How to decide on stent insertion or surgery in colorectal obstruction?

Abstract

Colorectal cancer is one of the most common cancers in western society and malignant obstruction of the colon accounts for 8%-29% of all large bowel obstructions. Conventional treatment of these patients with malignant obstruction requiring urgent surgery is associated with a greater physiological insult on already nutritionally replete patients. Of late the utility of colonic stents has offered an option in the management of these patients in both the palliative and bridge to surgery setting. This has been the subject of many reviews which highlight its efficacy, particulary in reducing ostomy rates, allowing quicker return to oral diet, minimising extended post-operative recovery as well as some quality of life benefits. The uncertainity in managing patients with malignant colonic obstructions has lead to a more cautious use of stenting technology as community equipoise exists. Decision making analysis has demonstrated that surgeons’ favored the use of stents in the palliative setting preferentially when compared to the curative setting where surgery was preferred. We aim to review the literature regarding the use of stent or surgery in colorectal obstruction, and then provide a discourse with regards to the approach in synthesising the data and applying it when deciding the appropriate application of stent or surgery in colorectal obstruction.

Key Words: Self-expanding metallic stent; Stenting; Surgery; Colorectal cancer; Large bowel obstruction; Radiology

Core tip: Despite the accumulation of data on stent insertion, the choice of stent or surgery as the most appropriate modality in the management of colorectal obstruction presents a constant decision dilemma. When cure is possible we want that, but with minimal morbidity. In a group of patients who are prone to higher rates of morbidity and mortality, this can be problematic and full of uncertainty. This review takes an approach to review the primary and secondary outcomes established in the literature regarding the use of stent or surgery in colorectal obstruction, and then create discourse and a structured approach in regards to synthesising the data and applying it when deciding the appropriate application of stent or surgery in colorectal obstruction.

INTRODUCTION

Despite the accumulation of data on stent insertion, the choice of stent or surgery as the most appropriate modality in the management of colorectal obstruction presents a constant decision dilemma. When cure is possible we want that, but with minimal morbidity. In a group of patients who are prone to higher rates of morbidity and mortality, this can be problematic and full of uncertainty.

This review takes an approach to review the primary and secondary outcomes established in the literature regarding the use of stent or surgery in colorectal obstruction, and then create discourse and a structured approach in regards to synthesising the data and applying it when deciding the appropriate application of stent or surgery in colorectal obstruction.

Colorectal cancer is one of the most common cancers in western society and malignant obstruction of the colon accounts for 8%-29% of all large bowel obstructions[1]. Other causes of large bowel obstruction include uterine, ovarian, gastric, breast, bladder and kidney malignancies[2]. Conventional treatment of these patients with malignant obstruction requiring urgent surgery is associated with a greater physiological insult on already nutritionally replete patients. This is associated with mortality in 15%-34% of patients and morbidity in 32%-64% of patients[3]. Of late the utility of colonic stents has offered an option in the management of these patients. The first being used by Dohmoto et al[4] in 1991. Tejero et al[5] described the use of colonic stents as a “bridge to surgery” in 1994. This has been the subject of many reviews which highlight its efficacy, particulary in reducing ostomy rates, allowing quicker return to oral diet, minimising extended post-operative stay and some quality of life (QoL) benefits[6]. Xinopoulos et al[7] demonstrated that self-expanding metallic stent (SEMS) placement represents an alternative approach to colostomy for patients with inoperable malignant colonic strictures. The uncertainity in managing patients with malignant colonic obstructions has lead to a more cautious use of stenting technology as community equipoise exists[8].

