 |
|
|
| |
Predicting outcomes in people with HIV and HCV
|
| |
| |
Report 6 from the 44th ICAAC, October 30-November 2, 2004, Washington, DC
Written for NATAP by Mark Mascolini
Reflecting a sharpened research focus on HIV/HCV coinfection, the 44th ICAAC meeting featured a rich array of studies in this field. Several of them dissected results of recent or ongoing clinical trials to probe for predictors of sustained virologic response, treatment side effects, fibrosis, or quitting antiretroviral therapy. One sobering report chronicled the complications of HIV infection while coinfected people sit on liver transplant waiting lists.
Predicting who won't respond to PegIFN
Analyses of two pegylated interferon(PegIFN)/ribavirin trials in coinfected people found that a poor HCV response after 4 or 12 weeks of treatment usually means a low chance of sustained virologic response after therapy ends. But good early responses don't predict good long-term responses nearly as well.
In people infected with HCV but not HIV, a poor 12-week response to PegIFN/ribavirin predicted a poor long-term response 24 weeks after treatment ended (1). Researchers who studied the same regimen in people with HCV and HIV—in APRICOT trial—wondered if the same prediction held true in coinfected people (2). It did.
APRICOT randomized coinfected people to one of three regimens:
• Peginterferon alfa-2A (40 KD) 180 µg/week plus ribavirin 800 mg/day
• Peginterferon alfa-2a (40 KD) 180 µg/week plus ribavirin placebo
• Interferon alfa-2a 3 MIU three times per week plus ribavirin 800 mg/day
The study found a sustained virologic response rate in 40% taking PegIFN/ribavirin for 48 weeks versus a 12% sustained response in the other groups (P < 0.0001) (3). Looking only at 289 people who took PegIFN/ribavirin, the APRICOT team identified those with a:
• Rapid virologic response: at least a 100-fold drop in HCV RNA by treatment week 4
• Early virologic response: an HCV RNA below 50 IU/mL or at least a 100-fold drop in HCV load at week 12
• Sustained virologic response: an HCV RNA below 50 IU/mL 24 weeks after treatment ended (study week 72)
Then they calculated the negative predictive value and positive predictive value of rapid or early responses in foretelling a sustained virologic response. Negative predictive value meant the probability that a person WITHOUT an on-treatment virologic response WOULD NOT have a sustained virologic response after treatment. Positive predictive value meant the probability that a patient WITH an on-treatment virologic response WOULD have a sustained response after treatment.
In the 289-person study group, 176 (61%) had hard-to-treat HCV genotype 1 and 95 (33%) had HCV genotype 2 or 3. Overall, 116 people (40%) had a sustained virologic response, including 59 (62%) with genotype 2 or 3 and 51 (29%) with genotype 1. The positive predictive value of a rapid virologic response at week 4 measured only 66%—not very helpful. But the negative predictive value of the week 4 response measured 88%, and the negative predictive value of a response at week 12 climbed all the way to 98%. Negative predictive values at weeks 4 and 12 were equally robust with HCV genotype 1 and with genotype 2 or 3:
| Sustained virologic response at week 72, n (%) | Negative predictive value at weeks 4, 12, and 24 (%) | | All study participants | 116 (40) | 88, 98, and 99 | | HCV genotype 1 | 51 (29) | 90, 98, and 100 | | HCV genotype 2 or 3 | 59 (62) | 84, 100, and 100 |
Because of the mediocre positive predictive value of responses at weeks 4 and 12, the APRICOT researchers reasoned that good responses at those points do not reliably predict a sustained virologic response. The negative predictive value rose from 88% with the week 4 response to 98% with the week 12 response. But it improved no further with the week 24 response (see table above, column 3). As a result the APRICOT analysts proposed that the best time to figure negative predictive value is week 12.
