Thoracic epidural analgesia reduces gastric microcirculation in the pig

Thoracic epidural analgesia (TEA) is used for abdominal surgery to reduce intra- and postoperative pain, improve postoperative respiratory function, and to attenuate the surgical stress response [1]. Also, epidural analgesia has been shown to be associated with reduced risk of postoperative mortality and has beneficial effects on postoperative cardiovascular complications [2]. However, the effect of TEA on the splanchnic microcirculation remains unclear [3] and a concern is whether TEA affects splanchnic blood flow, since poor surgical outcome is correlated to ischemia at the anastomosis [4].

TEA-mediated reduced blood pressure could be of importance for the splanchnic microcirculation, since probes at the mesenteric arteries have demonstrated a TEA-mediated reduced blood flow [5, 6]. However, a reduced splanchnic blood flow in response to TEA has not been confirmed in experimental studies with focus on splanchnic microcirculation and oxygenation [7, 8]. Furthermore, laser Doppler flowmetry indicates increased postoperative gastric mucosal microcirculation with the use of TEA [9]. Thus, the effects of TEA on splanchnic microcirculation are inconsistent and conflicting results likely reflect different approaches to evaluate the microcirculatory flow and oxygenation. For example, both evaluation of tissue oxygenation and laser Doppler flowmetry require contact to the surface of the tissue [7, 9] and that could affect the measurements.

The aim of the present study was to evaluate the effect of TEA on splanchnic microcirculation in pigs with the use of microspheres [10, 11] and laser speckle contrast imaging (LSCI) for non-contact assessment [12]. We hypothesized that TEA would maintain splanchnic microcirculation.

The study was conducted under supervision of the veterinarians at the Department of Experimental Medicine, The Panum Institute, University of Copenhagen and was approved by the Danish Animal Experiments Inspectorate (2012-15-2934-00186). According to the replacement, reduction, and refinement principle [13], the current trial was performed after evaluation of the inter-observer reproducibility of LSCI determined microcirculation [12], and followed by another study comparing the microcirculation assessed by the LSCI, microspheres, and indocyanine green technique (Nerup et al., unpublished).

There was a decrease in systolic blood pressure following induction of TEA (p = 0.030), but no statically significantly changes in diastolic (p = 0.161) or mean arterial pressure (p = 0.154), cardiac output (p = 0.326), or heart rate (p = 0.236) (Fig. 2 and Table 1).

Without contact to the tissue, this study evaluated the effect of thoracic epidural analgesia (TEA) on liver, gastric, and small intestinal microcirculation by use of laser speckle contrast imaging (LSCI) and determination of regional blood flow by microspheres. In contrary to our hypothesis of maintained splanchnic microcirculation following the activation of TEA, microcirculation and blood flow in the arterioles appeared to decrease at both the antrum and corpus of the stomach following the induction of TEA and was accompanied by a drop in systolic blood pressure. Also, regional blood flow determined by microspheres at the antrum decreased following the induction of TEA, while flow to the liver and the small intestine was maintained.

Reduced systolic blood pressure and parallel decreases in microcirculation at the antrum and corpus of the stomach following induction of TEA are in line with studies in humans, indicating a TEA-mediated reduced splanchnic circulation [5, 6, 16‐18]. Blockade of sympathetic nerve fibres with TEA results in peripheral and splanchnic vasodilatation and reduction in vascular resistance [19, 20], which we consider would increase the microcirculatory flow. However, the results indicate a reduced gastric microcirculation, while liver and small intestine microcirculation were sustained following induction of TEA. We cannot provide a simple explanation for these findings, but a rationalization may be the distribution and function of the intestinal vasculature with first- and second-order arterioles located in the serosal and submucosal layers, respectively, supplying a system of smaller third-order arterioles and terminal arterioles [21]. Each terminal arteriole supplies one or several villi, the muscle layer that overlays the perfused region and crypt regions associated with each villus [22]. The villi are sensitive to changes in oxygen supply and oxygen dependency may manifest with as little as a 30 % reduction in blood flow [23]. Therefore, it might be that blood is shunted away from the proximal to the distal part of the gastrointestinal tract (i.e., from the stomach to the small intestine) that has the largest number of villi. Such redistribution of flow could compensate for a reduced microcirculation to the villi by attenuated blood pressure and blood flow following induction of TEA. The stable microcirculation in the liver may be due to its complex blood supply with the largest part of flow derived from the portal circulation [24]. Splanchnic vasodilation following induction of TEA increases venous capacity, with subsequent increase in portal flow and therefore maintained liver microcirculation (the hepatic arterial buffer response) [25].

