recurrence.128,129 Application of the same surgical principles to colon cancer, including the removal of

all the lymph node–bearing mesentery to the origin of the named blood vessels – an operation called

complete mesocolic excision or CME – is also associated with a lower rate of local tumor recurrence in

colon cancer patients.130 Differences in expertise and surgical technique probably explain surgeon and

institutional differences in outcomes among patients treated for CRC. A recent study from Denmark has

shown that CME surgery is associated with better disease-free survival than conventional cancer

resection for patients with stage I to III colon adenocarcinoma.131

TREATMENT OF LOCALIZED COLON CANCER

Malignant Colonic Polyps

The identification of a focus of invasive carcinoma in a polyp removed by snare excision during

colonoscopy represents a surgical dilemma. These types of polyps are often referred to as “malignant

polyps,” to distinguish them from lesions that are almost exclusively invasive adenocarcinoma with

exophytic or polypoid morphology. Malignant polyps are encountered more often because tumors are

diagnosed at an earlier stage, as a result of screening programs. Although there are some lymphatic

channels in the lamina propria, tumors superficial to the muscularis mucosa (also known as intramucosal

carcinoma, carcinoma in situ, or Tis) are considered to carry no risk of lymph node metastasis. When

the malignant cells penetrate beyond the muscularis mucosa they can access the lymphatic channels of

the submucosa and metastasize to the regional lymph nodes. This risk of nodal metastasis depends on

the morphology of the polyp, the depth of the invasive component, the presence of unfavorable

histologic features (grade 3 or 4, angiolymphatic invasion, perineural invasion) and the margin of

resection.132 Imaging studies are usually of no help in these patients, except to exclude distant

metastases (which are uncommon). In pedunculated polyps with focus of invasive cancer located in the

head, neck, or stalk of the polyp, with favorable histologic features, resected with negative margins, the

risk of nodal metastasis is small (Fig. 68-9). In these circumstances additional surgery is considered

unnecessary, as the risk of surgery may outweigh the risk of residual tumor in the bowel wall or nodal

metastasis. On the other hand, patients with pedunculated polyps with focus of invasive

adenocarcinoma reaching the base of the polyp, sessile polyps with focus of invasive adenocarcinoma,

the presence of adenocarcinoma at the resection margin, and malignant polyps with unfavorable

histologic features, are at risk of regional nodal metastasis, and should be offered surgical

resection.133,134 Other factors such as the location of the polyp in the colon or rectum,135 the patient’s

comorbidities and performance status, and the patient’s desires, should be taken into consideration.

Therefore, a final decision should only be made after full review of the pathology reports, and

disclosure to the patient of the risks and benefits of each approach.

Figure 68-9. Malignant polyp. Haggitt levels of invasion. Level 0, Noninvasive, severe dysplasia; Level 1, Invading through

muscularis mucosa but limited to the head of pedunculated polyp; Level 2, Invading the neck of pedunculated polyp; Level 3,

Invading the stalk of pedunculated polyp; Level 4, Invading into submucosa of the bowel wall below the stalk of a pedunculated

polyp. All sessile polyps.

Surgery for Localized Colon Cancer

Patients with biopsy-proven adenocarcinoma of the colon without evidence of distant metastasis, and

1782

without contraindications to major surgery, should be treated with surgical resection. The extent of a

colon cancer resection is dictated by the location of the primary tumor. Tumors in the cecum and

ascending colon require a right hemicolectomy that implies division of the ileocolic, right colic, and

sometimes the right branch of the middle colic vessels (Fig. 68-10). The portion of the omentum

attached to the removed segment of colon should be resected en bloc with the colon and mesentery.

Tumors located at the hepatic flexure and in the right side of the transverse colon require an extended

right colectomy that, in addition to the ileocolic and right colic vessels, requires division of the middle

colic vessels at their origin (Fig. 68-10). Tumors in the mid portion of the transverse colon can be

treated with an extended right colectomy, or a transverse colectomy. A transverse colectomy involves

division of the middle colic vessels only, but often requires mobilization of both the hepatic and splenic

flexure to ensure a tension-free anastomosis (Fig. 68-10). Tumors located in the distal transverse colon,

splenic flexure, and proximal descending colon can be treated with a left hemicolectomy, which requires

division of the left branch of the middle colic and left colic artery at its branch point from the inferior

mesenteric artery (Fig. 68-10). Locally advanced tumors located in the transverse colon can metastasize

to the regional lymph nodes located along the greater omentum and gastroepiploic arcades, leading

some surgeons to recommend removal of the omentum with the gastroepiploic arcade. In these patients,

an omentectomy with division of the gastroepiploic vessels at their origin may be necessary for

complete nodal control.136 However, the benefit of such an extended lymphadenectomy is debated.

