Less Is ( Sometimes ) More: Laparoscopic Peritoneal Lavage and Drainage for Diverticulitis

Introduction

Diverticulitis is one of the most common gastrointestinal diseases in the United States, and its incidence has steadily increased in recent years due in part to a rapidly aging population. ¹ Although historically a disease only affecting the elderly, there has been a notable increase in recent years in the incidence of diverticulitis in those younger than 50 years for reasons that are not fully elucidated.^2–5 Considering that about 20% of patients who develop diverticulitis will have at least one recurrent episode, this disease poses a significant burden to the health care system.^1,5 However, as diverticulitis has evolved throughout the years, so have the strategies for its treatment.

Currently, there are many unanswered questions regarding the best surgical treatment for those patients with acute diverticulitis (AD) who require emergency surgery. Patients with AD requiring emergency surgery typically present with peritonitis from colonic perforation. Traditionally, surgical treatment in these patients was performed in two steps. First, an emergent surgical resection of the diseased colon and creation of an end colostomy and rectal stump, known as a Hartmann’s procedure (HP) was performed. This initial procedure was followed by elective surgery to reverse the ostomy and reestablish colonic continuity after the resolution of the AD. However, HP is associated with a high rate of morbidity and mortality, and patients are often left with permanent ostomies due to the risk of reversal.^6,7 In an effort to mitigate these risks in the advent of minimally invasive surgery, laparoscopic peritoneal lavage and drainage (LLD) emerged as an attractive alternative to HP in patients with complicated AD.

LLD was initially considered a “safe and easy” procedure with minimal side effects and favorable rates of morbidity and mortality. ⁸ But through the years, LLD has been the subject of multiple randomized controlled trials (RCT), retrospective studies, systematic reviews, and meta-analyses with widely variable results, sparking controversy and scrutiny. The contradicting data and subsequent lack of consensus regarding the benefits of LLD precluded its widespread adoption. ⁹ However, LLD may still play a role in the surgical treatment of AD. The purpose of this narrative review is to provide a brief overview of the available literature regarding the role of LLD in the management of AD, and its indications, techniques, and observed outcomes.

Trends in the Management of Acute Diverticulitis

The management of AD is largely dependent on the disease severity and the clinical presentation of each patient, which can fall within a wide spectrum. While some patients may present with only mild localized disease and minimal symptoms, others may present with colonic perforations, generalized peritonitis, and sepsis. The disease severity can be clinically characterized using the Hinchey classification system, originally developed by Hinchey and colleagues in 1978. ¹⁰ The system is still preferentially used by surgeons despite more recent modifications and proposed alternative grading systems. ¹¹ This classification system is divided into four stages: Hinchey I and II refer to diverticulitis with a small, contained pericolic abscess or a distant (pelvic, retroperitoneal, or intra-abdominal) abscess, respectively. Hinchey III and IV are more severe forms with either purulent (Hinchey III) or feculent (Hinchey IV) peritonitis. Historically, there was a low threshold for surgical intervention in patients with AD based on the notion that recurrent episodes of diverticulitis compounded the risk and complications of each episode. However, this notion has since been disproven.^5,12 Current practices shifted toward more conservative approaches with nonoperative management indicated in a larger proportion of cases, causing a decrease in rates of emergent surgical interventions performed for AD.^5,12 Nonoperative management is largely successful in most cases due to advancements in diagnostic tools, antibiotic therapies, and treatment adjuncts such as percutaneous abscess drainage. ¹³ Some studies suggested that antibiotics might not even be necessary in milder cases of AD.^13,14 Nonetheless, current guidelines still recommend emergent surgical intervention in patients with generalized peritonitis, namely patients with Hinchey III and IV disease. ¹⁵

