Recent Advances in the Management of Colon and Rectal Cancer1

One as yet ill-utilized aspect of epidemiological research is the translation of the generated data into meaningful health actions to control the incidence of colorectal cancer. This may involve complex decisions to alter the lifestyles of the community and may include among others government legislation, taxation policy and participation of industry.

The adenocarcinoma sequence was initially proposed in the 1960s. Dramatic advances in molecular genetics led to our understanding of colorectal neoplasia development described in a now classic paper by Vogelstein et al.¹⁴ Briefly, specific chromosomal abnormalities (Chromosomes 5, 17 and 18) were shown to be important in colorectal carcinogenesis and also the classic histological progression was mirrored indicating an accumulation of genetic abnormalities.¹⁴

Colorectal cancers are monoclonal and arise by clonal expansion, with serial mutations, each providing a growth advantage to the affected cells daughter line.¹⁵ Two distinct molecular pathways involving three classes of genes have now been identified that lead to colorectal cancer development. The RER pathway (Replication Error) or the microsatellite instability pathway is mediated by mutation of DNA mismatch repair genes (hMSH2, hMLH1, others). The LOH (Loss of Heterozygocity) pathway is initiated by mutation in the APC gene located on chromosome 5q. When the mutations in above genes are inherited (Germ-line mutations as opposed to somatic) the familial syndromes result (RER pathway leads to HNPCC and APC gene pathway leads to the FAP syndrome).

The three classes of genes involved are DNA mismatch repair genes, oncogenes (Dominant genes) and tumor-suppressor genes (Recessive genes). The most important oncogene in colorectal cancer is K-ras which is mutated, in approximately 40 percent of sporadic cases.¹⁶ Other oncogenes implicated include myc and src. Important tumor-suppressor genes include APC, p53, DCC (Deleted in Colorectal Carcinoma) and MCC (Mutated in Colorectal Carcinoma).

A detailed description of the hereditary syndromes is clearly out of the scope of this chapter and the interested reader has several excellent sources to refer to.¹³

Vast accrual of data from well-designed prospective randomized trials or high quality case-control studies adhering to the rigorous standards of evidence based medicine approach allow us to draw firm conclusions at this stage regarding the development of standards for screening and surveillance of colorectal cancer. Randomized studies have shown a reduction in mortality of 15 to 33 percent in those who undergo routine screening for colorectal cancer.^20–22 Several prominent medical and surgical gastrointestinal professional societies and the American Cancer Society now endorse evidence-based practice guidelines for the screening of this cancer. Responding to a call to develop uniform national guidelines, a consortium of five societies (American College of Gastroenterology, American Gastroenterological Association, The American Society of Colon and Rectal Surgeons, American Society for Gastrointestinal Endoscopy and the Society of American Gastrointestinal Endoscopic Surgeons) drew an interdisciplinary panel of 16 experts that studied 3500 peer-reviewed articles and published recommendations for screening.²³ These guidelines made recommendations for two levels of risk—average risk and increased risk individuals. All screening strategies were found to have a net benefit and the cost was within the range of cost-effectiveness commonly accepted for other tests such as mammography.

The three proven screening strategies are digital rectal examination combined with one of the following–fecal occult blood testing plus flexible sigmoidoscopy or total colon examination (colonoscopy or double contrast barium enema plus proctosigmoidoscopy). Mandel et al, in a randomized trial (The Minnesota colon cancer control study) showed that annual fecal occult blood testing with rehydration of the samples decreased the 13-year cumulative mortality from colon cancer by 33 percent.²⁰ Double-contrast barium enema can examine the entire colon with relatively high sensitivity and specificity for larger polyps (greater than 1 centimeter in diameter) and cancers and is less expensive than colonoscopy.²³ However, it is not possible to biopsy or remove neoplasms during the same procedure, so that patients with abnormalities must undergo an additional examination by colonoscopy to establish the diagnosis and provide treatment. Adding flexible sigmoidoscopy to double-contrast barium enema increases the sensitivity, but the magnitude in clinical importance of the additional sensitivity is uncertain.¹⁸ Colonoscopy permits visualization of the entire colon directly, along with detection and removal of polyps and biopsy of cancers throughout the colon.²³

The American Society of Colon and Rectal Surgeons website (http://www.fascrs.org) has published the practice parameters for the detection of colorectal neoplasms. The screening guidelines are presented in a table format and consider three levels of risk–low or average, moderate risk and high-risk. The interested reader is referred to the website for specific details regarding the levels of risk and the appropriate choice of the test, the age at which testing should commence and the interval/frequency of testing.

