- Email: [email protected]
Esophageal solid bolus transit: studies using concurrent videofluoroscopy and manometry
THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 1999 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 94, No. 6, 1999 ISSN 0002-9270/99/$20.00 PII S0002-9270(99)00182-3
ORIGINAL CONTRIBUTIONS
Esophageal Solid Bolus Transit: Studies Using Concurrent Videofluoroscopy and Manometry Philippe Pouderoux, M.D., Guoxiang Shi, M.D., Ph.D., Roger P. Tatum, M.D., and Peter J. Kahrilas, M.D. Department of Medicine, Division of Gastroenterology and Hepatology, Northwestern University Medical School, Chicago, Illinois
OBJECTIVE: Our aim was to assess the efficacy and mechanism of solid bolus transit through the esophagus. METHODS: Eight healthy volunteers were studied with concurrent manometry and videofluoroscopy while swallowing 5 ml liquid barium, a 5– 6 mm diameter bread ball, and 4 g chewed bread in both a supine and upright posture. As many as four successive swallows were performed until clearance was achieved. RESULTS: The esophageal clearance of liquid barium was 100% with the first swallow. Clearance of the unchewed bread ball occurred with the first swallow in only 6.7% of trials in the upright posture and 5.9% in the supine posture. After four swallows, clearance was 100% and 52.9% in the upright and supine postures, respectively. Chewed bread was more readily cleared than unchewed bread, with 100% clearance after two swallows in the upright posture and 91% clearance after four swallows in the supine posture. The most common locus of bread stasis was at the aortic arch and carina. The bread boluses were noted to move more effectively when localized in the head as opposed to the tail of the bolus composite. Nonocclusive contractions often occurred at the bolus tail despite the increased peristaltic amplitude seen with the chewed bread. Failed peristalsis, a frequent cause for solid clearance failure, was observed during 30% of all bread swallows. This usually occurred distal to the stopping point of the bolus, suggesting it to be the result rather than the cause of impaired transit. CONCLUSIONS: Although infrequently perceived by these normal subjects and in contradistinction to liquid clearance, bread is rarely cleared from the esophagus with a single swallow. Mastication and an upright posture facilitate the esophageal transport of solids. Bolus composition and impaired bolus transit alter the amplitude and conductance of peristalsis. Manometric data pertaining to liquid clearance through the esophagus do not readily apply to bread. (Am J Gastroenterol 1999;94:1457–1463. © 1999 by Am. Coll. of Gastroenterology)
INTRODUCTION The determinants of esophageal liquid clearance (which relate to acid clearance) are well described (1– 4), whereas
those of esophageal solid food transit (which relate more closely to dysphagia) are less well defined. Fluoroscopy and manometry are the two techniques most commonly used for evaluating esophageal motor function. Fluoroscopy images the transit of a swallowed bolus to define the effectiveness of esophageal clearance, and manometry quantifies esophageal contractility. Fluoroscopic studies have revealed that a solid bolus traverses the esophagus more slowly than liquid but the method does not allow determination of the cause of this delay (5, 6). Slowed solid bolus transit has been corroborated with scintigraphic studies but localization and characterization of transit abnormalities are less specific than with fluoroscopic techniques (7, 8). Food (9 –13) and special solid bolus ingestion (14, 15) have been evaluated using esophageal manometry as well. These studies have shown that manometric recordings during solid bolus swallows differ from those obtained during water swallows. Although these different responses may be relevant to the prolonged transit of solids seen during fluoroscopy, the mechanism of delayed transit of solids through the esophagus cannot be defined without correlating imaged bolus transit with the manometric events. Such correlation was attempted in a case report by Breumelhof et al. in which fluoroscopy combined with manometry during bread swallows not only confirmed the existence of significant transit impairment, but also revealed esophageal spasm that had been missed during conventional manometry, ambulatory manometry, and radiographic studies using liquid barium (16). However, no study to date has used this technique to systematically evaluate the normal mechanism of esophageal solid clearance. Thus, the aim of this study was to use concurrent manometry and videofluoroscopy to compare the esophageal clearance of chewed and unchewed bread with that of liquid, and to correlate bolus clearance with manometric recordings in normal individuals.
MATERIALS AND METHODS Eight healthy volunteers (six male, two female), with no history of dysphagia, odynophagia, or reflux symptoms were recruited. One additional subject was excluded upon the discovery of aperistalsis on manometry. The mean age of the volunteers was 27 6 8 yr. The study protocol was
1458
Pouderoux et al.
approved by the Northwestern University Institutional Review Board, and written informed consent was obtained from all subjects. Concurrent Manometry and Videofluoroscopy Esophageal manometry was performed with an 8-lumen, water-perfused assembly connected to a low-compliance pneumohydraulic perfusion pump (Dentsleeve Pty. Ltd., Parkside, South Australia, Australia) and a computerized polygraph (Neomedix Systems Pty. Ltd., Warriewood, New South Wales, Australia). Side-hole recording sites on the manometry catheter were marked with radiodense tantalum to ensure fluoroscopic visibility. Each lumen was perfused with water at 0.6 ml/min and had a maximal rise rate . 200 mm Hg/s. Fluoroscopically imaged swallows were recorded with a videotape recorder (Sony VO 9800; Sony, Tokyo, Japan) at 30 frames/s. Manometric tracings recorded on the polygraph were synchronized with video images using a video timer (model VC 436; Thalner Electronics Laboratories, Ann Arbor, MI), which encoded time in hundredths of a second on each video frame and sent a 1-V, 10-ms pulse to an instrumentation channel of the polygraph at whole-second intervals. Swallowing Studies Subjects were fasted for at least 6 h before the study. The manometric catheter was passed transnasally and positioned so that the most distal side-hole recording site was intragastric. After a 15-min adaptation period, the swallowing study was begun. Three different test boluses were used: 5 ml liquid barium, a 5– 6 mm diameter bread ball, and 4 g chewed bread. Both bread boluses were first rolled into a ball shape, then dipped in a barium suspension to ensure fluoroscopic visibility. The esophagus was imaged fluoroscopically during swallowing to follow bolus transit, identify the sites of residue, and estimate the amount of residue after each swallow. Study subjects were asked if they perceived any residue after each swallow. When the bread ball or chewed bread was tested, as many as four sequential swallows were performed at 1-min intervals, depending upon the rapidity of clearance. Initially, 2 ml liquid barium was swallowed along with the bolus. If the bolus was not cleared after the first swallow, an additional 2 ml liquid barium was given for the second swallow. If the bolus was still not cleared, 5 ml additional liquid barium was used for the third swallow. Finally, 10 ml liquid barium was used for the fourth swallow if the bolus still had not cleared. Test sequences were done in both upright and supine postures, with different test boluses tested in a random order. In all, eight 5-ml liquid barium swallow studies were done in an upright posture and eight in a supine posture. Fifteen sequences of bread ball swallows were done in an upright posture, and 17 in a supine posture. Twenty-four sequences of chewed bread swallows were done in an upright posture, and 22 in a supine posture.
