Attention: Restrictions on use of AUA, AUAER, and UCF content in third party applications, including artificial intelligence technologies, such as large language models and generative AI.
You are prohibited from using or uploading content you accessed through this website into external applications, bots, software, or websites, including those using artificial intelligence technologies and infrastructure, including deep learning, machine learning and large language models and generative AI.

Evaluation and Management of Testosterone Deficiency (2024)

Using AUA Guidelines

This AUA guideline is provided free of use to the general public for academic and research purposes. However, any person or company accessing AUA guidelines for promotional or commercial use must obtain a licensed copy. To obtain the licensable copy of this guideline, please contact Keith Price at kprice@auanet.org.

Published 2018, Reviewed and Validity Confirmed 2024

To cite this guideline:
Mulhall JP, Trost LW, Brannigan RE et al: Evaluation and management of testosterone deficiency: AUA guideline. J Urol 2018; 200: 423.

The Evaluation and Management of Testosterone Deficiency AUA Guideline provides guidance to the practicing clinician on how to diagnose, treat and monitor the adult male with testosterone deficiency. The care of testosterone deficient patients should focus on accurate assessment of testosterone levels, symptoms and signs as well as proper on-treatment monitoring to ensure therapeutic testosterone levels are reached and symptoms are ameliorated. Guidance is also given on the management of patients with cardiovascular disease, men who are interested in preserving their fertility and men who are at risk for or have prostate cancer.

Guideline as it appears in The Journal of Urology® [pdf]
Unabridged version of this Guideline [pdf]
Español translated guideline courtesy of Confederacion Americana de Urologia (CAU) [pdf]
Algorithm associated with this guideline: Diagnostic [pdf]
Algorithm associated with this guideline: Treatment [pdf]
Table 1: AUA Nomenclature [pdf]
Table 2: Dosing Profiles of Testosterone Formulations [pdf]
Table 3: Prevalence of Testosterone Deficiency [pdf]
Table 4: Lab Assays for the Diagnosis of Testosterone Deficiency [pdf]
Table 5: Symptoms and Signs of Testosterone Deficiency [pdf]
Table 6: Alternative Therapies [pdf]
Table 7: Follow-up Laboratory Testing
Table 8: Adjunctive Testing [pdf]

Panel Members

John P. Mulhall, MD; Landon W. Trost, MD; Robert E. Brannigan, MD; Emily G. Kurtz, MD; J. Bruce Redmon, MD; Kelly A. Chiles, MD, MSc; Deborah J. Lightner, MD; Martin M. Miner, MD; M. Hassan Murad, MD, MPH; Christian J. Nelson, PhD; Elizabeth A. Platz, ScD, MPH; Lakshmi V. Ramanathan, PhD; Ronald W. Lewis, MD

Executive Summary

Testosterone testing and prescriptions have nearly tripled in recent years; however, it is clear from clinical practice that there are many men using testosterone without a clear indication.1-3 Some studies estimate that up to 25% of men who receive testosterone therapy do not have their testosterone tested prior to initiation of treatment.2, 3 Of men who are treated with testosterone, nearly half do not have their testosterone levels checked after therapy commences.2, 3 While up to a third of men who are placed on testosterone therapy do not meet the criteria to be diagnosed as testosterone deficient,2, 3 there are a large percentage of men in need of testosterone therapy who fail to receive it due to clinician concerns, mainly surrounding prostate cancer development and cardiovascular events, although current evidence fails to definitively support these concerns. Given the clinical and commercial testosterone landscape, the American Urological Association (AUA) identified a need to produce an evidence-based document that informs clinicians on the proper assessment and management of patients with testosterone deficiency. The AUA and the Testosterone Panel were committed to creating a Guideline that ensures that men in need of testosterone therapy are treated effectively and safely.

Methodology

A systematic review utilized research from the Mayo Clinic Evidence Based Practice Center and additional supplementation by the authors. Literature searches included Ovid Medline In-Process & Other Non-Indexed Citations, Ovid MEDLINE, Ovid EMBASE, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, and Scopus. Controlled vocabulary supplemented with keywords was used to search for studies according to each defined question. This search included articles published between January 1, 1980 – February 6, 2017 and yielded 15,217 references, 546 (enrolling approximately 350,000 men) of which were used to support guideline statements.

An updated evidence search was subsequently performed in 2022. The original guideline search strategy was updated and used to systematically search PubMed for new evidence published between the previous search end date and February 2022. This search covered the topic areas of all Key Questions listed previously. The literature search returned 11,537 citations after deduplication. During title first pass review, 11,060 citations were excluded. In second pass, an additional 419 studies were excluded.

When sufficient evidence existed, the body of evidence for a particular treatment was assigned a strength rating of A (high), B (moderate) or C (low) for support of Strong, Moderate, or Conditional Recommendations. In the absence of sufficient evidence, additional information is provided as Clinical Principles and Expert Opinions.

Guideline Statements

Diagnosis of Testosterone Deficiency

1. Clinicians should use a total testosterone level below 300 ng/dL as a reasonable cut-off in support of the diagnosis of low testosterone. (Moderate Recommendation; Evidence Level: Grade B)

2. The diagnosis of low testosterone should be made only after two total testosterone measurements are taken on separate occasions with both conducted in an early morning fashion. (Strong Recommendation; Evidence Level: Grade A)

3. The clinical diagnosis of testosterone deficiency is only made when patients have  low total testosterone levels combined with symptoms and/or signs.(Moderate Recommendation; Evidence Level: Grade B)

4. Clinicians should consider measuring total testosterone in patients with a history of unexplained anemia, bone density loss, diabetes, exposure to chemotherapy, exposure to testicular radiation, HIV/AIDS, chronic narcotic use, male infertility, pituitary dysfunction, and chronic corticosteroid use even in the absence of symptoms or signs associated with testosterone deficiency. (Moderate Recommendation; Evidence Level: Grade B)

5. The use of validated questionnaires is not currently recommended to either define which patients are candidates for testosterone therapy or to monitor symptom response in patients on testosterone therapy. (Conditional Recommendation; Evidence Level: Grade C)  

Adjunctive Testing

6. In patients with low testosterone, clinicians should measure serum luteinizing hormone levels. (Strong Recommendation; Evidence Level: Grade A)

7. Serum prolactin levels should be measured in patients with low testosterone levels combined with low or low/normal luteinizing hormone levels. (Strong Recommendation; Evidence Level: Grade A)

8. Patients with persistently high prolactin levels of unknown etiology should undergo evaluation for endocrine disorders. (Strong Recommendation; Evidence Level: Grade A)

9. Serum estradiol should be measured in testosterone deficient patients who present with breast symptoms or gynecomastia prior to the commencement of testosterone therapy. (Expert Opinion)

10. Men with testosterone deficiency who are interested in fertility should have a reproductive health evaluation performed prior to treatment. (Moderate Recommendation; Evidence Level: Grade B)

11. Prior to offering testosterone therapy, clinicians should measure hemoglobin and hematocrit and inform patients regarding the increased risk of polycythemia. (Strong Recommendation; Evidence Level: Grade A)

12. PSA should be measured in men over 40 years of age prior to commencement of testosterone therapy to exclude a prostate cancer diagnosis. (Clinical Principle)  

Counseling Regarding Treatment of Testosterone Deficiency

13. Clinicians should inform testosterone deficient patients that low testosterone is a risk factor for cardiovascular disease. (Strong Recommendation; Evidence Level: Grade B)

14. Patients should be informed that testosterone therapy may result in improvements in erectile function, low sex drive, anemia, bone mineral density, lean body mass, and/or depressive symptoms. (Moderate Recommendation; Evidence Level: Grade B)

15. Patients should be informed that the evidence is inconclusive whether testosterone therapy improves cognitive function, measures of diabetes, energy, fatigue, lipid profiles, and quality of life measures. (Moderate Recommendation; Evidence Level: Grade B)

16. The long-term impact of exogenous testosterone on spermatogenesis should be discussed with patients who are interested in future fertility. (Strong Recommendation; Evidence Level: Grade A)

17. Clinicians should inform patients of the absence of evidence linking testosterone therapy to the development of prostate cancer. (Strong Recommendation; Evidence Level: Grade B)

18. Patients with testosterone deficiency and a history of prostate cancer should be informed that there is inadequate evidence to quantify the risk-benefit ratio of testosterone therapy. (Expert Opinion)

19. Patients should be informed that there is no definitive evidence linking testosterone therapy to a higher incidence of venothrombolic events. (Moderate Recommendation; Evidence Level: Grade C)

20. Prior to initiating treatment, clinicians should counsel patients that, at this time, it cannot be stated definitively whether testosterone therapy increases or decreases the risk of cardiovascular events (e.g., myocardial infarction, stroke, cardiovascular-related death, all-cause mortality). (Moderate Recommendation; Evidence Level: Grade B) 

21. All men with testosterone deficiency should be counseled regarding lifestyle modifications as a treatment strategy. (Conditional Recommendation; Evidence Level: Grade B)

Treatment of Testosterone Deficiency

22. Clinicians should adjust testosterone therapy dosing to achieve a total testosterone level in the middle tertile of the normal reference range. (Conditional Recommendation; Evidence Level: Grade C)

