Surgical Re-interventions Following Colorectal Surgery: Open Versus Laparoscopic Management of Anastomotic Leakage

Abstract

Background:

Increasing numbers of colorectal resections are performed laparoscopically each year. In 2010, 42% of all colorectal procedures in The Netherlands were performed laparoscopically. Although the anastomotic leakage rate is 3%–19% of all patients, little is known about laparoscopic options for re-intervention. Our study aims to evaluate the safety and feasibility of laparoscopic re-intervention compared with open surgery following colorectal surgery.

Patients and Methods:

All patients who required a surgical re-intervention for an anastomotic leak, bowel perforation, or abscess after laparoscopic colorectal surgery between January 2008 and June 2012 were analyzed retrospectively. Demographic data, operative management, morbidity, hospital stay, and mortality were collected and analyzed for each patient.

Results:

Fifty-six patients were included. Eighteen patients had a laparotomy following laparoscopy, and 38 patients had a laparoscopic re-intervention following laparoscopy. The median age was 65 years, with a median body mass index of 26 kg/m². Four patients had a previous laparotomy, and 73% had surgery for malignant colorectal disease. The length of hospital stay was 20 days in the laparoscopic group versus 31 days in the open group (P=.044). Six out of 38 versus 7 out of 18 patients required an additional re-intervention (P=.056). Fewer patients developed fascial dehiscence in the laparoscopic group (P=.033). In-hospital mortality was 4 out of 18 in the open group compared with 2 out of 38 in the laparoscopic group (P=.077).

Conclusions:

Laparoscopic re-intervention could be a safe and feasible treatment for anastomotic leakage after laparoscopic colorectal surgery. These promising results need to be further investigated in a prospective study to reduce uncertainty in the patient's condition and perioperative findings.

Introduction

Historically, abdominal surgery implied laparotomy, while today colorectal surgery can be performed using laparoscopy, robotic surgery, or hand-assisted laparoscopic surgery as well. Data from the Dutch Surgical Colorectal Audit provide insight into the contemporary surgical treatment of colorectal cancer in The Netherlands; over 9000 oncological colorectal surgical procedures are performed each year, with laparoscopic surgery accounting for 42% of all elective procedures.¹

Previous research has proven laparoscopic techniques to be superior to laparotomy in regard to intraoperative blood loss, recovery, hospital stay, abdominal wall complications, and number of re-interventions.^2,3

Anastomotic leak, the most important complication in colorectal surgery with a prevalence of 3%–19%, is associated with a 30-day mortality of 6%–22%.⁴ Oncological outcomes are also affected in the case of an anastomotic leak, as rates of local recurrence and distant metastases are higher, and cancer-specific survival is reduced.^5–8 Therefore, complicated colorectal surgery has major short- and long-term consequences for patients. Although a significant amount of research has been done for prevention of anastomotic leakage, not much is known about the preferred approach for management of leaks. In the absence of peritonitis, successful conservative treatment and radiological drainage for limited anastomotic leaks and abscesses have been described for low rectal anastomoses.⁹ When surgery is indicated, the choice between resection and diversion of the anastomosis with a loop ileostomy remains unclear. Although most surgeons advocate resection of the anastomosis and end colostomy, some data suggest that management with lavage and a loop ileostomy is feasible and associated with less mortality and morbidity if no sepsis or fecal contamination is present.^9–14

With the aim of less invasive treatment of the postoperative acute abdomen, our study aims to assess safety and feasibility of laparoscopic re-intervention compared with open re-intervention in case of the suspected anastomotic leakage after colorectal surgery.

Patients and Methods

Patients

In this retrospective study, all consecutive patients who underwent surgical re-intervention following laparoscopic elective benign and malignant colorectal surgery in two teaching hospitals between January 2008 and June 2012 were assessed for inclusion. Patients who underwent colorectal re-intervention following hepatobiliary, abdominal vascular, gynecological, or urological surgery were excluded. Patients who underwent acute surgery, patients with a continuity restoration procedure, and patients with Dukes' D carcinoma were excluded. Only patients requiring surgical re-intervention for clinical or radiologically suspected anastomotic leakage, bowel perforation, or abscesses after laparoscopic colorectal resection were included.

