A SENSITIVE ASSAY FOR CYCLOPHOSPHAMIDE IN HUMAN PLASMA UTILIZING MASS-SPECTROSCOPY

This study describes a highly sensitive, solid-phase extraction-HPLC/M S/MS assay for cyclophosphamide in human plasma, with a minimum quanti fiable level of 0.025 mu g mL(-1). Plasma samples (0.5 mL) were mixed with 0.01 M ammonium acetate buffer (1.5 mt, pH 4.9) and internal stan dard (D4 deuterated compound, 50 mu L, 5 mu g mL(-1)), and then extrac ted using Isolute CH(EC) cartridges (100 mg). Contaminants were washed from the cartridges with 0.01 M ammonium acetate:methanol (90:10, 1 m t), prior to the elution of the compounds of interest with 0.1 M ammon ium acetate:methanol (50:50, 300 mu L). An aliquot (200 mu L) of the e luent was injected onto the chromatography system, which consisted of a LiChroCART 4-4 RP-select B, 5 mu m cartridge precolumn, protecting a LiChrospher 60 RP-select B, 5 mu m (250 x 4 mm id) cartridge analytic al column. The mobile phase tvas methanol:0.1 M ammonium acetate buffe r (pH 4.9, 60:40), run at a flow rate of 1 mt, min(-1) Detection was b y mass spectroscopy using a Finnigan MAT TSQ 700 triple stage quadrupo le machine, fitted with an APCI interface. For quantification, the par ent ions at m/z 261 and 265 (cyclophosphamide and D4 internal standard respectively) were selected by quadrupole 1, collisonally dissociated in the octapole, and daughters ions at m/z 120 and 124 then monitored via quadrupole 3. The calibration lines were linear over the range 0. 025 to 1.0 mu g mL(-1), with no evidence of any systematic deviation. The within- and between-day precision and accuracy were examined at fi ve levels (0.025, 0.05, 0.5, 0.8 and 1 mu g mL(-1) n=6), with values o f <14% and within +/-5% respectively. Loss of compound was not observe d (0.05 and 0.8 mu g mL(-1), n=6) after 3 freeze/thaw cycles, storage of the extracted samples for 24 h at room temperature and at 0 to 5 de grees C, and after 4 weeks storage at -15 to -25 degrees C. Diluting s amples from 80 and 10 mu g mL(-1) to within the range of the standard curve was shown not to affect the assay. Although the recovery of cycl ophosphamide and the internal standard was only 20%, the use of the de uterated compound ensured an accuracy and precision within the validat ion criteria. The utility of the assay was demonstrated using plasma f rom a patient who had received an IV infusion of cyclophosphamide at 1 000 mg m(-2), given over 1 h.