As previously stated, HP is the standard of care for the emergency surgical treatment of complicated AD. However, given the high morbidity and mortality rates associated with an emergency HP and the low rates of colostomy reversal, surgical alternatives or modifications to the procedure were explored.^6,7 One proposed modification is the creation of a primary anastomosis at the time of resection with or without a proximal diverting ostomy. Though the resulting studies showed improved long-term outcomes, quality of life, and rates of ostomy reversal, the short-term morbidity and mortality rates were not significantly different from HP, and thus the practice was not widely adopted. ¹³

In the advent of laparoscopy, minimally invasive approaches were proposed for the surgical management of AD. In fact, laparoscopic HP improved postoperative outcomes with higher rates of subsequent ostomy reversal versus open HP, but this procedure is technically challenging and is often overlooked in an emergency by surgeons who may lack the procedural experience.^16,17 However, in the 1990s, LLD emerged as a relatively accessible minimally invasive treatment option for patients with generalized peritonitis secondary to complicated AD. LLD was specifically intended for patients with Hinchey III disease, without intraoperative evidence of fecal contamination. ¹⁸

The principle of LLD is to decrease the inflammatory environment around the bowel by irrigating the abdomen with warm saline to remove purulent fluid and reduce levels of endotoxins in the peritoneal cavity. ¹⁹ Theoretically, this dialytic effect reduces the inflammatory burden and enhances the peritoneal landscape in septic patients. LLD was intended to serve as a definitive treatment or as a bridge to delayed colonic resection.^8,19 Though this might intuitively appear to be a reasonable option for patients with peritonitis secondary to AD and an attractive option for surgeons attempting to avoid an emergent HP, multiple LLD studies reported mixed results. The proposed reasons for the notable conflicts in the data include the variability in patient selection, selection bias, and the lack of a standardized surgical technique.

Indications for Laparoscopic Peritoneal Lavage and Drainage

Although diffuse peritonitis secondary to complicated AD is an indication for emergent surgical intervention, there are currently no clear indications for pursuing LLD over colonic resection.^11,15 Some authors indicate that patients with Hinchey III AD may be candidates for LLD, but patient selection is critical for favorable outcomes. ¹¹ Naturally, the same contraindications to exploratory laparoscopy apply to LLD. Additionally, the data from several studies demonstrated that significant comorbidities, intraoperative evidence of frank perforation, feculent peritonitis, no prior history of AD, high American Society of Anesthesiologists grade, and high Mannheim’s Peritoneal Index (MPI) scores were associated with increased risk of LLD failure and should therefore be considered potential contraindications to LLD.^9,20–22 Therefore, the American Society of Colon and Rectal Surgeons actively recommends against LLD for patients with feculent peritonitis (Hinchey IV) and recommends individual considerations for performing LLD in those with purulent peritonitis (Hinchey III). ¹⁵

Operative Techniques and Perioperative Care

A comprehensive review published in 2019 by Gregori et al. looked at the operative techniques employed in a total of 28 studies, including 3 RCTs. ¹¹ Their review showed that, in all studies, preoperative broad-spectrum antibiotics were initiated upon diagnosis and continued for 5–7 days postoperatively. The types of antibiotics used varied and were mostly subject to surgeon preference. Once in the operating room, initial access into the abdomen was achieved via the Hasson technique in most cases. However, the total number of trocars and their position in the abdomen varied greatly depending on surgeon preference (though in most cases only three trocars were placed).

