See editorial on page 1001.
Postoperative ileus (POI) remains a challenge for surgeons because of its frequency and the potential severity of its consequences. Enhanced recovery programs (ERPs) reduce the occurrence of POI and the delay in GI motility recovery, but within patients following these programs, POI still occurs in <24.4% of patients, and the mean GI motility recovery can be <4 days.1 , 2
POI increases the cost of care because it increases the length of stay and can also lead to severe morbidity, such as inhalation pneumopathy or cardiac arrhythmia, which can potentially be severe for the patient.3 , 4 More recently, the link between the physiopathology of POI and anastomotic leakage has been highlighted.5 There is therefore a real need to improve POI prevention and treatment.
POI involves many pathways, such as neuronal pathways, inflammatory pathways, or vagal nerve inhibition. To improve ERPs, recent clinical trials assessed the efficiency of new therapies, such as vagal nerve stimulation, by chewing6 , 7 or nutritional improvement.8 , 9 However, the results were not as relevant as had been expected. Research has continued, and Stakenborg et al10 proposed recently that using electric abdominal vagus nerve activation with a significant anti-inflammatory effect would lead to a significant improvement of GI motility recovery.
However, all of the therapies involving vagal nerve stimulation have an effect on the small bowel, but their effect on the colon is more uncertain because it has not been assessed, and, more importantly, because vagal nerve density decreases along the rostrocaudal gradient, leading to a near missing of vagal roots on the colon.11
Because inflammation pathways in the small bowel are reduced by vagal nerve stimulation, it would therefore be interesting to find a mean value for the reduction of inflammation in the colon. To do this, pharmacological and nutritional approaches must be developed, but these are not yet feasible in clinical practice.
However, reducing colon inflammation using an instrumental approach for other indications such as ulcerative proctitis was reported in 1 case.12 Indeed, sacral nerve stimulation significantly reduced inflammation in a case of refractory proctitis.12 Sacral nerve stimulation is invasive and requires ≥2 surgical procedures for implantation and therefore cannot be used to prevent POI. However, other devices used in urinary and fecal incontinence or constipation can be used to stimulate the nerve endings in this nerve,13 , 14 such as the posterior tibial nerve. One of these devices, UROstim2 (Schwa-Medico, Ehringshausen, Germany), is used in transcutaneous tibial nerve stimulation. The efficacy and safety of using transcutaneous tibial nerve stimulation (TENS) have been established for the treatment of incontinence or constipation but have yet to be determined in the context of postoperative care.13 , 15 We hypothesized that perioperative sacral nerve stimulation via the tibial nerve would improve postoperative care in terms of the delay in GI motility recovery.
The primary aim of this study was to assess the efficacy of TENS in reducing the delay in GI motility recovery compared with a placebo. The secondary aims were to assess the safety of TENS and its efficacy in reducing the delay in the first passage of flatus and in reducing the occurrence of POI, to improve the quality of the first stool, to reduce postoperative morbidity and mortality, to reduce the length of stay, to improve the postoperative quality of life in patients, and to reduce inflammatory parameters in the muscularis propria.
PATIENTS AND METHODS
This was a preliminary, randomized, single-blind monocentric clinical trial, which involved patients undergoing a colectomy or high anterior resection between October 1, 2016, and May 2, 2017. The study has been approved by the Personal Protection Committee and was registered at ClinicalTrials.gov under the number NCT02815956.
Inclusion criteria were patients undergoing a colectomy or high anterior resection with anastomosis using laparotomy or laparoscopy. Exclusion criteria were as follows: 1) emergency surgery, 2) chronic IBD, 3) documented irritable bowel syndrome, 4) surgical history of gastrectomy or esogastrectomy (risk of vagal nerve lesion), 5) treatment of a disease that would modify the metabolism of acetylcholine, 6) enterostomy, or 7) pacemaker, because of the risk of interference with the neurostimulator device.
Irritable bowel syndrome and chronic IBD have some overlapping mechanisms16 and result in GI motility dysfunction. They involve inflammation mediators that could modify clinical and molecular results.
Patients were randomly assigned online and stratified according to the location of the colon resected (right vs left/upper rectum) and according to the surgical access size (laparoscopy vs laparotomy). Patients who participated in the study were randomly assigned treatment by placebo (group P) or by TENS (group T). Surgical resection of the colon was performed in accordance with the standard techniques.
