Relative Benefit of ddI and AZT Depends on Duration of Patients' Previous AZT Use
For certain HIV-infected individuals, the relative effectiveness of the drug didanosine (ddI) and zidovudine (AZT) in slowing the progression of HIV disease depends on the duration of their previous AZT therapy, according to a recently completed study sponsored by the National Institute of Allergy and Infectious Diseases (NIAID) and Bristol-Myers Squibb Company.
In this study, AZT appeared to be the more effective drug among HIV-infected patients with advanced disease who had not taken AZT previously, while ddI appeared to be the more effective drug among those with previous AZT use of at least eight to 16 weeks," say Raphael Dolin, M.D., of the University of Rochester School of Medicine and Dentistry and principal investigator of the study.
The study results are based on data from 617 HIV-infected patients, all of whom had previously never taken AZT, or who had taken and tolerated AZT for 16 or fewer weeks. The study assessed whether ddI and AZT were similarly effective in these patients in delaying new AIDS-defining conditions or death.
Physicians who care for HIV-infected patients with advanced disease should discuss these new data with their patients to plan the best treatment for their individual needs", says Anthony S. Fauci, M.D., NIAID director. "Further discussions in the scientific community will help us to understand the implications of these results and their impact on patient care."
The study was carried out at 33 sites of the AIDS Clinical Trials Group (ACTG), four sites supported by Bristol-Myers Squibb Company and one regional center funded by the National Hemophilia Foundation. The ACTG is a nationwide network of AIDS clinical research centers funded by NIAID. Burroughs Wellcome Company provided AZT for the study, and Bristol-Myers Squibb Company provided ddI for the study.
Investigators enrolled patients with AIDS or advanced AIDS Related Complex (ARC) with fewer than 300 CD4+ T cell counts per cubic millimeter (mm) of blood and patients with no AIDS or ARC-related symptoms and 200 or fewer CD4+ T cells/mm. CD4+ T cells are the crucial immune cells depleted during HIV infection. ARC is a symptomatic stage of HIV infection before AIDS when a patient has a symptoms such as a yeast infections of the mouth, unexplained fever of diarrhea, recurrent herpes outbreaks or loss of more than 10 percent of his or her body weight.
Participants were randomly assigned to receive daily either 600 milligrams (mg) AZT, 500 mg ddI or 750 mg ddI. Neither the investigators nor the patients knew which treatment a patient received. Sachet (powdered buffer) formulation was used for ddI. A 500 mg ddI dose in the sachet form is equivalent to a 400 mg dose in tablet form because of differences in absorption. Patients were followed for a median of 85 weeks.
Among 380 patients in the study who had no prior AZT use, 18 percent of those taking AZT developed a new, previously undiagnosed AIDS-defining condition or died within one year. In comparison, 31 percent of the AZT-naive patients on the 750 mg ddI dose and 29 percent of those on the 500 mg ddI dose developed new AIDS-defining conditions or died within one year.
Among the 118 patients with eight to 16 previous weeks of AZT therapy, the study investigators found ddI to be the more effective drug. For those patients assigned to AZT, 33 percent developed a new AIDS-defining condition or died within one year, while only 11 percent of patients taking 500 mg ddI and 17 percent taking 750 mg ddI did so.
The treatment comparisons between AZT and the two doses of ddI were inconclusive among the 119 patients with eight or fewer weeks of prior AZT therapy. Overall, no significant differences between the effectiveness of the two doses of ddI were noted.
In the study, patients assigned to AZT were more likely than those on ddI to suffer lowered blood cell counts, including granulocytopenia, a decrease in a type of white blood cell. Patients assigned to AZT had a low risk, similar to that seen among ddI patients, of developing severe anemia, a decrease in red blood cells.
Patients on 750 mg ddI were more likely to develop pancreatitis within one year than those on the 500 mg ddI dose or those on AZT. No significant differences in the rate of development of peripheral neuropathy, a painful nerve condition, were observed between AZT and ddI recipients.
"These data suggest that 500 mg ddI is better tolerated and support the continued use of ddI 500 mg per day (400 mg per day in tablet form)," say Margaret Fischl, M.D., of the University of Miami School of Medicine and co-investigator of the study.
The study, know as ACTG 116A, began enrollment in October 1989 and follow-up ended in May 1992. Physicians and others caring for HIV-infected patients can call 1-800-TRAILS-A to request further information on the study findings, including a question and answer document.
NIAID, a component of the National Institutes of Health, supports research on allergy, immunology and infectious diseases. NIH is an agency of the U.S. Public Health Service Department of Health and Human Services.
