C 2005) Digestive Diseases and Sciences, Vol. 50, No. 9 (September 2005), pp. 1611–1615 (
DOI: 10.1007/s10620-005-2904-2

Acid Perfusion Test: A Useful Test for Evaluating Esophageal Acid Sensitivity ALISTAIR L. KING, PhD,* ANGELA ANGGIANSAH, PhD,† ROY ANGGIANSAH, BSc,† and TERRY WONG, MD‡

The acid perfusion test (APT) is a test of esophageal acid sensitivity. In gastroesophageal reflux disease (GERD), the APT has largely been superseded by indexes derived from 24-hr esophageal pH monitoring: symptom index (SI), symptom sensitivity index (SSI), and symptom association probability (SAP). To evaluate the role of APT in patients with GERD we compared values of SI, SSI, and SAP in 126 APT+ and 146 APT− patients. Median values for SI, SSI, and SAP were higher in the APT+ compared with the APT− group (P < 0.001). More patients had a positive SI, SSI, and SAP in the APT+ compared with the APT– group (SI, P < 0.0001; SSI and SAP, P < 0.001). Patients with a negative APT were unlikely to have symptoms caused by acid reflux (negative predictive value, 86%). APT may be useful in detecting acid sensitivity in patients who do not report symptoms on 24-hr esophageal pH monitoring. KEY WORDS: acid perfusion test; Bernstein test; gastroesophageal reflux disease; esophageal acid sensitivity; pH monitoring.

Gastroesophageal reflux disease (GERD) is a common condition, affecting approximately 20% of people in the United States on at least a weekly basis (1). “Typical” symptoms of GERD are heartburn and regurgitation, however, GERD may also present with atypical symptoms (chest pain, asthma/cough, hoarseness, sore throat) or with complications (ulceration, strictures, Barrett’s metaplasia) (2). In many cases investigation is unnecessary because the history is sufficiently revealing to identify the presence of troublesome reflux disease (3). However, a proportion of patients presents more of a diagnostic dilemma. Endoscopy may be performed to look for evidence of mucosal injury, however, approximately 75% of patients will have no endoscopic evidence of esophagitis (4, 5). One of the first investigations used in GERD was the acid per-

Manuscript received November 8, 2004; accepted January 6, 2005. From the *Department of Gastroenterology, Hemel Hempstead General Hospital, Hillfield Road, Hemel Hempstead, Herts HP2 4AD, and †Oesophageal Laboratory and ‡Department of Gastroenterology, St Thomas’ Hospital, Lambeth Palace Road, London SE1 7EH, UK. Address for reprint requests: Alistair King, Department of Gastroenterology, Hemel Hempstead General Hospital, Hillfield Road, Hemel Hempstead, Herts HP2 4AD, UK; [email protected].

fusion test, first described by Bernstein in 1958 (6). Up to 100 ml 0.1 N HCl is instilled into the distal esophagus, alternating with the instillation of normal saline. Reproduction of the patient’s symptoms with the instillation of acid is evidence that there is esophageal sensitivity to acid and that acid is the cause of the reflux symptoms. The development of 24-hr esophageal pH monitoring in the 1970s allowed for accurate documentation of reflux episodes as manifest by a fall in the distal esophageal pH to <4.0 (7). However, a certain number of short-lived reflux episodes occur in normal subjects (known as “physiologic” reflux). Several definitions have been published of what constitutes “abnormal” reflux. Richter et al. (8) studied data from a total of 110 healthy volunteers and defined abnormal reflux as total % time pH <4 is greater than 5.78%. However, generally, a study is considered abnormal when total % time pH <4 is greater than 4.2%. The DeMeester score is a composite scoring system that quantitates the degree of departure of the patient’s reflux pattern in the day and nighttime segments from the physiologic reflux pattern found in control individuals. An alternative definition of abnormal reflux is a DeMeester score of >14.7 (9).

Digestive Diseases and Sciences, Vol. 50, No. 9 (September 2005) C 2005 Springer Science+Business Media, Inc. 0163-2116/05/0900-1611/0


1611

KING ET AL.

Fig 1. Correlating reflux episodes with symptoms. The symptom event markers (♥) sometimes correlate with significant refluxes and sometimes do not.

