Question

Benefits of stenting or colostomy vs. acute resection with or without primary anastomosis in acute obstruction due to left-sided colon carcinoma.

Recommendation

Patients who present themselves with obstruction symptoms most likely caused based on an obstructing colon carcinoma are recommended a CT abdomen in order to determinate the extent of the primary tumor and potential metastases.

 

For a left-sided obstructive colorectal carcinoma in a patient without increased operative risk and curative intent an acute resection with end colostomy or acute resection with anastomosis and optionally ileostomy should be carried out.

 

For a left-sided obstructive colorectal carcinoma a colostomy, for example through a small transverse incision in the right upper abdomen, can be considered as a bridge to elective resection with an increased surgical risk.

 

The use of a stent as a bridge to elective resection should be considered at increased surgical risk, if it appears technically feasible and appropriate expertise is available, while the risk of perforation with possible oncological disadvantage should be taken into consideration.

 

The placement of a stent in a patient with an obstructive left-sided colon carcinoma and extensive metastases with a limited life expectation should be considered as palliation. If such a patient is still a candidate for systemic therapy a colostomy is preferred.

Conclusions

In patients with an obstruction due to left-sided colon carcinoma:

  • there is evidence of moderate quality that a staged operation less often results in a permanent colostomy in comparison with an acute resection.
  • there is evidence of moderate quality that peri-procedural mortality and cancer-specific mortality in the long term do not differ between patients who have undergone a staged resection or an acute resection.
  • there is evidence of moderate quality that as many peri-procedural complications occur in patients who undergo a staged resection as in those who undergo an acute resection.
  • there is evidence of poor quality that stenting less often results in a permanent colostomy than acute surgery does.
  • there is evidence of poor quality that stenting results in less peri-procedural morbidity than an acute operation, and if a period longer than the peri-procedural period is taken into account there is an equal amount of morbidity on comparison with an acute operation.
  • there is evidence of poor quality that stenting results in just as much peri-procedural and long-term mortality as an acute operation.
  • there is evidence of poor quality that stenting results in perforation in 3% of patients within 30 days of placement, and in 14% of patients in the longer term.

 

General quality of evidence on staged resection vs. acute resection = moderate

General quality of evidence on stenting vs. acute resection = poor

 

Literature summary

In a search for systematic reviews published from 2005 onwards, seven systematic reviews were identified (De Salvo 2004; Breitenstein 2007; Cirocchi 2012; Sagar 2011; Tan 2012; Watt 2007). One systematic review compared primary resection with staged resection (De Salvo 2004). The other six reviews compared stenting with acute resection (Breitenstein 2007; Cennamo 2012; Cirocchi 2012; Sagar 2011; Tan 2012; Watt 2007). These five reviews included a total of nine randomized controlled trials (Alcantara 2011; Cheung 2009; Fiori 2004; Ho 2012; Pirlet 2011; Xinopoulos 2004; Sankararajah 2005; van Hooft 2008; van Hooft 2011). However, inclusion varied from one trial to eight trials and the reasons for exclusion were often not reported. For this reason we decided not to describe the systematic reviews themselves, but the nine trials included in these reviews. In a search for randomized controlled studies from 2005 onwards, no new studies were found.

 

The systematic review of De Salvo and colleagues sought randomized and non-randomized controlled studies that compared primary resection with staged resection in patients with acute obstruction due to left-sided primary colon carcinoma (De Salvo 2004). One randomized study was found which was excluded by the authors because they did not regard the findings of this trial to be valid (Kronborg 1995). The reasons were that this trial

  1. did not do an a priori sample size calculation;
  2. did not describe standard treatment;
  3. had a very long inclusion period (15 years);
  4. gave no information about the number of patients excluded or the reasons for exclusion;
  5. 14% of patients included proved not to have carcinoma; and
  6. the long-term outcomes were apparently not adequately reported.

