Esophagectomy with reconstruction of a gastric tube is associated with significant morbidity, i.e., leakage and stenosis of the gastroesophageal anastomosis.1 A major cause of impaired anastomotic healing is insufficient local blood flow, which is attributed to poor arterial inflow and venous congestion caused by partial ligation of the vasculature.2 As a result, the uppermost 20% of the gastric tube is only perfused through the microcirculation.3
Several studies have demonstrated that improving venous drainage enhances the blood flow at the anastomotic site.4–6 However, the effect of arterial perfusion pressure on gastric tube tissue blood flow has never been evaluated. Although increasing arterial blood pressure with the administration of vasoconstrictive drugs does not improve tissue perfusion in tissues of septic patients,7 it can be hypothesized that in tissue with a severely compromised vascularization, such as the anastomotic site of the gastric tube, increasing perfusion pressure might increase tissue blood flow.
For this purpose, we used laser speckle imaging (LSI), which allows determination of the microvascular blood flow in a large area and with a high sampling rate without any physical contact. In addition, an indirect assessment of tissue blood flow was achieved by temperature measurement of the gastric tube by videothermography. These techniques also allow noncontact measurements in a large tissue area with a high sampling rate, and both techniques are frequently used to reflect tissue perfusion. LSI measures blood flow at a certain depth in the tissue (approximately 0.5-1 mm depending on various factors such as hemoglobin content),8 whereas thermography gives a more global impression of total tissue blood flow, with the blood flow of the most superficial layers (with respect to the position of the camera) contributing most to the surface temperature measurement.
With these techniques, we aimed to investigate the effect of increasing perfusion pressure on blood flow in the gastric tube, with a focus on the future anastomotic site. For this purpose, we used a pig model for gastric tube reconstruction.
Nine female pigs (Yorkshire, Landrace of Durok) with a mean weight of 30.6 ± 0.6 kg (mean ± sd) were included in this study. The study was approved by the institutional animal investigation committee, and the care and handling of the animals were in accordance with the European Community guidelines. All animals received premedication with azaperone (2.5 mg/kg IM). Anesthesia was induced with ketamine hydrochloride (30 mg/kg IM) and midazolam (1 mg/kg IM). All animals were intubated through a cervical midline tracheostomy and ventilated in a volume-controlled mode (Servo 300, Siemens-Elema, Solna, Sweden). Neuromuscular blockade was induced with pancuronium bromide (0.13 mg/kg, IV), and anesthesia was maintained with a continuous infusion of fentanyl (20 μg · kg−1 · h−1), midazolam (0.3 mg · kg−1 · h−1), and pancuronium bromide (0.3 mg · kg−1 · h−1). As maintenance fluid, crystalloid solution was administered at a rate of 10 mL · kg−1 · h−1. After surgery, hypovolemia was excluded with a fluid challenge consisting of 150 mL of crystalloid solution. When cardiac output (CO) increased by 15%, the animal was considered hypovolemic, and fluid challenges were repeated until CO did not increase further. The left carotid artery was cannulated for continuous arterial blood pressure monitoring and to obtain blood gas analyses. A pulmonary artery catheter (Edwards, Irvine, CA) was inserted through the right jugular vein for measurements of central venous, pulmonary artery, and pulmonary artery occlusion pressures. In addition, the CO was calculated using a Vigilance CO computer (Edwards). At the end of the experiment, the animals were killed with a bolus of potassium chloride.
After induction of anesthesia, the gastric tube was reconstructed as described by Schröder et al.9 In brief, after median laparotomy, the stomach was mobilized along the smaller and greater curvature. This included ligation of the 1 or 2 short gastric arteries, the gastric arteries, and the left gastroepiploic artery. The right gastroepiploic artery along the greater curvature was carefully preserved. After opening of the hiatus, the esophagus was dissected and transected at the gastroesophageal junction. Finally, a 3-cm-wide gastric tube was formed by dissection of the lesser curvature and the gastroesophageal junction with a linear stapler (Proximate 55 mm TLC linear cutter, Ethicon, Johnson & Johnson, The Netherlands). In this way, the perfusion of the gastric tube was completely dependent on the right gastroepiploic artery, comparable with the clinical setting.
