Correspondence
A Controlled Trial of Double versus Triple Therapy for HIV
N Engl J Med 1998; 338:196-197January 15, 1998
- Article
To the Editor:
With respect to the report by Hammer et al. (Sept. 11 issue),1 we are troubled by the randomization of patients with human immunodeficiency virus (HIV) infection and CD4 cell counts of 50 or fewer per cubic millimeter to a group without a protease inhibitor. Data indicating that the addition of ritonavir to the regimens of patients who have received antiretroviral therapy reduces AIDS-defining events and mortality had been reported at two international meetings2,3 and reviewed by the Food and Drug Administration's Antiviral Drugs Advisory Committee before closure of the randomization period and the first review. Was demonstrating that indinavir has the same effect worth the expected additional mortality in this group? If so, might this not have been more appropriately studied by the pharmaceutical industry than by the National Institutes of Health? Indinavir could have been compared with ritonavir.
Second, despite the prospective stratification of groups according to the CD4 cell count, the study was terminated in both groups, though there was no statistically significant difference in those with CD4 cell counts of 51 to 200 cells per cubic millimeter. The clinical benefit of aggressive antiviral therapy in this subgroup is unknown. Although one might assume that with longer follow-up, a statistically significant difference would have been seen in this subgroup, the overlapping confidence interval in the analysis of viral load in those studied at 40 weeks makes us uncomfortable with this assumption. It seems unwarranted to state in the abstract that “treatment with indinavir, zidovudine, and lamivudine as compared with zidovudine and lamivudine alone significantly slows the progression of HIV-1 disease in patients with 200 CD4 cells or fewer per cubic millimeter and prior exposure to zidovudine” and that “the effects of treatment were similar in both CD4 cell strata.” A more precise statement of the conclusions might be that for people with CD4 cell counts of 50 or fewer per cubic millimeter who are taking zidovudine, the addition of indinavir and lamivudine reduces mortality and slows progression of HIV disease as compared with the addition of lamivudine alone.
3 ReferencesMarta Feldmesser, M.D.
Albert Einstein College of Medicine, Bronx, NY 10461John Falkenberg, R.N., M.B.A.
360 E. 65th St., New York, NY 100211
Hammer SM ,Squires KE ,Hughes MD , et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. N Engl J Med 1997;337:725-733
Full Text | Web of Science | Medline2
Cameron B, Heath-Chiozzi M, Kravcik S, et al. Prolongation of life and prevention of AIDS in advanced HIV immunodeficiency with ritonavir. In: Program and abstracts of the 3rd Conference on Retroviruses and Opportunistic Infections, Washington, D.C., January 28–February 2, 1996:162. abstract.
3
Cameron DW, Heath-Chiozzi M, Kravcik S, et al. Prolongation of life and prevention of AIDS complications in advanced HIV immunodeficiency with ritonavir: update. In: Volume 1 of the Abstracts of the XI International Conference on AIDS, Vancouver, B.C., July 7–12, 1996:24. abstract.
To the Editor:
Hammer et al. indicate that substitution of stavudine for zidovudine was permitted in the event of drug-associated toxicity at any point after randomization or if clinical progression of HIV disease occurred, but they do not give the number of patients in whom this substitution occurred or the duration of treatment with zidovudine before the switch. The mean duration of zidovudine therapy before enrollment was 21 months (interquartile range, 10 to 42). Therefore, many patients may have had resistance to zidovudine at the time of enrollment.1 These patients received only two useful drugs if they were in the zidovudine, lamivudine, and indinavir group and only one useful drug if they were in the zidovudine, lamivudine, and placebo group. However, those who were switched from zidovudine to stavudine would have received real triple therapy if they were in the former group and double therapy if they were in the latter group. Were those patients equally distributed between the two groups in the study? If not, this is a possible source of bias.
