Flunisolide

To the Editors:

For a number of reasons, we were disappointed to see the article published by Nakanishi et al (1) in a recent issue of CHEST (September 2003). As two of the authors of a systematic review that addressed the specific question posed by Nakanishi et al, and was published in both The Cochrane Library (recently updated in May 2003) and in CHEST, (2) we were hoping for a change in direction in this field of research. While there are many problems with this study, we have focused on the main methodological limitations.

The question asked by these authors was whether therapy with inhaled corticosteroids (ICSs) can be used to replace therapy with oral corticosteroids after discharge from the emergency department in patients who are being treated for acute asthma. This topic has been studied previously by the authors of seven other trials, which are included in our systematic review. However, not all of these studies were mentioned by Nakanishi et al. In the conclusions of our review (2) published in CHEST last year, several recommendations for subsequent trials in this area were proposed, and none of these involved more small-scale studies using pulmonary function outcomes. Moreover, we were disappointed that Nakanishi et al did not use the current systematic reviews in the discussion of the results and their implications.

The primary outcome used by Nakanishi et al was percent predicted [FEV.sub.1]. Many investigators have suggested that pulmonary function test (PFT) results are a poor outcome to select for studies on patients who have been discharged from the emergency department with acute asthma. Not only are outcomes such as asthma relapse or quality of life more clinically important, but PFTs have poor psychometric properties in patients with acute asthma and may provide contradictory results. (3) In addition, the minimum difference in PFT results that would be clinically important is unclear. However, it is thought to be approximately 12%, (4) and the authors do not mention this important issue.

To appropriately answer the proposed research question, an equivalence trial is required. The authors present a sample size calculation based on detecting a 20% difference in [FEV.sub.1] First, the interpretation of differences in pulmonary function outcomes is problematic, although this 20% difference seems large. In the CHEST systematic review and in most trials on this topic, asthma relapse was proposed as the most clinically relevant primary outcome. For an equivalence trial based on asthma relapse, if overall relapse rates were 10% (they were 3% in this trial), 1,242 patients would be required to detect a 50% difference in the risk of relapse. This trial included only 58 patients.

To further confuse the issue, despite finding a statistically significant difference in favor of oral corticosteroids on this primary outcome, the authors then concluded that there was no difference in clinical examination findings between the groups (an outcome that we suspect is subject to bias, was not standardized or stated as an objective of the study, and was of questionable clinical importance) and state that therapy with ICSs was found to be useful. The validity of these claims is obviously lacking, and one wonders whether this is an illustration of author interpretation bias.

Finally, it is disheartening to see yet another small, inconclusive, industry-sponsored study in this area. In the preparation and update of the systematic review, we contacted all the major pharmaceutical companies that manufacture ICS agents, including Forest, requesting information on completed or ongoing studies. We were not made aware of this study, and while these results do not change the conclusions of the review, it does illustrate why many believe that the registration of and a central repository for clinical trials are necessary. (5) The ethics of continuing to conduct these trials without regard for previous evidence, using methods that are ill-designed to answer the stated objectives, needs to be better scrutinized by funding agencies, researchers, and editors. Hopefully, CHEST readers will not have been fooled.

Marcia L. Edmonds, MD, MSc

Brian H. Rowe, MD, MSc, FCCP

University of Alberta

Edmonton, AB, Canada

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: permissions@chestnet.org).

Correspondence to: Marcia L. Edmunds, MD, MSc, Division of Emergency Medicine, 1G1.43 WMC, 8440-112 St, Edmonton, AB, Canada T6G 2B7; e-mail: medmonds@ualberta.ca

REFERENCES

(1) Nakanishi AK, Klasner AK, Rubin BK. A randomized controlled trial of inhaled flunisolide in the management of acute asthma in children. Chest 2003; 124:790-794

(2) Edmonds ML, Camargo CA Jr, Brenner BE, et al. Replacement of oral corticosteroids with inhaled corticosteroids in the treatment of acute asthma following emergency department discharge. Chest 2002; 121:1798-1805

(3) Rowe BH, Bota GW, Fabris L, et al. inhaled budesonide in addition to oral corticosteroids to prevent asthma relapse following discharge from the emergency department: a randomized controlled trial. JAMA 1999; 281:2119-2126

(4) Karras DJ, Sammon ME, Terregino CA, et al. Clinically meaningful changes in quantitative measure of asthma severity. Acad Emerg Med 2000; 7:326-334

(5) Simes RJ. Publication bias: the ease for an international registry of trials. J Clin Oncol 1986; 4:1529-1541

To the Editor:

For a number of reasons we were disappointed in the letter by Drs. Edmonds and Rowe. These authors have published several meta-analyses assessing the use of inhaled corticosteroids (ICS) in the emergency department for the treatment of acute asthma (Cochrane database systemic reviews that they have referenced in their letter but also in the Annals of Emergency Medicine in August 2002 and in CHEST in 2002). (1,2) Each version concluded that ICS reduce hospital admission rates in children with acute asthma, but there was insufficient evidence that they resulted in clinically important changes in pulmonary function or clinical scores or that ICS alone are as effective as systemic steroids. They further suggested that additional research is needed to clarify if there is a benefit of ICS alone and if these are as effective as systemic steroids.