Recent observations suggest that aspirin and other NSAIDs, supplemental folate and calcium and postmenopausal hormone-replacement therapy (estrogen) have a chemopreventive benefit.²⁴ Of these the best studied are the aspirin and other non-steroidal anti-inflammatory class of drugs. These exert their effects by both COX dependent and COX independent mechanisms. COX-1 and COX-2 are enzymes (Cyclogenases) involved in the synthesis of prostaglandins. COX-2 expression is elevated in upto 90 percent of sporadic colon carcinomas and 40 percent of colonic adenomas, but is not elevated in the normal colonic epithelium.^25,26 Similar increase in levels of COX-2 and prostaglandins has been shown in adenomas in patients with FAP. Randomized trials of sulindac (a nonselective COX inhibitor) vs a placebo have shown significant decrease in the number of polyps in patients with FAP syndrome.^27–29 A selective COX-2 inhibitor, celecoxib, was shown in a randomized study to cause regression of polyps in patients with FAP syndrome.³⁰ Such level one evidence is not available at this time regarding the use of aspirin and other NSAIDs in the general (at average risk) population. The COX dependent mechanism appears to involve an increase in apoptosis, the regulation of angiogenesis or both.^31,32 The COX-independent mechanisms appear to involve inhibition of the activation of nuclear factor (NF-kB) or interference with the binding of the peroxisome- proliferator-activated receptor PPARδ to DNA.^33,34

Folic acid and its metabolites 5,10 methylenetetrahydrofolate and 5-methyltetrahydrofolate are critical components of DNA synthesis. MTHFR is an enzyme that converts 5,10 methylenetetrahydrofolate to 5-methyltetrahydrofolate. 5-methyltetrahydrofolate acts as a methyl donor for methionine synthase, an enzyme that catalyzes the conversion of homocysteine to methionine.²⁴ The role of folate in colon carcinogenesis has been elucidated by studies of patients who are homozygous for these enzymes. Patients with a diet adequate in folate with one of the above inherited enzyme polymorphisms have a decreased risk of colorectal cancer (relative risk, 0.5 for MTHFR and 0.51 for methionine synthase).³⁵

Drawing from the observation that during the past two decades mortality from colorectal cancer has decreased slightly in men but much more in women and that a possible explanation may be increasing use of hormone-replacement therapy, attention is focused on estrogens for their possible preventive role. The mechanism by which estrogens may exert their action is not entirely clear but may involve multiple pathways such as – decrease in the production of secondary bile-acids, decrease in the production of insulin-like growth factor I or by a direct effect on the colorectal epithelium.^36–38 Two recent meta-analyses have found an aggregate reduction in the risk of colorectal cancer of 20 percent with hormone replacement therapy.^39,40

6Other potential chemo-preventive measures such as ursodiol, eflornithine and oltipraz are undergoing evaluation in animal and clinical studies.

It is important to remember that chemoprevention is an alternate strategy and has not replaced the rigorously tested screening methods. The value and risk benefit ratio has not yet been confirmed in large double-blind, placebo-controlled, randomized studies. The data presented on chemoprevention does however provide tantalizing glimpses into the future approaches to reducing mortality from colorectal cancer.

The anatomic definition of the rectum is highly variable (discussed fully in the surgical management of rectum section) and therefore a preferred definition has been suggested as follows – colon is greater than 12 centimeters from the anal verge by rigid sigmoidoscopy.⁴¹

Surgical resection remains the gold standard to cure colon cancer. The principles of resection for cancer of the colon are well established. The reader may refer to standard textbooks and operative surgical atlases for descriptions of colon resection. The focus here will be on evidence based medicine approach to the current principles of colon resection for cancer. The extent of a colon resection is defined by the pattern of blood supply and venous drainage since the lymphatics run parallel with the named feeding vessels. The ideal extent is defined by removing the blood supply and the lymphatics at the level of the origin of the primary feeding arterial vessel. When the primary tumor is equidistant from two feeding vessels, both should be excised at their origin. The only still controversial issue is the level of ligation of the inferior mesenteric artery (IMA). Currently there is no data available to support ligation of the IMA at its origin.⁴¹

The length of bowel is usually not a concern with colonic primary and will be dictated by the ligation of the main feeding artery. Despite some debate, 5 cm of bowel margins on each side of the primary tumor appears to be adequate to minimize anastomotic recurrence.⁴² For cancers in the ascending colon where ileal resection is required the recommendation is to resect as short a length as possible to minimize the malabsorption syndromes since the length of ileum resected does not affect oncologic outcomes.^43,44 Patients with synchronous colon cancers or with HNPCC should be considered for total abdominal colectomy with ileorectal anastomosis. Patients with chronic ulcerative colitis with colon cancer should have total proctocolectomy with or without restoration.⁴¹

The lymphadenectomy has prognostic and therapeutic implications. An important value of lymph node resection is in staging. It is recommended that the TNM classification of tumors as described by the American Joint Committee on Cancer (AJCC) is universally adopted in lieu of the older Duke's and its variation of classification systems.^45,46 To achieve a high-degree of 7accuracy (> 90%), a minimum of 12 lymph nodes negative for metastatic disease must be examined to confirm that the N stage is truly N0.⁴⁷ The apical nodes (at the origin of the primary vessel) when possible should be removed and tagged since they may have additional prognostic significance.⁴⁸ The currently available data does not support extended lymphadenectomy beyond the main blood supply. One randomized clinical trial to assess the value of extended lymphadenectomy for left sided colon cancer failed to show oncologic benefit.⁴⁹

The management of rectal cancer has undergone significant evolution in the past decade. The optimal management of a patient with rectal cancer requires careful planning, individualization of treatment options and often a multi-disciplinary approach. The unique aspects of rectal cancer surgery are related to optimize the functional outcome–specifically preservation of the anal sphincters whenever possible, sexual and bladder function while not jeopardizing the oncologic outcome.

