- Email: [email protected]
Prevalence of Esophageal Disorders in Patients With Chest Pain Newly Referred to the Cardiologist
Prevalence of Esophageal Disorders in Patients With Chest Pain Newly Referred to the Cardiologist* Voskuil, MD; Maarten fan Cramer, MD; fan-Henk Ronald Breumelhof MD; Robin Timmer, MD; and MD Andre f.P.M. Smout,
Study objective: The prevalence of esophageal disorders (dysmotility and/or gastroesophageal re¬ study were to investigate the prevalence of esophageal abnormalities in these patients and to assess the value of medical history in predicting the origin of the patient's chest pain. Design: We evaluated 28 consecutive patients who were newly referred to the cardiologist because of angina-like chest pain. Patients with evidence of severe myocardial ischemia were excluded. Cardiologic evaluation included medical history, physical examination, ECG, and exercise testing; further cardiologic workup was carried out only when considered necessary. Gastroenterologic evaluation consisted of medical history, esophageal manometry, endoscopy, and 24-h ambulatory monitoring of esophageal pH and pressure. Measurements and results: In five patients a diagnosis of ischemic coronary artery disease was made. In only two of these five patients, the cardiologic history strongly suggested a cardiac origin of the pain. Twelve patients had a pathologic 24-h pH profile, four of whom also had reflux esophagitis. Ten patients had symptomatic reflux. In only three of these ten patients, the history was judged to be indicative of an esophageal origin of the chest pain. No motility disorders were found. Conclusions: Thirty-six percent of the patients with chest pain newly referred to a cardiologic out¬ patient clinic have symptomatic gastroesophageal reflux. Neither cardiologic nor gastroenterologic history data have a high predictive value with respect to the origin of the chest pain.
flux) in patients with chest pain newly referred to a cardiologic clinic is unknown. The aims of our
(CHEST 1996; 109:1210-14)
Key words: esophageal monitoring; gastroesophageal reflux disease; noncardiac chest pain Abbreviations: GER=gastroesophageal reflux; GERD=gastroesophageal reflux disease; LES=lower esophageal sphincter; GERD=gas ability; fSI=symptom index SAP=symptom association probability;
patients with angina-like chest pain consult a ]V/|*ost -*-*-¦to have a cardiac of the
cardiologist
origin
pain
identified and treated. However, angina-like retrosternal chest pain is not always caused by cardiac disease. For editorial comment
page 1139 abnormalities and gastroesoph¬ see
Esophageal motilityare reflux ageal
in those
(GER) likely to play an etiologic role patients in whom cardiac disease has been
excluded. These esophageal abnormalities can be detected by
24-h ambulatory esophageal endoscopy, manometry, monitoring, and provocation tests. Since 1986, many studies on esophageal monitoring in patients with (Drs. Voskuil, Departments of Breumelhof, and Timmer) and Gastroenterology Cardiologv (Dr.'Cramer), St. Antonius Hospital, Nieuwegein, the Netherlands, and the Depart¬ ment of Gastroenterology7 (Dr. Smout), University Hospital, Utrecht, the Netherlands.
*From the
Manuscript received April 26, 1995; revision accepted October 13. 1210
noncardiac chest pain were reported. The proportion of patients in whom a correlation was found between esophageal abnormalities and pain varied from 10 to 90%,1_6 differences in yield depending on differences in patient selection and diagnostic criteria. Until now, most studies have been performed in patients referred to the gastroenterologist in a late of the diagnostic pathway, after a cardiac origin phase of the symptoms had been excluded. In
a
recent
study by
Lam et
al,6 patients
were
studied in an early phase of the diagnostic pathway, ie, within 1 h after admission to a coronary care unit with acute chest pain. In 27 of 30 patients (90%) in whom cardiac ischemia was excluded, one or more chest pain events without ECG changes were found to be corre¬ lated with an esophageal abnormality. We hypothesized that the prevalence of an esopha¬ geal abnormality in unselected patients who were re¬ ferred for the first time to the cardiologic outpatient clinic of a large general hospital because of angina-like Clinical
Investigations
chest pain could also be substantial. If this hypothesis would be correct, this could have implications for the "standard" workup of these patients. Therefore, the first aim of this study was to investi¬ gate the prevalence of esophageal abnormalities in a very early phase in the diagnostic workup of patients referred for the first time to the cardiologic department of our hospital because of angina-like chest pain, im¬ mediately after life-threatening cardiac ischemia was excluded. The secondary aim of our study was to assess the value of medical history in predicting the origin of the patient's chest pain. Materials
and
Methods
patients with angina-like chest pain who had been newly referred to the outpatient cardiologic depart¬ ment of our hospital. The patients were 16 men and 14 women, with a median age of 53 years (range, 33 to 69 years). Patients with severe cardiac ischemia, congestive heart failure, previous cardiologic his¬ tory, thoracic trauma, or thoracic surgery were excluded. Patients with known esophageal or gastric disorders or surgery were also ex¬ cluded. The study protocol was approved by the medical ethics committee of our hospital. All patients gave written informed consent. In all patients, the cardiologic history was taken by one and the same investigator (M.J.C.). The incidence, duration, severity and character of the chest pain, and its relation to exercise, emotion, and temperature changes were assessed. In addition, the frequency of occurrence and severity of the symptoms orthopnoea, dyspnea, and palpitations were assessed. Furthermore, use of medication and the existence of cardiovascular risk factors were evaluated. Based on these history data, the probability of a cardiac origin of the patient's chest pain was scored on a scale ranging from 1 to 5 (l=very low, 2=low, 3=indifferent, 4=probable, 5=highly probable). Further cardiologic workup consisted of bicycle exercise testing. Additional evaluation such as cardiac ultrasound, myocardial perfusion scin¬ tigraphy, or coronary angiography was carried out only when con¬ sidered necessary by the patient's cardiologist. Within 3 days after the first visit to the cardiologist, the