Two groups were formed according to the technique of re-intervention. The first group included all patients who underwent open re-intervention, and the second group included the patients who underwent a laparoscopic re-intervention. Patients who required an additional open re-intervention after laparoscopic re-intervention and patients in whom laparoscopic re-intervention was converted to open surgery were included in the laparoscopic re-intervention group according to the intention-to-treat principle.

Patients' medical charts were reviewed, and data were collected on age, gender, body mass index, comorbidity, and primary procedure. Comorbidity was defined as diabetes, pulmonary disease, or cardiac disease with New York Heart Association class ≥2. Re-intervention data collected included the timing and indication of re-intervention, C-reactive protein and white blood cell levels, and radiology exams prior to re-intervention. Perioperative data including severity of peritoneal contamination, the condition of the anastomosis, and the procedure performed were collected from surgical reports. Conversion, defined as a laparotomy or extension of the extraction site incision, was recorded.

Clinical outcomes were length of hospital stay, intensive care unit (ICU) admission rate, ICU length of stay, presence of organ failure, fascial dehiscence, the number of additional surgical and radiological re-interventions, and in-hospital mortality following re-intervention. Severity of organ failure in the ICU was scored as minor (requiring surveillance only), single, or multi-organ failure. Surveillance only was defined as <24 hours of respiratory support and no hemodynamic or renal failure. Respiratory failure was defined as need for respiratory support for >24 hours, cardiovascular failure was defined as need for inotropic support, and renal failure was defined as need for hemodialysis or hemofiltration. Multi-organ failure was failure of two or more organ systems at the same moment in time.

Operative technique

Surgical re-interventions were performed with the patient under general anesthesia with additional epidural analgesia if hemodynamically stable. Pre- and postoperative antibiotics were provided to all patients. In both open and laparoscopic procedures, a urinary catheter and nasogastric tube were placed. In the case of laparoscopic re-intervention, patients were placed on a vacuum mattress and leg support for the lithotomy position. The standard supine position was used in open procedures. For laparoscopic procedures, access to the abdominal cavity was achieved by open introduction of trocars using the previous incision sites. A range of 14–20 mm Hg pressure to achieve pneumoperitoneum was accepted in laparoscopic re-interventions. Contraindications for laparoscopic re-intervention were a massive distended abdomen due to severe ileus and severe hemodynamic instability.

Laparoscopic procedures were converted to laparotomy in the case of a severely distended bowel or severe adhesions that prevented safe manipulation and sufficient overview.

In the case of anastomotic failure, the anastomosis was resected, and an end colostomy or ileostomy was created. When no or only a minor leak was seen, in the absence of fecal contamination, the decision between resection or diversion of the anastomosis with a loop ileostomy or drainage alone was made at the discretion of the surgeon. In all patients, a peritoneal lavage was performed extensively with at least 3 L of warm saline. Large-bore silicone drains were placed when indicated.

Statistical analysis

Continuous data were presented as median and interquartile range or mean±standard deviation values. Dichotomous and categorical data are presented as numbers with percentages. As the data were not normally distributed, continuous data were assessed using the Mann–Whitney U test. Pearson's chi-squared or Fisher's exact test was used for categorical and dichotomous data. A two-sided P value of less than .050 was considered statistically significant. Analyses were performed using Statistical Package for the Social Sciences version 17.0 software (SPSS, Chicago, IL).

Results

Patient characteristics

During the study period, in total, 56 patients with a surgical re-intervention and fitting the inclusion criteria were identified, of whom 38 had laparoscopic re-intervention and 18 had open re-intervention. During this period, the number of elective colorectal resections performed laparoscopically increased from 46% in the first year to over 90% in the last year. The overall anastomotic leakage rate during the study period was 5%. Characteristics of the two patient groups were compared in terms of age, gender, body mass index, American Society of Anesthesiologists classification, comorbidity, surgical indication, and type of resection (Table 1). The patients included had a median age of 65 years, and 59% were male. The median body mass index was 26 kg/m², and in both groups 4 patients had had a previous laparotomy. Six patients in the laparoscopic group were classified as American Society of Anesthesiologists III versus 1 in the open group (P=.318). The overall comorbidity rate was lower in the open group with 17% versus 47% in the laparoscopic group (P=.039). In the open group 83% of the patients had had surgery for malignant colorectal disease versus 68% in the laparoscopic group (P=.338). The surgical procedures performed were 15 right segmental colectomies, 29 left segmental colectomies, 11 rectal resections, and 1 subtotal colectomy. Of the 95% of patients with a primary anastomosis, 17% had a deviating ileostomy for protection of the anastomosis.

