Deep learning approach for differentiating indeterminate adrenal masses using CT imaging.

PURPOSE: Distinguishing stage 1-2 adrenocortical carcinoma (ACC) and large, lipid poor adrenal adenoma (LPAA) via imaging is challenging due to overlapping imaging characteristics. This study investigated the ability of deep learning to distinguish ACC and LPAA on single time-point CT images. METHODS: Retrospective cohort study from 1994 to 2022. Imaging studies of patients with adrenal masses who had available adequate CT studies and histology as the reference standard by method of adrenal biopsy and/or adrenalectomy were included as well as four patients with LPAA determined by stability or regression on follow-up imaging. Forty-eight (48) subjects with pathology-proven, stage 1-2 ACC and 43 subjects with adrenal adenoma >3 cm in size demonstrating a mean non-contrast CT attenuation > 20 Hounsfield Units centrally were included. We used annotated single time-point contrast-enhanced CT images of these adrenal masses as input to a 3D Densenet121 model for classifying as ACC or LPAA with five-fold cross-validation. For each fold, two checkpoints were reported, highest accuracy with highest sensitivity (accuracy focused) and highest sensitivity with the highest accuracy (sensitivity focused). RESULTS: We trained a deep learning model (3D Densenet121) to predict ACC versus LPAA. The sensitivity-focused model achieved mean accuracy: 87.2% and mean sensitivity: 100%. The accuracy-focused model achieved mean accuracy: 91% and mean sensitivity: 96%. CONCLUSION: Deep learning demonstrates promising results distinguishing between ACC and large LPAA using single time-point CT images. Before being widely adopted in clinical practice, multicentric and external validation are needed.

Differentiating between adrenocortical carcinoma and lipid-poor cortical adenoma: A novel cross-sectional imaging-based score.

BACKGROUND: Discrimination between adrenocortical carcinoma and lipid-poor cortical adenoma preoperatively is frequently difficult as these two entities have overlapping imaging characteristics. Differentiation will allow for the selection of the most appropriate operative approach and may help prevent over-treatment. We aimed to identify imaging features that could preoperatively differentiate adrenocortical carcinoma from lipid-poor cortical adenoma and use them in a novel imaging-based score. METHODS: We conducted a retrospective analysis of patients with pathologically proven adrenocortical carcinoma and lipid-poor cortical adenoma who underwent resection in a single tertiary referral center between March 1998 and August 2020. The inclusion criteria were diameter >1 cm, attenuation >10 Hounsfield units on nonenhanced computed tomography, and histopathologic diagnosis. Patients with metastatic or locally advanced adrenocortical carcinoma adenoma (stages 3-4) were excluded. We developed a score using binary logistic multivariate regression model in 5-fold derivation (∼70%) cohorts with stepwise backward conditional regression as feature selection. Standardized mean regression weight was used as variable score points. RESULTS: We identified 232 adrenals resected across 211 patients. By comparing the imaging characteristics of adrenocortical carcinoma (n = 56) and lipid-poor cortical adenoma (n = 156), we revealed statistically significant differences between the groups in 9 parameters: size, attenuation, thin and thick rim enhancement patterns, heterogeneity, calcification, necrosis, fat infiltration, and lymph node prominence. The score mean performance was 100% sensitivity for the exclusion of adrenocortical carcinoma, 80% specificity (95% confidence interval, 68.3-91.5), 66% positive predictive value (95% confidence interval, 52.3-78.7), and 100% negative predictive value with area under the curve of 0.974. CONCLUSION: We defined and evaluated a novel 9-variable, imaging-based score. This score outperformed any single variable and could facilitate safe preoperative discrimination of adrenocortical carcinoma and lipid-poor cortical adenoma.

The non-responding adrenal metastasis in melanoma: The case for minimally invasive adrenalectomy in the age of modern therapies.

BACKGROUND: Minimally invasive adrenalectomy has facilitated resection of resistant adrenal metastases. The adrenal gland may function as a sanctuary site for metastatic growth despite systemic therapy. The objective of the study was to assess the outcomes of selective minimally invasive adrenalectomy during immunotherapy. METHODS: Candidates included patients with adrenal metastases resistant to systemic therapy who underwent minimally invasive adrenalectomy. RESULTS: There were 15 patients undergoing 16 minimally invasive adrenalectomies. Patients received either immunotherapy or BRAF inhibition prior to surgery. The mean operative time was 130 min with a median length of hospital stay of 2 days. At a median follow up of 24 months, 7 patients have no evidence of disease, 6 patients had progression with eventual mortality, while another patients has stable disease with maintenance therapy. One was lost to follow up. CONCLUSION: Despite an increase in objective durable responses in metastatic melanoma, there is still some site-specific resistance in isolated areas like the adrenal where early minimally invasive adrenalectomy remains indicated.

