Effectiveness and economic impact of screening for colorectal cancer by mass fecal occult blood testing

OBJECTIVES

Fecal occult blood testing has been shown to reduce mortality from colorectal cancer in large randomized, controlled trials conducted in the United States, Denmark, and the United Kingdom, and mathematical simulation modeling found it to be cost-effective relative to other health care services. Before making a concerted effort to implement mass fecal occult blood testing based on this evidence alone, however, we considered it prudent to critically re-evaluate the effectiveness and economic impact of screening in the US population as a whole.

METHODS

To assess the effectiveness of screening, we projected published outcomes from each of the three large randomized controlled trials of fecal occult blood testing to the US population, as if each clinical trial had been done in the population as a whole. We then determined the resource costs of detection and treatment that would be associated with the outcomes predicted from each trial.

RESULTS

More than 1 million colorectal cancers could be expected to arise over 10 yr in the cohort of US residents eligible to enter a screening program in 1997, and trial outcomes indicate that > or = 60% of these cancers would be fatal. If the 60-67% compliance rate of the population-based randomized controlled trials were achieved, a fecal occult blood testing program would detect 30% of known colorectal cancers and save 100,000 lives over 10 yr. Screening would incur total costs of $3-4 billion over 10 yr, or $2,500 per life-year saved.

CONCLUSIONS

Mass fecal occult blood testing is cost-effective, and, although not inexpensive, many would consider the total cost acceptable. Even with a concerted effort to achieve compliance, however, the effectiveness of fecal occult blood testing would be limited to saving the lives of < or = 15% of those who otherwise would die from their cancer in the first 10 yr after beginning mass screening. The limitations of fecal occult blood testing suggest the need to further evaluate the role of endoscopy in screening, and to develop more effective, noninvasive screening tools.

Colorectal cancer screening

Colorectal cancer is an important problem in the United States, with over 130,000 new cases and 55,000 deaths each year. There is now strong evidence that screening for colorectal cancer with fecal occult blood testing can decrease mortality, and additional evidence that removing benign adenomas can decrease cancer incidence. Evidence-based screening guidelines depend on colorectal cancer risk. Individuals at higher risk because of a personal or family history deserve more intensive screening than asymptomatic individuals over age 50.

Aspirin and nonsteroidal anti-inflammatory agents and risk for colorectal adenomas

BACKGROUND & AIMS

Aspirin and nonsteroidal anti-inflammatory drugs have been reported to protect against the development of colorectal cancer. Because adenomas are precursors to most colorectal cancers, the aim of this study was to examine the relationship of these medications to the risk for colorectal adenomas in a colonoscopy-based case-control study.

METHODS

Study participants were drawn from patients who underwent colonoscopy at the University of North Carolina Hospitals. Medication use was assessed by telephone using a comprehensive list of prescription and nonprescription drugs as well as questions about dietary and lifestyle factors that might be relevant for adenoma development.

RESULTS

There were 210 patients with adenomas and 169 adenoma-free controls. After adjusting for potential confounders, regular users were half as likely to currently have adenomas compared with nonusers (adjusted odds ratio, 0.56; 95% confidence interval, 0.34-0.92). Regular users who stopped medication at least 1 year before colonoscopy were still protected (adjusted odds ratio, 0.59; 95% confidence interval, 0.21-1.67), although small numbers make this conclusion tentative. The protective effects of aspirin and the nonaspirin nonsteroidal anti-inflammatory drugs were similar.

CONCLUSIONS

The results suggest that aspirin and nonsteroidal anti-inflammatory drugs cause early disruption of the adenoma-carcinoma sequence. The challenge for the future will be to learn more about dose, duration, and mechanism of action.