Spanish researchers testing PegIFN/ribavirin in coinfected people also sized up the predictive prowess of a poor early response in their cohort, figuring that predicting which people have a low chance of a sustained response would spare them the rigors of useless therapy (4). Their study involved 60 people but they limited the analysis to 48 who had their HCV load measured at week 4. The trial called for 24 weeks of treatment for people with HCV genotype 2 or 3 and 48 weeks of treatment for people with genotype 1. The investigators defined a sustained virologic response as an undetectable load 24 weeks after treatment stopped.
In an intention-to-treat analysis, 13 of the 48 study participants (27%) had a sustained virologic response, including 3 with genotype 1, 8 with genotype 2 or 3, and 2 with genotype 4. Among 35 people who had an undetectable HCV load or at least a 1-log (10-fold) drop at week 4, only 13 achieved a sustained virologic response to yield a feeble positive predictive value of 37%. But among the 13 people who did not have an undetectable load or a 1-log drop at week 4, none had a sustained virologic response. So the negative predictive value hit 100%.
Looking at 25 people who had an undetectable load or a 2-log (100-fold) drop by week 4, the researchers counted 12 who had a sustained response—for a positive predictive value of only 48%. But 22 of 23 people who did not lower their HCV load 100-fold by week 4 also failed to have a sustained response—for a negative predictive value of 96%.
Echoing the APRICOT researchers, the Spanish team proposed that poor early responses are a good harbinger of poor long-term responses. Results of their much smaller analysis suggested that week 4 negative responses may be a useful predictor, whereas the APRICOT group found that the negative predictive power rose considerably when they waited for week 12 responses.
Good early results in another PegIFN trial
Early results of a large Spanish trial of PegIFN/ribavirin in HIV/HCV-coinfected people proved comparable to results in people infected only with HCV, even among those with HCV genotype 1 (5). The promising early findings may reflect the relative good health of the study participants and the high doses of ribavirin used.
PRESCO (Pegasys Plus Ribavirin for HCV Treatment in HIV/HCV Coinfection) is open to coinfected Spanish adults who meet several criteria including:
• CD4 count above 300 cells/µL
• Hemoglobin at or above 12 g/dL
• Leucocytes at or above 3000 cells/µL
• Neutrophils at or above 1500 cells/µL
• Platelets at or above 100,000 cells/µL
In the 198 people assessed so far, the median CD4 count measured 510 cells/µL (interquartile range 402 to 710 cells/µL). The group's median HIV load lay below 50 copies/mL, and 66% had a sub-50 load. All enrollees had a normal bilirubin, albumin, creatinine and prothrombin time. Study participants could not have hepatitis B virus infection, a history of decompensated cirrhosis, or any liver disease not related to chronic HCV infection. They didn't have to be taking antiretrovirals, but they couldn't be taking ddI. Sixty of these 198 people (30%) are women.
The HCV regimen consists of peginterferon alfa-2a (40 KD) at a dose of 180 µg once weekly plus ribavirin in weight-adjusted doses:
• 600 mg/day if <40 kg
• 800 mg/day if 40 to 65 kg
• 1000 mg/day if 65 to 85 kg
• 1200 mg/day if >85 kg
In contrast, everyone in APRICOT who took ribavirin got 800 mg daily. The higher ribavirin doses in PRESCO have been relatively well tolerated. Sixteen of the 198 people evaluated so far (8%) dropped out before 12 weeks of follow-up, 12 of them (6%) because of side effects.
