Laparoscopic surgery for failed medical treatment: Diverticulitis complicated by pericolic and/or pelvis abscess (Hinchey 1 and 2) are most often managed conservatively with intravenous antibiotics more or less combined to percutaneous drainage[1,38,39]. In case of failure, an emergent surgery may be indicated for sepsis control and resection of the diseased colon. Available data are, however, limited to a single-institutional retrospective series[40] which has been recently updated[41]. When comparing outcomes of the laparoscopic approach (24 patients) to those of laparotomy (18 patients), laparoscopy was associated with a 4-fold decrease in overall complication rate, faster bowel movement, shorter hospital stay and comparable rate of RPA despite a longer operative time. In the laproscopy group, RPA could be achieved in all but one patient and conversion to open surgery occurred in 2 patients. In regard to these encouraging results and motivated by the scarcity of data on this specific aspect of colonic diverticular surgery, RCTs are needed to confirm the suggested benefits of laparoscopy in this group of patients.

LLD: First described in 1996[42,43], the procedure consists of a laparoscopic exploration of the abdominal cavity followed by lavage with heated saline serum and drainage of the diseased colon for the next several days. The need for extensive adhesiolysis in search of the colonic perforation remains controversial[44]. In a prospective multi-institutional study[45], 92 patients underwent laparoscopic peritoneal lavage for Hinchey 2 or 3 diverticulitis. All patients but two had complete clinical resolution. One patient required subsequent colonic resection whereas the other required percutaneous drainage for pelvic abscess. Mortality and morbidity rates were 3% and 4%, respectively. After a median follow-up period of 36 mo, only two patients were readmitted with acute diverticulitis successfully treated with antibiotics. The authors concluded that LLD can be a reasonable alternative to HP for Hinchey 3 perforated diverticulitis. Another prospective study[46] found that LLD offers the advantages of shorter operative time and hospital stay, reduced estimated blood loss and better postoperative outcomes compared with laparoscopic HP. Including all stages of complicated diverticulitis (mainly Hinchey 3), only 2.1% of patients undergoing LLD were converted to open HP, and 6.4% were reoperated for uncontrolled sepsis. Almost half of the patients who received LLD underwent subsequent sigmoid resection, but the reasons were not mentioned. Such favorable results (low morbidity and mortality rates, high chance of sepsis control, avoidance of stoma and bowel salvage) have been constantly reported in case series and systematic reviews[37,44,47-53]. Soon after and despite robust evidence, LLD was cited in national and international scientific committees’ reports as a potential therapeutic option for purulent peritonitis of diverticular origin[1,54,55].

Preliminary results of the DILALA prospective randomized trial have been recently published after complete accrual[56]. Although the primary endpoint (the number of reinterventions) could not be assessed because of incomplete follow-up, this trial showed a higher than usual mortality rate of 7.7% after LLD. This result was comparable to that found in the open HP group. Both procedures provided comparable complication profiles, but LLD resulted in shorter operative time, shorter time in the recovery unit and shorter hospital stay. Recently, the enthusiasm toward LLD has been hindered by the results of the LOLA group within the Ladies trial confronting LLD to sigmoidectomy in Hinchey 3 diverticulitis[57]. This multicenter RCT has been prematurely terminated because of significantly increased in-hospital major morbidity or mortality in the lavage group compared with sigmoidectomy. Surgical reinterventions accounted for most of these adverse events. However, mortality rate was unaltered (9% in the LLD group vs 14% in the sigmoidectomy group), and sepsis was successfully controlled in 76% of cases (vs 90% in the sigmoidectomy group). In the long-term, there was no difference in the incidence of major morbidity or death between the two groups (37% vs 40% in the LLD and sigmoidectomy groups, respectively). LLD allowed salvage of the sigmoid in almost half of the cases and avoided stoma formation in three-quarters of the patients. In contrast to the resection procedure, LLD allows bowel salvage and avoids stoma but requires a multi-step careful surveillance and timely management. The authors conclude that peritoneal lavage cannot be favored over sigmoid resection as a routine intervention for purulent peritonitis of diverticular origin. Instead, it may be an alternative approach to sigmoidectomy with similar length of stay and long-term outcome in select patients. From this standpoint, in the Dutch Collaborative study group[58] as well as in another study[59], age older than 80 years, American Society of Anesthesiologists grade 3 or above, multiple comorbidities, immunosuppression, high C reactive protein and/or high Mannheim peritonitis index were associated with increased risk of failure after LDD. Currently, there are two ongoing RCTs comparing LLD to sigmoidectomy[60,61]. The results are eagerly awaited to provide further evidence about expected outcomes after LLD and to identify criteria for patients who would preferably benefit from each technique.

