Damage control surgery for the treatment of perforated acute colonic diverticulitis

Supplemental Digital Content is available in the text


Introduction
Acute colonic diverticulitis (ACD) is defined as an acute inflammation of one or more colonic diverticula. [1,2] Approximately 10% to 25% patients affected by colonic diverticulosis are going to develop ACD in their lifetime. [1][2][3] ACD complications arise in approximately 8% to 35% patients and the most common ones are represented by phlegmon or abscess (about 70% complications), followed by perforation, peritonitis, obstruction, and fistula. [1,4,5] Peridiverticular and pericolic infections stem from a microscopic or macroscopic perforation of one or more inflamed diverticula. [1,5] In accordance with current guidelines, patients affected by generalized peritonitis should undergo emergency surgery. [2] However, decisions on whether and when to operate ACD patients remain a substantially debated topic while algorithm for the best treatment has not yet been determined. [1,2] To date, no single treatment strategy has turned out as best method, in terms of efficacy and safety. [1,2] Krukowski et al and Vermeulen et al suggested a classification of surgical procedures to be performed in perforated ACD (Table 1), while neither the most recent laparoscopic lavage nor the more recent and less widespread damage control surgery (DCS) were mentioned. [2] DCS represents a well-established method in treating critically ill patients with traumatic abdomen injuries. [6] DCS strategy includes abbreviated source-control laparotomy followed by intensive care unit (ICU) transfer for physiology resuscitation and delayed surgery for definitive management. [6] At present, such surgical approach is also finding application in non-traumatic emergencies such as perforated ACD. [7] Thanks to a thorough systematic review of the literature, we aimed at achieving deeper knowledge of both indications and short-and long-term outcomes related to DCS in perforated ACD.

Methods
The protocol for this systematic review was registered on PROSPERO (CRD42020186958) and is available in full on the NIHR HTA programme website (https://www.crd.york.ac.uk/ prospero/display_record.php?RecordID=186958).

Search strategy
We carried out a systematic literature review, according to Preferred Reporting Items for Systematic Reviews and Meta-Analyzes (PRISMA) guidelines. [8] According to the gold standard for literature search for surgical reviews, [9]

Inclusion criteria
Only English-written scientific papers were selected, including case reports, case series, case-control studies, cohort studies, controlled clinical trials, and randomized clinical trials. Prior systematic reviews and meta-analyses were ruled out. We considered both comparative and non-comparative studies including adult patients (over 18 years of age) treated for peritonitis by perforated ACD through DCS strategy as defined in "Damage control surgery procedures" paragraph. Given the lack of scientific studies on this topic, all articles of qualitative interest have been selected despite population size, publication status, and lack of interesting parameters in some of them. In addition, references of relevant articles (previously published reviews, systematic reviews or meta-analyses, and the articles included in the qualitative analysis) were searched through, in order to identify further cases of interest. Table 1 Operative procedures [2] .

Quality assessment
The Newcastle-Ottawa quality assessment scale (NOS) was used to assess the quality of each study. Thresholds for converting the Newcastle-Ottawa scales to AHRQ standards (good, fair, and poor): i) good quality: 3 or 4 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/ exposure domain, ii) fair quality: 2 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain; iii) poor quality: 0 or 1 star in selection domain OR 0 stars in comparability domain OR 0 or 1 stars in outcome/exposure domain.

Search results and study characteristics
Final literature search, performed in March 2020, identified 108 potential items of interest (Fig. 1). After removing duplicate publications (42), 66 records were further analyzed. Twenty-four out of which were excluded as not relevant, while 42 full-text articles were assessed for eligibility. After removing full-text articles not complying with inclusion criteria (34), 8 articles were included into qualitative synthesis. [11][12][13][14][15][16][17][18] No item was included on the basis of other sources (e.g., references lists). The included articles were single-center retrospective studies (3), multicenter retrospective studies (4), and single-center prospective studies (1).