Myogenic mechanism for peristalsis in opossum smooth muscle esophagus

We studied the propagation of phasic contractions initiated by tetraethylammonium (TEA, 1-10 mM), high K+ concentration (10-30 mM), and bethanechol (10(-6) to 10(-2) M) in a whole organ in vitro preparation of the opossum smooth muscle esophagus. TEA initiated phasic contractions that began at all sites along the smooth muscle esophagus and propagated in both directions with a velocity similar to that of primary peristalsis. Blockade of neural transmission by tetrodotoxin (TTX, 10(-7) M) did not prevent contraction propagation. Although a majority of contractions initiated by TEA did not propagate the full length of the esophageal specimen, with the addition of TTX most contractions initiated by TEA did propagate the full specimen length in either direction. High K+ concentration and bethanechol elicited propagated contractions similar to those initiated by TEA. We conclude that 1) a myogenic mechanism exists for propagation of contractions along the smooth muscle esophagus and 2) intramural inhibitory nerves modulate the extent of myogenic propagation in the ascending as well as descending direction. We suggest that esophageal peristalsis may occur by myogenic propagation of contractions that are normally initiated in the proximal smooth muscle esophagus by excitatory nerves. Intramural inhibitory nerves may inhibit retrograde propagation as well as mediate descending inhibition in advance of the peristaltic wave.

Myogenic oscillatory mechanism for opossum esophageal smooth muscle contractions

We evaluated the control of phasic contractions in opossum esophageal circular smooth muscle by determining the contractile response in vitro to agents that cause membrane depolarization and excitation by different mechanisms. Transverse muscle strips taken from different sites along the length of the smooth muscle esophagus were exposed to progressively increasing concentrations of tetraethylammonium (1-30 mM), K+ (4.6-30 mM), or bethanechol (10(-6) to 10(-2) M). In normally inactive esophageal circular smooth muscle, tetraethylammonium and high K+ concentration elicited phasic contractions that were not blocked by atropine and tetrodotoxin. Bethanechol, an M2 muscarinic receptor agonist that acts selectively on smooth muscle, elicited phasic contractions that were not blocked by tetrodotoxin. We conclude that a latent myogenic oscillatory mechanism for control of phasic contractions exists in esophageal circular smooth muscle and that it may be activated by nonspecific excitation of the smooth muscle membrane. We suggest that this myogenic oscillatory mechanism is likely excited and modulated by nerves.

Regional esophageal distribution and clearance of refluxed gastric acid

Regional differences in the esophageal distribution and clearance of refluxed gastric acid was studied in seven asymptomatic volunteers and seven patients with reflux esophagitis. Intraluminal pH was recorded for 3 postprandial hours from the distal, middle, and proximal esophagus on two separate occasions (with subjects in upright and supine positions). With the subjects in a supine position, about half of the acid reflux episodes reached the proximal esophagus in patients as well as in controls. This percentage decreased to 25% in patients and 29% in controls when they were upright. In both groups, the pH drops in the distal esophagus were significantly greater than in the proximal esophagus for both the supine and upright positions. In both patients and controls, a 4-5-fold greater acid exposure occurred in the distal esophagus, than in the proximal esophagus. In both patient and control groups, acid exposure time, as well as the number of reflux episodes in the distal esophagus, were significantly greater than that of the proximal esophagus (P less than 0.05). Spontaneous acid clearance time in the distal esophagus was significantly longer than that of the proximal esophagus in both positions (P less than 0.05) for both subject groups. In conclusion, regional differences exist in the exposure of the esophageal mucosa to refluxed gastric acid. These regional differences are more pronounced when subjects are upright than supine. Regional differences also exist in esophageal acid clearance, with clearance taking longer in the distal esophagus than in the proximal esophagus. The net effect of these phenomena is that acid exposure time in the distal esophagus is greater than that in the remainder of the esophagus.

Abnormal esophageal motility. An analysis of concurrent radiographic and manometric findings

The findings of concurrent esophageal videofluoroscopy and manometry in 15 patients with major disturbances of esophageal motor function were evaluated and the data were analyzed from a fluid mechanical perspective. Each of 153 fluoroscopic barium swallow sequences was analyzed on a swallow-by-swallow basis. Two distinct pressure domains were identified: intrabolus pressure and pressure within a bolus-free contracting esophageal segment. Analyses in terms of these pressure domains showed specific and consistent correlations between the radiographic and manometric findings. Radiography was insensitive to contractions occurring in esophageal segments devoid of bolus fluid, whereas manometry was insensitive to contractions that did not occlude the lumen. It is concluded that using fluid mechanical principles of bolus transport allows meaningful comparison of esophageal motility as recorded by radiography and intraluminal manometry. However, the inherent limitations in the range of physical phenomena recorded by each modality make these techniques complementary for evaluating esophageal motor function.