A simple blind was ensured for the patient. One of the investigators (the one who allocated the device) was not blind, but the investigators who collected data were blind.
Data collection included the following:
- 1) Demographic information, including medical history, sex, BMI, and age;
- 2) Data related to patient compliance with the ERP, where ERP compliance was defined as the percentage of compliance with 18 elements assessed in the perioperative period (the elements used to assess the compliance with ERP are shown in Table 1);
- 3) Information on the surgical procedure (surgical access size, duration of procedure, and location of colectomy);
- 4) data from the postoperative period (morbidity classified according to the Dindo–Clavien classification),17 POI occurrence (see definition below), need for nasogastric tube (NGT) reinsertion, and length of stay; and
- 5) Gastro-Intestinal Quality of Life Index (GIQLI),18 regarding quality of first stool assessed using the Bristol stool form scale.19
The primary end point was the delay in GI motility recovery. GI motility recovery was defined as the passage of stools and tolerance to solid food, as described by van Bree et al.20
The secondary end points included the following: 1) occurrence of POI and need for NGT replacement (POI was defined as the absence of GI motility recovery after postoperative day 4), 2) quality of the first stool assessed using the Bristol stool scale,19 3) delay until the first passage of flatus, 4) length of stay, and 5) quality of life in patients assessed using the GIQLI.18 The inflammatory parameters, such as cyclo-oxygenase (COX)-2 mRNA expression, hematopoietic prostaglandin-D synthase (HPGDS), microsomal prostaglandin-E synthase (mPGES1), interleukin 6 (IL-6), inducible nitric oxide synthase (iNOS), and monocyte chemoattractant protein 1 (MCP1), were also compared between groups. Modality of tissue collection and storage, RNA extraction, and polymerase chain reaction performance and analysis have been described in a previous article.21 For this study, 6 patients from group P and 7 patients from group T had samples of healthy muscularis propria collected (from 1 of the extremities in the surgical specimen (colic margin of the surgical specimen in case of a right colectomy and upper colic margin of the surgical specimen in case of a left colectomy), immediately before the anastomosis was performed) and stored in the Biological Resources Center of Angers (BB-0033-00038). They were registered under the number CB-2011-05. Because biocollection was only open to patients with cancer, all of these patients underwent operations attributed to cancer. Twelve patients with cancer had no tissue collection because of problem of organization.
A UROstim2 device was used to perform the stimulation. The protocol for stimulation was the same as the one used for fecal incontinence. The first electrode was placed behind the internal malleolus, and the other was placed 10 cm above, on the internal side of the calf. Program P3 was used.
The amplitude of stimulation used ranged from 10 to 38 mA, with a 200-µs pulse width at a frequency of 10 Hz. The amplitude of stimulation was set according to patient feedback.
Stimulation was performed the day before surgery. On the day of surgery, stimulation was performed only once, ≥2 hours before surgery or ≥2 hours after surgery. In the days after surgery, stimulation was performed 3 times per day until GI motility was recovered. The duration of TENS was 20 minutes.
The protocol for group P was exactly the same as for group T, but the device did not deliver any electrical impulse, although it appeared to be a UROstim2 device. To ensure a simple blind, the device was similar in appearance but did not deliver an electrical impulse. The device indicated voltage and intensity.
Sample Size Calculation
Because this was a preliminary study, no information was found in the literature relating to the potential effect of TENS on POI prevention, and no sample size calculation had been performed. Twenty patients were assigned a priori in each group to assess this effect.
After descriptive analysis, we compared patients from group P with patients from group T. Data were expressed in numbers and percentages, and both the Fisher test and the χ2 test were used to compare categorical variables as appropriate. T tests or Mann–Whitney U tests were used to compare continuous variables as appropriate.
Data were expressed as mean ± SD or median (interquartile range); p < 0.05 was considered statistically significant. Data analysis was performed with version 15 of the Statistical Package for the Social Sciences software (SPSS Inc, Chicago, IL).
