THERAPEUTIC RATIO OF INHALED CORTICOSTEROIDS: FACT OR FICTION?

To the Editor :

We read with interest data published by Drs. Nielsen and Dahl (1), describing a ratio of topical to systemic potencies for inhaled fluticasone and budesonide. The concept behind the calculation of this ratio is sound and promises to assist therapeutic decisions. Unfortunately, results for both the topical and systemic potencies they used to calculate this ratio are invalid.

To assess topical potency, patients received doubling doses of fluticasone or budesonide at 2-wk intervals without washout and without dose randomization. Because effects of inhaled corticosteroids on airway responsiveness are still accumulating between 2-6 wk (2), this dosing scheme confounds effect of increasing dose with effect of duration of exposure. For example, the change in PC₂₀ between the end of the low and middle dose periods is due to both doubling of dose and the difference in total length of treatment (2 vs. 4 wk). The net effect is a spuriously steep dose-response slope. The method used by the authors to calculate relative potency works by dividing the difference in log-PC₂₀ response between the two formulations by the slope of the dose-response relationship (and antiloging the result). This produces a ratio of doses required to produce equal responses (i.e., relative potency) (6). When this slope is inaccurate (e.g., falsely steep), the resulting relative potency estimate and its confidence interval will also be inaccurate. Hence, the authors' topical potency estimate is invalid.

The authors minimize the importance of this time effect. However, equality of this inaccuracy of slope between the formulations does not change its effect on accuracy of relative potency. Neither does use of a 2-4 wk, low-dose inhaled steroid run-in period eliminate the problem. Effects of inhaled steroids progressively accrue for months after initiation (2).

The authors' systemic potency estimate, based on urinary cortisol excretion, is also troublesome because dose-response plots for the two formulations are not parallel. This violates one of the standard "validity criteria" of the statistical method used to calculate relative potency and invalidates the result (6). Nonparallelism indicates that the potency difference between formulations changes with dose level, which is incompatible with the definition of relative potency.

We applaud the authors' intent and rationality, but must conclude that neither their topical potency ratio estimate (based on log-PC₂₀ plot) nor their systemic potency estimate (based on 24-h urinary cortisol excretion) nor, therefore, their therapeutic ratio of these potencies are valid. Consequently, the authors' declaration that the therapeutic ratio of topical to systemic potency favors fluticasone over budesonide is not supported by their data.

University of Florida, Gainesville, Florida

Richard C. Ahrens, and William R. Clarke

The University of Iowa, Iowa City, Iowa

1. Nielsen LP, Dahl R. Therapeutic ratio of inhaled corticosteroids in adult asthma: A dose-range comparison between fluticasone and propionate and budesonide, measuring their effect on bronchial hyperresponsiveness and adrenal cortex function. Am J Respir Crit Care Med 2000; 162: 2053-2057 [Abstract/Free Full Text].

2. Juniper EF, Kline PA, Vanzieleghem MA. Ramsdale EH, O'Byrne PM, Hargreave FE. Long-term effects of budesonide on airway responsiveness and clinical asthma severity in inhaled steroid-dependent asthmatics. Eur Respir J 1990; 3: 1122-1127 [Abstract].

3. Booms P, Cheung D, Timmers MC, Zwinderman AH, Sterk PJ. Respiratory pathophysiologic responses: Protective effect of inhaled budesonide against unlimited airway narrowing to methacholine in atopic patients with asthma. J Allergy Clin Immunol 1997; 99: 330-337 [Medline].

4. Noonan MJ, Chervinsky P, Wolfe J, Liddle R, Kellerman DJ, Crescenzi KL. Dose-related response to inhaled fluticasone propionate in patients with methacholine-induced bronchial hyperresponsiveness: a double-blind, placebo-controlled study. J Asthma 1998; 35: 153-164 [Medline].

5. Hofstra WB, Neijens HJ, Duiverman EJ, Kouwenberg JM, Mulder PGH, Kuethe MC, Sterk PJ. Dose-responses over time to inhaled fluticasone propionate treatment of exercise- and methacholine-induced bronchoconstriction in children with asthma. Ped Pulmonol 2000; 29: 415-423 .

6. Finney DJ. Parallel line assays. In: Statistical methods in biological assay. 3rd ed. New York: MacMillan; 1978. p. 69-147.

From the Authors:

We thank Dr. Asmus and colleagues for their interest in our publication (1). Indeed, it is for the very points that they raise in their discussion of the time-course of effect of inhaled corticosteroids, that we chose not to include a washout period or dose-randomisation. Introduction of dose-randomisation would have required an adequate washout period that in patients with symptomatic asthma, would have been unacceptable. We therefore used a standard clinical pharmacology design, appropriate to studies of this type, that eliminates this need for washout, maintains the patients' disease stability, and minimises carryover effects. In addition, Asmus and colleagues state that the design of the study, a three-dose comparison of both the efficacy and safety of the molecules, leads to a spuriously steep dose-response. However, a steep dose-response can also suggest a well designed investigation using an appropriately sensitive endpoint (which for inhaled corticosteroids PC₂₀ is regarded) and the present study supports this (2, 3). Flat dose-response curves and use of a less sensitive efficacy endpoint, would make it more difficult to compare the relative potency of two molecules.

Moreover, Asmus and colleagues object to the dosing scheme, claiming an additional time effect on the response at higher doses, as effects of inhaled corticosteroids should still be accumulating between 2-6 wk. This is apparently supported by several papers. However, these studies were performed in patients who were either steroid-naive or reduced in corticosteroid treatment to the level of symptom worsening, and therefore different from the population of the present study.

Considering the observation that the dose-response plots for urinary cortisol excretion are nonparallel, we examined this in some detail and welcome the opportunity to clarify the issue further. In clinical studies, population dose-response curves are never exactly parallel and the focus is always on the clinical effect of the observed nonparallelism, aided by statistical tests. In our study the nonparallelism was diverging with increasing doses in favour of FP. This suggests that the estimate of safety potency (0.6, i.e. FP 40% less potent in reducing 24-h urinary cortisol than budesonide (p < 0.01)) is likely to be an overestimate at lower doses, and an underestimate at higher doses, which are more clinically relevant with regards to safety.

In fact, the interpretation of the study in the light of the points raised by Dr. Asmus and colleagues actually suggests that the estimate could be considered conservative and biased against fluticasone propionate.

Aarhus University Hospital, Aarhus, Denmark

1. Nielsen LP, Dahl R. Therapeutic ratio of inhaled corticosteroids in adult asthma. A dose-ranging comparison between fluticasone propionate and budesonide, measuring their effect on bronchial hyperresponsiveness and adrenal cortex function. Am J Respir Crit Care Med 2000; 162: 2053-2057 .

2. Meijer RJ, Kerstens HA, Arends LR, et al . Effect of inhaled fluticasone and oral prednisolone on clinical and inflammatory parameters in patients with asthma. Thorax 1999; 54: 894-899 [Abstract/Free Full Text].

3. Wilson AM, Lipworth BJ. Dose response evaluation of the therapeutic index for inhaled budesonide in patients with mild to moderate asthma. Am J Med 2000;108;269-275.