PATIENTS AND METHODS

Patient Selection

Consecutive patients treated at a Chinese national referral center for gastroenterologic surgery (Jinling Hospital) between January 2005 and December 2007 were included in this study if they met the following criteria: 1) chronic constipation according to Rome III criteria²⁶; 2) long disease history (more than 6 years) with severely jeopardized quality of life and strong desire for surgical intervention expressed by the patient; 3) conservative treatment with diet (1.5 L/d water and high-fiber diet), laxatives (including saline, osmotic, and stimulant laxatives), enemas, and biofeedback tried over the past 2 years without success; and 4) coexistence of obstructive defecation syndrome and slow transit constipation confirmed by a thorough examination. Patients were excluded if they had a contraindication to general anesthesia, immunocompromised status, physical or psychological problems precluding data collection, inflammatory bowel disease, or septic conditions of the anorectum.²⁷

Evaluation consisted of 2 colonic transit time studies, double-contrast barium enema (revealing the presence of dolichocolon), anorectal manometry with electromyography (ruling out the presence of Hirschsprung disease), defecography, and small intestine follow through. Colonoscopy was scheduled only in selected cases to exclude the presence of neoplasia.

The diagnosis of obstructive defecation syndrome was mainly based on evidence of lack of rectal contrast emptying on imaging studies (internal rectal intussusception and/or rectocele, assessed clinically and confirmed by dynamic imaging), and functional abnormalities (puborectalis dysfunction and/or pelvic floor dyssynergia as assessed by anorectal manometry and electromyography). There were 2 main principles for simultaneous assessment of total and segmental colonic transit time studies: 1) to follow the transit of the radio-labeled particles with either repeated examinations of the transit of a bolus along the colon²⁸ or 2) daily administration of the particles over a longer period followed by a single examination.²⁹ Each patient underwent both types of colonic transit study preoperatively, one in another tertiary-care hospital and the other in our hospital. Slow-transit constipation was defined as retention of at least 20% of a single ingestion of 20 radio-opaque markers (SGmark) in the colon over 120 hours. Patients undergoing the Jinling procedure all had generalized slow colonic transit, instead of segmental transit.

The study was approved by the Institutional Review Board of Jinling Hospital, and informed consent was obtained from all participants.

Surgical Procedure

All patients were operated on by the same surgical team. The patient was catheterized and placed in a supine position. This position was held during the laparoscopic procedure with a mild lateral rotation (to the right during left colon dissection and to the left during right colon dissection). A pneumoperitoneum with intraabdominal pressure of 12 mm Hg was established according to an “open technique” with a Hasson trocar. A 30° camera was utilized. A 12-mm port was placed just cephalad of the umbilicus. A 5-mm port was used in the lateral left flank. One additional 12-mm working port was placed on the right side, and another 5-mm port was placed between the xiphoid and umbilicus.

A lateral-to-medial approach was used to mobilize the entire colon. The mesocolon was divided using bipolar cautery (Ligasure, ValleyLab, Boulder, CO) with careful identification and division of the left, middle, and right colonic vessels. The operation started with mobilization of the hepatic flexure. After dissection of the great omentum from the transverse colon and mesocolon followed by a complete mobilization of the descending colon with identification of the left ureter and gonadic vessels, a transverse surgical incision (about 5~7 cm) was made. We made this short transverse abdominal incision instead of keeping the procedure completely laparoscopic because it allowed a large amount of abdominal lavage (200 mL/kg body weight) and the specimen could be easily pulled through the incision.

After the extraction of the entire colon out of the abdominal cavity, appendectomy and colonic resection were performed, preserving about 10 to 12 cm of ascending colon from the ileocecal valve. The actual length of remaining ascending colon was determined individually, depending on the blood supply from the ileocolic artery. The ileocolic trunk was preserved, while the right colic vessel was removed. After careful resection of the fatty appendages and preservation of the vascular pedicles close to the bowel, the anvil of a 29-mm circular stapler (Ethicon Endo-Surgery) was installed at the ascending colon stump by means of a purse-string suture. It was essential to make sure the colon stump could move downwards to the pelvis. Rapid dissection of the presacral space was finished with Bovie electrocautery, with special care to preserve the sacral nerves. The superior rectal artery was divided using ultrasonic shears. Anterior dissection of the rectum was not carried out. The rectum was transected at the level of the sacral promontory with a 60-mm linear stapler, leaving a 9- to 10-cm-long rectal stump. (Fig. 1A). The rectal stump was sewn up in a 2-layer maneuver. An end-to-side ascending colorectal anastomosis was established with the 29-mm curved circular stapler (Ethicon Endo-Surgery) at the posterior wall of rectum (2 cm above the dentate line; Figs. 1B and C). Finally, one arm of a 75-mm linear cutter (Ethicon Endo-Surgery) was inserted into the rectum and the other arm was inserted into the distal colon (Fig. 1D), and another side-to-side ascending colorectal anastomosis (approximately 5 to 6 cm) was established (Figs. 1E and F). The abdominal cavity was then lavaged with a large amount of warm sterile saline and drained at the pelvis.