Sigmoid tumors require sigmoid colectomy, which includes the superior rectal artery and its takeoff

from the inferior mesenteric artery (Fig. 68-10).

Patients with synchronous tumors, which occur in up to 5.3% of patients with CRC, should be

investigated for hereditary CRC syndromes or other predisposing conditions. Patients with sporadic

synchronous cancers can be treated with separate resections, or an extended resection incorporating

both lesions, depending on the location of the primaries.137 When performing a segmental resection, it

is important to preserve the blood supply to the intermediate segment of colon in order to avoid

ischemia, which can lead to perioperative complications. Synchronous tumors in patients with HNPCC,

or other risk factors such as inflammatory bowel disease, are indications for a subtotal or total

colectomy.

1783

Figure 68-10. Segmental colonic resections. A: Right colectomy. B: Extended right colectomy to transverse colon. C:

Transversectomy. D: Right colectomy extended to splenic flexure. E: Left colectomy. F: Sigmoidectomy.

A number of well-conducted prospective trials have proved that laparoscopic colectomy for cancer is

associated with short-term benefits and equivalent long-term oncologic outcomes, compared with the

traditional open surgical approach (Table 68-8). Patients treated laparoscopically have less

postoperative pain, earlier return of bowel function, shorter hospital stay, and a decreased rate of

complications. However, laparoscopic surgery is technically challenging and has a long learning curve.

Laparoscopy may be particularly difficult in patients with multiple adhesions from previous surgery, in

obese patients, and in patients with locally advanced disease. Late conversion from laparoscopic to open

surgery (also known as reactive conversion) is associated with higher complication rates compared to

open surgery, while early or preemptive conversion is associated with complication rates similar to

open surgery.138

Laparoscopic colectomy is typically performed using a number of ports, usually 3 to 5, and a

specimen extraction site. Additional expertise allows minimally invasive surgery to use fewer ports,

smaller instruments, and shorter incisions. This trend has resulted in the single port laparoscopic surgery

(SPLS) for the treatment of colon and rectal cancers. The operation is performed through a small port

located at the specimen extraction site. The camera and the operating ports are all positioned through

the SPLS device. A recent systematic review of 38 case series, including 565 patients operated by SPLS,

suggested that the procedure is feasible but technically challenging. Evidence regarding its safety was

limited.139

Table 68-8 Open versus Laparoscopic Colon Resections, Results of the Most

1784

Representative Series

The authors compared results of patients included in this review with those of a group of 3,526

patients treated with conventional laparoscopic colectomy and included in a Cochrane Review. They

found that, in spite of lower BMI and smaller tumors, SPLS did not result in less postoperative pain or

shorter hospital stay, compared to conventional laparoscopic surgery.139

The introduction of robotic platforms, which enhance the visualization and dexterity of the surgeon,

has added a new dimension to minimally invasive colon cancer surgery. The evidence accumulated so

far indicates that robotic colon resection is associated with lower conversion rates, similar short-term

outcomes, and higher costs, compared to conventional laparoscopic colon resection.140

Postoperative Adjuvant Therapy for Localized Colon Cancer

Many patients with what appears to be localized colon cancer at the time of diagnosis develop

recurrence after a curative-intent surgery. These recurrences are attributed to micrometastases that are

already present, but clinically undetectable, at the time of diagnosis. Postoperative adjuvant

chemotherapy aims to improve survival by eradicating micrometastatic disease. Numerous

chemotherapeutic agents are now considered for adjuvant therapy, mostly based on their efficacy as

observed in the setting of metastatic colon cancer. However, the mainstays of adjuvant therapy

following resection of localized colon cancer remain fluoropyrimidines (5-fluorouracil [5-FU] and

capecitabine) and oxaliplatin.

5-FU, a member of the family of antimetabolites, is a pyrimidine analog that works by irreversible

inhibition of thymidylate synthase (TS), the rate-limiting enzyme in pyrimidine nucleotide synthesis,

and by incorporation of its metabolites into RNA and DNA. 5-FU can only be delivered intravenously

and is commonly administered with leucovorin, a reduced folate that is thought to stabilize

fluorouracil’s interaction with the enzyme TS. Capecitabine is an oral prodrug of fluorouracil that is

absorbed intact through the gastrointestinal mucosa and undergoes enzymatic conversion to

fluorouracil. Response rates to fluoropyrimidine, as first-line therapy alone for metastatic CRC, are only

around 10% to 20%. Neutropenia, stomatitis and diarrhea are the most common side effects of 5-FU.

