Dyspepsia is a chronic or recurrent burning discomfort or pain in the upper abdomen that may be caused by diverse processes, such as gastroesophageal reflux, peptic ulcer disease, and “nonulcer dyspepsia,” a heterogeneous category that includes disorders of motility, sensation and somatization. Gastric and esophageal malignancies are less common causes of dyspepsia. Careful history-taking allows accurate differential diagnosis of dyspepsia in only about half of patients. In the remainder, endoscopy can be a useful diagnostic tool, especially in patients whose symptoms are not resolved by Helicobacter pylori treatment or an empirical trial of acid-reducing therapy. Endoscopy should be performed at the outset in patients with dyspepsia and alarm features, such as weight loss or iron-deficiency anemia.
GASTROESOPHAGEAL REFLUX DISEASE (GERD)
When classic symptoms of gastroesophageal reflux are present, such as water brash and substernal heartburn, presumptive diagnosis and empirical treatment are often sufficient. Endoscopy is a sensitive test for diagnosis of esophagitis (Fig. 315-48), but it will miss nonerosive reflux disease (NERD) since some patients have symptomatic reflux without esophagitis. The most sensitive test for diagnosis of GERD is 24-h ambulatory pH monitoring. Endoscopy is indicated in patients with reflux symptoms refractory to antisecretory therapy; in those with alarm symptoms, such as dysphagia, weight loss, or gastrointestinal bleeding; and in those with recurrent dyspepsia after treatment that is not clearly due to reflux on clinical grounds alone. Endoscopy should be considered in patients with long-standing (≥10 years) GERD, as they have a sixfold increased risk of harboring Barrett’s esophagus compared to patients with <1 year of reflux symptoms.
Causes of esophagitis. A. Severe reflux esophagitis with mucosal ulceration and friability. B. Cytomegalovirus esophagitis. C. Herpes simplex virus esophagitis with target-type shallow ulcerations. D. Candida esophagitis with white plaques adherent to the esophageal mucosa.
Barrett’s esophagus is specialized columnar metaplasia that replaces the normal squamous mucosa of the distal esophagus in some persons with GERD. Barrett’s epithelium is a major risk factor for adenocarcinoma of the esophagus and is readily detected endoscopically, due to proximal displacement of the squamocolumnar junction (Fig. 315-5). A screening EGD for Barrett’s esophagus should be considered in patients with a chronic (≥10 year) history of GERD symptoms. Endoscopic biopsy is the gold standard for confirmation of Barrett’s esophagus and for dysplasia or cancer arising in Barrett’s mucosa.
Periodic EGD with biopsies is recommended for surveillance of patients with Barrett’s esophagus. Endoscopic resection (EMR or ESD) and/or ablation are performed when high-grade dysplasia or intramucosal cancer are found in the Barrett’s mucosa. Although guidelines recommend observation and surveillance of low-grade dysplasia in Barrett’s mucosa, recent evidence suggests that endoscopic treatment may be appropriate in select patients. Radiofrequency ablation (RFA) is the commonest ablative modality used for endoscopic treatment of Barrett’s esophagus, and other modalities, such as cryotherapy, are also available.
Peptic ulcer classically causes epigastric gnawing or burning, often occurring nocturnally and promptly relieved by food or antacids. Although endoscopy is the most sensitive diagnostic test for peptic ulcer, it is not a cost-effective strategy in young patients with ulcer-like dyspeptic symptoms unless endoscopy is available at low cost. Patients with suspected peptic ulcer should be evaluated for H. pylori infection. Serology (past or present infection), urea breath testing (current infection), and stool tests are noninvasive and less costly than endoscopy with biopsy. Patients aged >50 and those with alarm symptoms or persistent symptoms despite treatment should undergo endoscopy to exclude malignancy.
Nonulcer dyspepsia may be associated with bloating and, unlike peptic ulcer, tends not to remit and recur. Most patients describe persistent symptoms despite acid-reducing, prokinetic, or anti-Helicobacter therapy, and are referred for endoscopy to exclude a refractory ulcer and assess for other causes. Although endoscopy is useful for excluding other diagnoses, its impact on the treatment of patients with nonulcer dyspepsia is limited.