Relationship of intraluminal pH and pressure within the lower esophageal sphincter

To determine the relationship between lower esophageal sphincter (LES) intraluminal pressure and its intraluminal pH, we studied six healthy volunteers. We recorded intraluminal pressure and pH concurrently using rapid pull-through, slow pull-through, and station pull-through, as well as, rapid push-through and slow push-through techniques. The results showed that LES length was 35 +/- 4 (SE) mm by RPT and 30 +/- 3 mm by SPT. The pressure was maximal in the proximal half of the LES. On rapid pull-throughs, the intraluminal pH rose from about 1.5 to reach a value of about 2.5 at the peak of the high pressure zone. With continued withdrawal into the esophageal body, the recorded pH rose minimally to about 3-4. On push-throughs, the pH recorded along the LES was the same as that of the esophageal body. After the electrode cleared the LES, the pH abruptly fell to gastric pH. During station pull-through with the electrode 0.5-1.5 cm proximal to the distal LES margin, transient pH drops were observed with swallows. With rapid swallows, however, the pH drop did not occur until after the last swallow. This finding suggests that the pH drops with swallows were due to axial LES movement rather than gastroesophageal acid reflux. We conclude that 1) the relationship of the gastroesophageal pH transition zone and LES high pressure zone is better defined by a sphincter push-through than a pull-through; 2) the transition between gastric and esophageal pH occurs either at or slightly distal to the distal LES margin; and 3) swallow-induced axial LES movement may cause spurious recording of acid reflux when the pH probe is positioned within the distal half of the LES.

Oesophageal acidification does not increase lower oesophageal sphincter pressure

We studied the effect of distal oesophageal acidification on lower oesophageal sphincter (LOS) pressure in normal human volunteers and in cats. The distal oesophagus was acidified by intraluminal injection of 0.1N HCl (pH 1.2). The LOS pressure was recorded by a sleeve device while pharyngeal and oesophageal pressures were monitored by nonperfused, water filled catheters. In normal human subjects, distal oesophageal acidification did not elicit a change in LOS pressure. In anaesthetised cats, injection of acid into the distal oesophagus elicited immediate LOS relaxation followed by a secondary peristaltic sequence. With propagation of the peristaltic sequence into the LOS, the LOS pressure abruptly increased 20-100 mmHg and gradually returned to the preinjection value over 15-180 s. In the one instance in which we were able to acidify the distal oesophagus without evoking secondary peristalsis, the LOS pressure did not change. Injection of saline into the distal oesophagus evoked a response in the LOS and oesophageal body that was indistinguishable from that seen with acid. We conclude that contrary to common belief, distal oesophageal acidification itself does not affect LOS pressure in man or the cat.

Intramural neural control of opossum sphincter of Oddi

We evaluated the intramural neural control of the opossum sphincter of Oddi (SO) in an in vitro preparation. Force transducers were used to record contractions at four sites along the sphincter segment. To stimulate intramural nerves, 10- to 120-s trains of pulses (4-10 V amplitude, 0.5 ms duration, and 5 Hz frequency) were delivered to one of three electrode pairs implanted along the SO. Electrical stimulation in the proximal, mid, or distal SO elicited phasic contractions that invariably originated in the proximal SO and propagated antegrade along the entire length of the sphincter segment. Stimulus-evoked contractions resembled spontaneous antegrade peristaltic contractions, but occurred at a higher rate (12-20/min). Atropine completely blocked this excitatory response to nerve stimulation. After atropine, nerve stimulation in the proximal, mid, or distal SO abolished spontaneous contractions at and distal to the site of stimulation for the duration of the stimulus. The inhibitory response to nerve stimulation was completely blocked by tetrodotoxin but was unaffected by phenoxybenzamine, tolazoline, or propranolol. We conclude that 1) the opossum SO is innervated by intramural cholinergic excitatory nerves and nonadrenergic noncholinergic inhibitory nerves; 2) cholinergic excitatory nerves are organized in ascending neural pathways, whereas nonadrenergic noncholinergic inhibitory nerves descend along the length of the SO; and 3) these neural pathways may modulate SO peristalsis in vivo and participate in ascending excitatory and descending inhibitory reflexes.

