New testosterone assays may become gold standard

Testosterone testing methods are finally providing sensitive and accurate measurements.

In recent years, the focus on female testosterone levels has risen dramatically. This focus has been hampered, however, by the fact that available testosterone assays have been neither sensitive nor accurate enough to assess testosterone levels in women.

Advances in research have mirrored the increased interest in and the clinical importance of androgen deficiencies in both men and women, and new tests have emerged that can accurately test testosterone levels in women, children and androgen-deficient men.

William Rosner, MD, was co-author of a position statement on testosterone testing. Photo by Najlah Feanny

Endocrine Today has interviewed top researchers and academics in the field of testosterone assay and androgen deficiency research, including William Rosner, MD, a professor of medicine at Columbia University in New York and a coauthor of a recent Endocrine Society position statement on testosterone testing.

Before discussing the methods of measuring testosterone (T), however, it is important to note that total testosterone (TT) does not tell the whole story.

In women, approximately 1% to 3% of testosterone is free in the blood, and it is this free testosterone that is generally considered biologically active.

About 65% to 75% of the remainder of T is bound to sex hormone-binding globulin (SHBG), and 25% to 35% is bound to albumin. Measurements of TT are significantly easier than direct measurements of free T, and calculations can be used to determine free T from measurements of TT, SHBG and albumin.

In fact, Margaret Wierman, MD, professor of medicine at the University of Colorado Health Sciences Center in Denver, said that measuring free T is inaccurate unless it is done via equilibrium dialysis. This method, however, requires a sensitive and specific TT assay as a baseline.

The most widely used testosterone assays in recent years have been radioimmunoassays (RIA). This technique involves the use of a radioactively labeled antigen bound to an antibody. When a patient sample is mixed with this radioactively labeled antigen, the amount that is displaced can be measured to give the testosterone level.

There are a number of advantages to this method. It is a relatively simple technique, and it can be automated resulting in a very fast turnaround time. It is also inexpensive, and depending upon the specific assay that is used, it is fairly accurate for levels above 300 ng/dL, the levels generally found in healthy men.

However, there are downsides to RIAs. Concentrations of T can often be overestimated, especially when levels sink below the 300 ng/dL mark. Normative ranges also vary based on the specific assay used. Plus, the technique produces radioactive waste.

RIAs can also be used following extraction of T molecules and chromatography. This method has been used extensively, and it has well-documented reference intervals in a number of populations. Also, the sensitivity of the assay is increased by the ability to use relatively large serum volumes.

A method related to RIA is called enzyme-linked immunosorbent assay, or ELISA. This technique removes the radioactivity from the method, instead utilizing enzyme reactivity as the measurable signal.

Tandem mass spectrometry

William Rosner

The newest method, which may soon become the standard assay, is tandem mass spectrometry following liquid chromatography (LC). Tandem mass spectrometry involves multistage mass spectrometry and is thus abbreviated MS/MS. Tandem mass spectrometers can quantify the amount of a compound in a sample by monitoring the mass-charge ratio of parent and product ions of T and determines the quantity of T in the sample. Rosner said that this method improves both the sensitivity and the accuracy compared with immunologic methods.

“With mass spectrometry, at the same time as allowing us to quantitate what is in the sample, it allows us to look precisely at what it is we’re measuring,” he told Endocrine Today.

“Because all the immunological methods are based on antibody specificity, there is always some cross-reactivity and interference. But because mass spectrometry can tell the difference between less than one mass unit, we are looking at precisely what we are measuring. This hasn’t come about until recently because the technology hadn’t afforded the sensitivity needed, but we’ve now reached that level,” although comparisons among various MS methods are still lacking, Rosner said.

Mark M. Kushnir, MS, senior scientist at ARUP Laboratories’ ARUP Institute for Clinical and Experimental Pathology in Salt Lake City, Utah, developed one of a growing number of LC-MS/MS assays. He said that his method uses very small sample volume – one-tenth of a milliliter; such a small sample size is beneficial in testing samples from children.

Mark M. Kushnir

Children, depending on Tanner stage, have even lower concentrations of T than women. Kushnir also noted that although LC-MS/MS tends to be slightly more expensive than RIAs, the cost is not prohibitive.

“This method has already become standard practice,” he said in an interview with Endocrine Today.

Questions remain, however, regarding whether or not LC-MS/MS is actually ready to be the standard for T testing. Rosner said that it is only in the last year or two that use of the technique has begun to skyrocket, and there are a number of steps to be taken before any new assay can be considered the gold standard.

“In theory this should work, but there is a lot of work that needs to be done,” he said. “The Endocrine Society and the Centers for Disease Control are working on this in order to make sure that LC-MS/MS will live up to its billing.”

The current objective is to standardize the testing and create normal ranges for men, women (pre- and postmenopausal) and children. “Before this technology, the numbers for women and children were down around the noise level of the available tests,” Rosner said.

“Normal ranges for this method will have to be established so we can trust that the range in California will be the same as that in New York. There will be a centralization of standards, and then we can move forward. I do, however, anticipate this becoming the gold standard,” Rosner said.

