Pre-analytical parameters have significant impacts on VEGF levels measured from human plasma

Abstract

Purpose: Numerous studies include measurements of plasma VEGF concentrations in their endpoint analyses. However, if measurement parameters are not standardized, the results from different studies may not be comparable. Even within one study, measured plasma VEGF levels may vary if methods of obtaining and processing of samples are not strictly defined. In this study, we investigated which pre-analytical parameters affect plasma VEGF measurements and should therefore be carefully controlled in clinical studies.

Methods: Blood samples were obtained at three different clinical sites from six healthy volunteers at two time points using precisely defined procedures and detailed instructions for sample acquisition and processing. Several pre-analytical factors were deliberately altered to investigate their respective effects on VEGF measurements. The samples were measured at the three clinical sites (ELISA), three additional measurement centers (ELISA) and one central reading center (ELISA and Luminex). The parameters analyzed were: cannula (butterfly vs. neonatal), filling level (completely filled vs. half-filled tubes), use of anticoagulant (EDTA vs. PECT / CTAD), type of centrifuge (swing-out rotor vs. fixed angle), delay before and after centrifugation, study center and method of detection (ELISA vs. Luminex). In addition, intrapersonal fluctuations over one week and possible sex differences were investigated. A linear regression model was used for analysis.

Results: Regression analysis identified the following parameters as independent confounders of VEGF levels measured in human plasma: Study center, method of detection, sex of the proband, anticoagulation agent and type of centrifuge used. In contrast, the following parameters had no detectable independent effect on the VEGF values measured: Intrapersonal variation over one week, type of cannula, time between sampling and centrifugation, time between centrifugation and separating plasma, filling level of sampling tubes.

Conclusions: In order to achieve reliable plasma VEGF readings, type of anticoagulant, type of centrifuge and measuring method need to be standardized, ideally in one central measuring center. Failure to standardize these parameters may significantly affect results of plasma VEGF measurements in clinical trials.