Stomach and Duodenum
3.1 Gastritis
Terminology for gastritis is difficult to follow and in its current state is of little clinical value. No satisfactory system that correlates endoscopic appearance, histologic findings, and clinical sx exists. What the endoscopist calls gastritis often bears no relationship to what the pathologist calls gastritis, which in turn often bears no relationship to the pt’s clinical complaint. More than 80% of cases of histologic gastritis are caused by Hp (Endoscopy 1991;23:289). Other causes include autoimmune gastritis, drugs, allergy, Crohn’s, sarcoid, vasculitis, and radiation, as well as other bacteria, viruses, or fungi. The Sydney system is currently the most widely used classification system for histologic gastritis. It was developed by European pathologists. It separates gastritis into 3 categories: (1) nonatrophic, (2) atrophic, and (3) special (Am J Surg Pathol 1996;20:1161). A second system that is frequently used in the literature classifies gastritis as: (1) type A (atrophic gastritis of the body, often autoimmune associated with pernicious anemia), (2) type B (diffuse antral gastritis usually associated with Hp), (3) type AB (atrophic gastritis involving body and antrum and a known risk for gastric cancer), and (4) type C (chemical gastritis). The term gastritis should probably be considered only a histologic dx given the lack of correlation between endoscopy, histology, and the pt’s sx (Gut 1994;35:1172).
3.2 Helicobacter Pylori
Cause: Helicobacter pylori (Hp). Humans are the major reservoir of Hp, though human isolates can infect cats and cause gastritis (Helicobacter 1998;3:225). Fecal-oral transmission has been hypothesized but is difficult to prove (Am J Med 1996;100:12S). Hp has been recovered from stool (BMJ 1997;315:1489) and from drinking water (GE 1996;110:1031). Oral transmission from stomach contents seems plausible. Oral-oral transmission from a mouth reservoir in adults is noteasy to demonstrate. Dentists (at risk for oral exposures) do nothave an increased prevalence (Scand J Infect Dis 1995;27:149; Am J Gastro 1992;87:1728), but endoscopists do (presumably from infected endoscopes, pH probes, etc) (Am J Gastro 1994;89:1987) and nurses do (at risk for both oral and fecal exposure) (Arch IM 1993;153:708).
Epidem: Hp is found worldwide, but prevalence varies widely and is greatest in developing countries. Risk factors for infection include crowded or unsanitary conditions. In developing countries, the infection is probably acquired in childhood, and childhood prevalence is very high. In developed countries such as the U.S., the prevalence is <20% in those under age 50 yr, but rises abruptly after that. This suggests that childhood acquisition was likely in those infected prior to 1950 and that since 1950, the infection has been slowly acquired through life (Jama 1995;274:1064). In the U.S., the prevalence is greater in blacks and in those with less education. However, there is a large geographic variation that has notbeen well explained by socioeconomic status alone (Am J Med 1996;100:12S). Smoking has little, if any, effect on prevalence. A dose- dependent, protective effect of alcohol has been observed. Drinking coffee is associated with a greater prevalence (BMJ 1997;315:1489).
Pathophys: Hp survives in the acidic gastric environment by metabolizing urea into ammonia. This creates an alkaline microenvironment. The bug attaches to the surface mucus-secreting cells and lives in the mucus layer, thereby escaping normal clearance by the immune system (Jama 1995;274:1064). The organism produces toxins that damage the mucosa and induce the inflammatory mediators that cause the gastritis. More virulent strains secrete vacuolating cytotoxin A (vacA), a cytotoxin that causes vacuolization of gastric epithelium. Some of the same virulent strains (about 60% of the total) possess the cytotoxin-associated gene A (cagA) that is associated with more severe inflammation and risk of subsequent development of ulcer, gastric cancer, and atrophic gastritis (Nejm 1996;334:1018).
Sx: Acute infection (rarely diagnosed) causes dyspepsia or is asymptomatic. Chronic sx arise from complications of Hp infection. These complications are DU, GU, MALToma, gastric cancer, and possibly nonulcer dyspepsia.