National Institute of Allergy and Infectious Diseases
NATIONAL INSTITUTES OF HEALTH
ACTG 116A QUESTIONS AND ANSWERS
1. WHAT WAS STUDY ACTG 116A?
ACTG 116A, a randomized, double-blind clinical trial, compared the efficacy and toxicity of the drug didanosine (ddI) at two different doses (750 milligrams per day and 500 milligrams per day, in sachet formulation) to standard treatment with zidovudine (ZDV; AZT) in patients with AIDS, AIDS-Related Complex (ARC), or asymptomatic HIV infection who had received no prior zidovudine therapy or who had taken zidovudine therapy for less than or equal to 16 weeks and were tolerating it at the time of study entry.
2. WHAT WAS THE PURPOSE OF STUDY ACTG 116A?
The primary objective was to assess whether didanosine had similar clinical efficacy to zidovudine in patients with little or no prior zidovudine experience. Another objective was to determine if the duration of prior zidovudine therapy affected the efficacy of didanosine compared to zidovudine. The study was also designed to compare the toxicities (side effects) of the two doses of didanosine (ddI) to those of zidovudine in terms of frequency and severity.
3. WHERE AND WHEN WAS THIS STUDY CONDUCTED?
Study enrollment opened in October 1989. The study was conducted at 33 sites of the AIDS Clinical Trials Group (ACTG), 1 regional center funded by the National Hemophilia Foundation, and 4 non-ACTG sites supported by Bristol-Myers Squibb Company, for a total of 38 sites. Accrual to the lower dose didanosine treatment arm (500 milligrams [mg] per day, sachet formulation) was ended in December 1990 in order to accelerate the completion of the study. Enrollment to the remaining two treatment arms ended in early April 1991, and follow-up of participants was stopped on May 1992.
4. WHO SPONSORED THIS STUDY?
ACTG 116A was sponsored as a collaborative effort between the AIDS Clinical Trials Group of the National Institute of Allergy and Infectious Diseases and by Bristol-Myers Squibb, the company that manufactures didanosine (ddI). Burroughs Wellcome Company provided zidovudine for the study.
5. WHY WAS THIS STUDY CALLED "ACTG 116A"?
The study was renamed "ACTG 116A" after an administrative amendment to protocols ACTG 116 and ACTG 117 reclassified patients in the two studies based on the duration of prior zidovudine therapy. Patients who had no prior zidovudine or less than or equal to 16 weeks of zidovudine were included in ACTG 116A.
6. WHO PARTICIPATED IN THE STUDY?
The study was open to patients who had AIDS or AIDS- Related Complex (ARC) with CD4 cell counts of 300 per milliliter or less, or who had HIV infection without symptoms and CD4 cell counts of 200 per milliliter or less. Patients either had not taken zidovudine previously or had taken zidovudine for less than or equal 16 weeks. Patients who had taken zidovudine previously had to be tolerating zidovudine at the time of enrollment.
Approximately one-fourth of the patients had AIDS at study entry, two-thirds had ARC, and the rest had HIV infection without symptoms and CD4 cell counts of less than 200 per milliliter. Thirteen percent of the patients were black and 12 percent were Hispanic. Women comprised 8 percent of the study participants. The median age was 35 years.
7. HOW WAS AIDS-RELATED COMPLEX (ARC) DEFINED FOR ELIGIBILITY FOR THIS STUDY?
ARC was defined on the basis of having one or more of the following clinical signs in addition to demonstrated antibodies to HIV: oral candidiasis (thrush), oral hairy leukoplakia, unexplained fever for a specific time period, unexplained chronic diarrhea for a specific time period, recurrent herpes zoster or multi-dermatomal herpes zoster, or weight loss of more than 10 percent of body weight.
8. HOW WAS PROGRESSION OF HIV DISEASE EVALUATED?
Progression of HIV disease was evaluated in terms of time to a new AIDS-defining event or death. Kaposi's sarcoma and early findings of AIDS dementia were not considered to be primary endpoints.
9. WAS THE STUDY DESIGNED TO SHOW WHICH MEDICATION WAS MORE EFFECTIVE?
No, ACTG 116A was designed to assess whether didanosine had clinical efficacy similar to or not greatly lower than that of zidovudine in patients with advanced HIV disease. Thus it was designed as a trial to show equivalence between the two therapies.
10. WHAT IS AN EQUIVALENCE TRIAL?
An equivalence trial is a randomized clinical trial that compares a new therapy (in this case, didanosine) with a standard therapy (in this case, zidovudine). The intent of an equivalence trial is to demonstrate if the efficacy of a new therapy is similar to or not much lower than that of a standard therapy. An equivalence trial is done when the new therapy has potential advantages, such as a lower rate or different type of toxicity than the standard therapy, that would make it a useful therapeutic option even if the new therapy might be somewhat less efficacious than the standard therapy.
11. WHY WAS ACTG 116A DESIGNED AS AN EQUIVALENCE TRIAL?
At the time the study was designed, zidovudine was the only approved antiviral therapy for HIV infection. Other antiviral therapies were urgently needed and didanosine was thought to be an effective agent. If it could be demonstrated that didanosine had efficacy similar to or not much lower than zidovudine, it would be useful as a new therapy for HIV infection.
