Acute colonic diverticulitis: Surgical management > 의학

INTRODUCTIONDiverticular disease of the colon is an important cause of hospital admissions and a significant contributor to health care costs in industrialized nations [1,2]. In Western countries, the majority of patients present with sigmoid diverticulitis [3,4].

Most patients with acute sigmoid diverticulitis are treated medically; surgery is only indicated when diverticulitis is either not amenable or refractory to medical therapy (algorithm 1) [5-8]. Approximately 15 percent of patients will require surgery for diverticular disease [7]. In the United States, diverticular disease is the leading indication for elective colon surgery [9]. For patients who require surgery for diverticulitis, the choice of techniques depends upon the patient's hemodynamic stability, extent of peritoneal contamination, and surgeon experience/preference [10].

Surgical treatment of acute colonic diverticulitis and its acute complications (perforation, abscess formation, or intractability) are described here. Diverticular fistulas, bleeding, and stricture/obstruction, which are typically subacute or chronic sequelae of diverticulitis, are discussed in other topics:

●(See "Diverticular fistulas".)

●(See "Colonic diverticular bleeding".)

●(See "Overview of mechanical colorectal obstruction".)

The diagnosis and medical management of acute diverticulitis are discussed separately. (See "Clinical manifestations and diagnosis of acute diverticulitis in adults" and "Acute colonic diverticulitis: Medical management".)

PERFORATIONWhile most patients with freely perforated diverticulitis require surgery, the choice of techniques largely depends upon the extent of peritoneal contamination as assessed by the Hinchey classification system [11]:

●Stage I – Pericolic or mesenteric abscess

●Stage II – Walled-off pelvic abscess

●Stage III – Generalized purulent peritonitis

●Stage IV – Generalized feculent peritonitis

In general, the majority of microperforations (not included in the Hinchey classification), Hinchey I, and Hinchey II perforations can be managed nonoperatively, while most Hinchey III and IV perforations require surgical intervention.

Free (frank) perforation — Acute diverticulitis with free (frank) perforation (Hinchey III or IV) is a life-threatening condition that mandates emergency surgery [7,8,12-14]. The primary goal of surgery is to obtain source control by removing the perforated colonic segment; the secondary goal of surgery is to restore intestinal continuity [15-19], the feasibility of which is predicated upon a patient's hemodynamic stability and the degree of peritoneal contamination as detailed below.

Unstable patients — For patients who are hemodynamically unstable due to perforated diverticulitis and who do not have the physiologic reserve to tolerate a colon resection and reconstruction, damage control surgery should be expedited to obtain source control of their sepsis, delaying less critical portions of the operation until after appropriate resuscitation [14,20,21]. (See "Overview of damage control surgery and resuscitation in patients sustaining severe injury".)

While damage control surgery always consists of peritoneal lavage, temporary abdominal closure, and a second look, some surgeons also perform a limited colonic resection, then either leave the colon stapled off in situ or construct a colostomy (ie, Hartmann’s procedure with temporary abdominal closure) [20]. In a study of 58 patients with perforated diverticulitis with generalized peritonitis, the use of the damage control strategies resulted in a 9 percent mortality rate, while 44 of 53 surviving patients were stoma-free [22].

Stable patients with feculent peritonitis — For stable patients with feculent peritonitis (ie, Hinchey IV diverticulitis), we suggest Hartmann's procedure [14]. Primary anastomosis is generally contraindicated because of the peritoneal contamination and inflammation.

Hartmann's procedure involves resecting the diseased colonic segment, creating an end colostomy and a rectal stump, and then reversing the colostomy in the future (figure 1) [16]. Because creating a mucous fistula by bringing the distal end of the transected bowel through the abdominal wall is often not possible after resecting the entire sigmoid colon, many surgeons mark the rectal stump with a long nonabsorbable suture and tack it to the anterior abdominal wall or sacral promontory to help identify the rectal stump at the second-stage operation.

Subsequent closure of the colostomy is a technically difficult operation associated with high morbidity and mortality rates [23,24]. As a result, colostomy closure is only performed in approximately 50 to 60 percent of all patients after a Hartmann's procedure [25-27]. In the author's practice, patients with fecal contamination of the abdominal cavity at the index operation are usually reversed after approximately one year; those without fecal contamination can usually be reversed sooner in three to four months. Colostomy reversal should be approached cautiously for obese patients with a short rectal remnant because the operation is technically difficult and the functional outcomes (eg, bowel control) are typically poor.

