Monitoring of Congenital Adrenal Hyperplasia by Microbore HPLC–Electrospray Ionization Tandem Mass Spectrometry of Dried Blood Spots

Congenital adrenal hyperplasia (CAH), a disorder caused by a deficiency of the 21-hydroxylase enzyme, is the most common inborn error of the adrenal steroid pathways. Early diagnosis of CAH can be lifesaving, and screening for CAH in newborns by measuring 17α-hydroxyprogesterone (17OHP) or other steroids has become a routine part of many programs (1)(2). These steroid hormones have been measured by fluorometry (3)(4), immunoassay (5)(6)(7)(8), and HPLC (4)(9)(10). Most methods are affected by interferences or cross-reactivity with other steroids. Currently, neonatal screening and monitoring for CAH use immunoassays (3)(4). This approach, although practical, lacks specificity because cross-reacting congeners are inseparable from 17OHP in the direct assay (4)(11)(12)(13). Electrospray ionization (ESI) has become an important method for the generation of gas-phase ions from biomolecules for mass spectrometric analysis, but the low proton affinity of natural steroids compromises their measurement by ESI. To improve sensitivity, we have derivatized steroids to form a covalent bond containing a permanent positively charged nitrogen atom. The carbonyl compound 17OHP was derivatized with a quaternary ammonium salt, Girard reagent P (GirP), to form water-soluble hydrazones with a permanently charged pyridine moiety. This derivative was selected for its introduction of a positive charge into the molecule of ketosteroid 17OHP and for the ease of its synthesis (14). The purpose of this study was to evaluate the applicability of liquid chromatography-tandem mass spectrometry (LC-MS/MS) to clinical analysis of 17OHP in dried filter-paper blood samples from patients with CAH caused by 21-hydroxylase deficiency. Although others have proposed the detection of several corticosteroids by LC-MS/MS (15)(16)(17)(18), large blood or urea sample volumes were needed, and published results of clinical LC-MS/MS analysis of steroids in whole blood are lacking. Glacial acetic acid, 17OHP, 6α-methylprednisolone (6MP), GirP, and related compounds were purchased from Sigma. HPLC-grade methanol and acetonitrile were obtained from LAB-SCAN Analytical Science (Labscan Ltd.). Blank human whole-blood samples were obtained from China Medical College Hospital (Taichung, Taiwan). Standardized filter-paper forms (Standardized S&S 903 filter paper; Schleicher & Schuell) impregnated with whole capillary blood from CAH patients or 2- to 5-day-old infants were collected from the Department of Genetics, China Medical College Hospital (Taichung, Taiwan). Patients with confirmed CAH were between 1 and 14 years of age (three girls and one boy). All patients or their parents gave informed consent. The National Taiwan University Hospital (Taipei, Taiwan) kindly provided five dried filter-paper blood samples from CAH infants. Samples were prepared from blood spots by simple solvent extraction. Four 3.175-mm (⅛-inch) circles from each blood spot (equivalent to 11.5 μL of whole blood) were excised from a 12.7-mm (½-inch) diameter dried-blood spot and placed in a flat-bottomed 96-well block automatically (individual 250-μL wells; Corning Incorporated) by a DELFIA DBS puncher (Wallac). A stock solution of extraction solvent (methanol) containing a known concentration of internal standard (50 μg/L 6MP) was prepared and added to each well (200 μL). The wells were capped and shaken on a Vibromix 203E flatbed shaker (Tehtnica Co.) for 50 min. Subsequently, the extracts were transferred, using a multichannel pipette, into a clean V-bottomed 96-well microplate (individual 220-μL wells; Corning). Each 96-well microplate was placed in an evaporator [Techne (Cambridge) Ltd], and the solutions were evaporated to dryness under a gentle stream of dry nitrogen. The residue in each well was derivatized with 160 μL of GirP solution (10 g/L in ethanol, 1 mL/L trichloroacetic acid as a catalyst), incubated at 65 °C for 50 min, and evaporated to dryness under a gentle stream of dry nitrogen. The GirP-derivatized 17OHP (GirP-17OHP) and 6MP (GirP-6MP) were reconstituted in 30 μL of 500 mL/L acetonitrile. The plate was covered with aluminum foil and placed on an autosampler tray for microbore HPLC-ESI-MS/MS analysis. The HPLC system consisted of two Perkin-Elmer Series 200 micropumps (PE-Sciex). HPLC analysis was performed in a 5-μm C4 microbore (Vydac) column [50 × 1.0 mm (i.d.)] operated at ambient temperature. A guard column (C4 cartridge; Vydac) was used to prolong the life of the HPLC column. The mobile phase was water–acetonitrile (50:50 by volume), and the flow rate was 50 μL/min. The autosampler was a Perkin-Elmer Series 200 autosampler fitted with a 10-μL loop (PE-Sciex) and equipped with a 96-well sample plate stack. We used an API 2000 