The Estimation of Lithium in Serum

Lithium was first used therapeutically for the treatment of gout (Daniels, 1914), but more recently it has been used in the treatment of certain mental disorders (Cade, 1949; Schou et al., 1954; Wharton and Fieve, 1966; Rice, 1956). Among the psychiatric disturbances reported to respond favourably to lithium therapy are manic depressive psychosis, repeated manic attacks, recurrent depression and Parkinsonism. The degree of success attributed to lithium therapy has been the subject of some controversy (Lancet, 1969; Whybrow, 1969; Goodwin et al., 1969) and recently it has been recommended that lithium be used only in those cases of mania which are resistant to phenothiazines and electro-convulsive therapy. Current medical opinion favours more rigorously controlled therapeutic trials and since it is the practice to regulate dosage with reference to serum lithium assays it seemed important to re-examine the technical problem of lithium estimation in serum. Preferably the therapeutic level should be maintained below 1.5 m-equivfl since toxic effects have been observed when the level is between 3.0 and 6.0 m-equiv/l (Amdisen, 1967). The need for accurate estimation was emphasised recently following a fatality in which an elevated serum lithium concentration was thought to be the major contributory cause of death (Hawkins and Dorken, 1969). A number of methods for the determination of lithium in serum have appeared (Barrow, 1960; van der Helm and Andriesse, 1961; Nevius and Lanchantin, 1965; Amdisen, 1967, Blijenberg and Leijnse, 1968; Tompsett, 1968; Little et al., 1968; Lehmann, 1968). Many of these use a 1 in 5 or 1 in 10 dilution with trichloroacetic acid (TCA), a reagent which is known to reduce the sensitivity of the assay. Blijenberg and Leijnse (1968) used 96 % ethanol as diluent, thereby increasing the sensitivity of the method and simultaneously decreasing the blank value. Although Amdisen (1967) using a Beckman D.U. atomic absorption spectrophotometer was able to use as little as 200 11.1 of serum at a 1 in 20 dilution, most methods utilise 1 ml of serum diluted 1 in 5. There is some discrepancy in the literature