12. WHY WAS ACCRUAL TO THE LOWER DOSE DIDANOSINE ARM (500 MG PER DAY) STOPPED IN LATE 1990?
Accrual to the lower dose didanosine arm (500 mg per day) was closed in December 1990 in order to accelerate the completion of accrual to the study. All patients who had already entered the study and who had been randomized to receive didanosine 500 mg per day continued to be followed. This decision was made without prior knowledge of the study results and only so as to complete the study more rapidly.
13. WAS THIS STUDY STOPPED EARLY BY A DATA AND SAFETY MONITORING BOARD (DSMB)?
No, ACTG 116A was not stopped before its planned completion. The last review of study ACTG 116A by the NIAID Data and Safety Monitoring Board (DSMB) occurred in February 1992, at which time the DSMB recommended completion of the study as originally planned. This means there were no significant safety problems to prevent the completion of the study.
14. DID THE RESULTS SHOW THAT DIDANOSINE AND ZIDOVUDINE HAVE SIMILAR EFFICACY?
No. The most important finding of ACTG 116A was that the efficacy of didanosine as compared to zidovudine was affected by the duration of prior zidovudine therapy. Zidovudine was more efficacious than didanosine among patients who had not taken zidovudine previously. In contrast, didanosine was more effective than zidovudine among those who were zidovudine-experienced, particularly in those patients who had more than 8 weeks of zidovudine therapy. It should be noted that patients in this study may not have been representative of all patients who have advanced HIV disease and who have had no or limited (less than 16 weeks) treatment with zidovudine.
15. WHAT WERE THE RESULTS IN PATIENTS WHO HAD NEVER TAKEN ZIDOVUDINE BEFORE?
The study showed that zidovudine was more efficacious than didanosine in delaying the occurrence of new AIDS-defining events or death among patients who had never taken zidovudine before.
16. WHAT WERE THE RESULTS IN PATIENTS WHO HAD PRIOR ZIDOVUDINE THERAPY?
The study showed that didanosine was more efficacious than zidovudine in delaying the occurrence of new AIDS-defining events or death in patients who had between 8 and 16 weeks of prior zidovudine therapy.
The results for patients who had some but less than 8 weeks of prior zidovudine therapy were inconclusive.
17. WAS A SURVIVAL DIFFERENCE SEEN BETWEEN PATIENTS ASSIGNED TO DIDANOSINE AND THOSE ASSIGNED TO ZIDOVUDINE?
The survival results were consistent with the results on clinical progression.
Among patients with no prior zidovudine therapy, those assigned to zidovudine tended to survive longer than those assigned to didanosine.
Among zidovudine-experienced patients (especially those with 8-16 weeks of prior zidovudine), patients assigned to didanosine tended to survive longer than those assigned to zidovudine.
These results should be interpreted with caution because some of the patients were followed on the study for a long period of time after protocol therapy was discontinued.
18. WHAT WERE THE MOST FREQUENT TOXICITIES OBSERVED WITH DIDANOSINE AND ZIDOVUDINE?
The major toxicities associated with didanosine were pancreatitis and elevated serum amylase levels (hyperamylasemia). The one-year rates of pancreatitis were 9 percent, 7 percent, and 4 percent for didanosine 750 mg per day, didanosine 500 mg per day and zidovudine respectively. The difference between didanosine 500 mg per day and zidovudine was not significant. The fact that pancreatitis and elevated serum amylase levels were also seen in some patients assigned to zidovudine suggest that these conditions may sometimes be a consequence of advanced HIV infection.
No significant differences in peripheral neuropathy were seen between patients taking didanosine as compared to zidovudine. However, more patients on didanosine 750 mg per day than on zidovudine experienced peripheral neuropathy; this difference was not significant.
Hematological toxicity (lowered blood cell counts) was more frequent in patients assigned to zidovudine. Specifically, patients on zidovudine were more likely to develop granulocytopenia, a decrease in a type of white blood cell. No differences in severe anemia (decreased red blood cells) were seen between zidovudine and didanosine.
Other toxicities were infrequent and were not significantly different between the study drugs.
19. WHAT HAPPENED TO THE CD4 CELL COUNTS OF STUDY PARTICIPANTS?
Among patients with no prior zidovudine experience, those assigned to zidovudine had significantly greater increases in CD4 cell count by week 8 when compared to either didanosine group. By week 48, however, the zidovudine group experienced the greatest decline in CD4 cell counts. Because of the appreciable drop-out rate by week 48, these long-term trends should be cautiously interpreted.
Among zidovudine-experienced patients, no significant differences in CD4 cell count increases or declines between the treatment groups were observed.
Public Health Service U.S. Department of Health and Human Services