Table 1.

Patient Characteristics

	Lap-Open (n=18)	Lap-Lap (n=38)	P value^a
Age (years)^b	65 (57–76)	65 (58–74)	.712^c
Male:female ratio	10:8	23:15	0.724^d
BMI (kg/m²)	26 (22.5–27.5)	26 (23–27)	0.923^c
ASA classification [n (%)]			0.318^d
ASA I	9 (50)	12 (32)
ASA II	8 (44)	20 (53)
ASA III	1 (6)	6 (15)
Previous laparotomy	1 (6)	3 (9)	1.000
Comorbidity [n (%)]	3 (17)	18 (47)	.039
Diabetes	1 (6)	4 (13)	.642
Pulmonary disease	1 (6)	9 (24)	.143
NYHA ≥2	3 (17)	9 (24)	.732
Malignancy [n (%)]	15 (83)	26 (68)	.338
Type of operation [n (%)]			.349^d
Right segmental	3 (17)	12 (32)
Left segmental	10 (56)	19 (50)
Rectal resection	4 (22)	7 (18)
Subtotal	1 (6)	0
Type of anastomosis [n (%)]			.858^d
Ileocolic	7 (39)	15 (39)
Colonic or colorectal	10 (56)	22 (58)
Stoma [n (%)]			.852^d
End colostomy	1 (6)	1 (3)
Loop ileostomy	3 (17)	6 (16)
Surgical interval (days)^b	6 (4–10)	5 (3–9)	.443^c
Within 7 days (%)	12 (67)	23 (61)	.658^d
Radiology exams done [n (%)]	13 (72)	31 (82)	.646^d
Exam shows leakage	11 (85)	28 (90)	.339^d
Inflammatory markers^b
C-reactive protein (mg/L)	242 (140–360)	206 (130–340)	.584^c
WBC count (10⁹/L)	10 (7–13)	10 (7–13)	.936^c

Calculated by Fisher's exact test unless indicated otherwise.

Data are median (interquartile range) values.

By Mann–Whitney U nonparametric test.

By chi-squared test.

ASA, American Society of Anesthesiologists; BMI, body mass index; Lap, laparoscopic; NYHA, New York Heart Association; WBC, white blood cell.

The interval until re-intervention was a median of 5–6 days for both groups (P=.443), with 63% of the re-interventions being performed within the first postoperative week.

In 79% of the patients a radiological exam was done because of the suspicion of an anastomotic leak. In most patients a computed tomography scan of the abdomen was done; some others had a conventional abdominal x-ray with a contrast enema or a plain chest x-ray showing a pneumoperitoneum. Of these examinations, 89% showed signs of an anastomotic leak or perforation.

Prior to re-intervention, the median level of C-reactive protein was 206 mg/dL in the laparoscopic group versus 242 mg/dL in the open group (P=.584). The white blood cell count was 10,000/mm³ in both groups (P=.936).

Characteristics of re-interventions

During re-intervention, the severity of peritoneal contamination was recorded in 47 of 56 patients (84%). Anastomotic dehiscence was seen in 71% of the patients in the open group versus 63% of the laparoscopic group (P=.629). Fecal contamination was more frequently present in the open group compared with the laparoscopic group (53% and 24% of patients, respectively) (P=.039) (Table 2).

Table 2.

Re-intervention Characteristics

	Lap-Open (n=18)	Lap-Lap (n=38)	P value^a
Perioperative findings [n (%)]			.629
Anastomotic dehiscence	12 (71)	19 (63)
Small bowel perforation	2 (12)	2 (7)
No leakage seen	3 (18)	9 (30)
Peritoneal contamination seen [n (%)]			.039
Clear fluid	1 (6)	6 (18)
Purulent	6 (35)	12 (35)
Fecal	9 (53)	8 (24)
Abcess	1 (6)	0
Main intervention [n (%)]			.001
End colostomy/ileostomy	12 (67)	15 (40)
Loop ileostomy	0	14 (37)
Closure of perforation	4 (22)	0
Abdominal lavage	2 (11)	9 (24)
Positive abdominal fluid culture (n/total n)	14/14	23/26	.342
Surgical time (minutes)	76 (56–99)	58 (45–75)	.043^b

Calculated by chi-squared test unless indicated otherwise.