AJG – Vol. 94, No. 6, 1999
Data Analysis Videofluoroscopic recordings were analyzed initially by slow-motion playback of the videotapes without reference to the manometric record to determine the esophageal clearance of the bolus, the distribution of bolus residue, and the clearance time. To define the location of bolus residue, the esophagus was divided into four anatomic regions (distal or ampullary, retrocardiac, aortic arch, and cervical), thus normalizing regional data among individuals of different height. The percentage of swallows during which bolus residue was evident in each esophageal segment was calculated for each bolus type and posture. Fragmentation of the chewed bread bolus was noted in many swallows. The size of each fragment was estimated to the nearest 25%, and these numbers were used in the calculation. The manometric tracings were then analyzed. For each subject, four manometric recording sites were chosen from among the eight sites along the catheter on the basis of their location on fluoroscopy to correspond with the four anatomic regions defined earlier. Thus, the most proximal site chosen (A, or cervical) was at the level of the clavicles, the next (B) was at the carina, hole C (retrocardiac) was midway between B and D, and D (ampullary) was chosen 2–3 cm above the upper limit of the lower esophageal sphincter. Again, this allowed for normalization of data among individuals of different height. Swallow sequences were judged to be peristaltic or nonperistaltic, and peristaltic amplitude was measured at recording sites A, B, C, and D. Thereafter, the occurrence of repetitive contractions and simultaneous pressure events was noted. Videofluorographic events were then correlated with the manometric record. Two-way analysis of variance, Student’s t test, or the x2 test was used for statistical analysis where appropriate. A p value , 0.05 was considered significant.
RESULTS Clearance of Liquid Barium In all swallows with liquid barium, both upright and supine, the barium was cleared during a single swallow by each subject. There was no significant residue remaining in the esophagus after this first and only swallow. Bread Clearance The clearance of unchewed bread balls in the upright and supine postures is shown in Figures 1 and 2, respectively. The first swallow cleared the unchewed bread ball in relatively few instances in both postures (6.7% and 5.9%, respectively). However, posture had a significant effect on clearance during subsequent swallows. In the upright posture, all unchewed bread balls were cleared in the second, third, or fourth swallow. On the other hand, in the supine posture, even after four swallows, the bread balls were cleared in only 52.9% of instances (p , 0.01, upright vs supine). The clearance of chewed bread in the upright and supine
AJG – June, 1999
Esophageal Solid Bolus Transit
1459
Figure 1. Esophageal clearance of an unchewed bread ball in the upright posture. A 5– 6 mm diameter wheat bread ball was tested during each test sequence, with as many as four sequential swallows (using 2 ml, 2 ml, 5 ml, and 10 ml liquid barium, respectively) at 1-min intervals. In all, 15 test sequences were performed. Small circles indicate the locations of bread balls in the esophagus after swallows. The percentage of test sequences in which the bread ball was cleared from the esophagus by each swallow is shown in the bottom of the drawing, corresponding to each swallow.
postures is shown in Figures 3 and 4, respectively. Compared with the unchewed bread balls, chewed bread was cleared much more easily in both the upright and supine postures (see Table 1). However, a significant posture effect
was still demonstrated; chewed bread was cleared in significantly fewer instances in the supine posture than in the upright posture after two swallows (72.8% vs 100%, respectively; p , 0.05). In the supine posture, the last two swal-
Figure 2. Esophageal clearance of an unchewed bread ball in supine posture in an experiment otherwise identical to that in Figure 1. In all, 17 test sequences were performed.
1460
Pouderoux et al.
AJG – Vol. 94, No. 6, 1999
Table 1. Clearance Efficacy of the Chewed Compared With Unchewed Bread: Data From All Eight Subjects
Upright swallow 1 Upright swallow 2 Supine swallow 1 Supine swallow 2 Supine swallow 4
Chewed Bread, Complete Clearance
Unchewed Bread, Complete Clearance
83.3% 100% 59.1% 72.8% 91.0%
6.7%* 53.3%* 5.9%* 17.6%* 52.9%†
* p , 0.01; † p , 0.05.
Figure 3. Esophageal clearance of a chewed bread ball in an upright posture. A bolus of 4 g chewed wheat bread was tested during each test sequence, with as many as four sequential swallows (using 2 ml, 2 ml, 5 ml, and 10 ml barium, respectively) at 1-min intervals. In all, 24 test sequences were performed. The percentage of test sequences during which the bread remained in each of four esophageal anatomic regions or stomach after each swallow is shown to the right of the drawing, corresponding to each swallow.
lows cleared the remaining chewed bread relatively infrequently. After four swallows, 9% of chewed bread still remained in the esophagus. Figure 5 summarizes the efficacy of clearance among bolus types.
Mechanisms of Solid Bolus Clearance Solid bolus clearance was significantly less effective than liquid clearance and had distinct mechanistic characteristics. Fluorographically, unchewed solid boluses were observed to travel segmentally and get caught either at the carina or at the ampulla. For a given swallow, the balls would traverse zero, one, or multiple segments of the esophagus before hanging up at one of these loci. In either posture, any one of three distinct patterns of propulsion was observed during each swallow: 1) the bolus did not move appreciably and the contractions simply rode over the bolus; 2) there was slight movement within an esophageal segment as the bolus moved with the tail of the peristaltic contraction, analogous to the mechanism of liquid clearance with the caveat that the contraction eventually rode over the bolus; or 3) the ball moved across one or more esophageal segments, in this case traveling with the leading edge of the bolus rather than the bolus tail. The third pattern was seen mainly in the upright posture when gravity facilitated movement of the solid bolus within the air-distended esophagus, accounting for
Figure 4. Esophageal clearance of a chewed bread ball in a supine posture in an experiment otherwise identical to that in Figure 3. In all, 22 test sequences were done.