During the period of inclusion, 55 patients underwent a colectomy or high rectal resection, and 40 patients were included in this trial (Fig. 1). The noninclusion of the 15 patients was attributed to problems related to organization in the case of 9 patients (eg, admission on the day of surgery, admission on a Sunday, or no investigator available), indication of IBD for surgery in 3, location of colectomy on the transverse colon in 2, and subtotal colectomy in 1 (the randomization process allocated did not allow for randomization of these colectomies).
Forty patients were recruited and randomly assigned. Of these patients, 6 were excluded because of issues with consent or contraindications (Fig. 1). The excluded patients did not follow the protocol and were not taken into account in the final analysis. In the end, 34 patients were analyzed.
Twenty-one patients were men (61.8%). Indication was cancer in 25 patients, diverticulosis in 8 patients, and an endoscopically nonresectable polyp in 1 patient. Surgery was completed in every case, and the median duration of the procedure was 167 minutes (range, 120–220 min). Twenty-seven patients had a left colectomy (79.4%), and 29 patients underwent a laparoscopic procedure.
Conversion to open was necessary in 9 cases (31%). Compliance with an enhanced recovery protocol was >70% in 26 patients (76.6%).
POI occurred in 9 patients (26.5%), and 4 patients required NGT replacement (11.8%). With the exception of POI, 1-month morbidity occurred in 11 patients (32.4%), and 1 of these patients experienced 2 complications. Morbidity collected included acute urinary retention (n = 3), pneumopathy (n = 2), postoperative bleeding (n = 2), epilepsy (n = 1), incisional hernia (n = 1), small-bowel obstruction occurring after GI motility recovery (n = 1), postoperative hyperthermia (n = 1), and urinary infection (n = 1).
Except for the patient experiencing epilepsy (group T), who had a Clavien–Dindo severity score of IVA, and the patient experiencing small-bowel obstruction (group P), who needed a reintervention and had a Clavien–Dindo severity score of IIIA (group T but device P allocated), all of the complications had a Clavien–Dindo severity score of ≤ 2. No deaths were reported.
Seventeen patients were assigned to each group. At baseline, the 2 groups were well matched (Table 2).
Mean GI motility recovery was 3.59 days in group P and 3.12 days in group T, but there was no significant difference (p = 0.40). Other secondary outcomes were not significantly different (Table 3), but the first passage of flatus occurred later in group P (2.24 d) than in group T (1.47 d; p = 0.06).
NGT replacement was more likely in group P (17.6%) than in group T (5.9%; p = 0.6), and POI was more likely to occur in group P (35.3%) than in group T (17.6%; p = 0.44). Morbidity was not significantly different between the groups (p = 0.27), but the rate of complications was higher in group P (41.2%) than in group T (23.5%).
The median quality of the first stool assessed using the Bristol stool form scale was not significantly different between the groups (p = 0.44). Medians of COX-2, mPGES1, MCP1, IL-6, and HPGDS mRNA expressions for muscularis propria were not significantly different between groups (p = 0.86, p = 0.17, p = 0.58, p = 0.86, and p = 0.41). mRNA expression of iNOS was measurable only in 2 patients from group P.
Because of the incorrect allocation of 2 devices, we also performed per-protocol (PP) analysis (Fig. 1). In the end, 19 patients received the placebo (group P), and 15 patients received stimulation (group T).
At baseline, the 2 groups were well-matched. There was no significant difference between the groups regarding the patients’ characteristics or their surgical characteristics (Table 4).
Mean GI motility recovery was 3.74 days in group P and 2.87 days in group T, but there was no significant difference (p = 0.10). Other secondary outcomes were not significantly different (Table 5), but the first passage of flatus occurred later in group P (2.16 d) than in group T (1.47 d; p = 0.07).
POI occurred significantly more in group P (n = 8 (42.1%)) than in group T (n = 1 (6.7%); p = 0.045), and NGTs had to be replaced in 4 patients (11.8%) from group P but in none in group T (p = 0.27). Morbidity was not significantly different between groups (p = 0.27), but the rate of complications was higher in group P (42.1%) than in group T (20%).
