Electrical stimulation of the gastrointestinal (GI) tract is an attractive concept. Since these organs have their own natural pacemakers, the electrical signals they generate can be altered by externally delivering electrical currents by intramuscular, serosal, or intraluminal electrodes to specific sites in the GI tract [1].

Over the past decade, some electrical stimulation methods for the treatment of gastric motility disorders have been developed, including implantable electrodes connected to a subcutaneous pacemaker [2]. Although electrical stimulation of the colon has been successful in improving constipation in animal models, there have been few evaluations of its use for constipation in humans. Sacral nerve modulation (SNM) has shown interesting results for the treatment of slow-transit constipation [3, 4], but there are no data about permanent colonic pacing.

Here we report on a procedure for electrical colonic pacing due to intramuscular electrode placement for slow-transit constipation and some preliminary results are given.

Materials and methods

From January 2011 to December 2012, consecutive patients affected by constipation and evaluated in our Pelvic Floor Center/general emergency and minimally invasive surgery unit were prospectively assessed. Patients who underwent colonic electrical stimulation were evaluated for the present study. Patients were considered for colonic electrical stimulation if they met the following criteria:

  • failed all the conventional therapies, including laxatives and/or enemas, dietary and habit modification, biofeedback, prucalopride, or transanal irrigation

  • failed SNM

  • had no evidence of obstructed defecation at defecographic (or magnetic resonance defecography) and clinical evaluation

  • was between 18 and 45 years old

  • had a slow transit time confirmed radiologically (gut transit time with radiopaque markers with plain abdominal radiography performed at 72 and 120 h after ingestion of the markers) without radiological rectal fecal impaction. The diagnosis of slow-transit constipation was suggested if >20 % of the markers were found in the colon after 120 h

  • had less than one bowel movement per week

  • had absence of bowel movement for more than 1 week without laxatives or enema

  • had absence of fecal residue in ampulla after 1 week without laxatives or enema

  • anorectal manometry had no signs of anorectal dyssynergia

  • had a history of constipation from a young age (or at least for more than 3 years)

  • irritable bowel syndrome (IBS) was excluded

  • psychological disturbance was excluded

Exclusion criteria included congenital anorectal malformations, external rectal prolapse, chronic inflammatory bowel disease, the presence of stoma, pregnancy, neurological disease, and psychiatric or physical inability to comply with study protocol. All patients selected for the study gave written informed consent.

For the slow-transit study, plain abdominal radiographs are taken 3 and 5 days after subjects ingest a capsule containing 20 radioactive markers. The presence in the colon (not accumulated in the rectum) of at least 8 markers on day 3 or at least 5 markers day 5 was considered abnormal [5]. Patients were instructed to abstain from the use of enemas, laxatives, or suppositories of any kind for the 5 days of the study. Radiography on day 5 was considered unnecessary for patients with fewer than 8 markers on day 3.

Before the procedure the patients underwent clinical investigation, including colonoscopy and proctography, to exclude other correctable causes of constipation. A bowel diary was kept from 3 months before to 3 months after the procedure, in which the patient recorded the number of bowel movements per week, the time spent in the bathroom, the presence or absence of a stimulus to assist defecation, and the use of laxatives or enemas. Patients were evaluated after 7 and 15 days, then every 30 days for the first 3 months and subsequently by telephone each month, evaluating the number of bowel movements per week and the maintenance of the effectiveness of the treatment.

Surgical technique

Patients were given a complete bowel preparation (with PEG or saline solution) and preoperative antibiotics (ceftriaxone 2 g+metronidazole 500 mg) were administered. Patients were placed on a beanbag table pad and padded stirrups were utilized. The anus should be lined up with the edge of the bed to facilitate endoscopy. Four trocars were placed: the umbilical port, the upper-right quadrant, the lower-right quadrant, and the lower-left quadrant (optional), as performed for left colonic laparoscopic procedures. The table was then turned on the right side in the Trendelenburg position. After exploration of the abdominal cavity, the sigmoidorectal junction was identified and the optimal area for placement of the electrodes was considered approximately at the level of the confluence of the taeniae anteriorly.

Two parallel 35-cm leads with a 10-mm electrode (Unipolar Intramuscular Lead model 4351, Medtronic, Minneapolis, MN, USA) (Fig. 1) were placed into the muscular layer, 1 cm apart, above the taenia and anchored with a silicone rubber fixation disc. Both the disc and the trumpet anchor on the other side of the electrodes were secured with individual sutures to the bowel wall. Correct electrode position was endoscopically verified before the fixation. A left low inguinal subcutaneous pocket was created (incision about 3 cm) and the leads were placed in an extraperitoneal tunnel, bringing them out at the level of the pocket. The tunneling near the site of insertion was done to avoid possible complications related to the intra-abdominal presence of the leads. The leads were then connected with the neurostimulator (Interstim neurostimulator model 3023, Medtronic) and the neurostimulator was placed in the pocket. The surgical procedure is summarized in Fig. 2. The stimulation parameters were activated by the physician and a remote control was given to the patients to turn stimulation on and off.