Effects of morphine on the human sphincter of Oddi

The effects of morphine on intraluminal pressures recorded from the sphincter of Oddi (SO) at endoscopic retrograde cholangiopancreatography in 19 patients who were without evidence of biliary or pancreatic disease were studied. Morphine was given in four successive doses of 2.5, 2.5, 5, and 10 micrograms/kg iv at five minute intervals. Morphine in subanalgesic doses increased the frequency of SO phasic pressure waves to a maximum of 10-12/min, caused the phasic waves to occur simultaneously along the sphincter segment, increased phasic wave amplitude from 72 (26) (SE) to 136 (31) mmHg, and increased SO basal pressure from 10 (1) to 29 (9) mmHg (p less than 0.05). The effects of morphine on the SO are mediated by more than one opioid receptor type, as naloxone competitively antagonised the increase in phasic wave frequency induced by morphine, but did not affect the increase in SO basal pressure elicited by morphine. When given after naloxone, morphine decreased phasic wave amplitude, an inhibitory effect that is normally masked by morphine's dominant naloxone sensitive excitatory effect. Mu receptors do not appear to be involved in control of spontaneous SO motor function, as naloxone alone did not affect SO motor activity. The excitatory effects of morphine on the SO are not mediated by cholinergic nerves, as they were not blocked by atropine. Cholinergic nerves, however, may have a role in regulating spontaneous SO motor function because atropine alone depressed phasic wave activity and basal pressure. Although morphine does cause 'spasm' of the human SO, its effects are more complex than is commonly believed.

Treatment of acute nontoxic megacolon during colonoscopy: tube placement versus simple decompression

The study compares the efficacy of colonoscopic decompression versus decompression and tube placement in the treatment of Ogilvie's syndrome. Nine patients were treated with a single colonoscopic decompression which resulted in four recurrences. In contrast, there were no recurrences observed in 11 patients who underwent decompression and subsequent tube placement (p less than 0.05). There was no morbidity observed from either decompression or tube placement. Tube placement added less than 10 min of additional procedure time to the colonoscopy. The tube utilized in this study was an enteroclysis tube with sideholes cut in the distal 20 cm. The tube was easily inserted over a Teflon-coated flexible guide wire inserted through the colonoscope into the cecum following decompression. This study demonstrates that colonoscopic decompression followed by tube placement is the preferred treatment modality for acute nontoxic megacolon.

Salivary response to esophageal acid in normal subjects and patients with reflux esophagitis

We studied the effect of esophageal acid perfusion on salivation in patients with reflux esophagitis and in normal subjects. Serial 10-min saliva collections were obtained by expectoration during perfusion of the esophagus with water, and then 0.1 N HCl (pH 1.2) for 50 min or 0.01 N HCl (pH 2.1) for 120 min. Within 1-5 min of beginning 0.1 N HCl perfusion, all 8 patients with esophagitis developed heartburn accompanied by an increase in saliva flow. By the time the severity of heartburn required discontinuation of HCl perfusion (10-40 min), saliva flow had increased nearly fourfold. With 0.1 N HCl perfusion, 8 of 10 volunteers developed mild heartburn after 22 +/- 3 min (mean +/- SE), whereas 0.01 N HCl induced heartburn in 6 of 10 volunteers after 57 +/- 12 min of perfusion. Saliva flow increased concurrently with the onset of heartburn and doubled in those volunteers who developed heartburn. Saliva flow did not change in those volunteers who were without heartburn. We conclude that esophageal acid perfusion unaccompanied by heartburn does not affect salivation. However, saliva flow increases concurrently with the onset of heartburn, a phenomenon called "water brash" when clinically evident. The increased saliva flow that accompanies heartburn may act as an endogenous antacid that serves as a protective response to symptomatic gastroesophageal reflux.

