Guideline 67: Management of colorectal cancer

Management of colorectal cancer
Section 7: Surgery

Surgery remains the definitive treatment for apparently localised colorectal cancer, offering the only chance of cure. In this section, recommendations for preparation for surgery are followed by technical considerations and recommendations for surgery for the emergency situation and for advanced disease. Specialisation and workload are also addressed.

7.1 Preoperative staging

Given the very different therapeutic strategies required for operable localised disease and locally advanced or disseminated disease, preoperative staging is indicated in most instances. Preoperative liver and chest imaging to detect metastases is advisable unless the patient’s management would not be altered by the findings.¹⁰⁹ For liver metastases, preoperative assessment with CT or magnetic resonance imaging is more sensitive than with transabdominal ultrasound, although the most accurate modality appears to be a combination of intraoperative ultrasound and palpation at the time of surgery.¹¹⁰ The accuracy of staging investigations for primary rectal cancer is improving, and the current body of evidence supports MRI (for more advanced tumours) and endorectal ultrasound (for early lesions) as the best modalities.^{111,
112} Evidence level 1+,2+,4

Another important aspect of preoperative staging is complete visualisation of the large bowel. Synchronous cancers occur in 5% of cases, and these may not be readily detectable at surgery.¹¹³ When a cancer has been diagnosed, a complete colonoscopy or barium enema should be carried out before surgery wherever possible. When this is impossible owing to obstruction or other emergency presentation, it should be performed within three months of resection. Evidence level 2+

All patients undergoing elective surgery for colorectal cancer should have preoperative imaging of the liver and chest.
In patients requiring emergency surgery intraoperative liver ultrasound or postoperative imaging is acceptable.

Intraoperative ultrasound is appropriate if a preoperative diagnosis of liver metastases would not alter the need for operative intervention. Preoperative imaging of primary rectal cancer may clarify operability and aid decisions regarding chemotherapy or radiotherapy delivered preoperatively (neoadjuvant chemo-irradiation).

Complete colonic examination by colonoscopy, CT pneumocolon or barium enema should be carried out, ideally preoperatively, in patients with colorectal cancer.

7.2 Preoperative preparation

Patients undergoing surgery for colorectal cancer are at risk of both venous thromboembolism and wound infection. It is therefore recommended that prophylactic measures are taken as outlined in the appropriate SIGN guidelines.^{114,
115} The use of antibiotic prophylaxis in colorectal surgery is further supported by a recent meta-analysis.¹¹⁶ Mechanical bowel preparation is also widely employed, but current evidence, consisting of three underpowered randomised trials does not support its routine use.^{117,
118, 119} All patients who will or might require a permanent or temporary stoma should be seen preoperatively by a stoma nurse specialist if possible (see section 3.5). Evidence level 4,1++,1-

Patients undergoing surgery for colorectal cancer should have:

venous thomboembolism prophylaxis,
antibiotic prophylaxis consisting of a single dose of antibiotics providing both aerobic and anaerobic cover given within 30 minutes of induction of anaesthesia.

Although there is no evidence that bowel preparation confers benefit, the quality of evidence suggesting no effect is too weak to make a definitive statement that it is not necessary.

The decision to use bowel preparation must be individualised according to the patient’s need and the surgeon’s experience.

7.3 Perioperative blood transfusion

Concern has been raised over the potential for increased risk of cancer recurrence following perioperative blood transfusion.¹²⁰ A meta-analysis of three randomised and two cohort studies where control groups received either leucodepleted or autologous blood transfusion found no significant difference in cancer recurrence. Due to the small number of patients taking part in the trials, the meta-analysis was insufficiently powered to detect a difference of less than 20% in risk. The inability of these studies to exclude a small effect is of less significance now that leucodepletion of blood for transfusion is universal in the UK.¹²¹ Evidence level 1+

If a patient undergoing colorectal cancer surgery is deemed to require a blood transfusion, this should not be withheld on account of a possible association with increased risk of cancer recurrence.