About 50% of patients presenting with difficulty swallowing have a mechanical obstruction; the remainder has a motility disorder, such as achalasia or diffuse esophageal spasm. Careful history-taking often points to a presumptive diagnosis and leads to the appropriate use of diagnostic tests. Esophageal strictures (Fig. 315-49) typically cause progressive dysphagia, first for solids, then for liquids; motility disorders often cause intermittent dysphagia for both solids and liquids. Some underlying disorders have characteristic historic features: Schatzki’s ring (Fig. 315-50) causes episodic dysphagia for solids, typically at the beginning of a meal; oropharyngeal motor disorders typically present with difficulty initiating deglutition (transfer dysphagia) and nasal reflux or coughing with swallowing; and achalasia may cause nocturnal regurgitation of undigested food.
Peptic esophageal stricture associated with esophagitis.
Schatzki’s ring at the gastroesophageal junction.
When mechanical obstruction is suspected, endoscopy is a useful initial diagnostic test, since it permits immediate biopsy and/or dilation of strictures, masses, or rings. The presence of linear furrows and multiple corrugated rings throughout a narrowed esophagus (feline esophagus) should raise suspicion for eosinophilic esophagitis, an increasingly recognized cause for recurrent dysphagia and food impaction (Fig. 315-51). Blind or forceful passage of an endoscope may lead to perforation in a patient with stenosis of the cervical esophagus or a Zenker’s diverticulum, but gentle passage of an endoscope under direct visual guidance is reasonably safe. Endoscopy can miss a subtle stricture or ring in some patients.
Eosinophilic esophagitis with multiple circular rings of the esophagus creating a corrugated appearance, and an impacted grape at the narrowed esophagogastric junction. The diagnosis requires biopsy with histologic finding of >15–20 eosinophils/high-power field.
When transfer dysphagia is evident or an esophageal motility disorder is suspected, esophageal radiography and/or a video-swallow study are the best initial diagnostic tests. The oropharyngeal swallowing mechanism, esophageal peristalsis, and the lower esophageal sphincter can all be assessed. In some disorders, subsequent esophageal manometry may also be important for diagnosis.
Various causes of dysphagia are amenable to endoscopic therapy. Benign strictures, rings, and webs can be dilated using a through-the-scope balloon (Fig. 315-52) or a polyvinyl dilator passed over a guide wire. In some instances, thin fibrotic strictures may respond to needle-knife electroincision (Fig. 315-53) when they prove refractory to dilation. Esophageal covered stents can be used to palliate dysphagia from malignant obstruction (Fig. 315-54), and flexible endoscopic myotomy is an option for Zenker’s diverticulum (Video V5-19). Recent advances in submucosal endoscopy have enabled the development of procedures, such as POEM (Video V5-20) and POET (Video V5-21) for the management of achalasia and select subepithelial esophageal tumors, respectively.
Endoscopic management of peptic stricture. A. Peptic stricture. B. Through-the-scope balloon dilation of stricture. C. Improvement in luminal diameter postdilation.
Endoscopic management of an esophagogastric anastomotic stricture. A. Recurrent anastomotic stricture despite periodic balloon dilation. B. Needle-knife electroincision of stricture. C. Improvement in luminal opening post therapy.
Palliation of malignant dysphagia. A. Obstructing distal esophageal cancer. B. Palliative stent placement.
ENDOSCOPIC TREATMENT OF OBESITY
The majority of Americans are overweight or obese, and obesity-associated diabetes has become a major public health problem. Bariatric surgery is the most effective weight-loss intervention, and it has been shown to decrease long-term mortality in obese persons, but many patients do not undergo surgery. Endoscopic treatments for obesity have been developed and include insertion of an intragastric balloon or duodenojejunal bypass liner, placement of a percutaneous gastric tube for aspiration of gastric contents after meals, or endoscopic sleeve gastroplasty, which utilizes endoscopic suturing to narrow the lumen of the gastric body (Video V5-22). Prospective trials show that these treatments induce total body weight loss of 7–20% and varying degrees of glycemic control. Additional endoscopic modalities are undergoing initial clinical trials. The long-term efficacy of endoscopic bariatric treatment is currently unknown.
TREATMENT OF MALIGNANCIES
Endoscopy plays an important role in the treatment of gastrointestinal malignancies. Early-stage malignancies limited to the superficial layers of the gastrointestinal mucosa may be resected using the techniques of EMR (Video V5-4) or ESD (Video V5-5). Photodynamic therapy (PDT) and RFA are effective modalities for ablative treatment of high-grade dysplasia and intramucosal cancer in Barrett’s esophagus (Video V5-23). Gastrointestinal stromal tumors can be removed en bloc by EFTR (Video V5-3). In general, endoscopic techniques offer the advantage of a minimally invasive approach to treatment but rely on other imaging techniques (such as CT, MRI, positron emission tomography [PET], and EUS) to exclude distant metastases or locally advanced disease better treated by surgery or other modalities. The decision to treat an early-stage gastrointestinal malignancy endoscopically is often made in collaboration with a surgeon and/or oncologist.