gastro¬ enterologic history was taken in all patients. All patients were interviewed by one and the same investigator (J.H.V.). The incidence, duration, severity and character of the chest pain, and its relation to exercise, emotion, and consumption of meals and/or were assessed. In addition, the frequency of occurrence beverages and the severity of the symptoms heartburn, regurgitation, dysphagia for solid and/or liquid food, and odynophagia were assessed. Furthermore, the patient's actions to relieve the symptoms (ie, rest, change of position, eating or drinking, and the use of antacids) were assessed. The interviewer then scored the probability of an esoph¬ ageal origin of the patient's symptoms on a scale ranging from 1 to 5 (low to high). Immediately after the interview, esophageal manometry, upper GI endoscopy and 24-h ambulatory esophageal pH and pressure monitoring were performed. At least 72 h before esophageal ma¬ nometry and 24-h ambulatory monitoring, treatment with any medication that could interfere with acid secretion or esophageal motility was stopped. A low-compliance perfusion pump with a three-lumen catheter with openings 5 cm apart and oriented in three different directions was used for conventional manometry. The lower esophageal sphincter (LES) was localized by the stationary pull-through method. Thereafter, the catheter was placed with the distal opening 5 cm above the upper border of the LES and ten wet swallows were given, 20 to 30 s apart. Amplitude, du¬ ration, and propagation of contractions at 5, 10, and 15 cm above the LES were assessed. We studied 30 consecutive
A video endoscopic system (Pentax) was used for esophageal endoscopy. Reflux esophagitis was described according to the Savary-Miller grading system (grades I through IV). Directly after¬ wards, a 24-h ambulatory pressure and pH recording system was installed. Intraesophageal pH was recorded with a glass electrode with incorporated reference (Ingold AG; Urdorf; Switzerland). In¬ traesophageal pressure was recorded using a 5F polyurethane catheter with three sensors 5 cm
pressure apart (Sentron 205; Sentron; Roden, the Netherlands). The pH electrode and the
pressure
distal pressure transducer were positioned 5 cm above the upper border of the LES. The signals were recorded with a digital porta¬ ble data logger (UPS 2020 GS; MMS; Enschede, the Netherlands). The three pressure signals were sampled at a rate of 4 Hz and the pH signal was sampled at a rate of 0.2 Hz. The data logger used al¬ lows the recording of five marker signals, of which only one was used, to indicate the episodes of chest pain. During the 24-h monitoring, periods of eating and drinking and periods spent in the supine position were registered by the patient in a diary. Episodes of chest pain were registered by pushing the event marker and writing down the time, nature, and characteris¬ tics of the symptoms. When the 24-h study period was ended, the recorded data were transferred to the computer, stored on a floppy disk, and analyzed in a fully automatic way using previously validated algorithms.' Esophageal acid exposure was calculated for the upright and su¬ pine positions separately and for the total time of registration. GER disease (GERD) was diagnosed when the percentage of time of pH below 4 exceeded the upper limit (95th percentile) of normal, ac¬ cording to the criteria proposed by Richter et al:8 pH less than 4, percent upright.8.15; pH less than 4, percent supine.3.45; and pH less than 4, percent total.5.78. In patients experiencing pain during the recording, an additional computerized analysis of the symptom episode was done using a 2-min window starting 2 min before pain onset and ending at the onset of pain.9 In the analysis of motility disorders, the patient's symptom-free episodes were used as his or her own control. In the analysis of reflux, a symptom episode was considered to be associated with reflux when a pH drop of more than 1 U occurred within 5 s, or when the pH fell below 4, or when both criteria were met during that episode. This computerized analysis of the symptom episode was published in detail elsewhere.6 The last step in the analysis consisted of the calculation of the symptom index (SI) for abnormal motility, and the symptom association probability (SAP) for reflux. The SI was calculated according to Wiener et al.10 A patient was considered to have a positive correlation between his or her pain and motor abnormality when the SI was more than 50%. The SAP was calcu¬ lated as described by Weusten et al.11 The SAP expresses the strength of the association between symptoms and reflux episodes taking all relevant variables (number of reflux episodes, number of symptom episodes, number of associated symptom episodes) into account. A patient was considered to have a positive correlation between his or her pain and reflux when the SAP was 95% or greater. Therapy with acid-secretion inhibitors (omeprazole, 40 mg daily) was started when the diagnosis of GERD, with or without reflux esophagitis, was made. A patient was considered to suffer from symptomatic GERD when there was a relief in chest pain after ini¬ tiating therapy. Finally, the chest pain of all patients was classified as cardiac (as proved by exercise test, scintigraphy, coronary angiog¬
or myocardial event), esophageal (as proved by a positive SI raphy, for dysmotility, a SAP of 95% or greater, and/or relief in chest pain after initiating therapy in patients with GERD), cardiac and esophageal, or unidentified (noncardiac, nonesophageal) in origin. Obviously, we performed a "routine" cardiologic workup com¬ pared with a more extensive gastroenterologic investigation. This was done to prove our hypothesis that the prevalence of esophageal disorders in angina-like chest pain is substantial. The statistical significance of differences in history scores be-
CHEST/109/5/MAY,
1996
1211
Table 1.Patient Data* Evaluation by Cardiologist
No.
Cardiac History Exercise
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19 20 21 22 23 24 25 26 28 29 30
Score
(1-5)
Test
Myocardial Perfusion Scintigraphy ND ND ND ND ND ND ND ND ND ND ND ND ND
Esophageal
Cardiac CAG Event
+
+
+
ND ND
5 0 0 3 0 0 0 0 0 0 0 0 0 0 6 8 1 0 1 5 0 0 0 2 0 11 1 3
PTCA PTCA
ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND
ND
(1-5)
of Pain
HistoryEpisodes Symptom Esophagitis Pathologic During 24-h SAP, Relief With Origin of Reflux (Grade) Study Omeprazole Chest Pain
ND ND
+
ND
Score
No.