In order to correctly choose the therapeutic modality for the treatment of rectal cancer, it is imperative to stage the cancer accurately. The TNM staging system remains not only the most important prognostic factor, preoperative staging is an important branch point in the decision tree analysis regarding therapeutic planning. The protected location of the rectum within the confines of the bony pelvis makes imaging by CT scan and conventional MRI scanning less optimal than in other locations. Endorectal ultrasound (ERUS) has emerged as an important tool for preoperative evaluation of the penetration of the rectal wall and identification of perirectal lymph node involvement. This technique involves visualization of the rectal walls, para-rectal tissues and the tumor using a 360° rotating ultrasound probe inserted via a rigid sigmoidoscope. The probe itself is covered with a balloon that is filled with variable quantities of water to provide a sonolucent window. The Beynon five-layer model is generally used and a modified ultrasonic classification system is employed.^50,51 The 8classification scheme emphasizes the importance of the middle white line which represents the submucosa. The modified ultrasonic staging for rectal tumors is as follows:

The Figure 1.1 shows the ERUS staging system adopted from the Beynon model and the datasheet in use at the Colon and Rectal Clinic of Orlando, Orlando, Florida, USA. ERUS has an accuracy of between 81 to 94 percent for determining the depth of wall invasion.^52,53 The accuracy to determine lymph node status is more limited and ranges between 58 to 83 percent.^50,54 Our own experience at the CRC of Orlando and accuracy rates for ERUS staging are summarized in Figure 1.2. ERUS is operator-dependent and a significant learning curve is required with the technique and interpretation of the images. Most errors result from over-staging of T2 lesions (overcall 10.68%, CRC of Orlando experience) due to inflammatory changes and also diathermy or biopsy artifacts. Despite its limitations ERUS remains a valuable tool in classifying rectal cancers into two broad groups – Early (uT1/uT2 with uN0) and advanced disease (uT3NO/uT4NO or any uT with uN1).²

Several factors impact upon the decision to save or not to save the sphincters in rectal cancer and the selection of the appropriate operation for the patient. The most significant is the level of the tumor. The level of the tumor from the anal verge should be measured by rigid procto-sigmoidoscopy for the same reasons as that for defining the rectum vs the colon. Upper and middle third rectal cancers will be usually adequately treated by a low anterior resection while the controversial region is the lower third of the rectum. The recommended length of the distal margins of resection has changed10 significantly in the past several years. In the era of APR an ideal distal margin was 5 centimeters.

The other factors that influence the choice of operative treatment include the body habitus, sex, obesity, presence of complications such as perforation or obstruction and the general medical condition. Poor sphincter control is a contraindication to sphincter-saving procedure. Thus, an APR is still indicated in morbidly obese patients, especially males with a narrow pelvis and with bulky or poorly differentiated tumors located less than 2 centimeters from the dentate line.

11One of the most significant advance in our understanding of the pattern of recurrence of rectal cancer is the appreciation of the role played by lateral/radial margin and the related concept of total mesorectal excision (referred to as TME from here on).⁶⁵ The mesorectum is defined as the lymphovascular, fatty and neural tissue that is circumferentially adherent to the rectum, starting at the level of the sacral promontory, where the superior hemorrhoidal vessel divides into right and left branches. The mesorectum tapers and then diminishes just below Waldeyer's fascia (the investing fascia of the levators) around the levator ani muscles at the level of the distal third of the rectum. The concept of TME has been lately championed by Heald who recommends sharp dissection of the avascular plane in the retrorectal space with excision of the entire mesorectal lympho-vascular tissues down to the levators. He and others also urges caution against cutting obliquely across the mesorectum, and ‘coning down’ the dissection plane.^65,66 The additional sound principles of TME as reiterated by Nelson et al, are, presacral dissection under direct visualization, preservation of the integrity of the mesorectal fascia propria integrity, at least 4 cm clearance of attached mesorectum distal to the tumor and pathologic confirmation of mesorectum attached to the rectum distal to the tumor.⁴¹ Enker and others reported on their experience adhering to the principles of TME for rectal cancer resection on 246 consecutive patients operated at the Memorial Sloan-Kettering cancer center, New York, NY.⁶⁶ The pelvic recurrence rates were low at 3 percent in stage II and 5.8 percent in the stage III cancers. Further, adjuvant therapy was not found to have any statistical benefit in this group of patients. The mortality rate was only 0.8 percent in this large series. TME does not add morbidity and allows preservation of the parasympathetic nerves (needed for normal urogenital function).⁶⁷ In a more recent pooled data analysis, outcomes from three large series of TME resection were compared to conventional (non-TME) rectal resections.⁶⁸ A 30 percent survival advantage was noted in the TME group (5-year cancer specific survival 75 to 80% vs 52% in the conventional group). The local recurrence rates ranged from 4 to 9 percent in the TME series vs 32 to 35 percent in the conventional surgery group. To summarize TME can be considered as a major advance in the surgical management of rectal cancer. As a result of TME 5-year survival figures have risen from 45 to 50 percent to 75 percent, local recurrence rates have declined from 30 percent to 5 to 8 percent, sphincter preservation has increased by at least 20 percent for mid and lower rectal cancers and the rates of impotence and bladder dysfunction have declined from 50 to 85 percent to 15 percent or less.⁶⁹

