Colonoscopy: A Guide to Endoscopic Screening and Therapy

Colonoscopy has become the preferred method for colorectal cancer screening in the United States. For patients at average risk, the procedure is typically performed every 10 years, beginning at age 50. Colonoscopy reduces colorectal cancer mortality by 2 mechanisms: diagnosis of cancer at an earlier (presymptomatic) stage and prevention of cancer development by resection of premalignant adenomas. In addition to colorectal cancer screening, colonoscopy is also used to diagnose and treat GI bleeding, obstruction, colitis, and other disorders of the colon and rectum (Table 1).

In the United States, most colonoscopies are performed by gastroenterologists and surgeons. However, adequately trained primary care physicians, physician assistants, nurse practitioners, and nurses perform colonoscopy in a variety of set-tings throughout the country. Case series from several medical centers have documented the safety and efficacy of colonoscopies performed by experienced family physicians in these centers.^1-3

PREPARATION OF THE BOWEL
Although colonoscopy can be performed without a bowel preparation in certain emergency situations, adequate visualization demands a thorough cleansing of the colon to remove solid and liquid stool. A variety of regimens can be used. These usually consist of a period of dietary restriction in combination with ingestion of a preparation agent.

Dietary restriction. An example of typical dietary restriction is consumption of a light breakfast of white bread and jelly the day before the procedure, followed by a clear liquid diet for the rest of the day. Alternatively, many centers prefer that patients take only clear liquids for the entire day before the procedure. Because most patients in the United States receive conscious sedation for colonoscopy, they should not eat during the 6 hours before the procedure or have any liquids for at least 2 hours before.

Discontinuation of certain medications. Iron supplements should be stopped at least 2 days before colonoscopy because iron can impair visualization. Decisions regarding the use of platelet inhibitors and anticoagulants can be complex. Many endoscopists are wary of performing polypectomy in patients who are currently taking these agents. Guidelines are available; however, in many cases the decision to discontinue therapy may need to be individualized.

Preparation agents. The afternoon before the procedure, the preparation agent is ingested. One commonly used regimen is 1 gallon of polyethylene glycol, which is consumed by drinking 1 cup every 10 minutes. Flavored polyethylene glycol preparations, sulfate-free preparations, and half-volume preparations with various additives are also commonly used. Polyethylene glycol is a nonabsorbable, osmotically balanced solution. Because it causes comparatively few electrolyte derangements and fluid shifts, it is relatively safe even for patients with renal insufficiency and other comorbidities. Major disadvantages of polyethylene glycol are the large volume that must be consumed and its poor palatability.

Another popular preparation is sodium phosphate (phosphosoda), which is available as a concentrated liquid or as tablets. Many patients find sodium phosphate more palatable than polyethylene glycol, but this agent can result in kidney injury, fluid shifts, and electrolyte imbalances.

SEDATION
Because colonoscopy is often moderately painful, it is usually performed under conscious sedation. Midazolam with either fentanyl or meperidine are commonly used, but deeper sedation with propofol is favored by some practitioners. Because propofol has a short half-life, patients typically recover quickly from the procedure; however, apnea can occur, and the treating team must be prepared to ventilate the patient if necessary.

A registered nurse typically administers sedation and monitors the patient. However, if propofol is used, then an anesthesiologist may be preferable or even required. A second assistant is usually present to assist with endoscopic accessories, to administer abdominal pressure to help advance the scope effectively, and to manage the endoscope shaft if you prefer to delegate this task.

FIRST STEPS. Start the procedure by performing a digital rectal examination while applying lubricant to the anus—usually with the patient lying on his or her left side. Then insert the endoscope into the rectum. To allow visualization, insufflate air into the rectum by covering the air/water button; if the lens is covered by debris, pressing down on the air/water button will clean it by streaming a small jet of water across the lens. Then advance the endoscope while maintaining a view of the lumen in front of the scope. Two dials on the scope allow you to steer the tip up and down and left and right.

DEALING WITH LOOPS
It is rarely possible to advance safely through the entire colon by simply steering the tip toward the center of the lumen and advancing the scope. This is because, after traversing several turns, the scope will no longer move forward smoothly. When the scope is simply advanced in this situation, there is a tendency for the loops in the colon to enlarge and for the tip to remain stationary rather than move forward as desired. This phenomenon is colloquially referred to as “looping.”

Figure 1 – These 2 pairs of images, obtained from a 3-dimensional magnetic scope device, are examples of scope looping and subsequent successful reduction of loops. All are standard anteroposterior views with the patient’s head at the top of the screen and the patient’s left side on the right of the screen. (A) Here the scope tip is at the splenic flexure; there is significant looping in the sigmoid colon. (B) After withdrawing a short length of scope while applying torque in a clockwise direction, the sigmoid loop has been eliminated and the scope tip has moved forward just past the flexure. (C) This is an example of more significant looping in the sigmoid colon. (D) Here the loop has been reduced by withdrawing a short length of scope while applying torque in a counterclockwise direction. The scope tip is then easily advanced to the proximal transverse colon.