Crs: Acute infection (documented by investigators intentionally infecting themselves! [Am J Gastroenterol 1987;82:192]) causes a dyspepsia-like illness, followed by achlorhydria, and a histologic gastritis with neutrophils. A period of 10-15 yr of chronic active gastritis follows, during which there is inflammation, but the stomach remains acidic. At the end stages, achlorhydria and atrophic gastritis develop. Of those infected, DU occurs in < 1% per yr. Only 20% of pts will ultimately develop a clinical consequence of infection (Lancet 1997;349:1020).
Cmplc: There have been 4 convincingly demonstrated consequences of Hp infection:
Duodenal ulcer: The most convincing evidence for the role of Hp as a cause of DU is the marked reduction in ulcer recurrence following eradication of Hp with antibiotics (eg, Lancet 1994;343:508; Nejm 1993;328:308). Until that evidence was available, causality was difficult to establish because of the lack of an animal model and because of the high prevalence of Hp compared to DU. It was also difficult initially to establish why an infection that is predominantly in the antrum causes DU (Jama 1994;272:65). However, it is likely that the Hp resides in areas of gastric metaplasia in the duodenum and induces ulceration by the inflammation induced by Hp cytotoxins. Acid hypersecretion is also likely the result of Hp infection, which reduces the number of somatostatin-producing cells in the antrum. This in turn lessens negative feedback on gastrin secretion, resulting in higher gastrin levels and greater acid output (Jama 1995;274:1064).
Gastric ulcer: The pathogenic role of Hp in GU is notas clear as its role in DU. However, Hp eradication speeds GU healing, reduces ulcer recurrence, and heals ulcers refractory to healing with acid suppression alone (Gut 1994;35:19).
Gastric cancer: Gastric cancer is more common in pts infected with Hp. Data from case control studies nested within prospective cohorts suggest a RR = 5.9 for noncardia gastric cancer in Hp-infected pts (Gut 2001;49:347). In a large Japanese cohort, gastric cancer was seen only in those with Hp infection and was most likely in those with atrophy, corpus gastritis, or intestinal metaplasia (Nejm 2001;345:784). Long-term Hp infection in Mongolian gerbils causes adenocarcinoma, adding strong support to this connection (GE 1998;115:642). Screening and treating Hp might be cost effective in preventing gastric cancer (Lancet 1996;348:150).
Mucosa-associated lymphoid tissue (MALT) lymphoma: Gastric mucosa does notordinarily contain organized lymphoid tissue. However, in response to Hp infection, lymphoid follicles develop and may evolve into low-grade B-cell lymphoma. About 90% of gastric MALTomas are associated with Hp infection, and Hp eradication cures the subset of pts with superficial and distal tumors (Ann IM 1999;131:88).
Nonulcer dyspepsia (NUD): Between 30% and 60% of pts with NUD have Hp infection. However, properly performed population-based studies do notshow a consistent association between Hp and NUD, and there is no convincing evidence of an effect of Hp on gastroduodenal motor or sensory function (GE 1997; 113:S67). The 2 best trials of rx have yielded conflicting results. One trial showed a benefit of Hp eradication (21% vs 7% sx relief at 1 yr [Nejm 1998;339:1869]). The other trial showed no benefit (26% vs 31% sx relief at 1 yr [Nejm 1998;339:1875]). However, the studies are similar in that the sx of NUD are notrelieved in 75% of pts in whom Hp is eradicated, suggesting a limited role for Hp in NUD. Most clinicians treat Hp-positive pts with NUD.
Other questionable associations: Recent data suggest an association between Hp infection and iron deficiency, but the cause and effect relationship is unproven. The argument has been made to test pts with unexplained iron deficiency for Hp and to treat if positive (Am J Gastro 2005;100:453). The ready availability of serologic tests seems to have inspired authors to search for non-GI-associated manifestations of Hp infection. Preliminary associations have been made between Hp infection and CAD, cerebrovascular disease, HTN, Raynaud’s phenomenon, migraine, sudden infant death syndrome, and ITP. However, the quality of the data is limited, and these associations remain uncertain, and in some cases biologically implausible (Arch IM 1999;159:925).