To ensure 1 cm margin in a lower third rectal cancer surgeon may need to transect the rectum at the levator plane and construct an anastomosis in the anal canal (Colo-anal anastomosis). As a result of the loss of reservoir function of the rectum these patients may experience urgency, frequency of evacuation and occasional incontinence. In order to improve the functional results after coloanal anastomosis, the construction of colonic J-shaped pouch was proposed in 1986. The pouch is constructed by folding the colon and fashioning a side-to-side anastomosis with a linear stapler introduced through the anti-mesenteric apex of the pouch. The pouch-anal anastomosis is then 12generally performed with a double-stapled circular stapling device or occasionally via a trans-anal hand-sewn technique. Some surgeons advocate performing routinely a temporary diverting ileostomy while others use it selectively in older patients, preoperative radiation, anemia, malnutrition or any concern regarding the integrity of the anastomosis. The optimal length of the pouch appears to be 5 centimeters. Colonic J-pouch appears to be associated with acceptable morbidity and superior functional results (as assessed by decreased frequency of bowel movements, improved continence scores and the ability to defer defecation more than 30 minutes) when compared to conventional straight coloanal anastomosis.^70–72 The functional superiority has been shown to be greatest at one year after surgery. Recent reports are also confirming that the late clinical outcome is also superior providing continued benefit to the patient for at least 2 years.^73,74 A second potential advantage observed of this technique is the almost four fold decrease in the incidence rate of anastomotic complications. Joos et al, observed an incidence of 4.5 percent anastomotic complications in the colonic J-pouch group while the incidence in the straight coloanal anastomosis group was 17.9 percent.⁷¹ The exact reason for this phenomenon is not known, however, one possible explanation is the obliteration of the dead-space in the pelvis as a result of the volume of the pouch.

In some patients, construction of a colonic J-pouch may be technically impossible. This may be secondary to a narrow pelvis with inability to fit the capacious pouch or difficulty in reach. A new technique of coloplasty has been described as an alternate option in such circumstances. It creates capacity in the segment of colon by the principle of stricturoplasty whereby a longitudinal incision is sutured in a transverse fashion. In a comparative series from the Cleveland Clinic Foundation, Cleveland, Ohio, USA, Mantyh et al, report similar functional outcome as well as similar pouch compliance with this technique.⁷⁵

At the other end of the spectrum of radical resections for rectal cancers are various surgical options for local management of the primary tumor. With very careful selection of patients early lower third rectal cancers with favorable tumor characteristics can be treated for cure by local full thickness excision.^76,77 Local excision may be expected to result in cure if the tumor has not metastasized to the lymph nodes and of course to more distant sites. Several tumor characteristics have been extensively studied to indicate these favorable features. T stage or the depth of invasion predicts lymph node metastasis. Morson et al, found lymph node involvement in 11 percent patients with T1 stage tumors while 58 percent of patients with T3 tumors had lymph node metastasis.⁷⁸ Only T1 or T2 stage tumors should be considered as candidates for curative local excision. These tumors are likely to be mobile on clinical examination. ERUS staging as discussed in detail earlier is at the time of writing the best method for evaluating the penetration of the rectal wall. Tumor differentiation has been identified as a significant prognostic indicator of lymph node metastasis and local failure. Cohen et al, found a 69 percent incidence of lymph node metastasis in poorly differentiated tumors.⁷⁹ High grade or poorly differentiated tumors are considered unsuitable for local excision.⁸⁰ Tumor size is an important predictor of 13lymph node involvement and local recurrence. Based on the series by Cohen et al, and Graham et al, it seems to be prudent to limit local excision to tumors less than 4 centimeter in diameter.^79,81 Mucinous pathology is an independent negative prognostic indicator in several studies of local excision.^82,83 To recapitulate curative local excision may be considered if following favorable criteria are present–T stage T1 or T2, tumor size less than 3 centimeters, well or moderately differentiated and nonmucinous pathology. Local full thickness excision is performed nowadays via a transanal approach. The major advantages of the trans-anal approach being low morbidity, lower anesthetic requirement and faster recovery, posterior approaches such as parasacrococcygeal (Kraske) or the transsphincteric (York-Mason) have largely been abandoned. The reader is referred to standard textbook descriptions of the operative technique of trans-anal full thickness local excision. Willett et al, showed that local excision alone achieved a 5-year actuarial disease-free survival of 87 percent when histology was favorable.⁸⁴ In a review of 10 published series using local excision alone the cancer-specific survival rate was 89 percent, with a local recurrence rate of 19 percent.⁸¹