23. Exogenous testosterone therapy should not be prescribed to men who are currently trying to conceive. (Strong Recommendation; Evidence Level: Grade A)

24. Testosterone therapy should not be commenced for a period of three to six months in patients with a history of a cardiovascular events. (Expert Opinion)

25. Clinicians should not prescribe alkylated oral testosterone. (Moderate Recommendation; Evidence Level: Grade B)

26. Clinicians should discuss the risk of transference with patients using testosterone gels/creams. (Strong Recommendation; Evidence Level: Grade A)

27. Clinicians may use aromatase inhibitors, human chorionic gonadotropin, selective estrogen receptor modulators, or a combination thereof in men with testosterone deficiency desiring to maintain fertility. (Conditional Recommendation; Evidence Level: Grade C)

28. Commercially manufactured testosterone products should be prescribed rather than compounded testosterone, when possible. (Conditional Recommendation; Evidence Level: Grade C)

Follow-up of Men on Testosterone Therapy

29. Clinicians should measure an initial follow-up total testosterone level after an appropriate interval to ensure that target testosterone levels have been achieved. (Expert Opinion)

30. Testosterone levels should be measured every 6-12 months while on testosterone therapy. (Expert Opinion)

31. Clinicians should discuss the cessation of testosterone therapy three to six months after commencement of treatment in patients who experience normalization of total testosterone levels but fail to achieve symptom or sign improvement. (Clinical Principle) 


Watch the guideline panel chair, John P. Mulhall, MD, discuss the new guideline in a series of new videos:


Purpose

Testosterone testing and prescriptions have nearly tripled in recent years; however, it is clear from clinical practice that there are many men using testosterone without a clear indication.1-3 Some studies estimate that up to 25% of men who receive testosterone therapy do not have their testosterone tested prior to initiation of treatment, and nearly half  do not have their testosterone levels checked after therapy commences.2, 3 While up to a third of men who are placed on testosterone therapy do not meet the criteria to be diagnosed as testosterone deficient,2, 3 there are a large percentage of men in need of testosterone therapy who fail to receive it due to clinician concerns, mainly surrounding prostate cancer development and cardiovascular events, although current evidence fails to definitively support these concerns.

The explosion in the use of testosterone in the past decade is multifactorial in its etiology, including the increased use of direct-to-consumer advertising, which has resulted in greater patient knowledge and demand; relaxation of the indications for testosterone prescribing by clinicians; and the establishment of clinical care centers devoted to men's health, testosterone treatment, and anti-aging strategies.

Given the growing concern and need for proper testosterone therapy, the AUA identified a need to produce an evidence-based document that informs clinicians on the proper evaluation and management of testosterone deficient patients. The goals of this document are to (i) guide clinicians in how to assess patients for testosterone deficiency and manage them with testosterone products, and (ii) educate clinicians in key areas of testosterone in which many clinicians are deficient (e.g., interpreting the testosterone literature, understanding testosterone laboratory testing). Ultimately, the AUA and the Testosterone Panel were committed to creating a Guideline that ensures that men in need of testosterone therapy are treated effectively and safely.

Definitions and Index Patient

The Panel chose to cease use of the term hypogonadism, a term introduced decades ago to signify low testosterone levels associated with infertility. Hypogonadism has more recently been used interchangeably with the idea of low testosterone production alone. To be scientifically accurate, the Panel chose the term testosterone deficiency. Testosterone deficiency does not imply simply a state of low testosterone production, but rather to be testosterone deficient is to have low testosterone levels combined with symptoms or signs that are associated with low serum total testosterone (henceforth referred to as 'low testosterone'). Thus, a patient is considered testosterone deficient and a candidate for testosterone therapy only when he meets both criteria. The challenge for clinicians is that the symptoms that have been associated with low testosterone levels are very non-specific and can be manifestations of other conditions (e.g., chronic fatigue, chronic stress, a depressed state).

It was decided that a cut-off value was critical to define testosterone deficiency and that this cut-off be based on at least two total testosterone levels drawn in an early morning fashion at the same laboratory using the same assay. The cut-off of 300 ng/dL was chosen based on the mean total testosterone levels cited in the best available literature with a view to maximizing the potential benefit from prescribing testosterone while minimizing the risks of such treatment.

The Panel explicitly uses the term testosterone therapy rather than testosterone replacement therapy or testosterone supplementation to be in keeping with the beliefs of the current thought leaders in the field. Testosterone therapy refers to all forms of treatment that are aimed at increasing serum testosterone, including exogenous testosterone as well as alternative strategies, such as selective estrogen receptor modulators (SERMs), human chorionic gonadotropin (hCG) or aromatase inhibitors (AIs).

The Panel defines success as achievement of therapeutic testosterone levels to the normal physiologic range of 450 -600 ng/dL (middle tertile of the reference range for most labs) accompanied by symptom/sign improvement/resolution.

The index patient for this guideline is the adult male with testosterone deficiency as defined above; however, the Panel included recommendations for three other patient types who are of great interest and concern for the practicing urologist: the patient with cardiovascular disease (CVD) or who has risk factors for CVD; men with testosterone deficiency who are interested in preserving their fertility; and men with testosterone deficiency who are at risk for or have prostate cancer.

Methodology

The guideline panel developed a priori 15 key questions from which guideline statements were derived. The questions were developed based clinical challenges faced by urologists in daily practice. A systematic review of the published literature was conducted to answer these key questions and provide the evidence base for the guideline. The searches included Ovid Medline In-Process & Other Non-Indexed Citations, Ovid MEDLINE, Ovid EMBASE, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, and Scopus. Controlled vocabulary supplemented with keywords was used to search for studies according to each defined question. This search included articles published between January 1, 1980 - February 6, 2017.

The search yielded 15,217 references, 546 (enrolling approximately 350,000 men) of which were used to support guideline statements. Of the outcomes included in the protocol of this systematic review, data were available on quality of life (QoL), sexual function, cardiovascular events, anemia, bone health, insulin resistance, cardiovascular risk factors, mood, cognitive function, body composition, and numerous adverse events. Minimal data were found regarding outcomes of frailty, risk of venous thromboembolism, hyperestrogenemia, sleep apnea, prostate biopsy, recurrence of treated prostate cancer, and incidence of breast cancer. Randomized controlled trials (RCTs) were sought for effectiveness questions, whereas both randomized and non-randomized studies were sought for adverse events and questions of association and risk factors. Random effects meta-analyses were performed when deemed appropriate. Evidence tables (for included studies) and evidence profiles (showing estimates of effect for the outcomes of interest) were generated and presented to the Panel.

Included Interventions. Direct testosterone therapies included the following: oral agents, transdermal agents (gels, creams, patches), buccal agents, trans-nasal agents, intramuscular (IM) agents (short- and long-acting), and subcutaneous (SQ) pellets. Alternative testosterone therapies included SERMs, hCG, and AIs.

Determination of Evidence Strength. The categorization of evidence strength is conceptually distinct from the quality of individual studies. Evidence strength refers to the body of evidence available for a particular question and includes not only individual study quality but consideration of study design, consistency of findings across studies, adequacy of sample sizes, and generalizability of samples, settings, and treatments for the purposes of the guideline. The AUA categorizes body of evidence strength as Grade A (well-conducted and highly-generalizable RCTs or exceptionally strong observational studies with consistent findings), Grade B (RCTs with some weaknesses of procedure or generalizability or moderately strong observational studies with consistent findings), or Grade C (RCTs with serious deficiencies of procedure or generalizability or extremely small sample sizes or observational studies that are inconsistent, have small sample sizes, or have other problems that potentially confound interpretation of data). By definition, Grade A evidence is evidence about which the Panel has a high level of certainty, Grade B evidence is evidence about which the Panel has a moderate level of certainty, and Grade C evidence is evidence about which the Panel has a low level of certainty.

AUA Nomenclature: Linking Statement Type to Evidence Strength. The AUA nomenclature system explicitly links statement type to body of evidence strength, level of certainty, magnitude of benefit or risk/burdens, and the Panel's judgment regarding the balance between benefits and risks/burdens (Table 1 - See button below). Strong Recommendations are directive statements that an action should (benefits outweigh risks/burdens) or should not (risks/burdens outweigh benefits) be undertaken because net benefit or net harm is substantial. Moderate Recommendations are directive statements that an action should (benefits outweigh risks/burdens) or should not (risks/burdens outweigh benefits) be undertaken because net benefit or net harm is moderate. Conditional Recommendations are non-directive statements used when the evidence indicates that there is no apparent net benefit or harm or when the balance between benefits and risks/burdens is unclear. All three statement types may be supported by any body of evidence strength grade. Body of evidence strength Grade A in support of a Strong or Moderate Recommendation indicates that the statement can be applied to most patients in most circumstances and that future research is unlikely to change confidence. Body of evidence strength Grade B in support of a Strong or Moderate Recommendation indicates that the statement can be applied to most patients in most circumstances but that better evidence could change confidence. Body of evidence strength Grade C in support of a Strong or Moderate Recommendation indicates that the statement can be applied to most patients in most circumstances but that better evidence is likely to change confidence. Body of evidence strength Grade C is only rarely used in support of a Strong Recommendation. Conditional Recommendations also can be supported by any evidence strength. When body of evidence strength is Grade A, the statement indicates that benefits and risks/burdens appear balanced, the best action depends on patient circumstances, and future research is unlikely to change confidence. When body of evidence strength Grade B is used, benefits and risks/burdens appear balanced, the best action also depends on individual patient circumstances, and better evidence could change confidence. When body of evidence strength Grade C is used, there is uncertainty regarding the balance between benefits and risks/burdens, alternative strategies may be equally reasonable, and better evidence is likely to change confidence.