Patients with a previous surgical history underwent an open Jinling procedure. In this group, a midline incision approximately 20 to 25 cm long was made to better visualize the hepatic and splenic flexure. After mobilization of the entire colon, the same subtotal colectomy and anastomosis procedures were performed as in the above laparoscopic procedure.

Postoperative Treatments

Early postoperative diarrhea was conservatively treated with dietary fiber supplement, Bifidobacterium, and opioids until resolution. If diarrhea was severe (>10 times daily) or if risk factors for anastomotic fistula (such as malnutrition, diabetes, glucocorticoid treatment) were present, somatostatin was given. Tenesmus was treated with hot sitz baths and local application of Chinese herbal medicine (Taining suppository). The Jackson-Pratt drain was removed when the drainage fluid was less than 50 mL for 2 consecutive days. Occasionally, dual-lumen–tube suction drainage (sump draining)³⁰ was used intraoperatively to prevent pelvic sepsis caused by anastomosis leakage noted intraoperatively. The tube was left in place until the drain output was clear, and it was removed on approximately the 7th to 9th postoperative day. Most patients remained hospitalized until the retention sutures were removed.

Data Collection and Follow-up

Clinical data were collected before the operation, and all patients were asked to attend follow-up clinic visits 1, 6, and 12 months after the operation, and during phone interviews 24, 36, and 48 months after the operation. All data were recorded using prospectively determined variables and entered into the Jinling Hospital Constipation Registry System, which is an interventional (open label, nonrandomized), single-center audit system that can be retrospectively queried. The registry was managed by an independent clinical technician; all data input occurred immediately after patient discharge or each follow-up visit.

Safety

Safety evaluation included morbidity, adverse events, and mortality. Urinary tract infections, transient atelectasis, and benign postoperative nausea and vomiting were not considered serious complications. A surgical site infection was defined as one requiring incision and drainage or draining spontaneously; minor wound erythema was not considered a surgical site infection. Bleeding, infection, small bowel obstruction, urgency, tenesmus, postsurgical anal stenosis requiring interventional dilatation, urine retention, and deep vein thrombosis/pulmonary embolism were recorded.

Effectiveness and Quality Of Life

Effectiveness and quality of life were evaluated with the Wexner constipation scale,³¹ the Gastrointestinal Quality of Life Index (GIQLI),³² and a satisfaction survey in which patients were asked whether they were very satisfied, satisfied, or dissatisfied. All questionnaires were administered orally by a technician in order to render the questions and the modalities of replying to them more comprehensible to the interviewees than they would be with self-administration in this population.

The Wexner constipation scale is a validated and internationally adopted questionnaire used to quantify the severity of constipation.³¹ It consists of questions examining the various clinical expressions of constipation, with scores ranging from 0 (best) to 30 (worst).

The GIQLI is a validated quality-of-life questionnaire designed to evaluate specific gastrointestinal symptoms and the impact of the disease on the physical, psychological, and social spheres of the patients.³² Most of the questions concern the 2 weeks immediately preceding administration of the questionnaire. The questionnaire consists of 36 questions scored on a 5-point Likert scale, from 0 (worst) to 4 (best). Final scores range from 0 to 144, with a mean of 125.8 (SD, 13) for healthy people obtained during validation studies.³² In our study, data regarding abdominal pain and urgency, measured on a frequency scale from 0 (never) to 4 (always), refer to questions 1 and 30 of the GIQLI.

Statistics

Statistical analysis was performed using paired t tests for continuous variables, Pearson’s χ² and the Fisher exact test, where appropriate. A p value less than 0.05 was regarded as statistically significant.

RESULTS

Patient Characteristics

A total of 117 patients treated between January 2005 and December 2007 met the study criteria and underwent the Jinling procedure. Of these, 56 patients (47.9%) underwent the Jinling procedure laparoscopically, and 61 (52.1%) underwent an open Jinling procedure.