Karen K. Miller, MD, an assistant professor of medicine at Harvard Medical School and faculty in the Neuroendocrine Unit at Massachusetts General Hospital in Boston, agreed that establishing normal ranges is a top priority in this field. “If you send a sample to a commercial laboratory, it will come back with a normal range for women,” she said.

“But that range will most likely be based on a small number of women, and it won’t be an accurate range. We need a large study to determine which assays are valid and what the normal ranges are. In establishing these ranges we need to take into account the time of day that samples are taken and the phase of the menstrual cycle, as T and free T increase by about 20% to 30% at mid-cycle,” she said.

Even among men, the normal ranges are not completely standardized. Traditionally, normal was considered to be between 300 ng/dL and 1,000 ng/dL, but some commercial laboratories differ on this, considering ranges down to 120 ng/dL as normal.

With women it becomes even more complicated. As Wierman said, old assays could potentially have been sensitive down to a range of 20 ng/dL, but most normal premenopausal women (and virtually all postmenopausal women) have T levels of below 20 ng/dL.

“The hope is that these new assays will be more sensitive and more specific, but only time will tell,” Wierman told Endocrine Today. “It is a frustrating issue. As endocrinologists, we often make diagnoses and form treatment strategies based on a careful history, a physical exam as well as sensitive laboratory data.

“We’re used to having very specific and sensitive assays, and lacking such assays has made it very difficult to make sure that we are treating the appropriate patients and targeting them to goals without excess or deficiency,” she said.

Hypogonadism and androgen deficiencies

With so much focus on improvements in assays and establishing normal ranges, the question arises: Why is testing for T levels important?

In men, low levels of T have been linked to sexual function and aspects of body composition for many years. The clinical implications of low T levels in women, however, are much less straightforward. Several studies have found that T replacement therapy can provide benefit to women with regard to sexual function and libido, but controversy remains over a definition of androgen deficiency.

“We are increasingly focusing on women who might be appropriate candidates for low-dose T replacement therapy, but if we cannot accurately determine who is deficient and to what levels the therapy restores T, this is clearly a problem,” Miller said.

“There are a few situations in which T replacement may be indicated and beneficial – such as in women who have had bilateral oopherectomy – but there are three caveats: The long-term safety has not been established, there is no FDA-approved T preparation for women in this country, and sexual dysfunction and decreased libido are multifactorial.

“T is certainly not a cure-all, and its effects on other endpoints, including aspects of brain function and body composition, are still active areas of research. We don’t have the answers yet,” Miller said.

Establishing sensitive and accurate assays with standardized normal ranges is clearly an important early step in moving toward the research goals Miller outlined. Christina Wang, MD, professor of medicine at UCLA School of Medicine and co-developer of another LC-MS/MS assay, noted that diagnoses of hypogonadism rely on both low T levels and presentation of symptoms.

Hypogonadal levels of T, even in men, can fall well below the thresholds at which RIAs lose accuracy, so improved assays are crucial. Once the diagnosis is made, Wang said, “you can improve the symptoms, especially in younger men. This includes sexual function, increases in bone mass and [other improvements].”

Rosner said that while the evidence for benefit with T therapy exists for only certain women to this point, some physicians now think that T therapy will be useful for all types of sexual dysfunction and that there is already evidence for so-called T-deficiency syndrome.

“There can’t be a T-deficiency syndrome unless you can show T deficiency,” Rosner said. “One of the things driving the need for good T assays is the fact that therapies are now available; but how do you make a decision about therapy if you can’t measure what is in a patient’s blood? There is a concatenation of forces that has pushed the field forward — the world has finally caught up.”

The future and clinical recommendations

What are the next steps to be taken in the field? The establishment of normal ranges for a number of different populations is an important step, but Wang said that the LC-MS/MS technique will continue to be improved.

“The trend will be to increase sensitivity of the assay and to decrease the volume needed to test,” she said. Also, the same technique may start to be used for other hormones and small molecules.

On the clinical side, large studies must be undertaken to determine who benefits from T replacement therapy and what the long-term effects are. Wierman said that clinicians need to use well-established assays and simply be aware that limitations still exist.

“Know your assay, be aware if it is among the better ones,” she advised. “And stay tuned. There is something coming down the pike that will be very good. Things move fast, and the commercial assays will be compared to research assays, and soon enough we will know which are the best assays for clinicians in practice to use.” – by Dave Levitan

Is there reliable evidence that female androgen deficiency syndrome exists?

For more information:

• Bhasin S. Female androgen deficieny syndrome–an unproven hypothesis. J Clin Endocrinol Metab. 2005;90:4970-4972.
• Kushnir MM, Rockwood AL, Roberts WL, et al. Performance characteristics of a novel tandem mass spectrometry assay for serum testosterone. Clinical Chemistry. 2006;52:120-128.
• Rosner W, Auchus RJ, Azziz R, et al. Utility, limitations, and pitfalls in measuring testosterone: an Endocrine Society position statement. J Clin Endocrinol Metab. 2007;92:405-413.
• Wang C, Catlin DH, Demers LM, et al. Measurement of total serum testosterone in adult men: comparison of current laboratory methods versus liquid chromatography-tandem mass spectrometry. J Clin Endocrinol Metab. 2004;89:534-543.
• Wierman ME, Basson R, Davis SR, et al. Androgen therapy in women: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2006;91:3697-3710.