Stable patients with purulent peritonitis — For stable patients with purulent peritonitis (ie, Hinchey III diverticulitis), the Hartmann's procedure is also the most commonly performed procedure [28]. However, the surgeon may choose to restore bowel continuity with or without fecal diversion based on patient and intraoperative factors as well as his or her own experience [10].

Primary anastomosis with proximal diversion — Some European authors have advocated primary anastomosis with proximal diversion (PAPD) (figure 2) in select patients with Hinchey III perforated diverticulitis [14]. Limited data suggest that it may have similar mortality, lower morbidity, and lower stoma rate at 12 months compared with Hartmann's procedure.

●In a randomized trial of 62 patients with left-sided colonic perforation due to Hinchey III or IV diverticulitis, patients treated with a primary anastomosis with diverting ileostomy, compared with patients treated with a Hartmann's procedure, had similar mortality (9 versus 13 percent) and morbidity rates (75 versus 67 percent) after the first operation [29]. However, a greater percentage of patients treated with a primary anastomosis with diverting ileostomy underwent stoma reversal (90 versus 57 percent), and reversal of the diverting ileostomy in those patients required less operative time (73 versus 183 minutes), length of hospital stay (six versus nine days), and resulted in fewer serious complications (0 versus 20 percent), compared with colostomy reversal in patients treated with a Hartmann's procedure.

●In another French trial (DIVERTI), 102 patients with Hinchey III or IV perforated diverticulitis were randomly assigned to primary anastomosis or Hartmann's procedure [30]. At 18 months, the mortality and morbidity rates were similar between the two groups, but patients who underwent primary anastomosis were much less likely to still have a stoma (4 versus 35 percent). Two-thirds of patients underwent primary anastomosis with a protective stoma and one-third without it. That decision was not randomized but made by surgeons intraoperatively. Although the morbidity rate was lower without a protective stoma, this result was likely biased, as all but one patient without a stoma had Hinchey III disease.

●A third multinational trial comparing primary anastomosis (without diversion) with Hartmann's procedure in patients with Hinchey III or IV perforated diverticulitis failed to reach any conclusion after closing prematurely due to poor accrual [31].

A 2018 systematic review and meta-analysis of randomized trials (including the three above) found that primary resection and anastomosis had similar major complication and mortality rates compared with Hartmann's procedure. However, patients were more likely to be stoma-free (relative risk [RR] 1.4, 95% CI 1.18-1.67) and to avoid major complications related to the stoma reversal procedure (RR 0.26, 95% CI 0.07-0.89) after primary resection and anastomosis than after Hartmann's procedure [32]. A 2019 systematic review and meta-analysis of 22 observational studies and the three trials reached a similar conclusion that both procedures were acceptable [33].

The two procedures have been compared in several large administrative database studies as well. In an analysis of over 130,000 patients undergoing either primary anastomosis with diverting loop ileostomy or Hartmann's procedure for acute diverticulitis (not stratified by Hinchey classification, but about 90 percent underwent surgery for perforation), the mortality (2.9 versus 7.6 percent) and morbidity rates (49 versus 55 percent) were comparable [34]. Although patients who underwent Hartmann's procedure were generally sicker (eg, more likely to be in septic shock [11 versus 5 percent]), primary anastomosis with diverting loop ileostomy was not associated with increased mortality or morbidity when compared with Hartmann's procedure in multivariate analysis. The majority of patients underwent Hartmann's procedure, and only 7.6 percent underwent primary anastomosis with diverting ileostomy.

However, an administrative database study of over 10,000 patients undergoing urgent or emergency colon surgery for diverticulitis found that both postoperative mortality (15 versus 7.4 percent) and morbidity rates (58.2 versus 39.5 percent) were significantly higher when noncolorectal surgeons performed primary anastomosis with proximal diversion compared with Hartmann's procedure, while the postoperative mortality (3.7 versus 5.3 percent) and morbidity rates (48.2 versus 43.4 percent) were comparable between the two procedures when performed by colorectal surgeons [35,36].