AJG – June, 1999
Esophageal Solid Bolus Transit
1461
Table 2. Effect of Posture and Bolus Consistency on Peristaltic Amplitude in Eight Healthy Subjects (mean 6 SE, mm Hg) Liquid Barium (5 ml) Region Region Region Region
1 2 3 4
Chewed Bread
Upright
Supine
Upright
Supine
37 6 6 70 6 7 82 6 7 97 6 10
55 6 9 93 6 10 89 6 13 88 6 11
46 6 7 75 6 9 74 6 7 80 6 9
98 6 12 111 6 13 140 6 21 117 6 23
p , 0.05 for bolus content, body posture, and posture p content (ANOVA).
bread was swallowed; no difference was found between the upright and supine postures when liquid barium was swallowed. Figure 5. Percentage of test sequences during which tested bolus was cleared from the esophagus after each swallow in each swallow condition; ✣ 5 liquid barium (upright and supine), } 5 unchewed bread ball (upright), { 5 unchewed bread ball (supine), F 5 chewed bread (upright), E 5 chewed bread (supine).
more rapid bolus clearance (particularly with the second swallow of a sequence). Chewed bread was cleared in the same manner as a viscous liquid; however, the apparent consistency of the chewed bolus varied substantially among subjects. Interestingly, regardless of the number of swallows required, subjects reported no perception of incomplete clearance of the chewed solid bolus. With the unchewed bolus, subjects were decreasingly aware of stasis with more distal locations. At the carina, stasis was suspected by subjects in only half of the instances in which it occurred with the first swallow. Combining the manometric records with the videofluorographic images revealed that failed peristalsis, observed in 30% of unchewed bread ball swallows, was frequently associated with incomplete solid bolus clearance. However, peristaltic failure often occurred distal to the stopping point of the solid bolus. With either solid bolus, non-lumen– obliterating contractions at the bolus tail occurred, especially with the chewed bread. Contractile pressures associated with these nonocclusive contractions were as great as 50 mm Hg. Furthermore, the peristaltic contraction propagated faster than the bolus moved, often passing beyond it, especially when the bread balls were in the region of the aortic arch and carina. The peristaltic amplitudes recorded from the four esophageal regions during swallows of liquid barium and chewed bread are shown in Table 2. In the upright posture, the amplitude of esophageal contraction at all recording sites was not significantly different between liquid barium swallows and chewed bread swallows. However, in the supine posture, the amplitude was significantly greater during chewed bread swallows, compared with liquid barium swallows at all recording sites (p , 0.05). In addition, peristaltic amplitudes were significantly greater (p , 0.05) in the supine posture than in the upright posture when chewed
DISCUSSION This study utilized concurrent manometry and videofluoroscopy to ascertain the efficacy of solid bolus clearance in normal individuals. The major finding was that the efficacy of esophageal clearance of solid boluses was significantly less than that of liquid barium; emptying of a solid bolus was frequently incomplete after one, two, and often three swallows. Surprisingly, the individuals studied had poor subjective perception of incomplete emptying. Furthermore, comparison of the manometric recordings to the simultaneous fluoroscopy revealed that the efficacy of bolus clearance and the site of stasis were poorly correlated with the manometry tracing for any given swallow. Although our focus in this investigation was on solid bolus transit through the esophagus, the swallowed bread was necessarily mixed with liquid (barium and saliva) and air. This mixture is initially propelled into the esophagus by the pharyngeal pump (2, 17). Analysis of ultrafast computed tomographic images suggests that 8 –32 ml air is swallowed along with the oral contents and that this mixture disperses along the esophagus to the level of the ampulla (18). The peristaltic clearing wave then acts to strip the tail of the bolus toward the stomach (2– 4, 19, 20). Thus, the rapidity of esophageal clearance of a solid bolus depends upon whether the solid travels with the head or the tail of the bolus composite, and the efficacy of solid transit within each domain. Consider each of these issues in turn. Consistent with earlier observations (20), we found that the pharyngeal pump, and hence the initial swallow, usually propelled a bread bolus only to the proximal esophagus, obliging it to travel with the bolus tail. The likelihood of traveling with the bolus head on subsequent swallows was clearly increased in an upright posture, when gravity worked in conjunction with the distension created by the pharyngeal swallow. When traveling with the head of the composite bolus, a bread bolus moved across one or more esophageal segment(s) ahead of the peristaltic stripping wave. In a supine posture, on the other hand, the bread was much more likely to reside with the tail of the bolus, in which case transit became dependent upon the action of the peristaltic
1462
Pouderoux et al.
stripping wave. Findings from the present study suggest that the peristaltic clearing mechanism is much less effective with a bread bolus than with liquid. With liquid barium, clearance is related mainly to peristaltic amplitude and only severely hypotensive contractions result in esophageal residue (2). In this study, peristaltic contractions were often observed to pass over the bread bolus, especially in the region of the aortic arch and carina. Evidently, these contractions (which at times exceeded 50 mm Hg) were not strong enough to overcome the frictional resistance associated with the viscous bread bolus. However, the apparent viscosity of the chewed bread was highly variable among individuals, to the point that in some individuals it appeared more like a viscous liquid than a solid. This variability is probably related to the degree of mastication and/or salivation. Previous manometric studies using solid boluses (food, gelatin, or marshmallow) have revealed that a solid bolus elicits distinct esophageal contractile responses, compared with liquid. Reported differences include a higher rate of nonconducted and/or nonperistaltic contractions (9, 11–14), lower peristaltic wave velocity (13, 14, 21), higher peristaltic amplitude (13, 14), and longer duration of wave contraction (21). Consistent with these observations, we observed failed peristalsis in 30% of swallows with the unchewed bread. However, failure commonly occurred past the stopping point of the bolus, perhaps due to the absence of distention of the distal esophageal segment. Thus, rather than being the cause of impaired clearance, failed peristalsis appeared to be the result of impaired clearance. Manometry during chewed bread swallows also demonstrated greater contraction amplitude in the supine as opposed to upright posture, again suggesting a feedback mechanism wherein the contractile response is influenced by the difficulty of bolus clearance. An interesting observation made during this study relevant to the topic of pill esophagitis was that subjects had poor perception of whether or not the bread boluses had cleared the esophagus on any given swallow. Similarly, Kikendall et al. noted that despite pills and capsules remaining in the esophagus of normal subjects for as long as 45 min, most individuals had no awareness of retention (22). Furthermore, there was no difference in pill retention between those individuals reporting pill dysphagia and those who did not (22, 23). These observations emphasize the poor relationship between subjective and objective measures of esophageal clearance. Also relevant to the subject of pill esophagitis is the movement pattern of unchewed bread observed in this study. Specifically, unchewed bread required a greater number of swallows for clearance, was facilitated by an upright posture, and tended to become lodged at the aortic arch or carina. Pill esophagitis is frequently associated with swallowing in the supine position, taking medications at bedtime (22, 24), and most commonly occurs in the region of the aortic arch and carina. Finally, it is noteworthy that the vast majority of cases of pill esoph-
AJG – Vol. 94, No. 6, 1999
agitis have been reported in patients who have no detectable abnormality of esophageal motility (22–24). In conclusion, data on the efficacy of liquid propulsion through the esophagus do not apply to solids. Unlike liquid, a bread bolus is rarely cleared from the esophagus with a single swallow despite individuals having normal manometric recordings. Furthermore, there was poor subjective awareness of incomplete bread transit. Mastication (and presumably salivation) facilitated the esophageal transport of bread by partially liquefying it. An upright posture also facilitated bolus transit. Further investigation into the mechanics of esophageal solid clearance will be necessary to determine how these differences between liquid and solid boluses relate to symptomatic dysphagia.
