A-056 Data Driven Approach to Ensure Accurate Total Testosterone Results in Consumer Initiated, Home-collected Capillary Samples

• Published in: Clinical chemistry (Baltimore, Md.) Vol. 69; no. Supplement_1
• Main Authors: Naranjo, A R, Carter, C, Chen, J, Hamdan, A, Krasowski, M D, O’Connor, K, Greene, D N
• Format: Journal Article
• Language: English
• Published: 27-09-2023
• ISSN: 0009-9147; 1530-8561
• DOI: 10.1093/clinchem/hvad097.054

Abstract
Background
Home-collected direct access testing (DAT) has become increasingly attractive as telehealth utilization advances. With DAT, consumers initiate their lab order and have access to clinical consulting staff. Steroid hormone concentrations are uniquely positioned for DAT due to their inherent stability and non-acute indications. Total testosterone (TT) is used to screen for or monitor treatment of hypogonadism. Once diagnosed, testosterone is administered by intramuscular or subcutaneous injection or via transdermal patch or topical gel. TT in adult men is typically between 200 and 900 ng/dL. However, in our DAT lab we were encountering a significant percentage of samples with supraphysiological TT concentrations (>1500 ng/dL). The objective of this study was to design a quality assurance workflow to differentiate whether the supraphysiological results were accurate or erroneous.
Methods
TT was measured using the Roche Cobas e801 immunoassay; basic assay parameters were verified per general CAP guidelines (precision, accuracy, linearity, method comparison, interferences, carryover, dilution). Additional validation experiments were performed to verify that capillary plasma was an acceptable sample type (n = 30 paired capillary/venous samples) and to determine if extreme summer temperature fluctuations would influence stability (n = 46 participants; samples subjected to temperature cycling for 72 h before re-evaluation). Self-collected capillary samples received for clinical testing with a TT concentration >1500 ng/dL were run on a 1:5 dilution (extending the clinical reportable range to 7500 ng/dL) and the sample was reflexed to luteinizing hormone (LH) and follicle stimulating hormone (FSH). Reference intervals in adult men for LH and FSH are 1.7–8.6 mIU/mL and 1.4–13.0 mIU/mL, respectively. When available, clinical notes were reviewed for additional information on exogenous hormone usage.
Results
All general assay parameters were within acceptable metrics. TT concentrations between capillary and venous samples had an average absolute bias of −9.2 ng/dL or −3.8% (n = 30; range = <20–823 ng/dL; y = 0.954x + 1.915; R = 0.995). Baseline and post-summer temperature cycling TT concentrations had an average absolute bias of −8.1ng/dL or −3.4% (n = 46; range = <20–716 ng/dL; y = 0.934x + 7.728; R = 0.999). Samples (n = 506) with a TT concentration >1500 ng/dL were reflexed to TT dilution and LH/FSH. Most (n = 413; 82%) had undetectable (<0.3 mIU/mL) LH/FSH and a mean TT concentration of 3356 ng/dL (SD = 1972). For 18.6% (n = 77) of this subset, we had medication information; 68 were using injectable testosterone. Elevated TT was likely a true result caused by overuse of anabolic steroids. A smaller subset of samples (n = 21) had detectable LH/FSH with mean concentration/SD of 6.1/5.2 mIU/mL for LH, 5.3/5.4 mIU/mL for FSH, and TT >7500 ng/dL. Medication notes were available for 5 of these samples and indicated that the TT results were likely caused by topical contamination (false positive). A final subset (n = 40) had detectable LH/FSH and TT between 1500 and 7500 ng/dL. These samples could not be binned into simple categories and instead were sent to the lab director for review.
Conclusion
Self-collected capillary specimens are acceptable for TT measurements. A quality assurance reflex to LH/FSH can support or refute the validity of supraphysiological results in a consumer initiated/DAT population.