The patients’ demographic, clinical, and operative characteristics are given in Table 1. Most women (77/87) had had 1 delivery, and 10 of them were multiparous. Conversion to an open procedure was performed in 4 patients (3.4%) who had extensive intraabdominal adhesions. The median duration was 18.1 (range, 5–94) days for the total hospital stay, which included a median of 12.1 (range, 6–82) days for the postoperative period. Most preoperative tests were performed on an outpatient basis. The median time to resolution of ileus, defined as time to toleration of oral fluids, was 3 (range, 1–9) days. The mean total colonic transit time was 75.3 (SD, 19.8; range, 68.2–97.2) hours before surgery (the upper limit of the reference range for total preoperative colon transit time is 66.2 hours). Of 117 patients, 107 patients had a rectocele. Of these, 19 (17.8%) had grade I; 55 (51.4%) had grade II; and 33 (30.8%) had grade III. Of patients with grade I rectocele, 84.2% (16/19) had a concomitant internal intussusception.

Follow-up data were available for 112 patients (95.7%) at 1 month, 110 patients (94.0%) at 6 months, 109 patients (93.2%) at 1 year, 108 (92.3%) at 2 years, 106 patients (90.6%) at 3 years, and 103 patients (88.0%) at 4 years. At the time of the last phone visit, 14 patients (12.0%) had been lost to follow-up.

Surgical and Safety Data

Morbidity and adverse events are summarized in Table 2. A total of 28 major complications or adverse events were reported for 117 procedures, giving an overall morbidity rate of 23.9%; 23 patients (19.7%) had 1 or more events.

During the first 2 weeks after the operation, 54 patients (46.2%) developed early postoperative diarrhea (Table 2), which was conservatively treated with a dietary fiber supplement, Bifidobacterium, and somatostatin and not regarded as major complication. Daily doses of antidiarrheal medications were used by 2 patients for 6 months postoperatively. Constipation recurred in 1 patient. She showed normal gastrointestinal function preoperatively and received a diagnosis of diabetes 3 months after the operation. She began to have evacuation difficulties at 6 months, and constipation was diagnosed 2 years after the operation. A double-contrast barium meal study showed gastrointestinal tract dysmotility. We attributed recurrence of constipation in this patient to autonomic neuropathy due to uncontrolled diabetes.

Postoperative urinary retention occurred in 16 patients (13.7%), with 8 patients (6.8%) requiring catheterization within 1 week; all recovered within 1 to 2 weeks, except 2 patients who recovered after 4 weeks. Of 11 patients (9.4%) who developed small-bowel obstruction, 8 patients recovered with nonsurgical management, and 3 patients recovered after reoperation. A total of 9 patients (7.7%) had anastomotic bleeding, which was minor (less than 200 mL of blood lost at postoperative day 1) in 8 patients and considered major in 1 patient. All 9 patients were successfully treated with local hemostatic gauze compression at the bedside and pharmacological therapy with somatostatin and hemostatic drugs (intravenous hemocoagulase 2 KU every 8 hours). Another major complication was anastomotic leakage, which occurred in 4 (3.4%) patients; 3 were successfully controlled conservatively, but 1 patient developed a serious septic complication requiring fecal diversion. Two male patients (1.7%) developed sexual dysfunction, but the symptoms improved after pharmacological therapy. No incontinence was reported, and no surgery-related deaths occurred.

Effects on Constipation

In the study patients, bowel movement frequency increased from a mean of 1.2 (SD, 0.2) per week preoperatively to 6.3 (SD, 1.9) per week at 1 month and 4.2 (SD, 2.1) per week at 6 months postoperatively (p < 0.001). The mean bowel movement frequency was maintained at 3.3 per day at 24 months and 3.4 per day at 48 months.

As shown in Figure 2, comparison of preoperative and postoperative Wexner constipation scores demonstrated a significant reduction between baseline (mean, 21.9; range, 5–30) and 1 month (mean, 9.8; range, 3–13), which was maintained up to the 48-month follow-up (mean, 5.8; range, 0–11; p < 0.001).

Quality of Life and Satisfaction

Improvement in constipation scores was matched by an overall improvement in quality of life as judged by the GIQLI scores after 6, 12, 24, 36, and 48 months of follow-up (Fig. 2). After 1 month of follow-up, GIQLI scores (mean, 46.7; range, 39–67) had decreased (indicating worsened quality of life) compared with the preoperative scores (mean, 62.4; range, 56–104), presumably because of the increased daily bowel movements during the early phase. However, GIQLI scores at 6-month follow-up were significantly increased (mean, 89.0; range, 56–95) compared with the preoperative and 1-month follow-up evaluation, and this increase in quality of life was sustained at 12 months (mean, 98.7), 24 months (mean, 108.0), 36 months (mean, 115.8), and 48 months (mean, 110.8) after the surgery (p < 0.01).