A variety of maneuvers can be used to reduce the loops. One is to withdraw the scope while applying torque to the shaft—in either a clockwise or a counterclockwise direction (Figure 1). When this maneuver (called a reduction) is successful, it results in a pleating of the colon over the scope, like a curtain being fitted over a curtain rod. Once the pleating occurs, it is often possible to start advancing the scope again without the loop redeveloping.

Another technique for avoiding looping is to reposition the patient onto his back, stomach, or right side. Having the assistant apply pressure strategically to various locations on the abdomen can also be helpful.

With these techniques, the cecum can be reached in about 98% of patients, although some patients require changing to a thinner scope (pediatric scope) or occasionally a longer scope (enteroscope). In most cases the cecum is reached in 3 to 10 minutes; however, significantly more time is needed in a minority of patients.

REACHING THE CECUM

Figure 2 – The puckered lips of the ileocecal valve are visible in the lower left of this photograph. The base of the cecum is seen at the center of the photo, in the distance.	Figure 3 – This image shows the appendiceal orifice as seen on entering the cecum. The orifice typically appears as a slit or a small opening visible at the base of the cecum.

The cecum is identified as a blind end to the colon, with the appendiceal orifice visible, even in patients who have had an appendectomy (Figure 2). The ileocecal valve entrance is visible on the proximal margin of the fold nearest the appendix (Figure 3). It is important to advance the scope into the cecum to maximize visualization of the cecum and to definitively identify the ileocecal valve and appendiceal orifice landmarks to ensure examination of the entire colon. Obtain photographs of the cecum and ileocecal valve for documentation purposes.

WITHDRAWAL OF THE SCOPE

Figure 4 – Diverticula are commonly seen even in asymptomatic patients. Two diverticula are visible near the top of this photograph. These are small and easily distinguished from the true lumen of the colon, which is visible on the lower right.	Figure 5 – The head of this pedunculated polyp is erythematous and consists of adenomatous tissue. The stalk is a tan color and contains feeding vessels that are cauterized during snare polypectomy.
Figure 6 – The scalloped mass on the left side of this image is an adenocarcinoma that penetrates at least into the deep submucosa and is appropriately treated by surgical resection.	Figure 7 – In ulcerative colitis, the surface of the colon is erythematous and friable. Here diminutive inflammatory nodules are visible in the center and at top right.

Figure 8 – A retroflexed view is important for optimal visualization of the distal rectum. Here the black scope shaft is visible entering through the anus with the instrument tip maximally deflected to obtain a backward view.

Once at the cecum, gradually withdraw the scope, carefully inspecting the colon as you do so. Usually considerable back and forth motion is required to inspect behind all of the many folds of the colon, as well as to lavage and suction fluid residue in areas that have not been adequately cleansed (a second button on the scope controls suction). A withdrawal time of at least 6 minutes is recommended to ensure an adequate examination. Common findings, such as diverticulosis, a pedunculated polyp, adenocarcinoma, and ulcerative colitis, are shown in Figures 4 through 7.

In the rectum, it is customary to retroflex the scope (turn the tip 180 degrees so that it faces the distal rectum) (Figure 8); this area is difficult to inspect fully without retroflexion. Once the examination is finished, suction excess air from the rectum, remove the scope, and monitor the patient for 30 minutes or more while the sedative wears off. Encourage the patient to pass gas to relieve the bloating that is nearly universal after colonoscopy.

 FINDINGS AND TREATMENT

Figure 9 – In this photograph, a snare is positioned around a diminutive polyp. The snare can now be retracted into the white plastic sheath, cutting the polyp off. It is also possible to apply electrocautery through the snare while cutting the polyp, although cautery is not necessary for diminutive lesions such as this.

The focus in screening colonoscopy is on finding premalignant adenomas. Most adenomas are polypoid; thus, they are usually visualized as protruding bumps of varying sizes surrounded by flat normal mucosa (see Figure 5). Some, however, are flat or depressed lesions (Box).

Small adenomas. Diminutive adenomas, less than 5 mm in diameter, are the most common. These are usually easily removed using either a biopsy forceps or a snare. Many experts now favor using a snare without cautery for diminutive lesions (Figure 9).

Large adenomas. Larger polyps, which can be sessile or pedunculated, are usually removed using a snare with monopolar cautery. The snare is positioned around the lesion and then retracted into a plastic sheath while monopolar cautery is applied (electric current travels through the snare to a grounding pad applied to the patient’s buttock or back). The current helps the snare cut through the tissue and also ensures some destruction of adenomatous cells at the resection margin. More advanced techniques are also widely used; these include clip application to prevent bleeding and submucosal saline injection to separate the lesion from the underlying muscle of the colon wall before resection. Since not all endoscopists are experienced in these techniques, patients may be referred to specialty centers for resection.

Nonadenomatous polyps. These are also frequently found in the colon. In the rectum and sigmoid, diminutive hyperplastic polyps are exceedingly common (Figure 10). These lesions have no malignant potential. Adenomatous polyps are usually redder than hyperplastic polyps and typically have a surface appearance reminiscent of brain, with sulci and gyri (see Figure 5).