ND ND ND
ND ND ND ND
ND
Evaluation by Gastroenterologist
i r
PTCA
MI
95
E E U
EandC
99 ND
21 98 2 92 8
ND
U U c u u E u u E u c E u E C u
EandC
U u u u E c E
13
95
sign (+)=positive; minus sign (-)=negative; plus/minus sign (±)=not conclusive; ND=not done; E=esophageal; C=cardiac; U=unidentified; PTCA=percutaneous transluminal coronary angioplasty; MI=myocardial infarction; CAG=coronary angiography.
*Plus
patients finally diagnosed as having cardiac and esophageal disorders was tested with the Wilcoxon test for unpaired data.
tween
Results Two of the 30 patients dropped out of the study; 1 pa¬
(patient 27) could not tolerate the catheters for more than 6 h and in 1 patient (patient 18), the 24-h monitoring was not successful because of technical failure. One patient had a myocardial infarction 14 h after starting the 24-h ambulatory monitoring, but the patient remained included in this study. All patients except the patient who developed the infarction performed an exercise test. In 18 myocardialresults of the exercise test were considered patients, normal; in 2 patients, results were abnormal; and in 7 patients, results were not conclusive. In five patients, a cardiac ultrasound study was performed, all with normal results. In seven patients, myocardial perfusion scintigraphy was performed. In five cases the result was considered normal, and in two cases it was suggestive of myocardial ischemia. In three patients, coronary arteriography was performed; in all three cases signif¬ tient
icant coronary stenosis 1212
was
diagnosed.
As can be seen in Table 1, five patients suffered from
ischemic coronary artery disease (proved by scintigraphy and/or coronary angiography; one patient had a myocardial infarction during ambulatory esoph¬ ageal monitoring). In only two of these five patients, the history was considered highly indicative of a cardiac origin of the chest pain (score 4 to 5). In only three of the seven patients with a cardiologic history score of 4 to 5, the chest pain was eventuallydiagnosed as cardiac (Fig 1). The cardiologic follow-up period of all patients was at least 6 months. The median time interval between the first visit to the cardiologist and gastroenterologic evaluation was 3 days (range,of 1 to 20 days). The gastroenterologic assessment all 28 patients is described in Table 1. In all patients, baseline esophageal manometric within normal limits. At endoscopy, four findings were patients were found to have a grade I reflux esophagi¬ tis (Savary-Miller). One of these patients also had Barrett's metaplasia. Seven patients had a sliding hiatal hernia. Eighteen patients had no abnormalities. Except for the one patient mentioned above, all severe
Clinical
Investigations
patients tolerated the 24-h monitoring procedure well. Twelve of the 28 patients showed pathologic GER duringthepHother monitoring: 4 of these had reflux esoph¬ 8 had no abnormalities at endoscopy. agitis, Ten of the 12 patients with pathologic reflux (includ¬ ing all patients with reflux esophagitis) were treated with omeprazole shortly after the 24-h monitoring. Nine patients were free of chest pain thereafter. Two of these patients were not treated with acid inhibition: one patient was already free of symptoms with betaand one patient developed a blocking medication, infarction myocardial during 24-h esophageal moni¬ toring. Eleven patients experienced one or more periods of pain during ambulatory monitoring of esophageal pH and pressure. In four of these patients, the pain and GER were associated, as reflected by a episodesSAP. There were no motility-related pain ep¬ positive isodes during 24-h pressure monitoring in any of the 28 patients. Based on the symptom relief upon acid inhibition, in nine patients the chest pain was consid¬ ered to be due to pathologic reflux. In only three of the ten patients with symptomatic reflux, the interviewing gastroenterologist classified the history as highly indicative of an esophageal origin of the chest pain (score 4 to 5). In only three of the six patients with a gastroenterologic history score of 4 or 5, the chest pain was finally diagnosed as esophageal (Fig 2). The history score for an esophageal origin was not significantly higher in patients finally diagnosed as having esophageal chest pain than in the patients otherwise (Wilcoxon test, p=0.18). diagnosed In 2 of the 28 patients, a combination of both symptomatic cardiac and esophageal abnormalities was found.
Discussion It has long been recognized that the differential di¬
Cardiac
Esophagea
Both
Unexplained
FINAL DIAGNOSIS
Figure 1. Relationship between the probability of a cardiac origin of the chest pain, as scored by the cardiologist on history data, and the final diagnosis.
chest pain was predicted correctly on the basis of the
history. All patients in our study were newly referred, by their general practitioner, with angina-like chest pain to the outpatient cardiologic department. These pa¬ tients represent a large proportion of the patients seen in daily practice. The results of our study confirm the notion that medical history often provides insufficient information to allow reliable differentiation between cardiac and esophageal causes and that further diag¬ nostic workup is always necessary. Common clinical practice is that gastroenterologic evaluation is started only after a cardiac origin of the symptoms has been excluded beyond reasonable doubt, requiring exten¬ sive cardiologic workup in some cases. To our knowledge, this study is the first in which evaluation was performed in a very gastroenterologic in the routine diagnostic workup of out¬ early phase with patients angina-like chest pain referred to a car¬ diologic department. We would like to stress the fact
agnosis of retrosternal chest pain can create difficulties. The "classic" symptoms of angina pectoris can also be found in patients with esophageal causes of retroster¬ nal chest pain. This can be explained by the fact that the esophagus and the heart have a similar innervation. Based on history only, it is difficult to distinguish be¬ tween cardiac and esophageal abnormalities.1213 In our study, the medical history was taken by a
cardiologist and a gastroenterologist separately. Based history alone, the probability of either a cardiac or esophageal origin of the patient's chest pain was assessed. The results were astonishing: in only 40% (two of five) of the patients with severe ischemic cor¬ onary artery disease the probability of a cardiac origin of the chest pain was scored as "high" (score 4 to 5). In only 37% (three of eight) of the patients with esophageal abnormalities the esophageal cause of the on an
Cardiac
Esophageal
Both
Unexplained
FINAL DIAGNOSIS
Figure 2.