1 ReferencesJosé Alberto Arranz Caso, Ph.D.
Hospital Universitario Príncipe de Asturias, 28805 Alcalá de Henares, Spain1
Coffin JM . HIV population dynamics in vivo: implications for genetic variation, pathogenesis, and therapy. Science 1995;267:483-489
CrossRef | Web of Science | Medline
To the Editor:
Hammer et al. conclude that “treatment with indinavir, zidovudine, and lamivudine as compared with zidovudine and lamivudine alone significantly slows the progression of HIV-1 disease in patients with 200 CD4 cells or fewer per cubic millimeter and prior exposure to zidovudine.” However, the analysis stratified according to CD4 counts shows that among patients with CD4 counts of 51 to 200 cells per cubic millimeter who were treated with indinavir and both nucleoside analogues there was a trend toward fewer AIDS-defining events or deaths than among those treated with the two nucleoside analogues alone (P = 0.08). There were no differences in mortality between these two treatment groups. Hence, a better conclusion would be that treatment with indinavir, zidovudine, and lamivudine as compared with zidovudine and lamivudine alone slows the progression of HIV-1 disease in patients with 50 CD4 cells or fewer per cubic millimeter and prior exposure to zidovudine.
It was also surprising that neutropenia was significantly more frequent among patients treated with zidovudine and lamivudine (15 percent) than among those treated with zidovudine, lamivudine, and indinavir (5 percent) (P<0.001). How do the authors explain that finding?
Marcial Delgado Fernández, M.D.
José Luis Zambrana García, M.D., Ph.D.
Felipe Díez García, M.D., Ph.D.
Hospital de Poniente, 04700 El Ejido, Almería, SpainAuthor/Editor Response
The authors reply:
To the Editor: Feldmesser and Falkenberg do not appear to appreciate the difference in design between the AIDS Clinical Trials Group 320 study and the Abbott-sponsored M94-247 trial.1 In the latter study, ritonavir or placebo was added to the regimens of patients who were receiving stable nucleoside-analogue therapy or no therapy. In our study, all patients received therapy to which they had not previously been exposed: lamivudine or indinavir plus lamivudine. Although one would not currently advocate the simple addition of lamivudine to a regimen of zidovudine, this approach clearly represented active treatment in the control group, as evidenced by the rate of disease progression, the positive responses in terms of CD4 cell counts and plasma HIV-1 RNA concentrations, and the independent confirmation of the benefit from lamivudine.2
With respect to the two strata of CD4 cell counts, the primary analysis and the prespecified early-stopping guideline were planned for the overall study population for which the difference in disease progression between the two groups was clearly significant. That this difference was driven by a higher event rate in the lower CD4 cell stratum is to be expected. The hazard ratios for the time to progression to AIDS or death for the overall population (0.50) and the two CD4 cell strata (0.49 and 0.51) were very similar. This finding, along with the similar CD4 cell and plasma HIV-1 RNA responses in the two strata, provides evidence of a similar treatment effect across the entire study population. However, we were careful to state that the possibility of differential effects could not be ruled out, since the study did not have the power to address this question.
In response to Arranz Caso: we did provide the numbers of patients for whom stavudine was substituted for zidovudine: 109 of 1156 patients (3 at randomization and 106 after randomization). Sixty-nine of the post-randomization substitutions occurred in the dual-nucleoside group and 37 in the indinavir-containing group, at a respective median of 4.3 and 3.6 months of follow-up. Only three of these patients (all in the dual-nucleoside group) had disease progression. Thus, it is unlikely that the use of stavudine had a major impact on the results or the conclusions of the study. The relation of the presence of zidovudine resistance at base line to the outcome is currently under study.
In response to Delgado Fernández et al.: we do not have an explanation for the difference in the proportion of patients with neutropenia but can only postulate that the hematologic toxicity of zidovudine in the group that included indinavir may have been mitigated by the greater antiretroviral activity of this combination, which may have positively affected marrow function. A similar observation was made in a study of zidovudine and didanosine as compared with zidovudine alone.3
3 ReferencesScott M. Hammer, M.D.
Harvard Medical School, Boston, MA 02115Kathleen E. Squires, M.D.
University of Alabama at Birmingham, Birmingham, AL 35294Margaret A. Fischl, M.D.
University of Miami, Miami, FL 331011
Cameron DW, Heath-Chiozzi M, Kravcik S, et al. Prolongation of life and prevention of AIDS complications in advanced HIV immunodeficiency with ritonavir: update. In: Volume 1 of the Abstracts of the XI International Conference on AIDS, Vancouver, B.C., July 7–12, 1996:24. abstract.
2
CAESAR Coordinating Committee
CrossRef | Web of Science | Medline3
Collier AC ,Coombs RW ,Fischl MA , et al. Combination therapy with zidovudine and didanosine compared with zidovudine alone in HIV-1 infection. Ann Intern Med 1993;119:786-793
Web of Science | Medline