7.4 techniques in colorectal cancer surgery

7.4.1 RECTAL CANCER

There is now evidence from large cohort studies using historical controls that the use of total mesorectal excision (TME) reduces the risk of local recurrence after rectal cancer surgery, and improves survival.^{122,
123, 124} This appears to be due to good circumferential clearance of tumour. It is unlikely that tumours of the upper rectum will benefit from total excision of the mesorectum, as long as the principles of careful dissection in the plane immediately outside the mesorectum are applied.¹²⁵ The low anastomosis necessitated by TME results in poorer functional results than a higher anastomosis, and should be avoided unless doing so would compromise adequate mesorectal exision.¹²⁶ It is also important to preserve the autonomic nerves in the pelvis to minimise bladder and sexual dysfunction.¹²⁷ Evidence level 2++,2+,4

Mesorectal excision is recommended for most rectal cancers where the patient is fit for radical surgery. The mesorectal excision should be total for tumours of the middle and lower thirds of the rectum, and care should be taken to preserve the pelvic autonomic nerves wherever this is possible without compromising tumour clearance.

7.4.2 COLON CANCER

In contrast to rectal cancer surgery, there is little evidence relating to the radicality of colon cancer surgery. Two underpowered randomised controlled trials were unable to demonstrate a beneficial effect of “no touch” technique¹²⁸ or formal left hemicolectomy¹²⁹ respectively. There is no evidence that the radicality of excision has an effect on outcomes in patients with colon cancer. Evidence level 1-

Where a resectable organ, (eg kidney, ureter, duodenum, liver, stomach, bladder, uterus or vagina) is involved by the primary tumour, careful consideration should be given to removal (partial or total as appropriate) of that organ.

7.4.3 ANASTOMOSES

Anastomotic leakage is an important and potentially fatal complication of colorectal cancer surgery, and measures to minimise it should be taken. There is no high quality evidence to support any specific technique, but a recent meta-analysis indicated that the only difference between hand-sewn and stapled anastomoses is a slightly increased risk of anastomotic stricture with stapling.¹³⁰ Evidence level 1++

Risk factors for anastomotic dehiscence are well known and include male sex, increasing age and obesity, but in anterior resection leakage is increased with a low (<5 cm from anorectal junction) anastomosis.¹³¹ Evidence level 2++

It has also been shown that a defunctioning stoma reduces the risk of a clinically evident leak in low colorectal anastomoses.¹³² Another disadvantage of the low anastomosis is poor function, and there is good evidence from randomised trials to support the use of a colopouch in this situation.^{133,
134, 135} Evidence level 2+,1+

With a low rectal anastomosis, consider giving a defunctioning stoma.

With a low rectal anastomosis after TME, consider a colopouch.

Not all patients will benefit from a low rectal anastomsis, and if the patient is deemed to be at unacceptable risk of anastomotic breakdown or poor function, then a permanent colostomy (low Hartmann’s procedure) should be employed.

7.5 Local excision of colorectal cancers

Certain rectal cancers are technically amenable to local excision, and there is evidence from a randomised trial that this is associated with less morbidity than radical surgery.¹³⁶ There is also non-randomised evidence that local excision is associated with higher rates of local recurrence than radical surgery, presumably owing to residual tumour in lymph nodes.¹³⁷ Evidence level 1+,4

Adjuvant radiotherapy and chemotherapy may reduce local recurrence rates, but a reliable and widely accepted regimen has not yet been developed.¹³⁷ T1 tumours (those with the smallest local spread) are often deemed suitable for local excision, but it must be stressed that extensive involvement of the submucosa is associated with a 17% rate of lymph node involvement. Minimal involvement of the submucosa (T1 sm1 tumours) appears to be associated with minimal risk of lymph node involvement.¹³⁸ Colon (and some rectal) cancers may be excised by polypectomy at colonoscopy (polyp cancers), and cohort studies indicate that such lesions do not require further surgery unless there is histopathological evidence of tumour at the margin (incomplete excision), lymphovascular invasion or the invasive tumour is poorly differentiated.^{139,
140} Evidence level 2+,4

Currently, it is not possible to identify a subgroup of rectal cancer patients in whom regional lymph node involvement can be comprehensively excluded thus allowing unreserved recommendation for local excision, although T1 sm1 tumours may be suitable.

