimland, et al.
3. Abstract 659. Tedaldi, et al.
Are you vaccinated properly?
The IDSA/USPHS recommends that everyone who is coinfected with HCV get vaccinated against hepatitis A as acute hepatitis A can cause fulminant liver failure in this group. In a cohort of 14,000 outpatients, 1933 had HCV (13%). Only 14% had received the hepatitis A vaccine.
The take home message on the survival impact of HCV/HBV coinfection is that it causes much more liver-related morbiditiy and mortality, and it may cause more mortality overall, but the main problem is that coinfected patients, particularly HCV coinfected patients are more likely to not be on antiretroviral therapy, which we know significantly affects survival. Is this the physicians not prescribing them, the patients not wanting them, or not tolerating them. This is a question for future conferences.