Mechanisms of Adrenal Insufficiency in the Perioperative Period
Adrenal insufficiency can be primary, in which the adrenal gland itself does not function properly. It can also be secondary or central, in which the hypothalamic/pituitary axis is functioning at sub-par levels. It is, in fact, secondary adrenal insufficiency that is most commonly encountered in clinical practice; this condition is caused by HPA axis suppression due to negative feedback from exogenous administration of corticosteroids.
Who Needs Perioperative Steroid Replacement?
Following are situations in which perioperative steroid replacement should be considered:
- Any patient who has received more than 20 mg of prednisone daily (or an equivalent dose of other glucocorticoids) for more than five days in the previous year is at risk for HPA axis suppression;13,14
- Any patient who has clinical Cushing’s syndrome from any steroid dose should be considered to have a suppressed HPA axis;6 and
- Any steroid dose given at bedtime—even physiologic doses—is more likely to suppress the HPA axis than a similar dose given in the morning because, at nighttime, the negative feedback from steroids to ACTH and corticotropin-releasing hormone (CRH) is more prominent.6
Patients receiving doses equivalent to 5 mg of prednisone in the morning for any length of time are not associated with HPA axis suppression.13 If these patients are on steroid replacement for primary adrenal insufficiency, however, it is vital to realize that they do not have the capacity to compensate for increased stress, surgical or otherwise, and may need additional replacement depending upon the nature and duration of their stressors.14
Doses greater than physiologic range but less than 20 mg/day of prednisone can cause suppression of the HPA axis, but the demarcation of the time frame is not absolutely clear. It is said that doses closer to the physiologic range cause suppression after about one month, whereas doses closer to 20 mg/day can do so in five days or more.13, 14
For patients who cannot be clearly identified by these criteria, provocative testing is recommended, as discussed below.
Laboratory Testing for HPA Axis Function
The ACTH stimulation test is recommended to test for HPA axis adequacy in perioperative patients.5,13,14 It has been shown to correlate well with the insulin tolerance test. Exogenous corticosteroids have to be held for 24 hours prior to testing because they affect the measured cortisol level, with the exception of dexamethasone, which does not affect the results.13
The test involves the administration of 250 mcg of cosyntropin—synthetic ACTH—intramuscularly or intravenously, followed by measurement of peak cortisol levels after 30 minutes.5 Traditionally, a plasma cortisol level greater than 18–20 mcg/dl defines adequate adrenal function, as quoted from Salem and colleagues.5
Basal values and the calculation of delta max—the difference between basal and peak values—have not been shown to correlate with clinical outcome in routine surgical patients.13 Although these values have been demonstrated to be useful in the setting of septic shock, the subject is still under debate, and readers interested in this aspect of adrenal insufficiency may refer to articles by Annane, Hollenberg, Gonzalez and colleagues.15,16,17
Use of a 250-mcg dose of cosyntropin has often been criticized for being supraphysiologic, and use of a 1-mcg dose has been suggested instead. This test has not yet been characterized well enough, however.13,18 In addition, a standardized, commercial preparation of the 1-mcg dose is not available.13 Thus, the low-dose test is, as yet, not routinely advocated.
To Replace or Not to Replace?
It is interesting to note that biochemical evidence of HPA axis dysfunction does not necessarily translate into clinical problems with surgical stress.19 Additionally, patients on chronic steroid supplementation higher than 5 mg/day of prednisone—or equivalent—may not necessarily suffer inappropriate biochemical responses to stress.20 With these observations comes the following question: Even if HPA axis dysfunction exists, when does it matter clinically, or does it matter? The difficulty of this situation has been discussed in detail by Levy and Shaw.21,22