Fig. 1
figure 1

Lead detail with muscular electrode, trumpet anchor, and fixation disk

Full size image
Fig. 2
figure 2

Procedure description. A Insertion of the needle on the muscular layer. B Fixing of the disk with two clips. C Definitive results of the double electrode placement. D Extraperitoneal tunneling

Full size image

Stimulation parameters

After the first activation, for 30 days the stimulation parameters were a pulse width of 150 μs, a rate of 10 Hz, and a voltage of 2 V in continuous mode. After 30 days the mode was cyclically changed, with 2 min ON and 20 min OFF. Voltage could be lowered if required by the patient.

Results

In the study period, 256 patients were evaluated as outpatients for constipation (Fig. 3). A total of 149 patients (58 %) had a main diagnosis of outlet obstruction or obstructed defecation syndrome related to rectal prolapse, rectocele, or intussusception; 70 patients (27.3 %) had IBS or mixed forms; 10 (4 %) patients were identified as having pelvic floor dyssynergia or paradoxical puborectalis muscle contraction; and 27 patients (10.5 %) were diagnosed with slow-transit constipation.

Fig. 3
figure 3

Constipated population evaluated in our center from January 2011 to December 2012

Full size image

The diet and habits of all 27 patients were reviewed and medical therapy (with fiber or PEG) was proposed. Fifteen patients (55.5 %) failed to benefit from this approach. Six more patients solved their problem with laxatives and enemas (3), transanal retrograde irrigation (2), and biofeedback, pelvic floor rehabilitation, and manual therapies (1). Nine patients (33.3 %) underwent a temporary stimulation period (6–8 weeks) with SNM, and five of those patients (55.5 %) were definitively implanted after the success of the treatment. Of the remaining four patients, one refused any more treatment and one was affected by a neurological syndrome and demolishing surgical treatment was proposed. Two patients, candidates for colectomy, agreed to undergo colonic electrical stimulation before resection. Both patients were females, aged 34 and 29 years, and suffered from severe constipation since childhood.

Follow-up was 19 and 6 months and data of the two patients is given in Table 1. The number of bowel movements per week increased from 0.3 to 3.5 in the first patient and from 0.5 to 2.5 in the second patient. Both patients no longer needed laxatives, enemas, or any other assistance. The hospital stay was 4 days. The mean operative time was 120 min. No perioperative or postoperative complications were reported. The patients did not require changes in the stimulation parameters during the follow-up period.

Table 1 Patient data
Full size table

Discussion

Patients with slow-transit constipation account for 5–15 % of the constipated population [6], and severe constipation (e.g., bowel movements only twice a month) is seen almost exclusively in young women [7], as was seen in our data.

During the past decades, much attention has been paid to abnormalities in autonomic nerves that are associated with colorectal motility disorders. The autonomic nervous system can be divided into parasympathetic and sympathetic components. The parasympathetic, general visceral efferent innervation of the large bowel is derived from the dorsal motor nucleus of vagi and the sacral parasympathetic nucleus. Even though the extent of colonic innervation is still under debate, it is generally believed that vagal innervation to the large bowel terminates at the level of the splenic flexure [8, 9], while the remainder of the colon, including the rectum, receives parasympathetic input from the pelvic nerves [10–12]. A pattern of dual, coordinated, parasympathetic innervation in the left colon may regulate motor activity between the proximal colon and the rectum [13]. The distal colon and rectum also receive sympathetic input from the hypogastric nerves (HGN), derived mainly from the lumbar preganglionic outflow that runs to the inferior mesenteric ganglia (hypogastric ganglion). The innervation and functional roles of the HGN on the internal anal sphincter have been well studied [14]. However, it still remains unclear how the HGN regulates colorectal motility.

Most of the colonic motor activity is represented by single nonpropagated contractions, rarely organized in bursts; this activity is maximal during the day, especially after waking and following meals. In addition, a specialized propagated activity with propulsive features is detectable, represented by high- and low-amplitude propagated contractions. In the severe form of constipation, represented by the slow-transit type, this motor activity is completely deranged. In fact, both basal segmental activity and propagated activity are usually decreased [15].