An important difference with human anatomy is the left dominancy of the gastroepiploic artery.9 Ligation of the left gastroepiploic artery and short gastric arteries results in profound ischemia of the gastric fundus, limiting the length of the gastric tube. For this reason, it is impossible to perform a transthoracic gastric pull-up to restore continuity in the neck, and measurements were performed with the gastric tube in the abdomen. The gastric tube was carefully positioned on top of the abdominal content and covered to prevent evaporation and cooling.
After construction of the gastric tube, baseline values of blood flow (LSI) and temperature (videothermography) were obtained from 4 positions on the serosal side of the gastric tube: the top (Fig. 1; location A), the virtual anastomotic site (Fig. 1; location B), the medial part (Fig. 1; location C), and the base of the gastric tube (Fig. 1; location D). A small suture was placed centrally in each region of interest to assure that each measurement was taken from exactly the same place. These landmarks correspond with the letters in Figures 1 and 3. Depending on baseline arterial blood pressure, the norepinephrine dose was adjusted to attain a mean arterial blood pressure (MAP) of 50 mm Hg. When baseline MAP was higher than the desired values, additional propofol was titrated to decrease MAP. Subsequently, norepinephrine was administered to attain a MAP of 60, 70, 80, 90, 100, and 110 mm Hg. After each step of increase in MAP, a 15-min stabilization period was included before both LSI and thermography measurements were performed. The total duration of a single experiment was 6-7 h.
Laser Speckle and Thermographic Imaging and Analysis
LSI was performed with a full-field laser perfusion imager (Moor Instruments, Axminster, UK), which was mounted perpendicular to the gastric tube, at a distance of 30 cm. Scans were captured with a sampling frequency of 25 Hz. The device uses a Class 1 near-infrared laser source with a wavelength of 785 ± 10 nm. The charge-coupled device camera incorporates a bandpass filter, which attenuates other wavelengths. The raw speckle images are used to compute the speckle-contrast image. The software calculates the speckle contrast k for any given square of 5 × 5 pixels and assigns this value to the central pixel of the square. This process is then repeated across the image of 576 × 768 pixels to obtain the contrast map. For each pixel in the speckle contrast map, the relative velocity of blood flow is obtained.
The temperature of the gastric tube tissue was registered with a computer-assisted infrared thermograph (ThermaCAM SC2000, Flir Systems, Danderyd, Sweden). The thermal sensitivity is 0.05°C at 30°C, the spectral range 7.5-13 mm, and the built-in digital video 320 × 240 pixels (total 76,800 pixels). Data were obtained through a high-speed (50-Hz) analysis and recording system coupled with a desktop personal computer (ThermaCAM Researcher 2001 HS). Thermograms were stored on a hard disk (14-bit resolution) awaiting further analysis. With an interval of −40°C until 120°C, this results in a resolution of 9.8 × 10−3°C per bit, which fits well in the range of the thermal sensitivity. The thermograph camera produces a matrix of temperature values. These temperature values each represent a pixel in the image measured. The distance between the objective and the gastric tube being measured was adjusted to 68 cm. Thereby, the resolution on the gastric tube was 0.8 × 0.8 mm2.
Values are reported as median and interquartile ranges. Each variable was analyzed using analysis of repeated measurements. When appropriate, post hoc analyses were performed using the Dunn multiple comparison test. Comparison of LSI blood flow values and thermographic temperature values was done by linear regression analysis and Pearson correlation analysis. P < 0.05 was considered statistically significant. All analyses were performed using GraphPad Prism (version 5.0, GraphPad Software, San Diego, CA).
During the stepwise increase of MAP, all other global hemodynamic values remained constant throughout the experiment. Systemic hemodynamic data and norepinephrine dosages after surgery and for each step of MAP are shown in Table 1. The volume of fluids administered was 2170 mL (1830-2177 mL) during the experiments. CO and right atrial and pulmonary wedge pressures remained constant throughout the experiments, indicating that fluid status remained constant as well.
Laser Speckle Imaging
In Figure 1, the laser speckle images of a typical gastric tube are shown. The gastric tube tissue blood flow is significantly lower in the top than in the base and the medial part of the gastric tube (Fig. 2). MAP levels were categorized as <70 mm Hg (hypotension), 70-90 mm Hg (normotension), and >90 mm Hg (hypertension). Increasing MAP did not have a significant effect on blood flow at any location of the gastric tube.
Figure 3 shows the thermographic images of the gastric tube at the different MAP levels. The size of the black/blue-colored area (lower temperature, less blood flow) even appears to increase at higher MAP levels. Temperature measurements show a similar pattern as LSI, although differences were not significant (Fig. 4). An increase in MAP did not change temperature values at any location.