Relationship between the probability of an esophageal
origin of the chest pain, as scored by the gastroenterologist on his¬ tory data, and the final diagnosis. CHEST/109/5/MAY,
1996
1213
that our study was deliberately meant not to interfere with normal daily practice in the cardiologic workup. Therefore, the cardiologic evaluation ended when the cardiologist believed there was sufficient proof for the diagnosis. To perform a coronary angiography in every patient would, in our opinion, not have been ethical. However, this approach might have led to a slight un¬ derestimation of the number of patients with cardiac and overlapping cardiac/esophageal problems. The results show that in a population of patients with chest pain, the prevalence of reflux-related disease is high: 43% (12 of 28) of the patients showed pathologic reflux during 24-h ambulatory pH monitoring. In 36% (10 of 28), the reflux was symptomatic, as proved by a relief of symptoms after acid inhibition (omeprazole) or an SAP of 95% or more. In this study, only 39% (11 of 28) of the patients experienced symptoms during the 24-h esophageal monitoring. This finding limits the value of the analy¬ sis of the association between symptoms and reflux. Therefore, we considered the chest pain also to be re¬ flux related in case of symptom relief after acid inhi¬ bition. The results of this study indicate that it is both feasible and worthwhile to perform a diagnostic workup for esophageal causes of chest pain in patients newly referred with angina-like chest pain to a cardiologic department. In our studypopulation, the prevalence of reflux-related disease was high (43%). Therefore, the combination of endoscopy and 24-h ambulatory pH monitoring is probably the most cost-effective ap¬ this condition. One could argue that endos¬ proach innot did contribute to the final outcome, since all copy four patients with esophagitis had pathologic reflux and symptom relief with omeprazole. However, it is well known that esophagitis, pathologic reflux, and symp¬ tomatic response to treatment do not correlate per¬ serves to exclude other fectly. Furthermore, endoscopy causes of chest esophageal pain, such as a tumor. Therefore, we believe that endoscopy is a mandatory part of the workup in these patients. The prevalence of motility disorders in patients with noncardiac chest pain is notoriously low. The preva¬ lence in our study (0%) confirms this low prevalence. This result is in contrast to the study by Lam et al6 who found similar prevalences of motility disorders and reflux in a group of patients who were admitted to a coronary care unit because of chest pain. Possibly, part of the motility disorders observed were secondaiy to the stress of the coronary care unit admission. The discrepancy might also be explained by the difference in time interval between first presentation and esoph¬ ageal evaluation; in the study by Lam et al,6 the eval¬ uation was performed within 1 h after hospital admis¬ sion, whereas in our study, this period varied between 1214
et al6 also found a days. The fact that Lam of pain episodes during substantially higherprevalence registration supports the latter hypothesis. Based on the low prevalence of motility disorders, we conclude that conventional manometry does not contribute much to the diagnostic workup. However, conventional manometry is the best technique to localize the upper border of the LES. In this study, possible or definite signs of coronary artery disease were ultimately diagnosed only in the of with either abnormal or
1 and 20
nonconclugroup patients sive results of the exercise test. Therefore, we propose that the diagnostic search for an esophageal cause should begin early in patients with chest pain who have relatively normal resting ECG, normal results of car¬ diac enzyme tests, and a normal result of an exercise test.
References 1 Soffer
EE, Scalabrini P, Wingate DC. Spontaneous noncardiac
pain: value of ambulatory esophageal pH and motility monitoring. Dig Dis Sci 1989; 34:1651-55 2 Janssens J, Vantrappen G, Ghillebert G. 24-h recording of esophageal pressure and pH in patients with noncardiac chest pain. Gastroenterology 1986; 90:1978-84 3 Breumelhof R, Nadorp JHSM, Akkermans LMA, et al. Analysis of 24-hour esophageal pressure and pH data in unselected patients with noncardiac chest pain. Gastroenterology 1990; 99: chest
1257-64
4 Wu WC, Castell DO, Richter JE.
Spontaneous noncardiac chest pain: evaluation by 24-h ambulatory esophageal motility and pH monitoring. Gastroenterology 1988; 94:878-86 5 Nevens F, Janssens J, Piessens J, et al. Prospective study on prevalence of esophageal chest pain in patients referred on an elective basis to a cardiac unit for suspected myocardial ischemia Dig Dis Sci 1991; 36:229-35 6 Lam HGTh, Dekker W, Kan G, et al. Acute noncardiac chest pain in a coronary care unit: evaluation by 24-hour pressure and pH recording of the esophagus. Gastroenterology 1992; 102:453-60 7 Smout AJPM, Breedijk M, van der Zouw C, et al. Physiological gastroesophageal reflux and esophageal motor activity studied with a new system for 24-hour recording and automated analysis. Dig Dis Sci 1989; 34:372-78 8 Richter JE, Bradley LA, De Meester TR, et al. Normal 24-hour ambulatory esophageal pH values: influence of study center, age and gender. Dig Dis Sci 1992; 37:849-56 9 Lam HGTh, Breumelhof R, Roelofs JMM, et al. What is the op¬ timal time window in symptom analysis of 24-hour esophageal pressure and pH data? Dig Dis Sci 1994; 39:402-09 10 Wiener GJ, Richter JE, Copper JB, et al. The symptom index: a clinically more important parameter of ambulatory 24-h esopha¬ geal pH monitoring. Am J Gastroenterol 1988; 83:358-61 11 Weusten BLAM, Roelofs JMM, Akkermans LMA, et al. The symptom association probability: an improved method for symp¬ tom analysis of 24-h esophageal pH data. Gastroenterology 1994; 107:1741-45 12 Alban-Davies H, Jones DB, Rhodes J. Angina-like esophageal chest pain: differentiation from cardiac pain by history. J Clin Gastroenterol 1986; 7:477-81 RE, Stenson WF, Avioli LV. Esophageal motility disor¬ ders and chest pain. Arch Intern Med 1985; 145:903-06
13 Clouse
Clinical Investigations
Study objective: The prevalence of esophageal disorders (dysmotility and/or gastroesophageal re¬ study were to investigate the prevalence of esophageal abnormalities in these patients and to assess the value of medical history in predicting the origin of the patient's chest pain. Design: We evaluated 28 consecutive patients who were newly referred to the cardiologist because of angina-like chest pain. Patients with evidence of severe myocardial ischemia were excluded. Cardiologic evaluation included medical history, physical examination, ECG, and exercise testing; further cardiologic workup was carried out only when considered necessary. Gastroenterologic evaluation consisted of medical history, esophageal manometry, endoscopy, and 24-h ambulatory monitoring of esophageal pH and pressure. Measurements and results: In five patients a diagnosis of ischemic coronary artery disease was made. In only two of these five patients, the cardiologic history strongly suggested a cardiac origin of the pain. Twelve patients had a pathologic 24-h pH profile, four of whom also had reflux esophagitis. Ten patients had symptomatic reflux. In only three of these ten patients, the history was judged to be indicative of an esophageal origin of the chest pain. No motility disorders were found. Conclusions: Thirty-six percent of the patients with chest pain newly referred to a cardiologic out¬ patient clinic have symptomatic gastroesophageal reflux. Neither cardiologic nor gastroenterologic history data have a high predictive value with respect to the origin of the chest pain.