The toxicity profile for capecitabine is similar to 5-FU, with hand-foot syndrome being more frequent.

Oxaliplatin produces cytotoxic effects by forming both inter- and intra-strand cross links leading to

disruption of DNA replication and apoptosis. As a single agent, oxaliplatin has limited effect in CRC; but

in combination it enhances the cytotoxic effect of fluoropyrimidines, possibly through downregulation

1785

of TS. The most relevant side effect of oxaliplatin is a progressive and often irreversible peripheral

neuropathy. Other chemotherapy agents commonly used in patients with metastatic CRC have not

shown proven benefit in the adjuvant setting.

The use of adjuvant therapy for stage III colon cancer was first shown to provide measurable benefit

in the National Cancer Institute (NCI) Cooperative Intergroup trial INT-0035. Patients who received 5-

FU and levamisole for 1 year after surgical resection had a 33% risk-reduction in death or recurrence,

compared to those who underwent surgery alone.141,142 Since then, we have come to recognize that

levamisole is overall an inactive agent; therapeutic combinations have evolved and have been tested

rigorously.143 A number of multi-institutional prospective, randomized trials have demonstrated a

survival advantage with postoperative chemotherapy in selected patients after curative-intent surgery,

compared to surgery alone. Pooled analyses of these trials have reported an approximately 30%

reduction in recurrence, and 26% improvement in survival, in patients with stage III colon cancers

treated with 6 months of postoperative fluoropyrimidine-based chemotherapy.144,145 More recently, the

landmark MOSAIC (Multicenter International Study of oxaliplatin, fluorouracil, and leucovorin in the

Adjuvant Treatment of Colon Cancer) trial, which randomized 2,246 stage II/III colon cancer patients to

5-FU/leucovorin with or without oxaliplatin, reported an additional 20% reduction in recurrence and

5% improvement in 5-year disease-free survival (73.3% vs. 67.4%; p = 0.003) in patients treated with

5-FU/leucovorin plus oxaliplatin, compared to 5-FU/leucovorin alone.146,147 Based on these results, the

current recommendation for patients with stage III colon cancer following curative-intent surgery

consists of 6 months of 5-FU/leucovorin and oxaliplatin (FOLFOX), or capecitabine and oxaliplatin

(CAPEOX).

Several studies have proven that the addition of irinotecan – a topoisomerase I inhibitor that is

effective in metastatic CRC – to standard fluoropyrimidines is not superior to fluorouracil/leucovorin

combinations alone in stage III colon cancers.148 Similarly, bevacizumab (anti–VEGF-A) and cetuximab

(anti-EGFR), two biologic agents that have shown efficacy in metastatic disease, do not improve

survival, compared to FOLFOX alone.149–151 Therefore, at the present time there is no role for

irinotecan, bevacizumab, or cetuximab in the adjuvant treatment of CRC.

The use of adjuvant chemotherapy after curative resection for stage II CRC has been controversial for

more than a decade. Three prospective, randomized trials have addressed the subject using 5-FU–based

regimens,142,152,153 all of them failing to show a clear benefit of adjuvant treatment over observation

alone. The addition of oxaliplatin to the adjuvant regimen does not seem to improve the results of

pyrimidine-based regimens for stage II. Most of the evidence for adjuvant treatment in stage II CRC

comes from pooled analyses, in which stage II and III CRC patients, undergoing different chemotherapy

regimens and controls, are analyzed. The most influential one, including 37 trials and 11 meta-analyses,

found a significant improvement in DFS for 5% to 10%, favoring adjuvant chemotherapy; nevertheless,

no difference in OS was observed.154 Therefore, postoperative adjuvant chemotherapy is still not

routinely recommended after curative resection for stage II CRC. However, there is evidence that some

node-negative patients at high risk for recurrence may benefit from postoperative adjuvant

chemotherapy. Determining which patients are at high risk has most commonly included consideration

of features such as tumor perforation, T4 category tumors, poorly differentiated histology,

lymphovascular or perineural invasion, and patients with fewer than 12 nodes in the surgical specimen

(who are considered inadequately staged).155 At the present time, the use of adjuvant therapy for stage

II disease ought to be weighed carefully by a multidisciplinary team, with adequate counseling of the

patient. As mentioned earlier, some predictive nomograms may help stratify risk in patients with stage

II disease, enabling us to more effectively select those more likely to benefit from adjuvant therapy;

however, these tools have not been widely incorporated into clinical practice.

