madman
Super Moderator
Use of calculated free testosterone in men: advantages and... : Current Opinion in Endocrinology, Diabetes and Obesity
hting its relevance in preventing misdiagnosis and overtreatment of male hypogonadism. Recent findings While there is consensus on measuring total T – comprising sex hormone-binding globulin (SHBG)-bound, albumin-bound, and free T – as a first step in diagnosing male hypogonadism, evidence...
journals.lww.com
INTRODUCTION
Clinical practice guidelines on the diagnosis of male hypogonadism focus on clinical signs and symptoms of androgen deficiency, as well as biochemical assessment of low circulating testosterone (T). However, there is a longstanding debate, as well as a persisting controversy concerning biochemical assessment of serum T and in particular the use (and misuse) of free T. Free T, as advocated by the free hormone hypothesis, represents circulating unbound T (2%) and is postulated to be the biologically active fraction of T [1]. In contrast, 98% of circulating T is bound to binding proteins, mainly sex hormone-binding globulin (SHBG) (44%) and albumin (50%), and to a lesser extent cortisol-binding globulin (CBG) (4%), thereby gatekeeping T bioavailability [2,3&&]. An alternative marker of androgen exposure is bioavailable T, in which the loosely albumin-bound fraction of T, together with free T, is also taken into consideration. Debates on the use and suitability of total, free and bioavailable T stem from variations in approaches to measure androgen exposure, leading to differing opinions even among experts [4–6]. Recent clinical practice guidelines from the Endocrine Society, the European Academy of Andrology and the European Society of Endocrinology recommend assessing free T in addition to total T, particularly in patients with conditions that alter serum SHBG levels and in men with borderline low total T concentrations (8– 12 nmol/l) [4,5,7].
On the one hand, the gold standard for total T measurement is liquid chromatography tandem mass spectrometry (LC-MS/MS). The use of this method is particularly important for low-range T measurements, as LC-MS/MS surpasses traditional commercially available total T immunoassays by providing greater sensitivity, accuracy and precision [6,8]. On the other hand, measurement of free T requires equilibrium dialysis coupled LC–MS/MS. Although gold standard and thus recommended, this method is labour-intensive, costly and requires highly trained staff and advanced equipment,which limits its use in routine clinical diagnostics. Alternatively, immunoassays have been developed to directly measure serum-free T. These assays however have significant drawbacks, including inadequate accuracy, bias and dependence on total T rather than free T. Therefore, their use is strongly advised against [3&&].
Due to the complexity of measuring free T, clinicians typically rely on calculators to estimate free T concentrations, using calculated free T (cFT) as a proxy for free T. Using total T and binding protein concentrations as well as historically determined affinity constants, the ‘Vermeulen’ method is currently accepted for use in clinical practice [9]. While the use of cFT is valuable in preventing misdiagnosis and overtreatment of hypogonadism, clinicians should be aware of potential limitations when interpreting cFT values [3&&]. In this review, we will highlight clinical use of cFT, focusing on its significance as a second-line assessment, together with known shortcomings of cFT calculations.
Although physiologically relevant, assessment of free androgen concentrations in women falls beyond the scope of this review.
IS TOTAL T MEASUREMENT SUFFICIENT TO ASSESS ANDROGEN STATUS?
The free hormone hypothesis remains a matter of debate. Experimental data have shown that transgenic mice overexpressing human SHBG exhibit higher total T and lower free T concentrations than wildtype mice lacking circulating SHBG postnatally. Despite high total T levels, these transgenic mice manifested features suggestive of mild androgen deficiency. Conversely, low free T levels were consistent with these features, supporting the view that total T does not always accurately reflect androgen exposure [10].
This is further supported by accruing human data. For instance, in a case study involving a young man with an SHBG mutation causing complete SHBG deficiency, total T levels were low, but freeT levels measured by equilibrium dialysis coupled LC-MS/MS were normal. Despite decreased total T levels, this patient maintained normal pituitary function, with normal luteinizing hormone levels, as well as normal sexual development and secondary sexual characteristics [11].
It has been shown that total T levels do not reliably predict free T levels in men with total T levels near the lower limit of normal. Only when total T exceeds 12 nmol/l, it can consistently predict normal free T levels. Inversely, total T must be below 5.2 nmol/l to reliably indicate low free T levels [12]. Among a sample of 3334 community-dwelling men, 96 out of 416 men with low total T levels had normal cFT, and 277 out of 2918 men with normal total T levels had low cFT. Relying solely on total T levels to diagnose low T thus resulted in a 23.1% false-positive rate (96/416) and a 9.5% false-negative rate (277/2918). Importantly, the group of men with low total T and normal free T did not exhibit features of androgen deficiency [13]. These findings were confirmed in longitudinal analyses. Obese men only developed symptoms suggestive of male hypogonadism when both total T and free T declined below the normal range. On the other hand, when free T remained normal, obese men with low total T did not develop hypogonadal symptoms [14]. Observations within the Veteran Affairs system also showed that obesity and the use of opioids were the strongest predictors to receiving a testosterone prescription [15].
Findings from the recent TRAVERSE Fracture Trial revealed that men with symptoms suggestive of hypogonadism and borderline T levels (<10.4 nmol/l) who received testosterone replacement therapy (TRT) experienced a higher incidence of fractures compared to those receiving a placebo [16]. This outcome is paradoxical, as TRT is generally believed to reduce fractures in men with hypogonadism by enhancing bone density and quality. In this study, most participants were obese and had diabetes, conditions associated with decreased SHBG levels. However, SHBG levels were not considered, as the diagnosis of hypogonadism was based solely on the assessment of total T. Questions have therefore arisen whether the participants were really hypogonadal in the first place, as participants may have had low total T and low SHBG, resulting in normal free T [17&].
This preclinical and clinical evidence reinforces the notion that total T levels may not always accurately reflect androgen exposure, especially when SHBG levels are altered. The argument to interpret T in light of SHBG and free T assessment is reinforced by recent individual participant data meta-analyses (IPDMA) that showed an association between lower T concentrations and higher all-cause mortality when SHBG concentration was normal or high, but not when SHBG was low. These results suggested that higher SHBG concentrations could modulate bioavailability of T [18&&].
Therefore, solely depending on total T may lead to underdiagnosis or overdiagnosis of hypogonadism, particularly in men with decreased or increased SHBG serum levels due to ageing, obesity or type 2 diabetes [13,19].
WHEN IS FREE T ASSESSMENT USEFUL?
Serum SHBG levels increase shortly after birth, remain high during childhood and decline during puberty to reach adult levels. SHBG concentration remains relatively stable until the sixth decade, after which SHBG levels rise gradually [20]. Recent IPD-MA also show an age-associated increase in SHBG. However, this increase appears to occur earlier in adulthood and exhibits a more pronounced magnitude, specifically between the ages of 60 and 70 years [21&&]. As summarized in Table 1, SHBG levels are influenced by various factors. A recent prospective study identified the increasing effect of advancing age (>60 years) on SHBG as a significant factor contributing to erectile dysfunction, a symptom of androgen deficiency, in patients with normal total T levels but low cFT [22]. Moreover, obesity and type 2 diabetes have been found to be negatively associated with SHBG levels [3&&,22]. Men with obesity exhibit lower SHBG levels compared to their nonobese counterparts. SHBG levels can also be reduced in hypothyroidism, exposure to exogenous substances with androgenic properties, as well as in nephrotic syndrome.
Inversely, determinants like liver disease, HIV-infection, hyperthyroidism and antiepileptics intake can lead to increased SHBG [3&&]. Results from recent IPDMA clearly show that conditions like obesity and diabetes influence not only SHBG levels but also total T [21&&]. As shown in Fig. 1, on the one hand, lower SHBG levels lead to lower total T levels, resulting in potential overdiagnosis of hypogonadism, although free T levels are unaltered [23]. On the other hand, increased SHBG and normal total T concentrations could lead to the underdiagnosis of hypogonadism despite low free T. Unlike total T, free T takes SHBG levels into consideration, arguably making it a more reliable indicator of androgen exposure [3&&].
WHAT ARE THE LIMITATIONS OF CALCULATED FREE TESTOSTERONE?
Although free T, if accurately measured, may be physiologically and clinically relevant, the complexity of directly measuring free T limits its introduction into routine clinical practice. Alternatively, clinicians rely on cFT as an acceptable estimate, and thus proxy,of free T concentrations. Existing calculators, including models by Vermeulen, Ly-Handelsman and Zakharov, use calculation methodologies based on total T and SHBG concentrations [3&&]. Free T calculator performance was investigated by comparing cFT values using these three different calculators against measured free T values obtained through the gold standard LC/MS-MS coupled with equilibrium dialysis. The Vermeulen formula appeared to perform best across a wide range of SHBG levels, whereas the Ly-Handelsman model showed significant divergence from measured free T at lower SHBG levels [9]. However, the Vermeulen formula exhibits suboptimal accuracy and tends to overestimate measured free T by 20–30%. Despite this, the current model remains a widely accepted tool for free T calculation due to its ability to integrate a broad range of SHBG, total T and albumin concentrations. This advantage is particularly important in conditions where SHBG concentrations are impacted and/or when total T concentrations are in the borderline range of the lower limit of normal [9].
The use of free T calculators in clinical routine is, however, hindered by a number of imperfections of which clinicians should be aware of when interpreting cFT values (Table 2). Firstly, quality of cFT results depends on the performance of assays used to measure total T, SHBG and albumin. For instance, automated SHBG immunoassays lack standardization[24]. Furthermore, these models are simplified representations of the true binding milieu and may not account for all variables influencing the equilibrium between total and free T, such as SHBG-binding affinity variability and stoichiometry [3&&].
This could particularly be important in men with SHBG polymorphisms. These genetic variations can potentially influence binding affinity between T and SHBG, which is not taken into account in calculators that use a constant binding affinity. In a recent study focusing on the impact of relatively common SHBG single nucleotide polymorphisms (allelic prevalence between 0.5 and 58.2%), healthy men who were heterozygotes for rs6258 had lower serum SHBG levels, while those who were heterozygotes for rs6259, homozygotes for rs727428 and carriers of rs1799941 had higher serum SHBG levels compared to healthy men with wild-type SHBG. These SHBG polymorphisms influenced both SHBG and total T levels, with total T being higher in rs727428 homozygotes and in carriers of rs5934505, rs1799941 and rs6259. Interestingly, these variants did not influence cFT or measured free T concentrations [25&&].
As cFT is a calculated variable, its validity is debated and limited by a lack of standardization and quality control resulting in variable reference ranges. The Vermeulen model, for instance, overestimates free T by 20–30%. Moreover, there is no consensus on a universal cut-off between low and normal cFT values. A thorough review detailing the pitfalls of various methodologies for total and free T assessment was recently published [3&&].
There is a need to enhance the measurement of free T, as cFT values are only approximations. There is also a pressing requirement to reassess current freeT calculators to improve their accuracy and alignment with direct measurement methods. Moreover, additional research is necessary to optimize existing commercially available assays for SHBG, as well as studying SHBG-binding affinity in specific patient groups (e.g. obesity and diabetic individuals) to accurately reflect the true binding environment. Standardizing and validating cFT calculators is also crucial to establish harmonized reference ranges and achieve consensus on cutoff values between low and normal cFT levels. Promising recent developments include the establishment of age-stratified reference ranges for free T in healthy nonobese adult men using the gold standard equilibrium dialysis coupled to LC-MS/MS, showing the expected age-related decline in serum-free T concentrations [26,27]. These efforts represent a significant step towards improving the accuracy of free T measurements and calculations in clinical practice.
HOW TO IMPLEMENT FREE T ASSESSMENT IN CLINICAL PRACTICE?
While there is an international consensus on measuring total T as the initial step in diagnosing male hypogonadism, preclinical and clinical evidence support a comprehensive hormonal assessment that includes total T, SHBG and a method accounting for SHBG alterations, such as free T [14,15,28]. In the latest clinical practice guidelines, most societies recommend assessing free T alongside total T, especially in patients with conditions that alter SHBG levels and in men with borderline low total T concentrations [4,5,7]. The use of free T (measured and cFT) as a second-line hormonal assessment, as recommended by Endocrine Society, the European Academy of Andrology and the European Society of Endocrinology, however faces significant opposition from within the Endocrine Society of Australia [6]. This opposition stems from concerns regarding the limitations of free T calculators. Like other societies, experts in Australia recommend considering SHBG abnormalities by evaluating SHBG levels in conjunction with total T levels. However, Australian guidelines advise against the use of cFT for clinical decision making; instead, they suggest interpreting T levels in the context of SHBG. Where SHBG is low, total T levels may also be low without confirming the presence of androgen deficiency [6].
CONCLUSION
Experimental in-vitro and in-vivo data have provided support for the free hormone hypothesis[10,29]. Accruing clinical data corroborate the notion that free T levels hold greater physiological significance than total T concentrations [10,28]. Furthermore, SHBG gene polymorphisms have been found to affect serum concentrations of both total T and SHBG, but not free T. This evidence highlights the importance of using reliable methodologies that account for SHBG level alterations in diagnosing male hypogonadism, such as free T, and reinforces current guidelines.
Due to the complexity of measuring free T, clinicians typically rely on calculators to estimate its concentrations. The widely accepted Vermeulen method for free T calculation accepts a broad concentration range of SHBG, total T and albumin. This is particularly important in conditions where SHBG concentrations are affected, or in men with borderline total T concentrations, which is common in men with obesity and type 2 diabetes. Using free T as an additional assessment in obese men with low total T can preventover diagnosis and overtreatment with TRT.
While the use of cFT may be of value in preventing misdiagnosis and overtreatment of hypogonadism, it has its limitations. Therefore, reassessing cFT calculators to enhance their accuracy and alignment with equilibrium dialysis measurement of free T is needed. Additionally, standardizing and validating cFT calculators, as well as optimizing available assays for total T and SHBG are crucial steps.
