Correspondence
Oral Ganciclovir for Cytomegalovirus Retinitis
N Engl J Med 1996; 334:55-56January 4, 1996
- Article
To the Editor:
We are troubled by the strong conclusion of Drew et al. (Sept. 7 issue)1 that “oral ganciclovir is safe and effective as maintenance therapy for cytomegalovirus retinitis.” This statement implies equivalence between intravenous and oral ganciclovir, which we believe is not convincingly supported by the data.
When there is no difference between the treatment groups, one does not accept the null hypothesis but rather fails to reject it. In other words, if the data show there is no difference between intravenous and oral ganciclovir, it does not imply they are equivalent but rather that there is not sufficient evidence to conclude that oral is worse than intravenous ganciclovir.
The sample size was too small to detect the stated difference of 25 days in the time to progression between treatment groups with 80 percent power. With the use of a mean time to progression of 70 days for the intravenous-ganciclovir group and 45 days for the oral-ganciclovir group, 120 patients (60 per group), a two-sided type I error of 0.05, an accrual period of 15 months, and a 20-week follow-up period, the power is 67 percent.2 With the inclusion of only 115 patients capable of being evaluated, the power drops to 65 percent. To achieve a power of 80 percent, 162 patients needed to be evaluated.
Given the above calculations along with the conflicting results of the funduscopic and photographic evaluations with respect to the time to progression and the occurrence of new cytomegalovirus retinitis in the previously uninvolved eye, we believe the authors have overstated their conclusions.
2 ReferencesMark L. Van Natta, M.H.S.
Janet Holbrook, M.S., M.P.H.
Johns Hopkins University School of Public Health, Baltimore, MD 212051
Drew WL ,Ives D ,Lalezari JP , et al. Oral ganciclovir as maintenance treatment for cytomegalovirus retinitis in patients with AIDS. N Engl J Med 1995;333:615-620
Full Text | Web of Science | Medline2
Schoenfeld DA ,Richter JR . Nomograms for calculating the number of patients needed for a clinical trial with survival as an endpoint. Biometrics 1982;38:163-170
CrossRef | Web of Science | Medline
To the Editor:
Why did Drew et al. calculate the intravenous dose on the basis of body weight (i.e., milligrams per kilogram of body weight) and give the oral maintenance therapy in an arbitrary dose of 3000 mg per day in six divided doses? The authors do not provide information on the specific antiretroviral agents used in each group, and marrow suppression and sepsis may be related to the additive toxic effects induced by the combination of zidovudine and ganciclovir.
Failures after successful induction therapy may reflect not the natural history of the disease, but the failure of physicians to prescribe an adequate maintenance dose of ganciclovir to suppress the virus effectively. For several years I have prescribed maintenance doses of 8 to 10 mg per kilogram daily and generally do not see a recurrence of cytomegalovirus retinitis in less than 9 to 12 months when this dose is tolerated. As a rule, this dose is tolerated. I generally avoid using concomitant zidovudine therapy and have not seen frequent septic episodes. I monitor platelet counts because they rise with ganciclovir therapy, reflecting the suppression of these marrow elements by systemic cytomegalovirus.
That the patients were instructed to take 500 mg of ganciclovir six times per day rather than 1000 mg three times per day, as is now generally prescribed and is clearly more convenient, suggests that the authors are aware of the decreased bioavailability of the drug as the bulk of drug exposed to the gut is increased. Missed doses, in practice, may promote viral resistance, and this may become a bigger problem as use of the newly available oral form of the drug increases.
Josh Torgovnick, M.D.
31 Washington Sq. W., New York, NY 10011Author/Editor Response
The authors reply:
To the Editor: Dr. Torgovnick raises several interesting questions. There is no rationale for varying the oral dose according to weight, since pharmacokinetic studies of oral ganciclovir have shown no clear relation between weight and plasma concentrations. When this study began, we thought that oral ganciclovir needed to be given in a dose of 500 mg six times per day to maximize absorption. Clinical-trial results (unpublished data) also support the efficacy of a 1000-mg dose given three times daily. Either regimen can be used. The question of whether higher doses may be even more efficacious is the subject of a study currently under way that is comparing doses of 3000, 4500, and 6000 mg daily. The standard dose of intravenous ganciclovir as maintenance treatment for cytomegalovirus retinitis is 5 mg per kilogram per day, but we are aware of studies evaluating other dosing regimens.
The use of antiretroviral agents was similar in the treatment groups; 65 percent of all subjects took some antiretroviral medication, with didanosine used most frequently. Users of zidovudine did not have appreciably greater neutropenia or anemia in our study.
Van Natta and Holbrook question the statistical basis of our conclusions. Although we reported a nonsignificant P value for the difference in the time to progression as assessed by studying fundus photographs in the two treatment groups, this was not the basis of our conclusion. Rather, it was based on a quantification of the difference between the two formulations. We estimated that the average difference in the time to progression was -5 days (oral-ganciclovir group minus intravenous-ganciclovir group), with a 95 percent confidence interval of -22 to +12 days. Because this interval does not include day -25, we can conclude with 95 percent confidence that the difference is less than 25 days. This fact supersedes any considerations of post hoc power.
In our concluding paragraph, we acknowledge that there were “small differences in efficacy” between oral- and intravenous-ganciclovir maintenance treatment — differences that would result in a risk of earlier progression of retinitis with oral ganciclovir. This risk is balanced by the absence of the need for hour-long daily intravenous infusions and by a lower risk of sepsis. Oral ganciclovir is therefore a useful option for those undergoing lifelong treatment for cytomegalovirus retinitis. Careful clinical judgment must be exercised in cases in which even limited progression of retinitis can have major visual consequences.
The names of Dr. M.-H. Heinemann and Dr. K. Squires, of Cornell University Medical College, New York, were omitted from the Appendix. We regret the omission.
W. Lawrence Drew, M.D., Ph.D.
University of California, San Francisco–Mt. Zion Medical Center, San Francisco, CA 94120Mary Jean Stempien, M.S., M.D.
Roche Bioscience, Palo Alto, CA 94303Anna Shadman, M.Sc.
Otsuka America Pharmaceutical, Palo Alto, CA 94306- Citing Articles (1)
Citing Articles
1
Jan Erik Otterstad. (2005) Guidelines and registries: Secondary prophylaxis after AMI with emphasis on the use of beta-blockers and ACE inhibitors. Scandinavian Cardiovascular Journal 39:1-2, 10-12
CrossRef