ACKNOWLEDGMENT This study was supported by grant RO1 DC00646 (PJK) from the Public Health Service. Reprint requests and correspondence: Peter J. Kahrilas, M.D., Northwestern University Medical School, Division of Gastroenterology and Hepatology, Department of Medicine, Passavant Pavilion, Suite 746, 303 East Superior Street, Chicago, IL 606113053. Received Nov. 17, 1998; accepted Feb. 24, 1999.
REFERENCES 1. Helm JF, Dodds WJ, Riedel DR, et al. Determinants of esophageal acid clearance in normal subjects. Gastroenterology 1983;85:607–12. 2. Kahrilas PJ, Dodds WJ, Hogan WJ. Effect of peristaltic dysfunction on esophageal volume clearance. Gastroenterology 1988;94:73– 80. 3. Ott DJ, Chen YM, Hewson EG, et al. Esophageal motility: Assessment with synchronous video tape fluoroscopy and manometry. Radiology 1989;173:419 –22. 4. Massey BT, Dodds WJ, Hogan WJ, et al. Abnormal esophageal motility. An analysis of concurrent radiographic and manometric findings. Gastroenterology 1991;101:344 –54. 5. Davies HA, Evans KT, Butler F, et al. Diagnostic value of “bread-barium” swallow in patients with esophageal symptoms. Dig Dis Sci 1983;28:1094 –100. 6. van Westen D, Ekberg O. Solid bolus swallowing in the radiologic evaluation of dysphagia. Acta Radiologica 1993; 34:372–5. 7. Kjelle´n G, Svedberg JB, Tibbling L. Solid bolus transit by esophageal scintigraphy in patients with dysphagia and normal manometry and radiography. Dig Dis Sci 1984;29:1–5. 8. Eriksen CA, Holdsworth RJ, Sutton D, et al. The solid bolus oesophageal egg transit test: Its manometric interpretation and usefulness as a screening test. Br J Surg 1987;74:1130 –3. 9. Mellow MH. Esophageal motility during food ingestion: A physiologic test of esophageal motor function. Gastroenterology 1983;85:570 –7. 10. Allen M, Mellow MH, Robinson M. Manometry during food ingestion aids in the diagnosis of diffuse esophageal spasm. Am J Gastroenterol 1992;87:568 –71. 11. Howard PJ, Maher L, Pryde A, et al. Systematic comparison
AJG – June, 1999
12. 13. 14. 15.
16. 17.
of conventional oesophageal manometry with oesophageal motility while eating bread. Gut 1991;32:1264 –9. Howard PJ, Pryde A, Heading RC. Oesophageal manometry during eating in the investigation of patients with chest pain or dysphagia. Gut 1989;30:1179 – 86. Johnston BT, Collins JSA, McFarland RJ, et al. A comparison of esophageal motility in response to bread swallows and water swallows. Am J Gastroenterol 1993;88:351–5. Keren S, Argaman Golan M. Solid swallowing versus water swallowing: Manometric study of dysphagia. Dig Dis Sci 1992;37:603– 8. Sears VW, Castell JA, Castell DO. Comparison of effects of upright versus supine body position and liquid versus solid bolus on esophageal pressures in normal humans. Dig Dis Sci 1990;35:857– 64. Breumelhof R, Timmer R, van Hees PAM, et al. Low-amplitude distal esophageal spasm as a cause of severe dysphagia for solid food. Am J Gastroenterol 1996;91:143– 6. Inglefinger FJ. Esophageal motility. Physiol Rev 1958;38: 533– 84.
Esophageal Solid Bolus Transit
1463
18. Pouderoux P, Ergun GA, Lin S, et al. Esophageal bolus transit imaged by ultrafast computerized tomography. Gastroenterology 1996;110:1422– 8. 19. Li M, Brasseur JG, Dodds WJ. Analyses of normal and abnormal esophageal transport using computer simulations. Am J Physiol 1994;266:G525– 43. 20. Buthpitiya AG, Stroud D, Russell COH. Pharyngeal pump and esophageal transit. Dig Dis Sci 1987;32:1244 – 8. 21. Dooley CP, Di Lorenzo C, Valenzuela JE. Esophageal function in humans. Effect of bolus consistency and temperature. Dig Dis Sci 1990;35:167–22. 22. Kikendall JW, Friedman AC, Oyewole MA, et al. Pill-induced esophageal injury. Dig Dis Sci 1983;28:174 – 81. 23. Minocha A, Greenbaum DS. Pill-esophagitis caused by nonsteroidal antiinflammatory drugs. Am J Gastroenterol 1991; 86:1086 –9. 24. Coates AG, Nostrant TT, Wilson JA, et al. Esophagitis caused by nonsteroidal anti-inflammatory medication: Case reports and review of the literature on pill-induced esophageal injury. South Med J 1986;79:1094 –7.