The presence of abnormal reflux does not necessarily prove that reflux is the cause of the patient’s symptoms. However, during a 24-hr esophageal pH study patients also have the opportunity to log their symptoms, and a good correlation between symptom episodes and reflux episodes is evidence that the symptoms are due to GERD (see Figure 1). A number of indexes have been described that correlate reflux and symptom episodes. The symptom index (SI) (10) is defined as the number of reflux-related symptom episodes divided by the total number of symptom episodes expressed as a percentage. The SI is the symptom correlation parameter used by most study software packages and, hence, is the parameter used in most esophageal laboratories. However, the SI has the disadvantage that it does not reflect the total number of reported symptoms or the total number of reflux episodes, so SI = 50% could represent 50 of 100 or merely 1 of 2. The symptom sensitivity index (SSI) (11) is defined as the percentage of symptom-associated reflux episodes ([symptom-associated reflux episodes/total number of refluxes] ×100%). Like the SI, the SSI does not take account of total number of symptom episodes. For both SSI and SI the cutoff for positive is purely arbitrary, and symptoms may be associated with reflux purely by chance. The symptom association probability (SAP) (12) applies the Fishers’ exact test to analyze the four potential associations between symptoms and reflux (reflux and pain, reflux and no pain, pain and no reflux, and no reflux and no pain) to produce a p value. The SAP is then defined as: SAP = (1.0 − p) × 100%. The SAP has the advantage that it takes into account both the total number of symptom

1612

episodes and the total number of reflux episodes. The SAP also controls for chance association and, for these reasons, is regarded by many as the best index of symptom/reflux association. Patients with reflux symptoms referred to our laboratory have an acid perfusion test (APT) (6) performed at the time of their esophageal manometry. This study aimed to investigate if the APT has a role in evaluating typical reflux symptoms. We thus sought to validate the APT by comparing SI, SSI, and SAP in APT+ versus APT− patients with typical reflux symptoms. METHODS Patients. The Esophageal Laboratory at St Thomas’ Hospital, London, is a regional referral center. We included in this study patients referred to the laboratory between January 2002 and September 2003 with symptoms of heartburn or acid regurgitation. In that period 272 patients had both successful APT and 24-hr esophageal pH monitoring. Of these 126 were APT+ (72 male, 54 female; mean age, 44.9 years), and 146 were APT− (68 male, 78 female; mean age, 48.3 years). Esophageal Manometry. All patients referred to the Esophageal Laboratory undergo static manometry and 24-hr ambulatory esophageal pH monitoring as a first-line investigation. Static manometry is performed in all patients, including those with reflux symptoms, for two reasons. The first reason is the manometric localization of the lower esophageal sphincter (LES) so that the pH sensor can be accurately placed, and the second is to exclude a primary motility disorder such as achalasia (13). Acid Perfusion Test. A 6F enteral feeding tube (Portex Ltd., UK) was bonded to a solid-state pressure catheter with a diameter of 2.3 mm (Gaeltec Ltd., UK) to ensure one intubation for the manometry and the APT. Patients with reflux symptoms after Digestive Diseases and Sciences, Vol. 50, No. 9 (September 2005)

ACID PERFUSION TEST exclusion of achalasia have an APT performed after completion of their esophageal manometry. The tip of the perfusion tube was placed 10 cm above the manometrically defined LES. Previous experience (unpublished data) showed that instillation of normal saline was unlikely to produce heartburn but was likely to abolish heartburn after acid perfusion. Thus in order to reduce the procedure time, the APT was performed as a modified version of the test described by Bernstein et al. (6) as follows: Up to 100 ml 0.1 N HCl was instilled at a rate of 10 ml/min. A test result was considered positive if the patient spontaneously perceived pain or burning during acid instillation. The range of volume of HCl to produce a positive response was between 5 and 100 ml. Failure of the HCl to reproduce the patient’s reflux symptoms after perfusion of 100 ml was deemed a negative result. The addition of the APT to the standard tests prolonged the testing procedure by a maximum of 10 min. However, while the APT was being performed, instructions were given to patients on what they should do with the pH recorder and what food to avoid, etc., for the 24-hr pH test. Overall, the addition of the APT only added a few minutes to the total procedure time. Twenty-Four-Hour Esophageal pH Monitoring. Twentyfour-hour esophageal pH monitoring was performed after the completion of the APT. The antimony pH catheter was calibrated in buffer solutions of pH 7 and 1 according to the manufacturer’s instructions. The pH catheter was passed transnasally until the sensor registered low pH to ensure that the electrode was in the stomach and not curled in the esophagus. It was then withdrawn and placed 5 cm above the manometrically defined upper border of the LES and connected to a pH recorder (Digitrapper; Synectics Medical). Patients were asked to avoid acidic drinks and acidic foods during the test. After 24 hr the pH data were downloaded onto a computer for data analysis. During the 24-hr recording period patients were asked to press the event marker on the pH recorder whenever a symptom occurred. They were also asked to record the exact time of these symptoms in a diary, and these episodes were later confirmed on the pH data tracing. Symptom Correlation. The esophageal pH data were analyzed using the Polygram for Windows version 2.03 software (Syntenics Medical). A symptom episode was held to be associated with a reflux event if it coincided with the reflux event (pH <4) or occurred within the subsequent 2-min period (14). For each of the 272 patients the SI and SAP were calculated from the 24-hr esophageal pH record using the study software. For the SSI the number of symptom-associated reflux episodes was determined by examining the pH tracing, and the total number of reflux episodes was determined by the study software. The SSI was calculated from these values using the formula of Breurnelhof et al. (11). For patients who reported no symptoms, all three indexes were recorded as zero. SI = 50%, SSI = 10%, and SAP > 95% were defined as positive. Statistical Analysis. Median values for SI, SSI, and SAP in APT+ versus APT− patients were compared using the Mann– Whitney U test. The numbers of patients with a positive SI, SSI, or SAP were compared in APT+ and APT− groups using the chi-square test.

