Emergency surgery for malignant large bowel Assessing management options and outcomes

Authors: Terence N Moyana

Abstract

Malignant large bowel obstruction accounts for a disproportionately high percentage of colorectal cancer emergencies. Traditionally, it was treated by emergency surgery, which, depending on the circumstances, could involve primary resection or staged procedures. However, this was associated with considerable morbidity and mortality. Qiu et al sought to improve this by creating a nomogram that can be used as a benchmark in the management of such patients. Although the nomogram is meant to be a predictive model for recurrence, it is only based on a snapshot of parameters at 2 years. To be of maximum benefit to patients consenting for surgery and their caregivers, the performance of the model should be assessed over both the short- and long-term intervals (e.g., 30, 60, and 90 days as well as 1, 2, and 5 years or longer). Moreover, the heterogeneity of colorectal cancer (e.g., right-sided vs left-sided cancers vs rectal cancers) limits the nomogram’s applicability in certain situations, as it was constructed using a one-size-fits-all approach. It is also noteworthy that the increasing acceptance of self-expanding metal stents as an option to emergency surgery provides significant benefits for patients with malignant large bowel obstruction. Lastly, it is important to distinguish residual disease from recurrence, as conflating the two may confound parameters and study endpoints. This distinction has gained renewed interest with recent advances in liquid biopsies and genomics and how they can better define minimal residual disease.

Core Tip: Malignant large bowel obstruction was traditionally managed by emergency surgery, a strategy associated with considerable morbidity and mortality. The introduction of self-expanding metal stents in the 1990s provided an option to emergency surgery with significant benefits for these patients. Looking ahead, advances in liquid biopsies and molecular biomarkers are poised to enhance cancer diagnostics and the monitoring of minimal residual disease, promising improved outcomes in the future.

INTRODUCTION

Colorectal cancer (CRC) is the third most common visceral malignancy and the second leading cause of cancer-related deaths in many countries[1,2]. Malignant large bowel obstruction (MLBO) occurs in 10%–15% of patients with CRC but accounts for 75% of CRC-related emergencies[3-7]. It is associated with high postoperative morbidity and mortality, and population studies have shown that 25% of all postoperative deaths after surgery for CRC occur in patients who present with MLBO[4,5]. The longer-term survival rates (5 and 10 years) for MLBO are also considerably worse compared to non-obstructive CRC[8,9]. Therefore, it is essential to conduct an in-depth analysis of MLBO, as it may provide additional insights for quantifying surgical risk and assessing clinical outcomes.

This editorial discusses the study by Qiu et al[10], which aimed to investigate early post-operative recurrence (defined as recurrence within 2 years following surgery) in patients with MLBO. Five key parameters, namely (1) Preoperative carcinoembryonic antigen; (2) Preoperative systemic immune-inflammation index; (3) Tumor–node–metastasis stage; (4) Tumor differentiation; and (5) Ki-67 expression, were used to construct a nomogram prediction model to guide patient management. To achieve this objective, Qiu et al[10] analyzed 181 patients with MLBO who underwent emergency surgery at their institution, comparing those who developed recurrence with those who did not.

However, for a tool intended to provide guidance and practical clinical value, it would be prudent to consider other treatment options besides emergency surgery. These options should take into account not just recurrence rates at 2 years but also morbidity and mortality in the short term (e.g., at 30, 60, and 90 days) and long term (e.g., overall survival at 5 years or beyond), as is customary for CRC assessments[5,7,11-13]. In addition, the heterogeneity of the cohort in the study by Qiu et al[10] (e.g., demographics, tumor site, anatomy, stage, histology, and molecular biology) makes it challenging to construct a nomogram that applies to all MLBO cases.

