Considerations for Clinical Management, Options for Treatment, and Future Research
It is critical to collect a thorough medication history in patients exhibiting signs and symptoms of Cushing's syndrome. Additionally, it is important to distinguish between lipodystrophy, caused by some antiretroviral medications, and iatrogenic Cushing's syndrome, as they have many overlapping characteristics (e.g. central obesity, weight gain and dorsocervical fat pad). However, abdominal striae, easy bruising, facial plethora, rapid weight gain, increased appetite and facial hirsutism are commonly related to Cushing's syndrome and peripheral atrophy is more likely to be noted in lipodystrophy. For the diagnosis of adrenal insufficiency, an early morning plasma cortisol level should be obtained followed by a 30-minute synthetic ACTH or cosyntropin stimulation test. A cortisol level > 11 μg/dL is unlikely to be associated with significant adrenal insufficiency; however, levels < 3 μg/dL are strongly suggestive of adrenal insufficiency. A decreased response to the ACTH stimulation test is diagnostic of adrenal insufficiency.
In view of the staggering number of case reports, a previous literature review by Foisy and colleagues clearly advised against the co-administration of fluticasone with ritonavir. However, despite these well-documented cases and the fact that electronic prescribing systems generally provide warnings against the combination of most PIs and some corticosteroids (e.g. fluticasone, mometasone and budesonide), patients continue to inadvertently receive these medications. Most importantly, there is a scarcity of guidance and recommendations regarding the medical management of patients requiring treatment for HIV infection and asthma or allergic rhinitis. Our review suggests that possible options include: (1) initiating or changing the PI to another antiretroviral agent that does not inhibit the CYP3A4 isozyme (such as nonnucleoside reverse transcriptase inhibitors or integrase inhibitors), if this is allowable based on the drug resistance and treatment history; (2) substituting the inhaled/intranasal corticosteroid with another corticosteroid with less potential for drug interactions (we will discuss these options below); and (3) using an alternative medication, such as a leukotriene receptor antagonist (e.g. montelukast), a mast cell stabilizer (e.g. cromolyn), an anticholinergic agent (e.g. ipratropium) or an antihistamine. It is unclear if significant dose reductions in the inhaled/intranasal corticosteroid dose will result in complete resolution or avoidance of this drug–drug interaction and there are reports of cases with an inhaled fluticasone dose as low as 200 μg/day.[8,13] Those diagnosed with Cushing's syndrome should be evaluated for a potential need to initiate oral steroid replacement and gradual taper, being careful to reduce the steroid dose if the PI is continued. Adrenal function recovery may take 9–12 months[35,37] and serum cortisol levels can be checked every 4–6 weeks to reassess the need for ongoing steroid replacement therapy.
It is noteworthy that this drug–drug interaction is not limited to inhaled and intranasal corticosteroids. Cushing's syndrome has previously been reported when injectable[38,39] and topical corticosteroids[40,41] were used in conjunction with CYP3A4 inhibitors. In a recent cohort study, the prevalence of iatrogenic Cushing's syndrome in HIV-positive individuals receiving a ritonavir-containing antiretroviral regimen in addition to inhaled, intranasal or topical corticosteroids (for ≥ 1 month and using ≥ 4.5% of total body surface area) was examined. Approximately 35% of individuals met the diagnosis criteria of iatrogenic Cushing's syndrome, including one patient using topical triamcinolone. The injection of triamcinolone in patients on ritonavir has similarly resulted in the development of Cushing's syndrome.[38,39] Therefore, this is a potential risk for all individuals on PIs who are chronically using corticosteroids via any route.
Results of one of the few studies to specifically evaluate the drug–drug interaction between corticosteroids and PIs were recently presented.[43,44] This was a randomized open-label study in HIV-negative healthy adults. The investigators evaluated whether ritonavir alone or darunavir/ritonavir could significantly increase plasma concentrations of beclomethasone 17-monopropionate (17-BMP), the more active metabolite of beclomethasone dipropionate (BDP). Additionally, they investigated whether inhaled BDP when co-administered with ritonavir or darunavir/ritonavir could significantly influence adrenal function in these participants. For 14 days, 30 subjects received orally inhaled BDP 160 μg twice daily and on day 15, they were randomized (1:1:1) to three groups: group 1 (control) remained on BDP alone for 14 days; group 2 received BDP and ritonavir (100 mg twice daily) for 14 days; and group 3 received BDP and darunavir/ritonavir (600/100 mg twice daily) for 14 days. It was noted that darunavir/ritonavir did not significantly increase the area under the concentration–time curve of 17-BMP and did not cause significant adrenal suppression. Ritonavir alone resulted in a twofold increase in 17-BMP, which was statistically significant; however, given that this increase did not result in significant adrenal suppression, it was not considered clinically consequential. The authors concluded that the use of BDP is preferable to fluticasone in patients receiving a PI and requiring an inhaled corticosteroid. Given the fact that PIs exhibit varying degrees of CYP3A4 inhibition, these promising results with the use of BDP should be interpreted with caution and warrant further research.
However, as we await additional research in this area, we can to some extent be reassured by the fact that evidence from pooled pharmacokinetic data (Table 1), cases reports (Table 2) and prior research[43,44] suggests that inhaled or intranasal beclomethasone is a relatively safe option in HIV-positive individuals who are receiving PIs. Additionally, based on the pharmacokinetic comparison of these compounds (Table 1), flunisolide appears to have a low risk of drug–drug interactions and may be another potential option in view of its similarly low glucocorticoid RRA, low lipophilicity, short elimination half-life, and weak CYP3A4 metabolism.
The drug–drug interaction between corticosteroids, especially fluticasone, and PIs is significant and deserves close attention and evaluation. Physicians, pharmacists and other health care providers should be further educated about the risks associated with this drug–drug interaction and be provided with alternative options for the treatment and resources for the management of patients. Electronic prescribing systems should frequently be updated to ensure accuracy of warnings for the use of corticosteroids via any route and PIs. Pharmacists should regularly generate lists of patients who may have inadvertently received inhaled/intranasal corticosteroids and PIs and advise prescribing clinicians of appropriate substitutions. Lastly, future research should examine the use of inhaled/intranasal corticosteroids, such as beclomethasone and flunisolide, in HIV-positive individuals receiving commonly prescribed PIs.
HIV Medicine. 2013;14(9):519-529. © 2013 Blackwell Publishing