ECE2018 Oral Communications Genomic and clinical aspects of endocrine tumours (5 abstracts)
Urine steroid metabolomics as a diagnostic tool for detection of adrenocortical malignancy – a prospective test validation study
Irina Bancos 1, , Angela Taylor 2 , Vasileios Chortis 2 , Alice Sitch 2 , Katharina Lang 2 , Alessandro Prete 2 , Massimo Terzolo 3 , Martin Fassnacht 4 , Marcus Quinkler 5 , Darko Kastelan 6 , Dimitra Vassiliadi 7 , Felix Beauschlein 8, , Urszula Ambroziak 10 , Michael Biehl 11 , Jonathan Deeks 2 & Wiebke Arlt 2
1Mayo Clinic, Rochester, USA; 2University of Birmingham, Birmingham, UK; 3University of Turin, Turin, Italy; 4University of Wuerzburg, Wuerzburg, Germany; 5Endocrinology Charlottenburg, Berlin, Germany; 6University Hospital, Zagreb, Croatia; 7Evangelismos Hospital, Athens, Greece; 8Ludwig-Maximilians-University, Munich, Germany; 9University of Zuerich, Zuerich, Switzerland; 10Medical University of Warsaw, Warsaw, Poland; 11University of Groningen, Groningen, Netherlands.
Background: Adrenal masses are discovered in 5% of abdominal imaging scans. Accuracy of currently available imaging tests to diagnose malignancy is poor. In a proof-of-concept study (JCE&M 2011;96(12):3775-84), we had demonstrated 90% sensitivity and specificity in detecting adrenocortical carcinoma (ACC) for urine steroid metabolomics, the combination of mass spectrometry-based steroid profiling and machine learning-based data analysis. This diagnostic performance is superior to costly imaging procedures currently used for differentiating benign from malignant adrenal masses, which lead to a high rate of unnecessary surgery. Implementation of our novel test in routine practice requires prospective validation.
Methods: We undertook a prospective multi-center international test validation study, powered to achieve recruitment of 2000 patients with an anticipated ACC rate of 5%, with prospective recruitment of patients with newly diagnosed adrenal mass >5 mm, biochemical exclusion of pheochromocytoma and 24-h urine collection; recruitment was carried out in 13 centers (11 countries) of the European Network for the Study of Adrenal Tumors (ENSAT). Urinary steroid excretion was quantified by high-throughput liquid chromatography-tandem mass spectrometry and results processed by an algorithm based on generalized matrix relevance learning vector quantization (GMLVQ). Reference standard (benign/malignant) was based on histology and imaging follow-up.
Results: We enrolled 2017 patients, 1767 (87.6%) with a benign adrenocortical adenoma (ACA), 98 (4.9%) with ACC, and 87 (4.3%) and 65 (3.2%) with other benign and malignant adrenal masses, respectively. Risk of ACC was highest in patients <40 years (13%; vs 4% in >40 years, P<0.0001) and adrenal masses >4 cm (20%; vs 0.13% in <4 cm, P<0.0001). Unenhanced CT imaging of the adrenal mass was available for 1328/1767 patients with ACA; 68% of masses had a radiodensity <10 HU indicative of a benign lesion; 17% had borderline results (10–20 HU) and 15% were suspicious of ACC (>20 HU). MRI with chemical shift indicated suspicion of ACC in 22% of 273 benign ACA. Adrenalectomy was performed in 21% (370/1767) of ACA patients. Urine steroid metabolomics demonstrated an excellent diagnostic performance with AUROC of 94.6% for 15 steroids (Sens=Spec 87.1%).
Conclusions: Overall risk of ACC in adrenal tumors is 4.9% and almost exclusively relates to adrenal masses >4 cm. ACAs are frequently misclassified as malignant by routine imaging, resulting in a high rate of imaging and unnecessary adrenalectomies. Urine steroid metabolomics demonstrates high accuracy for detection of ACC and should become standard-of-care in patients with indeterminate adrenal tumors.
Article tools
My recent searches
My recently viewed abstracts
Authors
Endocrine Abstracts
ISSN 1470-3947 (print) | ISSN 1479-6848 (online)
© Bioscientifica 2024 |
Privacy policy |
Cookie settings
BiosciAbstracts
Bioscientifica Abstracts is the gateway to a series of products that provide a permanent, citable record of abstracts for biomedical and life science conferences.
Find out more