Where gaps in the evidence existed, the Panel provides guidance in the form of Clinical Principles or Expert Opinion with consensus achieved using a modified Delphi technique if differences of opinion emerged. A Clinical Principle is a statement about a component of clinical care that is widely agreed upon by urologists or other clinicians for which there may or may not be evidence in the medical literature. Expert Opinion refers to a statement, achieved by consensus of the Panel, that is based on members' clinical training, experience, knowledge, and judgment for which there is no evidence.

For additional information on the challenges associated with reviewing the literature on testosterone deficiency, refer to the Appendix A section in the left menu.

Table 1: AUA Nomenclature


Name Brand Pharmaceutical Agents

The AUA has a policy that all pharmaceutical and biological agents are referred to only by their chemical compound formulation in guidelines, white papers, and best practice statements and not by their brand or generic name. This allows the AUA to eliminate any notion of allegiance to industry in general, or to any product in particular. Additionally, identifying drugs solely by their chemical compound formulation allows guidelines to remain current, despite the dynamic nature of the marketplace.

The AUA has made an exception for this guideline. For most pharmaceutical products, the usage, dosage, and application is consistent across brands, and identification by chemical compound is sufficient to communicate to the reader when to use a given medication. The testosterone therapeutic space is relatively unique. While all products contain the same medication (testosterone), each product and modality has distinct pharmacokinetic and application attributes based on the excipient agents and the permeator components.

For example, there are several testosterone gels available in 1%, 1.62%, and 2% formulations, each marketed under a different brand or generic name. Within this modality family alone, there are three different application sites, including upper body, thigh, and axilla, with four different dosing ranges for each gel.

Identifying injectable drugs can be likewise confusing. While there are three injectable drugs, two of them are short-acting and one is long-acting. Finally, testosterone pellets are also available in branded form, with no generic agents currently available.

To merely refer to injectable or gel testosterone formulations without differentiation does not impart complete and accurate information to the reader. Considering the inherent confusion surrounding testosterone therapy in the current prescribing landscape, the AUA believes it is imperative to be as explicit as possible and present the reader the most complete information, which will optimize the efficacy and safety of testosterone therapy.

Please refer to Table 2 (See button below) for more information on pharmaceutical products discussed in this guideline. A detailed profile of the therapeutic agents discussed in this guideline can be found in Appendix B (in the Appendix B section in the left menu).

Table 2: Dosing Profiles of Testosterone Formulations


Prevalence

The prevalence of testosterone deficiency in the American male population is difficult to quantify. A review by Millar et al.4 searched MEDLINE and Embase databases from January 1966 to July 2014 for studies that compared clinical indication of low testosterone along with a measurement of serum testosterone in men. Among the 40 studies (N=37,565; age 43-82 years) that met the inclusion criteria, the prevalence of low testosterone, defined as the lower normal limit of the assay used, ranged from 2-77%.4 The authors commented on the discrepant testosterone thresholds used to define low testosterone as well as conflicting approaches to measuring testosterone: total testosterone was used in 29 studies, range 200-433 ng/dL; bioavailable testosterone was used in 9 studies, range 69.4-198.4 ng/dL; and free testosterone was used in 4 studies, range 4.6-7.0 ng/dL.

Other population-based studies have attempted to measure prevalence, but have not used standard methodology, which makes arriving at a definitive number of testosterone deficiency difficult. Across the prevalence literature, the cut-off values used to define low testosterone vary widely, heterogeneity exists in the populations studied, the forms of testosterone used to measure testosterone (total and/or free) are not consistent, and the assays utilized to measure testosterone differ. Given these inconsistences, prevalence of low testosterone has varied dramatically among studies, with statistics reporting 2 - 50%.5-8 A summary of findings from four large-scale contemporary prevalence studies can be found in Table 3 (See button below).

Table 3: Prevalence of Testosterone Deficiency


Testosterone Measurement

Testosterone is the predominant androgen in males and is involved in a multitude of physiological and biochemical processes throughout the body. It is bound to albumin (50%, loosely-bound), sex hormone-binding globulin ([SHBG], 44%, tightly-bound), corticotropin-binding globulin (4%, loosely-bound), and approximately 2% circulates as free testosterone.9 The free and loosely-bound testosterone fractions combined are known as bioavailable testosterone.

Testosterone assays are plagued by variability in results. This variability is expressed as a coefficient of variation (CV), which is a measure of precision.10 In order to express this precision of assay test results, two measures of the CV are typically reported: the inter-assay CV and the intra-assay CV. Inter-assay CV measures the agreement between tests using the same method of measurement on identical samples, in the same laboratory, by the same operator using the same equipment within a short interval of time. Intra-assay CV is the degree of variation between repeated measurements of the same sample under different conditions. These parameters are calculated by analyzing normal and abnormal control specimens that have known values of the substance being measured.

The differences in testosterone methodologies have led to considerable effort by a variety of parties including the Centers for Disease Control (CDC) and the College of American Pathologists towards harmonization of assays. Part of this effort includes the availability of serum-based reference material from pooled sera available from the National Institute for Standards and Technology for testosterone and a hormone standardization program using liquid chromatography/mass spectrometry (LCMS) offered by CDC. Laboratories that perform testosterone assays that have a CV that falls within ±6.4 % of samples tested by the CDC (with testosterone values ranging from 2.5-1,000 ng/dL) are certified. The names of these laboratories are available on the CDC website.11-14

An overview of the assays available to aid in the diagnosis of testosterone deficiency is available in Table 4 (See button below).

Table 4: Assays for the Diagnosis of Testosterone Deficiency


Reference Ranges

Well-established reference ranges constitute the essential basis for identifying whether the circulating levels of a particular analyte, testosterone in this case, are normal or low. Due to the challenges in testosterone methodology, there is considerable variability in testosterone reference ranges.13  The specific reference ranges used to diagnose testosterone deficiency are discussed in more depth later in this document. However, practicing clinicians who review testosterone lab results will commonly face the dilemma of whether to use the reference ranges published by their specific lab or the absolute measure itself. As an example, a total testosterone value of 250 ng/dL may be considered low based on the current guideline but be marked within the normal range by the laboratory. This situation commonly occurs as reference laboratories often define a normal value as ranging within the 5th (or 2.5th) and 95th (or 97.5th) percentiles of a sampled population. However, as the testosterone literature uses absolute values to define low testosterone, the absolute value is ultimately the most important factor to determine whether patients may expect to achieve benefits with testosterone therapy. In cases of discrepancy between laboratory reference ranges and this guideline, clinicians are recommended to utilize the absolute value with the understanding that all labs (including CDC-certified LCMS) include some degree of variability. Clinicians wishing to identify laboratories meeting CDC standards are encouraged to refer to the list of sites currently meeting CDC requirements listed on the CDC Hormone Standardization Program.

Diagnosis

Guideline Statement 1

1. Clinicians should use a total testosterone level below 300 ng/dL as a reasonable cut-off in support of the diagnosis of low testosterone. (Moderate Recommendation; Evidence Level: Grade B)

Discussion


Guideline Statement 2

2. The diagnosis of low testosterone should be made only after two total testosterone measurements are taken on separate occasions with both conducted in an early morning fashion. (Strong Recommendation; Evidence Level: Grade A)

Discussion


Guideline Statement 3

3. The clinical diagnosis of testosterone deficiency is only made when patients have  low total testosterone levels combined with symptoms and/or signs.(Moderate Recommendation; Evidence Level: Grade B)

Discussion and Table 5


Guideline Statement 4

4. Clinicians should consider measuring total testosterone in patients with a history of unexplained anemia, bone density loss, diabetes, exposure to chemotherapy, exposure to testicular radiation, HIV/AIDS, chronic narcotic use, male infertility, pituitary dysfunction, and chronic corticosteroid use even in the absence of symptoms or signs associated with testosterone deficiency. (Moderate Recommendation; Evidence Level: Grade B)

Discussion


Guideline Statement 5

5. The use of validated questionnaires is not currently recommended to either define which patients are candidates for testosterone therapy or monitor symptom response in patients on testosterone therapy. (Conditional Recommendation; Evidence Level: Grade C)

Discussion


Adjunctive Testing

Guideline Statement 6

6. In patients with low testosterone, clinicians should measure serum luteinizing hormone levels. (Strong Recommendation; Evidence Level: Grade A)

Discussion


Guideline Statement 7

7. Serum prolactin levels should be measured in patients with low testosterone levels combined with low or low/normal luteinizing hormone levels. (Strong Recommendation; Evidence Level: Grade A)

Discussion


Guideline Statement 8

8. Patients with persistently high prolactin levels of unknown etiology should undergo evaluation for endocrine disorders. (Strong Recommendation; Evidence Level: Grade A)

Discussion


Guideline Statement 9

9. Serum estradiol should be measured in testosterone deficient patients who present with breast symptoms or gynecomastia prior to the commencement of testosterone therapy. (Expert Opinion)