SURGICAL OPTIONS

The traditional approach to managing MLBO has been emergency exploratory laparotomy. For proximally located CRCs (i.e., in the right or transverse colon), primary resection and anastomosis should be considered if the patient is hemodynamically stable, fit for resection, and the colon appears viable[14,15]. For more distally located tumors, various options can be considered depending on the exact site of the tumor, the viability of the obstructed bowel, and the overall patient condition. The classical three-stage approach involves creating a colostomy for obstruction relief, followed by a second surgery for cancer resection, and a third for colostomy reversal. This approach has fallen out of favor because it resulted in prolonged hospital stays without necessarily good prognostic outcomes[14,15]. Another approach is primary tumor resection with closure of the rectal stump and formation of proximal-end colostomy (Hartmann’s procedure). However, this approach carries considerable morbidity and mortality[6,14,15]. Furthermore, colostomy reversal is not possible in ≥ 40% of patients owing to advanced disease and/or comorbidities[6,14,15]. Therefore, it is not surprising that, in carefully selected patients who are hemodynamically stable and meet appropriate criteria (e.g., American Society of Anesthesiologists classification)[16], primary resection and anastomosis have been proposed as an alternative to Hartmann’s procedure[5,6,17]. Advocates of this approach reported a reduced incidence of complications and favorable long-term survival benefits when compared with staged procedures[14,15]. It is particularly suitable when performed by an experienced colorectal surgeon treating a younger patient with a freely mobile, easily resectable tumor[18]. Nonetheless, the threshold for fecal diversion should be low, as an anastomotic leak can have profound consequences on postoperative recovery and the timely administration of adjuvant chemotherapy or radiotherapy[18,19].

Overall, these various surgical approaches (primary anastomosis or staged procedures) have different morbidity and mortality risks at 30, 60, and 90 days, as well as at 2 years, 5 years, or more. Incorporating this information into Qiu et al’s study[10] would improve the utility of this nomogram and expand the range of informed treatment choices available to patients and their caregivers.

SELF-EXPANDING METAL STENTS

Patients with MLBO often present with advanced disease, are usually elderly, and are in poor medical condition due to the underlying tumor, malnutrition, dehydration, electrolyte imbalances, and a friable colonic mucosa resulting from prolonged bowel distension[6,7,14,15,20]. These metabolic derangements can be quite significant because, unlike acute obstruction (e.g., volvulus), the underlying process with MLBO is more chronic, and the obstruction is merely a final step that triggers the emergency. All this adversely affects surgical risk, especially with an operative emergency on an unprepared colon. Before the 1990s, most of these patients underwent emergency surgery[6,14,15]. Since then, self-expanding metal stents (SEMSs; Figure 1) have become an increasingly acceptable option for restoring luminal patency and decompressing the bowel before surgical resection, i.e., bridge-to-surgery (BTS) or for palliative purposes[3,5,14,21].

SEMS technology has advanced considerably in the last 30 years and includes covered, uncovered (Figure 1), partially covered, and uni-flanged or bi-flanged stents of variable length or shape that can be deployed depending on the clinical circumstances[14,21,22]. When used as BTS, SEMSs allow patients to be optimized for surgery by correcting fluid and electrolyte imbalances, improving nutritional status, treating sepsis, and permitting mechanical preparation of the bowel (Table 1)[7,14,15,20]. Furthermore, stenting allows the physician to perform a preoperative colonoscopy and/or imaging to rule out the possibility of synchronous tumors.

For rectal cancers in which neoadjuvant therapy is considered, it is important to note that preoperative chemoradiation can still be administered even after insertion of SEMs[23-25]. Stenting also provides the opportunity for accurate staging of the tumor and determining the best therapeutic option for the patient, considering that up to 50% of patients who undergo emergency laparotomy are not candidates for curative surgery[6,12,21]. In these patients who cannot be cured because of their advanced cancer or rapidly progressive disease, the ability to avoid emergency surgery is a major advantage of SEMSs, given their life expectancy is relatively short[6,7,25].