The search for molecular prognosticators aimed at stratifying stage II patients according to the risk of

relapse, is an active area of research. MMR deficiency has been associated with an improved prognosis,

but a decreased benefit from fluoropyrimidine-based adjuvant therapies, in patients with stage II

disease.156 Because of this, MMR testing is recommended for patients with stage II disease. Current

guidelines recommend no use of single-agent fluoropyrimidine treatment for patients with stage II

tumors exhibiting MSI or MMR deficiency. The multi-gene molecular signatures discussed previously

were developed, in large part, to stratify risk in patients with stage II CRC, and to identify those most

likely to benefit from adjuvant chemotherapy. While initial results are promising, there is insufficient

evidence to recommend their use in selecting adjuvant therapy for patients with stage II CRC.

1786

TREATMENT OF LOCALIZED RECTAL CANCER

The treatment of rectal cancer presents different challenges, but also different opportunities, compared

to colon cancer. Given the location of the rectum within the narrow bony pelvis surrounded by the

urogenital organs, large blood vessels, and autonomic nerves, and its proximity to the anal sphincters,

rectal cancer surgery is technically challenging, potentially associated with perioperative complications,

and often followed by loss of urinary, sexual, and bowel function that permanently impairs quality of

life. In addition, rectal cancer has been associated with a higher rate of local recurrence after curativeintent surgery, compared to colon cancer. But the extraperitoneal location of most of the rectum and its

proximity to the anal orifice make it accessible to explorations (DRE, endorectal ultrasound [ERUS]) and

interventions (external beam radiation therapy, brachytherapy, and local excision [LE]) not currently

available for colon cancer.

The selection of the best treatment option for each rectal cancer patient is a complex process and

requires consideration of the clinical tumor stage, location of the tumor within the rectum, the impact

of treatment upon anorectal and genitourinary function, and the consequences for the patient in terms

of prognosis and quality of life. Some rectal cancers can be treated with LE. Others require a

transabdominal resection (TAR) of the rectum, with or without sphincter preservation. Finally, many

rectal cancer patients require chemotherapy and radiation to reduce the risk of local recurrence and

distant metastasis.

Anatomy of the Rectum

The rectum represents the distal portion of the large bowel, extending from the rectosigmoid junction –

the area where the tenia coli (characteristic of the sigmoid colon) splay and become the longitudinal

muscular layer of the rectum – to the anorectal ring, a palpable anatomical landmark that corresponds

to the imprint of the puborectalis muscle on the bowel wall. The anal canal extends from the anorectal

ring to the anal verge, the palpable groove between the distal edge of the internal sphincter and the

subcutaneous portion of the external sphincter (Fig. 68-11). For practical purposes and standardization,

the rectum is defined in terms of distance in centimeters from the anal verge. The distance of a tumor

from the anal verge is best measured using a rigid proctoscope, which allows the simultaneous

visualization of the anal verge and the centimeter marks on the outside of the scope. In European

countries, tumors with the distal edge located within the last 15 cm of the large bowel are considered

rectal cancers; in the United States, the limit is 12 cm. The location of tumor in relation to the

peritoneal reflection and the promontory, and its distance from the anal verge, can also be determined

by MRI, in particular sagittal views.

The mesorectum is the visceral mesentery of the rectum containing the terminal branches of the

superior rectal vessels and the rectum’s lymphatic drainage. The upper portion of the rectum is located

above the anterior peritoneal reflection; it is covered with peritoneum in the front and on both sides,

and has a posterior mesorectum attached to the concavity of the sacrum, which is a continuation of the

mesentery of the sigmoid colon. Below the peritoneal reflection the rectum is completely

extraperitoneal and fully surrounded by the mesorectum. The mesorectum is covered by a thin,

glistening membrane called the mesorectal fascia (MRF). Posteriorly, the mesorectum is separated from

the presacral fascia by an avascular plane of loose areolar tissue that is the natural plane of dissection

during a radical proctectomy. Anteriorly, the mesorectum is separated from the urogenital organs by a

remnant of the fusion of two layers of the embryologic peritoneal cul-de-sac known as Denonvilliers

fascia – an important anatomical landmark in rectal cancer surgery. Below the peritoneal reflection, the

lateral ligaments connect the mesorectum to the pelvic sidewall. The mesorectum tapers off distally as

the rectum funnels toward the anorectal ring, where the longitudinal layer of the muscularis propria

becomes the internal anal sphincter.

1787