For rectal cancers that are otherwise suitable for local full thickness excision but situated in the middle-third or the proximal rectum the technique of transanal endoscopic microsurgery (TEM) is available. Professor Gerhard Buess in Germany developed the instruments and described this technique originally.⁸⁵ The technique involves using long instruments inserted through a long operating proctoscope. The shaft of stereoscopic microscope fits through the operating scope. A gas tight system allows carbon dioxide to be insufflated throughout the procedure to keep the rectal walls distended. Surgeons who have attempted to learn this technique agree that it is the most difficult new skill they have tried to master.⁸⁶ Buess et al, reported 74 carcinomas resected by TEM.⁸⁷ In his series major complications occurred in nine percent. Two rectovaginal fistulas resulted requiring temporary colostomy. This technique has not been widely adopted due to the high cost of the instrumentation, long learning curve involved to master a very difficult technique and the limited number of patients where the technique would be truly indicated.

An extension of the concept of sphincter preservation for patients with lower third rectal cancer is other described techniques of local destruction of the cancer. These include electrocoagulation, cryodestruction, endocavitary irradiation and laser ablation. While these techniques are principally used for palliation, curative intent in selected patients with favorable tumor characteristics has also been described. Salvati et al, report 65 percent 5-year survival in selected individuals with electrocoagulation.⁸⁸ These techniques may have a role in highly selected elderly patients with significant comorbidity. The major disadvantage of these techniques is the inability to obtain a pathologic specimen for adequate staging.

Modifications to standard surgical procedures are occasionally indicated in the presence of additional factors. Thus, a Hartmann's procedure, which involves removal of the rectosigmoid with closure of the rectal stump and an end colostomy may be indicated in perforated or obstructed rectosigmoid cancers or when a palliative 14resection is performed. One of the recent technique used for the management of obstructed rectal cancers is the use of metallic self-expanding stents and Nd:YAG laser photoablation.^89–91 This technique can be used not only with palliative intent but also in select patients with curative intent as well. It helps to tide over the emergent scenario, allows resolution of the micro-circulatory changes that accompany acute large bowel obstruction, allows proper work-up of the patient and acts as a bridge to a semi-elective curative resection.

Adjacent organ resection may be required in the treatment of T4 rectal cancers in approximately five percent of patients with rectal cancers.⁹² Rectal cancer most often involves the uterus, adnexa, posterior vaginal wall and the urinary bladder. This adherence may be inflammatory or may represent true malignant invasion. Since it is not possible to reliably differentiate between these two processes it is recommended to perform concomitant en bloc resection for tumors adherent to these structures.⁴¹ Several articles in the literature report improved survival when en bloc extended resection achieves histologically clear margins.^92,93

Grossly involved ovaries should be removed at the time of primary resection. However, prophylactic oophorectomy has been a controversial issue. Ovarian metastases occur in upto 6 percent of patients and are associated with poor prognosis and generally widespread disease.⁹⁴ Cutait et al, in a retrospective study showed no survival advantage with oophorectomy.⁹⁵ Ballantyne in a similar study reported a 5 percent improvement in the oophorectomy group however this was not statistically significant.⁹⁶ Overall the current recommendation based upon available data does not support routine oophorectomy.⁴¹

Jacobs et al, Cooperman et al, initially described laparoscopic colon resections in the literature in 1991 followed by Monson et al, in nineteen ninety two.^97–99 Larach et al, described the technique of laparoscopic abdomino-perineal resection and concluded that an APR is ideally suited for the laparoscopic approach.¹⁰⁰ The feasibility of laparoscopic colectomy was then established for nearly every benign and malignant disease process of the colon and rectum.^101,102 It was soon realized that laparoscopic colon surgery is much different and vastly more complex than the operation of laparoscopic cholecystectomy. Laparoscopic colectomy involves operating on mobile segments of the bowel, dissection in multiple quadrants, necessity to ligate major blood vessels, need to perform bowel anastomosis, removal of a bulky specimen, need to be familiar with a diverse types of pathology and diverse operative procedures. This translates to a long learning curve estimated to be between 35 to 50 procedures.^103–105 When a laparoscopic resection is performed with safety, the benefits of laparoscopic approach have been well documented in the literature.^{102,105–109} The benefits are related to minimal abdominal wall trauma, less postoperative pain, better cosmesis, earlier return of gastrointestinal function, earlier postoperative feeding and shorter length of 15hospital stay. As to the issue of cost-effectiveness while the operating room costs are typically higher due to expensive disposable instrumentation and longer operation times secondary to the learning curves overall costs have been shown to be comparable due to the benefits of shorter hospital length of stay and diminished requirement of pain medications.¹¹⁰

What is the issue then when we consider laparoscopic resection for colon cancer? The central issue is the oncologic equivalency of the laparoscopic procedure and this includes adequacy of the resection, staging limitations if any, the issue of port-site recurrences and the survival outcome. Laparoscopic colon resection must be able to achieve the same oncologic resection achieved by open resection. Lord SA and Larach et al, showed that an equivalent oncologic resection measured by the number of lymph nodes harvested and tumor free margins can be achieved by laparoscopy.¹¹¹ Other investigators have confirmed these findings and it is now accepted that laparoscopy does not compromise oncological resection.^112–116 As far as the issue of staging is concerned it has been shown that regional staging is adequately and accurately achieved by laparoscopy in so far as the lymphatic dissection and the average number of lymph nodes are harvested.¹¹¹ However, as far as distant staging is concerned, specifically the assessment of hepatic metastasis, the value of laparoscopic ultrasound has not yet been shown to be equivalent to open intraoperative ultrasound.

