Optimal collection of blood samples for the measurement of tumor necrosis factor alpha.

We have examined how delayed separation of plasma from cells affects the recovery of recombinant human tumor necrosis factor alpha (rhTNF alpha) from whole blood. Storage of heparinized whole blood samples at room temperature for 1 hr results in a significant (p = 0.036) fall in recovery of plasma TNF alpha from 788 +/- 119 pg/mL to 472 +/- 77 pg/mL, measured by specific enzyme-linked immunosorbent assay (ELISA). Storage of whole blood samples at 4 degrees C for 1 hr reduces but does not prevent the fall in recovery of plasma TNF alpha: 725 +/- 82 pg/mL at time 0, 472 +/- 81 pg/mL after 1 hr, p = 0.038. Recovery of bioactive TNF alpha (cytotoxocity for L929 cells) after 1 hr at room temperature is also significantly reduced from 576 +/- 139 pg/mL to 450 +/- 154 pg/mL, p = 0.036. Studies with 125I-rhTNF alpha confirmed the fall in plasma activity and revealed a rapid commensurate increase in 125I-rhTNF alpha activity in the cell fractions. We recommend that clinical samples for the measurement of cytokines should be kept at 4 degrees C and separated rapidly (within half an hour) before storing the plasma at -70 degrees C.