For purposes of this guideline it was decided to include this RCT, and to describe it including its weak points. Between 1978 and 1993 Kronborg. randomized 121 patients: in 58 patients a laparotomy was carried out and a colostomy created. A resection with an anastomosis was then carried out in the second instance. In the other 63 patients, resection with a colostomy was directly carried out, and an anastomosis was done later in the second instance. In seven of the 63 patients an anastomosis was created during the resection (breach of protocol). In 17 of the 121 patients the diagnosis of cancer had been made incorrectly (11 vs. 6 patients).

 

Nine controlled, randomized trials compared stenting with surgery in patients with an obstruction due to left-sided colon carcinoma. In addition there is the CReST trial which is still running and which is intending to randomize 400 patients between stenting and acute resection. One of the nine trials compared stenting followed by resection with acute resection with a primary anastomosis (Alcantara 2011). This trial was prematurely ended due to high perioperative morbidity in the acute resection group. In addition, six of the nine trials compared stenting followed by elective surgery with acute surgery (Cheung 2009; Ho 2012; Pirlet 2011; Sankararajah 2005; van Hooft 2008; van Hooft 2011). There are a number of interesting points concerning these six trials:

  • Cheung compared stenting followed by laparoscopic resection with acute open resection (Cheung 2009). One quarter of the 48 patients (12 patients) had a stage IV tumor, nine of these patients were in the open surgery group and three in the stenting group (p = 0.02).
  • Ho compared stenting followed by elective laparoscopic or open resection with acute resection (Ho 2012). One quarter of the patients had a stage IV carcinoma; three were in the stenting group and seven in the acute resection group - a difference that was not statistically significant.
  • The trial of Pirlet was prematurely discontinued as during stent placement two perforations occurred (Pirlet 2011).
  • The trial of Sankararajah was an interim analysis, and was described only in an abstract (Sankararajah 2005).
  • The trial of van Hooft in 2008 randomized 21 patients with a stage IV colon carcinoma and a threatening obstruction between stenting and palliative surgery (resection or creation of colostomy) (van Hooft 2008). This trial was stopped prematurely due to the large number of complications in the stenting group.
  • In 2011 van Hooft randomized 98 patients between stenting followed by elective surgery or acute surgery (van Hooft 2011). This trial was prematurely discontinued due to high morbidity in the stent group (which was not statistically significant) during two interim analyses.

 

The last two of the nine trials evaluated palliative stenting compared with the elective creation of a colostomy in patients with threatening obstruction (Fiori 2004; Fiori 2012; Xinopoulos 2011). Originally Fiori wanted to include 30 patients, but after two years there had been no complications associated with stenting so it was decided to discontinue recruitment as stenting apparently involved no more complications that the elective creation of a colostomy. The trial of Xinopoulos randomized 30 patients with a partial obstruction of the colon due to inoperable malignancy (24 patients with colon carcinoma and 6 patients with cervical carcinoma) between stenting or the creation of a stoma (Xinopoulos 2004).

 

Quality of the evidence

Two of the outcomes (colostomy and peri-procedural mortality) of van Kronborg, who compared staged resection with acute resection, were devalued due to the small number of events which can lead to fragile results. The outcome cancer-specific mortality in the long term was devalued because the data were not given. The outcome peri-procedural morbidity was devalued due to the high risk of bias. All outcomes were of moderate quality only.

 

As five of the nine trails that compared stenting with acute surgery were prematurely ended for diverse - and sometimes contradictory – reasons, all outcomes were devalued. The outcomes morbidity and permanent colostomy were also devalued due to a high risk of bias meaning that the quality of the evidence transpired to be low. The outcome mortality was also devalued due to low mortality and small trial size, meaning that the results were not robust.

 

Desirable effects

Effect on colostomy: staged resection vs. acute resection

After a follow-up period of between 4 months and 15 years, a permanent colostomy was seen less often in patients who had undergone staged resection for cancer and survived (3/35 vs. 14/50, p=0.05).