The median quality of the first stool assessed using the Bristol stool form scale and the postoperative quality of life assessed using GIQLI were not significantly different between groups (p = 0.44 and p = 0.44). Medians of mRNA expression for mPGES1 tended to be higher in group P than in group T (p = 0.056; Fig. 2). There was no significant difference in mRNA expression between HPGDS and iNOS, but there was no measured expression for these enzymes in group T (Figs. 3 and 4). Medians of COX-2, MCP1, and IL-6 mRNA expressions were not significantly different between groups (p = 0.93, p = 0.63, and p = 0.88).
Adverse Effects Potentially Linked With TENS and Tolerance to TENS
No complication directly linked to TENS was reported, but 1 patient from group T experienced epilepsy. As it is recommended that TENS is used with caution in the case of patients with epilepsy, we secondarily considered epilepsy as a criterion for exclusion.
No patient withdrew from the trial because of a poor tolerance to TENS. Also, the GIQLI was not different between the 2 groups (p = 0.36 in intention-to-treat (ITT) and p = 0.40 in PP).
This preliminary randomized controlled trial (RCT) included 40 patients undergoing colectomies; however, 34 patients completed the trial and were included in the final analysis. Differences observed between groups regarding GI motility recovery were not significant in ITT or PP analysis. However, although the difference was not significant, we observed a reduction in the delay time for recovery from 13.0% (ITT) to 23.3% (PP). Despite the limitations inherent to the preliminary nature of this RCT, but also inherent to the selection bias of procedure (wrong allocation of devices and secondary exclusion), our study achieved its first aim, which was to assess the effect of TENS on patient recovery after colorectal surgery to allow sample size calculation for additional studies.
Interestingly, and despite the fact that we did not expect to find a significant difference between the groups in terms of our secondary aims, POI occurrence was significantly higher in group P (n = 8 (42.1%)) than in group T (n = 1 (6.7%); p = 0.045) in PP analysis. This result was not as significant in ITT analysis, but we noted a trend of POI reduction in group T (35.3% vs 17.6%; p = 0.42). Occurrence of POI in our study (26.5%) was similar to that reported in previous literature on a population of patients who followed ERPs.1 However, the rate of POI in group P was higher than that found in previous literature, although it was expected to have been the same. This was probably because of the small size of our population, which resulted in a bias. However, there was no difference between groups regarding the usual risk factors for POI, such as male sex,22–24 history of abdominal surgery,25 open access (p = 0.63) and conversion to open,24 right-side colectomy,26 duration of surgery,27 or ERP compliance.1 This confirmed that the groups were well matched in ITT and PP analysis.
As for the consequences of POI, NGTs were more likely to be replaced in group P than in group T (p = 0.60 in ITT analysis and p = 0.27 in PP analysis). Although this was not statistically significant, this result is interesting, because it confirms that high-grade POIs27 were more likely to occur in group P.
Data regarding the delay in the first passage of flatus were of interest in PP and in ITT analysis (p = 0.07 and p = 0.06). The reduction in this delay was 32% in PP and 34% in ITT. These data provide a sample size calculation for more powerful analyses.
Considering that anastomotic leakage is associated with POI occurrence (OR = 12.57),5 we believed that TENS could have an effect on postoperative morbidity. Our results support this hypothesis, showing a nonsignificant reduction in morbidity for ≈50% of group T compared with group P (in PP and also in ITT analysis). However, the morbidity occurring in our study was low grade, and no fistulas were reported.
To provide additional evidence for the beneficial effect of TENS on postoperative care after colorectal resection, the mRNA expression for some enzymes involved in inflammation pathways at an early stage was quantified in patients with cancer. The fact that only this population of patients was studied in a molecular point of view represents a bias but, despite the small number of patients, low rates of POI occurrence, and results that were not statistically significant, a reduction in mPGES1 mRNA expression was observed in PP analysis (p = 0.056) and in ITT analysis (p = 0.17). Targeting this enzyme has previously been found useful for colon motility improvement in patients with cancer.21 Interestingly, other enzymes, such as iNOS or HPGDS, were also reduced, but the level of expression was very low, and the risk of bias was high in such a small population. Previous literature has reported on the involvement of these enzymes in POI physiopathology,21 , 28 but, although no significant difference was found between groups, our results cannot support the idea that TENS decreases inflammation pathways in POI. Molecular studies could have been performed too early after surgery to allow the inflammation to be induced and thus to observe significant difference between groups, or other pathways can be involved in this effect. Indeed, it has been shown in the literature that sacral nerve neuromodulation can reduce epithelial barrier permeability of the colon in pigs,29 whereas an increase of permeability is involved in POI.30
Finally, TENS was successfully applied to every patient after the trial, and we did not report any severe adverse effect directly because of the use of TENS. No patient experienced local cutaneous inflammation or pain. Only 1 patient experienced an episode of epilepsy. The treatment of patients who had a severe history of epilepsy was suspended in the perioperative period. However, even if the link between this episode and the TENS was uncertain, we chose to contraindicate TENS in patients with a medical history of epilepsy.