Discussion
Diverticular perforation is an extremely important occurrence in ACD natural history. Mortality following complicated ACD (abscess, perforation, or fistula) has increased, if compared to mortality in patients affected by uncomplicated ACD. [1] It records the highest rate among patients with perforation or abscess. [1] A UK cohort study reported a 20% 1-year mortality rate for patients with perforated ACD, against 4% controls matched by age and gender. [1] After first diverticulitis acute attack, 20% to 30% patients go to surgery, being about half of them performed at emergency. [1][2][3] Fifteen to 40% out of these cases involve people younger than 50. [1][2][3] To date, HP is the most performed method in Hinchey III and IV patients. [2] Despite being a relatively simple and ideally safe surgical procedure and given absence of intestinal anastomosis, its morbidity and mortality are not negligible. Keep in mind that Hartmann's reversal is typified by a 49% to 55% morbidity and 20% mortality rates. [2] In addition, a large amount of patients will never undergo stoma reversal (48-74%), although patients affected by diverticular disease show high stoma reversal rates (83%). [2] LADIES, a multicenter, parallel, randomized, open-label superiority trial identified a 12-month stoma-free rate of 94.6% and 71.4% (Hinchey III: PADS 95.3% vs HP 79.8%; Hinchey IV: PADS 92.2% vs HP 51.9%) with a median interval of reversal of 101 days and 186 days for PADS and HP, respectively. [4] In intention-to-treat analysis, no statistically significant discrepancy was identified between HP and PADS, as concerned perioperative mortality (3% vs 6%) and overall morbidity (HP 44% vs PADS 39% -Hinchey III: HP 37% vs PADS 37%; Hinchey IV: HP 60% vs 44%). [4] In stoma reversal analysis, 68% HP patients and 83% PADS patients underwent stoma reversal with a median interval of reversal of 133 days and 113.5 days, for their respective groups. [4] Overall morbidity recorded a statistically significant discrepancy between HP patients and PADS ones (30% vs 8%). [4] DIVERTI, a multicenter, prospective, randomized controlled trial reported no statistically significant difference between HP and PADS, in terms of mortality and overall morbidity (42.3% vs 54%) at emergency surgery analysis. [19] In stoma reversal analysis, discrepancy among HP patients and PADS ones turned out statistically significant (64.6% vs 96%), while overall morbidity recorded no statistically significant difference (21.2% vs 12.5%). [19] Above mentioned findings are in accordance with those gathered by recent meta-analyzes. [19][20][21][22][23] In general, overall postoperative morbidity, mortality and stoma-free survival rates following HP were equivalent or inferior to those following PADS. [1,[20][21][22][23][24][25] According to recommendation 19 of 2016 World Society of Emergency Surgery (WSES) Guidelines, Hartmann resection is still advised for managing diffuse peritonitis in critically ill patients and in patients with multiple comorbidities. However in clinically stable patients with no co-morbidities primary resection with anastomosis with or without a diverting stoma may be performed (Recommendation 1 B). [26] This instruction is shared by many guidelines, although WSES Guidelines offer an additional choice in treating critically ill patients. Indeed, Recommendation 21 stated: Damage control surgery strategy may be suggested for clinically unstable patients Table 2 Demographic and clinical data of reported cases/series of DCS for perforated acute colonic diverticulitis.  Table 3 Intraoperative and perioperative outcomes data of reported cases/series of DCS for perforated acute colonic diverticulitis. Sohn et al/2018 [14] None 74 (100) None 37 (50) 25 (34)  Sohn et al/2018 [15] None 58 (100) None 34 (59) 14 (24)  with diverticular peritonitis (severe sepsis/septic shock) (Recommendation 1 B). [26] Nevertheless, no general agreement has yet been reached on DCS in perforated ACD. [7,27] According to our analysis, patients treated with DCS showed a 23% to 74% overall morbidity rate, a 0% to 20% 30-day mortality rate and 7% to 33% follow-up overall mortality rate, in addition to a 0% to 33% definitive stoma rate. [11][12][13][14][15][16][17][18] The latter result would suggest a potential advantage of DCS over HP.
However, our findings must be carefully taken into account. Indeed, retrospective quality of most analyzed studies and population heterogeneity become clear in patient selection criteria. Although Hinchey III/IV represented inclusion criteria in all studies, clinical presentation was openly sepsis/septic shock or organ failure in less than a quarter individual populations (see Table 2, Clinical presentation), with the only exception of Tartaglia et al, who just enrolled patients with septic shock/organ failure. [11][12][13][14][15][16][17][18] Differences in sepsis/septic shock terminology need to be taken into account. Two out of 8 studies adopted The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). [17,18,28] Three studies followed German Sepsis Association S-2k guidelines based on the definitions of