Introduction: Free flap failure often results from venous thrombosis. We developed a rabbit flap model of partial venous obstruction and evaluated the ability of 3 noninvasive monitoring devices to detect partial and full venous outflow obstruction. Methods: 4 x 8 cm myocutaneous flaps (n=9) based on the lateral thoracic vessels were elevated in New Zealand white rabbits. Arterial pedicle inflow was measured using a quantitative volumetric Doppler probe. A pedicle vein branch was cannulated to measure venous outflow pressure. An adjustable tension ligature was used to create partial or full venous obstruction. After an equilibration period, partial venous obstruction was applied to keep venous pressure between 10 and 20 mmHg (mean 13.5±2.2 mmHg) for 25 minutes, followed by full obstruction for 25 minutes (mean 39.1±7.4 mmHg), and then release from obstruction for 25 minutes. Flap skin was assessed with transcutaneous oxygen tension (TcPO2), transcutaneous carbon dioxide tension (TcPCO2), near-infrared spectroscopy (NIRS), and scanning laser Doppler imaging (LDI). Device measurements, arterial inflow and venous outflow pressure were taken at equilibrium and every 5 minutes throughout all interventions. Data were analyzed using Friedman's repeated measures test for nonparametric data with Dunnets post hoc analysis. Results: All values are given as means ± standard deviation. Flap arterial flow decreased significantly with both partial (p<0.01) and full venous obstruction (p<0.01) from a baseline mean value of 5.9 ± 3.0 ml/min blood flow to values of 4.1 ± 2.4ml/min and 0.3 ± 0.4 ml/min respectively. The arterial flow remained lower than the baseline equilibrium period flow even after release from obstruction (p<0.05). TcPO2, TcPCO2, NIRS, and LDI all changed significantly with full obstruction. TcPO2 decreased from a mean of 38.7 ± 17.5 mmHg to 7.9 ± 16.5 mmHg; TcPCO2 increased from a mean of 56.8 ± 14.5mmHg to 96.0 ± 39.1 mmHg; NIRS decreased from a mean saturation of 67.0% ± 6.6% to 43.1% ± 9.0%; and the Doppler perfusion decreased from a mean of 222.8 ± 77.3 to 47.2 ± 16.8 perfusion units. The LDI was the only device that decreased significantly (p<0.01) with partial obstruction from 222.8 ± 77.3 to 186.5 ± 73.2 perfusion units. After release from venous obstruction NIRS (p<0.01) and LDI (p<0.05) remained below baseline, mirroring the decrease in arterial inflow measured during this period, while TcPO2 rose above baseline (p<0.01). Conclusion: We successfully established a partial venous outflow obstruction flap model able to directly measure flap venous back pressure, and evaluated 3 clinically relevant devices in the setting of venous partial and full occlusion. The data show all devices were capable of detecting full occlusion, but only LDI detected partial occlusion. Therefore, LDI monitoring may allow early warning of impending venous outflow obstruction. NIRS varied the least between flaps and thus may be the most easily interpreted based on an absolute number rather than a change in baseline equilibrium values. These are both important criteria for clinical application.
View Synopsis (.doc format, 2351.0 kb)