Some limitations have to be discussed. Indeed, blinding patients in studies using TENS is imperfect, because the amplitude of stimulation is set according to the patient’s sensation. Some patients could be aware of their group. However, the investigators did not give any information on the device allocated, leading to doubt, probably reducing the part of placebo effect. Also, despite some limitations inherent to its preliminary nature and several deviations from the protocol (secondary exclusion of patients), this preliminary RCT provides interesting information that will be useful for calculations involving sample sizes of patients from future multicenter RCTs. Indeed, we observed some differences between the groups in the delay in GI motility recovery, first flatus recovery, and the occurrence of morbidity and POI. Finally, considering that mean time for GI function recovery was 3.6 in group P and 3.11 in group T, with a global SD of 1.59, the sample size calculation would suggest the need for 144 patients in each group for an α risk of 0.05 and a power of 0.8. This RCT fulfilled its objectives in showing some difference in GI function recovery between groups and in allowing the sample size calculation of patients for an additional multicenter RCT, but also by confirming that the use of TENS is safe in the perioperative period.
The use of TENS is feasible and safe during the perioperative period when contraindications are respected. TENS could have an effect on GI recovery and on morbidity occurrence, but the power of this study was too weak to conclude to a significant effect. These results have yet to be confirmed and characterized in highly recruited, multicenter, double-blinded RCTs. Results of this preliminary study will be used for the sample size calculations of these RCTs.
The authors thank Henry Guepin and Amine Amroun for their generous help in data collection, Elodie Bamboux for her generous help in data collection quality assessment, and Drs Elsa Parot-Schinkel and Bruno Viel for their help in methodology and statistics.
1. Grass F, Slieker J, Jurt J, et al. Postoperative ileus
in an enhanced recovery pathway-a retrospective cohort study. Int J Colorectal Dis. 2017;32:675–681.
2. Barbieux J, Hamy A, Talbot MF, et al. Does enhanced recovery reduce postoperative ileus
after colorectal surgery? J Visc Surg. 2017;154:79–85.
3. Asgeirsson T, El-Badawi KI, Mahmood A, Barletta J, Luchtefeld M, Senagore AJ. Postoperative ileus
: it costs more than you expect. J Am Coll Surg. 2010;210:228–231.
4. Bragg D, El-Sharkawy AM, Psaltis E, Maxwell-Armstrong CA, Lobo DN. Postoperative ileus
: Recent developments in pathophysiology and management. Clin Nutr. 2015;34:367–376.
5. Peters EG, Dekkers M, van Leeuwen-Hilbers FW, et al. Relation between postoperative ileus
and anastomotic leakage after colorectal resection: a post hoc analysis of a prospective randomized controlled trial. Colorectal Dis. 2017;19:667–674.
6. Liu Q, Jiang H, Xu D, Jin J. Effect of gum chewing on ameliorating ileus following colorectal surgery: a meta-analysis of 18 randomized controlled trials. Int J Surg. 2017;47:107–115.
7. Atkinson C, Penfold CM, Ness AR, et al. Randomized clinical trial of postoperative chewing gum versus standard care after colorectal resection. Br J Surg. 2016;103:962–970.
8. Stein K, Stoffels M, Lysson M, et al. A role for 12/15-lipoxygenase-derived proresolving mediators in postoperative ileus
: protectin DX-regulated neutrophil extravasation. J Leukoc Biol. 2016;99:231–239.
9. Rees J, Bobridge K, Cash C, Lyons-Wall P, Allan R, Coombes J. Delayed postoperative diet is associated with a greater incidence of prolonged postoperative ileus
and longer stay in hospital for patients undergoing gastrointestinal surgery. Nutr Diet. 2018;75:24–29.