Rarely performed procedures — Drainage procedures are rarely performed because they do not definitively address the underlying diverticular disease, and some studies suggest poorer outcomes in patients undergoing lavage. However, they may be useful in treating septic patients who are too ill to tolerate a resectional procedure.

Laparoscopic lavage — In the 1990s, laparoscopic lavage and drainage were introduced to avoid laparotomy and fecal diversion in patients with complicated diverticulitis [37-40]. Compared with other surgical options, laparoscopic lavage has been shown to decrease stoma rate within 90 days (RR 0.18; 95% CI 0.12-0.27); however, it does not decrease one-year mortality rate and actually increases short-term morbidity rate due to higher complication and reintervention rates [14,41]. Therefore, laparoscopic lavage should only be performed in selected patients with Hinchey III (purulent) perforated diverticulitis and by experienced surgeons, after they are counseled regarding the higher complication and reintervention rates. Hinchey IV (feculent) perforated diverticulitis should be treated with Hartmann’s procedure, while Hinchey I or II (abscess) diverticulitis is likely to respond to nonoperative management. (See 'Localized perforation (ie, abscesses)' below.)

Although earlier retrospective studies found a low mortality rate of 2 percent and avoidance of a permanent stoma in the majority of patients who underwent laparoscopic lavage [42], subsequent randomized trials performed against resectional procedures in the 2000s reported conflicting results:

●In the SCANDIV trial, 199 patients suspected of having perforated diverticulitis based upon detection of free air by abdominal computed tomography (CT) scan were randomly assigned to undergo emergency laparoscopic lavage or sigmoidectomy [43]. Compared with sigmoidectomy, laparoscopic lavage achieved similar mortality (13.9 versus 11.5 percent) and severe morbidity rates (30.7 versus 26 percent) at 90 days. However, patients who were treated with laparoscopic lavage were more likely to require reoperation (20.3 versus 5.7 percent) for complications such as secondary peritonitis (six versus zero patients) or missed sigmoid cancer (four versus zero patients).

●The LOLA trial, which included 90 patients with purulent perforated diverticulitis, showed that laparoscopic lavage produced a higher combined major morbidity and mortality rate within 30 days compared with sigmoidectomy (39 versus 19 percent) [44]. At 12 months, the rates were comparable between the two groups (65 percent for lavage versus 63 percent for sigmoidectomy).

●The DILALA trial randomly assigned 83 patients to laparoscopic lavage or Hartmann's procedure after a laparoscopic diagnosis of purulent perforated diverticulitis [45,46]. The mortality rates were similar at both 90 days (8 versus 11 percent) and one year (14 versus 15 percent); the major morbidity rates were similar at 30 (13 versus 18 percent) and 90 days (21 versus 25 percent). The reoperation rates were similar at 30 days (13 versus 17 percent). At one year, however, fewer patients required reoperation after laparoscopic lavage (28 versus 63 percent). In addition, laparoscopic lavage resulted in shorter operative time (1 versus 2.5 hours) and hospital stay (6 versus 9 days for index admission; 8 versus 14 days at one year). At two years, still fewer patients required one or more reoperations after laparoscopic lavage than after Hartmann's procedure (42 versus 68 percent), while the mortality rate (14 versus 18 percent) and total days of hospital stay (18 versus 24) were statistically similar between the two groups [47].

Although laparoscopic lavage has purported advantages of reduced morbidity and mortality, a systematic review and meta-analysis of randomized trials (including the three above) found that major complications were more frequent after laparoscopic lavage than sigmoidectomy, whereas postoperative mortality rates (RR 1.03; CI 0.45-2.34) were not different between the two procedures [32,41]. Weighing the lower stoma rate against the higher risks of complication and reoperation, we do not recommend routine laparoscopic lavage for Hinchey III or IV perforated diverticulitis. Hartmann's procedure remains our standard treatment for Hinchey III or IV perforated diverticulitis.

Laparoscopic lavage was originally intended for purulent (Hinchey III) but not feculent (Hinchey IV) perforated diverticulitis. But in common practice, it is often difficult to exclude fecal peritonitis and/or sigmoid carcinoma during the preoperative evaluation. If a surgeon chooses to perform laparoscopic lavage, he/she must exclude fecal perforation (Hinchey IV) or a visualized perforation by diagnostic laparoscopy, and either colon cancer or ongoing colonic air leak (from perforation) by intraoperative sigmoidoscopy.