By Mann–Whitney U nonparametric test.

Lap, laparoscopic.

In all cases of fecal contamination, the anastomosis was resected, and an end ileostomy or colostomy was created. In 14 laparoscopic patients, a loop ileostomy was created versus none in the open group (P=.001). Abdominal fluid cultures were taken in 71% of all patients and tested positive for abdominal contamination in all open patients and 88% of the patients in the laparoscopic group. Surgical time for the re-intervention procedure was shorter in the laparoscopic group with a median of 76 minutes for the open group versus a median of 58 minutes for the laparoscopic group (P=.043). Five patients (13%) in the laparoscopic group required conversion during inspection because of adhesions and poor visualization.

Clinical outcomes

The total length of hospital stay was shorter for the laparoscopic group at 20 days versus 31 days for the open group (P=.044) (Table 3). In the laparoscopic group, 58% of the patients were admitted to the ICU compared with 72% in the open group. Those who were admitted stayed a median of 3–4 days in the ICU (P=.830). In both groups, about half of the patients were admitted for postoperative surveillance without any organ failure. Four patients in the open group and 9 patients in the laparoscopic group developed multi-organ failure (P=.861). The in-hospital mortality was 22% in the open group and 5% in the laparoscopic group (P=.077).

Table 3.

Results

	Lap-Open (n=18)	Lap-Lap (n=38)	P value^a
Hospital stay (days) following re-intervention^b	24 (12–35)	13 (9–23)	.118^c
Total hospital stay	31 (19–44)	20 (14–28)	.044^c
Intensive care unit admission [n (%)]	13 (72)	22 (58)	.268^d
Total days^b	3 (2–11)	4 (2–5)	.830^c
Severity of organ failure [n (%)]			.861^d
ICU surveillance only	6 (46)	9 (41)
Single organ failure	3 (23)	4 (18)
Multi-organ failure	4 (31)	9 (41)
Patients with additional re-interventions [n (%)]
Surgical	7 (39)	6 (16)	.056^d
Radiological	4 (22)	4 (11)	.416
Fascial dehiscence [n (%)]	4 (22)	1 (3)	.033
Stoma closure [n (%)]	4 (40)	14 (52)	.714^d
In-hospital mortality [n (%)]	4 (22)	2 (5)	.077

Calculated by Fisher's exact test unless indicated otherwise.

Data are median (interquartile range) values.

By Mann–Whitney U nonparametric test.

By chi-squared test.

ICU, intensive care unit; Lap, laparoscopic.

In total, 13 patients underwent one or more additional surgical re-interventions, 39% of the patients in the open and 16% of the patients in the laparoscopic group (P=.056). Of the 7 patients in the laparotomy group with additional reinterventions, 4 developed a fascial dehiscence versus 1 in the laparoscopic group after conversion (P=.033). Four patients in both groups required an additional radiological re-intervention such as abscess drainage (P=.416). The loop ileostomy could be closed in 40% of the patients in the open group versus 52% in the laparoscopic group (P=.714).

Patients required more additional re-interventions in the open group compared with the laparoscopic group. Seven (3 patients had one additional re-invention, and 4 others had two additional re-interventions) in the open group compared with 6 (5 patients had one additional re-intervention, and 1 patient had two re-interventions) in the laparoscopic group.

Discussion

Laparoscopic colorectal surgery is widely practiced, although its use is more common in the elective setting rather than in the emergency setting. Because of improvements in both laparoscopic skills as well as technology, laparoscopy could be a logical next step for the treatment of patients presenting with acute colorectal conditions. Furthermore, the minimally invasive nature of the laparoscopic technique and the accompanying increased conservation of abdominal wall integrity can be beneficial in postoperative recovery and prevent short- and long-term complications such as fascial dehiscence and incisional hernias.

