New Cushing’s Saliva Test Does Not Meet Standard Methods, Study Shows

Steve Bryson, PhD avatar

by Steve Bryson, PhD |

Share this article:

Share article via email
saliva test

An alternative diagnostic test to measure the levels of late-night cortisol in the saliva of patients with Cushing’s disease did not meet the standards of currently used methods, a study demonstrates. 

The study, “Prospective evaluation of late-night salivary cortisol and cortisone by EIA and LC-MS/MS in suspected Cushing’s syndrome,” was published in the Journal of the Endocrine Society.

Cushing’s disease is characterized by an increase of cortisol levels in the bloodstream and saliva, a condition known as hypercortisolism. It is caused by the release of too much adrenocorticotropic hormone (ACTH), which usually is caused by a tumor on the pituitary gland.

The condition can be challenging to diagnose because its symptoms develop gradually and are similar to those of Cushing’s syndrome, which refers to any number of symptoms caused by hypercortisolism. 

One diagnostic strategy is the late-night salivary cortisol (LNSC) test. As cortisol levels typically drop in the evening, samples are collected in the late evening. Samples from a healthy person will be low in cortisol, whereas levels will be high in a person with Cushing’s.

The standard method to measure LNSC is an enzyme immunoassay (EIA-F), in which antibodies that specifically bind to cortisol are used. 

An alternative validated technique to measure cortisol is called liquid chromatography-mass spectrometry (LCMS), in which the substances in a liquid are first separated using liquid chromatography, followed by mass spectrometry, which identifies and measures each substance separately. 

Researchers based at the Medical College of Wisconsin designed a study to evaluate the potential of measuring late-night cortisol by LCMS (LCMS-F) in comparison to the standard EIA test. Levels of the cortisol metabolite cortisone also were tested (LCMS-E) as an enzyme in the salivary gland can convert cortisol to cortisone.  

The team tested saliva samples previously collected over one year and extracted clinical data from medical records.  

A total of 1,453 late-night salivary samples were collected from 705 patients. Of those, 501 samples from 283 patients had at least one test with elevated cortisol. 

The remaining 952 samples were from 422 patients who had no elevated salivary cortisol result, and of these, 391 did not have tumor-related hypercortisolism. From those 391 patients, 121 were selected randomly to be a control group as a comparison. 

Of the 283 patients with elevated cortisol, 35 were confirmed with Cushing’s disease (ACTH-dependent), and five had high levels of ACTH caused by tumors elsewhere in the body (ectopic ACTH). An additional five patients had hypercortisolism caused by adrenal Cushing’s syndrome (ACTH-independent) triggered by a tumor in the adrenal gland. 

In patients with Cushing’s disease the test performance showed that the EIA test had the highest sensitivity (97.5%). The sensitivity of LCMS to measure cortisol was 67.5%, and 92.5% for cortisone. 

In contrast, the EIA test had the lowest specificity (69.3%), whereas the specificity in measuring cortisol by LCMS was 84.6%, and 76.2% with cortisone. 

In patients with ACTH-independent Cushing’s syndrome, the EIA test had a sensitivity of only 31.3%. The sensitivity of LCMS to measure cortisol was 25%, and 18.8% for cortisone. 

Regarding specificity, the results followed the same trend as those in patients with Cushing’s disease.

A detailed evaluation of the 35 patients with Cushing’s disease showed fewer abnormal measurements using EIA compared to LCMS. Furthermore, the percentage of negative tests in patients with confirmed Cushing’s disease (false-negatives) was 3% using the EIA test, but 34% measuring cortisol by LCMS. False-positives for cortisone was similar to EIA. 

Finally, a comparison of 23 patients with persistent Cushing’s disease after pituitary surgery to 12 patients who did not have surgery found lower levels of cortisol by either method, although still higher than control subjects. 

“LNSC measured by EIA is a sensitive test for ACTH-dependent Cushing’s syndrome but not for ACTH-independent [Cushing’s syndrome],” the researchers wrote. 

“We suggest that neither LCMS-F nor LCMS-E improves the sensitivity of late-night EIA-F for [Cushing’s syndrome], they concluded.