In contrast to the enthusiasm for LLD in purulent peritonitis, this approach is not commonly admitted for Hinchey 4 diverticulitis. The presence of a visible perforation in the colon has always represented the limit of application of the technique[42,62]. In his prospective study, Myers et al[45] stated that stercoral peritonitis constitutes an indication for conversion to open HP. Conversely, Liang et al[46] reported very encouraging results despite the inclusion of more than 10% of Hinchey 4 peritonitis in his prospective study. Many authors agree that LLD cannot be accepted unless perforation of the colon is formally ruled out[45,53]. Conversely, others propose suturing the colonic hole if stercoral peritonitis is not evident[59]. The diagnosis of stercoral peritonitis on explorative laparoscopy excluded patients from both of the available randomized controlled trials[56,57]. This condition is presumed to significantly modify the results observed with purulent peritonitis. In fact, the high reintervention rate found in the Ladies trial is mostly attributed to the misdiagnosis of stercoral peritonitis. To optimize outcomes after peritoneal lavage, the authors pushed toward a meticulous search for colon perforation. In all cases, the adoption of LLD in emergency settings, abdominal exploration for generalized peritonitis, pelvic dissection in inflammatory conditions and possible suture of a diseased colon require that the surgeon have a minimum of colorectal and minimally invasive skills before he can propose this conservative approach[51,58].

Sigmoidectomy: Primary anastomosis and HP: Supported by considerable improvement in the perioperative care, RPA (without or without DI) has been proposed as an alternative to HP in emergent situations. A comparison of these two techniques has mostly enrolled patients undergoing open procedures before the widespread application of laparoscopy in emergent colorectal surgery. In fact, several comparative studies, systematic reviews and meta-analyses favored RPA over HP in respect to reduced mortality and morbidity rates, shorter cumulative operative time and hospital stay, more frequent stoma reversal and reduced cost[63-76]. Even without DI, RPA was shown to be preferable than HP for purulent peritonitis[70,72,77]. These studies, however, suffer from marked heterogeneity and selection bias with low-risk patients mainly undergoing RPA, whereas HP is offered to high-risk elderly patients[63,74]. A recent RCT showed that for Hinchey 3 and 4 diverticular disease, the main differences between RPA with DI and HP occur during the stage of stoma reversal[78]. When both stages (colonic resection and stoma reversal) were combined, the rates of overall complications, severe complications and mortality (13% in HP vs 9% in PA) were similar in both groups. In contrast, when the reversal procedure was considered alone, HP was associated with lower stoma reversal rate (58% vs 90%), more frequent severe complications (20% vs 0%), and longer operative time and hospital stay. The main flaw of this RCT would be the lack of information about the adopted approach (laparoscopic or open) for the reversal of the stoma. This issue is of paramount importance because laparoscopy has been proved to decrease the morbidity of HP reversal in several case series and systematic reviews[79-86]. The elevated rate of severe complications during stoma closure in the HP group may have been overemphasized by the open approach per se.

This evidence raises the question whether the laparoscopic HP may offer advantages over the open approach in terms of reduced morbidity and mortality in the acute setting of perforated diverticulitis. To improve the outcome after open HP, we have been among the first to show the feasibility of a laparoscopic two-staged strategy for complicated diverticular disease. Despite conversion in 19% of cases, this approach offered adequate control of sepsis with low rates of mortality (3%) and morbidity (23%)[33]. Similarly, in a small case series, Agaba et al[34] described favorable outcomes after laparoscopic HP for Hinchey 3 and 4 diverticulitis. Recently, a propensity-matched analysis of the ACS NSQIP database failed to show a decrease in postoperative morbidity and mortality when HP was performed under laparoscopy compared with the open approach. This study, however, suffers from substantial imperfections in methodology such as retrospective data acquisition and lack of analysis of pertinent variables that might substantially interfere with the results[26].