The satisfaction rate (very satisfied or satisfied) was 80.4% (90/112) at 1 month, 90.9% (100/110) at 6 months, 92.7% (101/109) at 1 year, 94.4% (102/108)at 2 years, 93.4% (99/106) at 3 years, and 93.2% (96/103) at 4 years.

DISCUSSION

Multiple options exist for the treatment of pure obstructive defecation syndrome or for slow-transit constipation; however, none is generally accepted or even recommended by a guideline or a consensus statement for mixed constipation, in which refractory slow-transit constipation is associated with concomitant obstructive defecation. We developed the Jinling procedure to treat both conditions in one operation, in which we modified the classic procedure of subtotal colectomy with colorectal anastomosis by adding a new cecorectal anastomosis. The rationale for the Jinling procedure includes 2 main points: 1) Subtotal colectomy eliminates the delayed colonic transit. 2) The rectum stump is pulled straight by the distal mesenterium, acting as a rectopexy, and therefore the pelvic floor musculature is elevated, which significantly relieves rectocele, intussusception, and prolapse.

We began performing the Jinling procedure in March 2000 and conducted a retrospective analysis of follow-up data from 81 patients at the end of 2005, which we published in Chinese.²² Because the early results were promising, with a high satisfaction rate, we enrolled more patients. The current study examined the durability of results at 4-year follow-up. To our knowledge, this study of 117 patients represents one of the largest reported series with a long term follow up of patients receiving surgery for constipation. As a leading Chinese tertiary-care gastroenterology surgical center and teaching hospital, Jinling Hospital receives patients nationwide and worldwide. Thus, 98.3% (115/117) of the study patients came from or were referred from other cities in China.

In the 4-year follow-up study presented here, the Jinling procedure continued to provide promising outcomes. No severe adverse events occurred, and most complications could be managed conservatively. The overall satisfaction rate was also significantly improved after the procedure and sustained up to 48 months. An overall improvement in Gastrointestinal Quality of Life Index (GIQLI) further verified its effectiveness. In addition, most patients stated that they would undergo the procedure again.

In our clinical experience, slow-transit constipation and obstructive defecation seldom exist in isolation.²² Patients with a long course of constipation who come to our surgical clinic, are willing to accept the risk of potential severe complications after surgery, and ask for the surgeon’s help, are likely to have developed severe chronic constipation of the mixed type.²³ In patients with slow-transit constipation, prolongation of transit time through the colon may cause impaired propagated colonic contraction, force the patient to overexert during evacuation, and ultimately cause tenesmus and excessive straining, which are the characteristics of obstructive defecation. Conversely, in patients with typical symptoms of obstructive defecation, long-time difficulty or incomplete evacuation might also cause overperistalsis and a disorder of the myenteric plexus, thus unavoidably resulting in myopathy or neuropathy of the whole colon. Dinning et al³³,³⁴ found that in patients with severe obstructed defecation without an obvious mechanical cause such as rectocele or levator weakness, abnormal colonic waveforms are found in the distal as well as the proximal colon.³³,³⁴

Total abdominal colectomy with ileorectal anastomosis represents the procedure of choice in patients with pure slow-transit constipation and relieves constipation in most patients.³⁵ However, a significant proportion continues to experience other symptoms, particularly urgency and diarrhea. According to Ternent et al,² studies have shown that, postoperatively, 41% of patients are affected with abdominal pain, 65% with bloating, 29% require assistance with bowel movements, 47% have some incontinence to gas or liquid stool, and 46% may be affected with diarrhea. Results after subtotal colectomy are good for pure slow-transit constipation and are sustained over follow-up, but patients must be carefully warned about potential operative morbidity, and even the small possibility of a permanent ileostomy.^36–38 In the Jinling procedure, we chose subtotal colectomy with cecorectal anastomosis (SCCA) as another alternative to more widely adopted surgical approaches for the treatment of slow-transit constipation.³⁹ We believe the preservation of the terminal ileum, cecum, and ileocecal valve is of particularly great importance for Asian populations and theoretically represents a physiological advantage over ileorectal anastomosis, avoiding malabsorption and colonic bacterial contamination of the small bowel. Moreover, the cecal reservoir and its water-absorbing function would seem to provide an additional preventive factor regarding diarrhea, urgency, and incontinence, without affecting the efficacy of SCCA in alleviating constipation.