Endoscopic palliation of gastrointestinal malignancies relieves symptoms and in many cases prolongs survival. Malignant obstruction can be relieved by endoscopic stent placement (Figs. 315-17, 315-54, and 315-55; Videos V5-16, V5-17), and malignant gastrointestinal bleeding can often be palliated endoscopically as well. EUS-guided celiac plexus neurolysis may relieve pancreatic cancer pain.
Biliary and duodenal self-expanding metal stents (SEMS) for obstruction caused by pancreatic cancer. A. Endoscopic retrograde cholangiopancreatography (ERCP) demonstrates a distal bile duct stricture (arrow). B. A biliary SEMS is placed. C. Contrast injection demonstrates a duodenal stricture (arrow). D. Biliary and duodenal SEMS in place.
ANEMIA AND OCCULT BLOOD IN THE STOOL
Iron-deficiency anemia may be attributed to poor iron absorption (as in celiac sprue) or, more commonly, chronic blood loss. Intestinal bleeding should be strongly suspected in men and postmenopausal women with iron-deficiency anemia, and colonoscopy is indicated in such patients, even in the absence of detectable occult blood in the stool. Approximately 30% will have large colonic polyps, 10% will have colorectal cancer, and a few patients will have colonic vascular lesions. When a convincing source of blood loss is not found in the colon, upper gastrointestinal endoscopy should be considered; if no lesion is found, duodenal biopsies should be obtained to exclude sprue (Fig. 315-56). Small bowel evaluation with capsule endoscopy (Fig. 315-57), CT or MR enterography, or device-assisted enteroscopy may be appropriate if both EGD and colonoscopy are unrevealing.
Scalloped duodenal folds in a patient with celiac sprue.
Capsule endoscopy images of a mildly scalloped jejunal fold (left) and an ileal tumor (right) in a patient with celiac sprue. (Images courtesy of Dr. Elizabeth Rajan; with permission.)
Tests for occult blood in the stool detect hemoglobin or the heme moiety and are most sensitive for colonic blood loss, although they will also detect larger amounts of upper gastrointestinal bleeding. Patients with occult blood in normal-appearing stool should undergo colonoscopy to diagnose or exclude colorectal neoplasia, especially if they are over 50 years of age or have a family history of colonic neoplasia. Whether upper endoscopy is also indicated depends on the patient’s symptoms.
The small intestine may be the source of chronic intestinal bleeding, especially if colonoscopy and upper endoscopy are not diagnostic. The utility of small bowel evaluation varies with the clinical setting and is most important in patients in whom bleeding causes chronic or recurrent anemia. In contrast to the low diagnostic yield of small bowel radiography, positive findings on capsule endoscopy are seen in 50–70% of patients with suspected small intestinal bleeding. The most common finding is mucosal vascular ectasia. CT or MR enterography accurately detects small bowel masses and inflammation, and are also useful for initial small bowel evaluation. Deep enteroscopy may follow capsule endoscopy for biopsy of lesions or to provide specific therapy, such as argon plasma coagulation of vascular ectasias (Fig. 315-58).
A. Mid-jejunal vascular ectasia identified by double-balloon endoscopy. B. Ablation of vascular ectasia with argon plasma coagulation.
COLORECTAL CANCER SCREENING
The majority of colon cancers develop from preexisting colonic adenomas, and colorectal cancer can be largely prevented by the detection and removal of adenomatous polyps (Video V5-24). The choice of screening strategy for an asymptomatic person depends on personal and family history. Individuals with inflammatory bowel disease, a history of colorectal polyps or cancer, family members with adenomatous polyps or cancer, or certain familial cancer syndromes (Fig. 315-59) are at increased risk for colorectal cancer. An individual without these factors is generally considered at average risk.
Numerous colon polyps in a patient with familial adenomatous polyposis syndrome.