Discussion


Guideline Statement 10

10. Men with testosterone deficiency who are interested in fertility should have a reproductive health evaluation performed prior to treatment. (Moderate Recommendation; Evidence Level: Grade B) 

Discussion


Guideline Statement 11

11. Prior to offering testosterone therapy, clinicians should measure hemoglobin and hematocrit and inform patients regarding the increased risk of polycythemia. (Strong Recommendation; Evidence Level: Grade A)

Discussion


Guideline Statement 12

12. PSA should be measured in men over 40 years of age prior to commencement of testosterone therapy to exclude a prostate cancer diagnosis. (Clinical Principle)

Discussion


Counseling Regarding Treatment of Testosterone Deficiency

Guideline Statement 13

13. Clinicians should inform testosterone deficient patients that low testosterone is a risk factor for cardiovascular disease. (Strong Recommendation; Evidence Level: Grade B)

Discussion


Guideline Statement 14

14. Patients should be informed that testosterone therapy may result in improvements in erectile function, low sex drive, anemia, bone mineral density, lean body mass, and/or depressive symptoms. (Moderate Recommendation; Evidence Level: Grade B)

Discussion


Guideline Statement 15

15. Patients should be informed that the evidence is inconclusive whether testosterone therapy improves cognitive function, measures of diabetes, energy, fatigue, lipid profiles, and quality of life measures. (Moderate Recommendation; Evidence Level: Grade B)

Discussion


Guideline Statement 16

16. The long-term impact of exogenous testosterone on spermatogenesis should be discussed with patients who are interested in future fertility. (Strong Recommendation; Evidence Level: Grade A)

Discussion


Guideline Statement 17

17. Clinicians should inform patients of the absence of evidence linking testosterone therapy to the development of prostate cancer. (Strong Recommendation; Evidence Level: Grade B)

Discussion


Guideline Statement 18

18. Patients with testosterone deficiency and a history of prostate cancer should be informed that there is inadequate evidence to quantify the risk-benefit ratio of testosterone therapy. (Expert Opinion)

Discussion


Guideline Statement 19

19. Patients should be informed that there is no definitive evidence linking testosterone therapy to a higher incidence of venothrombolic events. (Moderate Recommendation; Evidence Level: Grade C)

Discussion


Guideline Statement 20

20. Prior to initiating treatment, clinicians should counsel patients that, at this time, it cannot be stated definitively whether testosterone therapy increases or decreases the risk of cardiovascular events (e.g., myocardial infarction, stroke, cardiovascular-related death, all-cause mortality). (Moderate Recommendation; Evidence Level: Grade B) 

Discussion


Guideline Statement 21

21. All men with testosterone deficiency should be counseled regarding lifestyle modifications as a treatment strategy. (Conditional Recommendation; Evidence Level: Grade B)

Discussion


Treatment of Testosterone Deficiency

Guideline Statement 22

22. Clinicians should adjust testosterone therapy dosing to achieve a total testosterone level in the middle tertile of the normal reference range. (Conditional Recommendation; Evidence Level: Grade C)

Discussion


Guideline Statement 23

23. Exogenous testosterone therapy should not be prescribed to men who are currently trying to conceive. (Strong Recommendation; Evidence Level: Grade A)

Discussion


Guideline Statement 24

24. Testosterone therapy should not be commenced for a period of three to six months in patients with a history of cardiovascular events. (Expert Opinion)

Discussion


Guideline Statement 25

25. Clinicians should not prescribe alkylated oral testosterone. (Moderate Recommendation; Evidence Level: Grade B)

Discussion


Guideline Statement 26

26. Clinicians should discuss the risk of transference with patients using testosterone gels/creams. (Strong Recommendation; Evidence Level: Grade A) 

Discussion


Guideline Statement 27

27. Clinicians may use aromatase inhibitors, human chorionic gonadotropin, selective estrogen receptor modulators, or a combination thereof in men with testosterone deficiency desiring to maintain fertility. (Conditional Recommendation; Evidence Level: Grade C)

Discussion and Table 6


Guideline Statement 28

28. Commercially manufactured testosterone products should be prescribed rather than compounded testosterone, when possible. (Conditional Recommendation; Evidence Level: Grade C)

Discussion


Follow-Up of Men on Testosterone Therapy

Guideline Statement 29

29. Clinicians should measure an initial follow-up total testosterone level after an appropriate interval to ensure that target testosterone levels have been achieved. (Expert Opinion)

Discussion


Guideline Statement 30

30. Testosterone levels should be measured every 6-12 months while on testosterone therapy. (Expert Opinion)

Discussion and Table 7


Guideline Statement 31

31. Clinicians should discuss the cessation of testosterone therapy three to six months after commencement of treatment in patients who experience normalization of total testosterone levels but fail to achieve symptom or sign improvement. (Clinical Principle)

Discussion


References

References


Tools and Resources

Future Research

There are several areas in the testosterone deficiency space, more specifically, epidemiology, diagnosis, treatment and adverse events, which warrant more detailed investigation.

Epidemiology

  1. Are there age-specific reference ranges for testosterone? This question should be answered for both total and free testosterone and should likely be focused on two assays, LCMS for total testosterone and equilibrium dialysis for free testosterone.
  2. What is the impact of changes in testosterone levels over the life-span of an individual patient? Specifically, in a man who has high-normal testosterone levels in his younger years, does a drop in testosterone as he ages (with resultant testosterone levels remaining in the normal range) put him at risk for symptoms and deleterious effects of low testosterone levels?
  3. Is there a threshold testosterone level that is linked to specific symptoms (e.g., fatigue, low sex drive, ED, depression, reduced physical function, cognitive effects) or signs (e.g., bone density loss, elevation of HbA1C, MACE)?
  4. Is there a genotype that predicts testosterone deficiency later in life?
  5. Are there biomarkers for the development of MACE in men with testosterone deficiency?
  6. Does exposure to chemotherapy put a man at increased risk for development of testosterone deficiency later in life?

Diagnosis of Testosterone Deficiency

  1. Further research is needed to understand the role of free testosterone level in the diagnosis of men with testosterone deficiency.
  2. What is the role of androgen receptor sensitivity (CAG repeat analysis) in the diagnosis of testosterone deficiency?
  3. There is a great need for the development of robust and reliable patient-reported outcome tools (e.g., questionnaires) in the screening and follow-up of response to therapy in men with testosterone deficiency.

Treatment of Testosterone Deficiency

  1. Long-term analysis is needed on the impact of weight loss and exercise on testosterone levels and reversal of testosterone deficiency.
  2. There is great need for longer term studies looking at symptom and sign improvement.
  3. Trials are currently ongoing to develop orally administered exogenous testosterone agents. Approval of such agents would give men further (and likely palatable) therapeutic options.
  4. Further analysis of the role of subcutaneously administered testosterone. Trials are currently ongoing in this space.
  5. Greater study is needed on the roles of the different non-traditional testosterone therapies (e.g., SERMs, hCG, AIs).
  6. Long-term study is needed of the recovery of endogenous testosterone production after short, medium and long-term exogenous testosterone therapy.
  7. Study of long-term effects of testosterone therapy on MACE is needed.

 Adverse Events of Testosterone Deficiency

  1. Greater study is needed on the prevalence of adverse events (e.g., polycythemia, VTE, gynecomastia, MACE).
  2. A more detailed analysis is required of the mechanisms of polycythemia development in men on testosterone therapy.
  3. Further exploration of the impact of long-term testosterone therapy in prostate cancer patients is a great need. In which sub-populations is testosterone therapy safe?
  4. There is a need for greater study of the bipolar testosterone concept.
  5. Longer-term studies are needed on spermatogenesis recovery strategies in men who have been on testosterone therapy.

Appendix A: Testosterone Literature Methodology Review

As with all AUA guideline documents, recommendations are based where possible on data extracted from the evidence report, which was generated by methodologists from Mayo Clinic. The development of the evidence report was particularly challenging in the testosterone space due to the heterogeneity in the literature resulting in difficulties comparing data across studies. As the reader delves into this guideline, and more importantly reads the literature, it should be borne in mind that studies have varied significantly in areas, such as patient age, failure to control for concomitant comorbidities associated with low testosterone levels, use of total versus free testosterone, and the testosterone cut-offs used to define low levels. This is further complicated by laboratory methodology issues, such as time of day for the blood draws analyzed, number of levels checked, and assays used.

Thousands of articles on testosterone deficiency and testosterone therapy have been published over the past several decades. To accurately interpret the published testosterone literature, it is important to critically evaluate various aspects of study design, including the population evaluated, study inclusion/exclusion criteria, duration of follow-up, primary endpoints, adverse event reporting, statistical reporting, and clinical relevance of findings.

Study Design. Study design is one of the most important aspects of any investigation because it defines the reliability of outcomes and the extent to which they may be extrapolated to other groups. Meta-analyses of RCTs and cohort studies provide the highest levels of evidence and reliability, followed by individual RCTs, prospective cohorts, retrospective cohorts, and observational studies.