bench-top triple quadrupole mass spectrometer (PE-Sciex) operated in ion evaporation mode with a TurboIonSpray ionization probe source (operated at 5 kV). The TurboIonSpray ionization probe was operated with the turbo gas on (5 L/min; sensor temperature, 300 °C). The collision energy (Q0-RO2) was varied from −30 to −40 V. The orifice (OR) and ring (RNG) voltages were set at 50 and 360 V, respectively. Sample control (Ver. 1.4), TurboQuant (Ver. 1.0), and Microsoft Excel (Ver. 6.0) were used for data processing and statistical analysis. Background subtraction and a three-point smoothing algorithm were applied to all ion chromatograms and viewed using MultiView (Ver. 1.4) software. Immunoassay of 17OHP was carried out with the IMMULITE analyzer (DPC), and the procedures for preparation, setup, dilutions, adjustment, assay, and quality control procedures as given in the IMMULITE operator’s manual (19). For GirP-17OHP and GirP-6MP, the [M]2+ (m/z 299 and 321) ion was the most abundant in full-scan mode. The derivatized steroids had 10-fold higher ESI-MS/MS sensitivity than did the underivatized steroids because the precharged property of GirP-derivatized steroids (data not shown). We used multiple reaction monitoring, with Q1 transmitting the [M]2+ parent ions and Q3 monitoring the daughter ion signals. Under optimal isocratic conditions (acetonitrile–water, 50:50 by volume), GirP-17OHP and GirP-6MP were observed by LC-MS/MS. The monitored reactions and multiple reaction monitoring chromatograms for 17OHP and 6MP in dried-blood specimens are shown in Fig. 1. Mass chromatograms of GirP-17OHP and internal standard GirP-6MP (10 ng in 30 μL of mobile phase) obtained from dried-blood specimen (containing 500 μg/L 17OHP). Instrument in multiple reaction monitoring mode. Mobile phase: H2O–acetonitrile (50:50 by volume). Because no buffers were used in the mobile phases, ∼300 samples were analyzed without instrument cleaning. Each run required 3 min, and intersample time delay was not needed. The retention times of the analytes were highly reproducible (CV = 2.6%; mean = 1.50 min; n = 50 for 17OHP), indicating that chromatographic stability was not sacrificed by either the lack of buffer or the short column equilibration. We replaced the guard column after analysis of 200–300 samples. The column pressure for the microbore HPLC column remained normal in the present study (400∼500 samples). The R2 of the calibration curve (30–500 μg/L) was 0.991–0.997 (median, 0.995). The limit of detection was 10 μg/L (∼12 μL of whole blood) based on a signal-to-noise ratio of 3. Analytical recovery of added 17OHP was 76–82% (30–250 μg/L added). CVs for dried-blood spots of uniform size under these conditions were <7%, suggesting that methanol is a suitable eluant for removing 17OHP from blood spots. The interassay CV for 17OHP was 4.3–10%, and the intraassay CV was <12%. The results of the quantification of 17OHP from dried-blood spots in the control and CAH groups by LC-MS/MS and RIA are shown in Table 1. 17OHP in all dried-blood samples could be quantified by immunoassay. In the control groups, measured 17OHP was 3–12 μg/L in the RIA (median, 6 μg/L). In the LC-MS/MS analyses, 17OHP was <20 μg/L in the control groups, although the internal standard (6MP) was readily detected in all cases. 17OHP concentration in dried-blood spots from healthy controls and CAH patients by LC-MS/MS and RIA. Values in parentheses are for dried blood spots obtained after therapy with hydrocortisone. 17OHP concentration in dried-blood spots from healthy controls and CAH patients by LC-MS/MS and RIA. Values in parentheses are for dried blood spots obtained after therapy with hydrocortisone. We used LC-MS/MS to monitor the treatment of four CAH patients (Table 1, patients 1–4). Repeated blood sampling on filter paper can assist in improving the monitoring of CAH treatment. After treatment, 17OHP decreased in all patients. 17OHP was increased in dried filter-paper blood samples from five CAH infants (Table 1, patients 5–9). The detection of 17OHP by the present method was inferior to that obtained by the immunoassay, but was adequate for determining 17OHP in CAH patients. The high-throughput CAH screening method described in this report, although it includes a derivatization step, does not require long and complicated preparation of samples. Measurement of 17OHP in dried blood on filter paper by LC-MS/MS is simple to carry out. The method provides analytical specificity and eliminates handling of radioactive materials. The LC-MS/MS assay appears useful not only for diagnosis and monitoring of treatment of CAH in all age groups, but also for screening for CAH in infants. The study was funded by a grant from China Medical College Hospital (DMR-91-101). We thank National Taiwan University Hospital for providing five dried filter-paper blood samples from CAH infants.