By far the most ominous concern raised so far has been the issue of port-site recurrences. The initial report by Berends et al created a stir in the laparoscopic surgical community.¹¹⁷ He reported 21 percent incidence of port-site recurrences and raised the possibility that the pneumo-peritoneum somehow altered the pattern of spread and local wound biology. However, it must be realized that for his series the denominator was of 14 patients. Nevertheless other anecdotal reports of trochar site recurrences raised the awareness of surgeons about the potentially adverse effects on outcome.^118,119 Recent data examined from large series (Single institutions with over 100 patients and multi-institution series with over 400 patients) however shows that the implantation rate at trochar sites is low between 0 to 2.4 percent.^{110,120–122} This rate is comparable to that reported in open resections.¹²³

Prospective and retrospective reviews of large series of patients who have had laparoscopic resection for colorectal cancer suggest that the preliminary survival rates are similar to historical series of open resections for cancer.^120–122 Several small randomized, controlled trials have produced early data suggesting that laparoscopic colectomy for cancer is safe, offers improved early recovery and has no adverse oncological outcomes.^124–126 Laparoscopic colectomy for cancer is currently being rigorously studied by well designed randomized prospective clinical trials in the United States, Europe, Scandinavia and Australia.^116,127,128 Reporting of oncological outcomes from these trials is expected to answer definitively the role laparoscopic colectomy for cancer will play in future.

The most common pattern of failure of rectal cancer is characterized by locoregional recurrence. Approximately one-third of rectal cancer patients undergoing radical surgery will develop a local recurrence.¹³⁰ It is generally agreed that the recurrence rate of APR and sphincter-saving radical resection are virtually identical.¹³¹ As many as 70 to 80 percent of patients have pelvic recurrence alone or in combination with distant metastasis, while only 20 to 30 percent have distant metastasis alone. Local recurrence of rectal cancer causes significant suffering to the patient and poses a challenge to the colon and rectal surgeon. This discussion will focus on the recent advances in the identification and management of this complex problem.

Local recurrence can be classified into following types: (a) Axial –includes those occurring at the suture line on the mucosa generally as a result of inadequate distal margins, perianastomotic by extrinsic disease with an inward growth pattern generally from inadequate mesorectal clearance and perineal recurrences occurring after an APR generally from direct tumor implantation at the time of surgery. (b) Posterior involving the sacrum and the coccyx (c) Anterior involve the genitourinary tract, in the proximity of prostate, seminal vesicles, urinary bladder, vagina and/or uterus and (d) Lateral involving the pelvic side walls such as iliac vessels, distal ureters, lymph nodes, nerves and muscles.¹³⁰ The classification helps in standardization of treatment and comparison of outcomes from heterogeneous series.

The recurrence may be suspected on the basis of an elevated carcinoembryonic antigen (CEA). On the other hand, patients may present with symptoms such as pelvic pain, pressure, new onset vaginal bleeding or discharge, dysuria or rectal bleeding. The evaluation must be thorough and systematic, the first step being a complete physical examination including rectal and pelvic examination, rigid/flexible sigmoidoscopy, full colonoscopy and imaging studies. The imaging modalities helpful in the preoperative workup of local recurrence of rectal cancer are endoluminal ultrasound (EUS), CT scan, MRI, nuclear imaging and positron emission tomography (PET).

Endoluminal ultrasound can detect pelvic masses and help to direct needle biopsy for pathological confirmation. Differentiation between recurrent cancer and fibrosis is possible since fibrosis appears hyperechoic while recurrent cancer appears hypoechoic.¹³² Metallic clips do not cause interference. The limitations are–highly operative dependent study and postoperative changes alter the ultrasound anatomy.

CT Scan findings suspicious for local recurrence include asymmetric thickening of the bowel wall, bileteration of perianastomotic fascial or fat planes and presacral or pelvic side-wall mass. In addition 17extrapelvic disease may be identified as well. The major limitation is the inability to distinguish between scar and recurrent cancer.

