Gastrointestinal Motility Following Thoracic Surgery: The Effect of Thoracic Epidural Analgesia

A Randomised Controlled Trial

Argyro Zoumprouli; Aikaterini Chatzimichali; Stamatios Papadimitriou; Alexandra Papaioannou; Evaghelos Xynos; Helen Askitopoulou


BMC Anesthesiol. 2017;17(139) 

In This Article


The key finding of the present controlled randomised study was that all patients undergoing major thoracic surgery had significantly reduced GI motility both on the first and the third postoperative days regardless of the postoperative analgesic technique used. GI motility recovered faster in patients who received TEA with ropivacaine and morphine in a standardised recovery programme compared to TEA with ropivacaine alone or IV morphine. The objective OCTT test of GI mobility revealed that although the effect of the TEA was beneficial from the first postoperative day, it became more pronounced on the third postoperative day.

POI is an important common clinical problem following abdominal surgery,[20,21] but also extra-abdominal procedures[2,22] or noxious stimuli.[23] Since the early nineteenth century, it has been known that stressful, centrally acting stimuli have marked effects on the GI tract. Incision of the peritoneum inhibits the migrating myoelectric complex (MMC) activity, while prolonged inhibition is present after bowel manipulation.[23] To avoid the direct effect on the GI tract from bowel manipulation and local bowel inflammation factors, we only enrolled patients scheduled for major thoracotomy.

All general anaesthetics and short-acting opioids used for induction and maintenance of anaesthesia depress GI motility, but their effects are not prolonged or significant.[1] It is well established that the systemic or epidural administration of opioids decreases gastric emptying, affects MMC activity of the small bowel and decreases propulsive waves in the colon.[1,23,24] Equally, the epidural administration of local anaesthetics blocks afferent and efferent inhibitory reflexes, increases splanchnic blood flow and exhibits anti-inflammatory effects via the systemic absorption.[10] However, although extensive epidural blockade with LAs can prevent the endocrine and metabolic responses to surgery in the pelvis and lower limbs, in thoracic surgery it is not possible to block completely the stress response even with a block up to the C6 dermatome.[5]

Data about the epidural administration of LAs and the combination of LAs and opioids on GI function following non-abdominal surgery are sparse. The only study of mid-TEA with fentanyl and bupivacaine following thoracotomy is that of Guha et al., who demonstrated reduced gastric emptying, using the paracetamol absorption technique.[6] Several studies and reviews have concluded that the epidural administration of local anaesthetics in patients undergoing different types of abdominal surgery provides a faster recovery of the GI hypomotility compared with the systemic or epidural administration of opioids without any increased risk of GI complications.[3,9,11,12,14,20] However, the study by Liu et al. showed no difference between epidural LAs and the combination of epidural LAs and opioids following colon surgery, but both groups showed faster return of GI function than those with systemic or epidural opioids.[3] This was one of the first studies to include a standardised recovery programme to control non-analgesic factors that may influence the rate of GI recovery. Furthermore, a systematic review by Shi et al. of the effect of thoracic epidural analgesia vs. systemic analgesia on the recovery of GI function following GI surgery presented evidence that TEA (compared to systemic analgesia) improved the recovery of GI function after GI procedures even when the analgesic regime included opioids in combination with LAs.[9] In addition, it showed that for the TEA to have a beneficial effect on the motility of the gut, it should be administered for at least 2–3 days after surgery.[9] These findings are in agreement with our findings, which showed that both TEA groups with ropivacaine or ropivacaine and morphine were superior to IV morphine in particular on the third postoperative day.

The present study also demonstrated that patients receiving TEA with ropivacaine and morphine had faster GI recovery compared to those receiving TEA with ropivacaine alone. This finding can be explained by the effect of epidural morphine on the central nervous system, where opioids suppress hypothalamic and pituitary hormone secretion. Although the primary sites of opioid inhibition of the GI function are on the μ-receptors in the peripheral nervous system,[24] the central analgesic and hormonal effects of opioids may also be important. It is possible that epidural morphine decreases the activity of the adrenocortical system and blocks the stress response directly at the hypothalamic level, while epidural LAs cannot block it completely. The present study did not explore other outcomes that may be related to the overall stress response.[5,25]

Perioperative factors of importance in the control of postoperative recovery as well as a multimodal approach to postoperative care should be considered in all studies of the effects of postoperative analgesia on surgical outcomes.[12,20] In this study, factors known to affect postoperative gut recovery were controlled, with the only exception being the administration of the anticholinergic glycopyrrolate and the acetylcholinesterase inhibitor neostigmine at the end of the operation. Although a fixed dose of neuromuscular reversal was administered, further comparison of glycopyrrolate−1 and neostigmine−1 with OCTT did not reveal any correlation. The single dose of these two agents was so small that it is unlikely to have produced a clinically significant sustained effect in the postoperative period. All patients of this study participated in a structured postoperative programme that included early oral intake, early mobilisation and reduction of postoperative administration of IV fluids, with postoperative analgesia aimed to achieve a VAS score of 5. Although patients in the Ep-R group had large variations in VAS score, no correlation was found between pain scores and GI motility.

The difficulty in comparing results from different studies on GI motility is increased by problems of methodology and design, lack of reporting the level of sensory block and failure to control factors known to affect the GI tract. Clinical assessment of the overall GI motility also presents difficulty. The time to first bowel sounds is not a specific test, while the time to passage of first flatus, although a measure of coordinated bowel function, is an insensitive marker, as patients usually overlook the sensation when influenced by wound pain and analgesic medications.[17] Whilst the time to the first passage of stools represents a clear clinical endpoint, it may indicate only distal bowel emptying and not the function of the entire gut.[1,17] Other tests, such as the scintigraphic imaging investigations of radio-labelled meals used for the assessment of GI recovery have been shown to be relatively independent of the clinical markers used to evaluate resolution of the ileus.[15,17,18] For this reason, in the present study we assessed GI motility not only by clinical signs of bowel recovery (flatus, bowel sounds, defecation), but also by measuring the OCTT and the migration of radiopaque markers by abdominal radiographs. It is noteworthy that our findings showed no correlation between the clinical bowel signs and the OCTT and CTT data.

The lactulose H2-breath test is a simple, non-invasive method widely used for quantifying the OCTT, where the substance (lactulose) and dose (10 g) used in a controlled diet setting improves the reliability of the test.[18] In addition, testing subjects preoperatively allowed identification of potential, pre-existing gut abnormalities (H2-non producers) and confirmed the homogeneity of the cohort. The main limitation of the second objective test used, the CTT, is the physiological variability in colonic movement. The preoperative administration of 20 radiopaque markers allows for the evaluation of the baseline colonic transit time with an exponential rate of disappearance of the markers. The mean number of markers retained in the healthy colon 3 days after indigestion is reported as 2.[19] In the present study although there was a difference between preoperative and postoperative CTT, no difference was shown between groups. This may be because we performed only one X-ray (rather than a series of daily X-rays) in order to minimise radiation exposure and discomfort to our patients or that the study was underpowered to detect the difference.

Several other limitations must be acknowledged. The study was monocentric and half blinded. Although there were no demographic differences between the groups, the study included mainly men (men 27: women 3). Women have slightly different rates of gut motility especially regarding the CTT. Also, the study was underpowered to assess the secondary outcomes (VAS) scores which could have an effect on gut motility.