When reviewing results from meta-analyses, it is important to recognize that the overall reliability is dependent on the quality of the weakest study included in the analysis. For example, outcomes of meta-analyses using RCTs alone are generally more robust than those that also include cohort studies. Meta-analyses that are limited to only including RCTs may be restricted to a small number of studies and relevant studies may be excluded that could provide sufficient power to make alternative conclusions. This is particularly relevant for the current guideline as it provides context to situations where the pooled odds ratios and mean differences may contradict or fail to support published meta-analyses.

Study Population. Individual study factors, such as the heterogeneity and demographics of the study population, the comorbidities of the study population and how they are controlled in the analysis, and confidence intervals also impact overall study quality. Study populations in individual trials included in any meta-analysis have a significant impact on the reliability of outcomes.  Differences in age, geography, date of initial testing (testosterone immunoassay testing was more commonly used before 2005), comorbid conditions, and baseline and therapeutic testosterone levels across studies introduce heterogeneity in the pooled population. Readers should recognize that guideline statements have been generalized in an attempt to provide a clinically useful document with the understanding that certain populations and clinical scenarios will fall outside of the initial criteria upon which the studies were based. Actual patient scenarios will require individual adaptation with variability in expected outcomes. It also highlights that treating clinicians should have specific endpoints for treatment in mind, with regular monitoring of these outcomes to assure that ongoing therapy is warranted and effective.

Criteria for Testosterone Deficiency and Study Duration. The current guideline only included studies in the meta-analysis that used morning total testosterone <350 ng/dL as an inclusion criterion. Despite this standardization, there remains significant variability among studies. The mean testosterone in the patient population across all the studies was 249 ng/dL but this does not take into account that there were some patients who entered with lower baseline testosterone levels and may have been more likely to demonstrate greater improvements in symptoms compared to those who entered with testosterone levels closer to eugonadal levels, while still meeting the inclusion criteria.411

Study duration is also a significant factor. Testosterone therapy likely yields rapid improvements in some symptoms, while others require a longer time course to show improvement.297 RCTs are commonly of short duration (<1 year), which leads to study results that might have a bias towards lack of benefit in certain symptoms/signs while failing to identify therapeutic benefit, which can require additional time to manifest.

Primary Endpoints, Adverse Events, and Statistical Measures. One important aspect of study design is the specific endpoints and objective measures used to identify outcomes. Studies are often specifically powered and designed to address a key efficacy endpoint, such as a particular symptom improvement, and not to address secondary symptom improvement or adverse events. Although one objective of meta-analyses is to increase study power to identify significant results, this often results in an amalgamation of studies that may have different primary and secondary endpoints, thereby reducing the reliability of the outcomes.

Beyond statistical significance, clinical relevance is another key factor. In analyzing the literature, it is imperative to determine whether or not statistically significant results are clinically meaningful.  For example, a particular study might show that testosterone therapy is correlated with a statistically significant improvement in the IIEF scores in a given population; however, the clinician may not feel that this has any clinical meaning for the patient in terms of his QoL or sexual function. 

Appendix B: Therapeutic Agents for Treatment of Testosterone Deficiency

Transdermal Agents - Gels and Solutions

Pharmacokinetics and Pharmacodynamics. The unique pharmacokinetic profiles of transdermal testosterone preparations relate to several factors, including the delivery system (alcohols or other penetration enhancers), concentration, surface area applied, and location of application.228, 412 Transdermal drug solubility has been variably estimated, with most studies reporting systemic absorption rates ranging from 13-20%.413-415 Repeated application to the same site (after washing) does not reduce uptake, and the use of an occlusive dressing has been shown to increase absorption by approximately 2.5 fold.415, 416

Topical gels and liquids generally demonstrate less variability in absorption uptake when compared to other therapies.417 After application, steady state levels are achieved within 24-72 hours, with testosterone levels returning to baseline within 4 days of discontinuation.418, 419

Dosing Strategies. Liquids and gels should be applied to clean, dry skin, and the treatment site should not be washed until the time of next application to optimize delivery. If insufficient testosterone levels are achieved with one topical agent, including with dose adjustments, substitution with another topical agent is a viable treatment strategy.420

Efficacy. Topical liquid and gel formulations are able to achieve testosterone levels in the normal range in 74-87% of men and are relatively similar among the various preparations.421-423 Given the variable absorption profiles among patients, dose adjustments may be required to achieve appropriate therapeutic delivery. There is no consistent data at this time that demonstrate that one agent achieves higher serum levels than others.

Adverse Effects. Adverse effects specific to topical preparations include application site reactions (3-16% erythema or rash), and risk of transference. Patients should be particularly cautioned against contact with women and children after application of the gel to limit the possibility of inadvertent transmission. Transference may be mitigated by washing hands, covering the application site with clothing, and washing the region prior to anticipated direct contact with others.

Transdermal Agents - Patches

Pharmacokinetics and Pharmacodynamics. Testosterone patches consist of a mixture of testosterone, penetration agents, and a gelatinous matrix separated from the skin by a microporous membrane. Initial studies of testosterone patches demonstrated increases in total testosterone from a baseline 167 ng/dL to a peak of 1,154 ng/dL at 5.7 hours, with a decrease to 490 ng/dL over the next 12 hours.424 Following removal, the observed testosterone half-life was 116 minutes.425, 426 A multicenter, open label study confirmed mirroring of the circadian rhythm when the patch is applied in the evening with a morning peak of 740 ng/dL and a night-time trough of 213 ng/dL.427

When applied to the abdomen, the patch exhibits slightly lower minimum testosterone values (over 24 hours) compared to other methods of delivery and some gels, with bioequivalence noted for average and maximum testosterone values.428 Two RCTs compared patches to IM testosterone administration and demonstrated improved maintenance of testosterone values within normal physiologic levels and closer representation of the natural circadian rhythm.181, 429 However, it is noteworthy that pharmacokinetic comparisons of any one modality to another are dependent on the dosing and schedule of administration. In the case of topical patches, the testosterone levels achieved directly relate to the amount of surface area exposed to drug.430

Dosing Strategies. Patches are currently available in 2 and 4 mg formulations, with a 4mg starting dose recommended and titration to 6 mg permitted.

Efficacy. Patches are able to achieve testosterone levels within normal physiologic ranges (2 patches every 24-48 hours) in 77-100% of individuals with >85% achieving values >300 ng/dL.181, 429

Adverse Effects. The most common adverse effect with patches is application site reactions, which have been historically reported in up to 60% of patients.181 Other adverse effects include pruritus, application site vesicles, and back pain.431 Compared to topical gels and solutions, the rate of transference is likely minimal.

Buccal Agents

One of the oral alternatives for testosterone therapy is the 30 mg sustained-release muco-adhesive buccal pellet applied to the upper gums above the incisor teeth twice daily.432

Pharmacokinetics. Absorption through the oral mucosa avoids liver deactivation that is experienced by other formulations. Testosterone is released from the tablet in a manner similar to the normal daily rhythm of endogenous testosterone, with serum levels rising rapidly after buccal absorption and peak levels reached by the second 12-hour daily dose. It restores the circulating testosterone level to the physiological range. Removal of the system results in a rapid drop in testosterone levels.433

Dosing. The progressive hydration tablet with a matrix containing 30 mg of testosterone is placed in position on the gum above the right or left canine and is held in position for approximately 30 seconds. It adheres to the buccal surface as it slowly hydrates, becoming soft and gelatinous. It is administered twice daily, 12 hours apart.432

Efficacy. In a 12-week study in 82 men, 72.6% of patients achieved a total testosterone concentration within the physiological range at steady state.434 Men treated with the agent were compared to a group of patients given 5 mg of a testosterone gel formulation, and no differences in mean testosterone serum levels were observed between the two groups.435 The study showed 92% of buccal versus 83% of gel patients achieved testosterone levels in the physiologic range.

Adverse Effects. In clinical trials up to two years long, the most common side effects were gum-related disorders: gum or mouth irritation‎ (‎9.2%), gum tenderness‎ (3.1%), gum pain‎ (‎3.1%), and gum edema‎ (2%).432 Another limitation of this therapy includes displacement of the buccal device during exertion activities.

Intranasal Gel

An intranasal testosterone gel applied topically into the nose was approved by the FDA in 2014.

Pharmacokinetics. Following application, testosterone is absorbed through the nasal mucosa to achieve maximum concentrations in about 40 minutes, with a serum half-life of 10-100 minutes.

Dosing. The product is provided in a metered pump that supplies 5.5 mg of testosterone per actuation. The recommended dose is two pumps (one to each nostril) applied three times daily.

Efficacy. In a 90-day open label trial of 306 testosterone deficient men using two actuations (11mg) of the drug applied three times daily, results were reported for 73 men at day 90. Sixty-nine of 73 men (90%) had an average testosterone concentration within the specified normal range for the study (300-1,050 ng/dL). The mean testosterone concentration was 421 ng/dL.436

Adverse Effects. In the clinical trial leading to FDA approval, side effects related to nasal delivery included nasopharyngitis, rhinorrhea, and epistaxis occurring in 7-10% of men.436

Injectable Agents (Short-Acting)

Injectable testosterone is available in several forms, including short acting and long-acting preparations. Although  IM injections are the traditional route for injectable agents, the SQ route has also been described with short-acting agents.437

Pharmacokinetics and Pharmacodynamics. The pharmacokinetics of short-acting testosterone therapy depends on the dose, interval, and method of delivery (SQ versus IM). In a study directly comparing the pharmacokinetics of 2 doses of SQ testosterone enanthate injected weekly (50 or 100 mg) and 1 concentration of IM testosterone enanthate injected once (200 mg), the IM testosterone achieved the highest peak testosterone (mean 2,261 ng/dL) followed by SQ 100 mg (1,345 ng/dL) and SQ 50 mg (622 ng/dL).437 The time-to-peak level was slightly faster with IM testosterone (33 hours) compared to SQ 100 mg (36 hours) and SQ 50 mg (45 hours). The half-life for IM testosterone was also shorter at 173 hours versus 240 hours for SQ testosterone. Mean testosterone values over a 7-day time period were 1,659, 896, and 422 ng/dL for IM testosterone SQ 100, and SQ 50, respectively.

