|Home | About | Journals | Submit | Contact Us | Français|
Colorectal cancer is the second most common malignancy in the western world. In the UK 32 000 new cases are diagnosed annually, of which approximately 45% occur in the rectum and rectosigmoid. Although the management of colonic malignancies has not altered greatly in recent years this is not true of rectal cancer: advances in surgical technique, molecular biology, radiological imaging and adjuvant therapy have dramatically altered the way patients are treated. This review gives a brief overview of these changes.
Since 1932 when Sir Cuthbert Dukes described his system for staging cancers of the rectum, attempts have been made to identify other factors that affect outcome in the hope that management can be tailored accordingly. Until recently, little progress had been made, but the advent of new genetic and molecular pathology techniques including DNA chip technology, in which numerous DNA sequences, can be analysed simultaneously, yielded much new information on the biology of malignancy.
Germline mutations responsible for inherited conditions that predispose to the development of colorectal cancers, such as the APC gene in familial adenomatous polyposis coli (FAP) and the mismatch repair genes in hereditary non-polyposis colon cancer (HNPCC), have been known for some time, allowing mutation carriers to be identified; these individuals are either enrolled into a surveillance programme or offered prophylactic surgery before malignancy develops. But FAP and HNPCC account for only about 5% of all colorectal cancers, and the identification of genetic polymorphisms that predispose to the development of colorectal cancer may be more important in the population as a whole. Polymorphisms are normally functioning genes that contain a region slightly different in sequence from that found in the majority of the population. A polymorphism is commonly identified in the APC gene of Ashkenazi Jews and increases the risk of colorectal cancer by 1.5-1.7 times1. The identification of other similar polymorphisms may be hastened by DNA chip technology.
Molecular techniques have been employed in the diagnosis of colorectal cancer, and mutated copies of the APC, K-ras and P53 genes have been detected in the stools of patients with malignancies as have DNA sequences demonstrating microsatellite instability2. To date this remains a research tool but has the potential to be employed for population screening.
The ability to predict the response of a tumour to chemotherapy or radiotherapy may also be governed by its molecular pathology. Rectal cancers expressing the mutant P53 protein tend to be resistant to radiotherapy, whilst those expressing the p21 gene seem more sensitive to its effects3. Cancers that express high levels of the enzyme thymidylate synthase show resistance to 5-fluorouracil4 whilst those demonstrating microsatellite instability seem to respond well to 5-fluorouracil5. The prognostic value of identifying different genetic mutations has been widely studied but its importance remains controversial. Cancers with mutations of the K-ras, DCC and p53 genes tend to have a poor prognosis6,7,8. Microsatellite instability or reduced expression of COX-29,10 correlate with better survival. In the future, a knowledge of the underlying genetic and molecular abnormalities in each cancer should allow treatment to be targeted specifically.
Preoperative evaluation of rectal cancer has become increasingly important, to identify patients with locally advanced disease who may benefit from neoadjuvant chemoradiotherapy and those suitable for local resection. Endoscopic rectal ultrasound (ERUS) accurately indicates the extent of tumour spread through the wall of the rectum in over 90% of cases (T stage) and the involvement of perirectal lymph nodes in up to 80% (N stage)11,12. It is an essential investigation for those being considered for local resection. Currently, examination by ERUS is limited to tumours within the distal rectum but the recent development of colonoscopic ultrasound may prove useful to evaluate more proximal disease. However, CT and MRI define the extent of local spread, tissue infiltration and lymph node involvement better than ERUS. The detection of recurrent disease by CT or MRI is highly effective for hepatic and extrahepatic disease, but changes in tissue appearances on CT scans following irradiation are similar to those seen with recurrent tumour and can lead to diagnostic uncertainty, particularly within the pelvis. The use of positron emission tomography with fluorine-18-labelled deoxyglucose (a glucose substitute preferentially taken up by cancer cells) is more sensitive than CT scanning for the detection of both hepatic and extrahepatic disease including pelvic recurrence13,14 and even identifies previously unsuspected metastasis.