 

Effect on colostomy: stenting vs. acute surgery

A permanent colostomy (not statistically significant) was seen less often in the stenting groups (Odds ratio (OR) 0.66, 95% CI, 0.36-1.22 – not including trials that compared stenting with elective colostomy).

 

Effect on survival: staged resection vs. acute resection (crucial outcome)

In the trial of Kronborg et al. the postoperative mortality was 11 vs.8 (after two laparotomies in both groups). Concerning cancer-specific survival in the long term (follow-up four months to 15 years), no significant difference was seen (data not reported).

 

Effect on survival: stenting vs. acute surgery (crucial outcome)

Table 1 gives an overview of peri-procedural mortality (in which the mortality resulting from elective surgery was included in the stenting groups). In a meta-analysis of peri-procedural mortality in the nine trials, no difference was seen between the two treatment groups (OR 1.09, 95% CI, 0.43-2.72). No meta-analysis of long-term mortality was carried out, as this was reported differently in the six trials that reported it. None of the individual trials showed any differences in mortality in the longer term (Table 2).

 

Undesirable effects

Effect on morbidity: staged resection vs. acute resection (crucial outcome)

The trial of Kronborg et al. showed no difference between these two groups concerning the number of patients who had one or more complication/s (31 vs. 42, p=0.19).

 

Effect on morbidity: stenting vs. surgery (crucial outcome)

Table 1 gives an overview of the peri-procedural morbidity (whereby morbidity associated with elective surgery is included in the stenting groups), as described in the nine trials that compared stenting with or without elective surgery with acute surgery. In a meta-analysis of eight of the nine trials, less peri-procedural morbidity was seen in the stenting groups (OR 0.45, 95% CI, 0.23-0.90). Four of the trials described long-term morbidity and five trials only described peri-procedural morbidity (Table 2). In a meta-analysis of all data on morbidity available during the longest follow-up period, no more differences between patients treated by stenting and patients treated by surgery were seen (OR 1.0, 95% CI, 0.4- 2.7). The two trials that went on to look at the peri-procedural period observed a great deal of morbidity in the stent groups:

  • van Hooft 2008: nine of the 11 patients who were randomized to the stent group had a complication; six of these patients had a perforation (four during the peri-procedural period).
  • Xinopoulos 2011: in six of the 15 patients who were randomized to the stenting groups there was ingrowth of the tumor into the stent and in five of these patients it was not possible to free the lumen with a laser; in one patient the stent migrated and a colostomy had to be created. Perforation did not occur.

 

Perforations by the stent ≤30 days after placement occurred in two of the seven trials that described this: two patients in one trial (7%) and two patients in the other (18%) (Pirlet 2011; van Hooft 2008), i.e. in four of the 126 stents positioned (3.2%). Two of the four trials that went on past the peri-procedural period reported perforation by the stent in the long-term in six (55%) and six (13%) of the 84 (14%) patients, respectively (van Hooft 2008; van Hooft 2011). Also, one trial found three occult perforations in the resected specimen. 

Considerations

No considerations are described.

Authorization date and validity

Last review : 16-04-2014

Last authorization : 16-04-2014

The validity of this guideline and its associated modules is five years. For various reasons, it may be necessary to edit modules sooner than intended. The National Working Group on Gastrointestinal Cancers therefore annually assesses the content of the guideline and its associated modules. By 2016 it is decided whether a new multidisciplinary working group should be installed to revise the entire guideline.

Initiative and authorization

Initiative : Nederlandse Vereniging voor Radiotherapie en Oncologie

Authorized by:
  • Nederlandse Internisten Vereniging
  • Nederlandse Vereniging van Maag-Darm-Leverartsen
  • Nederlandse Vereniging voor Heelkunde
  • Nederlandse Vereniging voor Klinische Geriatrie
  • Nederlandse Vereniging voor Nucleaire geneeskunde
  • Nederlandse Vereniging voor Pathologie
  • Nederlandse Vereniging voor Radiologie
  • Nederlandse Vereniging voor Radiotherapie en Oncologie
  • Vereniging Klinische Genetica Nederland

General details

All members were mandated by a scientific, professional or patient association. In the composition of the working group we tried to take national distribution, input from participants from both academic and general hospitals and representatives of various disciplines into account. Patients are also represented by delegation into the working group, as well as a focus group meeting.