Vol. 94, No. 6, 1999 ISSN 0002-9270/99/$20.00 PII S0002-9270(99)00182-3
ORIGINAL CONTRIBUTIONS
Esophageal Solid Bolus Transit: Studies Using Concurrent Videofluoroscopy and Manometry Philippe Pouderoux, M.D., Guoxiang Shi, M.D., Ph.D., Roger P. Tatum, M.D., and Peter J. Kahrilas, M.D. Department of Medicine, Division of Gastroenterology and Hepatology, Northwestern University Medical School, Chicago, Illinois
OBJECTIVE: Our aim was to assess the efficacy and mechanism of solid bolus transit through the esophagus. METHODS: Eight healthy volunteers were studied with concurrent manometry and videofluoroscopy while swallowing 5 ml liquid barium, a 5– 6 mm diameter bread ball, and 4 g chewed bread in both a supine and upright posture. As many as four successive swallows were performed until clearance was achieved. RESULTS: The esophageal clearance of liquid barium was 100% with the first swallow. Clearance of the unchewed bread ball occurred with the first swallow in only 6.7% of trials in the upright posture and 5.9% in the supine posture. After four swallows, clearance was 100% and 52.9% in the upright and supine postures, respectively. Chewed bread was more readily cleared than unchewed bread, with 100% clearance after two swallows in the upright posture and 91% clearance after four swallows in the supine posture. The most common locus of bread stasis was at the aortic arch and carina. The bread boluses were noted to move more effectively when localized in the head as opposed to the tail of the bolus composite. Nonocclusive contractions often occurred at the bolus tail despite the increased peristaltic amplitude seen with the chewed bread. Failed peristalsis, a frequent cause for solid clearance failure, was observed during 30% of all bread swallows. This usually occurred distal to the stopping point of the bolus, suggesting it to be the result rather than the cause of impaired transit. CONCLUSIONS: Although infrequently perceived by these normal subjects and in contradistinction to liquid clearance, bread is rarely cleared from the esophagus with a single swallow. Mastication and an upright posture facilitate the esophageal transport of solids. Bolus composition and impaired bolus transit alter the amplitude and conductance of peristalsis. Manometric data pertaining to liquid clearance through the esophagus do not readily apply to bread. (Am J Gastroenterol 1999;94:1457–1463. © 1999 by Am. Coll. of Gastroenterology)
INTRODUCTION The determinants of esophageal liquid clearance (which relate to acid clearance) are well described (1– 4), whereas
those of esophageal solid food transit (which relate more closely to dysphagia) are less well defined. Fluoroscopy and manometry are the two techniques most commonly used for evaluating esophageal motor function. Fluoroscopy images the transit of a swallowed bolus to define the effectiveness of esophageal clearance, and manometry quantifies esophageal contractility. Fluoroscopic studies have revealed that a solid bolus traverses the esophagus more slowly than liquid but the method does not allow determination of the cause of this delay (5, 6). Slowed solid bolus transit has been corroborated with scintigraphic studies but localization and characterization of transit abnormalities are less specific than with fluoroscopic techniques (7, 8). Food (9 –13) and special solid bolus ingestion (14, 15) have been evaluated using esophageal manometry as well. These studies have shown that manometric recordings during solid bolus swallows differ from those obtained during water swallows. Although these different responses may be relevant to the prolonged transit of solids seen during fluoroscopy, the mechanism of delayed transit of solids through the esophagus cannot be defined without correlating imaged bolus transit with the manometric events. Such correlation was attempted in a case report by Breumelhof et al. in which fluoroscopy combined with manometry during bread swallows not only confirmed the existence of significant transit impairment, but also revealed esophageal spasm that had been missed during conventional manometry, ambulatory manometry, and radiographic studies using liquid barium (16). However, no study to date has used this technique to systematically evaluate the normal mechanism of esophageal solid clearance. Thus, the aim of this study was to use concurrent manometry and videofluoroscopy to compare the esophageal clearance of chewed and unchewed bread with that of liquid, and to correlate bolus clearance with manometric recordings in normal individuals.
MATERIALS AND METHODS Eight healthy volunteers (six male, two female), with no history of dysphagia, odynophagia, or reflux symptoms were recruited. One additional subject was excluded upon the discovery of aperistalsis on manometry. The mean age of the volunteers was 27 6 8 yr. The study protocol was
1458
Pouderoux et al.
approved by the Northwestern University Institutional Review Board, and written informed consent was obtained from all subjects. Concurrent Manometry and Videofluoroscopy Esophageal manometry was performed with an 8-lumen, water-perfused assembly connected to a low-compliance pneumohydraulic perfusion pump (Dentsleeve Pty. Ltd., Parkside, South Australia, Australia) and a computerized polygraph (Neomedix Systems Pty. Ltd., Warriewood, New South Wales, Australia). Side-hole recording sites on the manometry catheter were marked with radiodense tantalum to ensure fluoroscopic visibility. Each lumen was perfused with water at 0.6 ml/min and had a maximal rise rate . 200 mm Hg/s. Fluoroscopically imaged swallows were recorded with a videotape recorder (Sony VO 9800; Sony, Tokyo, Japan) at 30 frames/s. Manometric tracings recorded on the polygraph were synchronized with video images using a video timer (model VC 436; Thalner Electronics Laboratories, Ann Arbor, MI), which encoded time in hundredths of a second on each video frame and sent a 1-V, 10-ms pulse to an instrumentation channel of the polygraph at whole-second intervals. Swallowing Studies Subjects were fasted for at least 6 h before the study. The manometric catheter was passed transnasally and positioned so that the most distal side-hole recording site was intragastric. After a 15-min adaptation period, the swallowing study was begun. Three different test boluses were used: 5 ml liquid barium, a 5– 6 mm diameter bread ball, and 4 g chewed bread. Both bread boluses were first rolled into a ball shape, then dipped in a barium suspension to ensure fluoroscopic visibility. The esophagus was imaged fluoroscopically during swallowing to follow bolus transit, identify the sites of residue, and estimate the amount of residue after each swallow. Study subjects were asked if they perceived any residue after each swallow. When the bread ball or chewed bread was tested, as many as four sequential swallows were performed at 1-min intervals, depending upon the rapidity of clearance. Initially, 2 ml liquid barium was swallowed along with the bolus. If the bolus was not cleared after the first swallow, an additional 2 ml liquid barium was given for the second swallow. If the bolus was still not cleared, 5 ml additional liquid barium was used for the third swallow. Finally, 10 ml liquid barium was used for the fourth swallow if the bolus still had not cleared. Test sequences were done in both upright and supine postures, with different test boluses tested in a random order. In all, eight 5-ml liquid barium swallow studies were done in an upright posture and eight in a supine posture. Fifteen sequences of bread ball swallows were done in an upright posture, and 17 in a supine posture. Twenty-four sequences of chewed bread swallows were done in an upright posture, and 22 in a supine posture.