Screening strategies are summarized in Table 315-3. While stool tests for occult blood have been shown to decrease mortality rate from colorectal cancer, they do not detect some cancers and many polyps, and direct visualization of the colon is a more effective screening strategy. Either sigmoidoscopy or colonoscopy may be used for cancer screening in asymptomatic average-risk individuals. The use of sigmoidoscopy was based on the historical finding that the majority of colorectal cancers occurred in the rectum and left colon, and that patients with right-sided colon cancers had left-sided polyps. Over the past several decades, however, the distribution of colon cancers has changed in the United States, with proportionally fewer rectal and left-sided cancers than in the past. Large American studies of colonoscopy for screening of average-risk individuals show that cancers are roughly equally distributed between left and right colon and half of patients with right-sided lesions have no polyps in the left colon. Visualization of the entire colon thus appears to be the optimal strategy for colorectal cancer screening and prevention.
TABLE 315-3Colorectal Cancer Screening Strategies ||Download (.pdf) TABLE 315-3 Colorectal Cancer Screening Strategies
| ||Choices/Recommendations ||Comments |
|Average-Risk Patients |
|Asymptomatic individuals ≥50 years of age (≥45 years of age for African Americans) ||Colonoscopy every 10 yearsa ||Preferred cancer prevention strategy |
| ||Annual FIT or FOBT, multiple take-home specimen cards ||Cancer detection strategy, does not detect most polyps; colonoscopy if results are positive |
| ||CT colonography every 5 years ||Colonoscopy if results are positive |
| ||Flexible sigmoidoscopy every 5 years ||Does not detect proximal colon polyps and cancers; colonoscopy if results are positive |
| || |
Double-contrast barium enema every 5 years
Less sensitive than colonoscopy or CT colonography, misses some cancers and polyps; colonoscopy if results are positive
| || |
Stool DNA test every 3 years
Does not detect many polyps; colonoscopy if results are positive
|Personal History of Polyps or CRC |
|1 or 2 small (<1 cm) adenomas with low-grade dysplasia ||Repeat colonoscopy in 5–10 years ||Assuming complete polyp resection. Interval may vary based on prior history, family history |
|3–10 adenomas, or any high-risk adenomab ||Repeat colonoscopy in 3 years; subsequent colonoscopy based on findings ||Assuming complete polyp resection |
|>10 adenomas ||Repeat colonoscopy in <3 years based on clinical judgment ||Consider evaluation for FAP or HNPCC; see recommendations below |
|Piecemeal removal of a sessile polyp ||Exam in 2–6 months to verify complete removal || |
|Small (<1 cm) hyperplastic polyps of sigmoid and rectum ||Repeat colonoscopy in 10 years ||Those with hyperplastic polyposis syndrome merit more frequent follow-up |
Sessile serrated adenoma/polyp <10 mm, without dysplasia
Repeat colonoscopy in 5 years
Sessile serrated adenoma/polyp ≥10 mm or with dysplasia, or ≥2 serrated polyps
Repeat colonoscopy in 3 years
|Serrated polyposis syndrome merits more frequent follow-up |
|Incompletely removed serrated polyp ≥1 cm ||Exam in 2–6 months to verify complete removal || |
|Colon cancer ||Evaluate entire colon around the time of resection, then repeat colonoscopy in 1 year ||Subsequent colonoscopy in 3 years if the 1-year examination is normal |
|Inflammatory Bowel Disease |
|Long-standing (>8 years) ulcerative pancolitis or Crohn’s colitis, or left-sided ulcerative colitis of >15 years’ duration ||Colonoscopy with biopsies every 1–2 years ||Consider chromoendoscopy or other advanced imaging techniques for detection of flat dysplasia during colonoscopy |
|Family History of Polyps or CRC |
|First-degree relatives with only small tubular adenomas ||Same as average risk || |
|Single first-degree relative with CRC or advanced adenoma at age ≥60 years ||Colonoscopy every 10 years starting at age 40 || |
|Single first-degree relative with CRC or advanced adenoma at age <60 years, OR two first-degree relatives with CRC or advanced adenomas at any age ||Colonoscopy every 5 years beginning at age 40 years or 10 years younger than age at diagnosis of the youngest affected relative, whichever is earlier || |
|FAP ||Sigmoidoscopy or colonoscopy annually, beginning at age 10–12 years ||Consider genetic counseling and testing |
|HNPCC ||Colonoscopy every 2 years beginning at age 20–25 years (or 10 years younger than the youngest affected first-degree relative) until age 40, then annually thereafter ||Consider histologic evaluation for microsatellite instability in tumor specimens of patients who meet modified Bethesda criteria; consider genetic counseling and testing |
Virtual colonoscopy (VC) is a radiologic technique that images the colon with CT following rectal insufflation of the colonic lumen. Computer rendering of CT images generates an electronic display of a virtual “flight” along the colonic lumen, simulating colonoscopy (Fig. 315-60). Findings detected during VC often require subsequent conventional colonoscopy for confirmation and treatment.