Lab: (Am J Gastro 1998;93:2330)
Whom to test: It only makes sense to test those pts who will be treated if the test is positive. Pts with DU, GU, and MALTomas should be tested. Gastric polyps in an Hp-infected stomach may disappear with Hp rx, so that group should be tested (Ann IM 1998;129:712). The data are conflicting on the utility of eradicating Hp in NUD, but most clinicians test and treat the positive pts. Some consider testing those with a family hx of gastric cancer, but the value of that is speculative. Testing may be considered prior to long-term NSAID use (p 71) or prior to long-term PPI use (p 39). No large group has yet endorsed screening and treating asymptomatic pts, but that day may come. The choice of test depends on the clinical circumstances.
Urease tests of bx specimens: These tests depend on the urease activity in endoscopically obtained bx and show positive results by a color change in an indicator when the urease is present. The 3 FDA-approved tests are Clotest, Pyloritek, and Hp-fast. Sensitivity is 88-95%,
and specificity is 95-100%. Sensitivity is reduced by prior use of PPIs (Aliment Pharmacol Ther 1996;10:289), H2RAs, bismuth, antibiotics, and alcohol consumption (Am J Gastroenterol 1997;92:1310). Bx from the antrum during active GI bleeding greatly reduces sensitivity of urease testing (Gastrointest Endosc 1999;49:302), so breath testing or serology should be used instead. Serology should also be used if testing is necessary in a pt who has been on antibiotics, PPIs, or H2RAs within the last month. Two antral bx are usually sufficient (Gastrointest Endosc 1994;40:342), but sensitivity can be improved using an additional bx from the body in pts on H2RAs or PPIs (Am J Gastro 1997;92:1310).
Histology: With this method, both the presence of Hp and the inflammation can be detected. If there is no gastritis histologically, then Hp infection is reliably excluded. Specialized stains such as Giemsa or immunostains (J Clin Pathol 2000;53:756) can be used if H&E is negative and gastritis is present. The same conditions decrease sensitivity for histology as for urease bx testing (see previous bullet item). The sensitivity is 93-96%, and specificity is 98-99%. One strategy to reduce cost and maximize sensitivity is to run urease testing first and to proceed with histology only if urease is negative.
Urea breath tests: (Am J Gastro 1998;93:2330; Gut 1994;35:723) The pt swallows a small amount of 13C (nonradioactive) or 14C (radioactive) urea. Hp urease converts the urea to bicarbonate and ammonia. The bicarbonate is converted to CO2, which is recovered in a breath sample and detected on mass spectroscopy (13C) or scintillation counting (14C). In the U.S., the FDA has approved a 13C test (Meretek) and a 14C test (Tri-Med). The 14C has the advantage of lower expense and the disadvantage that it cannot be given to children or pregnant women because of the radioactivity. Both tests perform well, with sensitivity of 90-96% and specificity of 88-98%. Antibiotics lower sensitivity, and the test should be performed at least 4 weeks after completion of eradication rx. PPIs may reduce sensitivity for 2 weeks, and H2RAs do so for 5 days (Ann IM 1998;129:547).
Fecal antigen test: These tests identify bacterial antigens in stool that suggest active infection. Both polyclonal and monoclonal tests are available, and the latter may be more specific when performed posttreatment (Helicobacter 2004;9:347). The tests have similar limitations to breath testing regarding recent use of PPIs, antibiotics, and bismuth. Overall, they work about as well as breath testing (regarding sensitivity and specificity) but have the difficulties associated with handling stool.
Serology: IgG for Hp by ELISA is 86-94% sensitive and 78-95% specific. It is readily available and is useful in detection when endoscopy is notbeing performed, but it does notdistinguish remote from active infection. A fall in titer after rx is notconsistent enough to be useful for determining response to rx, unless the titer falls to zero (Jama 1998;280:363). The positive predictive value of serology is low in populations with a low pretest probability of infection, and breath testing or fecal antigen testing is preferred.
Culture: Culture is notreadily available in many hospital labs and is notcommonly used in clinical practice. It is 80-98% sensitive, 100% specific. Its major use would be for detecting resistant strains in pts who have failed rx. As a practical matter, an alternative rx is chosen without sensitivity information.