MRI often complements the CT findings by adding further anatomical detail regarding the extent of involvement in up- to 40 percent of patients. The additional information helps in the preoperative planning. MRI has been shown to have higher sensitivity, specificity and accuracy (91-100%, 83-100% and 93-95% respectively) than CT scan (82%, 50% and 69%) for detecting locoregional cancer recurrence.^133,134

Immunoscintigraphy is a nuclear imaging technique that uses radiolabeled antibodies to tumor specific antigens. This technique is useful when rising CEA levels are accompanied by negative conventional radiological imaging studies. Overall this technique has been reported to have a sensitivity of 77 percent and a positive predictive value of 78 percent.¹³⁵

PET scanning is a functioning imaging technique with picomolar sensitivity. It is based upon the principle that cancer cells have an increase in the expression of glucose transporter molecules and therefore accumulate the 18-fluoro-analog of glucose tracer that emits positron on decay. PET scanning is especially useful in distinguishing scar from cancer. In addition it also helps in identifying the whole body recurrence status. PET scanning has been shown to be superior to conventional imaging studies for the identification of loco-regional recurrence, with sensitivity of 90 percent and specificity of 90 percent as compared to CT scans sensitivity of 71 percent and specificity of 85 percent.¹³⁶

Once recurrent pelvic rectal cancer has been identified the selection of modality of treatment needs to be individualized. The only curative option for patients with isolated resectable pelvic recurrence is surgical resection. The selection of this modality of treatment demands careful planning and preoperative patient selection. The factors that need to be taken into consideration include age of the patient, symptoms, comorbid diseases, presence of extrapelvic recurrence and resectability of local disease. Thus, an elderly patient with significant comorbidity may not be a candidate for a radical resectional salvage procedure. Similarly, a patient with diffuse systemic disease also is not a candidate for a surgical option.

In carefully selected patients, the operative approach begins with a cystoscopy and placement of bilateral ureteric stents. This is followed by an exploratory laparotomy at which time 25 to 50 percent patients may be found to have unresectable disease. The curative intent operative options for recurrences based upon the classification schema outlined earlier are as follows—Axial recurrences–low anterior resection or APR or wide local excision, anterior recurrences – pelvic exenteration, posterior recurrences en bloc resection with periosteal elevation/abdomino-sacral resection and lateral recurrences—en bloc resection. The interested reader is referred to two excellent reviews on the subject.^130,131 Wanebo et al, have reported 33 percent 4-year survival rate and median overall survival of 39 months in this challenging subset of patients.¹³⁷

Radiation therapy (that may be combined with chemotherapy) may be not only palliative but be associated with long-term survival as well in these patients. However, patients who have had prior pelvic radiation may not be able to receive additional external-beam radiation. Intraoperative radiation therapy (IORT) and brachytherapy are two 18alternate approaches that have promising application in this scenario. Harrison et al, reporting on memorial Sloan-Kettering cancer center, New York experience with IORT achieved 63 percent 2-year actuarial local control and 71 percent disease free survival if negative margin was obtained.¹³⁸ Suzuki et al, reported on the morbidity of this technique and point out that sciatic nerve, ureter and iliac vessels are the major structures at risk from radiation damage.¹³⁹ The technique of brachytherapy is based upon implantation of radioactive sources close to the areas of residual disease with the goal of achieving a safer and possible more effective higher dose of radiotherapy.¹⁴⁰ Brachytherapy appears to improve the local control of recurrent rectal cancer.¹³⁰

Once again the location of the primary cancer, i.e. colon vs rectum impacts significantly upon the selection of appropriate therapy and will be discussed separately.

Currently patients with stage III colon cancer are regarded as candidates for adjuvant chemotherapy that consists of administration of 5-FU and leucovorin for a period of 6-months postoperatively.¹⁴¹ This important milestone recommendation was achieved in several small steps. The 1st major intermediate milestone was the recommendation by national institutes of Health, USA sponsored consensus development conference in 1990s that accepted a combination of 5-FU and levamisole for one year as standard regimen.¹⁴² This recommendation was based upon the then accumulated evidence from large co-operative groups that were able to accrue sufficiently large number of patients to achieve statistical power to show significance in outcomes. The most notable is the national intergroup trial consisting of pooling of data from the NCCTG (North central cancer treatment group), The Mayo clinic, The south western 19oncology group and the eastern cooperative oncology group.¹⁴³ This trial recruited dukes stages B2 and C (Stages II and III) patients and had three arms of study – observation alone, treatment with Levamisole alone and adjuvant treatment with 5-FU plus levamisole. A total of 1247 patients were analyzed. The study reported that at 3 years median follow up for patients with dukes stage C disease administration of 5-FU and levamisole resulted in 41 percent reduction in disease recurrence and a 33 percent reduction in the overall death rate as compared to the observation alone arm. The use of levamisole alone showed no significant benefit. A subsequent analysis of dukes C patients at six and half-year follow-up showed unchanged results.¹⁴⁴

The next milestone was the demonstration of the efficacy of fluorouracil plus leucovorin based regimes. The national surgical adjuvant breast and bowel project (NSABP) protocol C-03 randomized trial accrued 1081 patients with stage II and III colon cancer and at 3-years follow-up reported significantly improved disease free and overall survival in the arm receiving combination of 5-FU plus leucovorin.¹⁴⁵ In this study high-dose of leucovorin was used at 500 mg/m² and the duration of adjuvant therapy was 48 weeks.