Compared to emergency surgery, BTS stenting allows elective surgery with primary anastomosis in a higher percentage of patients, shorter hospital stays, and reduced stoma rates without adversely affecting overall survival (Table 1)[7,11,14,25]. Nonetheless, it should be noted that there are certain situations in which placement of a SEMS is contraindicated (e.g., suspected or impending perforation, intra-abdominal abscess, or peritonitis; Table 2)[6,25,26]. Technical difficulties may also arise during SEMS insertion. These (1) Patients with fibrous adhesions from previous abdominal surgery; (2) Those with peritoneal carcinomatosis with concomitant colonic narrowing, tortuosity, or extrinsic compression; (3) Stenting of the right colon, which has a larger diameter and thinner wall. Although early experiences favored left-sided stenting, technical advancements have largely equalized outcomes, especially in experienced hands[25]; and (4) Tumors located < 5 cm from the anal verge, where proximity to the dentate line can cause severe tenesmus and increase the risk of expulsion of the stent[21,25,27]. SEMSs are also associated with complications, including bowel perforation, stent migration, bleeding, occlusion, re-stenosis due to tumor ingrowth or overgrowth, and stool impaction (Table 1)[7,21,24,25]. Patient optimization for BTS or palliation also includes pertinent radiologic assessment (e.g., contrast-enhanced computed tomography), as other causes of large bowel obstruction should also be considered[14,28].

OTHER CAUSES OF LARGE BOWEL OBSTRUCTION

Large bowel obstruction usually presents with abdominal pain, distention, and obstipation[14,26]. The majority of these cases (> 50%) are due to neoplasms, most often CRC, but a significant proportion of cases arise from other causes. In this regard, radiologic procedures such as computed tomography are very important in establishing the etiology of the obstruction. Other causes of large bowel obstruction include extra-intestinal tumors (e.g., ovarian, uterine, prostatic, or urinary bladder cancers). They can cause extrinsic compression on the large bowel owing to either the mass effect of the primary tumor or metastases. Non-neoplastic causes of large bowel obstruction include diverticular disease, especially when complicated by peri-diverticulitis or strictures. Volvulus arises from the axial rotation of the bowel around the mesentery and most commonly occurs in the sigmoid colon or cecum. Other causes include internal hernias, fibrous adhesions, intussusception, Crohn’s disease, fecal impaction, foreign bodies, and infections. Radiology is also important in ruling out functional types of large bowel obstruction, such as Ogilvie syndrome (acute colonic pseudo-obstruction) and toxic megacolon[29,30]. Differentiation between total mechanical obstruction and partial obstruction or pseudo-obstruction is important because complete obstruction is typically treated surgically, whereas the others can be treated with medication or simple bowel rest. Misdiagnosis can result in considerable morbidity and mortality[14,26,29,30]. Computed tomography is particularly helpful in identifying pseudo-obstruction or megacolon, which typically lacks a transition point or definitive area of caliber change[14].

RESIDUAL DISEASE VS RECURRENT DISEASE

Qiu et al’s[10] nomogram is based on the presence or absence of recurrent disease. Since it is such a critical parameter of the study, it is important to define “recurrence”, because what is sometimes described as recurrence could simply be “residual tumor” left behind at the time of surgery. Recurrence is defined by the American Joint Committee on Cancer and the Union for International Cancer Control as the appearance of cancer after a disease-free interval. Although recurrence is presumably the result of residual cancer, the defining feature is that the cancer must be undetectable in the immediate aftermath of the treatment[31-34]. Thus, it is generally easier to envisage tumor eradication in patients who undergo resection for primary non-metastatic CRC with curative intent[35,36]. In contrast, residual disease denotes the presence of microscopic or macroscopic tumor after treatment. The treatment modality can be surgical, chemo-/radiotherapy, or other modalities/combinations. The residual disease can be at the level of the primary tumor (e.g., positive surgical resection margins), in regional lymph nodes, or at distant sites. Based on the American Joint Committee on Cancer and the Union for International Cancer Control, residual disease can be classified into three R0, R1, and R2[31-34]. The R0 classification applies to cases in which residual tumor cannot be detected by conventional diagnostic methods, i.e., it is the desired outcome in surgical resection with curative intent. The R1 category applies to residual disease found by histologic examination either at the resection site or distant sites at the time of surgery[31,35]. In this regard, R1 disease tends to be more common in rectal cancer than in right- or left-sided colon cancers. R2 refers to macroscopically visible residual tumor, either clinically or pathologically. The reporting of residual disease (R status) is not clear in Qiu et al’s study[10].