flux) in patients with chest pain newly referred to a cardiologic clinic is unknown. The aims of our
(CHEST 1996; 109:1210-14)
Key words: esophageal monitoring; gastroesophageal reflux disease; noncardiac chest pain Abbreviations: GER=gastroesophageal reflux; GERD=gastroesophageal reflux disease; LES=lower esophageal sphincter; GERD=gas ability; fSI=symptom index SAP=symptom association probability;
patients with angina-like chest pain consult a ]V/|*ost -*-*-¦to have a cardiac of the
cardiologist
origin
pain
identified and treated. However, angina-like retrosternal chest pain is not always caused by cardiac disease. For editorial comment
page 1139 abnormalities and gastroesoph¬ see
Esophageal motilityare reflux ageal
in those
(GER) likely to play an etiologic role patients in whom cardiac disease has been
excluded. These esophageal abnormalities can be detected by
24-h ambulatory esophageal endoscopy, manometry, monitoring, and provocation tests. Since 1986, many studies on esophageal monitoring in patients with (Drs. Voskuil, Departments of Breumelhof, and Timmer) and Gastroenterology Cardiologv (Dr.'Cramer), St. Antonius Hospital, Nieuwegein, the Netherlands, and the Depart¬ ment of Gastroenterology7 (Dr. Smout), University Hospital, Utrecht, the Netherlands.
*From the
Manuscript received April 26, 1995; revision accepted October 13. 1210
noncardiac chest pain were reported. The proportion of patients in whom a correlation was found between esophageal abnormalities and pain varied from 10 to 90%,1_6 differences in yield depending on differences in patient selection and diagnostic criteria. Until now, most studies have been performed in patients referred to the gastroenterologist in a late of the diagnostic pathway, after a cardiac origin phase of the symptoms had been excluded. In
a
recent
study by
Lam et
al,6 patients
were
studied in an early phase of the diagnostic pathway, ie, within 1 h after admission to a coronary care unit with acute chest pain. In 27 of 30 patients (90%) in whom cardiac ischemia was excluded, one or more chest pain events without ECG changes were found to be corre¬ lated with an esophageal abnormality. We hypothesized that the prevalence of an esopha¬ geal abnormality in unselected patients who were re¬ ferred for the first time to the cardiologic outpatient clinic of a large general hospital because of angina-like Clinical
Investigations
chest pain could also be substantial. If this hypothesis would be correct, this could have implications for the "standard" workup of these patients. Therefore, the first aim of this study was to investi¬ gate the prevalence of esophageal abnormalities in a very early phase in the diagnostic workup of patients referred for the first time to the cardiologic department of our hospital because of angina-like chest pain, im¬ mediately after life-threatening cardiac ischemia was excluded. The secondary aim of our study was to assess the value of medical history in predicting the origin of the patient's chest pain. Materials
and
Methods
patients with angina-like chest pain who had been newly referred to the outpatient cardiologic depart¬ ment of our hospital. The patients were 16 men and 14 women, with a median age of 53 years (range, 33 to 69 years). Patients with severe cardiac ischemia, congestive heart failure, previous cardiologic his¬ tory, thoracic trauma, or thoracic surgery were excluded. Patients with known esophageal or gastric disorders or surgery were also ex¬ cluded. The study protocol was approved by the medical ethics committee of our hospital. All patients gave written informed consent. In all patients, the cardiologic history was taken by one and the same investigator (M.J.C.). The incidence, duration, severity and character of the chest pain, and its relation to exercise, emotion, and temperature changes were assessed. In addition, the frequency of occurrence and severity of the symptoms orthopnoea, dyspnea, and palpitations were assessed. Furthermore, use of medication and the existence of cardiovascular risk factors were evaluated. Based on these history data, the probability of a cardiac origin of the patient's chest pain was scored on a scale ranging from 1 to 5 (l=very low, 2=low, 3=indifferent, 4=probable, 5=highly probable). Further cardiologic workup consisted of bicycle exercise testing. Additional evaluation such as cardiac ultrasound, myocardial perfusion scin¬ tigraphy, or coronary angiography was carried out only when con¬ sidered necessary by the patient's cardiologist. Within 3 days after the first visit to the cardiologist, the gastro¬ enterologic history was taken in all patients. All patients were interviewed by one and the same investigator (J.H.V.). The incidence, duration, severity and character of the chest pain, and its relation to exercise, emotion, and consumption of meals and/or were assessed. In addition, the frequency of occurrence beverages and the severity of the symptoms heartburn, regurgitation, dysphagia for solid and/or liquid food, and odynophagia were assessed. Furthermore, the patient's actions to relieve the symptoms (ie, rest, change of position, eating or drinking, and the use of antacids) were assessed. The interviewer then scored the probability of an esoph¬ ageal origin of the patient's symptoms on a scale ranging from 1 to 5 (low to high). Immediately after the interview, esophageal manometry, upper GI endoscopy and 24-h ambulatory esophageal pH and pressure monitoring were performed. At least 72 h before esophageal ma¬ nometry and 24-h ambulatory monitoring, treatment with any medication that could interfere with acid secretion or esophageal motility was stopped. A low-compliance perfusion pump with a three-lumen catheter with openings 5 cm apart and oriented in three different directions was used for conventional manometry. The lower esophageal sphincter (LES) was localized by the stationary pull-through method. Thereafter, the catheter was placed with the distal opening 5 cm above the upper border of the LES and ten wet swallows were given, 20 to 30 s apart. Amplitude, du¬ ration, and propagation of contractions at 5, 10, and 15 cm above the LES were assessed. We studied 30 consecutive
