In a recent communication, Wille and coworkers110 compared the efficacies of ion exchange sorbents and mixed mode, neutral polymeric and silica-based reverse phase sorbents. The analytes consisted of 13 new generation antidepressants with pKa values ranging from 6.7 to 10.5 and log P values ranging from 0.04 to 7.10. The authors utilized derivatization with hexafluorobutyryl imidazole to facilitate GC/MS quantitation of SPE recoveries and to assess the purity of the extracts. They also utilized HPLC for optimizing SPE and for investigating the protein binding of these antidepressant drugs to plasma proteins. Since water absorption and/or retention by the sorbents is not compatible with derivatization and GC/MS analysis, although Oasis HLB and Focus sorbents retained all the drugs well, they were excluded after initial screening because they are very hydrophilic. The inability to withstand 100% methanol wash for all the drugs tested eliminated silica-based nonpolar sorbents and neutral polymer strata-X and Certify.
Although Certify is a mixed mode sorbent with C8 and sulfonic acid moieties, the authors rationalized that the hydrophobic retention on this sorbent is more dominant and caused the nonretention of certain drugs during methanol wash. The weak WCX ion exchanger was also excluded for similar reasons. Both the mixed mode strata-X-C and the ion exchange sorbent SCX were found to be most amenable for the derivatization-based GC/MS analysis and both yielded pure extracts. However, the yields were consistently lower with strata-XC than SCX and the authors hypothesized that this was due to the inability of the 5% ammonia/methanol eluent to completely disrupt the hydrophobic and dipolar interactions between the analytes and XC.
Wille et al.110 made interesting observations on the protein binding of the 13 antidepressant drugs investigated. These drugs were divided into two groups—one consisting of desmethylmirtazapine, O-desmethylvenlafaxine, desmethylcitalopram, didesmethylcitaloporam, reboxetine, paroxetine, maprotiline, fluoxetine, norfluoxetine, and m-chlorophenylpiperazine. The other group included mirtazapine, viloxazine, desmethylmianserin, citalopram, mianserin, fluvoxamine, desmethylser-traline, sertraline, melitraen, venlafaxine, and trazodone. They tested protein precipitation by four methods: dilution with (1) pH 2.5 or pH 6.5 phosphate buffer, (2) glycine hydrochloride, (3) 2% phosphoric acid, and (4) organic solvents (methanol and acetonitrile). Since the sorbents used for SPE were cation exchangers, Willie's group did not investigate inorganic salts.
After equilibration at 4°C overnight, the samples were vortexed and centrifuged and the super-natants subjected to SPE on SCX and strata-X-C. HPLC analysis of eluates indicated that glycine HCl and dilution with acidic phosphate buffer yielded 89 and 88% recoveries, which result was interpreted on the basis of negligible drug binding by a-1-acid glycoprotein of the plasma (isoelectric point 3.0) at pH 2.5 for both reagents. On the other hand, the lower recoveries for acetonitrile (62%) and methanol (78%) were interpreted as arising from the hydrophobic binding of the drugs to albumin and lower solubility of the drugs in acetonitrile. Phosphoric acid gave 73% recovery. The importance of load pH and disruption of hydrophobic interactions while using ion exchange and mixed mode sorbents is thus emphasized.
Of particular interest is the comparison of the performance of cation exchange and mixed mode sorbents for their efficacy in cleaning up endogenous phospholipids. Unlike the protein-related materials that are eluted in the very early stages of HPLC, these phospholipids elute in the hydro-phobic region and interfere with drug peaks which also elute around the same time.
Shen and coworkers111 compared SPEC SCX disks with SPEC MP1 disks and Oasis MCX. SPEC-SCX is a phenylsulfonic acid, while MP1 is a mixed mode C8/sulfonic acid and MCX is a polymeric sulfonic acid on a divinylbenzene-vinylpyrrolidone polymer backbone. The sorbents were conditioned with methanol and then with 2% formic acid. The sample was loaded in 2% formic acid solution and washing was done with 2% formic acid, followed by acetonitrile:methanol (70:30). Analytes were eluted with two aliquots of methanol:acetonitrile:water:ammonia (45:45:10:4% v/v/v/v). The eluent was dried under nitrogen and the residue reconstituted in the mobile phase (80% 10mM ammonium formate containing 0.2% formic acid and 20% 10mM ammonium formate in methanol with 0.2% formic acid). Their data on desloratadine and its 3-hydroxy analog (see Figure 1.7), along with data on phosphatidylcholine indicates that MCX retains about seven times as much phospholipids as SCX does and MP1 retains around 60 times more than SCX (see Figure 1.8). Post-column infusion experiments with blank plasma extracts showed ion suppression in the hydrophobic region for MP1 and MCX, but not for SCX (see Figures 1.9 through 1.11). The observations were rationalized through hydrophobic retention of the phospholipids by the mixed mode sorbents; SCX did not exhibit such retention mechanisms.
Was this article helpful?