There are several reasons why we performed another side-to-side cecorectal anastomosis on the basis of SCCA, which successfully relieved the symptoms of obstructive defecation. First, the rectal stump was fixed together with the distal mesenterium after the colorectal anastomosis. We assumed that postoperative sterile inflammation would develop at the pelvis and form a pulling force between the rectum stump and mesenterium. The force would act as a rectopexy, which would significantly relieve the symptoms of obstructive defecation (rectocele, intussusception, or prolapse). These assumptions were supported by analysis of defecography data at 1-month follow-up after the Jinling procedure in a later cohort of patients: defecography showed 98% disappearance of rectocele, prolapse, and intussusception (Fig. 3).²⁵ Second, end-to-side colorectal anastomosis alone, although maintaining the ileocecal valve, was often complicated by cecal distention¹⁷,¹⁸ and anastomosis stenosis. In the Jinling procedure, another side-to-side cecorectal anastomosis was added so that the large pouch could act as a reservoir to decrease the occurrence of diarrhea, urgency, and incontinence. This concept was supported by the follow-up results in the current study. Twelve patients who had experienced recurrence of constipation after STARR (Table 1) successfully underwent the Jinling procedure in our department. The surgical staples in the low rectum did not affect the accomplishment of Jinling procedure because the staple line was re-cut by the linear cutter during the side-to-side anastomosis.

At first glance, the overall morbidity/adverse event rate of 23.9% (1 or more events in 19.7% of patients) seems high, but most complications could be managed conservatively. Anastomotic leakage occurred in 4 (3.4%) patients, however, only 1 needed a temporary stoma, and most were successfully controlled by dual-lumen–tube suction drainage (sump draining) and total parenteral nutrition. As the leading treatment center for intestinal fistula in China, our department treats more than 200 patients for fistula annually from all over the world, including Canada, Africa, and Southeast Asia. Therefore, the low reoperation rate of fistula was not surprising. Constipation reoccurred in 1 patient who had coexisting gastrointestinal tract dysmotility attributable to severe autonomic neuropathy due to uncontrolled diabetes. Only 11 (9.4%) of our patients had small –bowel obstruction. In contrast, a review by Pfeifer et al⁴⁰ found a high rate of postoperative small-bowel obstructions, occurring in as many as 50% of patients in some series. Pikarsky et al⁴¹ found 20% of patients required admission for small-bowel obstruction after total abdominal colectomy with ileorectal anastomosis for colonic inertia. We believe that the low incidence of small-bowel obstruction can be attributed to a small surgical incision, large amount of intraoperative abdominal lavage fluid (200 mL/kg body weight), and careful closure of the retroperitoneum after removal of the colon. We did observe some seepage (about 20%) in the first week after the operation, especially in the older women. It was corrected after early ambulation and use of antidiarrheal drugs. Anorectal manometry 1 month after the Jinling procedure (data not shown) indicated that the patients retained normal sphincter pressures at rest and squeeze, and no incontinence was found at follow-up.

Extensive clinical and physiological preoperative assessment of patients with constipation is essential before considering surgery, including an assessment of small-bowel motility and the psychological state of the patient. Surgery should only be considered in patients who have severe refractory constipation. From our clinical practice, careful preoperative work-up and meticulous selection of patients is critical for obtaining good functional results.

Although this study suggests a favorable long-term outcome for patients who undergo the Jinling procedure, it has several limitations. First, conclusions drawn from these results are limited by the fact that this was not a comparative trial, and it was conducted in only 1 center. Future randomized controlled trials are warranted to compare the different surgical procedures, such as SCCA + biofeedback or SCCA + STARR, with the Jinling procedure. Second, by the time of the last follow-up at 48 months, 14 (12%) of the patients had been lost. Thus, follow-up issues may have influenced results if, for example, patients with poor function or complications were more likely to be lost. Third, patients in this study did not undergo postoperative ultrasound examination or MRI, which would have been useful to show the effects of the surgery on the puborectal muscle. Namely, because we found in a different cohort that anterior rectocele, prolapse, intussusception, and descending perineum were significantly improved at defecography 6 months after the Jinling procedure,²⁵ we presumed that the puborectal muscle, responsible for most cases of puborectalis syndrome, might be partially divided by the circular stapler. Thus, the ultralow location of the anastomosis (about 2 cm above the dentate line) would avoid the influence of the puborectal muscle. A longer term follow-up and more comprehensive evaluation of pelvic morphologic correction after the Jinling procedure is warranted.

CONCLUSIONS

The optimal surgical treatment for slow-transit constipation associated with concomitant obstructive defecation is controversial. Based on the pathophysiologic changes of mixed constipation, we have proposed a new surgical procedure (Jinling procedure), which adds a new side-to-side anastomosis to the colorectal posterior anastomosis after subtotal colectomy. Our clinical practice indicates that the Jinling procedure is safe and effective for refractory slow-transit constipation associated with obstructive defecation, with minimal major complications, significant improvement of quality of life, and a high satisfaction rate after 4-year follow up.