Correspondence
Preventing Fungal Infections in Chronic Granulomatous Disease
N Engl J Med 2003; 349:1190-1191September 18, 2003
- Article
To the Editor:
To determine whether itraconazole prevents fungal infections, Gallin et al. (June 12 issue)1 used treatment alternation as the basis for inference but did not consider that the group with greater exposure would, by chance alone, tend to have more events (severe fungal infections) than the other group. Of the 39 patients, 21 received placebo during odd years. Using as the base not 0.5 but rather the probability that an event occurs during exposure to the treatment received during odd years increases the P value to somewhat more than 0.102, since P = 2[(0.5)5 + (5)(0.5)8] = 26, and 26 ÷ 256 = 0.102. The numerical difference is small, and the overall conclusions do not change. However, the power of the study could have been increased if the investigators had randomized cycles within patients,2 forgone the interim examinations, and used the most informative end point possible. With use of the information-preserving composite end point3 and use of the cycle as the unit of analysis, there would have been seven severe infections, five nonsevere infections, and 51 cycles of exposure without any infection in the placebo group and one severe infection, no nonsevere infections, and 60 cycles without any infection in the itraconazole group. This approach would have yielded P=0.0021 (by the exact Smirnov test4). Gallin and colleagues' study design, possibly augmented with alternative analyses, is useful for studying rare diseases.
4 ReferencesVance W. Berger, Ph.D.
National Cancer Institute, Bethesda, MD 20892-7354
vb78c@nih.gov 1
Gallin JI ,Alling DW ,Malech HL , et al. Itraconazole to prevent fungal infections in chronic granulomatous disease. N Engl J Med 2003;348:2416-2422
Full Text | Web of Science | Medline2
Honkanen VEA ,Siegel AF ,Szalai JP ,Berger V ,Feldman BM ,Siegel JN . A three-stage clinical trial design for rare disorders. Stat Med 2001;20:3009-3021
CrossRef | Web of Science | Medline3
Berger VW . Improving the information content of categorical clinical trial endpoints. Control Clin Trials 2002;23:502-514
CrossRef | Medline4
Berger VW ,Permutt T ,Ivanova A . Convex hull test for ordered categorical data. Biometrics 1998;54:1541-1550
CrossRef | Web of Science | Medline
To the Editor:
In the report by Gallin et al., there is no mention of review and approval of the protocol for the study by an institutional review board. In addition, the article describes seven serious infections in patients with chronic granulomatous disease while they were receiving placebo and one serious infection in a patient during treatment with itraconazole. I believe that the study was unethical. The code should have been broken, and all patients treated immediately with itraconazole.
Michael J. Franzblau, M.D.
University of California School of Medicine, San Francisco, San Francisco, CA 94143
mfranzblau@aol.com To the Editor:
Gallin et al. report that itraconazole is effective in preventing fungal infections in patients with chronic granulomatous disease. The one serious fungal infection that occurred in a patient during treatment with itraconazole was probably a result of that patient's noncompliance with the treatment. Moreover, in none of the patients was the development of clinically significant resistance observed. In contrast, we have noted serious fungal infections in two patients with X-linked chronic granulomatous disease who received itraconazole as prophylaxis. In one, an 11-year-old boy who received itraconazole for six years, invasive pulmonary aspergillosis due to a multidrug-resistant strain of Aspergillus fumigatus developed, as previously reported.1 The other patient, an eight-year-old boy who received itraconazole as prophylaxis for three years, presented with an inguinal abscess and lymphadenopathy due to Fusarium solani infection. The strain was resistant to itraconazole in vitro, and the patient was successfully treated with voriconazole and radical surgical excision of the infected tissue. These cases illustrate that serious fungal infections due to molds that are resistant to itraconazole may develop in patients with chronic granulomatous disease who undergo long-term prophylaxis with itraconazole.
1 ReferencesPaul E. Verweij, M.D.
Adilia Warris, M.D.
Corry M. Weemaes, M.D.
