Laboratory detective work identifies a mishandling problem in sample aliquoting.

Data from a recent ovarian cancer biomarker study using serum aliquots from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial Biorepository showed that CA125II concentrations in these aliquots were significantly lower than those previously measured in the same subjects from the same blood draw. We designed an experiment to investigate whether samples used in the study (reference study) were compromised during the aliquoting process. We measured CA125II in the "sister" vials created during the same aliquoting process as the reference study aliquot, and in "cousin" vials newly aliquoted from another parent vial from the same blood draw, from 15 healthy controls in the study. Because the sister vials were created in a specific order, we also assessed whether there was a CA125II concentration gradient among the sisters. The Wilcoxon signed-rank test was used to test the statistical significance of the observed differences. Mean CA125II concentration (volume-averaged) was greater in the sisters than the cousins in all 15 subjects (p<0.001). The mean coefficient of variation was 0.25 (range: 0.12-0.43) in the sisters and 0.11 (range: 0.-1.1) in the cousins (p<0.008). The mean ratio of CA125II in the 5(th) aliquoted versus the 3(rd) aliquoted sister vial was 1.66 (1.25-2.5, p<0.001). These data suggest that the parent vials were not adequately mixed before they were aliquoted. CA125II in serum can partially precipitate to form a concentration gradient in long-term storage. Rigorous vortexing after thawing and before aliquoting is thus critical.