Abstract
Inductively coupled plasma-mass spectrometry (ICP-MS) is a comparatively new technique which in the last few years has been increasingly used for the analysis of biological samples. The multi-element capabilities of ICP-MS have been demonstrated for the analysis of several biological fluids including serum, whole blood and urine. This thesis evaluates the use of ICP-MS for the routine trace element analysis of cerebrospinal fluid (CSF). Although it is possible to introduce CSF directly into the pneumatic nebulizer of the ICP-MS a wet digestion method, using nitric acid with microwave or water bath heating, has been assessed in view of the safety precautions required when handling CSF (Appendix I). International standard reference material NIST SRM-909 Human Serum was used to test the validity of this procedure. The results obtained were comparable to the certified data for the elements Ca, Cr, Cu, Fe and Mg. Additionally Rb and Zn results agreed with other literature values. Good recoveries (92 to 105%) were obtained for the addition of Ca, Cu, Fe, Mg, Rb, Sr and Zn to CSF. Other elements were below the detection limit in the non-spiked CSF. A preliminary study of 163 CSF samples, obtained from patients undergoing routine lumbar punctures, gave elemental concentrations within the literature range. For multi-element scan mode many of the elements lay below the detection limit: Cd, Co, Sc, Sn and V at concentrations less than 1 ng ml-1 and Hg less than 2 ng ml-1. The reported values for Ba, Cs, Pb and Sb were less than 1 ng ml-1. Elemental concentrations for Al, Cu, Rb, Sr and Zn were in the ng ml-1 range and Mg, P, Ca and Br in the mug ml-1 range. However, as a result of the lumbar puncture procedure only small sample volumes of CSF were obtained (from 0.4 - 1.5 ml). Thus two sample volumes were used in the analysis 0.2 and 1.0 ml. The 1.0 ml sample volumes gave significantly lower elemental concentrations at the 95% confidence limit than the 0.2 ml volumes for the elements Mg, P, Ca, Cu and Br. These differences were not attributable to sample heterogeneity or pipetting errors. As a result of this and due to the necessity of changing ICP-MS instruments during this study the analytical performance of the ICP-MS was assessed. Detection limits at the ng ml-1 level were obtained for most elements. The instrumental accuracy was checked by analyzing NIST SRM 1643b Trace Elements in Water which gave comparable results to the certified data for V, Cr, Mn, Co, Zn, Cu, Cd and Pb. A linear dynamic range from 1 ng ml-1 to 5 mug ml-1 was established for the elements analyzed. It was found that the analytical performance of ICP-MS was impaired by the introduction of matrix elements (Na, K, Ca, Mg, P and Br). Some spectral interferences were caused by the addition of these matrix elements, for example 40Ar23Na+ on 63Cu+. However, suppression and enhancement of the analyte signals were observed depending on instrument operating parameters, such as nebulizer flow rate (NFR). As a result of these studies the optimum NFR for analyzing biological fluids (such as CSF and serum) was determined to be between 700 and 950 ml min-1. Although this NFR range did not provide the maximum analyte signal particularly for the low mass analytes, it did limit the spectroscopic and non-spectroscopic interferences. Preliminary studies in synthetic solutions showed that it was possible to correct for the matrix interferences with internal standards of similar mass to the analyte element. Several methods for correcting spectroscopic interferences were attempted: internal standards; standard addition; sample dilution; isotope dilution mass spectrometry and matrix matched standards. No one method was successful for all elements. Several elements could be determined namely. Mg, Rb, Cu and Sr and others depending on instrument fluctuations, blank contamination levels and their level in a particular CSF sample. These determinations could not be thought of as routine as correction procedures for the data varied from day to day as a result of instrument instability and signal drift, perhaps caused by the sample matrix. This study highlights the difficulty of using pneumatic nebulizer ICP-MS for the routine multi-element analysis of CSF.