Acute limb ischemia in cancer patients: aggressive treatment is justified.

BACKGROUND: The outcome of cancer patients with acute limb ischemia (ALI) is not well defined. The purpose of this study is to report our experience treating patients with active malignancy who developed ALI and compare their outcome with non-cancer patients. METHODS: A retrospective review of patients treated for ALI between 2009 and 2012 with ALI. We identified those patients who suffered from ALI and compared the outcome of those with active malignancy to those without malignancy. RESULTS: Of 147 patients treated for ALI (122 lower extremity, 25 upper extremity), 24 (16%) were cancer patients. Mean follow-up was 9.8 months for the malignancy group and 13.4 months for the control. Perioperative mortality rates were similar among cancer and non-cancer patients (20% vs. 16%, respectively, NS). Freedom from major amputation at 30 months was similar (95% vs. 89%, NS). Long-term survival rates of cancer patients were significantly lower compared to non-cancer patients (45% vs. 77% respectively, P<0.05). CONCLUSIONS: Treatment of ALI among cancer patients can be achieved with perioperative mortality and limb salvage rates comparable to non-cancer patients. Aggressive treatment is justified when treating cancer patients with ALI.

Endovascular Repair of Abdominal Aortic Aneurysms in the Presence of a Transplanted Kidney.

PURPOSE: To present our experience performing endovascular repair of abdominal aortic aneurysms in kidney transplanted patients. METHODS: A retrospective review of all patients who underwent endovascular aneurysm repair (EVAR) for abdominal aortic aneurysms (AAA) performed at our institution from 2007 to 2014. We identified all patients who had previously undergone a kidney transplant. Data collected included: comorbidities, preoperative imaging modalities, indication for surgery, stent graft configurations, pre- and postoperative renal function, perioperative complications, and survival rates. RESULTS: A total of 267 EVARs were performed. Six (2%) had a transplanted kidney. Mean age was 74 (range, 64-82) years; five were males. Mean time from transplantation to EVAR was 7.5 (range, 2-12) years. Five underwent preoperative planning with noncontrast modalities only. Devices used included bifurcated (n = 3), aortouniiliac (n = 2), and tube (n = 1) stent grafts. Technical success was achieved in all patients. None experienced deterioration in renal function. Median follow-up was 39 (range, 6-51) months. Four patients were alive at the time of the study. Two patients expired during the period of follow-up from unrelated causes. CONCLUSIONS: EVAR is an effective modality for the management of AAAs in the coexistence of a transplanted kidney. It can be performed with minimal morbidity and mortality without harming the transplanted kidney. Special consideration should be given to device configuration to minimize damage to the renal graft.

Endovascular treatment of lower extremity ischemia in chronic renal failure patients on dialysis: early and intermediate term results.

BACKGROUND: Peripheral arterial occlusive disease is common in patients with chronic renal failure requiring dialysis. Despite the increasing use of endovascular revascularization for lower extremity ischemia, the success rates of treating lower extremity ischemia in this challenging population remain obscure. OBJECTIVES: To assess the results of endovascular revascularization for lower extremity ischemia in dialysis patients. METHODS: We conducted a retrospective review of all dialysis patients who underwent endovascular treatment for critical limb ischemia (CLI) in our institution between 2007 and 2011. Data collected included comorbidities, clinical presentation, anatomic distribution of vascular lesions, amputation and survival rates. RESULTS: We identified 50 limbs (41 patients). Indications included gangrene in 22%, non-healing wounds in 45%, rest pain in 31%, and debilitating claudication in 4%. Mean follow-up was 12 months (1-51 months). Nineteen patients required amputations. Freedom from amputation at 5 years was 40%. Factors associated with amputation included non-healing wounds or gangrene (68% and 36% respectively) and diabetes (P < 0.05). The survival rate was 80% after 5 years. CONCLUSIONS: Despite improvement in endovascular techniques for lower extremity revascularization, the incidence of limb salvage among dialysis patients remains poor, resulting in a high rate of major amputations.