AJG – Vol. 94, No. 6, 1999
Data Analysis Videofluoroscopic recordings were analyzed initially by slow-motion playback of the videotapes without reference to the manometric record to determine the esophageal clearance of the bolus, the distribution of bolus residue, and the clearance time. To define the location of bolus residue, the esophagus was divided into four anatomic regions (distal or ampullary, retrocardiac, aortic arch, and cervical), thus normalizing regional data among individuals of different height. The percentage of swallows during which bolus residue was evident in each esophageal segment was calculated for each bolus type and posture. Fragmentation of the chewed bread bolus was noted in many swallows. The size of each fragment was estimated to the nearest 25%, and these numbers were used in the calculation. The manometric tracings were then analyzed. For each subject, four manometric recording sites were chosen from among the eight sites along the catheter on the basis of their location on fluoroscopy to correspond with the four anatomic regions defined earlier. Thus, the most proximal site chosen (A, or cervical) was at the level of the clavicles, the next (B) was at the carina, hole C (retrocardiac) was midway between B and D, and D (ampullary) was chosen 2–3 cm above the upper limit of the lower esophageal sphincter. Again, this allowed for normalization of data among individuals of different height. Swallow sequences were judged to be peristaltic or nonperistaltic, and peristaltic amplitude was measured at recording sites A, B, C, and D. Thereafter, the occurrence of repetitive contractions and simultaneous pressure events was noted. Videofluorographic events were then correlated with the manometric record. Two-way analysis of variance, Student’s t test, or the x2 test was used for statistical analysis where appropriate. A p value , 0.05 was considered significant.
RESULTS Clearance of Liquid Barium In all swallows with liquid barium, both upright and supine, the barium was cleared during a single swallow by each subject. There was no significant residue remaining in the esophagus after this first and only swallow. Bread Clearance The clearance of unchewed bread balls in the upright and supine postures is shown in Figures 1 and 2, respectively. The first swallow cleared the unchewed bread ball in relatively few instances in both postures (6.7% and 5.9%, respectively). However, posture had a significant effect on clearance during subsequent swallows. In the upright posture, all unchewed bread balls were cleared in the second, third, or fourth swallow. On the other hand, in the supine posture, even after four swallows, the bread balls were cleared in only 52.9% of instances (p , 0.01, upright vs supine). The clearance of chewed bread in the upright and supine
AJG – June, 1999
Esophageal Solid Bolus Transit
1459
Figure 1. Esophageal clearance of an unchewed bread ball in the upright posture. A 5– 6 mm diameter wheat bread ball was tested during each test sequence, with as many as four sequential swallows (using 2 ml, 2 ml, 5 ml, and 10 ml liquid barium, respectively) at 1-min intervals. In all, 15 test sequences were performed. Small circles indicate the locations of bread balls in the esophagus after swallows. The percentage of test sequences in which the bread ball was cleared from the esophagus by each swallow is shown in the bottom of the drawing, corresponding to each swallow.
postures is shown in Figures 3 and 4, respectively. Compared with the unchewed bread balls, chewed bread was cleared much more easily in both the upright and supine postures (see Table 1). However, a significant posture effect
was still demonstrated; chewed bread was cleared in significantly fewer instances in the supine posture than in the upright posture after two swallows (72.8% vs 100%, respectively; p , 0.05). In the supine posture, the last two swal-
Figure 2. Esophageal clearance of an unchewed bread ball in supine posture in an experiment otherwise identical to that in Figure 1. In all, 17 test sequences were performed.
1460
Pouderoux et al.
AJG – Vol. 94, No. 6, 1999
Table 1. Clearance Efficacy of the Chewed Compared With Unchewed Bread: Data From All Eight Subjects
Upright swallow 1 Upright swallow 2 Supine swallow 1 Supine swallow 2 Supine swallow 4
Chewed Bread, Complete Clearance
Unchewed Bread, Complete Clearance
83.3% 100% 59.1% 72.8% 91.0%
6.7%* 53.3%* 5.9%* 17.6%* 52.9%†
* p , 0.01; † p , 0.05.
Figure 3. Esophageal clearance of a chewed bread ball in an upright posture. A bolus of 4 g chewed wheat bread was tested during each test sequence, with as many as four sequential swallows (using 2 ml, 2 ml, 5 ml, and 10 ml barium, respectively) at 1-min intervals. In all, 24 test sequences were performed. The percentage of test sequences during which the bread remained in each of four esophageal anatomic regions or stomach after each swallow is shown to the right of the drawing, corresponding to each swallow.
lows cleared the remaining chewed bread relatively infrequently. After four swallows, 9% of chewed bread still remained in the esophagus. Figure 5 summarizes the efficacy of clearance among bolus types.
Mechanisms of Solid Bolus Clearance Solid bolus clearance was significantly less effective than liquid clearance and had distinct mechanistic characteristics. Fluorographically, unchewed solid boluses were observed to travel segmentally and get caught either at the carina or at the ampulla. For a given swallow, the balls would traverse zero, one, or multiple segments of the esophagus before hanging up at one of these loci. In either posture, any one of three distinct patterns of propulsion was observed during each swallow: 1) the bolus did not move appreciably and the contractions simply rode over the bolus; 2) there was slight movement within an esophageal segment as the bolus moved with the tail of the peristaltic contraction, analogous to the mechanism of liquid clearance with the caveat that the contraction eventually rode over the bolus; or 3) the ball moved across one or more esophageal segments, in this case traveling with the leading edge of the bolus rather than the bolus tail. The third pattern was seen mainly in the upright posture when gravity facilitated movement of the solid bolus within the air-distended esophagus, accounting for
Figure 4. Esophageal clearance of a chewed bread ball in a supine posture in an experiment otherwise identical to that in Figure 3. In all, 22 test sequences were done.