On the other hand, some consider laparoscopy to be contraindicated in the case of peritonitis because of impaired overview, risk of increased bacteremia, and hypercapnia due to pneumoperitoneum. However, these assumptions have never been proven or disproven because of lack of comparative data.^15,16

Most experience with laparoscopic intervention in peritonitis is available for perforated diverticulitis. A systematic review by Toorenvliet et al.¹⁷ showed laparoscopic peritoneal lavage to be successful in over 95% of the patients with Hinchey's III purulent peritonitis. A delayed elective resection was feasible, and even those without a resection had a low percentage of recurrence.

For anastomotic leaks, Soeters et al.¹⁰ suggested conservative management of small anastomotic abscesses with only mild clinical symptoms to be a safe treatment. Small leaks with larger abscesses should be treated with transrectal or radiological drainage, and larger leaks needed surgical treatment with resection of the anastomosis. Only in the case of failure of less than 50% of the circumference and the absence of generalized peritonitis was the anastomosis expected to heal spontaneously if a diverting enterostomy was present and abscesses were adequately drained.¹⁰ These recommendations are not entirely supported by the data presented by Pera et al.⁹ and Fraccalvieri et al.¹¹ Pera et al.⁹ described successful salvage surgery with lavage and a diverting ileostomy in 6 out of 7 patients, including four laparoscopic procedures. Fraccalvieri et al.¹¹ described 39 patients with a salvage procedure and 54 with resection of the anastomosis, with less mortality (15% versus 37%) and better chances for reversal in the salvage group (91% versus 38%). Whether surgery was performed by laparoscopy or laparotomy was not described.

For the choice between laparoscopic and open surgical re-intervention, Wind et al.¹⁴ did provide some evidence that laparoscopic re-intervention may be beneficial for patients with anastomotic leakage following laparoscopic colorectal surgery. It was shown that the morbidity rate, hospital stay, ICU admission, and incisional hernia rate were reduced in the laparoscopic re-intervention group. However, in this study, not all results were consistent, and the study population was small and consisted of young patients with inflammatory bowel disease as the predominant indication for surgery.¹⁴

Rosin et al.¹³ further substantiated this concept. In their study, complications after laparotomy were treated laparoscopically. In 14 cases, 1 patient died, and 1 required a conversion to laparotomy. These data suggest that laparoscopic reintervention is a feasible option even after laparotomy, although data are not compared with open re-intervention.¹³

This study was able to show that laparoscopic re-intervention results in a shorter hospital stay and less fascial dehiscence compared with open re-intervention in selected patients. In addition, there is a trend seen for fewer ICU admissions, fewer additional re-interventions, and less mortality in the laparoscopic group compared with the open group.

There was a shorter surgical time for the laparoscopic group compared with the open group. This is not only a financial benefit, but a shorter procedure is expected to be beneficial for the recovery of these severely ill patients.

But despite these promising results, we should review these results carefully because of some differences in the perioperative findings and treatment. The lack of standardization of choice of procedure, moment of re-intervention, and peroperative findings are important limitations of this study. A more complete clinical score in combination with a peritonitis severity score such as the APACHE score or the Mannheim Peritonitis Index could be able to give a more objective description of severity of leakage. In this retrospective design, unfortunately, there were not sufficient data available to reliably determine these scores for a sufficient number of patients.

In most aspects, no differences were seen in the baseline for both groups. The patients in the laparoscopic group had a higher overall comorbidity rate, and a trend was seen for a higher American Society of Anesthesiologists rate in the laparoscopic group and more malignant disease in the open group. The higher comorbidity rate in the laparoscopic group would be expected to extend the hospital stay and the ICU stay and admission instead of the reduction seen.