As most tumors in Qiu et al’s study[10] were advanced and required emergency surgery, some cases would not have been treated with curative intent. In fact, 90 patients (i.e., approximately 50%) underwent palliative resection, which implies the presence of residual disease. Parenthetically, none of the patients were classified as stage IV, which is somewhat unusual considering the large size of the primary tumors (mean diameter approximately 5 cm). Studies have also shown that the likelihood of residual tumor after cancer therapy increases progressively with higher disease stage[31,34].

While the established classification of residual disease (R0–R2) is useful, this approach has limitations, as highlighted by the increasingly promulgated concept of minimal residual disease (MRD). MRD is defined as clinically undetectable microscopic disease remaining after curative treatment[37,38]. As such, it holds the potential to precipitate disease relapse. Recently, there has been growing interest in the potential of circulating biomarkers in detecting MRD[37-39].

CIRCULATING BIOMARKERS

The traditional circulating biomarker used for CRC is carcinoembryonic antigen, which was one of the biomarkers evaluated by Qiu et al[10]. However, its levels can be elevated in benign conditions such as inflammatory bowel disease, diverticulitis, pancreatitis, and liver disease, as well as in cigarette smokers and individuals who consume alcohol[39,40]. This limits its role as a screening tool for CRC. However, when initiated in the preoperative setting, it has been validated as a surrogate of disease burden for patients undergoing CRC resection as well as a predictor of cancer recurrence[40].

More recent advances in genomics and cell biology have stimulated renewed interest in liquid biopsy techniques such as circulating tumor DNA, cell-free DNA, circulating tumor RNA, circulating total nucleic acid, tumor-educated platelets, and plasma proteomics[39-42]. They offer distinct advantages such as simplicity in sampling, minimal invasiveness, and improved ability to capture intratumor heterogeneity. Additionally, they detect real-time cancer dynamics compared to archival material from tissue biopsies or resection specimens. For tumors that have not yet metastasized, liquid biopsy can be helpful in identifying MRD, potentially guiding adjuvant treatment and reducing overtreatment. In the metastatic setting, liquid biopsy is useful for real-time treatment monitoring, tracking of clonal evolution, and assessing mechanisms of resistance. The integration of these techniques into clinical practice could enhance therapeutic strategies and curtail unnecessary interventions[41,42]. Collectively, these emerging techniques promise to revolutionize noninvasive cancer diagnostics and monitoring. However, there is currently wide variability in how the biomarker assays are developed and validated. Standardization is needed to ensure consistency, reliability, and wider clinical acceptance of these approaches[39-42].

HETEROGENEITY OF THE MLBO GROUP

The prevalence and distribution of the histologic subtypes of CRC vary according to tumor location in the large bowel. For example, the proportion of microsatellite instability-high, mucinous, medullary, and signet-ring cell carcinomas is higher on the right side, whereas conventional adenocarcinoma is relatively more prevalent on the left side[43-47] (Figure 2). Notably, Qiu et al’s study[10] had a disproportionately large percentage of mucinous tumors (57%), even though this subtype typically accounts for only approximately 10% of all CRCs[43,48]. Mucinous adenocarcinomas are more likely to dissect tissue planes, aiding their spread and exacerbating tumor dissemination and progression. The incidence of peritoneal carcinomatosis is higher in right-sided cancers, while that of hepatic and pulmonary metastases is higher in left-sided cancers[44,49,50]. These biological and anatomical differences may influence disease behavior and outcomes, potentially affecting the functionality of Qiu et al’s[10] nomogram.

Medullary carcinomas (Figure 2B) are commonly described as poorly differentiated or undifferentiated; however, they have a significantly better prognosis than conventional poorly differentiated adenocarcinomas[43,51,52]. This underscores a limitation of Qiu et al’s[10] nomogram, which incorporates tumor grade without taking into account the tumor subtype.