To clarify the role of laparoscopy in emergent sigmoidectomy, Mbadiwe et al[87] retrospectively analyzed the ACS NSQIP database. In a total of 11981 patients, a bivariate analysis showed that patients undergoing laparoscopy experienced lower rates of complications with both RPA (14% vs 26%, P < 0.001) and HP (30% vs 37%, P = 0.02). The laparoscopic approach was associated with decreased mortality rate for patients undergoing RPA (0.24% vs 0.79%, P < 0.001). The reduced complication rate after laparoscopic RPA was confirmed in the multivariate analysis[87].

To provide a high level of evidence in the present era of widespread use of laparoscopy for colorectal disease, a well-built RCT is highly desirable comparing laparoscopic RPA (with or without DI) with a two-step laparoscopic HP (sigmoid resection and stoma closure) in perforated complicated diverticulitis. Not only outcomes of both procedures will be clarified but also identification of precise criteria would define the subgroups of patients who will benefit more from each technique. Based on this perspective, the results of the ongoing DIVA section of the Ladies trial are keenly awaited to provide us with level 1 evidence about the preferable laparoscopic attitude in Hinchey 4 complicated diverticulitis[57].

Based on solid proofs from a large-scale meta-analysis[88] and RCTs[89,90], international committees have adopted laparoscopy as the preferred approach for elective sigmoidectomy after acute diverticulitis[1,23,24]. Compared with an open procedure, the laparoscopic modality offers a significant decrease in major complications and morbidity, less blood loss, fewer analgesic requirements, shorter hospital stay and improved quality of life.

Conversely, in early experience, laparoscopy was contraindicated for the treatment of diverticular chronic complications (stricture, fistula and persistent phlegmon) because severe inflammation and distorted anatomy exposes the patient to high risks of bleeding and adjacent organ trauma (bladder, left ureter, female genital organs)[11,91]. With increasing experience, the laparoscopic approach has been progressively accepted as an alternative to open surgery but its routine usage in chronic complications remains controversial[92].

Because presently available data mostly rely on small retrospective series and case reports, several controversies cannot be fully elucidated. In a descriptive case series, Le Moine et al[93] suggested that chronic complicated diverticulitis increases the risk of conversion to laparotomy in elective laparoscopic sigmoidectomy. This effect has been recently disproved by two comparative studies[94,95] which stated that surgeons’ expertise in minimally invasive and colorectal surgery is the principal determinant of morbidity and conversion rates in complicated cases. This evidence outlines that accumulating experience during the last ten years has inevitably challenged previously accepted knowledge.

Colonic fistula is the most reported late complications of diverticular disease, although available data is scarce and confined to retrospective case series. In recent systematic reviews[96,97], the laparoscopic approach was judged to be feasible and safe for the treatment of colovesical fistulae. However, due to a lack in methodology and/or a limited number of patients, the studies failed to show superiority of laparoscopy over the open approach. Furthermore, the conversion rate could not be determined and the predictive factors of its occurrence were not discussed[96]. This issue has been addressed in a previous case series including 31 patients conducted over 10 years. The overall conversion rate was approximately 30% but declined to 10% during the second half of the study period[98]. This finding is consistent with the statement of Abbass et al[99] that similar rates of conversion, morbidity and mortality can be expected with laparoscopy regardless of the presence of colonic fistula. Therefore, the current surgical management of colonic fistula relies on poor data derived from empirical experiences in open surgery. The most salient example is the adoption of a single-stage procedure as the preferred surgical option despite the absence of solid proofs confirming its superiority[98].

During this study, we noticed a striking lack of high quality papers dedicated to the issues of inflammatory phlegmon and stricture. Information on the most appropriate treatment in these situations is thus absent. If low incidence of such cases is the cause of this shortage, prospective enrollment of patients in large multi-institutional databases might be the solution to build precise therapeutic algorithms on which tomorrow’s general surgeons can use for the treatment of their patients.

Finally, with the emergence of LLD as a conservative modality for complicated perforated diverticulitis, controversies have arisen about the need for systematic elective sigmoidectomy after the diseased organ has been saved. In a recent systematic review[44], rehospitalization was observed in 7% of patients who underwent LLD. More than half of them presented with a new episode of diverticulitis, 21% with generalized peritonitis, 10% with colovesical fistula and 7% with undetected colon cancer. At this stage of experience, the present data lack sufficient proof to know whether LLD can be regarded as a definitive treatment or a bridge to elective laparoscopic sigmoidectomy[100]. If LLD becomes widely admitted, well-built RCTs will be necessary to answer these questions and to determine the best candidates for bowel sparing.