Statistics from Altmetric.com
The recent letter from Meijer et al1concludes that measuring serum cortisol by RIA severely underestimates serum cortisol suppression in the presence of oral prednisolone. This is rather a sweeping statement as the underestimation will, of course, depend on the degree of the cross reactivity with the particular assay. For example, in another study where inhaled fluticasone and oral prednisolone were compared in asthmatic patients and the cross reactivity of the RIA was quoted at 11%, it was found that 1 μg inhaled fluticasone (pMDI plus spacer) was equivalent to 8.5 mg (95% CI 5.7 to 11.2) oral prednisolone for suppression of 08.00 hour plasma cortisol.2
From the data from Meijer et al for HPLC morning serum cortisol levels, prednisolone 30 mg per day produced 72% suppression compared with 38% suppression for fluticasone 2 mg per day (by DPI). Extrapolating between these two values, it seems that 1 mg per day inhaled fluticasone produces equivalent serum cortisol suppression to 7.9 mg per day oral prednisolone. This is similar to our own estimated ratio of 8.5 μg:1 using RIA. Furthermore, in another dose ranging study by Casale et al3 in asthmatics which compared the effects of inhaled fluticasone and prednisone on 22.00 hour serum cortisol levels (area under the curve) using HPLC, the relative degree of suppression was 15% for fluticasone MDI 440 μg daily compared with 55% for prednisone 7.5 mg daily, which extrapolates to 1 mg fluticasone MDI being systemically equivalent to 4.6 mg prednisone for adrenal suppression. As the addition of a large volume spacer doubles the adrenal suppression with fluticasone via MDI,4 the ratio reported by Casale et al in asthmatic patients equates to 1 mg fluticasone via MDI plus spacer producing equivalent suppression to 9.2 mg prednisone, which is similar to our own ratio of 8.5 μg:1.2
Taking all these data together clearly suggests that inhaled fluticasone is highly systemically bioavailable and produces systemic adverse effects at high doses which are equivalent to those produced by low doses of oral prednisolone.
We thank Dr Lipworth for his comments. The ratio of systemic effects of fluticasone to prednisolone cannot be deduced reliably from our data, but we agree that the suppression we found is probably not markedly different from the one found by his group or from that of others in the literature.
However, this was not the content of—or the reason for—our note in Thorax. Our attention was drawn at a rather late stage to the fact that assessing prednisolone induced cortisol suppression by conventional radioimmunoassay (RIA) could lead to underestimation of suppression due to cross reactivity in the assay.1,2 We therefore subsequently compared cortisol results measured by conventional RIA with values measured by HPLC and, indeed, a significant underestimation in the presence of prednisolone was detected. Other researchers and clinicians might not be aware of this problem when assessing cortisol suppression by systemic corticosteroids.