A video endoscopic system (Pentax) was used for esophageal endoscopy. Reflux esophagitis was described according to the Savary-Miller grading system (grades I through IV). Directly after¬ wards, a 24-h ambulatory pressure and pH recording system was installed. Intraesophageal pH was recorded with a glass electrode with incorporated reference (Ingold AG; Urdorf; Switzerland). In¬ traesophageal pressure was recorded using a 5F polyurethane catheter with three sensors 5 cm
pressure apart (Sentron 205; Sentron; Roden, the Netherlands). The pH electrode and the
pressure
distal pressure transducer were positioned 5 cm above the upper border of the LES. The signals were recorded with a digital porta¬ ble data logger (UPS 2020 GS; MMS; Enschede, the Netherlands). The three pressure signals were sampled at a rate of 4 Hz and the pH signal was sampled at a rate of 0.2 Hz. The data logger used al¬ lows the recording of five marker signals, of which only one was used, to indicate the episodes of chest pain. During the 24-h monitoring, periods of eating and drinking and periods spent in the supine position were registered by the patient in a diary. Episodes of chest pain were registered by pushing the event marker and writing down the time, nature, and characteris¬ tics of the symptoms. When the 24-h study period was ended, the recorded data were transferred to the computer, stored on a floppy disk, and analyzed in a fully automatic way using previously validated algorithms.' Esophageal acid exposure was calculated for the upright and su¬ pine positions separately and for the total time of registration. GER disease (GERD) was diagnosed when the percentage of time of pH below 4 exceeded the upper limit (95th percentile) of normal, ac¬ cording to the criteria proposed by Richter et al:8 pH less than 4, percent upright.8.15; pH less than 4, percent supine.3.45; and pH less than 4, percent total.5.78. In patients experiencing pain during the recording, an additional computerized analysis of the symptom episode was done using a 2-min window starting 2 min before pain onset and ending at the onset of pain.9 In the analysis of motility disorders, the patient's symptom-free episodes were used as his or her own control. In the analysis of reflux, a symptom episode was considered to be associated with reflux when a pH drop of more than 1 U occurred within 5 s, or when the pH fell below 4, or when both criteria were met during that episode. This computerized analysis of the symptom episode was published in detail elsewhere.6 The last step in the analysis consisted of the calculation of the symptom index (SI) for abnormal motility, and the symptom association probability (SAP) for reflux. The SI was calculated according to Wiener et al.10 A patient was considered to have a positive correlation between his or her pain and motor abnormality when the SI was more than 50%. The SAP was calcu¬ lated as described by Weusten et al.11 The SAP expresses the strength of the association between symptoms and reflux episodes taking all relevant variables (number of reflux episodes, number of symptom episodes, number of associated symptom episodes) into account. A patient was considered to have a positive correlation between his or her pain and reflux when the SAP was 95% or greater. Therapy with acid-secretion inhibitors (omeprazole, 40 mg daily) was started when the diagnosis of GERD, with or without reflux esophagitis, was made. A patient was considered to suffer from symptomatic GERD when there was a relief in chest pain after ini¬ tiating therapy. Finally, the chest pain of all patients was classified as cardiac (as proved by exercise test, scintigraphy, coronary angiog¬
or myocardial event), esophageal (as proved by a positive SI raphy, for dysmotility, a SAP of 95% or greater, and/or relief in chest pain after initiating therapy in patients with GERD), cardiac and esophageal, or unidentified (noncardiac, nonesophageal) in origin. Obviously, we performed a "routine" cardiologic workup com¬ pared with a more extensive gastroenterologic investigation. This was done to prove our hypothesis that the prevalence of esophageal disorders in angina-like chest pain is substantial. The statistical significance of differences in history scores be-
CHEST/109/5/MAY,
1996
1211
Table 1.Patient Data* Evaluation by Cardiologist
No.
Cardiac History Exercise
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19 20 21 22 23 24 25 26 28 29 30
Score
(1-5)
Test
Myocardial Perfusion Scintigraphy ND ND ND ND ND ND ND ND ND ND ND ND ND
Esophageal
Cardiac CAG Event
+
+
+
ND ND
5 0 0 3 0 0 0 0 0 0 0 0 0 0 6 8 1 0 1 5 0 0 0 2 0 11 1 3
PTCA PTCA
ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND
ND
(1-5)
of Pain
HistoryEpisodes Symptom Esophagitis Pathologic During 24-h SAP, Relief With Origin of Reflux (Grade) Study Omeprazole Chest Pain
ND ND
+
ND
Score
No.