AJG – June, 1999
Esophageal Solid Bolus Transit
1461
Table 2. Effect of Posture and Bolus Consistency on Peristaltic Amplitude in Eight Healthy Subjects (mean 6 SE, mm Hg) Liquid Barium (5 ml) Region Region Region Region
1 2 3 4
Chewed Bread
Upright
Supine
Upright
Supine
37 6 6 70 6 7 82 6 7 97 6 10
55 6 9 93 6 10 89 6 13 88 6 11
46 6 7 75 6 9 74 6 7 80 6 9
98 6 12 111 6 13 140 6 21 117 6 23
p , 0.05 for bolus content, body posture, and posture p content (ANOVA).
bread was swallowed; no difference was found between the upright and supine postures when liquid barium was swallowed. Figure 5. Percentage of test sequences during which tested bolus was cleared from the esophagus after each swallow in each swallow condition; ✣ 5 liquid barium (upright and supine), } 5 unchewed bread ball (upright), { 5 unchewed bread ball (supine), F 5 chewed bread (upright), E 5 chewed bread (supine).
more rapid bolus clearance (particularly with the second swallow of a sequence). Chewed bread was cleared in the same manner as a viscous liquid; however, the apparent consistency of the chewed bolus varied substantially among subjects. Interestingly, regardless of the number of swallows required, subjects reported no perception of incomplete clearance of the chewed solid bolus. With the unchewed bolus, subjects were decreasingly aware of stasis with more distal locations. At the carina, stasis was suspected by subjects in only half of the instances in which it occurred with the first swallow. Combining the manometric records with the videofluorographic images revealed that failed peristalsis, observed in 30% of unchewed bread ball swallows, was frequently associated with incomplete solid bolus clearance. However, peristaltic failure often occurred distal to the stopping point of the solid bolus. With either solid bolus, non-lumen– obliterating contractions at the bolus tail occurred, especially with the chewed bread. Contractile pressures associated with these nonocclusive contractions were as great as 50 mm Hg. Furthermore, the peristaltic contraction propagated faster than the bolus moved, often passing beyond it, especially when the bread balls were in the region of the aortic arch and carina. The peristaltic amplitudes recorded from the four esophageal regions during swallows of liquid barium and chewed bread are shown in Table 2. In the upright posture, the amplitude of esophageal contraction at all recording sites was not significantly different between liquid barium swallows and chewed bread swallows. However, in the supine posture, the amplitude was significantly greater during chewed bread swallows, compared with liquid barium swallows at all recording sites (p , 0.05). In addition, peristaltic amplitudes were significantly greater (p , 0.05) in the supine posture than in the upright posture when chewed
DISCUSSION This study utilized concurrent manometry and videofluoroscopy to ascertain the efficacy of solid bolus clearance in normal individuals. The major finding was that the efficacy of esophageal clearance of solid boluses was significantly less than that of liquid barium; emptying of a solid bolus was frequently incomplete after one, two, and often three swallows. Surprisingly, the individuals studied had poor subjective perception of incomplete emptying. Furthermore, comparison of the manometric recordings to the simultaneous fluoroscopy revealed that the efficacy of bolus clearance and the site of stasis were poorly correlated with the manometry tracing for any given swallow. Although our focus in this investigation was on solid bolus transit through the esophagus, the swallowed bread was necessarily mixed with liquid (barium and saliva) and air. This mixture is initially propelled into the esophagus by the pharyngeal pump (2, 17). Analysis of ultrafast computed tomographic images suggests that 8 –32 ml air is swallowed along with the oral contents and that this mixture disperses along the esophagus to the level of the ampulla (18). The peristaltic clearing wave then acts to strip the tail of the bolus toward the stomach (2– 4, 19, 20). Thus, the rapidity of esophageal clearance of a solid bolus depends upon whether the solid travels with the head or the tail of the bolus composite, and the efficacy of solid transit within each domain. Consider each of these issues in turn. Consistent with earlier observations (20), we found that the pharyngeal pump, and hence the initial swallow, usually propelled a bread bolus only to the proximal esophagus, obliging it to travel with the bolus tail. The likelihood of traveling with the bolus head on subsequent swallows was clearly increased in an upright posture, when gravity worked in conjunction with the distension created by the pharyngeal swallow. When traveling with the head of the composite bolus, a bread bolus moved across one or more esophageal segment(s) ahead of the peristaltic stripping wave. In a supine posture, on the other hand, the bread was much more likely to reside with the tail of the bolus, in which case transit became dependent upon the action of the peristaltic
1462
Pouderoux et al.
stripping wave. Findings from the present study suggest that the peristaltic clearing mechanism is much less effective with a bread bolus than with liquid. With liquid barium, clearance is related mainly to peristaltic amplitude and only severely hypotensive contractions result in esophageal residue (2). In this study, peristaltic contractions were often observed to pass over the bread bolus, especially in the region of the aortic arch and carina. Evidently, these contractions (which at times exceeded 50 mm Hg) were not strong enough to overcome the frictional resistance associated with the viscous bread bolus. However, the apparent viscosity of the chewed bread was highly variable among individuals, to the point that in some individuals it appeared more like a viscous liquid than a solid. This variability is probably related to the degree of mastication and/or salivation. Previous manometric studies using solid boluses (food, gelatin, or marshmallow) have revealed that a solid bolus elicits distinct esophageal contractile responses, compared with liquid. Reported differences include a higher rate of nonconducted and/or nonperistaltic contractions (9, 11–14), lower peristaltic wave velocity (13, 14, 21), higher peristaltic amplitude (13, 14), and longer duration of wave contraction (21). Consistent with these observations, we observed failed peristalsis in 30% of swallows with the unchewed bread. However, failure commonly occurred past the stopping point of the bolus, perhaps due to the absence of distention of the distal esophageal segment. Thus, rather than being the cause of impaired clearance, failed peristalsis appeared to be the result of impaired clearance. Manometry during chewed bread swallows also demonstrated greater contraction amplitude in the supine as opposed to upright posture, again suggesting a feedback mechanism wherein the contractile response is influenced by the difficulty of bolus clearance. An interesting observation made during this study relevant to the topic of pill esophagitis was that subjects had poor perception of whether or not the bread boluses had cleared the esophagus on any given swallow. Similarly, Kikendall et al. noted that despite pills and capsules remaining in the esophagus of normal subjects for as long as 45 min, most individuals had no awareness of retention (22). Furthermore, there was no difference in pill retention between those individuals reporting pill dysphagia and those who did not (22, 23). These observations emphasize the poor relationship between subjective and objective measures of esophageal clearance. Also relevant to the subject of pill esophagitis is the movement pattern of unchewed bread observed in this study. Specifically, unchewed bread required a greater number of swallows for clearance, was facilitated by an upright posture, and tended to become lodged at the aortic arch or carina. Pill esophagitis is frequently associated with swallowing in the supine position, taking medications at bedtime (22, 24), and most commonly occurs in the region of the aortic arch and carina. Finally, it is noteworthy that the vast majority of cases of pill esoph-
AJG – Vol. 94, No. 6, 1999
agitis have been reported in patients who have no detectable abnormality of esophageal motility (22–24). In conclusion, data on the efficacy of liquid propulsion through the esophagus do not apply to solids. Unlike liquid, a bread bolus is rarely cleared from the esophagus with a single swallow despite individuals having normal manometric recordings. Furthermore, there was poor subjective awareness of incomplete bread transit. Mastication (and presumably salivation) facilitated the esophageal transport of bread by partially liquefying it. An upright posture also facilitated bolus transit. Further investigation into the mechanics of esophageal solid clearance will be necessary to determine how these differences between liquid and solid boluses relate to symptomatic dysphagia.