Patients who undergo laparoscopic lavage should be advised that a reoperation (usually Hartmann's procedure) may be necessary if laparoscopic lavage fails to control the sepsis or a sigmoid carcinoma is later found. In a multicenter, prospective, noncomparative study (the LLO study), 212 patients underwent laparoscopic lavage for laparoscopy-confirmed Hinchey III perforated diverticulitis [48]. The short- and long-term success rates of laparoscopic lavage in this cohort of patients were 74.5 percent (discharged without further surgery or readmission in 60 days) and 65.4 percent (no surgery for the initial admission or recurrence), respectively. Only one patient had a colon cancer, which required reoperation. Since one-quarter of patients will not have sepsis control with laparoscopic lavage alone, those who undergo laparoscopic lavage must be able to tolerate persistent or recurrent infection. Thus, frail, septic patients or those with major comorbidities are not good candidates for laparoscopic lavage [40]. A fit patient with previous acute diverticulitis and without severe sepsis may be a better candidate for the procedure [48].

Three-stage procedure — A legacy procedure for colonic perforation includes three stages: the first stage of draining but not resecting the diseased segment and construction of a proximal diverting stoma, the second stage of resecting the diseased segment with a primary anastomosis under the protection of the proximal stoma, and the third stage of closing the proximal stoma. Until the 1980s, the three-stage procedure was felt to be the safest approach to perforated diverticulitis.

Since then, the three-stage procedure has been largely replaced by other procedures (eg, the Hartmann's procedure) with lower postoperative mortality rates (26 versus 7 percent) [49,50]. In contemporary practice, the three-stage procedure is only performed when inflammation precludes safe pelvic dissection of the colon from critical sidewall structures (eg, iliac vessels and ureters), or when the patient is unstable. Drainage and fecal diversion in these situations can serve as a temporizing measure to allow treatment of infection and inflammation before further surgery or transfer to a more experienced center. (See "Overview of mechanical colorectal obstruction", section on 'Three-stage'.)

Localized perforation (ie, abscesses) — Localized perforations present acutely as a mesocolic or pelvic abscess (Hinchey I or II). Hinchey I or II diverticulitis is characterized by one or more localized abscesses in the pericolonic, mesenteric, or pelvic locations. In contemporary practice, diverticular abscesses are typically treated with percutaneous image-guided drainage or with intravenous antibiotics if the abscess is too small (<4 cm) or inaccessible to percutaneous drainage. Surgery may be indicated for patients who deteriorate or fail to improve within two to three days of percutaneous intervention or antibiotic therapy as a persistent intra-abdominal abscess is unlikely to respond to further nonoperative management. (See "Acute colonic diverticulitis: Medical management", section on 'Abscess'.)

Patients with a localized perforation can usually tolerate a preoperative bowel preparation. Thus, if the phlegmon or abscess can be resected with the colonic segment, a primary anastomosis can be performed in these patients. (See 'Colon resection with primary anastomosis' below.)

If there are concerns about either contamination or inflammation involving the surrounding tissue (eg, with a large pelvic abscess) but the bowel is not edematous, a primary anastomosis with or without a protective ostomy, depending upon the condition of the local tissue, can be performed. This is preferred to a Hartmann's procedure as a protective stoma is easier to reverse than an end colostomy with a rectal stump [51,52]. (See 'Primary anastomosis with proximal diversion' above.)

Microperforation — Microperforation, usually indicated by one or a few extraluminal air bubbles on CT images, and phlegmon are not considered complicated diverticulitis and thus can be managed nonoperatively with intravenous antibiotics and bowel rest [10]. (See "Acute colonic diverticulitis: Medical management", section on 'Microperforation'.)

Those with extraluminal air bubbles only on initial CT may develop an abscess on subsequent studies. In one study, the rate was 19 percent [53]. Patients with microperforation and an associated abscess should be treated accordingly as having complicated disease. (See "Acute colonic diverticulitis: Medical management", section on 'Abscess'.)