Ann Rivard (Presenter)
The Mayo Clinic
Bio: Ann Rivard. I am a development technologist for the Clinical Mass Spectrometry Development Laboratory in the Department of Laboratory Medicine and Pathology, at The Mayo Clinic.
Authorship: Ann L. Rivard, Robert L. Taylor, Jolaine M. Hines, Irina Bancos M.D., Stefan K. Grebe M.D., Ph.D., Ravinder J. Singh Ph.D.
The Mayo Clinic- Department of Laboratory Medicine and Pathology
| Short Abstract Incidental adrenal tumors are found in approximately 5% of the 80 million CT scans performed in the U.S each year1. While most masses are benign adrenal adenomas, those with indeterminate imaging characteristics will require additional diagnostic workup to exclude adrenocortical carcinoma. To address these issues we developed a liquid chromatography-high resolution accurate mass spectrometry 24-hour urine panel of 26 steroid metabolites to distinguish ACC from benign ACAs. This non-invasive method allows for an accurate, rapid, and cost-effective means of diagnosis. Here we present the analytical validation. This method decreases the need for more aggressive diagnostic procedures which most often includes surgery as well as biopsies, provocative hormone driven testing, and repeated imaging. This method brings significant improvement to both physician diagnosis and patient safety. |
Long Abstract
Introduction
Incidental adrenal tumors are found in approximately 5% of the 80 million CT scans performed in the U.S. each year1. Incidental tumors are described as tumors discovered by chance in CT scans that are performed for a variety of different reasons. Those with indeterminate imaging characteristics will require additional diagnostic workup to exclude adrenal cortical carcinoma (ACC), a rare, but highly aggressive disease. Current testing options might include: further imaging using different modalities such as FDG PET or repeat CT scanning at 3-12 month intervals, provocative hormone testing, adrenal biopsy, and not infrequently adrenalectomy. Given that less than 2,000 of the approximately 250,000 suspicious cases ultimately prove to be ACC, this creates enormous healthcare costs and exposes large numbers of patients to diagnostic procedural risks. Many patients ultimately undergo surgery, which more often than not yields the diagnosis of a benign and typically non-functioning adrenocortical adenoma (ACA). Conversely, ACC discovery may be delayed in some instances due to the lengthy conventional diagnostic process; such delays, adversely affect patient outcome.
To improve this situation, there has been a strong interest in rapid, non-invasive and accurate laboratory tests that may distinguish ACC from ACA. Urine steroid metabolite panels, using gas chromatography- mass spectrometry (GC-MS) 2 have been created showing promising clinical performance in reasonable sized initial studies. These GC-methods, however, suffer from high procedural complexity and relatively low throughput, and limitations imposed by single quadrupole MS in terms of analyte specificity when multiple closely related compounds are monitored, as is the case with steroid profiles. While liquid chromatography tandem mass spectrometry (LC-MS/MS) can potentially address the first two limitations of GC-MS, its limited mass resolution and, compared to GC, inferior chromatographic resolution does not address the specificity issues, in particular for steroids, which tend to fragment into a limited number of product ions in single reaction monitoring experiments.
To address these issues we developed a liquid chromatography-high resolution accurate mass spectrometry (LC-HRAM) clinical assay for the differential diagnosis of ACC vs. ACA, and for ACC follow-up. Twenty-six steroids and steroid metabolites are analyzed in urine of patients with adrenocortical tumors. Liquid chromatographic separation coupled with the high resolution capabilities of the Q-Exactive (QE) allows for unequivocal identification of all twenty-six steroid metabolites while maintaining a high throughput workflow. The clinical validation and implementation of this method should allow for an accurate, rapid, cost-effective means of diagnosis, but more importantly, this non-invasive test can replace the more aggressive diagnostic procedures that are currently in use. Here, we report the analytical performance of this assay.
Methods
Steroid assays are typically designed with optimized hydrolysis conditions for one to three individual analytes; however, our panel is designed to give optimized hydrolysis conditions for twenty-six analytes. A 24 hour urine specimen is incubated with β-Glucuronidase which also contains sulfatase activity. Following hydrolysis, specimens undergo a liquid/liquid extraction using ethyl acetate to separate the steroids from the urine matrix and hydrolysis byproducts. The supernatant is dried down and the residue is reconstituted in mobile phase leaving a purified mixture of the unconjugated neutral steroids. The steroids are separated using an Agilent SB-C18 column with a biphasic gradient (10% Acetonitrile with 0.1% formic acid/90%Acetonitrile with 0.1% formic acid) over a 25 minute run time (series of steps and ramps starting at 98% of the aqueous mobile phase). A Thermo Scientific Ultimate 3000 LC system interfaced with a Thermo Scientific Q Exactive mass spectrometer with a Heated Electron Spray Ionization probe (HESI) is used for full scan HRAM detection of each metabolite in our panel (mass scan range 250-390 m/z). The resolution is set at 70,000 m/z with a mass accuracy set at 10 ppm. Seven calibrators (0 to 5000 ng/mL) and eight internal standards are used to quantify the panel. The entire analytical validation was performed using actual patient urine specimens, in order to ensure the effectiveness of the hydrolysis step. We determined the limit of detection and quantitation for all 26 analytes using the following approach. The lower limit of detection (LOD) or analytical sensitivity is the lowest concentration that can be distinguished from the blank with greater the 95% certainty. It is obtained by replicating (> 20 ) a low concentration specimen that is greater than the limit of the blank. LOD = Limit of the Blank + unidirectional Z-score for 95% probability (=1.65) x SDlow-sample. The Limit of quantitation (LOQ) or functional sensitivity is the lowest concentration that can be reliably measured based on predefined goals of inter-assay accuracy and inter-assay repeatability. 20 different assays were evaluated daily to generate accuracy and precision data, bias and imprecision should be less than 20% coefficient of variation. The LOQ (20 ng/mL) is the lowest concentration that will be reported for each metabolite in this assay, and cannot be lower than the LOD.
