Utility of Directly Measured Free Testosterone in Predicting Benefit of Testosterone Therapy in Men with Hypogonadal Symptoms and Normal Total Testosterone

Introduction

Hypogonadal symptoms, including decreased libido, erectile dysfunction, and decreased energy and fatigue, often related to a low total testosterone (TT) level, are common among aging men, affecting at least 20% of men age ≥45.^1–3 Testosterone exists in two main forms: bound to proteins (primarily sex hormone binding globulin [SHBG] and albumin) and as unbound free testosterone (FT). Only 2% of the hormone circulates as FT. ⁴

As SHBG increases with age, and because FT varies linearly with SHBG, FT proportionally decreases to a greater extent than TT, as evidenced by older men with low FT who also exhibit hypogonadal symptoms.^5–7

There are several biochemical tests commonly used to investigate testosterone deficiency: TT, FT, calculated FT, bioavailable testosterone, and free androgen index. ⁸ Current guidelines from multiple professional associations advise clinicians to use the presence of hypogonadal symptoms along with low serum TT to determine the need for Testosterone Therapy (TTh).^9,10

The American Urological Association (AUA) guidelines state that the clinical diagnosis of testosterone deficiency is only made when patients have low TT <300 ng/dL in combination with hypogonadal symptoms and/or signs.^9,10 The International Society for Sexual Medicine states that symptomatic men with TT lower than 350 ng/dL should be offered TTh.^10,11

Although these guidelines are helpful in ensuring the proper selection of men who stand to benefit from TTh, they discount the potential utility of TTh in men with low-normal TT with low values in the alternate measures of T status. The use of FT as a stand-alone measure for diagnosis for treatment is not well established. Equilibrium dialysis is the gold standard for measuring FT, but it is uncommonly used clinically due to practical limitations including decreased availability, increased cost, variations in technique, and dependence on the accuracy of TT levels.^5,11 LabCorp uses a electrochemiluminescence immunoassay to measure TT.

Many guidelines also discourage the use of commercially available directly measured FT by analog enzyme-linked immunoassay (EIA) due to variability and accuracy concerns. ¹¹ Thus, in guideline-based practice, symptomatically hypogonadal men with normal TT levels but low EIA-FT are often not offered TTh, despite evidence that these men may benefit from treatment. ⁵ In addition, patients may have varying responses to TTh based on the severity of testosterone deficiency, hypogonadal symptoms, and baseline TT levels. ¹²

In our study, we attempt to examine men's subjective response to TTh and the overall safety of TTh in men with hypogonadal symptoms who demonstrate a normal initial TT level but low directly measured FT using EIA. We hypothesize that hypogonadal men with normal TT, and low EIA-FT will respond well to TTh, and that such treatment will prove safe and efficacious in improving bothersome hypogonadal symptoms.

Patients and Methods

This study prospectively collected data for 86 consecutive men presenting to our specialty urologic clinical practice with hypogonadal symptoms, normal TT (>300 ng/dL), and low FT (<7.9 pg/mL). Exclusion criteria included men with prior TTh use, men with untreated prostate cancer, and men with fewer than 6 months of follow-up. The TT and FT were measured by the “free and total testosterone” serum assay from a single national laboratory (LabCorp), which measures FT using EIA. Subjects included in the study then began a standardized regimen of testosterone cypionate intramuscular injections at a dose of 200 mg every 2 weeks.

Baseline blood tests were performed for TT, FT, estradiol (E2), luteinizing hormone, prolactin (PRL), hematocrit (HCT), and prostate-specific antigen (PSA) before initiation of TTh. Men with previously uninvestigated elevated PSA and/or PRL levels were excluded.

Our cohort had follow-up visits at 3-month intervals. Follow-up blood tests included TT, FT, E2, HCT, and PSA. The population was divided into two groups based on initial TT levels: Group 1: 300–399 ng/dL; Group 2: 400 ng/dL. Clinical response to TTh was based on documentation of patient responses to improvement in hypogonadal symptoms, including erectile dysfunction (ED) as measured using the validated International Index of Erectile Function (IIEF-5) questionnaire, as well as subjective patient reports of changes in libido, energy, and vitality, as asked by a single provider (N.R.S.) in a standard single-question presentation format (“Have you noticed any change in your libido, energy, and/or general vitality?”).

