Current sample handling methods for measurement of platinum-DNA adducts in leucocytes in man lead to discrepant results in DNA adduct levels and DNA repair.

DNA adduct levels were measured with atomic spectroscopy in white blood cells (WBCs) from patients with solid tumours who were treated with six weekly courses of cisplatin. In 21 patients (I) the WBCs were collected after thawing frozen whole-blood samples according to a previously described method. In 32 other patients (II) WBCs were collected immediately after blood sample collection. The two methods for WBC collection were also compared in vitro. The maximal DNA adduct levels in vivo after the first course were in I 2.48 +/- 1.14 and in II 1.28 +/- 0.40 pg of platinum per microgram of DNA (P < 0.0001). The DNA 'repair' in the first course (DNA adduct level at the end of the infusion minus the level 15 h post infusion) was in I 40% +/- 29% and in II 18% +/- 29% (P = 0.009). These differences were consistent in all measured courses. In vitro, the DNA adduct levels in the freshly prepared WBCs were significantly lower at 0, 1 and 4, but not 24 h, after start of the incubation with cisplatin than in the WBCs collected after freezing and thawing the blood sample. The same experiment with carboplatin in vitro also resulted in significantly lower adducts in freshly isolated WBCs. The higher DNA adduct levels and DNA 'repair' in I are caused by remaining unbound cisplatin in the sample tubes, which can form DNA adducts ex vivo. The same results in vivo can be anticipated when carboplatin is used.