10. Stakenborg N, Wolthuis AM, Gomez-Pinilla PJ, et al. Abdominal vagus nerve stimulation as a new therapeutic approach to prevent postoperative ileus
. Neurogastroenterol Motil. 2017;29:e13075.
11. Berthoud HR, Carlson NR, Powley TL. Topography of efferent vagal innervation of the rat gastrointestinal tract. Am J Physiol. 1991;260(1 pt 2):R200–R207.
12. Brégeon J, Neunlist M, Bossard C, et al. Improvement of refractory ulcerative proctitis with sacral nerve stimulation. J Clin Gastroenterol. 2015;49:853–857.
13. Rimmer CJ, Knowles CH, Lamparelli M, et al. Short-term outcomes of a randomized pilot trial of 2 treatment regimens of transcutaneous tibial nerve stimulation for fecal incontinence. Dis Colon Rectum. 2015;58:974–982.
14. Iqbal F, Collins B, Thomas GP, et al. Bilateral transcutaneous tibial nerve stimulation for chronic constipation. Colorectal Dis. 2016;18:173–178.
15. Iqbal F, Thomas GP, Tan E, et al. Transcutaneous sacral electrical stimulation for chronic functional constipation. Dis Colon Rectum. 2016;59:132–139.
16. Barbara G, Cremon C, Stanghellini V. Inflammatory bowel disease and irritable bowel syndrome: similarities and differences. Curr Opin Gastroenterol. 2014;30:352–358.
17. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240:205–213.
18. Eypasch E, Williams JI, Wood-Dauphinee S, et al. Gastrointestinal quality of life index: development, validation and application of a new instrument. Br J Surg. 1995;82:216–222.
19. Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit time. Scand J Gastroenterol. 1997;32:920–924.
20. van Bree SH, Bemelman WA, Hollmann MW, et al. Identification of clinical outcome measures for recovery of gastrointestinal motility in postoperative ileus
. Ann Surg. 2014;259:708–714.
21. Venara A, Duchalais E, Dariel A, et al. Anti-inflammatory effects of enhanced recovery programs on early-stage colorectal cancer surgery. World J Surg. 2018;42:953–964.
22. Millan M, Biondo S, Fraccalvieri D, Frago R, Golda T, Kreisler E. Risk factors for prolonged postoperative ileus
after colorectal cancer surgery. World J Surg. 2012;36:179–185.
23. Chapuis PH, Bokey L, Keshava A, et al. Risk factors for prolonged ileus after resection of colorectal cancer: an observational study of 2400 consecutive patients. Ann Surg. 2013;257:909–915.
24. Vather R, Josephson R, Jaung R, Robertson J, Bissett I. Development of a risk stratification system for the occurrence of prolonged postoperative ileus
after colorectal surgery: a prospective risk factor analysis. Surgery. 2015;157:764–773.
25. Kronberg U, Kiran RP, Soliman MS, et al. A characterization of factors determining postoperative ileus
after laparoscopic colectomy enables the generation of a novel predictive score. Ann Surg. 2011;253:78–81.
26. Moghadamyeghaneh Z, Hwang GS, Hanna MH, et al. Risk factors for prolonged ileus following colon surgery. Surg Endosc. 2016;30:603–609.
27. Venara A, Slim K, Regimbeau JM, et al. Proposal of a new classification of postoperative ileus
based on its clinical impact-results of a global survey and preliminary evaluation in colorectal surgery. Int J Colorectal Dis. 2017;32:797–803.
28. The FO, Bennink RJ, Ankum WM, et al. Intestinal handling-induced mast cell activation and inflammation in human postoperative ileus
. Gut. 2008;57:33–40.
29. Meurette G, Blanchard C, Duchalais-Dassonneville E, et al. Sacral nerve stimulation enhances epithelial barrier of the rectum: results from a porcine model. Neurogastroenterol Motil. 2012;24:267–73, e110.
30. Schwarz NT, Beer-Stolz D, Simmons RL, Bauer AJ. Pathogenesis of paralytic ileus: intestinal manipulation opens a transient pathway between the intestinal lumen and the leukocytic infiltrate of the jejunal muscularis. Ann Surg. 2002;235:31–40.