Changes in IIEF over the study period in the whole cohort and within subgroups were compared using Student's t-test. Patient responses to changes in subjective symptom improvements were categorized according to significance of improvement; that is substantial improvement, mild but noticeable improvement, or no improvement. Erythrocytosis was defined as HCT >51.

All research was conducted in accordance with the Helsinki Declaration with regard to patient privacy, consent, and data utilization.

Results

A total of 86 men with a mean age of 62.0 years (range 34–86) were included. Mean age and baseline blood test results of the total patient population are depicted in Table 1. Before TTh, mean TT was 391.7 ng/dL (interquartile range [IQR] [332.8–425.5 ng/dL]) and mean FT was 5.1 pg/mL ([4.2–6.0 pg/mL], normal range 8.0–24.0 pg/mL). At the 3-month follow-up, mean TT was 672.6 ng/dL (407.0–878.5 ng/dL) and FT was 11.6 pg/mL (6.9–14.9 pg/mL). At the 6-month follow-up, mean TT was 990.0 ng/dL (609.0–1425.0 ng/dL) and FT was 26.9 pg/mL (12.8–30.1 pg/mL).

Mean IIEF-5 score before TTh was 12.8 (5.0–19.0) for the entire cohort. At 3 and 6 months after TTh initiation, mean IIEF scores were 14.1 (p = 0.13) and 14.7 (p = 0.07), respectively. Table 2 depicts mean baseline, 3- and 6-month post-TTh TT, FT, and IIEF-5 score for the entire patient population. Within the whole cohort, 79 patients (91.9%) reported subjective improvement in overall energy and libido after 3 months of TTh, with 61 patients (70.9%) reporting substantial improvement and 18 patients (20.9%) reporting more mild but still noticeable improvement. The remaining 6 patients (7.0%) reported no improvement. After 6 months of TTh, 77 patients (89.5%) reported substantial improvement compared with pre-TTh baseline, 7 (8.1%) with more mild but noticeable improvement, and only 2 (2.3%) with no improvement.

Our cohort was further divided into two groups based on initial TT levels. Group 1 (TT 300–399 ng/dL) had 56 patients, and group 2 (TT 400–660 ng/dL) had 30 patients. Mean age and baseline blood results of Groups 1 and 2 are shown in Table 3. Composite results are summarized in Figure 1.

OPEN IN VIEWER

FIG. 1.

Percentage of patients reporting improvement in hypogonadal symptoms aside from erectile dysfunction (e.g., energy, libido, vitality), qualified as substantial improvement, mild improvement, no improvement. TTh, testosterone therapy.

Discussion

In recent decades, there has been controversy over the criteria by which hypogonadism (and therefore eligibility to receive TTh) is determined. TT has always been the gold standard, with each group having slight variations in cutoff values.^9–11 The use of EIA-measured FT has a reputation for substandard accuracy and high variability, ¹¹ despite some evidence showing excellent concordance with equilibrium dialysis. ¹³ Therefore, current guidelines-based practice uses low TT along with hypogonadal symptoms to determine TTh.

However, many men fall into the “borderline” or “low-normal” testosterone category; likewise, current literature and guidelines provide limited guidance on when to treat and the general effectiveness of TTh in this very common patient population. ¹² The AUA states that the clinical diagnosis of testosterone deficiency is only made when patients have low TT <300 ng/dL combined with symptoms and/or signs.^9,10 We decided to utilize the readily available “free and total testosterone” assay at a single U.S. laboratory due to its universal availability, cost effectiveness, ease of interpretation, and the fact that other variables used in many calculations of FT, SHBG in particular, are rarely modifiable. As such, we sought to determine whether the simplest-possible test to order and interpret could demonstrate utility in predicting a symptomatic response to TTh.

