Critical illness related corticosteroid insufficiency
(CIRCI) is a phenomenon occurring in conditions of severe inflammation. CIRCI
is a complex, syndrome like, constellation of disruptions in normal homeostasis
that can derange the normal compensatory mechanisms of the
hypothalamic-pituitary axis (HPA). The prevalence may be as high as one in five
critically ill patients, higher in severe sepsis. Adrenal dysfunction related
to trauma, burns, DIC, among other known and unknown factors can decrease
production of steroids by the gland. Cortisol is not stored, so as organ
dysfunction (including adrenal dysfunction) develops in severe sepsis, adrenal
insufficiency is not surprising. Cortisol and ACTH production falls in severe
sepsis. 90% of cortisol is bound to cortisol binding protein, levels of which
can drop by half during acute illness. Free cortisol is the active form but the
decrease in total cortisol due to binding protein loss may contribute to CIRCI.
Some authors further suggest that severe acute inflammation may cause tissue
cortisol resistance similar to that seen in chronic inflammatory conditions. Glucocorticoid
receptor down-regulation by endotoxin and inflammatory cytokines may be a
mechanism for tissue resistance.
Diagnosis of CIRCI has changed. Historically, if a patient
was resistant to crystalloid resuscitation and vasopressors, a dose of
dexamethasone was administered empirically, a baseline serum cortisol was
measured and a cosyntropin stimulation test was administered. This approach was based upon the notion
that if ACTH was administered and the adrenal responded appropriately or if the
baseline (random) cortisol was ≥10 mcg/dL, additional cortisol would not
be helpful. The limitations of this approach are multiple and this approach has
been abandoned for acutely ill patients. Currently there is no readily
available assay for free cortisol, the active form. Neither free nor total
cortisol levels take into account tissue resistance. Several randomized trials have
examined the validity of cosyntropin stimulation test in patients with septic
shock and the most important finding has been that administration of
hydrocortisone was associated with more rapid reversal of shock, independent of
the results of the stimulation test.
So one must ask ‘when is it appropriate to use steroids in
the shock state’? In severe sepsis, start with adherence to principles of
resuscitation outlined by the Surviving Sepsis Campaign (See Sepsis and Septic
Shock Guideline http://uktraumaprotocol.blogspot.com/2015/02/sepsis-and-septic-shock.html).
Once sepsis is suspected, aggressive crystalloid resuscitation should be
initiated and broad-spectrum antibiotics started after collecting appropriate
cultures. If there is not brisk response to appropriate volume expansion,
vasopressors should be initiated and hemodynamic monitoring devices placed. The
Campaign suggests that certain goals of resuscitation be met in the first 6
hours of treatment. As soon as it is clear that the above measures are not
helping to achieve these goals, administration of exogenous glucocorticoids
should be considered. Realistically, it will take 2-3 hours to get to this
point in the algorithm, but steroids should be certainly be considered and
acted upon within 6 hours of persistent shock.
Treatment should start with hydrocortisone and should last
until there is resolution of severe sepsis. Dexamethasone should not be used. Resolution
of severe sepsis is evidenced by resolution of vasopressor requirement, and
improvement of organ dysfunction. Previous timelines suggesting seven days of
treatment were arbitrary and have been abandoned. Tapering is probably best
left to clinical judgment, but is expected if glucocorticoid therapy has been
ongoing for five to seven days.
In 2008, the American College of Critical Care Medicine
published recommendations for the use of corticosteroids in the ICU, including CIRCI
and some reference to acute respiratory distress syndrome (ARDS). The recommendations with regards to sepsis have been
addressed and updated by the Surviving Sepsis Campaign, but no comments were
made with regards to the use of corticosteroids in ARDS without severe sepsis.
Though there are some data to suggest increased ventilator free days if
steroids are initiated for ARDS that has not improved in 7 days and the process
has not been going on for more than 14 days, the overall utility of steroids in
ARDS appears to be minimal and their use in general cannot be recommended.
Some authors advocate using corticosteroids as well as multiple
hormonal treatments in brain dead organ
donors. Brain death is associated with a catecholamine storm that can produce
myocardial dysfunction and lead to hemodynamic instability and organ or donor
loss. Vasopressin, insulin, corticosteroids and thyroid hormone have all been
reported in organ donors. Retrospective data suggest that early administration
of steroids benefits hemodynamics and oxygenation. Whether combination hormonal
therapy improves organ recovery is controversial. Currently the only data
supporting the use of combination therapy to improve graft function is in heart
recipients from one study. In this retrospective cohort study, one-year recipient
survival was improved, and the incidence of early graft dysfunction was
reduced. At present, there are inadequate data to support or reject the
prophylactic use of corticosteroids in brain dead organ donors because studies
are dissimilar and difficult to compare, with most showing neutrality in
outcomes. Overall, using a lower threshold for the administration of
corticosteroids to unstable brain dead donors is probably appropriate.
Finally, patients receiving therapeutic corticosteroids do not require stress doses of steroids
in the perioperative period as long as they continue to receive their usual
daily dose. Adrenal function testing is also not indicated as it is not
predictive of adrenal crisis (Marik P. Arch Surg 2008).