Scope and target group

Goal

This guideline and its associated modules are - as much as possible - based on scientific research and / or consensus. It is a document with recommendations to support the daily practice of health care professionals involved in patients with (possible) colon cancer, rectal cancer or colorectal liver or lung metastases. It provides recommendations for diagnosis, treatment, follow-up and organization of care. The guideline and its associated modules are thus seeking to improve the quality of care, to increase transparency of choice for treatment and reduce diversity.

 

Target population

Each year colorectal cancer is identified in approximately 13,000 new patients. Rectal carcinoma occurs in about 1 in 3 patients of this. In the Netherlands, the colorectal cancer in both men and women rank third place in incidence of oncological diseases. The expected number of patients diagnosed with colorectal cancer are increased in 2020 to about 17,000, reflecting a slight increase in incidence (especially in men), population growth and aging.

Colorectal cancer is slightly more common in men than in women and ninety percent of patients 55 years or older. More information about the Dutch population can be found at the Netherlands Cancer Registry: www.cijfersoverkanker.nl

This guideline is applicable to all adult patients with (suspected) a primary colorectal carcinoma and patients with metastatic disease. Particular attention is given to the elderly. A separate guideline is available for adult patients with an increased risk of hereditary colon cancer.

Target Audience

This guideline and its associated modules are intended for all professionals involved in the diagnosis, treatment and rehabilitation of patients with (metastatic) colorectal cancer, such as surgeons, general practitioners, consultants, internists, gastroenterologists, (specialist) nurses, clinical geneticists, paramedics, pathologists, radiologists and radiotherapists. The complete guideline is used to develop a patient education text from the Dutch patients Consumer Federation (NPCF).

 

Members of the guideline panel

Name

Function

Hospital

Mandated

Mw. prof. dr. C.A.M. Marijnen chair

Radiotherapist

LUMC Leiden

NVRO

Mw. prof. dr. R.G.H. Beets-Tan

Radiologist

MUMC Maastricht

NVVR

Mw. S. de Bruijn

Specialist nurse

Renier de Graaf hospital Delft

V&VN

Mw. dr. A. Cats

Gastroenterologist

NKI-AVL Amsterdam

NVMDL

Prof. dr. E.F.I. Comans

Nuclear doctor

VUMC Amsterdam

NVNG

Dr. A.R. van Erkel

Intervention radiologist

LUMC Leiden

NVVR

Mw. dr. M.A.M. Frasa

Clinical chemist

Groene Hart hospitalGouda

NVKC

Mw. C. Gielen

Specialist nurse

MUMC Maastricht

V&VN

Dr. E.J.R. de Graaf

Surgeon

IJsselland hospital Capelle a/d IJssel

NVVH

Mw. dr. M. Hamaker

Geriatrician

Diakonessenhuis Utrecht

NVKG

Mw. dr. J.E. van Hooft

Gastroenterologist

AMC Amsterdam

NVMDL

Mw. H.J.A.M.. Kunneman

Researcher

LUMC Leiden

n.v.t.

Mw. dr. M.E. van Leerdam

Gastroenterologist

NKI-AVL Amsterdam

NVMDL

Dr. H. Martijn

Radiotherapist

Catharina-hospital Eindhoven

NVRO

Mw. dr. A.M. Mendez

Romero

Radiotherapist

EMC Cancer Insititute Rotterdam

NVRO

Mw. prof. dr. I.D. Nagtegaal

Pathologist

UMCN St Radboud Nijmegen

NVVP

Dr. L.A. Noorduyn

Pathologist

Lab. Voor Pathologie Dordrecht e.o.