PATIENT AND DISEASE FACTORS

Regarding the location of obstructing colonic malignancy, Fiori et al[9] in 2004 reported that 63.6% of obstructing malignancies occur in the rectum and 36.3% in the rectosigmoid/sigmoid colon. Sankararajah et al[10] in 2005 observed 37% in the rectosigmoid, 21% in the sigmoid colon, 16% at the splenic flexure, 16% in the descending colon, 5% in the rectum and 5% in the ascending colon. van Hooft et al[11] in 2008 observed 76% obstruction in the rectosigmoid and 24% obstruction in the descending colon. With the majority of obstructing pathology being on the left side, this makes these lesions amenable to endoscopic intervention. Sankararajah et al[10] demonstrated malignant stricture length to be in the range of 3-7 cm meaning that all these lesions are within “stentable” range. Fiori et al[9] and van Hooft et al[11] collected data on patient ASA level with all patients included in their trails being ASA 1 to 3. The majority of the patients were in the ASA 2 category.

MORBIDITY OF SURGERY V STENT

While decision making with regards to the utility of stents in patients with metastatic disease may be easier for the treating clinician, this decision is more difficult to make for patients with local disease. A recent randomised controlled trial (RCT) by Young et al[6] reported that in a population of patients with incurable metastatic large bowel obstruction, stent use was associated with faster return to diet, decreased stoma rates, reduced post-procedure stay, and some QoL benefits.

The decision with regards using a stent in patients with non-metastatic malignant bowel obstruction is one that is fraught with indecision due to the theoretical risk of perforation converting a once potentially curable disease to incurable[12,13]. However this risk needs to be balanced with multiple other factors, principally being the patients pre-existing morbidities and the need for emergent surgical intervention. In this day, with highly trained endoscopists, the more imminent risk of perforation is much lower in some centres than the reported 4%.

Efficacy

The efficacy of SEMSs as a tool in the treatment of malignant colonic obstruction has been demonstrated well over the past few years. Many randomised control trials have supported their use and hence should be considered a valid option in the treatment of this condition (Table 1).

Table 1 Summary of the studies included in this review.

Ref.	Centres and aim	No. of patients (stenting/surgery)	Stenting morbidity	Surgery morbidity	Stenting mortality	Surgery mortality	Stenting efficacy
Fiori et al[9]	Single centre, palliation	11/11	0	1/11 (9%)	0	0	100%
Xinopoulos et al[7]	Single centre, palliation	15/15	0	0	0	0	93%
van Hooft et al[11]	Multicentre, palliation	11/10	11/11 (100%)	5/10 (50%)	3/11 (27%)	0	82%
Sankararajah et al[10]	Single centre	9/9	2/9 (22%)	6/9 (67%)	1/9 (11%)	1/9 (11%)	78%
Cheung et al[23]	Single centre, bridge to surgery	24/24	2/24 (8%)	17/24 (71%)	0	0	83%
van Hooft et al[24]	Multicentre, bridge to surgery	47/51	25/47 (53%)	23/51 (41%)	9/47 (19%)	9/51 (18%)	70%
Ho et al[15]	Single centre, bridge to surgery	20/19	7/20 (35%)	11/19 (35%)	0	3/19 (16%)	70%
Young et al[6]	Multicentre, palliative	26/26	10/26 (38%)	14/26 (54%)	0	0	79%

The 2011 review by Sagar et al[14] reported an clinical relief of obstruction in the colonic stenting group to be approximately 0.66 d compared to 3.55 d in the emergency surgery group, with an overall success rate of 86%. In Ho’s review in 2012, the placement of self expanding metallic stents took a median time of 35 min (range, 20-80 min). Seventy percent patients (14/20) had been stented successfully. Following stent placement, they resumed a diet after approximately day 2 and were discharged about day 4. Six out of 20 patients failed stenting with the main cause being the inability to pass the guide wire across the stenotic cancer (4/6 cases)[15]. This technical success was also noted in the review by Khot et al[16]. It may be overcome with the use of a pediatric nasogastroscope[17].

Both Tan et al[18] and Zhang et al[19] reviews demonstrated that a higher primary anastomosis rate and lower morbidity rate was achieved in the group receiving colonic stents.

In the study by Ho et al[15] stented patients were sent home significantly sooner than in the emergency surgery groups, with medial length of stays at 6 d vs 8 d respectively (P = 0.028). Furthermore, they demonstrated significantly better outcomes for the stenting group that went on to have elective surgery compared to the group randomized to have emergency surgery[14].

