e21016 Background: Currently there is a very limited data reporting response of adrenal metastases to immunotherapy in patients with metastatic malignant melanoma. Certain sites in the body may serve as sanctuary sites due to tumor microenvironment. Lower response rates and shorter PFS are reported in malignant melanoma with liver metastases while higher objective response rates are seen with lung metastases. We hypothesized that adrenal glands are sanctuary sites for melanoma metastases and thus would have poor response to immunotherapy due to steroid production in the adrenals making microenvironment less optimum for the anti-tumor effects of immunotherapy. Methods: A retrospective single-institution chart review of all malignant melanoma with metastasis to adrenal gland treated at University Colorado Hospital from 2008 till 2018, with age ≥ 18 yrs of age who received at least one dose of immunotherapy. Immune-related response criteria (irRC) was used to assess the response to immunotherapy on the imaging. Kaplan Meier estimate, and Chi square test were used for statistical analysis. Results: Sixty-five patients met the inclusion criteria. Mean age of 61 yrs. with range from 25 to 90 yrs. 71% were males and 95% were Caucasians. Median duration of follow up was 51.3 months with range of 1.5-212 months. Forty-three percent were cutaneous, 35% unknown primary, 11% each with mucosal and ocular melanoma. Based on the category of malignant melanoma, 60% were metastatic, 12% nodular, 8% choroidal, 8% superficial spreading, 6% acral lentiginous and 6% other. BRAF mutations were present in 38%, wild type in 54% and unknown in 8%. NRAS was mutated in 18% and wild type in 17% and unknown in 65% of the patients. Comparison of response between adrenal metastasis and non-adrenal metastasis, showed that 71% adrenal metastases had progressive disease compared to 54% of non-adrenal metastases. The overall response rate (ORR) to immunotherapy was 15% in adrenal metastases compared to 20% of non-adrenal metastases. Disease control rate (DCR) was 29% in adrenal metastases compared to 46% non-adrenal metastases with p value = 0.015. Conclusions: The adrenal glands appear to be a sanctuary site for metastases in patients with malignant melanoma receiving immunotherapy. The reasons behind this are unclear but may be related to the microenvironment and local corticosteroid production suppressing the immune response. Preliminary data examining cellular responses in nine surgically removed adrenal metastases support this hypothesis and will be presented.