Virtual colonoscopy image of a colon polyp (arrow). (Image courtesy of Dr. Jeff Fidler; with permission.)
Most cases of diarrhea are acute, self-limited, and due to infections or medication. Chronic diarrhea (lasting >6 weeks) is more often due to a primary inflammatory, malabsorptive, or motility disorder, is less likely to resolve spontaneously, and generally requires diagnostic evaluation. Patients with chronic diarrhea or severe, unexplained acute diarrhea often undergo endoscopy if stool tests for pathogens are unrevealing. The choice of endoscopic testing depends on the clinical setting.
Patients with colonic symptoms and findings such as bloody diarrhea, tenesmus, fever, or leukocytes in stool generally undergo sigmoidoscopy or colonoscopy to assess for colitis (Fig. 315-8). Sigmoidoscopy is an appropriate initial test in most patients. Conversely, patients with symptoms and findings suggesting small bowel disease, such as large-volume watery stools, substantial weight loss, and malabsorption of iron, calcium, or fat, may undergo upper endoscopy with duodenal aspirates for assessment of bacterial overgrowth and biopsies for assessment of mucosal diseases, such as celiac sprue.
Many patients with chronic diarrhea do not fit either of these patterns. In the setting of a long-standing history of alternating constipation and diarrhea dating to early adulthood, without findings such as blood in the stool or anemia, a diagnosis of irritable bowel syndrome may be made without direct visualization of the bowel. Steatorrhea and upper abdominal pain may prompt evaluation of the pancreas rather than the gut. Patients whose chronic diarrhea is not easily categorized often undergo initial colonoscopy to examine the entire colon and terminal ileum for inflammatory or neoplastic disease (Fig. 315-61).
Bright red blood passed with or on formed brown stool usually has a rectal, anal, or distal sigmoid source (Fig. 315-62). Patients with even trivial amounts of hematochezia should be investigated with flexible sigmoidoscopy and anoscopy to exclude polyps or cancers in the distal colon. Patients reporting red blood on the toilet tissue only, without blood in the toilet or on the stool, are generally bleeding from a lesion in the anal canal; careful external inspection, digital examination, and proctoscopy with anoscopy may be sufficient for diagnosis in such cases.
Internal hemorrhoids with bleeding (arrow) as seen on a retroflexed view of the rectum.
About 20% of patients with pancreatitis have no identified cause after routine clinical investigation (including a review of medication and alcohol use, measurement of serum triglyceride and calcium levels, abdominal ultrasonography, and CT or MR). Endoscopic assessment leads to a specific diagnosis in the majority of such patients, often altering clinical management. Endoscopic investigation is particularly appropriate if the patient has had more than one episode of pancreatitis.
Microlithiasis, or the presence of microscopic crystals in bile, is a leading cause of previously unexplained acute pancreatitis and is sometimes seen during abdominal ultrasonography as layering sludge or flecks of floating, echogenic material in the gallbladder. EUS may identify previously undetected microlithiasis.
Previously undetected chronic pancreatitis, pancreatic malignancy, or pancreas divisum may be diagnosed by either ERCP or EUS. Autoimmune pancreatitis is often suspected on the basis of CT, MR, or serologic findings, but it may first become apparent during EUS and may require EUS-guided pancreatic biopsy for histologic diagnosis.
Severe pancreatitis often results in pancreatic fluid collections. Symptomatic pseudocysts and areas of walled-off pancreatic necrosis can be drained into the stomach or duodenum endoscopically, using transpapillary and transmural endoscopic techniques. Pancreatic necrosis can be treated by direct endoscopic necrosectomy (Video V5-2) via an endoscopically created transmural drainage site.
Local staging of esophageal, gastric, pancreatic, bile duct, and rectal cancers can be obtained with EUS (Fig. 315-19). EUS with fine-needle aspiration (Fig. 315-20) currently provides the most accurate preoperative assessment of local tumor and nodal staging, but it does not detect many distant metastases. Details of the local tumor stage can guide treatment decisions including resectability and need for neoadjuvant therapy. EUS with transesophageal needle biopsy may also be used to assess the presence of non-small-cell lung cancer in mediastinal nodes.