A recent metaanalysis by Zhao et al[20] emphasized that there is limited data on the long term survival of patients with malignant left sided colonic obstruction when comparing emergency surgery with semi-elective use of stents. With limited data, recommendation was made for more studies on the topic[20].

BLOCKAGE

Blockage of stents principally affects patients who have long term stent insertion in the palliative setting. In the review by Khot et al[16], the overall, reobstruction occurred in 52 of 525 (10%) cases with only three patients in the “bridge to surgery” group having reobstruction. The reasons of reobstruction in these patients included tumour in-growth in 32 (62%), stent migration in seven (13%) and faecal impaction in 13 (25%)[16]. These issues with obstruction of the stent can be managed expectantly with surveillance being tailored to the patient’s condition. In general, patients who are having the stent as a bridge to surgery would very rarely experience obstruction. Patients with palliative stent insertion who are not candidates for surgery would present the main group with tumor related blockage and this may be managed expectantly with re-stenting of the lesion.

STOMA RATES

A major advantage of colonic stent placement is the reduction of stoma formation rates[5,6,9,11,21]. This represents a significant improvement in the patient outcomes with relation to physical recovery and overall QoL issues. In the meta-analysis by Cennamo et al[22], the permanent stoma creation rate was 38/152 (25%) in the stent group and 78/162 (48.1%) in the surgical group; the pooled analysis showed a significantly higher rate in the surgical group[20]. In the RCT by Young et al[6], none of the 19/26 patients in the stent group who were successfully stented required a stoma while 24/26 in the surgery group required a stoma to be fashioned (P < 0.001).

PERFORATION RATES

The decision of using a stent in patients with non-metastatic malignant bowel obstruction is one that is frought with indecision due to the theoretical risk of perforation converting a once potentially curable disease to incurable. However this risk needs to be balanced with multiple other factors, prinicipally being the patients pre-existing morbidities and the need for emergent surgical interventions. In four trials, no stent related perforation was noted (Young et al[6] 2015, Cheung et al[23] 2009; Fiori et al[9] 2004; Sankararajah et al[10] 2005). In two of the RCTs by Khot et al[16] 2011 and van Hooft et al[24], a perforation rate of 4% was noted. Khot et al[16] states that this rate was significantly associated with balloon pre-dilatation. With Van Hooft’s study the large number of centres[24] involved in the study may not have allowed a standardisation in the technique and also local expertise may vary considering that some centres contributed one patient over the two year period.

DEATH

In malignant obstruction of the colon, emergency surgery is associated with a high mortality rate of 10%-30%, when compared to < 5% rate in elective surgery for colorectal cancer[25,26]. Three meta-analyses[14,18,19], did not show any advantage in terms of post- operative mortality between the emergency surgery and stenting groups. In the recent RCT by Young et al[6], similar mortality figures were noted in both groups, noting that this patient population was palliative. A review of the United Kingdom National Audit showed that patients undergoing surgery for left-sided colonic obstruction had an operative mortality rate of 12.9%[27]. The mortality rate with stenting being a lot lower at 1%, giving evidence that it is a safe method to decompress a patient as a bridge to surgery[16].

COST

The cost of stents utility needs to be weighed up against many factors. They may represent an expensive option in isolation, however overall they represent a cost-effective option in the treatment of malignant obstruction of the colon. A study from the United Kingdom demonstrated the cost of a palliative stent was fifty percent less than surgical decompression and that the expense of ‘bridge to surgery was reduced by twelve percent with compared to a two stage procedure[28]. In the review by Fiori et al[9], the median hospital stay was 2.6 d for stent group and the median hospital stay was 8.1 d for the stoma group.

Other factors such as QoL, faster return to normal bowel function and significantly less physiological insult make stenting a much more cost-effective option. Further, the additional costs of outpatient stoma care should also not be forgotten[16].