NVVP

Mw. A. Ormeling

Patient

Stomavereniging

NFK

Drs. T.A.M. van Os

Klinisch Geneticus

AMC Amsterdam

VKGN

Dr. F.T.M. Peters

Gastroenterologist

UMCG Groningen

NVMDL

Mw. J. Pon

Patient

NFK/SPKS

NFK

Mw. dr. J.E.A. Portielje

Medical oncologist

Haga hospital Den Haag

Gerionne

Prof. dr. C.J.A. Punt

Medical oncologist

AMC Amsterdam

NIV

Mw. dr. H. Rütten

Radiotherapist

UMCN Radboud Nijmegen

NVRO

Prof. dr. H.J.T. Rutten

Surgeon

Catharina-hospitalEindhoven

NVVH

Prof. dr. J. Stoker

Radiologist

AMC Amsterdam

NVVR

Dr. P.J. Tanis

Surgeon

AMC Amsterdam

NVVH

Dr. J.H. von der Thüsen

Pathologist

MC Haaglanden Den Haag

NVVP

Prof. dr. H.M.W. Verheul

Medical oncologist

VUMC Amsterdam

NIV

Prof. dr. C. Verhoef

Surgeon

EMC Cancer Insititute Rotterdam

NVVH

Dr. Tj. Wiersma

General practitioner

NHG

NHG

 

 

Name

Function

Location

Mw. drs. A.Y. Steutel

process manager

Utrecht

Drs. T. van Vegchel

process manager

Amsterdam

Mw. S. Janssen-van Dijk

secretary

Rotterdam

M.P.  van den Berg

Researcher

Bilthoven

P.F. van Gils

Researcher

Bilthoven

Mw. J. Robays

Methodologist

Brussels

Mw. drs. Y Smit

Methodologist

Germany

Mw. A. Suijkerbuijk

Researcher

Bilthoven

Mw. dr. L. Veerbeek

Methodologist

Groningen

Mw. dr. L. Verheye

Methodologist

Brussels

Mw. dr. G.A. de Wit

Researcher

Bilthoven

 

Patient involvement

Two patient experts have been part of the guideline development group. One on behalf of the Dutch Ostomy Association and on behalf of SPKS/NFK. Based on a focus group meeting experiences of patients regarding care were collected. The guideline also has been used to develop a patient education text for the Dutch patients Consumer Federation (NPCF).

Method of development

Evidence based

Implementation

Promoting the use of recommendations begins with a broad (digital) distribution of the guideline, using direct mailing and an article published in the Dutch Journal of Oncology. In other journals or training sessions, for example, the guideline is brought to the attention. An implementation plan for this guideline contains the key recommendations and an overview of barriers and facilitators for implementation.

Methods and proces

The working group met in July 2012 for the first time. Based on an initial list of problems by working group members a survey among professionals involved in patients with colorectal carcinoma was held. Through this survey, sixty professionals supplied and prioritized possible subjects for revision. Also, a focus group meeting was held, collection patients' experiences. The eleven most relevant questions were answered, and translated into English. Also, alle recommendations were translated into English.

 

Each clinical question was andwered by a subgroup within the development group. External methodologists provided the literature search, review, critical assessment, evidence tables and a draft literature review. Workgroup members suggested other considerations and recommendations.

Responsibility

The Comprehensive Cancer Organisation the Netherlands (IKNL) promotes that people with cancer and their families have access to a consistent and qualitatively adequate care as close to home as possible. IKNL was established to improve treatment, care and clinical research in oncology. It also has a role in setting up and supporting networks for palliative care. IKNL supports multidisciplinary guideline development for oncology and palliative care and facilitates the maintenance, management, implementation and evaluation of these guidelines.

AGREE was used to check the methodological quality of the guideline.

Search strategy

Searches are available upon request. Please contact the Richtlijnendatabase.