ND ND ND
ND ND ND ND
ND
Evaluation by Gastroenterologist
i r
PTCA
MI
95
E E U
EandC
99 ND
21 98 2 92 8
ND
U U c u u E u u E u c E u E C u
EandC
U u u u E c E
13
95
sign (+)=positive; minus sign (-)=negative; plus/minus sign (±)=not conclusive; ND=not done; E=esophageal; C=cardiac; U=unidentified; PTCA=percutaneous transluminal coronary angioplasty; MI=myocardial infarction; CAG=coronary angiography.
*Plus
patients finally diagnosed as having cardiac and esophageal disorders was tested with the Wilcoxon test for unpaired data.
tween
Results Two of the 30 patients dropped out of the study; 1 pa¬
(patient 27) could not tolerate the catheters for more than 6 h and in 1 patient (patient 18), the 24-h monitoring was not successful because of technical failure. One patient had a myocardial infarction 14 h after starting the 24-h ambulatory monitoring, but the patient remained included in this study. All patients except the patient who developed the infarction performed an exercise test. In 18 myocardialresults of the exercise test were considered patients, normal; in 2 patients, results were abnormal; and in 7 patients, results were not conclusive. In five patients, a cardiac ultrasound study was performed, all with normal results. In seven patients, myocardial perfusion scintigraphy was performed. In five cases the result was considered normal, and in two cases it was suggestive of myocardial ischemia. In three patients, coronary arteriography was performed; in all three cases signif¬ tient
icant coronary stenosis 1212
was
diagnosed.
As can be seen in Table 1, five patients suffered from
ischemic coronary artery disease (proved by scintigraphy and/or coronary angiography; one patient had a myocardial infarction during ambulatory esoph¬ ageal monitoring). In only two of these five patients, the history was considered highly indicative of a cardiac origin of the chest pain (score 4 to 5). In only three of the seven patients with a cardiologic history score of 4 to 5, the chest pain was eventuallydiagnosed as cardiac (Fig 1). The cardiologic follow-up period of all patients was at least 6 months. The median time interval between the first visit to the cardiologist and gastroenterologic evaluation was 3 days (range,of 1 to 20 days). The gastroenterologic assessment all 28 patients is described in Table 1. In all patients, baseline esophageal manometric within normal limits. At endoscopy, four findings were patients were found to have a grade I reflux esophagi¬ tis (Savary-Miller). One of these patients also had Barrett's metaplasia. Seven patients had a sliding hiatal hernia. Eighteen patients had no abnormalities. Except for the one patient mentioned above, all severe
Clinical
Investigations
patients tolerated the 24-h monitoring procedure well. Twelve of the 28 patients showed pathologic GER duringthepHother monitoring: 4 of these had reflux esoph¬ 8 had no abnormalities at endoscopy. agitis, Ten of the 12 patients with pathologic reflux (includ¬ ing all patients with reflux esophagitis) were treated with omeprazole shortly after the 24-h monitoring. Nine patients were free of chest pain thereafter. Two of these patients were not treated with acid inhibition: one patient was already free of symptoms with betaand one patient developed a blocking medication, infarction myocardial during 24-h esophageal moni¬ toring. Eleven patients experienced one or more periods of pain during ambulatory monitoring of esophageal pH and pressure. In four of these patients, the pain and GER were associated, as reflected by a episodesSAP. There were no motility-related pain ep¬ positive isodes during 24-h pressure monitoring in any of the 28 patients. Based on the symptom relief upon acid inhibition, in nine patients the chest pain was consid¬ ered to be due to pathologic reflux. In only three of the ten patients with symptomatic reflux, the interviewing gastroenterologist classified the history as highly indicative of an esophageal origin of the chest pain (score 4 to 5). In only three of the six patients with a gastroenterologic history score of 4 or 5, the chest pain was finally diagnosed as esophageal (Fig 2). The history score for an esophageal origin was not significantly higher in patients finally diagnosed as having esophageal chest pain than in the patients otherwise (Wilcoxon test, p=0.18). diagnosed In 2 of the 28 patients, a combination of both symptomatic cardiac and esophageal abnormalities was found.
Discussion It has long been recognized that the differential di¬
Cardiac
Esophagea
Both
Unexplained
FINAL DIAGNOSIS
Figure 1. Relationship between the probability of a cardiac origin of the chest pain, as scored by the cardiologist on history data, and the final diagnosis.
chest pain was predicted correctly on the basis of the
history. All patients in our study were newly referred, by their general practitioner, with angina-like chest pain to the outpatient cardiologic department. These pa¬ tients represent a large proportion of the patients seen in daily practice. The results of our study confirm the notion that medical history often provides insufficient information to allow reliable differentiation between cardiac and esophageal causes and that further diag¬ nostic workup is always necessary. Common clinical practice is that gastroenterologic evaluation is started only after a cardiac origin of the symptoms has been excluded beyond reasonable doubt, requiring exten¬ sive cardiologic workup in some cases. To our knowledge, this study is the first in which evaluation was performed in a very gastroenterologic in the routine diagnostic workup of out¬ early phase with patients angina-like chest pain referred to a car¬ diologic department. We would like to stress the fact
agnosis of retrosternal chest pain can create difficulties. The "classic" symptoms of angina pectoris can also be found in patients with esophageal causes of retroster¬ nal chest pain. This can be explained by the fact that the esophagus and the heart have a similar innervation. Based on history only, it is difficult to distinguish be¬ tween cardiac and esophageal abnormalities.1213 In our study, the medical history was taken by a
cardiologist and a gastroenterologist separately. Based history alone, the probability of either a cardiac or esophageal origin of the patient's chest pain was assessed. The results were astonishing: in only 40% (two of five) of the patients with severe ischemic cor¬ onary artery disease the probability of a cardiac origin of the chest pain was scored as "high" (score 4 to 5). In only 37% (three of eight) of the patients with esophageal abnormalities the esophageal cause of the on an
Cardiac
Esophageal
Both
Unexplained
FINAL DIAGNOSIS
Figure 2.
