July 25, 2007
The SMART (the Strategic Management of Antiretroviral Therapy) study may have been one of the most important studies of 2006.1 Initiated by the U.S. National Institutes of Health (NIH)-funded Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA), this large clinical trial opened in early 2002 in 33 countries. HIV-infected patients were randomized to a virologic suppression strategy (VS, n=2,752); which maintained viral load as low as possible, versus a drug conservation strategy (DC, n=2,720); which stopped HIV treatment, or in those who were drug naive, offered HIV treatment only when the CD4 count was below 250/mm3, and then provided episodic HIV treatment thereafter to increase the CD4 count to more than 350/mm3.
The trial was stopped early by the study's Data and Safety Monitoring Board when significantly more patients developed an AIDS-defining illness or death in the drug conservation arm compared to the virologic suppression arm.
At IAS 2007 we saw results from two SMART substudies. The first one I'll discuss was from Tedaldi and colleagues2 who looked at the rates of opportunistic infections and death in those patients who were co-infected with hepatitis B virus (HBV) or hepatitis C virus (HCV) in the SMART study.
The rationale for the current analysis is that in general HBV and HCV co-infection have been associated with increased medical complications in people with HIV infection. The current analysis looked at whether people with viral hepatitis were more significantly impacted if they stopped there HIV medications.
Patients were defined as having HBV infection if they had a positive HBV surface antigen test for more than six months, and having HCV infection if they were HCV- antibody positive. Rates of events were defined as per 100 person-years.
Of those enrolled in SMART, 922 (16.8%) were HBV and/or HCV infected (110 HBV, 2%) and 798 HCV, 14.6%) and HBV/HCV infected (14, 0.25%). Time of follow-up was 1,467 person years for those with HBV or HCV infection and 5,901 person years without those infections.
The median baseline CD4 count was over 550/mm3 for patients with HBV/HCV infection and the 65% had an undetectable HIV viral load, 27% had prior AIDS, 3% were antiretroviral naive, and on average this group had six years of HIV treatment experience.
There was no significant difference between the virologic suppression or drug conservation groups with respect to these variables. The overall rate of opportunistic infection and death was significantly higher in patients who had HBV or HCV compared to those who did not, but the rate of opportunistic infection (not including death) was the same. Deaths that were not associated with opportunistic infections were significantly greater (3.8 times higher) in the HCV/HBV- infected group. Deaths in this group were mostly caused by substance abuse, non-AIDS cancers and unknown causes.
There did not appear to be a significant difference in opportunistic infections or death when analyzed by virologic suppression or drug conservation, regardless of whether patients had HBV or HCV or not. There were too few HBV-infected patients to analyze any difference in risks between those with HBV or HCV infection.
The conclusions of this study found that the risk of opportunistic infections was comparable regardless of hepatitis status. Although this group constituted 17% of the total study group, they accounted for almost half of the non-AIDS associated deaths. The risk of developing an opportunistic infection or non-AIDS death was the same in both arms of the study.
In other words, interrupting treatment for a period of time did not increase risk compared to staying on treatment. However, given the overall results of the SMART study, it is not recommended to stop treatment once a certain CD4 count threshold is reached and allow the CD4 count to go back down again before restarting treatment. This is particularly true for those with viral hepatitis where the risk of medical complications in general is much higher than those without hepatitis. In the second oral presentation from the SMART study group at IAS, Dore and colleagues3 looked at the rates of restarting HIV treatment in those patients who were co-infected with HBV or HCV in the SMART study.
They specifically looked at the drug conservation strategy arm; those who stopped or in those who were drug naive offered HIV treatment only when the CD4 count was below 250/mm3, and then had episodic HIV treatment thereafter to increase the CD4 count to >350/mm3. Because HIV infection can result in increased HBV or HCV viral loads compared to those with viral hepatitis without HIV infection, and because co-infection with HBV or HCV can result in increased liver-related problems and death in those with HIV infection, this group wanted to know whether having viral hepatitis caused the restarting of HIV medications sooner or not.
There was no exclusion in this study for those with viral hepatitis, although it recommended that those patients who required continued antiviral therapy for hepatitis should not be enrolled.
Patients were defined as having HBV infection if they had a positive HBV surface antigen test for more than six months, and having HCV infection if they were HCV antibody positive.
Various factors such as age, gender, prior AIDS diagnosis, baseline and nadir CD4 count and viral load were examined to determine whether they affected the rate of restarting HIV medications. Data was available from 2,669 patients in the drug conservation arm enrolled in SMART, (65 HBV and 402 HCV infected).
Several HIV medications are used to treat HBV, at baseline: 45% were receiving lamivudine (3TC, Epivir) or emtricitabine (FTC, Emtriva), 7% tenofovir (TDF, Viread), and 17% lamivudine or emtricitabine and tenofovir. About 70% had an undetectable HIV viral load and the mean CD4 count was over 550/mm3.
Median time to restarting HIV medications was shorter for those with HBV infection (7.5 months) compared to those with HCV infection (17 months) and no HBV/HCV infection (18.2 months). The percentage of patients who had to reinitiate HIV medications was significantly higher in the HBV-infected patients. In a multivariate analysis, factors that were associated with the need to restart HIV medications included being infected with HBV, having a lower baseline and nadir CD4 count, higher viral load, being older, and having a prior AIDS diagnosis.
HIV- infected patients with HBV infection who stopped treatment were found to lose CD4 cells more quickly than those with HCV or without viral hepatitis. This accounted for the need to restart treatment sooner than the other groups. The exact explanation for why this happened is not clear. Data was not presented on whether those with HBV infection saw a rebound in HBV viral load or worsening of their hepatitis, which could have happened when the HIV drugs that are active against HBV are stopped. This once again stresses the need to not stop HIV treatment, particularly in patients coinfected viral hepatitis.