Primary adrenal insufficiency (PAI) is caused by the failure of the adrenal gland to produce steroid hormones - glucocorticoids and mineralocorticoids - and is a potentially lethal disease. Synthetic steroid hormones have transformed PAI from a lethal condition to a chronic one. However, management of PAI is still challenging for patients and clinicians as the current regimens do not restore or replicate normal cortisol secretion in normal conditions and during illness and stress. Hence with current treatment mortality remains higher than the control population and quality of life is poor. The goal is to treat PAI patients with adrenocortical cells in immunoisolation devices to restore physiological steroid hormone secretion without needing immunosuppression.

We studied the survival and function of isolated human adrenal cells in vitro and in vivo in immunodeficient NOD mice (NODRag1nullIL2Rgcnull). About 3 x10⁸ adrenocortical cells/ human adrenal gland were routinely obtained with > 80% viable cells, with survival and function in vitro for more than 14 days. A cohort of 10 immunodeficient mice was implanted with immunoisolation devices into the epididymal or ovarian fat pad.  After 4 weeks, when the vascularisation of the device was established, mice underwent bilateral adrenalectomy and 5 x10⁶ human adrenocortical cells were loaded into the device. Two mice died 4 weeks after adrenalectomy. The remaining mice were healthy, secreting cortisol and responding to stimulation with synthetic ACTH 1-24 (Synacthen) for >10 weeks. Mice were culled after 12 weeks, and harvested devices were found to be well-vascularised and filled with transplanted cells. Results indicate the survival and function of encapsulated human adrenal cells.

To scale up, devices were implanted into sheep. After 4 weeks, sheep underwent bilateral adrenalectomies, which were confirmed by the absence of cortisol response to Synacthen. Sheep were maintained on exogenous cortisol and DOCA until the loading of devices with sheep adrenal cells. We routinely isolated an average of 50 x10⁶ sheep adrenal cells/gland. While device vascularisation was seen in the sheep, the long-term function of the encapsulated cells was not observed. We found that sheep adrenal cells were more difficult to isolate than human cells because of the thick capsule resulting in less viability. Therefore, to study the function and long-term survival of adrenal cells in a large animal model, we will use human adrenal cells in pre-implanted encapsulation devices in immunosuppressed sheep.