Flow measured by LSI and temperature measured by thermographic imaging were shown to be positively correlated (Pearson r = 0.84, P < 0.001). This was supported by linear regression analysis, as shown in Figure 5 (r2 = 0.71).
The results of our study demonstrate that increasing perfusion pressure does not improve the gastric tube tissue blood flow, especially at the site of the future anastomosis, where blood flow is extremely low.
Our results are in accordance with the results in septic patients, in whom no increase in splanchnic perfusion was seen with increasing MAP.7 However, our model is typical for the local anatomical disruption of the vasculature, which compromises tissue blood flow. It has been shown that improvements can be made either with surgical interventions or by pharmacologically influencing the microvascular bed. Surgical techniques such as “supercharging” the anastomosis of arteries and veins of the gastric tube improve local blood flow, but are technically difficult and require an additional operating time of >1 h.6,10 Studies using pharmacological interventions were able to decrease venous congestion and improve tissue blood flow at the site of the anastomosis.4,11
Therefore, the residual function of the vascular bed, although compromised, should be able to supply the top of the gastric tube with sufficient blood. Increasing perfusion pressure, however, does not contribute to an increase in blood flow, as shown by our results. This also implies that venous congestion plays an important role during gastric tube formation, especially when improving tissue perfusion is the aim. This is underlined by the beneficial effect of venous drainage by transient bloodletting on blood flow of the gastric tube.5 However, Schröder et al.12 have demonstrated that, regardless of the occurrence of postoperative complications, all gastric tubes had substantially impaired postoperative tissue perfusion. Independent of systemic hemodynamic variables, tissue perfusion recovered to baseline values after 4 days. Considering that this happens in all patients and that anastomotic healing is impaired in only a fraction, it is likely that tissue blood flow is not the only factor to play a role in anastomotic healing.
Our study demonstrates not only that aiming for supranormal arterial blood pressures has no beneficial effects on gastric microvascular blood flow in the anastomotic area but also, maybe even more importantly, that the use of vasoconstrictive substances to induce deliberate short-term hypertension under normovolemic conditions has no acute detrimental effect on already impaired serosal-muscularis gastric tube microvascular blood flow. A previous study has suggested the opposite, but it is very likely that hypovolemic conditions played an important role in those observations.13 Our results are supported by the study by Banic et al.,14 who demonstrated that increasing arterial blood pressure by 30% with phenylephrine did not have adverse effects on blood flow in free musculocutaneous flaps.
Our results emphasize that thermography is a suitable method to evaluate blood flow of the gastric tube during surgery, as is demonstrated by Nishikawa et al.15 The noninvasiveness of the technique makes it especially suitable in a surgical setting. However, thermography provides a global impression of blood flow, which is attributable to the fact that temperature is a derivate of blood flow and can be influenced by other factors. This could possibly contribute to the variation in the measurements and the fact that there is no significant difference between the several locations of the gastric tube. Another possible explanation for the increased variation in temperature at supranormal arterial blood pressures could be that thermography measures the result of overall blood flow. It is possible that some heterogeneity occurred, although not significantly, between the different layers of the gastric wall, with LSI measuring exclusively the serosal-muscularis layer and thermography reflecting, in addition, part of the submucosal-mucosal layer. This is underlined by the absence of a linear correlation, which suggests that the 2 techniques do not assess exactly the same depth and perhaps complement each other. Because of different “penetration depths,” blood flow changes in the mucosa cannot be excluded. Although LSI and thermography showed a good correlation, we consider LSI to be a more sensitive technique. LSI detects perfusion dynamics in a volume of tissue (approximately 0.5- to 1-mm depth) in a fashion similar to laser Doppler flowmetry (LDF). As a matter of fact, our LSI data are similar to those of previous studies using LDF in the gastric tube.4,16,17 Although LDF provides sensitive tissue perfusion variables, it measures only small areas, whereas blood flow alterations, especially in the case of gastric tube formation, occur more heterogeneously. Moreover, LDF requires tissue contact for optimal performance, which could compromise reliability by the introduction of a potential pressure artifact and reduce its usefulness during surgery.
Potential limitations of our study, in addition to those concerning the techniques, include differences in the vascular anatomy of the stomachs of humans and pigs, the study's nonrandomized design without washout periods, and the relatively long duration of a single experiment.