Dosing Strategies. The optimal dosing strategy has not been defined for short-acting IM testosterone preparations. In general, smaller dosages at more frequent intervals are preferred over high, less frequent administrations to limit the duration of time spent outside (above or below) the normal reference range. As an example, a starting dose of 100 mg weekly is preferred to 200 mg every 2 weeks or 300-400 mg monthly. The best time to obtain monitoring blood tests for IM testosterone has not been definitively established. Given the half-life of approximately seven days, it is reasonable to obtain testosterone levels four weeks after starting therapy. Historically, testosterone levels have been measured mid-cycle (day three to four); however, such a measurement protocol misses the ability to define peak and trough levels. While mid-cycle testing is convenient for patients, there may be value in assessing peak level (18-36 hours after injection) as the adverse events (e.g., polycythemia, hyperestrogenism) are likely at least partially related to the peak level. Likewise, there might be value in defining the trough level (measured prior to injection on day one) to ensure patients remains therapeutic throughout the entire cycle. Furthermore, the concept of testosterone 'crash' is well recognized by clinicians, with large differences between peak and trough levels potentially leading patients to become symptomatic towards the end of the cycle despite having therapeutic trough testosterone levels.

Efficacy. In contrast to topical agents where a percentage of men have difficulty achieving therapeutic levels within standard dosing ranges, injectable testosterone preparations are able to achieve therapeutic levels in almost any clinical scenario. However, compared to other agents, short-acting injections can result in longer times in the supra-therapeutic and sub-therapeutic ranges, which may impact overall efficacy and rates of adverse events. This may be overcome by altering injection dose and frequency.

Adverse Effects. Adverse effects that are more common with short acting injectable agents include local site reactions (7-33%) and abnormally elevated Hb/Hct (19-44%; mean 1.6 mg/dL in the current meta-analysis of RCTs).181, 182, 191, 194, 195, 201, 203-214, 216-218, 220, 299-301, 310

Injectable Agents (Long-Acting)

Testosterone undecanoate is the only currently available long-acting injection therapy available as a 750 mg (3 mL) preparation and must be administered in the office under supervision.

Pharmacokinetics and Pharmacodynamics. The pharmacokinetic profile of long-acting IM testosterone therapy has been detailed in several studies.438-440 Morgentaler et al. reported outcomes of the currently available preparation (750 mg in 3 mL oil) administered at weeks 0, 4, and every 10 weeks thereafter.438 Peak concentrations were achieved at a mean 7 days after injection (range 4-42 days). Results after the third injection demonstrated median peak and trough T levels of 813 ng/dL and 317 ng/dL, respectively, with overall median values of 476 ng/dL during the 10-week period. These data are notable as they demonstrate far less variability between peak and trough levels compared to shorter-acting preparations.441, 442

No RCTs have compared the current formulation of IM testosterone undecanoate available in the United States to other therapies. One study reported comparative pharmacokinetics between IM testosterone enanthate (250 mg every 3 weeks) and IM testosterone undenaconate (1,000 mg every 9 weeks, a dosage that is only available outside the United States).440 Results demonstrated that IM testosterone enanthate achieved trough levels of 239 ng/dL compared to 470 ng/dL with IM testosterone undecanoate at the end of the 10-week cycle.

Dosing Strategies. The manufacturer-recommended dosing of IM testosterone undecanoate is 750 mg administered at weeks 0, 4, and every 10 weeks thereafter. The dosing at 0 and 4 weeks represents the loading period followed by regular dosing is every 10 weeks. Further individualization may be considered based on trough testosterone levels at the end of a 10-week injection cycle. For trough total testosterone values <300 ng/dL, the interval may be decreased by 1 week (9 weeks) until values >300 ng/dL are achieved at the end of an injection period. For trough total testosterone values >600 ng/dL, the interval may be prolonged by 1 week (11 weeks) until total testosterone levels <600 ng/dL are noted at the end of an injection period.  

Efficacy. Administration of 750 mg of IM testosterone undecanoate at weeks 0, 4, and every 10 weeks thereafter maintained total testosterone levels between 300-1,000 ng/dL for 94% of men.438 No men experienced maximal values <300 ng/dL, and only 5% had mean concentrations <300 ng/dL during the 10 weeks. Mean peak levels were 890 ng/dL, with 92% of men remaining below 1,500 ng/dL.

Adverse Effects. The most common adverse effects associated with IM testosterone undecanoate 750 mg administered every 10 weeks include injection site pain, acne, and fatigue (all ≤5%).438, 439 Coughing related to the oil-based preparation has been reported secondary to pulmonary microemboli (POME) and reactions to the benzyl benzoate component.438, 443, 444 Due to these risks, the FDA has issued a boxed warning and requires that providers be risk evaluation and mitigation strategy certified to administer the therapy. Patients should be monitored for 30 minutes in a healthcare setting after injection to monitor for POME or anaphylactic-type symptoms.

Although the absolute risks of POME and anaphylaxis require ongoing study, data from 342 patients undergoing 3,022 injections (1,000 mg in 4 mL) over a period of 3.5 years demonstrated that POME occurred after 1.9% of injections (12% of patients experienced at least one POME), with coughing episodes lasting 1-10 minutes in duration.443 All episodes were managed conservatively in the clinic, with no supplemental oxygen required. No episodes of anaphylaxis occurred. Findings are similar to the previously cited pharmacokinetic study (750 mg in 3 mL) in which one patient in 130 (<1%) experienced coughing lasting 10 minutes.438 It is notable that similar findings have also been observed with other oil-based testosterone preparations that are currently most often self-administered at home (typically with lower volumes of injection).445

Long-acting IM testosterone injection may also result in higher rates of polycythemia when compared to topical therapies, which is consistent with other short-acting  IM testosterone therapies. In the current meta-analysis of RCTs, long-acting IM testosterone resulted in a mean increase in Hb levels of 1.4 mg/dL compared to 1.6 mg/dL with short-acting  IM testosterone, 0.9 mg/dL with transdermal preparations, and 0.7 mg/dL with topical patches.182, 194, 195, 201, 203-214, 216-218, 299-301, 310

Subcutaneous pellets

SQ testosterone pellets were initially developed and FDA approved in 1972 and were reformulated in the USA in 2008.

Pharmacokinetics and Pharmacodynamics. The unique pharmacokinetic profile of testosterone pellets is due to their crystalline structure, which dissolves slowly in SQ spaces. Individual pellets consist of 75 mg of testosterone and may be combined to deliver variable doses of testosterone therapy.

Initial pharmacokinetic data were provided by Kaminetsky et al.446 who selected the number of pellets inserted based on testosterone levels and BMI: six pellets for baseline total testosterone level <315 ng/dL and BMI <18.5, eight pellets for BMI 18.5-24.9, 10 pellets for BMI 25-30, and 12 pellets for BMI >30. Men with total testosterone level <225 ng/dL received 10 pellets for BMI <25 and 12 pellets for BMI ≥25. Of the 30 patients enrolled, none met criteria for 6 pellets, and a median of 10 pellets were implanted. Peak total testosterone levels were achieved 1 week after implantation (845 ng/dL) and were conserved until at least week 4 (838 ng/dL), with LH suppressed by week 4. The percentage of men with total testosterone values >315 ng/dL declined from 100% at 4 weeks to 86%, 75%, and 14% by 12, 20, and 24 weeks, respectively.

Mean peak total testosterone levels are dose-dependent, with a mean of 746, 866, and 913 ng/dL noted with 8, 10, and 12 pellets administered (not BMI adjusted).446 The duration of effect is similar, however, and is relatively independent of dosing. These findings are supported by a multi-institutional study that reported that with variable dosing and clinical protocols, most men required re-implantation after four months, with all men returning to sub therapeutic levels by six months.447

Based on these initial data, Kaminetsky and colleagues performed a follow-up re-dosing study with 2 fewer pellets administered if peak testosterone levels were >1,000 ng/dL and 2 additional pellets given for testosterone levels <500 ng/dL. With new dosing, mean testosterone levels declined to 275 ng/dL by week 16, with only 32% having levels >315 ng/dL at that time point. 