The PRESCO population consists of 104 people (52%) with HCV genotype 1, 2 (1%) with genotype 2, 73 (37%) with genotype 3, and 19 (10%) with genotype 4. After 12 weeks of treatment, response rates among people with genotype 1 only slightly lagged the overall response rate:
|
|
Drop in HCV RNA
|
|
|
|
>1 log, n (%)
|
>2 logs, n (%)
|
Undetectable, n (%)
|
|
Week 4
|
|
|
|
|
Genotype 1 (n = 65)
|
45 (69)
|
30 (46)
|
22 (34)
|
|
Genotype 2 or 3 (n = 48)
|
45 (94)
|
43 (90)
|
41 (85)
|
|
Genotype 4 (n = 14)
|
8 (57)
|
7 (50)
|
4 (29)
|
|
All (n = 127)
|
98 (77)
|
87 (63)
|
67 (53)
|
|
Week 12
|
|
|
|
|
Genotype 1 (n = 94)
|
78 (83)
|
73 (78)
|
56 (60)
|
|
Genotype 2 or 3 (n = 70)
|
68 (97)
|
66 (94)
|
65 (93)
|
|
Genotype 4 (n = 18)
|
16 (89)
|
13 (72)
|
12 (67)
|
|
All (n = 182)
|
162 (89)
|
152 (84)
|
133 (73)
|
The PRESCO team points out that these early response rates approximate those seen with PegIFN/ribavirin in people with only HCV infection (6). Sustained virologic response rates for the whole PRESCO cohort won't be reported for another year or so. Seeing how many people reach a sustained response will be interesting—not only to compare results with earlier trials, but also because other work clearly correlates sustained responses with long-term health. Vincent Soriano made that point in a study of all HIV/HCV-coinfected people treated for HCV at Madrid's Carlos III Hospital (7).
These 351 people had received standard interferon monotherapy, standard interferon plus ribavirin, or PegIFN plus ribavirin. Seventy-seven (22%) managed a sustained virologic response, with the highest sustained response rate, 30%, among the 126 who got PegIFN. After 4466 patient-months of follow-up, no one who notched a sustained response had an HCV RNA rebound. The sustained responders had no HCV-linked jumps in liver enzymes, no hepatocellular carcinoma, and no decompensated cirrhosis.
In contrast, after 15,344 patient-months of follow-up on the 274 people who did not achieve a sustained response, 90% had persistent liver enzyme elevations. Complications of cirrhosis developed in 11 of these 274 (4%), including encephalopathy in 5, jaundice in 6, ascites in 4, gastrointestinal bleeding in 3, spontaneous peritonitis in 1, and hepatorenal syndrome in 1. Two of these people died with end-stage liver disease.
Early ribavirin levels to predict anemia
Anemia, the primary risk of ribavirin therapy, may be predicted by early ribavirin levels, according to results of another Carlos III Hospital study presented by Ana Rendón (8). The ribavirin concentrations, she found, did not predict virologic response to anti-HCV therapy.
This analysis involved 49 HIV/HCV-coinfected people taking peginterferon alfa-2a plus ribavirin. People weighing less than 75 kg got 1000 mg of ribavirin daily and those weighing more got 1200 mg daily. Some people taking drugs that might interact with ribavirin took 800 mg daily.
Median ribavirin levels did not change significantly from treatment week 4 (2.19 µg/mL) to week 12 (2.13 µg/mL). Concentrations of the drug varied greatly from person to person—from 0.07 to 4.23 µg/mL at week 4, and from 0.7 to 3.78 µg/mL at week 12. Those varying levels did not correlate with weight, daily dose, or dose per kilogram of body weight.
By week 4 hemoglobin levels had dropped by a median 2.3 mg/dL. Higher ribavirin concentrations at that point correlated with lower hemoglobin—the higher the ribavirin level, the lower the hemoglobin (odds ratio 0.91, 95% confidence interval [CI] 0.6 to 1.2, P < 0.001). Ribavirin concentrations did not predict virologic response at week 4 for any HCV genotype. Nor did ribavirin levels differ in people taking or not taking methadone.
Liver toxicity risk with HAART and HCV
An ongoing debate in managing HIV/HCV coinfection involves when to start antiretroviral therapy (HAART). Starting HAART early has the obvious advantage of reining in runaway HIV replication, but the liver toxicity that comes with many antiretrovirals makes some clinicians wary of starting HAART until they first control HCV.
One study of 105 HIV/HCV-coinfected people found that treating HCV first lowered the risk of later liver toxicity-related drug discontinuations (9). Because of findings like these, the British HIV Association recommends treating HCV infection before starting HAART in people with CD4 counts above 200 cells/µL (10). But a chart review presented at ICAAC found that liver complications only rarely forced coinfected people to stop or switch antiretroviral regimens (11).
