Development and validation of methods for the determination of copper and iron in serum of dogs with canine visceral Leishmaniasis using multivariate optimization and GF AAS

In this work efficient methods to determine copper and iron in dog serum samples by graphite furnace atomic absorption spectrometry were developed. The samples were diluted at a 1 : 9 or 1 : 19 (for Cu and Fe, respectively) ratio with 1% (v/v) nitric acid containing a 0.1% Triton X-100 solution. Rhodium for copper and no modifier for iron proved to be the better permanent modifier. Optimization included fractional factorial planning using Pareto and the CCD designs. For both analytes the working linear range was 0-100 mg L-1 (r(2) > 0.99). The obtained LOQ was 19.3 +/- 2.8 mu g L-1 for Cu and 16.5 +/- 0.2 mu g L-1 for Fe. Aqueous and matrix matching calibration curves had average angular coefficients that were not statistically different, i.e. the matrix effect was absent for both analytes. The accuracy was checked by recovery tests with an average of 101 +/- 4% (n = 45) for Cu and 90 +/- 3% for Fe (n = 45). The certified reference material Seronorm (TM) Trace Elements Serum L-1 obtained was 1705 +/- 20 mg L-1 for Cu (certified: 1691 +/- 84 mu g L-1) and 1.40 +/- 1.22 mg L-1 for Fe (certified: 1.43 +/- 0.08 mg L-1). Cu and Fe levels in 39 canine serum samples - 9 uninfected, 19 symptomatic and 11 asymptomatic dogs naturally infected with Leishmania chagasi - ranged from 374 to 913 mu g L-1, and 1103 to 4260 mu g L-1, respectively. Cu was higher in symptomatic than in asymptomatic and control dogs, and Fe was significantly lower in symptomatic than in either asymptomatic or control dogs.