Dosing Strategies. Currently, the FDA recommends placement of two to six pellets every three to six months, which has been the recommendation since the approval of pellets in 1972. These recommendations, however, are not based on current testosterone pellet formulations and contrast with pharmacokinetic data available. Definitive dosing protocols have not been described.446,447, 448

Testosterone levels should be obtained at one to four weeks after insertion.446 These 'peak' levels facilitate adjustment of future pellet number: peak level >1,000 ng/dL, reduce by 2 pellets at next insertion; <500 ng/dL, increase by 2 pellets. Subsequent testosterone levels should be assessed around three months after implantation and re-checked every two to four weeks thereafter if persistently therapeutic levels are found. Although no consensus exists, it is reasonable to perform re-implantation when total testosterone levels are <400 ng/dL. Due to variations within the same individual, it is recommended to obtain end-of-cycle testosterone measurements prior to implantation to ensure that levels are sub-therapeutic.446 

Efficacy. Data from a large, multi-institutional series using varied protocols (inserted pellet number ranged from 6 to >10 pellets), demonstrated therapeutic levels in 100% of men at 4 weeks and maintained levels >300 ng/dL at 4 months. It is notable that the majority of providers elected to utilize ≥10 pellets (63%), with 27% of cases including 8-9 pellets, and only 10% of cases using 6-7 pellets. No providers utilized five or fewer pellets, which contrasts with the FDA recommended dosing.221

Adverse Effects. Mild level adverse events specific to SQ pellet insertion includes polycythemia (48-50%), ecchymosis (32-36%), tenderness (20-32%), pain (28-29%), and swelling (16-18%), all of which resolve by 4 months post-insertion.446 Moderate level adverse events were less common (e.g., pain 3%, erythema 3%, ecchymoses 7%) and improved within 1 week. Pellet extrusions are also possible and may be reduced by the use of a V technique whereby 2 channels are created for pellet insertion, thus keeping the most superficial pellet >1cm away from the skin.449 Results of the modified technique resulted in reduced extrusion (7.5% down to 0.8%), infection (5% down to 1.2%), pain (5% down to 1.2%), but an increase in hematoma occurrence (0% up to 1.2%). Of note, hematoma rates were not impacted by the use of anti-coagulants (1.7%), although many practitioners are cautious about pellet use in this population. In one retrospective comparative series, the rate of polycthemia (Hct >52%) with pellets was also higher (13%) than topical gels (5%) but lower than  IM testosterone agents (19%) (p=0.03).220

Appendix C: Adjunctive Testing

Table 8: Adjunctive Testing

TestIndicationClinician Response
Serum Luteinizing HormoneLH is an appropriate first-line test in conjunction with a repeat testosterone level to determine the etiology of the testosterone deficiency. A low or low/normal LH level points to a secondary, or central hypothalamic-pituitary defect, (hypogonadotropic hypogonadism); while an elevated LH indicates a primary testicular defect (hypergonadotropic hypogonadism). The location of the defect may be an important factor in deciding upon further evaluation of such a patient.Men with low testosterone and low to low/normal LH, should have a prolactin level measured.   Men with low to low/normal LH levels are candidates for the use of SERMs in the management of their testosterone deficiency.   Men with very high LH levels (without an obvious cause, such as chemotherapy) are at increased risk for KS, which can be diagnosed using a karyotype.
Serum Follicle-stimulating HormoneMen who are interested in preserving their fertility warrant a baseline FSH prior to the commencement of SERMs, hCG, or AI. The presence of an elevated FSH level suggests abnormal spermatogenesis.Men with elevated FSH levels should have a semen analysis.   Men with very high FSH levels (without an obvious cause, such as chemotherapy) are at increased risk for KS, which can be diagnosed using a karytotype.  
Serum HbA1CWhile data supporting the link between testosterone deficiency and diabetes is mixed, in the middle-aged or older testosterone deficient man with obesity, metabolic syndrome, or chronic exposure to corticosteroids, measuring a HbA1C level should be considered.An abnormal HbA1C level should prompt referral (primary care clinician, internist, endocrinologist) for further evaluation and management.  
Serum ProlactinMen with low testosterone level accompanied by a low/low-normal LH level warrant measurement of serum prolactin to investigate for hyperprolactinemia. If prolactin is mildly elevated (≤1.5 times the upper limit of normal) a repeat prolactin should be drawn to rule out a spurious elevation.If the prolactin level is mildly elevated, a repeat prolactin level should be measured to rule out a spurious elevation.   For persistently elevated prolactin levels referral to an endocrinologist should be considered.
Serum EstradiolSerum E2 levels should be measured in a patient with baseline gynecomastia or breast symptoms. For those men who develop gynecomastia or breast symptoms while on testosterone therapy, measuring a E2 level is optional.  If E2 is persistently elevated (>40 pg/mL) at baseline, referral to an endocrinologist should be made.   For gynecomastia/breast symptoms that develop while on testosterone therapy, a period of monitoring should be considered, as breast symptoms sometimes abate.   If gynecomastia/breast symptoms persist on testosterone therapy and the E2 level is elevated, reduction may be accomplished through dose adjustment of the testosterone therapy if the on-treatment testosterone levels are in the upper range of normal. If the on-treatment testosterone levels are low/normal, E2 level reduction can be accomplished by the use of AIs.
Pituitary MRIMen with sustained elevated prolactin levels, very low total testosterone levels (<150 ng/dL) and unexplained failure to produce LH/FSH warrant a pituitary MRI to identify sellar (pituitary adenoma, prolactinoma, infiltrative diseases of the pituitary) or parasellar processes.The clinician may decide to refer such patients to an endocrinologist prior to ordering an MRI or may order the MRI first and refer only for abnormalities.   For clinicians experienced in managing prolactinomas, bromocriptine or cabergoline may be prescribed without endocrinology input.  
Bone DensitometryMen with testosterone deficiency are at increased risk of bone density loss. Consideration of a baseline dual energy X-ray absorptionometry (DEXA) scan is warranted, particularly in middle-aged or older men with severe testosterone deficiency or in men with a history low trauma bone fracture.Results are used to assess baseline bone health and if abnormal to follow changes over time whether the patient opts for testosterone therapy or not.   Patients with osteoporosis should be referred to an endocrinologist.  
KaryotypeA karyotype should be considered in men with unexplained hypergonadotropic hypogonadism. The most common chromosomal abnormality identified is 47,XXY, also known as KS, although other chromosomal abnormalities can also be found. For those clinicians inexperienced in managing KS, referral to a more experienced clinician is advisable.
Hemoglobin/Hematocrit  Prior to initiation of testosterone therapy, all patients should undergo baseline assessment of Hb/Hct.If baseline Hct >50%, the clinician should with-hold testosterone therapy until the etiology of the high Hct is explained (polycthemia vera, living at altitude, obstructive sleep apnea, tobacco use).   While on testosterone therapy, a Hct ≥54% warrants intervention. In men with high on-treatment testosterone levels, dose adjustment should be attempted as first-line management. In men with low-normal on-treatment testosterone levels, measuring a SHBG level and a free testosterone level using a reliable assay such as equilibrium dialysis is suggested. If SHBG levels are low/free testosterone levels are high, dose adjustment of the testosterone therapy should be considered.   Men with on-treatment low/normal total and free testosterone levels should be referred to a hematologist for further evaluation.  
AI: aromatase inhibitor, E2: estradiol, FSH: follicle-stimulating hormone, hCG: human chorionic gonadotropin, Hg: hemoglobin, Hct: hematocrit, KS: Klinefelter syndrome, LH: luteinizing hormone, SERM: selective estrogen receptor modulator, SHBG: sex hormone-binding globulin

Appendix D: Testosterone Therapy and the Risk of Major Adverse Cardiac Events

Given the conflicting nature of the evidence, the Panel cannot definitively state that there is an association between testosterone therapy and subsequent MACE events nor can it be stated definitely that testosterone therapy is associated with reduced incidence of MACE. However, the FDA added a warning to testosterone product labeling after reviewing five observational studies and two meta-analyses of RCTs that examined the effects of testosterone therapy on MACE.

Two of the trials and one meta-analysis pointed to an increased risk of cardiovascular events,363, 364, 366 two revealed no cardiovascular risk,233, 367 and one was neutral with respect to risk.373 The Corona meta-analysis,372 which showed that there was no increased risk of cardiovascular events, was not officially reviewed but was taken into consideration in the final analysis. Although the committee reviewing the evidence concluded that there was not enough data to definitively state that testosterone therapy posed a significant cardiovascular risk, the FDA nonetheless required testosterone product manufacturers to add information to the labeling about a possible increased risk of myocardial infarction and cerebrovascular accidents in patients using testosterone therapy.

Two of the retrospective studies included in the FDA review pointed to an increased risk of cardiovascular events in men on testosterone therapy. Vigen et al. (2013)363 conducted a retrospective analysis of patients who received a prescription for testosterone therapy after coronary angiography. The end-points included all-cause mortality as well as cardiovascular events. Patients were excluded if they had previously been on testosterone therapy, had a total testosterone >300 ng/dL, had a baseline Hct >50%, a PSA ≥4 ng/mL, or had commenced testosterone therapy after a myocardial infarction. Over a mean duration of 27.5 months, 1,223 men received testosterone therapy, and 7,486 were placed on placebo. In the testosterone therapy group, the raw data revealed a 2% myocardial infarction rate and a 3% cerebrovascular accident rate compared to 6% and 6%, respectively, in those patients not receiving testosterone. Complex statistical analysis using a methodology known a stabilized inverse propensity treatment weighting was utilized to adjust for 50 potentially confounding variables. Following inverse propensity treatment weighting, the cumulative percentage of patients who met the primary outcome 3 years post-angiography was 25.7% on treatment and 19.9% in the placebo group. This resulted in a calculated OR for developing a cardiovascular event in the testosterone therapy group of 1.29 (CI: 1.04 - 1.58; p=0.02), after adjusting for the presence of CVD. Two errata were published because of significant data errors in the original dataset. It is also unclear if everyone receiving a testosterone prescription actually used the medication, considering that 17.6% of patients in the treatment group filled only a single prescription. There was also inadequate documentation of on-treatment testosterone levels with 40% of men having no documented laboratory testing performed after the prescribing of testosterone therapy.

