Adrenal insufficiency in thyroid cancer patients treated with tyrosine kinase inhibitors and detected by ACTH stimulation test

Abstract

Purpose: Advanced thyroid cancer patients treated with tyrosine kinase inhibitors (TKI) can develop several adverse events (AEs), including adrenal insufficiency (AI).

Methods: We studied 55 patients treated with TKI for radioiodine-refractory or medullary thyroid cancer. The adrenal function was evaluated during follow-up by performing serum basal ACTH, and basal and ACTH-stimulated cortisol.

Results: Twenty-nine/55 (52.7%) patients developed subclinical AI during TKI treatment as demonstrated by a blunted cortisol response to ACTH stimulation. All cases showed normal values of serum sodium, potassium and blood pressure. All patients were immediately treated, and none showed an overt AI. Cases with AI were all negative for adrenal antibodies and did not show any adrenal gland alteration. Other causes of AI were excluded. The onset time of the AI, as measured in the subgroup with a first negative ACTH test, was < 12 months in 5/9 (55.6%), between 12 and 36 months in 2/9 (22.2%) and > 36 months in 2/9 (22.2%) cases. In our series, the only prognostic factor of AI was the elevated, although moderate, basal level of ACTH when the basal and stimulated cortisol were still normal. The glucocorticoid therapy improved fatigue in most patients.

Conclusions: Subclinical AI can be developed in > 50% of advanced thyroid cancer patients treated with TKI. This AE can develop in a wide period ranging from < 12 to > 36 months. For this reason, AI must be looked for throughout the follow-up to be early recognized and treated. A periodic ACTH stimulation test, every 6-8 months, can be helpful.

Keywords: ACTH; Adrenal insufficiency; Adverse event; Cortisol; Thyroid cancer; Tyrosine kinase inhibitors.

Fig. 1

Panel A Serum basal and ACTH-stimulated cortisol in the nine patients who developed primary adrenal insufficiency during follow-up and with a previous normal ACTH stimulation test. The diagnosis of adrenal insufficiency was defined by a peak cortisol level below 18 µg/dL at either 30 or 60 min. Panel B ACTH values at the time of the normal ACTH stimulation test in the same nine patients of panel A: 6/9 had already elevated values of basal ACTH (> 50 ng/L, as indicated by the dashed line), thus precluding imminent AI subsequently demonstrated by a blunted ACTH test

Fig. 2

Prevalence of cases with elevated basal level of ACTH (mean 95 ng/L; median 101 ng/L, range 54–136 ng/L) at the time of the first normal ACTH stimulation test during TKI treatment, both in patients who further developed adrenal insufficiency (n = 9) and those who did not (n = 26). A significant higher number of cases with elevated ACTH was observed in the first group