Comparison of pseudoachalasia and achalasia

Malignancies involving the gastric cardia or distal esophagus can result in a clinical syndrome termed pseudoachalasisa that mimics idiopathic achalasia. If not promptly recognized, pseudoachalasia can result in inappropriate pneumatic dilatation of the lower esophageal sphincter segment and delay appropriate treatment of the underlying malignancy. During the past 14 years, six patients with pseudoachalasia and 161 patients with primary idiopathic achalasia were encountered. Pseudoachalasia occurred mainly in the elderly and represented about 9 percent of these patients over 60 years of age with suspected achalasia. Five of the six pseudoachalasia cases were secondary to adenocarcinoma that originated in the gastric fundus, and one was caused by a squamous cell carcinoma of the distal esophagus. Conventional esophageal manometry did not discriminate achalasia from pseudoachalasia. On the other hand, esophagogastroscopy with biopsy resulted in a diagnosis of pseudoachalasia in five of these cases and in 24 of 32 cases reported previously. Ominous endoscopic findings are mucosal ulceration or nodularity, reduced compliance of the esophagogastric junction, or an inability to pass the endoscope into the stomach. Radiographic evaluation, particularly in conjunction with amyl nitrite inhalation, was also useful in discriminating pseudoachalasia from primary achalasia. It is concluded that pseudoachalasia generally mimics idiopathic achalasia imperfectly and can usually be diagnosed prior to surgery by fastidious endoscopic and radiographic examination.

Esophageal acid clearance

Esophageal acid clearance is an important defense against the development of reflux esophagitis. Acid clearance normally occurs in two sequential steps. One or two peristaltic sequences empty virtually all acid volume from the esophagus, leaving a minimal residual that sustains a low pH, and then the residual acid is neutralized by swallowed saliva. The ability of saliva to neutralize acid is primarily due to bicarbonate. Saliva flow normally increases concurrent with the onset of heartburn. Hypersalivation with heartburn, commonly referred to as waterbrash, may be a protective response to gastroesophageal reflux. Gastroesophageal reflux and esophageal acid clearance normally do not occur during sleep. Acid clearance may be delayed markedly in the event that a subject falls asleep before clearance of acid from the esophagus is complete. Esophageal acid clearance may be prolonged by abnormal esophageal emptying or impaired salivation in patients with gastroesophageal reflux disease.

Integrity of cholinergic innervation to the lower esophageal sphincter in achalasia

The human lower esophageal sphincter (LES) is believed to be innervated by nonadrenergic, noncholinergic inhibitory nerves, and cholinergic excitatory nerves. In idiopathic achalasia, LES relaxation is abnormal because the inhibitory nerves to the sphincter are either absent or functionally impaired. The integrity of cholinergic excitatory nerves to the LES, however, has not been thoroughly evaluated. In 27 patients with untreated idiopathic achalasia, and 21 healthy volunteers, we investigated the hypothesis that postganglionic cholinergic nerves to the LES are functionally intact in achalasia. The LES responses to atropine, edrophonium, methacholine, amyl nitrite, and pentagastrin were assessed. In 2 achalasia patients, patterns of fasting motor activity in the LES were investigated during overnight manometric studies. Resting LES pressure was significantly greater in the achalasia patients, 41 +/- 4 mmHg (mean +/- SE), than in the normal subjects, 20 +/- 2 mmHg. Atropine significantly reduced LES pressure in both groups by 30%-75%. Edrophonium increased LES pressure in the achalasia patients but had negligible effect on the normal subjects. The LES in achalasia patients exhibited an increased sensitivity to both methacholine and pentagastrin compared with the normal subjects. In both patients who underwent an overnight manometric study, the LES exhibited cyclic phasic contractile activity synchronous with gastric contractions during the migrating motor complex. We conclude that the study findings support the hypothesis that postganglionic cholinergic LES innervation in achalasia patients is either normal or only minimally impaired, in contrast to the marked impairment of the inhibitory nerves governing LES relaxation.