Whole blood tests: These tests were designed for office use and have reduced sensitivity (67-88%), reduced specificity (75-91%), and reduced cost compared to serum-based ELISA testing. It does notseem worthwhile to use them given the small cost savings at the expense of accuracy.
Endoscopy: There are no endoscopic features that identify Hp. A grossly normal stomach at endoscopy can be infected, and bx are mandatory if Hp infection is to be diagnosed or excluded.
Rx: (Curr Gastroenterol Rep 1999;1:518; Am J Gastroenterol 1998;93:2330; J Gastroenterol 1998;33:48; Gut 1997;41:8; Am J Gastro 2007;102:1808)
Pts with a positive test should be treated if the test was obtained for an appropriate clinical reason. Since rx morbidity is minimal and the World Health Organization has called Hp a class 1 carcinogen, it becomes difficult notto treat the positives even if the test was obtained for inappropriate reasons. However, it has been argued that rx of all strains of Hp is premature since notall strains cause clinical disease (Lancet 1997;349:1020).
Characteristics of best therapies: A bewildering 1409 Hp rx arms have been identified in a meta-analysis of Hp rx (GE 1997;113:S131). In screening the literature for appropriate regimens, one needs to look for rx that is >80% effective in intent-to-treat analysis and >90% effective in per-protocol analysis, with reproducible results from different geographic areas. It should be well tolerated, simple, and cost effective. Regimens that are bid have greatly enhanced compliance. The regimens in the following list are all oral and use omeprazole 20 mg bid or lansoprazole 30 mg bid. Pantoprazole 40 mg po bid or rabeprazole 20 mg bid are probably equally effective (Aliment Pharmacol Ther 1999;13:741). Esomeprazole 40 mg qd is another alternative. PPIs are used to increase concentration of antibiotics in gastric juice, to lessen the viscosity of the mucus layer and to increase the stability of the antibiotics that act locally in the mucus layer (Am J Gastro 2009;104:26).
Unfortunately, resistance to rx is increasing and regimens are often less effective in practice nowthan when they were originally described. Guidelines from the ACG suggest (Am J Gastro 2007;102:1808):
For pts notallergic to penicillin who have notpreviously received a macrolide antibiotic: PPI + amoxicillin 1000 mg + clarithromycin 500 mg bid × 2 weeks. This is commonly called clarithromycin-based triple therapy.
For pts allergic to penicillin or who have previously received a macrolide antibiotic: Bismuth subsalicylate (Pepto Bismol) 525 mg qid + metronidazole 250 mg qid + tetracycline 500 mg qid for 10-14 days + acid suppression with PPI (qd or bid) or ranitidine 150 mg po bid. This is commonly called bismuth quadruple therapy.
For pts allergic to penicillin who have notpreviously received a macrolide antibiotic or who don’t tolerate bismuth quadruple therapy: PPI + metronidazole 500 mg + clarithromycin 500 mg po bid × 2 weeks.
Sequential first-line therapy: Because of the high frequency of failure of standard clarithromycin-based triple therapy (especially in Europe), sequential therapy was devised. In this rx, the first 5 days are treatment with a bid PPI and 1000 mg bid of amoxicillin. This is followed by 5 days of clarithromycin 500 mg bid and metronidazole 500 mg bid. Success rates are high (Am J Gastro 2008;103:2220) but U.S. data are lacking.
Treatment failures: (Am J Gastro 2007;102:1808) Pt compliance and antibiotic resistance are likely important factors in treatment failures. Unfortunately, antibiotic susceptibility data are notreadily available in most pts with treatment failure. Bismuth quadruple therapy × 7-14 days (see above) would be the first choice for pts who have failed clarithromycin-based therapies. Levofloxacin triple therapy with a bid PPI, amoxicillin 1 gm bid, and levofloxacin 500 mg qd for 10 days is a third-line regimen that is better tolerated and may be more effective than bismuth quadruple therapy (Am J Gastro 2006;101:488) but for which U.S. data are lacking.