Findings from our study demonstrate that most men with hypogonadal symptoms, normal TT, but low EIA-FT stand to benefit from TTh, based on their responses to the IIEF-5 questionnaire and reported subjective improvement in libido, energy, and overall vitality. The IIEF scores notably increased in the entire cohort nearly 2 points above baseline (from 12.8 to 14.7), but not quite to statistically significance (p = 0.07; Table 6). After 6 months of TTh, patients with “borderline” or “low-normal” (TT 300–399) in Group 1 showed a greater improvement in their IIEF-5 score compared with patients in Group 2 with TT >400, 13.8–16.1 (p = 0.07) versus 10.9–11.6 (p = 0.37), though neither met the cutoff value for significance of p ≤ 0.05. We suspect that with more patients being added to the cohort in prospective fashion, these changes in IIEF scores will become significant.

More than 80% of the men in both groups reported varying degrees of improvement in energy, libido, and vitality from mild to substantial in just 3 months and more than 90% of men in 6 months. These results have become a precedent in the literature: A retrospective study by Reyes-Vallejo et al demonstrated that subjective response to TTh correlated with initial TT levels, and men with borderline low T (TT >300) showed the most substantial improvement in erectile function. ¹² More than 70% of the patients with borderline low T also showed a positive response to libido and morning erections. ¹² Similarly, our study suggests that men with borderline low TT would benefit from consideration of initiating TTh, as well as added value in regards to quantitative FT values that can be utilized in the future as triggers for therapeutic initiation.

It should be noted that high levels of SHBG, often seen in aging men with hypogonadal symptoms, may falsely result in normal TT level but low bioavailable FT.^6,7,12 Other factors such as obesity, narcotics, and medical comorbidities also affect SHBG levels.^6,14–16 In a retrospective, cross-sectional study that compared SHBG, FT, and TT levels in obese patients, obese men had lower SHBG and TT concentration than nonobese men, but there were no changes in FT between the two groups. ⁶

This, in part, may explain the differences in IIEF increases between Groups 1 and 2 in that patients who are sicker at baseline, with related alterations in SHBG and TT, are more prone to suffer from organic ED, which is less likely to resolve with TTh only. It also seems logical that the closer a man is to the current definition of hypogonadal (i.e., the lower the starting TT level), the more likely he is to respond to TTh in the way that “truly hypogonadal” men do.

Although there are significant and proven benefits of TTh, such treatment is not without potential side effects; risks include erythrocytosis, venous thrombosis, increase in prostate volume and PSA, breast tenderness, and gynecomastia.^17,18 In our study, 16 patients developed new erythrocytosis by 6 months; as they were symptomatically improved with treatment, all men opted to perform periodic phlebotomy and continue TTh. Three patients developed a pathologic rise in PSA; however, none of the patients developed prostate cancer as evidenced by negative biopsies, demonstrating the overall safety and efficacy of TTh even in this under-studied patient population.

Although there are several exciting findings presented here, our study is not without limitations. An obvious weakness is the very small sample size. However, the prospective design and inclusion of consecutive patients meeting criteria help to minimize bias and confounding. Certain outcome measures, like the IIEF questionaries, are validated in the literature; however, many data were based simply on well-documented subjective patient responses to a standardized line of questioning by a single provider (N.R.S.); these responses, however, are susceptible to recall bias among others. We believe this work represents a brief, concept-proving pilot study requiring further longitudinal follow-up and ideally randomization, as well as a much longer follow-up period.

Conclusion

We hope to contribute to the current literary repository by examining the efficacy and safety of TTh in symptomatic men with low FT measured by EIA but normal TT. Our findings suggest that many men with hypogonadal symptoms, normal TT, and low EIA-FT respond well to TTh, and that such treatment is safe. Noteworthy, but not quite significant, improvement in erection function was demonstrated in the entire group, especially those with lower starting levels of TT (Group 1), after 6 months of TTh. They may represent the key group to benefit from our findings, as they generally would not be offered TTh per current guidelines.

Until the creation of specific guidelines regarding the use of FT to guide TTh treatment decisions, it may be of clinical benefit to treat men based on appropriate symptoms and low directly measured FT levels. This simple lab test simplifies the process of deciding which patients stand to safely benefit from TTh, thereby allowing providers of many specialties to offer TTh options to this otherwise undertreated cohort of men with hypogonadal symptoms but normal testosterone.