Background

The introduction of proteasome inhibitors and immunomodulatory agents to treat patients with multiple myeloma (MM) and AL amyloidosis have greatly improved survival in these patients and allowed for the use of relatively steroid-sparing regimens.  However, 40 mg of dexamethasone is still more than 50 times basal glucocorticoid secretion.  Therefore, intermittent dosing of dexamethasone, the longest acting oral corticosteroid, while beneficial in terms of reducing side effects from chronic glucocorticoid excess, may still compromise endogenous adrenal gland function.

Our hypothesis was that prolonged, intermittent use of high dose steroids can result in adrenal insufficiency in some patients.  The aim of this retrospective study was to determine the characteristics of patients who developed AI.

Methods

Inclusion criteria for this retrospective case series were patients who had plasma cell disorders and who had been diagnosed with AI based on symptoms concordant with a low serum cortisol level (normal 6.7 - 22.6 mcg/dL), an inadequate cortisol response on an ACTH stimulation test, or for those patients with a fulminant clinical presentation - a rapid clinical improvement upon initiation of low dose maintenance corticosteroid replacement therapy.

Exclusion criteria were patients who had serum cortisol levels checked in the setting of recent administration of corticosteroids, who had an adequate cortisol response to an ACTH stimulation test in the setting of a normal basal cortisol level, or who did not require replacement therapy to achieve resolution of symptoms.

Results

Sixteen patients met the inclusion criteria over a span of approximately 18 months.  Two patients had AL amyloidosis, 12 had MM, and 2 had both.  3 patients were excluded. The median age of patients at the time of AI diagnosis was 61.5 (Range: 44-76).

The median number of steroid-containing cycles taken before the diagnosis of AI was 10.5 (Range: 4-50) over a median of 27 months (Range: 3-129). The median cumulative steroid consumption was 1000 mg of dexamethasone. Of note, the 2 primary AL amyloid patients only received 4 and 8 cycles of corticosteroids and a lower amount of cumulative corticosteroids, 768 and 800 mg, respectively prior to being diagnosed with AI.

The symptoms and signs of AI at the time of diagnosis included fatigue (88% of patients), diarrhea (56%), hypotension (56%), orthostasis (44%) and weight loss (31%).  Other symptoms that appeared in multiple patients included nausea, diffuse myalgia, fever, and cardiovascular shock. The median time between the last steroid dose and the serum cortisol assay was 7 days (Range: 1-62), which resulted in a median serum cortisol of 3.7 mcg/dL (Range 0.5-21 mcg/dL).

Of the seven patients who had serum ACTH levels checked, only one patient (with primarly AL amyloid) had an elevated ACTH of 68 (normal 12-46 pg/mL), suggesting a possible component of primary AI. Five patients also underwent ACTH stimulation tests, two of which demonstrated an inappropriate response, defined as a lower rise in serum cortisol than expected.

To treat AI, patients received about 15-20 mg of hydrocortisone (equivalent to 0.6- 0.8 mg of dexamethasone) on days not receiving steroids for treatment of their malignancy. 2 patients requiring pressors for shock also required stress dose steroids.  For patients with AI symptoms, normalization of hypotension and weight required a median of 9 days (Range 2-43 days) and 2 months (Range 0.3-20 months) respectively to return to their pre-AI levels. Orthostasis resolved after a median of 33 days (Range: 22-72 days), however, orthostatics were not checked at each clinic visit.

Conclusions

Our study shows that chronic treatment with even intermittent high-dose steroids can lead to AI, typically characterized by such nonspecific symptoms as fatigue, diarrhea, and dizziness occurring 3-4 days after the last exogenous steroid administration. Unfortunately, the debilitating presentation of some patients with orthostasis/hypotension as well as an already demanding schedule for chemotherapy visits makes optimal cortisol and ACTH stimulation testing challenging. A high index of suspicion for AI is required to initiate diagnostic and therapeutic interventions in a timely fashion to minimize the morbidity and mortality of this condition.

Disclosures:

Chari:Millenium : Membership on an entity’s Board of Directors or advisory committees; Celgene: Consultancy, Membership on an entity’s Board of Directors or advisory committees; Onyx: Membership on an entity’s Board of Directors or advisory committees.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution