madman
Super Moderator
Diagnosis and Evaluation of Hypogonadism (2022)
Alvin M. Matsumoto, MD
INTRODUCTION
Over the last 20 years, large database surveys of several countries reported consistent two- to four-fold increases in the rates of testosterone testing and prescriptions from 2000 to 2014.1 Increases in prescription rates were temporally associated with the availability of transdermal testosterone formulations and direct-to-consumer advertising of testosterone treatment of age-related symptoms associated with low serum testosterone concentrations (“low T”). Following initial observational studies that reported possible increased cardiovascular events associated with testosterone treatment (although not confirmed in subsequent studies) and Food and Drug Administration warnings, testosterone prescription rates declined in 2014 but remained higher than in 2000.1,2 These findings suggested considerable overtesting for testosterone that likely resulted in an increased finding of low testosterone concentrations and subsequent overtreatment with testosterone.
The primary indication for testosterone treatment is a diagnosis of hypogonadism. According to evidence-based Endocrine Society clinical practice guidelines (published initially in 2006 and updated most recently in 2018), hypogonadism is defined as symptoms or signs of testosterone deficiency and consistently low serum testosterone concentrations on at least two occasions.3 However, numerous studies found that testosterone therapy was often initiated in the absence of a diagnosis of hypogonadism as specified by guideline recommendations.1 In these studies, a minority (10%–40%) of men had at least two testosterone measurements and 12% to 46% had no testosterone levels measured in the year before initiating testosterone therapy.
To avoid an inappropriate testosterone treatment, an accurate diagnosis of hypogonadism is imperative. However, for many practitioners, the nonspecific clinical manifestations of testosterone deficiency and complexities associated with serum testosterone measurements present challenges in making a proper diagnosis of hypogonadism. This review highlights these challenges and provides a systematic and practical approach to making a rigorous guideline-based diagnosis and evaluation of hypogonadism.
APPROACH TO THE DIAGNOSIS AND EVALUATION OF HYPOGONADISM
A structured approach is necessary to make a reliable diagnosis of hypogonadism and identify men who are the most appropriate candidates for testosterone replacement and the most likely to respond. The following is a summary of the orderly diagnostic strategy that is recommended to diagnose hypogonadism (Fig. 1):
*Establish the presence of symptoms and signs of testosterone deficiency.
*Consider other causes of largely nonspecific symptoms and signs that might be managed without testosterone treatment.
*Delay laboratory testing for testosterone until conditions that transiently suppress serum testosterone concentrations are completely resolved.
*Confirm the diagnosis of hypogonadism by measuring serum total testosterone in the morning after an overnight fast on at least 2 separate days, using an accurate and reliable standardized assay.
*Measure serum-free testosterone using an accurate and reliable method in patients who have conditions that alter sex hormone-binding globulin (SHBG) or if serum total testosterone concentrations are near the lower limit of the normal adult male reference range.
*In men found to have hypogonadism, measure serum luteinizing hormone (LH) and follicle-stimulating hormone (FSH) concentrations to distinguish primary versus secondary hypogonadism.
*Establish the specific cause of hypogonadism and determine whether the cause of hypogonadism is a potentially reversible or treatable (functional hypogonadism) or an irreversible congenital or destructive disorder (organic hypogonadism) to guide management.
PRESENCE OF CLINICAL MANIFESTATIONS OF TESTOSTERONE DEFICIENCY
Hypogonadism should be suspected in a man who has symptoms and signs of testosterone deficiency.3 Therefore, before measuring serum testosterone and pursuing a diagnosis of hypogonadism, it is important to assess whether a man has clinical manifestations of testosterone deficiency. However, the symptoms and signs of testosterone are mostly nonspecific and occur commonly, especially in older men with multiple comorbidities. The symptoms and signs of testosterone deficiency are broadly classified as sexual, physical, and psychological manifestations (Table 1).
Clinical manifestations depend on the severity and duration of testosterone deficiency. Although uncommon, many individuals with congenital or structural disorders of the testes (eg, Klinefelter syndrome), or hypothalamus or pituitary gland (eg, congenital hypogonadotropic hypogonadism, androgen deprivation therapy in older men with prostate cancer) who have severe testosterone deficiency demonstrate most of the symptoms and signs summarized in Table 1.
Although most symptoms and signs of testosterone deficiency are nonspecific, eunuchoidism (inadequate sexual development), decreased androgen-dependent hair, and very small testes are specific signs that are characteristic of severe testosterone deficiency.4 A clinical finding that is pathognomonic of severe prepubertal testosterone deficiency is eunuchoidism, which is characterized by extremely small penis and testes (prepubertal testis volume <2–4 mL), poorly developed scrotum, lack of androgen-dependent hair pattern (facial, axillary, chest, and pubic hair), a high-pitched voice, and a eunuchoidal body habitus (poor upper body muscle mass; prepubertal fat distribution in hips, chest, and face; and arms and legs >5 cm longer than height). Although usually detected in boys who present with delayed puberty, eunuchoidism may also be found in adult males who are not diagnosed at an earlier age. In men, decreased or loss of androgen-dependent hair is a specific sign of long-standing, severe testosterone deficiency. Decreased testes size is an indicator of impaired spermatogenesis because seminiferous tubules comprise 80% to 90% of the testes volume. However, very small testes (eg, <6 mL) are characteristic of Klinefelter syndrome, which is associated with impairments of testosterone and sperm production.
In contrast to these manifestations, other sexual, physical, and psychological symptoms and signs of testosterone deficiency are nonspecific and may be caused by coexisting comorbidities, illnesses, or medications. In addition to the severity and duration of testosterone deficiency, clinical manifestations of hypogonadism may be modified by previous testosterone therapy, age, comorbid conditions, medications, and variations in target-organ androgen sensitivity, all of which contribute to variability in clinical presentation.
Sexual symptoms, especially reduced libido (sexual interest or desire), loss of spontaneous (nighttime and morning) erections, erectile dysfunction, and reduced sexual activity, are common symptoms of testosterone deficiency that are responsive to testosterone treatment. In the middle-aged and older male participants of the European Male Aging Study (EMAS), sexual symptoms (decreased frequency of sexual thoughts, erectile dysfunction, and decreased frequency of morning erections) demonstrated a stronger syndromic association with low serum testosterone concentrations (total testosterone <230–320 ng/dL [8–11 nmol/L] and free testosterone <64 pg/mL [220 pmol/L]) than physical or behavioral symptoms.5 In healthy older men (aged 60–80 years) with experimental hypogonadism (induced by gonadotropin-releasing hormone [GnRH] agonist administration) who were treated with various doses of testosterone, sexual desire, and erectile dysfunction were progressively decreased at serum testosterone concentrations less than 300 ng/dL (10.4 nmol/L), but significantly only at testosterone level less than 100 ng/dL (3.5 nmol/L), compared with placebo-treated men.6 A meta-analysis of randomized, placebo-controlled clinical trials of testosterone treatment in men who had a morning total testosterone less than or equal to 300 ng/dL (10.4 nmol/L) and at least one symptom or sign of testosterone deficiency (a more rigorous diagnosis of hypogonadism than used in previous meta-analyses) found that testosterone treatment for 3 months or longer increased sexual desire, erectile function, and sexual satisfaction, but had no consistent significant effect on energy or mood.7
CONSIDER OTHER CAUSES OF SYMPTOMS AND SIGNS
Because the clinical manifestations of hypogonadism are nonspecific, it is important to consider the differential diagnosis or contribution of other causes of symptoms and signs to identify those that could be managed and treated independently of testosterone therapy.4,8 Comorbid illnesses and conditions (eg, depression), and medications may contribute significantly to symptoms and signs consistent with testosterone deficiency. Consideration of other causes of symptoms and signs is particularly relevant for a patient who has an isolated or predominant symptom or signs consistent with testosterone deficiency. For example, in an older man who presents with low bone mineral density, potential contributing causes in addition to hypogonadism might include immobility or reduced activity, smoking, excessive alcohol intake, low calcium intake, vitamin D deficiency, medications (eg, glucocorticoids, anticonvulsants), and chronic kidney disease. Of note, in healthy older men with experimental hypogonadism induced by GnRH agonist treated with various doses of testosterone, L4 trabecular bone mineral density by computed tomography was only decreased in men who had serum testosterone levels less than 200 ng/dL (6.9 nmol/L), compared with placebo-treated men.9
DELAY LABORATORY TESTING UNTIL CONDITIONS THAT TRANSIENTLY SUPPRESS TESTOSTERONE ARE RESOLVED
Acute illness or surgery, use of certain medications (eg, opioids or glucocorticoids), and nutritional deficiency that causes an energy deficit (eg, eating disorders, malnutrition, or excessive exercise associated with inadequate energy intake) can transiently suppress gonadotropin and testosterone production.3,4 Therefore, measurements should be delayed until these conditions are completely resolved to avoid an incorrect diagnosis of testosterone deficiency and hypogonadism and subsequent inappropriate testosterone therapy.
REPEATED FASTING MORNING TESTOSTERONE MEASUREMENTS USING AN ACCURATE ASSAY TO CONFIRM HYPOGONADISM
To confirm a diagnosis of hypogonadism in men with symptoms and signs of testosterone deficiency, serum total testosterone concentrations should be measured in the morning (eg, 7:00–10:00 AM) samples after overnight fasting, and measurements should be repeated on at least 2 separate days.3 For the most rigorous and accurate diagnosis of hypogonadism, an accurate and reliable total testosterone assay should be used, preferably an assay that is standardized and certified by an accuracy-based quality control program (eg, Centers for Disease Control and Prevention [CDC] Hormone Standardization Program).
Serum testosterone levels exhibit a circadian variation with higher values in the morning that is blunted but still present in older compared with young men.10 More than half of the men older than 65 years of age who had low testosterone in the afternoon were found to have normal levels measured in the morning. Testosterone concentrations also demonstrate substantial day-to-day variability. Approximately 30% to 35% of men who had a single low testosterone level on initial sampling were subsequently found to have a normal testosterone level on repeat sampling.11 In community-dwelling middle-aged to older men who had single initial serum testosterone less than 250 ng/dL (8.7 nmol/L), 20% had average testosterone greater than 300 ng/ dL (10.4 nmol/L) on repeated sampling over the subsequent 6 months; a single testosterone measurement was inadequate to define an individual’s serum concentrations.12 In contrast, no men who had initial average testosterone less than 250 ng/dL on two or more samples drawn on separate days had average testosterone greater than 300 ng/dL on repeated sampling over the subsequent 6 months. Finally, testosterone levels are suppressed by glucose administration and food intake.13,14 These findings provide a strong scientific rationale for using a standardized approach of repeated blood sampling in the morning after an overnight fast to reduce the influence of biologic variability of serum testosterone measurements that contributes to inaccurate diagnosis of hypogonadism.
Currently, most testosterone measurements are not standardized and performed by automated platform-based immunoassays and few are performed by mass spectrometry-based assays, which are more accurate but may not be standardized. As a result, testosterone assays exhibit considerable assay-to-assay variation in the values measured. A College of American Pathologists peer-based quality control sample from a hypogonadal man measured in 1133 laboratories using 14 different assays reported testosterone values that ranged from 45 to 365 ng/dL (1.6–12.7 nmol/L), that is, values that ranged from clearly hypogonadal to eugonadal concentrations.15 In contrast, the same sample measured in five different mass spectrometry-based assays ranged from 60 to 72 ng/dL (2.1–2.5 nmol/L).15 This extreme interassay and interlaboratory variability is caused by the lack of standardization of most total testosterone assays. As a result, measured values and reference ranges for total testosterone differ from assay-to-assay and laboratory-to-laboratory. The extreme variability in reference ranges is highlighted in two recent surveys of laboratories in the United States and the United Kingdom. In 120 laboratories in the United States, the lower limit of the reference range ranged from 160 to 300 ng/dL (5.5–10.4 nmol/L) with the lower limit in 50% of laboratories less than or equal to 241 ng/dL (8.4 nmol/L).16 In 60 laboratories in the United Kingdom, the lower limit of the reference ranges ranged 141 to 317 ng/dL (4.9–11 nmol/ L) with the lower limit in 50% of laboratories less than 231 ng/dL (8.0 nmol/L).17
Clearly, the use of a uniform lower limit of normal, such as total testosterone less than 300 ng/dL (10.4 nmol/L) as used by some clinical practice guidelines,18 is inappropriate for all total testosterone assays and results in an increased likelihood of overdiagnosis or underdiagnosis of hypogonadism. The Endocrine Society clinical practice guideline recommends the use of a standardized testosterone assay, such as a CDC-certified assay or similar accuracy-based assay standardization program.3 CDC-certified total testosterone assays (mostly mass spectrometry and some immunoassays) are listed and regularly updated at https://www.cdc.gov/labstandards/hs_ certified_participants.html. Most major commercial reference laboratories offer CDC-certified total testosterone assays. However, clinicians should be careful when ordering because some reference laboratories offer CDC-certified and non-CDC certified total testosterone assays.
A harmonized reference range that cross-calibrated testosterone values to CDC reference standards were established in 9054 community-dwelling men.19 The harmonized reference range for young (18–39-year-old), healthy, nonobese men was 264 to 916 ng/dL (9.2–31.8 nmol/L). This range is used as a uniform reference range for all CDC-certified total testosterone assays. If nonstandardized assays are used, total testosterone concentrations and reference ranges vary greatly depending on the specific assay and may not accurately identify men with hypogonadism.
Consistent use of accurate and reliable CDC-certified standardized assays in clinical trials of hypogonadism (eg, The Testosterone Trials20), clinical practice guidelines (eg, Endocrine Society guidelines3 ), and endocrine practice will facilitate translation of research findings to clinical practice. Standardized testosterone assays will also prevent an inaccurate diagnosis of hypogonadism with subsequent inappropriate testosterone treatment and potential adverse events and increased health care costs associated with treatment.
INDICATIONS FOR ACCURATE AND RELIABLE FREE TESTOSTERONE MEASUREMENTS
Serum total testosterone is affected by alterations in SHBG concentrations (Table 2) and the high level of assay imprecision in the total testosterone immunoassays in the low range increases the risk of misdiagnosis when the measured total testosterone concentrations are at or slightly higher or lower than the lower limit of the normal adult male reference range. Therefore, in men being evaluated for hypogonadism who have conditions that alter SHBG or serum total testosterone concentrations that are moderately higher or lower than the lower limit of the normal adult male reference range (eg, 200–400 ng/dL [6.9–13.9 nmol/L]), an accurate measurement of free testosterone (free testosterone by equilibrium dialysis or calculated free testosterone estimate) should be used to confirm testosterone deficiency.3
In circulation, testosterone is mostly bound to serum proteins, primarily tightly bound to SHBG with high affinity and loosely bound to albumin with low affinity, and only 2% to 4% circulating testosterone is unbound to proteins, that is, free testosterone.21 Free testosterone and loosely albumin-bound testosterone are referred to as bioavailable testosterone. According to the free hormone hypothesis, free testosterone is the biologically active circulating fraction and testosterone loosely bound to albumin can dissociate in capillaries and become potentially biologically available in some tissues with long capillary transit times (eg, liver and brain).
Serum total testosterone assays measure protein-bound and free testosterone. Therefore, conditions associated with alterations in SHBG concentrations (not necessarily outside the normal reference range) affect total testosterone levels in the same direction; conditions associated with low SHBG levels (eg, obesity) result in low total testosterone and conditions associated with high SHBG levels (eg, advanced old age) result in high total testosterone concentrations. Because free testosterone is the biologically active fraction that is regulated by negative feedback control of gonadotropin secretion, abnormalities in total testosterone caused solely by alterations of SHBG levels are typically not associated with abnormalities of free testosterone levels.
The importance of free testosterone levels on symptoms of testosterone deficiency is supported by findings in middle-aged to older men in EMAS.22 Compared with eugonadal men who had normal total and free testosterone (303 ng/dL [10.5 nmol/L] and 65 pg/mL [226 pmol/L], respectively), men who had low total testosterone but normal free testosterone levels were more obese, had lower SHBG, and lacked sexual or physical symptoms of testosterone deficiency; in contrast, men with normal total testosterone but low free testosterone concentrations were older and in poorer health, and reported sexual and physical symptoms of testosterone deficiency. In a subsequent prospective study, eugonadal men in EMAS who had normal total (303 ng/dL [10.5 nmol/L]) and free testosterone (49 pg/mL [170 nmol/L]) and normal LH (>9.4 IU/L) were followed for a median of 4.3 years.23 At follow-up, most (93.2%) remained persistently eugonadal, but based on total testosterone concentrations, 6.8% developed apparent secondary hypogonadism with low total testosterone and normal LH levels. However, of the men who had low total testosterone concentrations, only those who also had low free testosterone levels developed or had worsening of sexual symptoms of testosterone deficiency (decreased sexual thoughts, erectile dysfunction, and decreased morning erections), compared with those with low total testosterone but normal free testosterone concentrations and eugonadal men who remained persistently free of sexual symptoms. Finally, a man who demonstrated undetectable SHBG because of a homozygous missense mutation of SHBG and very low serum total, but normal free testosterone was found to have normal serum gonadotropins, semen analysis, and sexual development, supporting the importance of free rather than total testosterone on these sensitive objective indicators of testosterone action.24
In a large cohort of men (3672 male Veterans; mean age, 59.7 years) who had laboratory evaluation with a panel comprised of total testosterone, SHBG, and albumin measurements and calculated free testosterone, 61.7% of men with low total testosterone had normal calculated free testosterone (<34 pg/mL [118 pmol/L]), whereas only 38.3% of those with low total testosterone had low calculated free testosterone; 2.1% of men with normal total testosterone had low calculated free testosterone.25 These results suggest that reliance only on total testosterone concentrations could result in considerable overdiagnosis of testosterone deficiency and hypogonadism. In these men, low free testosterone could be excluded reliably only when total testosterone exceeded 350 to 400 ng/dL (12.1–13.9 nmol/L) and low free testosterone could only be reliably predicted when total testosterone was less than 150 to 200 ng/dL (5.2– 6.9 nmol/L). Therefore, measurement of free testosterone in men with total testosterone concentrations 200 to 400 ng/dL (6.9–13.9 nmol/L) should improve the accuracy of biochemical evaluation of hypogonadism. In men with very low total testosterone levels (eg, <150 ng/dL [5.2 nmol/L]), free testosterone measurements are not needed because the likelihood of finding a normal free testosterone is extremely low.
Accurate and reliable methods that are available to measure serum free testosterone concentrations include equilibrium dialysis and calculated free testosterone.3,21 Preferably, free testosterone should be measured by an equilibrium dialysis method, the gold standard method. If equilibrium dialysis is not accessible, clinically useful accurate estimates of free testosterone relative to equilibrium dialysis are calculated using measurements of total testosterone, SHBG, and albumin concentrations and various published formulae that use algorithms based on the binding affinity of testosterone to SHBG and albumin. However, it is important to recognize that the accuracy of calculated free testosterone values depends on the accuracy of total testosterone, SHBG, and albumin assays. Therefore, calculated free testosterone values should be cross-calibrated against those measured using equilibrium dialysis. Recent studies suggest that SHBG circulates as a dimer with allosterically coupled binding sites on each of the two monomers; the binding of testosterone to SHBG is a multistep process that involves an allosteric interaction between the two binding sites.21,26 Calculated free testosterone estimates using an ensemble allosteric model yielded free testosterone concentrations that closely approximated those measured by equilibrium dialysis.
Accurate and reliable free testosterone assays are not available in most local laboratories. Therefore, free testosterone by equilibrium dialysis and calculated free testosterone should be measured in a dependable reference laboratory. Limitations of using free testosterone by equilibrium dialysis and calculated free testosterone concentrations in practice are the lack of assay standardization, an accuracy-based quality control program, and a harmonized reference range. Until these limitations are addressed, free testosterone by equilibrium dialysis and calculated free testosterone should use reference ranges established by individual laboratories or their specific assay method.
Many local laboratories and some reference laboratories still measure direct free testosterone levels by a testosterone tracer analog immunoassay on an automated assay platform. Free testosterone immunoassays are inaccurate, resulting in values that are an order of magnitude lower than free testosterone by equilibrium dialysis and calculated free testosterone and they should not be used to evaluate men for hypogonadism.21 Bioavailable testosterone is measured by an ammonium sulfate precipitation method, which is technically demanding or calculated from total testosterone, SHBG, and albumin measurements using the same algorithms that are used for calculating free testosterone. The major limitation of using bioavailable testosterone concentrations for clinical evaluation is the relative lack of clinical studies of testosterone deficiency and hypogonadism using bioavailable compared with those using free testosterone levels.
MEASURE GONADOTROPIN CONCENTRATIONS TO DISTINGUISH PRIMARY VERSUS SECONDARY HYPOGONADISM
If a diagnosis of hypogonadism is confirmed, serum gonadotropin, LH, and FSH concentrations should be measured to determine whether the origin of hypogonadism is a disorder of the testes (primary hypogonadism), or pituitary or hypothalamus (secondary hypogonadism).3,4 Serum LH and FSH should be measured in the same sample as testosterone, usually together with a repeat testosterone measurement after an initial low testosterone level or less commonly with an initial testosterone measurement.
Men with primary hypogonadism exhibit repeatedly low testosterone with simultaneously high LH and FSH concentrations (FSH typically being higher than LH). High LH levels indicate reduced testosterone negative feedback and production by Leydig cells of the testes. High FSH levels indicate seminiferous tubule dysfunction (reflecting reduced inhibin B negative feedback) and impaired sperm production but is a more sensitive indicator of primary testicular dysfunction than high LH levels. If high LH and FSH are measured in the same sample as an initial testosterone measurement, men with normal serum testosterone with high LH and/or FSH concentrations might be identified. These men have mild or subclinical primary hypogonadism (also called compensated hypogonadism), analogous to subclinical hypothyroidism.
Men with secondary hypogonadism demonstrate repeatedly low testosterone with simultaneously low or inappropriately normal LH and FSH levels. Some causes of hypogonadism are associated with defects in the testes and pituitary or hypothalamus, which is combined primary and secondary hypogonadism. However, in most cases, a hormone profile of either primary or secondary hypogonadism predominates. For example, in men with hemochromatosis, iron overload results in defects in the testes and pituitary but the latter is the dominant defect that results in gonadotropin deficiency and a hormone profile of low serum testosterone and low gonadotropin concentrations consistent with secondary hypogonadism.27
Serum LH and FSH measurements are usually performed by automated platform-based immunoassays. Most LH and FSH assays have sufficient sensitivity to distinguish low-normal from low values but are susceptible to immunoassay interference (eg, by high-dose biotin use). Although gonadotropin assays are not standardized, differences in values and reference ranges are small. The reference ranges for serum LH and FSH concentrations in well-characterized, healthy fertile young men are 1.6 to 8.0 IU/L and 1.3 to 8.4 IU/L, respectively.28 Assays with upper limits of reference range higher LH and FSH levels probably included older men or men with unrecognized impairment of spermatogenesis.
Distinguishing whether a patient has primary from secondary hypogonadism is clinically important.3,4 Secondary hypogonadism is caused by a pituitary or hypothalamic tumor that might result in deficiency of or be associated with hypersecretion of other pituitary hormones and space-occupying tumor mass effects (eg, headaches, visual field defects, hydrocephalus, or cerebrospinal fluid rhinorrhea) that may require further management. Also, secondary hypogonadism is commonly caused by potentially reversible gonadotropin suppression in the presence of a functionally intact hypothalamic-pituitary-testicular axis (ie, functional hypogonadism), whereas causes of primary hypogonadism are usually caused by irreversible pathologic disease (ie, organic hypogonadism). Finally, infertility caused by impaired spermatogenesis caused by gonadotropin deficiency in secondary hypogonadism is treatable with gonadotropin (or GnRH)-replacement therapy. In contrast, infertility caused by primary hypogonadism is usually not treatable with hormone therapy and requires other fertility options (eg, assisted reproductive technologies).
It is essential that LH and FSH are measured before initiating testosterone therapy. However, it is not uncommon that testosterone therapy is started without the measurement of gonadotropin levels. In this situation, testosterone should be discontinued for at least 2 to 4 weeks for short-acting (transdermal, oral, transbuccal), 2 to 3 months for intermediate-acting (intramuscular testosterone cypionate or enanthate), and 6 to 12 months for long-acting (intramuscular testosterone undecanoate and testosterone pellets) before measuring testosterone and gonadotropins. However, even with more prolonged discontinuation of testosterone therapy, some men may experience persistent gonadotropin and testosterone suppression, especially older men and those receiving high dosages of exogenous testosterone for long periods of time.
FURTHER EVALUATION OF SPECIFIC CAUSE OF FUNCTIONAL VERSUS ORGANIC HYPOGONADISM
After determining whether a patient has primary or secondary hypogonadism, further evaluation should be performed to establish the specific cause of hypogonadism to guide further management, including the need for testosterone therapy.3,4
It is important to determine whether a man with hypogonadism has organic hypogonadism or functional hypogonadism to guide management (Table 3).3,4,8 Organic (also known as “classical”) hypogonadism is caused by irreversible structural, destructive, infiltrative, developmental, or congenital disorders of the reproductive axis. Generally, it results in severe symptoms and signs of testosterone deficiency and consistently and severely low serum testosterone and high (primary organic hypogonadism) or distinctly low (secondary organic hypogonadism) serum LH and FSH concentrations for which testosterone treatment is indicated. In contrast, functional hypogonadism is caused by potentially reversible or treatable testosterone or gonadotropin suppression. Functional hypogonadism is more common than organic hypogonadism and it usually results in mild symptoms and signs of testosterone deficiency and slightly low serum testosterone and slightly high (functional primary hypogonadism) or normal to low-normal (functional secondary hypogonadism) LH and FSH levels. Some causes of functional secondary hypogonadism, such as long-acting opioid use, result in severe clinical manifestations of testosterone deficiency and severely low serum testosterone, LH, and FSH concentrations. Management of functional hypogonadism should initially focus on the treatment of the underlying causative condition (eg, weight loss for obesity) or discontinuation of the offending medication (eg, glucocorticoids) rather than testosterone treatment. However, in men who have severe functional hypogonadism that is not readily reversible or treatable (eg, men taking methadone for opioid use disorder), testosterone treatment could be considered after acknowledgment and discussion of the inadequacy of high-quality evidence for potential benefits and risks of testosterone treatment.
Age-related hypogonadism in middle-aged to older men is mostly related to functional secondary hypogonadism and gonadotropin suppression caused by age-associated comorbidities (eg, obesity, illness, and use of medications). However, advanced age men (eg, >75 years of age) develop organic primary hypogonadism and testicular failure with elevated gonadotropin levels and reduced testicular responsiveness to LH and human chorionic gonadotropin.29
Evaluation to identify the specific cause of primary or secondary hypogonadism should begin with a careful history and physical examination.4 In men with primary hypogonadism, evaluation should include inquiry about a history of undescended testes; mumps with testicular involvement; testes damage, torsion, or surgery; medications that reduce testosterone production (eg, alkylating agents); and end-stage kidney disease. In men with secondary hypogonadism, assessment should include questioning regarding a history of delayed puberty; anosmia or hyposmia; tumor mass symptoms (eg, headache, peripheral vision loss); hypothalamic/pituitary disease or surgery; trauma brain injury; medications that suppress gonadotropin secretion (eg, long-acting opioids, anabolic steroids, glucocorticoids); morbid obesity; reduced energy intake; excessive exercise; wasting syndromes; alcohol use disorder; type 2 diabetes mellitus; and chronic liver, heart, or lung failure.
If no cause is apparent in a man with primary hypogonadism and very small testes less than 6 mL (normal testis volume is 15–30 mL), a karyotype should be ordered to diagnose Klinefelter syndrome.3 In men with secondary hypogonadism, initial laboratory evaluation should include serum prolactin (to exclude hyperprolactinemia) and iron saturation (to screen for iron overload syndromes, such as hemochromatosis). If there is clinical evidence of hypopituitarism or unprovoked polyuria, assessment of other pituitary or hypothalamic hormones (eg, free T4, morning cortisol, or corticotropin stimulation test if clinical suspicion of adrenal insufficiency, water deprivation test to exclude diabetes insipidus) should be performed. Sella MRI to exclude pituitary and/or hypothalamic tumors or infiltrative disease should be performed in men with severely low serum testosterone (eg, <150 ng/dL [5.2 nmol/L]), LH, and FSH concentrations; persistent hyperprolactinemia after discontinuation of medications that elevate prolactin; panhypopituitarism; or tumor mass symptoms or signs (eg, new-onset headache, visual impairment, visual field defects, cerebrospinal fluid rhinorrhea).
If fertility is an important concern and a man presents with infertility (inability of a sexually active couple to conceive after a year of unprotected intercourse) with or without cooccurring testosterone deficiency, a seminal fluid analysis should be performed on an ejaculated semen sample obtained by masturbation after a 2- to 7-day period of abstinence from ejaculation.3,4 Given the extreme variability in sperm concentrations, a seminal fluid analysis should be performed on at least two occasions (separated by at least 1–2 weeks). World Health Organization criteria (based on men whose partners became pregnant in 1 year or less) for normal semen parameters include: sperm concentration greater than or equal to 15 million/mL; volume greater than or equal to 1.5 mL; count greater than or equal to 39 million/ejaculate; sperm motility greater than or equal to 40%; and morphology greater than or equal to 4% strict normal forms.30
In older men (especially, men >70 years old) who are at risk for falls and bone fractures, assessment of bone mineral density to exclude the presence of osteoporosis is advisable.3,4
Alvin M. Matsumoto, MD
INTRODUCTION
Over the last 20 years, large database surveys of several countries reported consistent two- to four-fold increases in the rates of testosterone testing and prescriptions from 2000 to 2014.1 Increases in prescription rates were temporally associated with the availability of transdermal testosterone formulations and direct-to-consumer advertising of testosterone treatment of age-related symptoms associated with low serum testosterone concentrations (“low T”). Following initial observational studies that reported possible increased cardiovascular events associated with testosterone treatment (although not confirmed in subsequent studies) and Food and Drug Administration warnings, testosterone prescription rates declined in 2014 but remained higher than in 2000.1,2 These findings suggested considerable overtesting for testosterone that likely resulted in an increased finding of low testosterone concentrations and subsequent overtreatment with testosterone.
The primary indication for testosterone treatment is a diagnosis of hypogonadism. According to evidence-based Endocrine Society clinical practice guidelines (published initially in 2006 and updated most recently in 2018), hypogonadism is defined as symptoms or signs of testosterone deficiency and consistently low serum testosterone concentrations on at least two occasions.3 However, numerous studies found that testosterone therapy was often initiated in the absence of a diagnosis of hypogonadism as specified by guideline recommendations.1 In these studies, a minority (10%–40%) of men had at least two testosterone measurements and 12% to 46% had no testosterone levels measured in the year before initiating testosterone therapy.
To avoid an inappropriate testosterone treatment, an accurate diagnosis of hypogonadism is imperative. However, for many practitioners, the nonspecific clinical manifestations of testosterone deficiency and complexities associated with serum testosterone measurements present challenges in making a proper diagnosis of hypogonadism. This review highlights these challenges and provides a systematic and practical approach to making a rigorous guideline-based diagnosis and evaluation of hypogonadism.
APPROACH TO THE DIAGNOSIS AND EVALUATION OF HYPOGONADISM
A structured approach is necessary to make a reliable diagnosis of hypogonadism and identify men who are the most appropriate candidates for testosterone replacement and the most likely to respond. The following is a summary of the orderly diagnostic strategy that is recommended to diagnose hypogonadism (Fig. 1):
*Establish the presence of symptoms and signs of testosterone deficiency.
*Consider other causes of largely nonspecific symptoms and signs that might be managed without testosterone treatment.
*Delay laboratory testing for testosterone until conditions that transiently suppress serum testosterone concentrations are completely resolved.
*Confirm the diagnosis of hypogonadism by measuring serum total testosterone in the morning after an overnight fast on at least 2 separate days, using an accurate and reliable standardized assay.
*Measure serum-free testosterone using an accurate and reliable method in patients who have conditions that alter sex hormone-binding globulin (SHBG) or if serum total testosterone concentrations are near the lower limit of the normal adult male reference range.
*In men found to have hypogonadism, measure serum luteinizing hormone (LH) and follicle-stimulating hormone (FSH) concentrations to distinguish primary versus secondary hypogonadism.
*Establish the specific cause of hypogonadism and determine whether the cause of hypogonadism is a potentially reversible or treatable (functional hypogonadism) or an irreversible congenital or destructive disorder (organic hypogonadism) to guide management.
PRESENCE OF CLINICAL MANIFESTATIONS OF TESTOSTERONE DEFICIENCY
Hypogonadism should be suspected in a man who has symptoms and signs of testosterone deficiency.3 Therefore, before measuring serum testosterone and pursuing a diagnosis of hypogonadism, it is important to assess whether a man has clinical manifestations of testosterone deficiency. However, the symptoms and signs of testosterone are mostly nonspecific and occur commonly, especially in older men with multiple comorbidities. The symptoms and signs of testosterone deficiency are broadly classified as sexual, physical, and psychological manifestations (Table 1).
Clinical manifestations depend on the severity and duration of testosterone deficiency. Although uncommon, many individuals with congenital or structural disorders of the testes (eg, Klinefelter syndrome), or hypothalamus or pituitary gland (eg, congenital hypogonadotropic hypogonadism, androgen deprivation therapy in older men with prostate cancer) who have severe testosterone deficiency demonstrate most of the symptoms and signs summarized in Table 1.
Although most symptoms and signs of testosterone deficiency are nonspecific, eunuchoidism (inadequate sexual development), decreased androgen-dependent hair, and very small testes are specific signs that are characteristic of severe testosterone deficiency.4 A clinical finding that is pathognomonic of severe prepubertal testosterone deficiency is eunuchoidism, which is characterized by extremely small penis and testes (prepubertal testis volume <2–4 mL), poorly developed scrotum, lack of androgen-dependent hair pattern (facial, axillary, chest, and pubic hair), a high-pitched voice, and a eunuchoidal body habitus (poor upper body muscle mass; prepubertal fat distribution in hips, chest, and face; and arms and legs >5 cm longer than height). Although usually detected in boys who present with delayed puberty, eunuchoidism may also be found in adult males who are not diagnosed at an earlier age. In men, decreased or loss of androgen-dependent hair is a specific sign of long-standing, severe testosterone deficiency. Decreased testes size is an indicator of impaired spermatogenesis because seminiferous tubules comprise 80% to 90% of the testes volume. However, very small testes (eg, <6 mL) are characteristic of Klinefelter syndrome, which is associated with impairments of testosterone and sperm production.
In contrast to these manifestations, other sexual, physical, and psychological symptoms and signs of testosterone deficiency are nonspecific and may be caused by coexisting comorbidities, illnesses, or medications. In addition to the severity and duration of testosterone deficiency, clinical manifestations of hypogonadism may be modified by previous testosterone therapy, age, comorbid conditions, medications, and variations in target-organ androgen sensitivity, all of which contribute to variability in clinical presentation.
Sexual symptoms, especially reduced libido (sexual interest or desire), loss of spontaneous (nighttime and morning) erections, erectile dysfunction, and reduced sexual activity, are common symptoms of testosterone deficiency that are responsive to testosterone treatment. In the middle-aged and older male participants of the European Male Aging Study (EMAS), sexual symptoms (decreased frequency of sexual thoughts, erectile dysfunction, and decreased frequency of morning erections) demonstrated a stronger syndromic association with low serum testosterone concentrations (total testosterone <230–320 ng/dL [8–11 nmol/L] and free testosterone <64 pg/mL [220 pmol/L]) than physical or behavioral symptoms.5 In healthy older men (aged 60–80 years) with experimental hypogonadism (induced by gonadotropin-releasing hormone [GnRH] agonist administration) who were treated with various doses of testosterone, sexual desire, and erectile dysfunction were progressively decreased at serum testosterone concentrations less than 300 ng/dL (10.4 nmol/L), but significantly only at testosterone level less than 100 ng/dL (3.5 nmol/L), compared with placebo-treated men.6 A meta-analysis of randomized, placebo-controlled clinical trials of testosterone treatment in men who had a morning total testosterone less than or equal to 300 ng/dL (10.4 nmol/L) and at least one symptom or sign of testosterone deficiency (a more rigorous diagnosis of hypogonadism than used in previous meta-analyses) found that testosterone treatment for 3 months or longer increased sexual desire, erectile function, and sexual satisfaction, but had no consistent significant effect on energy or mood.7
CONSIDER OTHER CAUSES OF SYMPTOMS AND SIGNS
Because the clinical manifestations of hypogonadism are nonspecific, it is important to consider the differential diagnosis or contribution of other causes of symptoms and signs to identify those that could be managed and treated independently of testosterone therapy.4,8 Comorbid illnesses and conditions (eg, depression), and medications may contribute significantly to symptoms and signs consistent with testosterone deficiency. Consideration of other causes of symptoms and signs is particularly relevant for a patient who has an isolated or predominant symptom or signs consistent with testosterone deficiency. For example, in an older man who presents with low bone mineral density, potential contributing causes in addition to hypogonadism might include immobility or reduced activity, smoking, excessive alcohol intake, low calcium intake, vitamin D deficiency, medications (eg, glucocorticoids, anticonvulsants), and chronic kidney disease. Of note, in healthy older men with experimental hypogonadism induced by GnRH agonist treated with various doses of testosterone, L4 trabecular bone mineral density by computed tomography was only decreased in men who had serum testosterone levels less than 200 ng/dL (6.9 nmol/L), compared with placebo-treated men.9
DELAY LABORATORY TESTING UNTIL CONDITIONS THAT TRANSIENTLY SUPPRESS TESTOSTERONE ARE RESOLVED
Acute illness or surgery, use of certain medications (eg, opioids or glucocorticoids), and nutritional deficiency that causes an energy deficit (eg, eating disorders, malnutrition, or excessive exercise associated with inadequate energy intake) can transiently suppress gonadotropin and testosterone production.3,4 Therefore, measurements should be delayed until these conditions are completely resolved to avoid an incorrect diagnosis of testosterone deficiency and hypogonadism and subsequent inappropriate testosterone therapy.
REPEATED FASTING MORNING TESTOSTERONE MEASUREMENTS USING AN ACCURATE ASSAY TO CONFIRM HYPOGONADISM
To confirm a diagnosis of hypogonadism in men with symptoms and signs of testosterone deficiency, serum total testosterone concentrations should be measured in the morning (eg, 7:00–10:00 AM) samples after overnight fasting, and measurements should be repeated on at least 2 separate days.3 For the most rigorous and accurate diagnosis of hypogonadism, an accurate and reliable total testosterone assay should be used, preferably an assay that is standardized and certified by an accuracy-based quality control program (eg, Centers for Disease Control and Prevention [CDC] Hormone Standardization Program).
Serum testosterone levels exhibit a circadian variation with higher values in the morning that is blunted but still present in older compared with young men.10 More than half of the men older than 65 years of age who had low testosterone in the afternoon were found to have normal levels measured in the morning. Testosterone concentrations also demonstrate substantial day-to-day variability. Approximately 30% to 35% of men who had a single low testosterone level on initial sampling were subsequently found to have a normal testosterone level on repeat sampling.11 In community-dwelling middle-aged to older men who had single initial serum testosterone less than 250 ng/dL (8.7 nmol/L), 20% had average testosterone greater than 300 ng/ dL (10.4 nmol/L) on repeated sampling over the subsequent 6 months; a single testosterone measurement was inadequate to define an individual’s serum concentrations.12 In contrast, no men who had initial average testosterone less than 250 ng/dL on two or more samples drawn on separate days had average testosterone greater than 300 ng/dL on repeated sampling over the subsequent 6 months. Finally, testosterone levels are suppressed by glucose administration and food intake.13,14 These findings provide a strong scientific rationale for using a standardized approach of repeated blood sampling in the morning after an overnight fast to reduce the influence of biologic variability of serum testosterone measurements that contributes to inaccurate diagnosis of hypogonadism.
Currently, most testosterone measurements are not standardized and performed by automated platform-based immunoassays and few are performed by mass spectrometry-based assays, which are more accurate but may not be standardized. As a result, testosterone assays exhibit considerable assay-to-assay variation in the values measured. A College of American Pathologists peer-based quality control sample from a hypogonadal man measured in 1133 laboratories using 14 different assays reported testosterone values that ranged from 45 to 365 ng/dL (1.6–12.7 nmol/L), that is, values that ranged from clearly hypogonadal to eugonadal concentrations.15 In contrast, the same sample measured in five different mass spectrometry-based assays ranged from 60 to 72 ng/dL (2.1–2.5 nmol/L).15 This extreme interassay and interlaboratory variability is caused by the lack of standardization of most total testosterone assays. As a result, measured values and reference ranges for total testosterone differ from assay-to-assay and laboratory-to-laboratory. The extreme variability in reference ranges is highlighted in two recent surveys of laboratories in the United States and the United Kingdom. In 120 laboratories in the United States, the lower limit of the reference range ranged from 160 to 300 ng/dL (5.5–10.4 nmol/L) with the lower limit in 50% of laboratories less than or equal to 241 ng/dL (8.4 nmol/L).16 In 60 laboratories in the United Kingdom, the lower limit of the reference ranges ranged 141 to 317 ng/dL (4.9–11 nmol/ L) with the lower limit in 50% of laboratories less than 231 ng/dL (8.0 nmol/L).17
Clearly, the use of a uniform lower limit of normal, such as total testosterone less than 300 ng/dL (10.4 nmol/L) as used by some clinical practice guidelines,18 is inappropriate for all total testosterone assays and results in an increased likelihood of overdiagnosis or underdiagnosis of hypogonadism. The Endocrine Society clinical practice guideline recommends the use of a standardized testosterone assay, such as a CDC-certified assay or similar accuracy-based assay standardization program.3 CDC-certified total testosterone assays (mostly mass spectrometry and some immunoassays) are listed and regularly updated at https://www.cdc.gov/labstandards/hs_ certified_participants.html. Most major commercial reference laboratories offer CDC-certified total testosterone assays. However, clinicians should be careful when ordering because some reference laboratories offer CDC-certified and non-CDC certified total testosterone assays.
A harmonized reference range that cross-calibrated testosterone values to CDC reference standards were established in 9054 community-dwelling men.19 The harmonized reference range for young (18–39-year-old), healthy, nonobese men was 264 to 916 ng/dL (9.2–31.8 nmol/L). This range is used as a uniform reference range for all CDC-certified total testosterone assays. If nonstandardized assays are used, total testosterone concentrations and reference ranges vary greatly depending on the specific assay and may not accurately identify men with hypogonadism.
Consistent use of accurate and reliable CDC-certified standardized assays in clinical trials of hypogonadism (eg, The Testosterone Trials20), clinical practice guidelines (eg, Endocrine Society guidelines3 ), and endocrine practice will facilitate translation of research findings to clinical practice. Standardized testosterone assays will also prevent an inaccurate diagnosis of hypogonadism with subsequent inappropriate testosterone treatment and potential adverse events and increased health care costs associated with treatment.
INDICATIONS FOR ACCURATE AND RELIABLE FREE TESTOSTERONE MEASUREMENTS
Serum total testosterone is affected by alterations in SHBG concentrations (Table 2) and the high level of assay imprecision in the total testosterone immunoassays in the low range increases the risk of misdiagnosis when the measured total testosterone concentrations are at or slightly higher or lower than the lower limit of the normal adult male reference range. Therefore, in men being evaluated for hypogonadism who have conditions that alter SHBG or serum total testosterone concentrations that are moderately higher or lower than the lower limit of the normal adult male reference range (eg, 200–400 ng/dL [6.9–13.9 nmol/L]), an accurate measurement of free testosterone (free testosterone by equilibrium dialysis or calculated free testosterone estimate) should be used to confirm testosterone deficiency.3
In circulation, testosterone is mostly bound to serum proteins, primarily tightly bound to SHBG with high affinity and loosely bound to albumin with low affinity, and only 2% to 4% circulating testosterone is unbound to proteins, that is, free testosterone.21 Free testosterone and loosely albumin-bound testosterone are referred to as bioavailable testosterone. According to the free hormone hypothesis, free testosterone is the biologically active circulating fraction and testosterone loosely bound to albumin can dissociate in capillaries and become potentially biologically available in some tissues with long capillary transit times (eg, liver and brain).
Serum total testosterone assays measure protein-bound and free testosterone. Therefore, conditions associated with alterations in SHBG concentrations (not necessarily outside the normal reference range) affect total testosterone levels in the same direction; conditions associated with low SHBG levels (eg, obesity) result in low total testosterone and conditions associated with high SHBG levels (eg, advanced old age) result in high total testosterone concentrations. Because free testosterone is the biologically active fraction that is regulated by negative feedback control of gonadotropin secretion, abnormalities in total testosterone caused solely by alterations of SHBG levels are typically not associated with abnormalities of free testosterone levels.
The importance of free testosterone levels on symptoms of testosterone deficiency is supported by findings in middle-aged to older men in EMAS.22 Compared with eugonadal men who had normal total and free testosterone (303 ng/dL [10.5 nmol/L] and 65 pg/mL [226 pmol/L], respectively), men who had low total testosterone but normal free testosterone levels were more obese, had lower SHBG, and lacked sexual or physical symptoms of testosterone deficiency; in contrast, men with normal total testosterone but low free testosterone concentrations were older and in poorer health, and reported sexual and physical symptoms of testosterone deficiency. In a subsequent prospective study, eugonadal men in EMAS who had normal total (303 ng/dL [10.5 nmol/L]) and free testosterone (49 pg/mL [170 nmol/L]) and normal LH (>9.4 IU/L) were followed for a median of 4.3 years.23 At follow-up, most (93.2%) remained persistently eugonadal, but based on total testosterone concentrations, 6.8% developed apparent secondary hypogonadism with low total testosterone and normal LH levels. However, of the men who had low total testosterone concentrations, only those who also had low free testosterone levels developed or had worsening of sexual symptoms of testosterone deficiency (decreased sexual thoughts, erectile dysfunction, and decreased morning erections), compared with those with low total testosterone but normal free testosterone concentrations and eugonadal men who remained persistently free of sexual symptoms. Finally, a man who demonstrated undetectable SHBG because of a homozygous missense mutation of SHBG and very low serum total, but normal free testosterone was found to have normal serum gonadotropins, semen analysis, and sexual development, supporting the importance of free rather than total testosterone on these sensitive objective indicators of testosterone action.24
In a large cohort of men (3672 male Veterans; mean age, 59.7 years) who had laboratory evaluation with a panel comprised of total testosterone, SHBG, and albumin measurements and calculated free testosterone, 61.7% of men with low total testosterone had normal calculated free testosterone (<34 pg/mL [118 pmol/L]), whereas only 38.3% of those with low total testosterone had low calculated free testosterone; 2.1% of men with normal total testosterone had low calculated free testosterone.25 These results suggest that reliance only on total testosterone concentrations could result in considerable overdiagnosis of testosterone deficiency and hypogonadism. In these men, low free testosterone could be excluded reliably only when total testosterone exceeded 350 to 400 ng/dL (12.1–13.9 nmol/L) and low free testosterone could only be reliably predicted when total testosterone was less than 150 to 200 ng/dL (5.2– 6.9 nmol/L). Therefore, measurement of free testosterone in men with total testosterone concentrations 200 to 400 ng/dL (6.9–13.9 nmol/L) should improve the accuracy of biochemical evaluation of hypogonadism. In men with very low total testosterone levels (eg, <150 ng/dL [5.2 nmol/L]), free testosterone measurements are not needed because the likelihood of finding a normal free testosterone is extremely low.
Accurate and reliable methods that are available to measure serum free testosterone concentrations include equilibrium dialysis and calculated free testosterone.3,21 Preferably, free testosterone should be measured by an equilibrium dialysis method, the gold standard method. If equilibrium dialysis is not accessible, clinically useful accurate estimates of free testosterone relative to equilibrium dialysis are calculated using measurements of total testosterone, SHBG, and albumin concentrations and various published formulae that use algorithms based on the binding affinity of testosterone to SHBG and albumin. However, it is important to recognize that the accuracy of calculated free testosterone values depends on the accuracy of total testosterone, SHBG, and albumin assays. Therefore, calculated free testosterone values should be cross-calibrated against those measured using equilibrium dialysis. Recent studies suggest that SHBG circulates as a dimer with allosterically coupled binding sites on each of the two monomers; the binding of testosterone to SHBG is a multistep process that involves an allosteric interaction between the two binding sites.21,26 Calculated free testosterone estimates using an ensemble allosteric model yielded free testosterone concentrations that closely approximated those measured by equilibrium dialysis.
Accurate and reliable free testosterone assays are not available in most local laboratories. Therefore, free testosterone by equilibrium dialysis and calculated free testosterone should be measured in a dependable reference laboratory. Limitations of using free testosterone by equilibrium dialysis and calculated free testosterone concentrations in practice are the lack of assay standardization, an accuracy-based quality control program, and a harmonized reference range. Until these limitations are addressed, free testosterone by equilibrium dialysis and calculated free testosterone should use reference ranges established by individual laboratories or their specific assay method.
Many local laboratories and some reference laboratories still measure direct free testosterone levels by a testosterone tracer analog immunoassay on an automated assay platform. Free testosterone immunoassays are inaccurate, resulting in values that are an order of magnitude lower than free testosterone by equilibrium dialysis and calculated free testosterone and they should not be used to evaluate men for hypogonadism.21 Bioavailable testosterone is measured by an ammonium sulfate precipitation method, which is technically demanding or calculated from total testosterone, SHBG, and albumin measurements using the same algorithms that are used for calculating free testosterone. The major limitation of using bioavailable testosterone concentrations for clinical evaluation is the relative lack of clinical studies of testosterone deficiency and hypogonadism using bioavailable compared with those using free testosterone levels.
MEASURE GONADOTROPIN CONCENTRATIONS TO DISTINGUISH PRIMARY VERSUS SECONDARY HYPOGONADISM
If a diagnosis of hypogonadism is confirmed, serum gonadotropin, LH, and FSH concentrations should be measured to determine whether the origin of hypogonadism is a disorder of the testes (primary hypogonadism), or pituitary or hypothalamus (secondary hypogonadism).3,4 Serum LH and FSH should be measured in the same sample as testosterone, usually together with a repeat testosterone measurement after an initial low testosterone level or less commonly with an initial testosterone measurement.
Men with primary hypogonadism exhibit repeatedly low testosterone with simultaneously high LH and FSH concentrations (FSH typically being higher than LH). High LH levels indicate reduced testosterone negative feedback and production by Leydig cells of the testes. High FSH levels indicate seminiferous tubule dysfunction (reflecting reduced inhibin B negative feedback) and impaired sperm production but is a more sensitive indicator of primary testicular dysfunction than high LH levels. If high LH and FSH are measured in the same sample as an initial testosterone measurement, men with normal serum testosterone with high LH and/or FSH concentrations might be identified. These men have mild or subclinical primary hypogonadism (also called compensated hypogonadism), analogous to subclinical hypothyroidism.
Men with secondary hypogonadism demonstrate repeatedly low testosterone with simultaneously low or inappropriately normal LH and FSH levels. Some causes of hypogonadism are associated with defects in the testes and pituitary or hypothalamus, which is combined primary and secondary hypogonadism. However, in most cases, a hormone profile of either primary or secondary hypogonadism predominates. For example, in men with hemochromatosis, iron overload results in defects in the testes and pituitary but the latter is the dominant defect that results in gonadotropin deficiency and a hormone profile of low serum testosterone and low gonadotropin concentrations consistent with secondary hypogonadism.27
Serum LH and FSH measurements are usually performed by automated platform-based immunoassays. Most LH and FSH assays have sufficient sensitivity to distinguish low-normal from low values but are susceptible to immunoassay interference (eg, by high-dose biotin use). Although gonadotropin assays are not standardized, differences in values and reference ranges are small. The reference ranges for serum LH and FSH concentrations in well-characterized, healthy fertile young men are 1.6 to 8.0 IU/L and 1.3 to 8.4 IU/L, respectively.28 Assays with upper limits of reference range higher LH and FSH levels probably included older men or men with unrecognized impairment of spermatogenesis.
Distinguishing whether a patient has primary from secondary hypogonadism is clinically important.3,4 Secondary hypogonadism is caused by a pituitary or hypothalamic tumor that might result in deficiency of or be associated with hypersecretion of other pituitary hormones and space-occupying tumor mass effects (eg, headaches, visual field defects, hydrocephalus, or cerebrospinal fluid rhinorrhea) that may require further management. Also, secondary hypogonadism is commonly caused by potentially reversible gonadotropin suppression in the presence of a functionally intact hypothalamic-pituitary-testicular axis (ie, functional hypogonadism), whereas causes of primary hypogonadism are usually caused by irreversible pathologic disease (ie, organic hypogonadism). Finally, infertility caused by impaired spermatogenesis caused by gonadotropin deficiency in secondary hypogonadism is treatable with gonadotropin (or GnRH)-replacement therapy. In contrast, infertility caused by primary hypogonadism is usually not treatable with hormone therapy and requires other fertility options (eg, assisted reproductive technologies).
It is essential that LH and FSH are measured before initiating testosterone therapy. However, it is not uncommon that testosterone therapy is started without the measurement of gonadotropin levels. In this situation, testosterone should be discontinued for at least 2 to 4 weeks for short-acting (transdermal, oral, transbuccal), 2 to 3 months for intermediate-acting (intramuscular testosterone cypionate or enanthate), and 6 to 12 months for long-acting (intramuscular testosterone undecanoate and testosterone pellets) before measuring testosterone and gonadotropins. However, even with more prolonged discontinuation of testosterone therapy, some men may experience persistent gonadotropin and testosterone suppression, especially older men and those receiving high dosages of exogenous testosterone for long periods of time.
FURTHER EVALUATION OF SPECIFIC CAUSE OF FUNCTIONAL VERSUS ORGANIC HYPOGONADISM
After determining whether a patient has primary or secondary hypogonadism, further evaluation should be performed to establish the specific cause of hypogonadism to guide further management, including the need for testosterone therapy.3,4
It is important to determine whether a man with hypogonadism has organic hypogonadism or functional hypogonadism to guide management (Table 3).3,4,8 Organic (also known as “classical”) hypogonadism is caused by irreversible structural, destructive, infiltrative, developmental, or congenital disorders of the reproductive axis. Generally, it results in severe symptoms and signs of testosterone deficiency and consistently and severely low serum testosterone and high (primary organic hypogonadism) or distinctly low (secondary organic hypogonadism) serum LH and FSH concentrations for which testosterone treatment is indicated. In contrast, functional hypogonadism is caused by potentially reversible or treatable testosterone or gonadotropin suppression. Functional hypogonadism is more common than organic hypogonadism and it usually results in mild symptoms and signs of testosterone deficiency and slightly low serum testosterone and slightly high (functional primary hypogonadism) or normal to low-normal (functional secondary hypogonadism) LH and FSH levels. Some causes of functional secondary hypogonadism, such as long-acting opioid use, result in severe clinical manifestations of testosterone deficiency and severely low serum testosterone, LH, and FSH concentrations. Management of functional hypogonadism should initially focus on the treatment of the underlying causative condition (eg, weight loss for obesity) or discontinuation of the offending medication (eg, glucocorticoids) rather than testosterone treatment. However, in men who have severe functional hypogonadism that is not readily reversible or treatable (eg, men taking methadone for opioid use disorder), testosterone treatment could be considered after acknowledgment and discussion of the inadequacy of high-quality evidence for potential benefits and risks of testosterone treatment.
Age-related hypogonadism in middle-aged to older men is mostly related to functional secondary hypogonadism and gonadotropin suppression caused by age-associated comorbidities (eg, obesity, illness, and use of medications). However, advanced age men (eg, >75 years of age) develop organic primary hypogonadism and testicular failure with elevated gonadotropin levels and reduced testicular responsiveness to LH and human chorionic gonadotropin.29
Evaluation to identify the specific cause of primary or secondary hypogonadism should begin with a careful history and physical examination.4 In men with primary hypogonadism, evaluation should include inquiry about a history of undescended testes; mumps with testicular involvement; testes damage, torsion, or surgery; medications that reduce testosterone production (eg, alkylating agents); and end-stage kidney disease. In men with secondary hypogonadism, assessment should include questioning regarding a history of delayed puberty; anosmia or hyposmia; tumor mass symptoms (eg, headache, peripheral vision loss); hypothalamic/pituitary disease or surgery; trauma brain injury; medications that suppress gonadotropin secretion (eg, long-acting opioids, anabolic steroids, glucocorticoids); morbid obesity; reduced energy intake; excessive exercise; wasting syndromes; alcohol use disorder; type 2 diabetes mellitus; and chronic liver, heart, or lung failure.
If no cause is apparent in a man with primary hypogonadism and very small testes less than 6 mL (normal testis volume is 15–30 mL), a karyotype should be ordered to diagnose Klinefelter syndrome.3 In men with secondary hypogonadism, initial laboratory evaluation should include serum prolactin (to exclude hyperprolactinemia) and iron saturation (to screen for iron overload syndromes, such as hemochromatosis). If there is clinical evidence of hypopituitarism or unprovoked polyuria, assessment of other pituitary or hypothalamic hormones (eg, free T4, morning cortisol, or corticotropin stimulation test if clinical suspicion of adrenal insufficiency, water deprivation test to exclude diabetes insipidus) should be performed. Sella MRI to exclude pituitary and/or hypothalamic tumors or infiltrative disease should be performed in men with severely low serum testosterone (eg, <150 ng/dL [5.2 nmol/L]), LH, and FSH concentrations; persistent hyperprolactinemia after discontinuation of medications that elevate prolactin; panhypopituitarism; or tumor mass symptoms or signs (eg, new-onset headache, visual impairment, visual field defects, cerebrospinal fluid rhinorrhea).
If fertility is an important concern and a man presents with infertility (inability of a sexually active couple to conceive after a year of unprotected intercourse) with or without cooccurring testosterone deficiency, a seminal fluid analysis should be performed on an ejaculated semen sample obtained by masturbation after a 2- to 7-day period of abstinence from ejaculation.3,4 Given the extreme variability in sperm concentrations, a seminal fluid analysis should be performed on at least two occasions (separated by at least 1–2 weeks). World Health Organization criteria (based on men whose partners became pregnant in 1 year or less) for normal semen parameters include: sperm concentration greater than or equal to 15 million/mL; volume greater than or equal to 1.5 mL; count greater than or equal to 39 million/ejaculate; sperm motility greater than or equal to 40%; and morphology greater than or equal to 4% strict normal forms.30
In older men (especially, men >70 years old) who are at risk for falls and bone fractures, assessment of bone mineral density to exclude the presence of osteoporosis is advisable.3,4