In the past, cancers of the distal rectum were treated by abdominoperineal excision of the rectum, but recently ‘low’ cancers have been treated increasingly by anterior resection, which avoids the need for a permanent stoma. The old concept that the distal resection margin must be at least 5 cm clear of the cancer has been discredited. Unless the tumour is poorly differentiated a 2 cm clearance margin below the tumour edge is adequate since distal intramural spread seldom exceeds a few millimetres and distal lymphatic spread is rare. Tumours with intramural spread extending > 1 cm beyond the tumour are almost always advanced, with distant metastasis15,16. This change in approach, and the availability of stapling devices, allows restorative resection for well to moderately differentiated cancers only 2-3 cm above the anal sphincters without compromising oncological principles. While low colorectal or coloanal anastomoses avoid the need for a permanent stoma this method does result in increased frequency of bowel action, faecal urgency and incontinence (the anterior resection syndrome). The fashioning of a short 5-7 cm colon J pouch (similar in principle to the ileal pouch used in ulcerative colitis) improves the functional outcome of a low anastomosis without added morbidity17. Alternatively a coloplasty pouch, created by making an 8-10 cm longitudinal colostomy 4-6 cm above the anastomosis and closed transversely with two layers of sutures, has been advocated to improve the functional outcome following a low anastomosis18.
A major advance in surgical technique has been attention to cancer clearance. The mesorectum consists of fatty tissue surrounding the rectum and contains the lymphatics that drain the rectum. Isolated deposits of tumour can be found within the mesorectum up to 3 cm distal to the main tumour19. Total mesorectal excision has reduced local recurrence following surgery to less than 10%20,21 although anastomotic dehiscence rates as high as 19% have been reported22. Dehiscence is related to devascularization of the rectal stump, and a temporary defunctioning stoma (usually a loop ileostomy) is frequently employed for low resections to minimize the morbidity caused by a leak.
Failure to remove tumour that has extended into the pararectal tissue gives rise to high rates of local recurrence (78%), and clearance at the circumferential resection margins is another crucial component of rectal excision23. Pelvic lymphadenectomy, practised in Japan, offers no clear survival advantage over total mesorectal excision: it results in a high incidence of bladder and sexual dysfunction and has not been widely advocated in the West24.
In carefully selected patients, particularly the elderly or those with severe co-morbidity, local excision is a reasonable option. Tumours less than 10 cm from the dentate line may be resected under direct vision whilst more proximal lesions can be treated by transanal endoscopic microsurgery. The microsurgery is performed via a rectoscope in a manner similar to laparoscopic surgery, with the rectum insufflated with carbon dioxide. The lesions should be mobile on digital examination, less than 5 cm in diameter and of favourable differentiation. Endorectal ultrasound and CT scanning should be available to determine the depth of bowel wall involvement and local lymph node status. Local resection should be limited to T1 and T2 tumours. Recurrence rates following local excision are related to the tumour stage: 9.7% for T1, 25% for T2 and 38% for T3 tumours; with postoperative chemoradiotherapy the local recurrence rates are reduced to 9.5% for T1, 13.6% for T2 and 13.8% for T325. The role of prospective chemoradiotherapy has yet to be fully evaluated. Complication rates tend to be low but include local sepsis, rectovaginal fistula, faecal incontinence and urinary retention and infections. When pathological criteria are unfavourable because of poor differentiation or inadequate clearance, adjuvant chemoradiotherapy is essential and formal resection should be considered. Salvage surgery for local recurrence can be life-saving if performed early26.
The place of laparoscopic surgery in colorectal cancer is dubious because of concerns about intraperitoneal spread of malignant cells associated with pneumoperitoneum, malignant recurrence at port sites and the adequacy of resection. Results of studies have been inconsistent and large-scale trials are now underway.
The sentinel lymph node is the first node that receives lymph from a primary tumour and is thought to be the node most likely to contain metastatic cells. The sentinel node in rectal cancer can be accurately identified in up to 99%27 of cases. Theoretically examination of this node, possibly with molecular techniques, could improve the detection of micro-metastasis but experience with breast cancer and melanoma indicates that this is unlikely to have an important bearing on surgical technique for cancer clearance.
The prognosis for patients with hepatic metastasis is bleak. The median survival is 6-9 months and may increase to 12 months with 5-fluorouracil chemotherapy28. Although newer agents such as irinotecan and oxaliplatin may improve survival slightly, chemotherapy remains a palliative treatment.
Surgical resection offers the only hope of a cure although the proportion of patients with resectable disease is 10%29. The presence of diffuse disease or tumour that invades vital structures such as the porta hepatis is a contraindication to surgery.