Finkle et al. (2014)364 conducted a retrospective analysis of the Truven healthcare database (n=55,593) and compared non-fatal myocardial infarction events in men who were on testosterone therapy to those on PDE5 inhibitors. The authors compared the relative risk ratio (RRR) of developing a myocardial infarction within 90 days of receiving a testosterone or PDE5 inhibitor prescription compared to the year prior when patients were not using any medication. For men without a history of CVD, the RRR for having a myocardial infarction in those aged <65 years was 0.91 (CI: 0.60, 1.37) and for men ≥ 65 years of age 2.41 (CI: 1.12, 5.17). For men with a history of CVD, the RRR were more striking: <65 years of age 2.07 (CI: 1.05, 4.11) and ≥65 years 1.91 (CI: 0.66, 5.5). This analysis was limited in that it used an insurance claims database, had an abbreviated follow-up, and compared testosterone therapy to a class of medications (PDE5 inhibitors) known to be endothelial stabilizers and potentially cardioprotectants.

The Xu meta-analysis of 27 randomized placebo-controlled trials pointed to an increase in cardiovascular risk in men using testosterone therapy, although the results were statistically insignificant.366 A total of 2,994 men were randomized to either testosterone (n=1,773) or placebo (n=1,261) for 12 weeks to 3 years. There was a total of 115 cardiac events in men on treatment and 65 in the placebo arm (OR=1.54; CI 1.09, 2.18). Included in these events were 33 deaths, 22 of which were in men who were on testosterone therapy, and 11 in the placebo groups. A stratification of trials according to funding showed that those supported by pharmaceutical companies (n=13) showed decreased odds of having a cardiovascular event in testosterone patients (OR=0.89; CI: 0.50, 1.60), while those not funded by industry showed that the risk of a cardiovascular related event on testosterone therapy was greater (OR=2.06; CI: 1.34, 3.17). Despite the homogenous nature of the trials included, it was noted that there was a risk of publication bias since it is possible that trials favoring testosterone therapy might remain unpublished. Other limitations included the possible subjective nature in reporting some adverse events.

Conversely, the Shores, 367 Muraleedharan,233 and Baillargeon373 studies determined that there was no increased risk of MACE in men who were on testosterone therapy. The Shores study was an observational study of 1,031 men (mean age 62.1 years) with total testosterone level <250 ng/dL (mean 181 ng/dL). Over a mean period of 41 months, 398 were reported to be on testosterone therapy, while 631 were not. Mortality in testosterone treated men was 10.3% (mortality rate of 3.4 deaths/1,000 person-years) compared with 20.7% (5.7 deaths/1,000 person-years) in men who were not on treatment (p<0.0001). After adjustment for confounding factors, testosterone therapy remained associated with a decreased risk of death (HR: 0.61; CI 0.42, 0.88).

The Muraleedharan study looked at men with type 2 diabetes and stratified the population (n=581; mean age 59 years) into those who had normal testosterone levels (>300 ng/dL, n=343) and low testosterone levels (<300 ng/dL, n=238). Of the men with low testosterone, 64 received treatment via gel or IM injections for 42 months and were followed for nearly 6 years. At the 6-month time point, there were 34 deaths from CVD (17 in each group) and an overall death rate of 17.2% and 9% in the in the low and normal testosterone groups, respectively. When comparing the likelihood of survival, there was a significant decrease in survival in the low testosterone group compared to those with normal testosterone (adjusted HR=2.02; CI: 1.2, 3.4; p<0.01). Furthermore, a comparison of patients on treatment compared to those who were not showed that the HR for decreased survival was 2.3 (CI: 1.3, 3.9; p<0.01). The increases in mortality were found to be independent of age, BMI, pre-existing CVD, current smoking status, and statin therapy. The authors conceded that those patients treated had more severe testosterone deficiency, which may have resulted in treatment bias.

Finally, the Baillargeon study showed that testosterone therapy was not associated with an increased risk of myocardial infarction (HR=0.84; CI: 0.69 -1.02). The authors conducted a retrospective analysis of 6,355 Medicare beneficiaries who had at least 1 testosterone injection (mean number of injections over the entire study period 8.2) and matched them to 19,065 men who were testosterone therapy naïve for the preceding 12 months. Although confounders were accounted for in the analysis, concurrent medications that may have reduced the risk for myocardial infarction or other testosterone therapies used outside of the study protocol were not controlled for or assessed.

Since the FDA warning in 2015, other studies have failed to demonstrate a risk of cardiovascular events in patients on testosterone therapy. The T-Trials (2017)229 randomized 790 men (mean age 72 years) to either testosterone gel (n=395) or placebo (n=395) for 1 year to measure the effect of testosterone on physical and sexual function and patient vitality. Over half of the men were obese (BMI >30), and 70% had documented hypertension at baseline. At the end of the year-long treatment period, two men from the treatment arm had a definite myocardial infarction, and none were recorded in the placebo arm. During the subsequent year of follow-up, eight men from the placebo group and one man who had been on treatment were adjudicated to have had a definite myocardial infarction.  Although the study was not powered to detect cardiovascular events as a primary endpoint, the authors did not detect increased risk in the testosterone group.
 

Abbreviations

AACE    American Association of Clinical Endocrinologists
ADAMAndrogen Deficiency in the Aging Male
ADTAndrogen deprivation therapy
AIAromatase inhibitor
AMSAging Male Survey
ASCVDAtherosclerotic cardiovascular disease
AUAAmerican Urological Association
BMDBone mineral density
BMIBody mass index
CDCCenters for Disease Control
CVCoefficient of variation
CVDCardiovascular disease
DEXADual-energy X-ray absorptiometry
E2Estradiol
EDErectile dysfunction
EMASEuropean Male Aging Study
FSHFollice stimulating hormone
HbHemoglobin
HctHematocrit
hCGHuman chorionic gonadotropin
HDLHigh-density lipoproteins
IIEFInternational Index of Erecticle Function
IMIntramuscular
KSKlinefelter syndrome
LCMSLiquid chromatography/mass spectrometry
LHLuteinizing Hormone
LTBFLow-trauma bone fracture
MACEMajor adverse cardiac event
MMASMassachussetts Male Aging Study
PINProstatic intraepithelial neoplasia
POMEPulmonary microemboli
QoLQuality of life
RARheumatoid Arthritis
RCTRandomized controlled trial
RPRadial prostatectomy
RRRRelative risk ratio
RTRadiation therapy
SERMSelective estrogen receptor modulator
SHBGSex hormone-binding globulin
SQSubcutaenous
T:ETestosterone:estradiol count
TMSCTotal mobile sperm count
TOMTestosterone in Older Men
VTEVenothrombolic event

Disclaimer

This document was written by the Evaluation and Management of Testosterone Deficiency Guideline Panel of the American Urological Association Education and Research, Inc., which was created in 2016. The Practice Guidelines Committee (PGC) of the AUA selected the committee chair. Panel members were selected by the chair. Membership of the Panel included specialists in urology, cardiology, family medicine, and psychology with specific expertise on this disorder. The mission of the Panel was to develop recommendations that are analysis-based or consensus-based, depending on Panel processes and available data, for optimal clinical practices in the treatment of muscle-invasive bladder cancer. Funding of the Panel was provided by the AUA. Panel members received no remuneration for their work. Each member of the Panel provides an ongoing conflict of interest disclosure to the AUA. While these guidelines do not necessarily establish the standard of care, AUA seeks to recommend and to encourage compliance by practitioners with current best practices related to the condition being treated. As medical knowledge expands and technology advances, the guidelines will change. Today these evidence-based guidelines statements represent not absolute mandates but provisional proposals for treatment under the specific conditions described in each document. For all these reasons, the guidelines do not pre-empt physician judgment in individual cases. Treating physicians must take into account variations in resources, and patient tolerances, needs, and preferences. Conformance with any clinical guideline does not guarantee a successful outcome. The guideline text may include information or recommendations about certain drug uses ('off label') that are not approved by the Food and Drug Administration (FDA), or about medications or substances not subject to the FDA approval process. AUA urges strict compliance with all government regulations and protocols for prescription and use of these substances. The physician is encouraged to carefully follow all available prescribing information about indications, contraindications, precautions and warnings. These guidelines and best practice statements are not in-tended to provide legal advice about use and misuse of these substances. Although guidelines are intended to encourage best practices and potentially encompass available technologies with sufficient data as of close of the literature review, they are necessarily time-limited. Guidelines cannot include evaluation of all data on emerging technologies or management, including those that are FDA-approved, which may immediately come to represent accepted clinical practices. For this reason, the AUA does not regard technologies or management which are too new to be addressed by this guideline as necessarily experimental or investigational.