Post-rx follow-up: Confirmation of eradication 4 weeks after completion of rx is recommended in pts with MALToma, ulcer disease, and early gastric cancer after resection. Follow-up testing should also be done in pts with persistent dyspepsia despite rx as part of the “test-and-treat” approach to dyspepsia (p 2). Its importance in other pt groups is less well established. Breath testing is ideal unless endoscopy is needed for another reason.
3.3 Peptic Ulcer
Lancet 2002;360:933; Am J Gastro 1997;92:1255
Epidem:
GU: Incidence estimated at 0.1%/yr; much higher in NSAID users. About 3% of GUs turn out to harbor malignancy (Nejm 1995;333:32).
DU: M:F incidence ratio is 2:1 (confounded by a higher smoking rate in men); higher in smokers, RR = 2.2 (J Clin Gastroenterol 1997;24:2), though notconfirmed in a large prospective study (Epidemiology 1997;8:420); higher in those with a family hx of DU (J Clin Gastroenterol 1995;20:104). It is unlikely that caffeine or alcohol are important risk factors (Epidemiology 1997;8:420). The literature on psychological risks for DU is a quagmire, with purported risks ranging from life events stress (J Clin Gastroenterol 1995;21:185; GE 1986;91:1370) to a personality that prompts one to sport a half-mustache on the upper lip (J Clin Gastroenterol 1992;15:96). It seems unlikely that psychological factors are of major, definable, clinical relevance. Diet plays no clear role, and the traditional suggestion of a bland diet is notmade on solid ground (especially with the suggestion that chili is protective for DU! [Dig Dis Sci 1995;40:576]).
Pathophys:
GU: GU results when gastric mucosal defenses are overwhelmed by the harsh, intraluminal gastric environment rich in acid and pepsin. Gastric epithelium remains intact in the stomach by a series of prostaglandin-dependent defenses, including (1) the semipermeable mucus layer, (2) bicarbonate in an unstirred water layer, and (3) regeneration and repair (Gut 1997;41:425; Gut 1993;34:580). Hp infection causes epithelial injury by production of cytotoxins, urease, and other factors that cause cell injury and death, leading to ulceration. NSAIDs inhibit mucosal cyclooxygenase, thereby reducing prostaglandin production, inhibiting mucus and bicarbonate production, and decreasing mucosal blood flow. This prevents cell replication and leads to ulcer formation.
DU: The initiating event in most DU is Hp infection. Antral gastritis from Hp infection results in a decrease in somatostatin-producing cells (D cells), whose role is to inhibit
gastrin-producing G cells. This in turn leads to the increased parietal cell mass (perhaps also increased due to genetic or other environmental factors) that results in the increase in acid secretion seen in DU pts. Areas of intestinal metaplasia form, possibly in response to increased acid, and these are subsequently colonized by Hp. Duodenitis results from release of inflammatory mediators from Hp, and in the presence of acid, an ulcer forms. NSAIDs presumably cause ulceration by reducing mucosal prostaglandin-mediated defenses. The foregoing theory does notaccount for DU that is seen in Hp-negative pts nottaking NSAIDs. A meta-analysis of rigorously designed U.S. trials demonstrated 20% ulcer recurrence rate in pts cleared of Hp, suggesting that in a subset of DU pts, Hp infection is notthe critical factor (Am J Gastroenterol 1998;93:1409). These so-called “idiopathic ulcers” may be more difficult to treat and may be more likely to result in complications (Am J Gastro 2002;97:2950).
Sx: There is no sx complex diagnostic for GU or DU, and sx do notreliably distinguish DU from GU or from nonulcer dyspepsia (GE 1993;105:1378). PUD often presents with epigastric pain, sometimes described as burning, aching, or as a hunger pain. Classically it is described as postprandial and may be relieved with antacids, but these features are highly variable. GU, especially NSAID-induced, can be asymptomatic or present with a complication such as bleeding or perforation in the absence of prior pain (Am Fam Phys 1997;55:1323). If the ulcer is near the pylorus, poor gastric emptying may result in reflux sx, vomiting, or early satiety. Sudden severe sx may indicate perforation, and constant pain may suggest penetration. Weight loss is more common with giant GU (>3 cm).
Si: Epigastric tenderness.