Ottawa Hospital clinicians leafed through the records of everyone with HCV whom they treated with antiretrovirals from January 1996 through December 2003. Those rolls listed 186 people, 150 (81%) of them men, 139 (75%) with an injecting drug history, and 56 (30%) with a history of alcohol abuse. Most of them—144 (77%)—took a protease inhibitor (PI) regimen, and 76 (41%) took two PIs. Thirty-six (19%) took a nonnucleoside-based regimen.
A large majority of people either changed (57%) or stopped (19%) their HAART, and many did so because of antiretroviral side effects. But only 6 (3%) switched or stopped because of clinically apparent liver toxicity. All 6 were taking a PI, 4 were taking d4T, and 1 was drinking more than 50 g of alcohol daily. None of the 6 was taking low-dose ritonavir or a nonnucleoside.
Only a handful of people—7%—had an alanine aminotransferase (liver enzyme) elevation more than 5.1 times the upper limit of normal (40 IU/mL) during the first 6 months of HAART. The median ALT rose from 46 IU/mL before HAART to 67 IU/mL after 6 months, then waned 56 IU/mL after 12 months.
Statistical analysis linked three factors to reduced duration of the first HAART regimen:
• Injection drug use history (P = 0.004)
• Female gender (P = 0.007)
• PI-based therapy (P = 0.077)
Taking a nonnuke regimen improved the chance of longer treatment with the first regimen. Further analysis of 300 people who took a second HAART combo tied three variables to shorter treatment duration:
• Injection drug use history (P = 0.002)
• Pretreatment CD4 count below 200 cells/µL (P = 0.03)
• Pretreatment viral load above 100,000 copies/mL (P < 0.001)
A study of 107 coinfected people at Madrid's Ramón y Cajal Hospital also failed to tie antiretroviral therapy to liver toxicity (12). Everyone in this study had a liver biopsy—a sign that their clinicians already suspected liver damage. The biopsies showed that 35% had Knodel fibrosis score 1, 20% had fibrosis score 2, 32% had score 3, and 13% had score 4. Worse fibrosis correlated with a longer time since HCV diagnosis (r = 0.28, P = 0.003) and a higher ALT (r = 0.27, P = 0.005).
During 2 years of follow-up after biopsy, 25% of these people had HAART-related liver toxicity, defined as an ALT or aspartate aminotransferase (AST) elevation above 5 to 10 times the upper limit of normal in people who had normal liver enzymes at baseline, or an ALT or AST more than 3.5 times the upper limit of normal in people who already had high liver enzymes when the study started.
In a multivariate analysis two variables independently raised the risk of liver toxicity:
• Fibrosis score 3-4 versus fibrosis score 1-2: relative risk 2.75, 95% CI 1.08 to 6.97, P = 0.03
• Each higher ALT unit at baseline: relative risk 1.006, 95% CI 1 to 1.01, P = 0.04
The risk of fibrosis score 3-4 was significantly higher among people taking nevirapine than among those taking efavirenz (11.7 versus 8.6 cases per 100 patient-years, P < 0.05). And nevirapine trough concentrations were 40% higher in those with fibrosis score 4 than in those with less severe fibrosis, but that difference lacked statistical significance.
What happens while waiting for a new liver?
For coinfected people with the most advanced liver damage, a transplant is the last option. Studies of the first HIV-infected people to get new livers suggest they will survive as long as other transplant patients (13). But transplant waiting lists are lengthy for everyone, and waiting too long brings dire consequences. Everyone knows this, but quantifying these consequences could further the case for offering transplants to people with HIV. Vincent Soriano's group at Carlos III Hospital in Madrid did so (14).
Among 2256 people monitored regularly for liver disease at this hospital, 20 met criteria for getting on an orthotopic liver transplant waiting list: Child B or C score liver fibrosis with previous decompensated cirrhosis (ascites, encephalopathy, gastrointestinal bleeding, hepatorenal failure, or hepatocellular carcinoma), a CD4 count above 100 cells/µL, and a viral load below 50 copies/mL or—if not taking antiretrovirals—the opportunity to start successful therapy after transplantation.