Relationship between the probability of an esophageal
origin of the chest pain, as scored by the gastroenterologist on his¬ tory data, and the final diagnosis. CHEST/109/5/MAY,
1996
1213
that our study was deliberately meant not to interfere with normal daily practice in the cardiologic workup. Therefore, the cardiologic evaluation ended when the cardiologist believed there was sufficient proof for the diagnosis. To perform a coronary angiography in every patient would, in our opinion, not have been ethical. However, this approach might have led to a slight un¬ derestimation of the number of patients with cardiac and overlapping cardiac/esophageal problems. The results show that in a population of patients with chest pain, the prevalence of reflux-related disease is high: 43% (12 of 28) of the patients showed pathologic reflux during 24-h ambulatory pH monitoring. In 36% (10 of 28), the reflux was symptomatic, as proved by a relief of symptoms after acid inhibition (omeprazole) or an SAP of 95% or more. In this study, only 39% (11 of 28) of the patients experienced symptoms during the 24-h esophageal monitoring. This finding limits the value of the analy¬ sis of the association between symptoms and reflux. Therefore, we considered the chest pain also to be re¬ flux related in case of symptom relief after acid inhi¬ bition. The results of this study indicate that it is both feasible and worthwhile to perform a diagnostic workup for esophageal causes of chest pain in patients newly referred with angina-like chest pain to a cardiologic department. In our studypopulation, the prevalence of reflux-related disease was high (43%). Therefore, the combination of endoscopy and 24-h ambulatory pH monitoring is probably the most cost-effective ap¬ this condition. One could argue that endos¬ proach innot did contribute to the final outcome, since all copy four patients with esophagitis had pathologic reflux and symptom relief with omeprazole. However, it is well known that esophagitis, pathologic reflux, and symp¬ tomatic response to treatment do not correlate per¬ serves to exclude other fectly. Furthermore, endoscopy causes of chest esophageal pain, such as a tumor. Therefore, we believe that endoscopy is a mandatory part of the workup in these patients. The prevalence of motility disorders in patients with noncardiac chest pain is notoriously low. The preva¬ lence in our study (0%) confirms this low prevalence. This result is in contrast to the study by Lam et al6 who found similar prevalences of motility disorders and reflux in a group of patients who were admitted to a coronary care unit because of chest pain. Possibly, part of the motility disorders observed were secondaiy to the stress of the coronary care unit admission. The discrepancy might also be explained by the difference in time interval between first presentation and esoph¬ ageal evaluation; in the study by Lam et al,6 the eval¬ uation was performed within 1 h after hospital admis¬ sion, whereas in our study, this period varied between 1214
et al6 also found a days. The fact that Lam of pain episodes during substantially higherprevalence registration supports the latter hypothesis. Based on the low prevalence of motility disorders, we conclude that conventional manometry does not contribute much to the diagnostic workup. However, conventional manometry is the best technique to localize the upper border of the LES. In this study, possible or definite signs of coronary artery disease were ultimately diagnosed only in the of with either abnormal or
1 and 20
nonconclugroup patients sive results of the exercise test. Therefore, we propose that the diagnostic search for an esophageal cause should begin early in patients with chest pain who have relatively normal resting ECG, normal results of car¬ diac enzyme tests, and a normal result of an exercise test.
References 1 Soffer
EE, Scalabrini P, Wingate DC. Spontaneous noncardiac
pain: value of ambulatory esophageal pH and motility monitoring. Dig Dis Sci 1989; 34:1651-55 2 Janssens J, Vantrappen G, Ghillebert G. 24-h recording of esophageal pressure and pH in patients with noncardiac chest pain. Gastroenterology 1986; 90:1978-84 3 Breumelhof R, Nadorp JHSM, Akkermans LMA, et al. Analysis of 24-hour esophageal pressure and pH data in unselected patients with noncardiac chest pain. Gastroenterology 1990; 99: chest
1257-64
4 Wu WC, Castell DO, Richter JE.
Spontaneous noncardiac chest pain: evaluation by 24-h ambulatory esophageal motility and pH monitoring. Gastroenterology 1988; 94:878-86 5 Nevens F, Janssens J, Piessens J, et al. Prospective study on prevalence of esophageal chest pain in patients referred on an elective basis to a cardiac unit for suspected myocardial ischemia Dig Dis Sci 1991; 36:229-35 6 Lam HGTh, Dekker W, Kan G, et al. Acute noncardiac chest pain in a coronary care unit: evaluation by 24-hour pressure and pH recording of the esophagus. Gastroenterology 1992; 102:453-60 7 Smout AJPM, Breedijk M, van der Zouw C, et al. Physiological gastroesophageal reflux and esophageal motor activity studied with a new system for 24-hour recording and automated analysis. Dig Dis Sci 1989; 34:372-78 8 Richter JE, Bradley LA, De Meester TR, et al. Normal 24-hour ambulatory esophageal pH values: influence of study center, age and gender. Dig Dis Sci 1992; 37:849-56 9 Lam HGTh, Breumelhof R, Roelofs JMM, et al. What is the op¬ timal time window in symptom analysis of 24-hour esophageal pressure and pH data? Dig Dis Sci 1994; 39:402-09 10 Wiener GJ, Richter JE, Copper JB, et al. The symptom index: a clinically more important parameter of ambulatory 24-h esopha¬ geal pH monitoring. Am J Gastroenterol 1988; 83:358-61 11 Weusten BLAM, Roelofs JMM, Akkermans LMA, et al. The symptom association probability: an improved method for symp¬ tom analysis of 24-h esophageal pH data. Gastroenterology 1994; 107:1741-45 12 Alban-Davies H, Jones DB, Rhodes J. Angina-like esophageal chest pain: differentiation from cardiac pain by history. J Clin Gastroenterol 1986; 7:477-81 RE, Stenson WF, Avioli LV. Esophageal motility disor¬ ders and chest pain. Arch Intern Med 1985; 145:903-06
13 Clouse
Clinical Investigations