University Medical Center St. Radboud, 6500 HB Nijmegen, the Netherlands
p.verweij@mmb. umcn. nl 1
Warris A ,Weemaes CM ,Verweij PE . Multidrug resistance in Aspergillus fumigatus. N Engl J Med 2002;347:2173-2174
Full Text | Web of Science | Medline
Author/Editor Response
Dr. Berger correctly points out that slightly greater exposure to placebo than to itraconazole in our trial of prophylactic itraconazole makes the P value slightly larger than the stated value of 0.102. However, the imbalances in the number of days of treatment (on 48.5 percent of which itraconazole was given) and in the number of treatment courses (49.2 percent of which consisted of itraconazole) are substantially smaller than the imbalance in the treatment assignment in the initial year of randomization (when 46.2 percent of the patients were assigned to itraconazole). Therefore, the final adjustment to the P value would be quite small. The power of our study would have been increased if the interim examinations associated with sequential monitoring could have been avoided. However, at the time the study was designed we believed, as we do now, in retrospect, that this design was ethically necessary. A composite end point of invasive and superficial fungal infections would have been easier to use than the single end point of invasive fungal infection, but it would have been less robust. Serious fungal infections are the critical problem itraconazole was developed to address. It was gratifying that the results in patients with less severe fungal infections also supported our conclusions. Dr. Berger's elegant analysis of the composite end point, with an overall P value of 0.0021, strongly corroborates, in a quantified way, the efficacy of itraconazole.
Dr. Verweij and colleagues note that the fungal infections that occur during itraconazole prophylaxis may be resistant, suggesting that long-term daily itraconazole therapy may permit or create drug-resistant fungal infections. We and others have not made such an observation,1 possibly because our patients alternated between cycles with and without itraconazole. Additional observations and perhaps a trial in which continuous treatment is compared with intermittent treatment are warranted. Regardless of the drug-resistant infections noted by Dr. Verweij and colleagues, in our prospective trial, itraconazole markedly reduced the incidence of fungal infections in patients with chronic granulomatous disease.
Dr. Franzblau raises a difficult issue that frequently arises in clinical research: How much evidence is needed in favor of a therapy before it becomes untenable to continue giving some patients an alternative therapy? We were keenly aware of this problem. The study's sequential rule was designed so that the study would be stopped when the one-sided P value reached 0.051, the traditional level deemed to provide significant evidence of a difference. The decisive conclusion of our study — that itraconazole prevents fungal infections in chronic granulomatous disease — was possible only because we rigorously adhered to this design. This is the only justifiable basis on which the long-term use of potentially toxic and expensive treatments can be broadly recommended. Premature termination that endorses the initial hypothesis carries the risk of ending up with a conclusion that will withstand neither scientific nor ethical scrutiny.
1 ReferencesJohn I. Gallin, M.D.
Robert Wesley, Ph.D.
Steven M. Holland, M.D.
National Institutes of Health, Bethesda, MD 208921
Mouy R ,Veber F ,Blanche S , et al. Long-term itraconazole prophylaxis against Aspergillus infections in thirty-two patients with chronic granulomatous disease. J Pediatr 1994;125:998-1003
CrossRef | Web of Science | Medline
- Citing Articles (4)
Citing Articles
1
S. Blumental, R. Mouy, N. Mahlaoui, M.-E. Bougnoux, M. Debre, J. Beaute, O. Lortholary, S. Blanche, A. Fischer. (2011) Invasive Mold Infections in Chronic Granulomatous Disease: A 25-Year Retrospective Survey. Clinical Infectious Diseases 53:12, e159-e169
CrossRef2
Marieke E. B. Welzen, Roger J. M. Brüggemann, J. Merlijn Van Den Berg, Heleen W. Voogt, Jos H. Gilissen, Dasja Pajkrt, Nigel Klein, David M. Burger, Adilia Warris. (2011) A Twice Daily Posaconazole Dosing Algorithm for Children With Chronic Granulomatous Disease. The Pediatric Infectious Disease Journal 30:9, 794-797
CrossRef3
Venita DePuy, Vance W. Berger. 2005. Counterbalancing. .
CrossRef4
A. Warris, P. E. Verweij. (2005) Clinical implications of environmental sources for Aspergillus. Medical Mycology 43:s1, 59-65
CrossRef