Although all surgeons performed a thorough examination of the abdominal cavity to rule out other sources of peritonitis, there was a significant discrepancy in the extent of adhesiolysis and mobilization of the affected colon to identify the colonic perforation. ¹¹ Some authors favored aggressive adhesiolysis to appropriately assess the colon, localize any concealed colonic perforation, and avoid missing occult feculent contamination, while others favored minimal adhesiolysis to avoid compromising a contained or healed perforation and cause overt feculent contamination. Proponents of extensive adhesiolysis argued that thorough peritoneal lavage may be hindered by limited adhesiolysis, as the instilled saline may be unable to reach interloop abscesses or purulence contained between tightly adhered loops of the bowel. In fact, Gregori et al. argued that this likely caused the early cessation of one RCT (the laparoscopic lavage [LOLA] arm of the LADIES study, discussed later), as patients who underwent limited adhesiolysis had high rates of reoperation, leading to the early termination of the study. ¹¹ However, a prior retrospective multicenter international study, the LLO study published in 2018 by Binda et al., noted a significant association between adhesiolysis and operative outcomes, stating that those who underwent extensive adhesiolysis had higher rates of free perforation and LLD failure than those who had limited or no adhesiolysis. ²⁰ In light of this, they recommend limited adhesiolysis and careful visual inspection of the colon with the use of adjunct tools like flexible sigmoidoscopy and a hydropneumatic test if the presence of free perforation is unclear. ²⁰

In cases where a colonic perforation was identified, the appropriate management of the perforation was another point of controversy. Some authors argued that the presence of a free colonic perforation is an automatic indication to abort LLD and proceed with colonic resection, but others believed colorrhaphy (with absorbable suture) with an omental patch and/or fibrin glue was sufficient. ¹¹ Although there is not a study directly comparing these two approaches, it should be noted that the LLO study illustrated that the presence of a free perforation was a risk factor for the failure of LLD. ²⁰ The volume of saline typically used during lavage can also vary, with some studies using between 3 and 25 L.^9,11 However, most authors abide by the concept of continuing the lavage until the effluent runs clear.^11,20 Almost all surgeons utilized warm normal saline for the procedure, with few exceptions. ¹¹ Although all surgeons placed drains during LLD, the number, type, location, and duration of drains were also notably different between studies. Some authors placed drains near the site of the colonic perforation to facilitate early detection of a leak, whereas others placed drains in Douglas’ pouch to prevent the formation of a pelvic abscess.^9,11 A majority of surgeons placed a single drain during the procedure, but placement of up to four drains was reported. ¹¹ The drains stayed in place anywhere from 1 to 6 days (with one exception of a study reporting a maximum of 21 days), typically being removed once the outflow became serous.^11,23

Postoperative Outcomes

The numerous studies looking at outcomes after LLD showed highly variable results. Some authors argued that these variabilities may be attributed to study design flaws and small sample sizes. Additionally, the best level of evidence comes from three RCTs performed in Europe that used different study endpoints, patient selection criteria, randomization, and operative techniques, which renders direct comparison between them difficult.^19,22 Results from a fourth RCT (LapLAND) are still pending. ²⁴

The LADIES trial was a two-arm multicenter RCT performed across 42 hospitals in Belgium, Italy, and the Netherlands. ²³ The LOLA arm included 90 patients and directly compared LLD versus sigmoidectomy with primary anastomosis in patients with Hinchey III disease. The trial was terminated early by the data and safety monitoring board because of interim analyses demonstrating a significantly higher rate of major postoperative morbidity (defined as surgical reintervention, abdominal wall dehiscence, abscess requiring percutaneous drainage, urosepsis, respiratory insufficiency, myocardial infarct, and renal failure) in patients undergoing LLD. The combined 30-day major morbidity and mortality rate was 39% for those undergoing LLD versus 19% for those who underwent sigmoidectomy (P = .0427). ²³ However, 3-year follow-up data was later published, showing that, although short-term morbidity was higher for the LLD group, the overall cumulative morbidity and mortality rate was not different between the groups at 3 years. Additionally, they reported that patients who had LLD had significantly lower rates of reoperation and fewer of them had stomas at 3 years. ²⁵ The authors ultimately concluded that LLD was not superior to sigmoid resection in the treatment of Hinchey III disease. ²²