It should also be noted that within the context of left-sided colon cancers, rectal cancers are often regarded as a separate entity because stage II/III rectal tumors have a different treatment paradigm compared to other CRCs. For example, neoadjuvant or total neoadjuvant therapy is commonly deployed for this cohort[47,53,54]. Qiu et al’s[10] study, however, does not clarify whether neoadjuvant treatment was administered in any of the 61 rectal cancer cases included. In addition, the study does not provide sufficient information on the effect of adjuvant chemoradiation or immunotherapy on disease recurrence[47,55-57]. Furthermore, with regard to tumor biology, several parameters such as lymphovascular invasion, extramural vascular invasion, and tumor budding are regarded as very important parameters in CRC progression. Their omission from Qiu et al’s[10] predictive model could raise questions regarding its practicality and applicability for patients with MLBO. Overall, the heterogeneity within the study cohort of Qiu et al[10] could be a confounding factor that limits the generalizability of their findings and nomogram. Taken together, this means that comparisons between recurrence and non-recurrence groups in MLBO will carry more weight and provide deeper insights if the heterogeneity of disease biology is taken into account.

PERFORATION

Owing to the generation of intraluminal pressure as a result of obstruction, MLBO is more likely to be complicated by perforation compared to non-obstructive tumors. According to Laplace’s law, the right colon, which is more capacious and relatively thin-walled, is particularly prone to perforation[14]. In severe cases, especially when the ileocecal valve is competent, an outright cecal blowout may occur[28]. Perforation provides cancer cells direct access to the peritoneal cavity, thus upstaging the tumor and considerably raising the prospect of tumor recurrence. Notably, perforation is not mentioned as a complication or included as a parameter in the cases analyzed by Qiu et al[10].

FUTURE DIRECTIONS

MLBO nomograms based on larger studies with more statistical power

While Qiu et al’s study[10] is commendable, future studies could enhance such predictive models by enrolling larger numbers of patients. This would provide more statistical power to overcome the limitations arising from a small sample size/multiple parameters.

Randomized controlled studies on SEMS

More randomized controlled studies are required with respect (1) The type of SEMS deployed, including covered, uncovered, or partially covered designs[21,22,25,58,59]; (2) Optimal timing between insertion of SEMS and surgery; (3) Long-term oncological outcomes, as current research predominantly focuses on the well-documented short-term benefits of SEMS[25]; and (4) Use of neoadjuvant or adjuvant therapy with SEMS in situ (e.g., anti-angiogenic drugs such as bevacizumab)[23-25].

Liquid biopsy and MRD

Liquid biopsies are increasingly being integrated into the management algorithms of CRC, as they can offer insights into MRD through tumor genomics, transcriptomics, and proteomics. However, there is currently much variability in assay protocols, which complicates interpretation and clinical adoption[25,41,42]. Thus, there is a need to standardize methodologies and develop uniform clinical recommendations for serial sampling, testing frequency, and actionable thresholds.

CONCLUSION

MLBO accounts for a disproportionately large percentage of colorectal cancer emergencies. Traditionally, it was treated by emergency surgery, which, depending on the circumstances, could involve primary resection or staged procedures. However, this was associated with considerable morbidity and mortality. Qiu et al[10] sought to address this by creating a nomogram that can be used as a benchmark in the management of such patients. Although the nomogram is meant to be a predictive model for recurrence, it is only based on a snapshot of parameters assessed at 2 years. To be of maximum benefit to patients consenting for surgery and their caregivers, the performance of the model should be assessed over both the short and long term (e.g., 30, 60, and 90 days as well as 1, 2, and 5 years and beyond). Secondly, the heterogeneity of CRC, such as the differences between right-sided, left-sided, and rectal cancers, limits the applicability of a one-size-fits-all nomogram. Lastly, the increasing acceptance of SEMSs as an alternative to emergency surgery provides significant benefits for patients with MLBO.