Based on data generated from colonic manometry studies, there are some primary indicators of abnormal motility that emerge: reduced frequency of high-amplitude propagating sequences, diminished or absent response to eating a high-calorie meal or morning waking, and abnormal colonic response to chemical stimulation or rectal mechanical distension [16]. As the central nervous system is likely to play role in both the increase in propagating pressure waves after a meal and their nocturnal suppression, a diminished or absent response to these stimuli has been proposed as a possible indicator of myopathy or neuropathy [17]. With regard to chemical stimuli, a failed response may indicate an abnormality within the myenteric plexus [18], cholinergic pathways [19], or rectocolonic neural pathways [20].

Considerable progress has been made at the ultrastructural, molecular, and electrophysiological levels in understanding the normal functions of the muscles, nerves, and interstitial cells that generate and control colonic motility. Furthermore, abnormalities in these cell types, and in the interstitial cells of Cajal in particular, have been identified in a number of disease states [21]. In patients with slow-transit constipation, the number of interstitial cells of Cajal was significantly decreased in all layers except the outer longitudinal muscle layer and the myenteric plexus showed a variable grade of hypoganglionosis [21].

Several studies showed the positive effects of direct colonic stimulation in animal models, even if it has never been reported in humans [22–25]. Shafik et al. [26] reported on endoscopic placement of mucosal electrodes connected to a subcutaneous stimulator for direct pacing of the colon. However, direct stimulation in this manner is technically demanding and has never been duplicated by others. Moreover, we preferred the placement of the electrodes in the seromuscularis layer because of the guaranteed contact and direct effect on the targeted organ but the obvious disadvantage is the invasiveness. Laparotomy or laparoscopy under general anesthesia is required. This led us to the proposed procedure just before a colonic resection, considering the medical or minimally invasive treatments (e.g., SNS, colonic irrigation) as preferred treatment of this benign but disabling condition.

Sacral neuromodulation represents a promising alternative via indirect sacral roots stimulation [3, 4], but the results are inconsistent and the mechanism of action is still unclear. Sacral nerve stimulation did not induce major changes in the propulsive capacity of the GI tract in patients successfully treated for fecal incontinence [27], and even though some authors reported that it induces pancolonic propagating pressure waves [28], the clinical implications of this result remain unclear.

According to the colonic innervation, it was demonstrated in a canine model that sacral nerve stimulation elicited contractile movements propagating from the distal colon to the rectum, relaxation response in the rectum, and internal anal sphincter relaxation response, but no proximal colonic effective responses, suggesting its role in modulation of sacral reflexes and confirming its inconsistent outcome in the treatment of constipation [29].

The role of the sigmoid colon in the pathophysiology of slow-transit constipation is well known but not completely clear, considering that megacolon (mainly left) is the main clinical finding in Hirschsprung’s disease, that a dolicho-megacolon (mainly left) is a common feature of patients affected by chronic slow-transit constipation, that diverticular disease manifests in left and sigmoid colon, and an incomplete sigmoid resection for diverticulitis exposes the patient to a higher recurrence rate.

Connell [30] raised the possibility of a sigmoid “brake” by demonstrating that the duration but not the average amplitude of phasic motor activity in the sigmoid colon was greater in constipated subjects but lower in patients with functional diarrhea compared to controls. Similarly, Preston and Lennard-Jones [18] suggested that phasic pressure activity in the sigmoid colon was higher in normal-transit constipation than in controls or in patients with slow-transit constipation. Thus, we believed that the sigmoid colon plays a key role in slow-transit constipation, also taking into account the complex and variable innervation, different from that of the rest of the colon. Moreover, the visionary theories (and experiments) of Shafik et al. [31, 32] suggested the presence of a rectosigmoid junction pacemaker and a colosigmoid functional sphincter, regulated by rectosigmoid and rectocolonic reflexes.

All these observations suggest that slow-transit constipation (or certainly some cases) could be related to conflicting neurogenic input received by the left/sigmoid colon, maybe associated with (or the cause of) hypogangliosis and alteration of interstitial cells. For this reason, we belief that a direct left colonic modulation (maybe at the sigmoidorectal junction) could theoretically resolve the disease and restore a physiological (or regular) electrical transmission.

Regarding the stimulation parameter, considering the lack of consensus about optimal stimulation features and starting from the suggestion of Shafik et al. [33], the studies of Aellen et al. [24], Bertschi et al. [34], and Yin and Chen [35], the gastric Enterra pacing experience [2, 36], and some general principles of electrical stimulation [37], a pulse width of 150 μs, a rate of 10 Hz, and a voltage of 2 V in continuous mode are proposed, with a change in cyclic mode, with 2 min ON and 20 min OFF after 30 days, trying to restore a more physiological stimulation protocol.

Conclusions

Colonic pacing seems to be feasible and has shown positive results. Deeper neurophysiopathological studies need to be performed for a better understanding of colonic motor function, but electrical stimulation seems to be a promising solution for the treatment of slow-transit constipation. Further studies with a larger number of patients and a longer follow-up are required.