The international multicentre pooled analysis of colon cancer trials (IMPACT) reported on the combined analysis of three trials conducted in Italy, Canada and France.¹⁴⁶ Total of 1493 patients with stage II and stage III colon cancer were randomized to two arms after curative resection – no adjuvant treatment vs 5-FU plus leucovorin (200 mg/m²) for 6 months. Stage III patients particularly with 4 or more positive lymph nodes had a significantly improved disease free and overall survival with adjuvant therapy. There was 25 to 30 percent decrease in the odds of dying from colon cancer, translating to an absolute improvement in survival of 5 to 6 percent compared to controls. The intergroup trial of intensive course of 5-FU plus low-dose of leucovorin (20 mg/m²) for 6 months showed that at three and half-years follow-up the treatment group had a significant reduction in tumor recurrence.¹⁴⁷

What about the role of adjuvant treatment in stage II colon cancer? At the present time a standard recommendation does not exist however a recent pooled analysis of NSABP protocols C-01, C-02, C-03 and C-04 has indicated that the relative treatment benefit in stage II colon cancer is at least equal to the benefit in stage III disease.¹⁴⁸ However, if the proportional reductions in mortality are similar, the absolute benefit in terms of 5-year survival will be small in this group because of their lower risk of recurrence. The estimate is 2 to 3 lives saved for every 100 patients treated.¹⁴⁹ Defining subgroups within Stage II will be helpful to focus on appropriate patients for adjuvant therapy.

Chemotherapy, radiation therapy and the combination chemoRT has been used in order to improve further the oncological results of rectal cancer surgery. Based upon the timing of administration three modes currently exist. These are neoadjuvant, adjuvant and sandwich setting. In the neoadjuvant setting radiation therapy usually in combination with chemotherapy is administered first followed 5 to 6 weeks later by planned resection. In the adjuvant setting chemotherapy with or without radiation is administered following the surgical resection. As the name suggests sandwich setting follows the following sequence:

While the cornerstone of chemotherapy for rectal cancer remains 5-FU, as to the administration of radiation several different protocols exist with notable differences between the USA and Scandinavia/western Europe. Large cooperative group trials [The Gastrointestinal tumor study group and the north central cancer treatment group] have offered evidence supporting the value of this therapy.^151,152 These led to a national institutes of health consensus development conference, which concluded that a combination of postoperative chemotherapy and radiation therapy [adjuvant setting] offered improved local control as well as survival in stages II and III rectal cancer.¹⁴² The major advantages of adjuvant setting are that accurate pathological staging is available and the bowel is not irradiated. More recently doubts have been cast on the value of the addition of radiation therapy to chemotherapy in the postoperative or adjuvant setting. The NSABP R-02 protocol accrued 694 patients with rectal cancers staged II and III between 1987 and 1992.¹⁵³ This study reported that the addition of postoperative radiation therapy reduced loco-regional recurrence statistically significantly from 13 to 8 percent at 5 years compared to the chemotherapy alone arm while not altering the disease free or overall survival.

In the Scandinavian and Western Europe nations, the neoadjuvant setting has been more popular.¹⁴¹ There are several theoretical advantages to administering radiation therapy alone or as chemoRT in the preoperative setting. Radiation is most effective on well-oxygenated cells and the local environment including the blood supply is optimal prior to surgical handling of the tissues. Small bowel is not adherent by postoperative adhesions and can be displaced out of the pelvis by positioning and by bladder filling to protect it from radiation damage. Also the cancer burden of the primary tumor is decreased and is even sterilized in upto 22 percent of patients (Williamson et al, CRC of Orlando, Orlando, Florida, data presented at the northeastern society of colorectal surgeons meeting west Palm Beach, FL, November 4-7, 2001) thus minimizing the chances of seeding during surgical resection. Reduction of tumor bulk may increase the chance of sphincter preservation.¹⁵⁴ Chemotherapy agents such as 5-FU may have a radiosensitizing effect and thus enhance the potential for locoregional control.¹⁵⁵ The greatest caveat in the use of any neoadjuvant therapy comes from the limitations inherent in the accuracy of 21preoperative staging methods that were discussed earlier.

The Swedish Rectal Cancer Trial data published in 1997 has made a very important contribution to demonstrate the efficacy of preoperative radiation therapy.¹⁵⁶ This study randomized 1168 patients with resectable rectal cancer to two arms–surgery alone vs preoperative radiation (25 Gy in 5 fractions in 1 week) followed 1 week later by surgery. At 5-year follow-up the neoadjuvant therapy group had 11 percent local recurrence rate vs 27 percent in the surgery alone group, 5-year survival was 68 percent compared to 48 percent in the surgery alone group. This study conclusively demonstrated improved local control as well as survival. Several other trials are studying the most effective radiation therapy delivery protocols and other phase I/II trials are addressing the effectiveness of newer chemotherapeutic agents such as CPT-11 in association with 5-FU plus Leucovorin.

At the beginning of this new millennium we excitingly await the promise of dramatic new approaches to the diagnosis, staging and treatment of colon and rectal cancer.