ACKNOWLEDGMENT This study was supported by grant RO1 DC00646 (PJK) from the Public Health Service. Reprint requests and correspondence: Peter J. Kahrilas, M.D., Northwestern University Medical School, Division of Gastroenterology and Hepatology, Department of Medicine, Passavant Pavilion, Suite 746, 303 East Superior Street, Chicago, IL 606113053. Received Nov. 17, 1998; accepted Feb. 24, 1999.
REFERENCES 1. Helm JF, Dodds WJ, Riedel DR, et al. Determinants of esophageal acid clearance in normal subjects. Gastroenterology 1983;85:607–12. 2. Kahrilas PJ, Dodds WJ, Hogan WJ. Effect of peristaltic dysfunction on esophageal volume clearance. Gastroenterology 1988;94:73– 80. 3. Ott DJ, Chen YM, Hewson EG, et al. Esophageal motility: Assessment with synchronous video tape fluoroscopy and manometry. Radiology 1989;173:419 –22. 4. Massey BT, Dodds WJ, Hogan WJ, et al. Abnormal esophageal motility. An analysis of concurrent radiographic and manometric findings. Gastroenterology 1991;101:344 –54. 5. Davies HA, Evans KT, Butler F, et al. Diagnostic value of “bread-barium” swallow in patients with esophageal symptoms. Dig Dis Sci 1983;28:1094 –100. 6. van Westen D, Ekberg O. Solid bolus swallowing in the radiologic evaluation of dysphagia. Acta Radiologica 1993; 34:372–5. 7. Kjelle´n G, Svedberg JB, Tibbling L. Solid bolus transit by esophageal scintigraphy in patients with dysphagia and normal manometry and radiography. Dig Dis Sci 1984;29:1–5. 8. Eriksen CA, Holdsworth RJ, Sutton D, et al. The solid bolus oesophageal egg transit test: Its manometric interpretation and usefulness as a screening test. Br J Surg 1987;74:1130 –3. 9. Mellow MH. Esophageal motility during food ingestion: A physiologic test of esophageal motor function. Gastroenterology 1983;85:570 –7. 10. Allen M, Mellow MH, Robinson M. Manometry during food ingestion aids in the diagnosis of diffuse esophageal spasm. Am J Gastroenterol 1992;87:568 –71. 11. Howard PJ, Maher L, Pryde A, et al. Systematic comparison
AJG – June, 1999
12. 13. 14. 15.
16. 17.
of conventional oesophageal manometry with oesophageal motility while eating bread. Gut 1991;32:1264 –9. Howard PJ, Pryde A, Heading RC. Oesophageal manometry during eating in the investigation of patients with chest pain or dysphagia. Gut 1989;30:1179 – 86. Johnston BT, Collins JSA, McFarland RJ, et al. A comparison of esophageal motility in response to bread swallows and water swallows. Am J Gastroenterol 1993;88:351–5. Keren S, Argaman Golan M. Solid swallowing versus water swallowing: Manometric study of dysphagia. Dig Dis Sci 1992;37:603– 8. Sears VW, Castell JA, Castell DO. Comparison of effects of upright versus supine body position and liquid versus solid bolus on esophageal pressures in normal humans. Dig Dis Sci 1990;35:857– 64. Breumelhof R, Timmer R, van Hees PAM, et al. Low-amplitude distal esophageal spasm as a cause of severe dysphagia for solid food. Am J Gastroenterol 1996;91:143– 6. Inglefinger FJ. Esophageal motility. Physiol Rev 1958;38: 533– 84.
Esophageal Solid Bolus Transit
1463
18. Pouderoux P, Ergun GA, Lin S, et al. Esophageal bolus transit imaged by ultrafast computerized tomography. Gastroenterology 1996;110:1422– 8. 19. Li M, Brasseur JG, Dodds WJ. Analyses of normal and abnormal esophageal transport using computer simulations. Am J Physiol 1994;266:G525– 43. 20. Buthpitiya AG, Stroud D, Russell COH. Pharyngeal pump and esophageal transit. Dig Dis Sci 1987;32:1244 – 8. 21. Dooley CP, Di Lorenzo C, Valenzuela JE. Esophageal function in humans. Effect of bolus consistency and temperature. Dig Dis Sci 1990;35:167–22. 22. Kikendall JW, Friedman AC, Oyewole MA, et al. Pill-induced esophageal injury. Dig Dis Sci 1983;28:174 – 81. 23. Minocha A, Greenbaum DS. Pill-esophagitis caused by nonsteroidal antiinflammatory drugs. Am J Gastroenterol 1991; 86:1086 –9. 24. Coates AG, Nostrant TT, Wilson JA, et al. Esophagitis caused by nonsteroidal anti-inflammatory medication: Case reports and review of the literature on pill-induced esophageal injury. South Med J 1986;79:1094 –7.