American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference. [13][14][15]29] The remaining 3 studies did not report what definitions they had adopted. [11,12,16] Application of DCS principles is based on clinical assessment of a patient with trauma, who is physiologically decompensated, as it is determined by the phase "lethal triad" in hemorrhagic shock: acidosis, coagulopathy, and hypothermia. [27,30] Patients with decompensated trauma should be treated immediately, in order to avoid progression to irreversible physiological exhaustion and death. [27,30] In this case, abbreviated operations allow stabilization, correction, and re-evaluation of physiological imbalances at ICU. [27,30] Likewise, patients undergoing general emergency surgery might experience a decompensated, almost irreversible, physiological exhaustion, and subsequent death. [6,10,27,30] As in trauma lethal triad represents a combination related to patients with hemorrhagic shockless frequently to patients undergoing general emergency surgeryreference to it might be considered as inappropriate during decision making process for patients undergoing general emergency surgery. [6,10,27,30] For those patients, clinician's decision is mainly based on septic shock's consequences. [6,10,27,30] A deeper analysis of our results, however, highlights how DCS might have represented overtreatment in good portion of ACD general population. [31,32] Taking into account the small number of patients with sepsis/septic shock/organ failure and the great number of Hinchey III patients, we could assume that many enrolled patients belonged to Hinchey III, being hemodynamically stable and without sepsis/septic shock at clinical presentation. [31,32] Therefore, we deem it possible to assume that Authors have often chosen DCS as an alternative to HP or PADS, rather than considering it as an effective measure to overcome patient's potentially lethal criticality.
In Hinchey III patients, who are hemodynamically stable and without sepsis/septic shock at clinical presentation, laparoscopic lavage might represent a more correct method than DCS, which is a more invasive strategy. [33][34][35] Some Authors consider laparoscopic lavage as one possible strategy of damage control aimed at representing a bridge to definitive surgery. [7] In accordance with Moore et al, we believe that laparoscopic lavage should not be equated to DCS. [7] Moreover, as defined by Moore et al, DCS role in emergency surgery is not only controversial but it is often misconcepted as "planned relaparotomy". [27] Reoperations are performed every 48 hours for "washing," until abdomen is free from ongoing peritonitis. [27] Then abdomen is closed. [27] Such method probably prevents and/or provides early treatment of secondary infections, thus reducing multiple organ failures and deaths. [27] Increased use of resources and higher risk of both gastrointestinal fistulas and delayed hernias represent drawbacks of planned relaparotomy. [27] In the light of excellent preliminary results confirmed by literature, we underline the need to further analyze outcomes of DCS in patients with acute peritonitis from perforated colonic diverticulitispossibly randomized, controlled, multicenter trialsby assessing both potential benefits and drawbacks. These trials should analyze DCS patients in comparison with patient populations undergoing HP and / or PRA/PADS. However, correct patient selection is required. In particular, there is a need for: i) patient populations with sepsis/septic shock/organ failure and ii) the adoption of an international and standardized definition of sepsis (e.g., The Third International Consensus Definitions for Sepsis and Septic Shock -Sepsis-3).

Limitations
Our systematic review introduces some limitations: i) the literature search was not extended to non-English-written scientific papers; ii) reported events were mainly small retrospective series; iii) populations under analysis showed heterogeneity; iv) many relevant data were not thoroughly described by the Authors, as reported in Tables 2-4; v) overlapping of analyzed populations cannot be ruled out either by 3 Perathoner group's manuscripts [11,12,16] or by 3 Sohn group's studies; [13][14][15] vi) sepsis, septic shock, and organ failure definitions differed among studies or were missing; vii) data on age, MPI, operative time, ICU stay, and hospital stay were reported in median days or mean days. For all these reasons, direct comparison of results turned out difficult.

Conclusion
DCS represents a well known strategy for trauma surgeons. At present, it is spreading in general emergency surgery. Its application to ACD patients seems to offer good outcomes with a lower percentage of patients with definitive ostomy, if compared to HP. However, correct definition of DCS eligible patients is paramount in avoiding overtreatment. In accordance to 2016 WSES Guidelines, DCS remains an effective surgical strategy in critically ill patients affected by sepsis/septic shock and hemodynamical unstability. We strongly believe that further studies are required to refine indications, timing, techniques of DCS, and resuscitation approaches to patients in non-traumatic abdominal emergencies.