All of these 16 men and 4 women had HCV infection, and 15 had taken antiretrovirals for a median 9 years. The median CD4 count at the start of follow-up measured 252 cells/µL. Fourteen people had Child score B and 6 had score C.
Four of these 20 people died from liver complications in the first year of follow-up. Hospital admissions for liver failure numbered 22 and for other reasons 9. During follow-up gastrointestinal bleeding developed in 6 people, ascites in 4, encephalopathy in 3, hepatocellular carcinoma in 1, and renal failure in 1.
The Carlos III team believes their findings "suggest that assessment of cirrhotic HIV-positive patients and their inclusion in waiting lists for orthotopic liver transplantation should be considered much earlier than in HIV-negative patients."
Mark Mascolini writes about HIV infection (mailmark@ptd.net).
References
1. Fried MW, Shiffman ML, Reddy KR, et al. Peginterferon alfa-2a plus ribavirin for chronic hepatitis C virus infection. N Engl J Med 2002;347:975-982.
2. Rodriguez-Torres M, Dieterich DT, Lissen E, et al. Predictability of sustained virologic response in HIV-HCV co-infected patients treated with peginterferon alfa-2a (40 KD) + ribavirin in the APRICOT trial. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1751.
3. Torriani FJ, Rodriguez-Torres M, Rockstroh JK, et al. Peginterferon alfa-2a plus ribavirin for chronic hepatitis C virus infection in HIV-infected patients. N Engl J Med 2004;351:438-450.
4. Shaw E, Casanova A, Santin M, et al. Assessment of hepatitis C virus RNA levels at week 4 as an early predictor of sustained virological response in HIV-infected patients treated for chronic hepatitis C. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1752.
5. Camino N, Núñez M, Barreiro P, et al. Early HCV-RNA clearance in HIV-HCV co-infected patients who begin treatment with pegylated interferon plus ribavirin. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1753.
6. Hadziyannis SJ, Sette HJ, Morgan TR, et al. Peginterferon-alfa2a and ribavirin combination therapy in chronic hepatitis C: a randomized study of treatment duration and ribavirin use. Ann Intern Med 2004;140:346-355.
7. Soriano V, Maida I, Núñez M, et al. Long-term follow-up of human immunodeficiency virus-infected patients with chronic hepatitis C virus infection treated with interferon-based therapy. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1757.
8. Rendón A, Núñez M, Jiménez-Nácher, Soriano V. Early measurement of ribavirin plasma concentrations may predict anemia but does not predict virological response in HIV-HCV co-infected patients on anti-HCV therapy. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1754.
9. Uberti-Foppa C, De Bona A, Morsica G, et al. Pretreatment of chronic active hepatitis C in patients coinfected with HIV and hepatitis C virus reduces the hepatotoxicity associated with subsequent antiretroviral therapy. JAIDS 2003;33:146-152.
10. British HIV Association. Coinfection Guideline Committee on behalf of the British HIV Association. British HIV Association (BHIVA) guidelines for treatment and management of HIV and hepatitis C coinfection. October 2003. http://www.bhiva.org/guidelines/2003/HCV/index.html.
11. Cooper CL, Breau C, Mackle D, et al. Predictors of HAART discontinuation or changes in HIV-HCV co-infection. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1759.
12. Aranzabal L, Moya J, Casado JL, et al. HCV-related liver fibrosis as risk factor for HAART-associated hepatotoxicity in HIV-infected patients. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1760.
13. Ragni MV, Belle SH, Im K, et al. Survival of human immunodeficiency virus-infected liver transplant recipients. J Infect Dis 2003;188:1412-1420.
14. Sanchez-Conde M, Gil P, Romero M, et al. Complications in HIV+ patients with end stage liver disease in a waiting list for liver transplantations. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-1758.
|
| |
|
|
|
|
|