Control mechanism of spontaneous in vitro contractions of the opossum sphincter of Oddi

We evaluated the control mechanism of peristaltic contractions in the opossum sphincter of Oddi (SO) by means of an in vitro preparation. At each of four sites spaced uniformly along the sphincter segment, a force transducer recorded contractions while a monopolar electrode recorded myoelectric activity. Spontaneous myoelectric and contractile activity occurred in 15 of the 20 intact SO specimens studied. Electrical recordings showed characteristic control waves and response activity. Each control wave was invariably accompanied by a phasic contraction, irrespective of whether or not response activity was superimposed on the control wave. The predominant motor activity of the SO was antegrade peristalsis. Retrograde peristalsis occurred when antegrade peristalsis failed to traverse the entire sphincter. Spontaneous SO phasic contractions were not antagonized by tetrodotoxin. Muscle rings sectioned from the SO exhibited spontaneous phasic contractions with a proximal-to-distal gradient of intrinsic contraction frequencies. We conclude that a) the frequency of SO phasic contractions is determined by control wave frequency, b) spontaneous SO peristalsis is myogenic in origin and may be modeled by a linear array of bidirectionally coupled relaxation oscillators, c) the predominance of antegrade peristalsis may be explained by a high-frequency oscillator in the proximal SO that drives the slower, more distal oscillators, d) retrograde peristalsis is initiated by an ectopic oscillator in the distal SO when antegrade contractions fail to propagate the entire length of the SO, and e) ectopic SO contractions can propagate retrograde when the more proximal oscillators are not in their absolutely refractory state.

Multipeaked esophageal peristaltic pressure waves in patients with diabetic neuropathy

We evaluated esophageal function in 14 consecutive insulin-dependent diabetic patients who had evidence of peripheral and autonomic neuropathy, but no esophageal symptoms. One to three contraction waves immediately followed a primary peristaltic contraction wave. The majority of these multipeaked pressure complexes consisted of two peaks. Multipeaked contractions were observed with all peristaltic waves in 12 of the 14 diabetic patients and with most of the peristaltic complexes in the remaining 2 patients. Multipeaked peristaltic waves were present in 1 of 6 diabetic patients without neuropathy, in 1 of 100 consecutive nondiabetic patients referred for suspected esophageal disease, and in 1 of 10 healthy volunteers. Double-peaked peristaltic pressure complexes in the nondiabetic control subjects differed from those present in the insulin-dependent patients by their low incidence and by a tendency to be limited to the distal esophagus. Pharmacologic responses to edrophonium and atropine suggested a possible increased cholinergic tone as the basis of the multipeaked peristaltic waves in diabetics with autonomic neuropathy.

Relationship between swallow rate and salivary flow

Recent studies indicate that swallow-induced, primary peristalsis is a major determinant of normal esophageal acid clearance. However, factors that regulate the rate of spontaneous swallowing in normal subjects are incompletely understood. We postulated that the rate of salivary flow influences the rate of spontaneous swallows. To test this hypothesis, we did a total of 60 studies measuring salivary flow or the rate of spontaneous swallowing in 10 healthy volunteers, age 10-30 years. Saliva was collected by expectoration. Swallow rate was recorded by a small, transnasal catheter stationed in the pharynx and also, in some circumstances, by cervical electrodes. On separate sessions, we evaluated the effect of five test manipulations on salivary flow and swallowing rate, respectively. The test manipulations consisted of: (1) pharyngeal intubation, (2) sucking of a dummy lozenge, (3) sucking of a peppermint lozenge, (4) bethanechol injection (5 mg subcutaneously), and atropine administration (12 micrograms/kg intravenously). Pharyngeal intubation caused a small, but significant increase in the rate of salivation and spontaneous swallows. Sucking of a peppermint lozenge caused a sixfold increase in salivary flow while nearly doubling the swallowing rate whereas the dummy lozenge caused only a modest increase in salivary flow and swallowing. Cholinergic stimulation by bethanechol elicited a substantial increase in salivary flow and swallowing rate. In contrast, atropine caused a significant decrease in both salivary flow and swallowing. We conclude that in awake, normal subjects the rate of spontaneous swallows is influenced directly by salivary flow. Because oral lozenges substantially increase both swallowing rate and salivary flow, such agents merit investigation as a potentially useful ancillary treatment for the relief of heartburn.