In conclusion, we demonstrate in an experimental model of gastric tube formation that the impaired microvascular perfusion in the anastomotic area of the gastric tube after surgery cannot be improved with higher perfusion pressures. Other pharmacological or surgical interventions should be used to improve microcirculatory blood flow of the gastric tube to improve postoperative outcome.
1.Urschel JD. Esophagogastrostomy anastomotic leaks complicating esophagectomy: a review. Am J Surg 1995;169:634–40
2.Alanezi K, Urschel JD. Mortality secondary to esophageal anastomotic leak. Ann Thorac Cardiovasc Surg 2004;10:71–5
3.Liebermann-Meffert DM, Meier R, Siewert JR. Vascular anatomy of the gastric tube used for esophageal reconstruction. Ann Thorac Surg 1992;54:1110–5
4.Buise MP, Ince C, Tilanus HW, Klein J, Gommers D, van Bommel J. The effect of nitroglycerin on microvascular perfusion and oxygenation during gastric tube reconstruction. Anesth Analg 2005;100:1107–11
5.Kono K, Sugai H, Omata H, Fujii H. Transient bloodletting of the short gastric vein in the reconstructed gastric tube improves gastric microcirculation during esophagectomy. World J Surg 2007;31:780–4
6.Murakami M, Sugiyama A, Ikegami T, Ishida K, Maruta F, Shimizu F, Ikeno T, Kawasaki S. Revascularization using the short gastric vessels of the gastric tube after subtotal esophagectomy for intrathoracic esophageal carcinoma. J Am Coll Surg 2000;190:71–7
7.Ledoux D, Astiz ME, Carpati CM, Rackow EC. Effects of perfusion pressure on tissue perfusion in septic shock. Crit Care Med 2000;28:2729–32
8.Dunn AK, Bolay H, Moskowitz MA, Boas DA. Dynamic imaging of cerebral blood flow using laser speckle. J Cereb Blood Flow Metab 2001;21:195–201
9.Schröder W, Beckurts KT, Stahler D, Stutzer H, Fischer JH, Holscher AH. Microcirculatory changes associated with gastric tube formation in the pig. Eur Surg Res 2002;34:411–7
10.Sekido M, Yamamoto Y, Minakawa H, Sasaki S, Furukawa H, Sugihara T, Nohira K, Yajima K, Shintomi Y, Okushiba S, Kato H, Hosokawa M. Use of the “supercharge” technique in esophageal and pharyngeal reconstruction to augment microvascular blood flow. Surgery 2003;134:420–4
11.Matsuzaki Y, Edagawa M, Maeda M, Shimizu T, Sekiya R, Nakamura K, Onitsuka T. Beneficial effect of prostaglandin E1 on blood flow to the gastric tube after esophagectomy. Ann Thorac Surg 1999;67:908–10
12.Schröder W, Stippel D, Gutschow C, Leers J, Holscher AH. Postoperative recovery of microcirculation after gastric tube formation. Langenbecks Arch Surg 2004;389:267–71
13.Theodorou D, Drimousis PG, Larentzakis A, Papalois A, Toutouzas KG, Katsaragakis S. The effects of vasopressors on perfusion of gastric graft after esophagectomy. An experimental study. J Gastrointest Surg 2008;12:1497–501
14.Banic A, Krejci V, Erni D, Wheatley AM, Sigurdsson GH. Effects of sodium nitroprusside and phenylephrine on blood flow in free musculocutaneous flaps during general anesthesia. Anesthesiology 1999;90:147–55
15.Nishikawa K, Matsudaira H, Suzuki H, Mizuno R, Hanyuu N, Iwabuchi S, Yanaga K. Intraoperative thermal imaging in esophageal replacement: its use in the assessment of gastric tube viability. Surg Today 2006;36:802–6
16.Boyle NH, Pearce A, Hunter D, Owen WJ, Mason RC. Scanning laser Doppler flowmetry and intraluminal recirculating gas tonometry in the assessment of gastric and jejunal perfusion during oesophageal resection. Br J Surg 1998;85:1407–11
17.Miyazaki T, Kuwano H, Kato H, Yoshikawa M, Ojima H, Tsukada K. Predictive value of blood flow in the gastric tube in anastomotic insufficiency after thoracic esophagectomy. World J Surg 2002;26:1319–23