Figure 10 – A diminutive hyperplastic polyp is seen slightly to the left of the center of this image. It is slightly paler than the surrounding normal mucosa and contains numerous circular crypts, indicating that it is hyperplastic rather than a precancerous adenoma

Hyperplastic polyps are typically pale and usually have a regular pattern of circular or stellate pits at the surface (see Figure 10). Before the development of highresolution scopes, many of which now have various proprietary imaging modes that accentuate the surface features of the mucosa, it was difficult to accurately distinguish between hyperplastic and adenomatous polyps, so all visualized polyps were typically removed. This practice is evolving as endoscopic diagnosis progresses, and more selective removal may soon become the norm.

Carcinomas. Colorectal adenocarcinomas are usually obvious, appearing as large masses (often several centimeters in diameter) that frequently have an ulcerated surface. Multiple forceps biopsies are generally performed to facilitate tissue diagnosis before surgery. More unusual malignancies, such as lymphomas, can also be diagnosed with forceps biopsies. Rarely, infectious ulcers, Crohn disease, or ischemia can have the appearance of an ulcerated mass; in this setting, the tissue samples obtained on biopsy do not demonstrate malignancy. Other findings. These include diverticula, colitis, arteriovenous malformations, and surgical anastomoses.

COMPLICATIONS
The 2 main complications of colonoscopy are perforation and bleeding. Bleeding usually occurs within a few days of polypectomy. A second colonoscopy is often performed to investigate the source of the bleeding, and an ulcer is commonly seen at the initial polypectomy site. Visible vessels in the ulcer can be clipped or cauterized to prevent rebleeding. Severe blood loss can occur, and some patients may require blood transfusion.

When an attempt is made to advance the endoscope in the presence of a loop, perforation can occur if the pressure exerted on the colon wall (often the sigmoid) by the looped portion of the scope increases sufficiently. It is less common for the scope tip to perforate the colon directly by advancing into and through a diverticulum, a tumor, or a turn in the colon. Overinflation of the colon can also lead to perforation, but this is quite rare with modern equipment.

Perforations can also occur during snaring of polyps if the cut is too deep. In addition, they can occur as a delayed complication of cautery. A commonly quoted overall perforation rate for colonoscopy is 1 in 2000,^4,5 although reported rates vary widely. The risk is significantly higher in patients with certain conditions, such as acute diverticulitis or fulminant colitis; in fact, the presence of such a condition is a contraindication to colonoscopy (Table 2). Perforation is typically treated by emergency surgery.

Other complications may include aspiration or hypoventilation as a result of sedation, as well as cardiovascular events during the time that antiplatelet or anticoagulant medications are withheld.

In addition to these direct complications, many authorities also consider the development of advanced colon cancer within a few years of colonoscopy to be a type of complication. Unfortunately, as a result of missed lesions on colonoscopy, incomplete removal of visualized lesions, and possibly occasional rapid development of new tumors, cancer does develop in some patients between regularly scheduled colonoscopies. It is prudent to advise patients of these issues as part of the informed consent process.

TRAINING IN COLONOSCOPY
Proper training in colonoscopy generally requires a minimum of 100 supervised procedures in order to achieve acceptable success rates in the advancement of the scope to the cecum. The same number of procedures is typically required of clinicians who are experienced at sigmoidoscopy. Basic polypectomy training is generally integrated with training in diagnostic colonoscopy, since small polyps are encountered in a sizable percentage of cases.

A substantial portion of colonoscopy instruction is devoted to learning how and when to reduce loops, reposition the patient, and apply abdominal pressure in an effort to reach the cecum efficiently. Commercially available simulators and colon models provide a good introduction to the mechanics of handling and manipulating the colonoscope, and should be considered before supervised training on patients.

Flat and Depressed Colonic Lesions

Figure  – This flat neoplasm with a central depression is an example of a lesion that is more difficult to visualize and remove during colonoscopy. The central erythematous, slightly depressed region contained high-grade dysplasia. The lesion was removed in 1 piece by snare after submucosal saline injection.

These lesions are often particularly difficult to visualize. Most are slightly redder and slightly more elevated than the surrounding normal mucosa (Figure). The most worrisome have a slight central depression in the center. A significant percentage of flat—and particularly of depressed lesions—harbor high-grade dysplasia or early carcinoma.

 

A recent study at the Veterans Administration (VA) Palo Alto Health Care System demonstrated the prevalence and importance of these easy-to-miss lesions.⁶ Of 1819 patients in the VA Palo Alto study who underwent elective colonoscopy, 9% had flat or depressed lesions, and 0.8% had a flat or depressed lesion with carcinoma in situ or early invasive carcinoma.

Flat and depressed lesions were nearly 10 times more likely to contain carcinoma in situ or early invasive cancer than were polyps (odds ratio, 9.8). This study highlights the need for all endoscopists to be aware of these lesions. Detection of flat lesions requires excellent bowel preparation and unhurried examination of the mucosa. Newer high-resolution colonoscopes have substantial built-in image processing, which facilitates visualization of these lesions.