8
Figures
Figure 2. Proposed Approach to Systemic Glucocorticoid
Discontinuation
Taper down glucocorticoids to physiologic
replacement doses (avoid dexamethasone)
SCENARIO 1 (biochemical confirmation of normal
cortisol production is not desired/feasible):
1. Gradually taper down glucocorticoids and
eventually stop.
2. Closely monitor the patient for clinical
manifestations of adrenal insufficiency/
glucocorticoid withdrawal throughout the
tapering period and after glucocorticoid
discontinuation.
Patients receiving supraphysiologic glucocorticoid doses and at high risk of
adrenal insufficiency, in whom glucocorticoid therapy is no longer required
Continue
glucocorticoid
replacement
1
and re-
check early-
morning
cortisol after
a few months
Continue
glucocorticoid
replacement
1
and re-check
early-morning
cortisol after a
few weeks
Consider
dynamic
testing
Stop
glucocorticoid
replacement
2,3
In case of clinical concerns,
possible options are:
• Move to scenario 2.
• Temporarily increase the
glucocorticoid dose and taper
down over a longer period.
Important considerations:
• The proposed cortisol cut-offs
are only a guide and may vary
according to the cortisol assay
used and local protocols.
• The proposed cortisol cut-offs
only apply to patients who are
not under major stress (e.g.,
sepsis, trauma, surgery, or
other acute illness requiring
hospital admission).
• The proposed cortisol cut-offs
do not apply to subjects with
abnormal CBG and albumin
(e.g., use of oral estrogens,
pregnancy, advanced
liver cirrhosis, nephrotic
syndrome).
Likelihood of
adrenal insufficiency
Cortisol
300 nmol/L
10.0 µg/dL
150 nmol/L
5.0 µg/dL
LOW
HIGH
SCENARIO 2 (biochemical
confirmation of normal cortisol
production is desired/feasible):
Measure early-morning serum
cortisol 24 hours after the last
glucocorticoid dose.
1
Exogenous glucocorticoid should not be reduced below
the lower end of the physiologic replacement dose range
to ensure adequate replacement for adrenal insufficiency,
yet still providing a stimulus for HPA axis recovery.
Further significant dose reduction should only occur with
indication of HPA axis recovery.
2
Some patients with cortisol values close to the proposed 300 nmol/L (10µg/dL) cut-off may still
have a degree of supoptimal cortisol response when exposed to major stress (e.g., sepsis, trauma,
surgery, or other acute illness requiring hospital admission). Rely on clinical judgment and offer stress
glucocorticoid coverage if adrenal insufficiency is suspected in such cases. Dynamic testing may also
be considered.
3
Some patients may develop glucocorticoid withdrawal symptoms (e.g., those who have been on
supraphysiologic doses for a very long time) and may benefit from gradual tapering rather than an
abrupt discontinuation.