With improved understanding of hepatic anatomy and better surgical and anaesthetic techniques, perioperative mortality is now as low as 2%30,31. The liver consists of eight segments each with its own bile duct, portal vein and hepatic blood supply; thus each segment can be resected individually. Blood loss is reduced by the use of ultrasound parenchymal dissectors, argon diathermy and controlled anaesthesia in which the central venous pressure is lowered; sometimes blood transfusion can be avoided.
The 5-year survival rates after liver resection for metastatic colorectal cancer are around 30-50%, depending upon selection criteria32,33,34. Resection was previously limited to less than four metastases in a single lobe, but it is now apparent that the number of metastases is less important than the ability to excise them completely with a 1 cm margin of clearance. Ideally two disease-free segments, preferably in continuity, should be preserved. Other factors that influence prognosis after hepatic resection are the stage of the primary tumour, the disease-free interval before hepatic metastasis is recognized and the level of carcinoembryonic antigen.
Adjuvant chemotherapy in combination with liver resection seems to improve the results. Hepatic arterial infusion together with systemic chemotherapy gave better 2-year survival than surgical resection alone34. Systemic chemotherapy reduces the volume of metastatic disease, rendering inoperable tumours amenable to surgery in 16% of cases with subsequent 5-year survival comparable with that of patients undergoing a standard resection29. Currently a European Organization for Research and Treatment of Cancer (EORTC) trial is evaluating the role of preoperative and postoperative chemotherapy in patients with resectable hepatic metastases.
Newer modalities such as cryotherapy and radio-frequency ablation may be of value. Cryotherapy, in which cancer tissue is destroyed by repeated cycles of freezing and thawing, can be used to treat tumours of up to 8 cm in size including those located near major structures; however, the drawbacks include major haemorrhage, systemic responses to thermal injury and residual necrotic tissue. Radio-frequency ablation, in which radiofrequency radiation is used to produce heat in a local area of tissue, yields fewer complications than cryotherapy although the volume of tumour that can be destroyed is limited. Both techniques are capable of destroying cancer tissue but their impact on long-term survival remains controversial35.
In addition to their standard adjuvant role in Dukes stage C disease, chemotherapy and radiotherapy have been used in rectal cancer to reduce local recurrence, improve operability and allow anal sphincter preservation. The addition of preoperative radiotherapy can reduce the rate of local recurrence after surgery from 27% to 11% and improves long-term survival from 48% to 58%36. When used in combination with total mesorectal excision, preoperative radiotherapy may reduce the incidence of local recurrence to as little as 2%37. Preoperative radiotherapy is safe: although it tends to delay perineal wound healing and causes sexual dysfunction, it does not predispose to anastomotic dehiscence. Postoperative radiotherapy also reduces local recurrence, although it does not seem to improve long-term survival and is associated with more complications, particularly small-bowel obstruction and late-onset diarrhoea due to radiation enteritis38. An advantage of postoperative radiotherapy is that it is can be reserved for patients with unfavourable operative or pathological findings.
Combined chemotherapy and radiotherapy (chemoradiotherapy) is being used more widely in both the preoperative and postoperative management of rectal cancer. Cancer cells are sensitized to the effect of radiotherapy by 5-fluorouracil, and combined therapy achieves a greater reduction in local recurrence and better survival than seen with radiotherapy alone39. The US National Institutes of Health consensus recommends chemoradiotherapy for all stage II and III rectal cancers40. Preoperative chemoradiotherapy is particularly indicated for locally advanced or inoperable rectal cancers (T3 or T4). Although in the various studies selection criteria and treatment regimens have differed widely, over 70% of unresectable and locally advanced rectal cancers are ‘downstaged’, allowing excision41, and pathologists are commonly unable to demonstrate residual cancer tissue after chemoradiation. Patients with rectal cancers that respond to preoperative chemoradiotherapy have better 5-year survival, with lower local recurrence rates, than those in whom the tumour does not respond42.
This review is based on a joint meeting at which the principal speakers were Dr Rob Glynne Jones (Neoadjuvant chemoirradiation for T3, T4 rectal cancer), Dr Harpreet Wassan (Molecular predictors of treatment response), Mr Paul Finan (Surgical technique and outcome of rectal cancer), Dr Tim Maugham (Chemotherapy before or after liver metastasis resection), Mr Mike Thompson (Early diagnosis and identification of increased risk of colorectal cancer), and Professor Mike Richards (Bowel cancer and the National Cancer Plan).