The relative risks of operative morbidity and recurrence must be carefully weighed and explained fully to the patient so that an informed decision can be made regarding local excision and rectal cancer.

Further surgery for pedunculated polyp cancers is indicated if:

there is histological evidence of tumour at, or within 1 mm of, the resection margin;
there is lymphovascular invasion;
the invasive tumour is poorly differentiated.

7.6 Laparoscopic surgery for colorectal cancer.

Evidence from several randomised controlled trials, case control studies, and cohorts indicates that laparoscopic surgery for colorectal cancer is feasible, and can reduce postoperative pain, analgesia use, hospital stay and blood loss in the short term.¹⁴¹ Most of the randomised trials have the potential for bias, as blinding is impractical. Reliable outcomes are also lacking.

Laparoscopic surgery can be considered for colorectal cancer.

7.7 Management of malignant colonic obstruction

When a mechanical large bowel obstruction is suspected, a water-soluble contrast enema can confirm this and avoid operative intervention for pseudo-obstruction.¹⁴² Evidence level 2+

Mechanical large bowel obstruction should be distinguished from pseudo-obstruction before surgery.

There is evidence that, in suitable patients, and with sufficient surgical expertise, removal of the tumour at the first operation is feasible.¹⁴³ If primary resection is carried out, immediate anastomosis is feasible, again given a suitable patient and appropriate surgical expertise, and there is randomised evidence that segmental resection is preferable to subtotal colectomy in terms of functional outcome.¹⁴⁴ Evidence level 1++,2+

Colonic stenting is also possible, and can provide both palliation in patients with inoperable disease and relief of obstruction in those with operable disease prior to semi-elective resection.^{145,
146} Evidence level 3

Patients with malignant obstruction of the large bowel should be considered for immediate resection.

If immediate reconstruction after resection is deemed feasible, segmental resection is preferred for left-sided lesions.

Where facilities and expertise are available, colonic stenting should be considered.

7.8 Surgery for advanced disease

There is evidence from cohorts with historical controls that survival can be improved by hepatic resection for technically suitable metastatic disease,¹⁴⁷ and the same may be true of lung resection.¹⁴⁸ In situ ablation for liver metastases which are not suitable for resection is also feasible, but the benefit is less clear.¹⁴⁹ Evidence level 3,4

In the patient with locally advanced primary or recurrent disease, it must be remembered that surgical removal offers the only chance of cure, but that quality of life may be adversely affected by inappropriate attempts at resection.¹⁵⁰ For disease that is clearly inoperable, interventions such as stenting or laser ablation may provide useful palliation.¹⁵¹ Evidence level 4

Patients with liver and lung metastases should be considered for resection or, in the case of liver disease, in situ ablation.

In patients with advanced local or recurrent disease, careful consideration should to surgical excision or palliative intraluminal procedures.

7.9 Specialisation and work load in colorectal cancer surgery

There is evidence from cohorts and historical controls that morbidity and survival are affected by surgeon and hospital workload but the evidence is insufficient to recommend a specific yearly volume.¹⁵² Evidence from North America, where specific colorectal accreditation is available, indicates better outcomes from specialists,¹⁵³ and evidence from Europe convincingly demonstrates better outcomes after specialist training in rectal cancer surgery.¹²⁴ Evidence level 2++

Surgery for colorectal cancer should only be carried out by appropriately trained surgeons whose work is audited. Low rectal cancer surgery should only be performed by those trained to carry out TME.