• adrenal crisis
• high-dose fluticasone
• inhaled corticosteroids

In response to comments by Pearce and Mabin on Professor Russell’s editorial¹ on our paper.²

They doubt that our survey underestimated the true scale of the problem. I can inform themthat this is not the case. Since our survey was completed we have been notified of a further seven cases (five children, two adults). All but one of the children had been taking fluticasone in similar dosages to those reported in our survey. Three of these were critically ill in intensive care and an 8-year-old girl died due to adrenal crisis. The remaining child was only 20 months old and had been given budesonide in extremely high doses of 2000–8000 mcg/day.³ Both adults had been taking fluticasone (1000 mcg/day, 2250 mcg/day).

Case reporting clearly plays a much greater role than clinical studies in post license surveillance of new drugs. In an international 20 year study of drug safety discontinuations, nearly all occurred as a result of case reporting. Despite studies the authors concluded that “it is impossible to know fully all the facts about a drugs effects both beneficial and harmful at the time of approval”.⁴ Further, it is incorrect for Pearce and Mabin to say that “studies show no increased risk of hypothalamic pituitary axis (HPA) suppression with fluticasone propionate when compared with other inhaled steroids”. There are many studies arguing against this assertion.^5–9 Actually, there is a serious disparity between the results of different safety studies involving fluticasone which requires explanation. While the product monograph claims “mean plasma cortisol concentrations remained within the normal range for adults and children demonstrating that, even at high doses (2000 mcg), fluticasone propionate is well tolerated with regard to side effects”,¹⁰ many studies suggest otherwise. For example, as little as 88 mcg/day of fluticasone can produce 10% adrenal suppression ¹¹ and 352 mcg/day can produce 50% adrenal suppression, a considerably greater degree than the equipotent dosage of beclometasone,⁹ leading those authors to conclude that “increasing the dose beyond this point of maximum efficacy... resulted in increasing systemic effect, especially with fluticasone metered dose inhaler with its CFC propellant”—exactly what we have reported.

Pearce and Mabin correctly state that “individuals have differing sensitivities to inhaled corticosteroids. Idiosyncratic responses to inhaled corticosteroids may occur even at licensed doses”. I agree, however, such patients “disappear” in large multicentre studies, particularly when the pharmaceutical company will not make available data on individual patients.¹² This is of concern as there is evidence that fluticasone is associated with significantly more individual abnormally low cortisol values than other inhaled steroids.¹³ Also, over 75% of patients developed adrenal crisis greater than one year after starting fluticasone, and it is known that length of time taking inhaled steroids is a major factor in determining the frequency of side effects.¹⁴ How many studies of fluticasone have lasted greater than one year?

Pearce and Mabin correctly state that in recent years, when paediatricians decide that high doses of inhaled corticosteroids are necessary, more are choosing to prescribe fluticasone propionate. However, they need to explain why only 2 cases of adrenal crisis (both adults) in over 30 years of prescribing inhaled corticosteroids had ever been reported in literature before the introduction of fluticasone propionate allowing Russell to make a claim in 1994 that “there is no firm evidence that any child has ever come to harm as a result of adrenal suppression induced by inhaled corticosteroid therapy”.¹⁵ Further, some cases reported in our survey had previously been taking very high doses of either beclometasone or budesonide but only developed adrenal crisis some time after changing to fluticasone.¹⁶

Finally, it is unfair to blame doctors for prescribing fluticasone “off label”. Almost half of all drug prescriptions for children in hospital are either unlicensed or off label.¹⁷ Prescribers have every right to expect a reasonable margin of safety with a drug should they decide that off label dosages are necessary in children. Bearing in mind that there have now been two reported deaths and many intensive care cases, the risks of prescribing fluticasone off label appear greatly to exceed any possible benefits for patients, and will have serious medico-legal implications for doctors, particularly when there is not a single study showing better efficacy for fluticasone compared with other available inhaled corticosteroids.¹⁸