Effect of esophageal emptying and saliva on clearance of acid from the esophagus

We studied clearance of acid from the esophagus and esophageal emptying in normal subjects. A 15-ml bolus of 0.1 N hydrochloric acid (pH 1.2) radiolabeled with [99mTc]sulfur colloid was injected into the esophagus, and the subject swallowed every 30 seconds. Concurrent manometry and radionuclide imaging showed nearly complete emptying of acid from the esophagus by an immediate secondary peristaltic sequence, although esophageal pH did not rise until the first swallow 30 seconds later. Esophageal pH then returned to normal by a series of step increases, each associated with a swallow-induced peristaltic sequence. Saliva stimulation by an oral lozenge shortened the time required for acid clearance, whereas aspiration of saliva from the mouth abolished acid clearance. Saliva stimulation or aspiration did not affect the virtually complete emptying of acid volume by the initial peristaltic sequence. We conclude that esophageal acid clearance normally occurs as a two-step process: (1) Virtually all acid volume is emptied from the esophagus by one or two peristaltic sequences, leaving a minimal residual amount that sustains a low pH, and (2) residual acid is neutralized by swallowed saliva.

Determinants of esophageal acid clearance in normal subjects

In this study, we evaluated factors that affect esophageal acid clearance in normal subjects. A 15-ml bolus of 0.1 N HCl (pH 1.2) was injected into the esophagus, and the subject then swallowed every 30 s. Manometric and pH monitoring demonstrated that esophageal acid clearance occurred by a series of step increases in pH, each associated with a swallow-induced peristaltic sequence. Between peristaltic sequences, pH increase was minimal. Saliva stimulation by oral lozenge greatly improved acid clearance, while oral aspiration of saliva abolished the step increases in esophageal pH and markedly delayed acid clearance. Replacement of aspirated saliva with a bicarbonate solution reproduced the step increases in esophageal pH and restored acid clearance toward normal, while replacement with water alone failed to improve acid clearance. Similar to the effect of the oral lozenge, bethanechol (5 mg subcutaneously) improved esophageal acid clearance, but this improvement was reversed by oral aspiration of saliva, which markedly delayed acid clearance. A change from the recumbent to the sitting position tended to improve acid clearance slightly, but this improvement was not statistically significant. We concluded that in normal subjects (a) swallowing carries saliva into the esophagus and peristalsis empties intraesophageal fluid into the stomach, (b) the neutralization of acid by saliva carried into the esophagus with each swallow accounts for the occurrence of acid clearance by step increases in pH, (c) the improvement in acid clearance with bethanechol is due to saliva stimulation, and (d) gravity contributes little to esophageal acid clearance in the presence of normal peristaltic stripping waves.

Mechanisms of gastroesophageal reflux in patients with reflux esophagitis

We evaluated the mechanisms of gastroesophageal reflux in 10 patients with reflux esophagitis and compared the results with findings from 10 controls. The patients had more episodes of reflux (35 +/- 15 in 12 hours, as compared with 9 +/- 8 in the controls) and a lower pressure of the lower esophageal sphincter (13 +/- 8 mm Hg as compared with 29 +/- 9 in the controls) (P less than 0.001). Reflux occurred by three different mechanisms: transient complete relaxation of the lower esophageal sphincter, a transient increase in intra-abdominal pressure, or spontaneous free reflux associated with a low resting pressure of the lower esophageal sphincter. In controls 94 per cent of reflux episodes were caused by transient sphincter sphincter relaxation. In the patients 65 per cent of episodes of reflux accompanied transient sphincter relaxation, 17 per cent accompanied a transient increase in intra-abdominal pressure, and 18 per cent occurred as spontaneous free reflux. The predominant reflux mechanism in individual patients varied: some had normal resting sphincter pressure and reflux that occurred primarily during transient sphincter relaxation, whereas others with low resting sphincter pressures had spontaneous free reflux or reflux that occurred during an increase in intra-abdominal pressure.