QOL

Increasing evidence has been published with regards to the QoL of patients undergoing stents and surgical intervention for the management of malignant bowel obstruction. In the study by van Hooft et al[24] (2011), primary outcome of global health status was recorded and no significant difference was noted between the two groups. More recently, Young et al[6] (2015) observed that 15/26 (58%) patients in the stent group patients were recorded as having an increased QoL from baseline to one week compared to 7/26 (27%) of the surgery group (P = 0.02). The surgery group had significantly lowered QoL compared to the stent group from baseline to 1 and 2 wk (P < 0.001 and P < 0.012), and from baseline to 12 mo (P = 0.01) in favor of the stent group, while both reported reduced QoL[6]. There were no significant differences in whether the patient had an increased or decreased QoL at any other time point.

DECISIONS

The treatment of patients with senting technology is one that has traditionally being frought with concern by the treating clinician. A recent study by Suen et al[8] demonstrated that there would be limitations in conducting a future randomised controlled trial to assess the use of colonic stenting especially in the curative setting. Surgeons’ favored the use of stents in the palliative setting preferentially when compared to the curative setting where surgery was preferred (Table 2).

Table 2 Surgeons’ treatment preferences in different clinical scenarios (Suen et al[8]).

Clinical scenarios		Level of clinical certainty			Evidence of community equipoise?
		Surgery (%)	Undecided (%)	Stent (%)
1	70yo; partial obstruction; metastatic cancer; ASA score 4	8	12	80	N
2	70yo; complete obstruction; metastatic cancer; ASA 4	9	8	82	N
3	50yo; partial obstruction; metastatic cancer; ASA 4	15	10	75	N
4	50yo; complete obstruction; metastatic cancer; ASA 4	12	8	80	N
5	70yo; partial obstruction; metastatic cancer; ASA 1	51	19	30	Y
6	70yo; complete obstruction; metastatic cancer; ASA 1	40	13	47	Y
7	50yo; partial obstruction; metastatic cancer; ASA 1	60	17	23	Y
8	50yo; complete obstruction; metastatic cancer; ASA 1	51	14	35	Y
9	70yo; partial obstruction; curable cancer; ASA 4	66	15	19	Y
10	70yo; complete obstruction; curable cancer; ASA 4	41	13	46	Y
11	50yo; partial obstruction; curable cancer; ASA 4	73	10	17	N
12	50yo; complete obstruction; curable cancer; ASA 4	50	11	39	Y
13	70yo; partial obstruction; curable cancer; ASA 1	96	4	0	N
14	70yo; complete obstruction; curable cancer; ASA 1	79	12	9	N
15	50yo; partial obstruction; curable cancer; ASA 1	96	4	0	N
16	50yo; complete obstruction; curable cancer; ASA 1	87	9	4	N

In the management of physiologically poor patients (ASA > 3) with complete bowel obstruction, SEMS is the preferred initial intervention of choice. This allows the patient to be physiologically optimised for subsequent interventions and also increases the chance of a one-stage resection. The morbidity of emergency surgery can be as high as 51% with an associated mortality rate of 16%[29]. With the greatest concern of colonic perforation being reported at 4% in previous trials, and modern day trials are quoting this at 0% with increasingly experienced interventionalists and safe methodology[6]. This low rate of perforation and the benefits of stenting with lower stoma formation rates, lower perioperative morbidity and quicker recovery/return to community should make SEMS a valid tool in the management of malignant complete bowel obstruction[23].

In the fit patient with curable disease, surgery is more often preferred as the intervention of choice due to the improved physiologic paramenters aiding in a better outcome and potential for one stage resection.

Considering the myriad of clinical scenarios and variables, the overall judgement, stenting technology offers an alternate tool to the clinician in the management of large bowel obstruction, with safe and effective outcomes.

The present body of evidence regarding stent insertion demonstrates its role, but to more clearly define its use in areas of uncertainty and community equipoise would require large multi-centre RCT’s. Such trials may be necessary, but will be hard to complete with the difficulties of recruiting patients to trials where treating clinicians still hold conservative views as to the merits of stent or surgery.