Results
The table below lists the analytical performance parameters of our LC-HRAM assay. Data listed as not available reflects analytes with concentrations below our limit of quantitation in the urine samples that were available.
Listed in the following order- Metabolite Name, Abbreviation, Percent Expected Linearity, Percent Recovery, Inter-Precision % CV
Androsterone, An, 94, 96, 4.9
Etiocholanolone, Etio, 96, 96, 3.7
Dehydroepiandrosterone, DHEA, 95, 102, 13.8
11β-Hydroxy-androsterone, 11β-OH-An, 88, 98, 4.9
11β-Hydroxy-etiocholanolone, 11β-OH-Et, 96, 98, 5.6
11-Oxo-etiocholanolone, 11-OXO-Et, 92, 96, 6.9
16á-OH-DHEA, 16á-DHEA, 105, 81, 11.8
Pregnanediol, PD, 93, 89, 4.4
Pregnenediol, 5PD, 119, 87, 12.3
17-OH-pregnanolone, 17HP, 92, 101, 9.2
Pregnanetriol, PT, 94, 94, 7.2
5-Pregnenetriol, 5PT, 111, 100, 6.5
Pregnanetriolone, PTONE, 87, 100, 6.1
Tetrahydrodeoxycorticosterone, TH-DOC, not available, 111, 9.6
Tetrahydro-11-deoxycortisol, THS, 106, 101, 5.5
5á-Tetra-11-dehydrocorticosterone, 5áTHA, not available, 109, 13.2
Tetrahydrocortisone, THE, 87, 114, 8.8
Tetrahydrocorticosterone, THB, 105, 102, 11.6
Tetrahydrocortisol, THF, 91, 100, 5.6
5á-Tetrahydrocortisol, 5áTHF, 90, 110, 4.6
á-Cortolone, á-Cortolone, 83, 105, 5.1
β-Cortolone, β-Cortolone, 91, 103, 4.3
β-Cortol, β-Cortol, 99, 101, 5.0
6β-OH-cortisol, 6β-OH-cortisol, 85, 108, 5.5
Cortisol, Cortisol, 83, 106, 5.8
Cortisone, Cortisone, 91, 118, 6.6
Discussion
The analytical performance of our urinary steroid LC-HRAM panel is as anticipated. In some cases we did not detect certain metabolites in urine samples from normal individuals; however, in ongoing studies of patients with adrenal tumors these analytes can be detected. Normative data/reference ranges and disease cut-offs are currently being generated and analyzed. Based on preliminary studies our intent is to base diagnosis of ACC using steroidal Z-score deviation from the reference range mean. While we anticipate some diagnostic overlap, we have found in our preliminary studies that several metabolites are often 10-fold higher in confirmed ACC patients. Additional studies are being conducted to explore other applications of this panel, such as diagnosis of Cushing"s syndrome, primary aldosteronism, inborn errors of steroidogenesis, and polycystic ovary syndrome, as well as other disease states that might affect steroid production. Not only will these studies produce additional test utilizations, they will also be used to ensure that the representation see in an ACC profile is indeed specific for ACC. Furthermore, we are also in the process of screening urines from individuals with a range of non-adrenal diseases for potential aberrations in steroid metabolite excretion.
Conclusions
We have completed the analytical development and validation of a rapid and accurate LC-HRAM method to measure a 26 steroid metabolite panel in urine. This test is a simple and effective method for diagnostic work-up of patients with adrenal incidentalomas. As studies meet acceptance criteria, this test will significantly improve the differential diagnosis of ACC from ACA, along with follow up care.
References & Acknowledgements:
1.OECD. OECD Stat (database). 2015
2. J Clin Endocrinol Metab. 2011 Dec;96(12):3775-84. doi: 10.1210/jc.2011-1565. Epub 2011 Sep 14
| Description | Y/N | Source |
| Grants | no | |
| Salary | yes | Mayo Clinic |
| Board Member | no | |
| Stock | no | |
| Expenses | no |
IP Royalty: no
| Planning to mention or discuss specific products or technology of the company(ies) listed above: | no |