The SCANDIV trial was a two-armed multicenter randomized clinical superiority trial performed across 21 centers in Sweden and Norway. ²⁶ This study also compared LLD with colonic resection in patients with AD. AD was defined by CT imaging findings consistent with colonic perforation and clinical signs of peritonitis that required surgery. Patients determined to have Hinchey IV disease intraoperatively were treated with HP regardless of preoperative randomization. A total of 199 patients were enrolled and the initial published results showed no significant difference in rates of severe postoperative complications or mortality rates at 90 days. However, despite the LLD group having significantly shorter operative times, they showed higher rates of re-operation than those who underwent resection. ²⁶ Subsequent studies published by the group including 1- and 5-year follow-up results showed no difference in severe complications or disease-related mortality at either time point but did note significantly lower stoma rates for those who underwent LLD (though there was no significant difference in reported quality of life between groups).^27,28 It is important to note that surgeon preference and local practices were the deciding factors on the procedure performed for patients in the colonic resection group, and the procedures performed varied in approach (open versus laparoscopic) and type (HP versus resection with primary anastomosis, with or without diverting ostomy). ²⁶ Based on these results, the authors concluded that their findings did not support LLD as a treatment for perforated AD. ²²

Finally, the DILALA trial was another multicenter RCT performed in Sweden and Denmark across nine surgical departments. The study enrolled 83 patients with Hinchey III disease and compared outcomes after LLD with open HP. ²⁹ Their initial published results showed no difference in immediate postoperative morbidity or mortality, but did show that the LLD group had shorter operative times and hospital stay prompting them to determine that LLD was a feasible and safe alternative to HP. ²⁹ A 2-year follow-up study showed that the LLD group had a 45% reduced risk of reoperation (RR = 0.55, P = .012) and fewer operations overall (RR = 0.51, P = .024) at 2 years, but no difference was noted in rates of readmission or mortality rates when compared with the open HP group. ³⁰ Based on these results, they conclude that LLD is a better option than HP for the treatment of Hinchey III disease. ³⁰

Results from these RCTs, along with those of multiple retrospective studies, prompted several reviews and meta-analyses to attempt to arrive at a conclusion regarding the role of LLD in the treatment of AD. Interestingly, many of these meta-analyses include the same studies but their conclusions are very different. This is likely due to the incongruences in study design that make the available studies difficult to compare. Despite this, Kiely et al. attempted to draw some conclusions from several meta-analyses in their published review. ²² They noted that overall there was no significant difference in postoperative morbidity or mortality between the LLD group and those who underwent some form of resection, but that patients who underwent LLD were consistently shown to have higher rates of postoperative abscess formation and had higher rates of unplanned reinterventions within 30 days. ²² An important caveat is that some studies count nonsurgical procedures like percutaneous abscess drainage as a reintervention, which in some cases qualifies as a failure of LLD.

Conclusion

The role of LLD in the treatment of AD is still widely debated. Despite the abundance of data from RCTs and retrospective studies, the heterogeneity in study design makes generalized conclusions very challenging. Some reviews of meta-analyses demonstrated that rates of postoperative morbidity and mortality are similar between LLD and resection strategies, but that those who undergo LLD have higher rates of reintervention.^9,22 But this may not entirely disqualify LLD as a useful procedure. LLD still presents an attractive surgical option, despite high re-intervention rates, as it can be an initial attempt to treat a patient with peritonitis and potentially save them from a stoma, if successful.

It is still possible that LLD may play a significant role in the treatment of AD in the future after additional research, but several obstacles must be addressed to improve the quality and generalizability of the results. Firstly, objective criteria must be defined to select patients who should be considered for LLD to delineate a strict indication for LLD in the future. In the studies discussed, patient selection and treatment strategies relied heavily on the judgment and preferences of the surgeon, which have been criticized as a clear cause of selection bias. Additionally, the technical aspects of LLD should be standardized based on the available data, specifically, the degree of adhesiolysis and the management of colonic perforations when identified. By standardizing the surgical approach of LLD, it may be possible to observe consistent surgical outcomes. Finally, consistent endpoints are required to define success after LLD in order to define its role in the emergency surgical management of AD.