Reduced gastric microcirculation with TEA adds to the concern regarding the risk of anastomotic ischemia after, e.g., Ivor-Lewis esophagectomy where gastric continuity is re-established using the corpus of the stomach [26]. The integrity of an anastomosis depends on several factors, including surgical technique and viability of the muscle layers, but increased strain and local ischemia are believed to be main reasons for anastomotic leakage [27, 28]. A decrease in corpus microcirculation may question routine intraoperative use of epidural analgesia during esophagectomy, since diminished microcirculation has deleterious effect on the healing of the gastric conduit [4, 29]. An alternative may be that epidural anesthesia for postoperative analgesia after esophagectomy is activated only prior to the termination of general anesthesia.

The discrepancy between changes in regional blood flow as estimated by microspheres and microcirculation as assessed with LSCI likely reflects the different aspects of the microcirculation probed by the two techniques. The microspheres estimate a “snapshot” of absolute tissue blood flow rate (ml/min/g), while LSCI displays mean values of flux recorded (here over 30 s). Furthermore, the tissue samples obtained for calculations of regional blood flow were for the full wall thickness at the small intestine and stomach, as for samples from the liver. Thus, the microsphere-determined estimate of flow expresses blood flow rates through the whole sample thickness. LSCI on the other hand assesses the microcirculation to a depth of only approximately 1 mm [30]. This discrepancy suggests a compensatory reduction in microcirculation in serosal and muscular layers to counteract a reduced mucosal circulation following induction of TEA, as proposed by the increased mucosal microcirculation with TEA by Michelet et al. [9].

Limitations of the present study include uncertainty regarding the precise position of the epidural catheters in pigs. Also, we did not demonstrate that the sympathetic nervous innervation to the splanchnic area was blocked by TEA since, e.g., plasma catecholamines were not measured. However, the “loss of resistance” technique for establishing epidural anesthesia is a clinical standard procedure with low rate of failure [31] and has been used previously in porcine studies [7]. Therefore, based on the changes in the hemodynamic values following the induction of TEA (i.e., drop in systolic blood pressure and (non-significant) increase in heart rate), we believe that the catheters were positioned correctly. Nonetheless, we accept that, e.g., an epidurogram [32] could be used to identify the position of the catheters, which was used in subsequent investigations (Fig. 5). Another limitation to the study may be the dose of the local anesthetics used for epidural blockage, since the volume used was small and the effect on pigs is not well known. The chosen volume was based on studies using TEA in pigs [7, 32] and deduced from the experience with the doses and concentrations used at our department. Further, the current results seem to indicate a delayed block to the lower part of the spine, presented as a (non-significant) decrease in microcirculation at the small intestine 30 min following induction of TEA. Therefore, the volume used may be insufficient and/or the evaluation was made too early to reveal the long-term effects of epidural anesthesia on the microcirculation. Also, the study setup with the present study conducted after another investigation [12] may have affected the microcirculation and its regulation. This may, therefore, not fully resemble the effect of epidural analgesia during clinical situations, with the epidural block often activated prior to surgery. Furthermore, tissue oxygenation of the splanchnic organs was not assed and tissue oxygenation could have been maintained despite a reduced blood flow. However, reliable techniques for assessing tissue oxygenation requires contact to [33] or penetration of tissue surface [7], which could affect the microcirculation. In addition, the objective of the current study was to assess the LSCI technique for monitoring splanchnic microcirculation during TEA, in which we believe, was achieved.

The present study indicates that TEA may have an adverse effect on gastric microcirculation, as assessed by LSCI and microspheres in the pig. The results stress the need for randomized clinical trials to clarify the impact of TEA on gastric microcirculation. We suggest the non-touch technique offered by LSCI, to display changes in blood flow in real-time over a large area of tissue.