MSACL 2023 Abstract

Background. Androgen concentrations are measured in females to investigate conditions associated with hyper-/hypo-androgenism. The most commonly performed tests include total testosterone (Te), sex hormone binding globulin (SHBG) and free Te (FTe). Concentrations of Te or FTe outside of the age-specific reference intervals are considered key diagnostic features of biochemical hypo/hyper-androgenism. We compared distributions of Te, SHBG and FTe concentrations in association with various health conditions observed in post-menopausal females (PMF) and assessed diagnostic utility of the tests.
Methods. We reviewed data on Te, SHBG and FTe concentrations measured in consecutive routine serum samples from PMF (n=7,022), for which International Classification of Diseases (ICD-10) codes were provided with the test request. We evaluated a subset of specimens (age range 55-85 years; mean/median 65/63), where the ICD-10 codes corresponded to healthy PMF (medical examination without abnormal findings, n=194), and PMF with clinical conditions known to be associated with androgen excess or deficiency (n=3,365). Te was measured using LC-MS/MS method, reference interval (RI), 5-32 ng/dL; SHBG was measured using immunoassay (Roche Cobas e602, RI 17-125 nmol/L); concentrations of FTe (RI 0.6-3.8 pg/mL) were calculated (Vermeulen, JCEM 1999) using measured concentrations of Te, SHBG and albumin.
Results. Among the study samples, the most common conditions potentially associated with androgen status were decreased libido, drug therapy, hirsutism, hyperlipidaemia, hypothyroidism, malaise, menopause symptoms, nonscarring hair loss (NHL) and vitamin D deficiency. Mean/ median concentrations of Te, SHBG and FTe in samples of healthy and pathologic PMF were 32/21 and 50/23 ng/dL for Te; 81/103 and 78/68 nmol/L for SHBG; and 3.6/2.3 and 5.9/2.4 pg/mL for FTe. Statistically significant differences between the distributions of concentrations of Te in pathologic vs healthy PMF were in hyperlipidaemia, menopause symptoms and NHL. Statistically significant differences between the distributions of concentrations of SHBG in pathologic vs healthy PMF were in hirsutism and NHL. Statistically significant differences between the distributions of concentrations of FTe in pathologic vs healthy PMF were in hyperlipidaemia, menopause symptoms, hirsutism, and NHL. Of the groups with statistically significant differences compared to healthy women, lower concentrations were observed in PMF with NHL for Te; hirsutism and NHL for SHBG; and NHL for FTe. In the remaining groups, concentrations of Te and FTe were higher than in healthy PMF, while concentrations of SHBG were comparable to healthy PMF. To assess diagnostic utility of the three evaluated biomarkers for the ability to provide distinction between samples of healthy and pathologic PMF, we determined cut off concentrations which provide the best clinical specificity and sensitivity to differentiate samples from women with pathologies from samples of healthy PMF. Receiver operator characteristic analysis of the distributions of concentrations of Te, SHBG and FTe between the pathologic groups and healthy PMF, demonstrated that the disease-specific cut off concentrations may provide better diagnostic utility as compared to the RIs established using samples of healthy PMF.
Conclusions. We observed statistically significant differences in distributions of Te and FTe concentrations in specimens from healthy PMF and PMF with hyperlipidaemia, menopause symptoms and NHL (in addition, the distribution of FTe concentrations was different in PMF with hirsutism). Statistically significant differences in distributions of SHBG were observed in healthy PMF as compared to women with hirsutism and NHL. Among the evaluated biomarkers comparable clinical sensitivity and specificity were observed for Te and FTe (with exception of NHL, where Te provided greater specificity than FTe); greater clinical sensitivity was observed for SHBG than for Te and FTe in PMF with decreased libido, hirsutism and NHL. Our data suggest that measurement of FTe could provide additional diagnostic value as compared to measurement of Te or SHBG alone in associating androgens status with diagnosis of hirsutism.