an electron capture negative chemical ionization gas chromatography-mass spectrometry ... The ACS testosterone assay is a competitive chemiluminescent immuno-. VA Medical .... reagent gas for ECNCI at an ionizer pressure of 107 ± 40 Pa,.
ClinicalChemistiy 42:5 749-755 (1996)
Automation and Anaytica1 Techniques
Serum total testosterone: immunoassay compared with negative chemical ionization gas chromatography-mass spectrometry L.
ROBERT
We have developed ionization gas MS) procedure ically
relevant
cedure
to
range
assess
testosterone liquid-liquid
an electron
capture
chromatography-mass to quantify serum 0.69-69.3 Ciba
and DAVID A.
FITZGERALD*
negative
chemical
spectrometry (GCtestosterone in the clin-
and used this pro-
nmol/L
Corning
Diagnostics
pentafluorobenzyloxime/silyl ether one with excellent chromatographic properties. The ACS testosterone
ACS:180#{174}
method involves and synthesis of a
inununoassay. The GC-MS extraction of serum samples
derivative of testosterand electron capturing assay is the first fully
automated nonradioactive testosterone immunoassay approved by the US Food and Drug Administration. Patients’
(101, 57 males,
specimens
both techniques. A plot testosterone concentrations + 0.19 nmol/L), showing between
the two assays.
female specimens a poor correlation INDEXING
TERMS:
mentography. cent assay
The majority
44 females)
endocrinology.
of commercially
analyzed
by
of the two assays
Agreement
was poor (y (r2 = 0.31). methods
were
of the GC-MS (x) vs ACS (y) for men was linear (y = 1 .07x excellent correlation (r2 = 0.98) =
0.72x
comparison
+
#{149} steroids.
immunoassays
available
1.2 nmol/L),
for with
mass frag-
#{149} chemilumines-
HEROLD
tive imnmunoassay for the analysis of serum testosterone) The primary advantages of the Ciba Corning ACS:180#{174} are automation, high throughput, and chemiluminescent detection. Because immunoassays can be subject to cross-reactivity with structurally related compounds, it is necessary to have a reference method such as gas chromatography-mass spectrometry (GC-MS) to evaluate these methods. The purpose of this study was to develop a sensitive GC-MS procedure that could be used to validate commercially available assays for serum testosterone. Previous GC-MS methods for the analysis of serum testosterone involved rigorous purification procedures combining liquid-liquid extraction, chromatography, and (or) solid-phase extraction [1-4]. Furuta et al. reported good results for increased serum testosterone concentrations in the 34.7-62.4 nmol/L range with electron impact GC-MS, but did not provide accuracy and precision data in the normal male or female range [4]. Using a simple solid-phase extraction procedure and electron capture negative chemical ionization (ECNCI), Legrand et al. reported good accuracy and precision at concentrations of 8-25 nmol/L in serum and a limit of detection of 3.5 fiuiol per injection of pure reference material /5].However, there are no reported methods capable of accurately quantifying serum testosterone over the entire clinically important range of 0.6969.3 nmol/L. This report details a simplified, highly sensitive procedure for the analysis of serum testosterone by ECNCI GC-MS and uses this methodology to assess the new Ciba Corning ACS: 180 testosterone assay. The GC-MS procedure combines a simple one-step extraction of testosterone from human serum with the synthesis of a derivative that has excellent gas chromatographic and ECNCI mass spectral characteristics. Quantification is based on chromatographic peak areas of testosterone relative to the internal standard, trideuterated testosterone. The ACS testosterone assay is a competitive chemiluminescent immuno-
assays for quantification
of total testosterone in serum are based on competitive binding of radioactive testosterone and endogenous testosterone with testosterone antibodies. Recently, Ciba Corning Diagnostics, a Chiron business, received approval from the US Food and Drug Administration (FDA) for the first fully automated nonradioac-
VA Medical Center- 113, San Diego, CA 92161 and Department of Pathology, University of California-San Diego, La Jolla, CA 92093. * Address correspondence to this author at: VA Medical Center -113, 3350 La Jolla Village Dr., San Diego, CA 92161. Fax 619-552-7479; email rlfitzgerald@ vapop.ucsd.edu. Received October 27, 1995; accepted January 5, 1996.
Nonstandard abbreviations: FDA, Food and Drug Administration; GC-MS, gas chromatography-mass specrrometrv; ECNCI, electron capture negative chemical ionization; BSTFA, .V,O-bis(trimethylsilyl)trifluoroacetamide; 1)1-LEA, dehvdroepiandrosterone; HFBA, heptafluorobutvric anhydride: and PER, ptafluorobenzyl hydro.wlamine hydrochloride.
749
750
Fitzgerald
and Herold:
Serum
assay in which testosterone present in the patient’s sample competes with an acridinium ester-labeled testosterone for binding to polyclonal rabbit anti-testosterone antibody on a solid phase. Serum samples (101, 57 males and 44 females) were analyzed by both GC-MS and by the ACS testosterone assay.
Materials and Methods MATERIALS
Testosterone was obtained from Sigma Chemical Co. (lot no. 50H0345; St. Louis, MO) and the internal standard testosterone-l#{243},16,17-d3 was from MSD Isotopes (lot no. Md-3793; Montreal, Canada). HPLC-grade ethyl acetate was from Burdick and Jackson (Muskegon, MI). HPLC-grade pyridine was purchased from Aldrich (Milwaukee, WI). The derivatizing reagents N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) with 10 mLIL trimethylchlorosilane and Florox [O-(pentafluorobenzyl)hydroxylamine, 2.5 g/L in pyridine] were from Pierce Chemical Co. (Rockford, IL). Commercially prepared testosterone controls for GC-MS were from Bio-Rad Labs. ECS Division (Anaheim, CA) Lyphocheck immunoassay control serum, levels 1, 2, and 3. The following steroids were from Sigma: epitestosterone; dehydroepiandrosterone (DHEA); pregnenolone; 5-androstene-313, 17/3-diol; 4-androstene-3, I 7-dione; progesterone; 5-androstene-3 /3,1 6a-diol- 17-one; 5 a-androstane3/3,1 113-diol- 17-one; I 6-dehydropregnenolone; etiocholane17/3-ol-3 -one; and 16-dehydroprogesterone. All sera were from patients at the Veterans Affairs Medical Center-San Diego or at the University of California-San Diego Medical Center whose treating physician had ordered a total
testosterone as part of their clinical evaluation.
Any additional testing was performed on discard specimens according to our institutions’ human subject committee guidelines. Samples were frozen at