Iodinated Contrast Media Interfere with Gel Barrier Formation in Plasma and Serum Separator Tubes

Vacutainer® serum separator tubes (SST® II) and plasma separator tubes (PST® II) manufactured by Becton Dickinson are widely used for blood collection. These tubes contain an inert, thixotropic polymer gel with a specific gravity of ∼1.04 (1). Aspiration of blood into the tube and subsequent centrifugation displace the gel, and it forms a disk between the cells and the supernatant. We recently encountered blockages of sample needles on our clinical chemistry Modular automated analyzer (Roche Diagnostics) with samples collected in PST II tubes. The tubes were placed directly on the analyzer after centrifugation for primary-tube sampling. The problematic samples had been collected from patients shortly after radiologic examinations requiring administration of iodinated radio-opaque contrast medium (Visipaque® or Omnipaque®). Instead of forming a separation layer between the sediment and the plasma after centrifugation, the gels had floated to the top and thus hindered analytical processing. We investigated the influence of iodinated contrast media on gel barrier formation in PST II and SST II tubes. The most commonly used contrast media in our hospital are two nonionic contrast media, Omnipaque (iohexol; 350 g/L iodine; Nycomed) and Visipaque (iodixanol; 320 g/L iodine; Nycomed), and to a much lesser extent, two ionic contrast media, Telebrix® (sodium meglumine ioxithalamate; 350 g/L iodine; Guerbet) and Urografin® (sodium meglumine diatrizoate; 146 g/L iodine; Schering). In one experiment we investigated the interference in whole blood samples. On three separate occasions, a total of 43 blood samples were collected from a healthy male volunteer (age, 26 years) into PST II (cat. no. 367376) and SST II (cat. no. 367955) tubes. Immediately after blood collection, we added different amounts of Omnipaque (0–44 mL/L; equivalent to 0–16 g/L iodine, final whole blood concentration), Visipaque (0–33 mL/L; equivalent to 0–11 g/L iodine), Telebrix (0–44 mL/L; equivalent to 0–16 g/L iodine), and Urografin (0–91 mL/L; equivalent to 0–13 g/L iodine). The tubes were mixed, and the serum specimens were allowed to clot for 30 min at room temperature before centrifugation. After centrifugation at 1700g for 10 min, the tubes were inspected for the position of the separating gel. Density was measured by weighing 1 mL of plasma or serum, and osmolality was measured with an Advanced Osmometer 3D3 (Advanced Instruments). The concentrations at which the gel failed to form a barrier under the plasma after centrifugation are shown in Table 1A . Results of in vitro and in vivo experiments on gel barrier formation after injections of contrast media for radiography. Lowest concentration at which the gel failed to form a layer under plasma or serum after centrifugation. NN, not needed; ND, not done. Results of in vitro and in vivo experiments on gel barrier formation after injections of contrast media for radiography. Lowest concentration at which the gel failed to form a layer under plasma or serum after centrifugation. NN, not needed; ND, not done. In another experiment, we added different amounts of Omnipaque (0–100 mL/L; equivalent to 0–35 g/L iodine, final plasma concentration), Visipaque (0–108 mL/L; equivalent to 0–35 g/L iodine), Telebrix (0–111 mL/L; equivalent to 0–39 g/L iodine), and Urografin (0–267 mL/L; equivalent to 0–39 g/L iodine) to pooled plasma samples in PST II tubes. The tubes were centrifuged and inspected for the position of the gel barrier as above. The density and osmolality of the plasma were also measured. We found that the concentrations needed to cause the gel to float to the top were much lower for whole blood than for plasma (Table 1A ). The presence of high-density molecules such as iodinated contrast media increases the density of the plasma and serum. When the density of the plasma or serum exceeds that of the gel, the latter no longer forms a layer between the plasma or serum and the cells, but floats partially or completely to the top of the plasma or serum. Ionic substances affect osmolality to a greater extent than do nonionic substances. This explains the lower osmolality measured with the nonionic contrast media. Our in vitro experiments would suggest that injection of 60–80 mL (assuming the circulatory volume of an adult to be 3–4 L) of the nonionic contrast material might be sufficient to cause interference with gel barrier formation. However, in practice, because the distribution half-lives of all of the contrast media are relatively short (≤22 min) (2)(3)(4)(5), in vivo administration of up to 210 mL of Omnipaque and 250 mL of Visipaque followed by drawing of blood immediately after the radiographic procedure produced no interference, the densities being 1.035 and 1.026 kg/L, respectively (Table 1B ). By contrast, infusion of 350 mL of Visipaque and 320 mL of Omnipaque led to failure of appropriate gel barrier formation in samples taken immediately after radiography. We observed no interference in samples taken from the same individuals 2 h later (Table 1B ). The improvement at 2 h is consistent with the half-lives of ∼2 h for clearance of contrast agents in patients with normal renal function, e.g., 126 min for Omnipaque (2), 123 min for Visipaque (3), ∼ 120 min for Telebrix (4), and 60–120 min for Urografin(5). Blood samples collected in SST II tubes occasionally fail to form an appropriate gel barrier in patients who have not received any contrast material or other infusion (the densities being 1.023–1.024 kg/L and osmolalities 281–294 mmol/kg of H2O). In some such cases, the PST II samples obtained at the same time exhibited no interference. In these cases, the gel either remained lodged at the bottom of the SST II tube or was located elsewhere in the body of the clot. This appeared to reflect physical entrapment of the gel under or in the clot. Careful loosening of the clot from the sides of the tube followed by a second centrifugation yielded the usual location of the gel barrier at the interface of the serum and the clot. In summary, we report that the barrier gel in PST II tubes may not form an appropriate barrier in blood samples collected from patients who have had infusions of contrast media. Failure of gel barrier formation occurred only in those cases in which >300 mL of contrast medium was injected and blood was drawn immediately after the radiographic procedure. We recommend that in such cases blood sampling be delayed for at least one elimination half-life (∼ 2 h) after infusion to obviate interference. Although less practicable, theoretically, allowing four to five half-lives to elapse would avoid the interference completely regardless of the initial concentration of the contrast medium, and longer delays would appear necessary in patients with impaired renal function as clearance is inversely related to glomerular filtration rate (2)(3)(4)(5).