Crs:
GU: The vast majority of GUs will heal with initial rx. A small subgroup refractory to H2RAs will heal with a PPI (Aliment Pharmacol Ther 1996;10:381). If NSAIDs are eliminated and Hp is cured, recurrence is low. Chronic GU should lead to investigation for a missed gastric malignancy (adenocarcinoma or lymphoma) or surreptitious NSAID use.
DU: Before the advent of rx for Hp, DU was a chronic relapsing illness. With clearance of Hp, DU recurrence is uncommon (eg, Lancet 1994;343:508; Nejm 1993;328:308). However, recent analysis of U.S. trials suggests that the proportion of pts who relapse after Hp eradication may be as high as 20% (Am J Gastroenterol 1998;93:1409). At one time it was thought that ulcer disease largely burned itself out, but long-term follow-up studies of pts dx in the pre-Hp era show that 50% remain symptomatic (Gut 1996;38:812).
Cmplc: Bleeding, gastric outlet obstruction, perforation, penetration into adjacent structures without free perforation.
Diff Dx: See Dyspepsia (p 2). The diff dx of an endoscopically identified ulcer includes gastric adenocarcinoma, lymphoma, and Crohn’s disease.
Lab: CBC, CMP, amylase, lipase aid in diff dx. Testing for Hp is usually done at EGD for GU but may be done by serology or breath testing if DU was diagnosed by UGI rather than EGD.
X-ray: UGI barium x-ray can establish dx but is highly operator dependent and is much less sensitive and specific than EGD. A variety of radiographic signs have been proposed to separate the benign from the malignant GU. However, if the pt is medically fit, EGD should be performed to rule out malignancy if UGI shows GU (Jama 1996;275:622). UGI should generally be discouraged in the evaluation of dyspepsia unless EGD is thought unsafe or
unavailable. UGI should notbe used to dx DU (because it is a poor negative). However, if a DU is found on UGI, then EGD is notmandatory since the cancer risk in typical DU is negligible.
unavailable. UGI should notbe used to dx DU (because it is a poor negative). However, if a DU is found on UGI, then EGD is notmandatory since the cancer risk in typical DU is negligible.
Endoscopy: EGD is the gold standard for dx of ulcers. Ulcers are directly visualized, and features that predict risk for bleeding can be identified (p 75). A clean, punched-out appearance of a GU suggests that the ulcer is benign. At least 6 bx should be obtained to rule out malignancy. The endoscopic gross appearance is nota reliable negative, and 3 bx detect only 70% of cancers (GE 1982;82:228). Many experts believe that all GUs should be followed to healing by EGD (notUGI) to prove benignity by healing. Some have advocated foregoing follow-up EGD because initial EGD and bx is 99% accurate (Dig Dis Sci 1993;38:284). However, the yield of finding a curable cancer at follow-up EGD is high enough (1 cancer per 250 endoscopies [Scand J Gastroenterol 1991; 26:1193]) that the practice is felt appropriate by many (Dig Dis Sci 1994;39:442). DUs are usually identified in the first portion of the duodenum (often called the bulb or cap). Since they are at a very low risk for malignancy, follow-up EGD to document healing is rarely indicated.
Rx: (Jama 1996;275:622)
Suppression of acid: Though excess acid is notthe primary cause of GU, acid suppression heals ulcers, and about 90% of ulcers will be healed in 12 weeks. Acid suppression relieves sx. Cimetidine (800 mg qhs) or ranitidine (300 qhs) are cheaper but notas rapidly effective as PPIs. Consider initial use of a PPI at a high dose (eg, omeprazole 40 mg daily) in large or complicated ulcers (Am J Gastro 1996;91:2516). About 90-95% of DUs heal by 8 weeks with acid suppression, and GUs typically take 12 weeks because healing is slower and GUs are often larger.
NSAID ulcer: Stop NSAIDs. If NSAIDs are stopped, then an H2RA is appropriate rx unless the ulcer is very large. If NSAIDs cannot be stopped, the dosage might be reduced. If NSAIDs are continued, a PPI (eg, omeprazole 20-40 mg po qd) rather than an H2RA should be used to treat the ulcer (Nejm 1998;338:719).Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