In twice as many patients in the open group, fecal contamination was seen during re-intervention. A higher rate of fecal contamination does not necessarily predict a worse outcome compared with generalized purulent peritonitis. Because of the lack of an objective scoring system, fecal contamination is scored both if a closed pocket is opened during the procedure and in the case of generalized fecal peritonitis. Despite a careful review of surgical reports, we could not determine this difference. However, regarding infectious parameters such as C-reactive protein and white blood cell count, no high numbers of patients with generalized fecal peritonitis in the open group are to be expected. Depending on the extent of bowel and anastomotic manipulation during the procedure, a covered leak can stay closed or be revealed without any difference in the prior condition of the patient. Consequently, more leaks will be found, which will result in resection of the anastomosis, whereas if it stays covered, a loop ileostomy could be judged as sufficient. To complete this hypothesis, we expect the manipulation to be more extensive in the open group because the laparoscopic surgeon would be more careful to avoid unintentional bowel injury by manipulation with his instruments compared with manual manipulation during open surgery. This hypothesis could explain the differences in perioperative findings and procedures performed between the two groups, without compromising the results.

Even while less extensive salvage surgery is done in the laparoscopic group with a loop ileostomy or lavage only, and an impaired overview with missed diagnoses is feared, we did not see a higher number of additional re-interventions. In contrast, twice as many additional re-interventions were performed in the open group compared with the laparoscopic group.

Although acknowledging the previous differences and uncertainties, it is difficult to interpret the value of the observed outcomes. The shorter hospital stay and lower mortality will be both an effect of a beneficial minimally invasive technique but to an unknown extent influenced by a selection bias. The observed mortality in both groups is similar to the reported 11% mortality in the Dutch national registry.¹

These data suggest that laparoscopic re-intervention could be a safe and feasible technique in the case of suspected anastomotic leakage, bowel perforation, or abscesses following colorectal surgery in selected patients. These promising results should be investigated in a larger prospective multicenter study with better scoring systems for patient condition and perioperative findings.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

DSCA. Dutch Surgical Colorectal Audit Annual Report. 2010. www.clinicalaudit.nl

Veldkamp

et al. Laparoscopic surgery versus open surgery for colon cancer: Short-term outcomes of a randomised trial. Lancet Oncol, 2005; 6:477–484.

Guillou

et al. Short-term endpoints of conventional versus laparoscopic-assisted surgery in patients with colorectal cancer (MRC CLASICC trial): Multicentre, randomised controlled trial. Lancet, 2005; 365:1718–1726.

Rullier

et al. Risk factors for anastomotic leakage after resection of rectal cancer. Br J Surg, 1998; 85:355–358.

den Dulk

et al. Multicentre analysis of oncological and survival outcomes following anastomotic leakage after rectal cancer surgery. Br J Surg, 2009; 96:1066–1075.

McArdle

, McMillan

, Hole

. Impact of anastomotic leakage on long-term survival of patients undergoing curative resection for colorectal cancer. Br J Surg, 2005; 92:1150–1154.

Ptok

et al. Impact of anastomotic leakage on oncological outcome after rectal cancer resection. Br J Surg, 2007; 94:1548–1554.

Walker

et al. Anastomotic leakage is predictive of diminished survival after potentially curative resection for colorectal cancer. Ann Surg, 2004; 240:255–259.

Pera

et al. The management of leaking rectal anastomoses by minimally invasive techniques. Surg Endosc, 2002; 16:603–606.

10.

Soeters

et al. Colorectal surgery and anastomotic leakage. Dig Surg, 2002; 19:150–155.

11.

Fraccalvieri

et al. Management of colorectal anastomotic leakage: Differences between salvage and anastomotic takedown. Am J Surg, 2012; 204:671–676.

12.

Hedrick

et al.

Anastomotic leak and the loop ileostomy: Friend or foe?

Dis Colon Rectum, 2006; 49:1167–1176.

13.

Rosin

et al. Laparoscopic management of surgical complications after a recent laparotomy. Surg Endosc, 2004; 18:994–996.

14.

Wind

et al. Laparoscopic reintervention for anastomotic leakage after primary laparoscopic colorectal surgery. Br J Surg, 2007; 94:1562–1566.

15.

Navez

et al. Laparoscopic management of acute peritonitis. Br J Surg, 1998; 85:32–36.

16.

Neudecker

et al. The European Association for Endoscopic Surgery clinical practice guideline on the pneumoperitoneum for laparoscopic surgery. Surg Endosc, 2002; 16:1121–1143.

17.

Toorenvliet

et al. Laparoscopic peritoneal lavage for perforated colonic diverticulitis: A systematic review. Colorectal Dis, 2010; 12:862–867.