Although our study was completed and the manuscript submitted before the publication of these similar meta-analysis systematic reviews, we had addressed many of these concerns, based in part on a similar earlier meta-analysis of the same subject published by these same authors, which we did acknowledge. (3)

We recognized that acute treatment with oral corticosteroids in patients with moderately severe asthma who were well enough to be discharged to home might not produce an immediate change in pulmonary function but could be associated with hospital readmission. Therefore, we conducted and reported on a double-blind, double-dummy, randomized, prospective trial comparing ICS with oral prednisolone in children with moderately severe asthma. We objectively assessed pulmonary function and critically, we continued to monitor these children, following discharge over the 7-day study period. Pulmonary function was measured not only by peak expiratory flow (PEF), which is recognized to be highly variable and effort dependent, (4) but also with spirometry measured by an experienced technician. We evaluated readmission and retreatment rate of subjects. Repeat spirometry and symptom survey were conducted to determine the presence and resolution of symptoms. By the very criteria proposed by Dr. Edmonds, the study that we reported in CHEST would be considered "high quality."

Edmonds and Rowe claim that pulmonary function test measurements are a poor outcome for the assessment of acute asthma. They provide few data to support this contention which seems to be based primarily on the variability of PEF measurement. In contrast, there are a wealth of established clinical trial data suggesting that [FEV.sub.1] is one of the most valuable, widely accepted, and objective measures of response to asthma therapy. (5) Similar to earlier reports, we report no difference in morning PEF between groups. However, there was a significant difference in pulmonary function. Both groups had [FEV.sub.1] return to normal over the course of 2 weeks but the improvement in [FEV.sub.1] was significantly more rapid in those patients receiving oral corticosteroids. These data strongly suggest support for the observation that [FEV.sub.1] is more useful than PEF in asthma assessment.

Edmonds and Rowe also seem to have a problem with deciding what they want. They don't like pulmonary function testing, but they suggest that if this testing is used than the threshold for a difference should be set at 12% change. This 12% number seems to come from the standard assessment of bronchodilator responsiveness in a single patient and is unrelated to the evaluation of the effectiveness of therapy over time. We chose to be more conservative than dais and specifically powered the study to detect a 20% improvement hi [FEV.sub.1]. It confuses me that they would dislike data variability and dislike a conservative approach to minimize tiffs.

The authors of the letter further propose an outcome such as asthma relapse or quality of life (QOL) is more clinically important. To our knowledge, there are no validated QOL instruments that have been used for assessment of resolution of acute asthma following therapy. We did evaluate subsequent emergency department visits and hospital admission rates as well as conduct repeat examinations and symptom interviews to assess symptoms which are probably as good a surrogate for QOL related to asthma therapy as is available. Edmonds and Rowe erect a straw man by stating that to determine equivalence in a trial based on asthma relapse (specifically not or objective), 1,242 patients would be required to detect a 50% difference in risk of relapse. It is not clear why a 50% difference was chosen and absurd they are suggesting that studies with fewer than 1,200 patients are no longer of value.

Finally, the innuendo that this is an unethical, industry sponsored study it is extremely disturbing. We were fortunate that Forest Laboratories gave us all medication amt placebos (including pressurized metered dose inhaler dummies) as well as the holding chambers and flow meters used to conduct this study. No monetary support was provided to any investigator from any industry source. This was an investigator-initiated study designed, directed, and analyzed by the authors. The reason Forest Laboratories did not report this study to the authors of the letter is because they gave no monetary support and thus did not receive a copy of the results of this study before publication. Study safety and outcomes were monitored by a committee of the University Institutional Review Board and not by industry. Had the authors of the above letter taken even a moment to contact the authors of this study, this could have been quietly clarified before any accusations were made. This bias on part of Edmonds and Rowe is unlikely to fool the readers of CHEST. We hope that they will stop fooling themselves as well.

Albert K. Nakanishi, MD

Saint Louis University School of Medicine

St. Louis, MO

Brace Rubin, MD, FCCP

Wake Forest University School of Medicine

Winston-Salem, NC

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: permissions@chestnet.org).

Correspondence to: Albert K Nakanishi, MD, Department of Pediatrics, Cardinal Glennon Childrens, 1465 S Grand Blvd, St. Louis, MO 63119; e-mail: nakamink@sht.edu

REFERENCES

(1) Edmonds ML, Camargo CA, Brenner BE, et al. Replacement of oral corticosteroids in the treatment of acute asthma following emergency department discharge: a meta-analysis. Chest 2002; 121:1798-1805

(2) Edmonds ML, Camargo CA, Pollack CV, et al. The effectiveness of inhaled corticosteroids in the emergency department treatment of acute asthma: a meta-analysis. Ann Emerg Med 2002; 40:145-154

(3) Rowe BH, Keller JL, Oxman AD. Effectiveness of steroid therapy in acute exacerbations of asthma: a meta-analysis. Am J Emerg Med 1992; 10:301-310

(4) Cote J, Carrier A, Malo JL, et al. Compliance with peak expiratory flow monitoring in home management of asthma. Chest 1998; 113:968-972

(5) National Asthma Education and Prevention Program. Expert panel report 2: Guidelines for the diagnosis and management of asthma. Bethesda, MD National Institute of Health, April 1997; Publication 97-405

COPYRIGHT 2004 American College of Chest Physicians
COPYRIGHT 2004 Gale Group