TABLE 1. MEDIAN VALUES FOR SI, SSI, AND SAP

SI SSI SAP

APT+

APT−

Significance*

66.7% 3.75% 83.3%

0% 0% 7.55%

P < 0.001 P < 0.001 P < 0.001

*Mann–Whitney U test.

(22%) APT−patients had a positive SI (χ 2 = 44.9, P < 0.0001), 28 of 126 (22%) APT+ vs 8 of 146 (5%) APT− patients had a positive SSI (χ 2 = 16.5, P < 0.001), and 48 of 126 (38%) APT+ vs 20 of 146 (14%) APT− patients had a positive SAP (χ 2 = 21.5, P < 0.001). These results are shown graphically in Figure 2. Sensitivity, specificity, positive, and negative predictive values for the APT compared to the SI, SSI, and SAP are listed in Table 2. The negative predictive value of the APT using the SAP as the gold standard was 86%. Thirty seven of 126 APT+ patients (29%) had “physiological” levels of acid reflux defined by either a DeMeester score of <14.7 or a total % time pH <4 of ≤4.2%. Thirteen of these 37 patients had a positive SI (≥50%). Twenty-two of 126 APT+ patients (17%), vs 64 of 146 APT−patients (44%), reported no symptoms on 24-hr esophageal pH monitoring (χ 2 = 21.8, P < 0.001). DISCUSSION The APT is an old test that has been largely superseded by newer investigations, in particular, 24-hr esophageal pH monitoring with symptom event recording. Our laboratory routinely performs an APT at the time the patient attends for esophageal manometry and 24-hr pH monitoring. The APT is a test of esophageal sensitivity to acid and looks at reproduction of reflux symptoms in response to exogenous acid. Twenty-four-hour esophageal pH monitoring looks for a correlation between the patient’s symptoms and spontaneous episodes of reflux of endogenous acid.

RESULTS Median values for SI, SSI, and SAP were significantly higher in the APT+ compared with the APT− group (see Table 1). Seventy-eight of 126 (62%) APT+ vs 32 of 146 Digestive Diseases and Sciences, Vol. 50, No. 9 (September 2005)

Fig 2. Numbers of patients with a positive SI, SSI, and SAP in the APT+ compared with the APT− group.

1613

KING ET AL. TABLE 2. SENSITIVITY, SPECIFICITY, AND POSITIVE PREDICTIVE VALUE OF APT COMPARED WITH SI, SSI, AND SAP

Sensitivity Specificity Positive predictive value Negative predictive value

SI

SSI

SAP

0.70 0.70 0.01 0.78

0.78 0.58 0.22 0.94

0.71 0.61 0.38 0.86

Nonetheless, 24-hr pH monitoring with symptom correlation is not considered by many investigators to represent an endogenous Bernstein test (10). The sensitivity of the APT in identifying GERD patients is 32% to 100% and the specificity is 40–100% (15, 16). The wide variation among studies possibly reflects differences in methodology, patient selection, and criteria for a positive test. The aim of this study was to investigate the role of the APT in investigating patients with typical reflux symptoms of heartburn and regurgitation. To do this the APT was compared with three common symptom indexes derived from 24-hr esophageal pH monitoring. It was found that values for SI, SSI, and SAP are all significantly higher in patients who have a positive APT. This study also notes that APT+ patients are more likely to be positive for all three of these indexes. Jung et al. (17) recently reported a poor association between acid sensitivity as detected by APT and that detected at 24-hr esophageal pH monitoring and SI. They proposed that this is because heartburn following acid perfusion and spontaneous heartburn are induced by different stimuli. Their data show that only 18 of 34 patients (53%) with a positive APT also had a positive SI and that 9 of 22 patients (41%) with a negative APT had a positive SI. In this study with a larger number of patients (272 vs 93 patients), the APT has performed rather better, with a positive SI in 62% of APT+ patients and only 22% of APT− patients. The value of the APT compared with the SI is also supported by previous studies noting that the APT is consistently positive in patients with a SI ≥50% regardless of the total acid exposure time (18). The SI, however, has limitations in that it does not consider the total number of reflux or symptom episodes and does not control for chance association. Hence, when this study was designed it was also decided to compare the APT with the SSI and SAP. Somewhat disappointingly we find concordance of positive APT and positive SSI of only 22% and of positive APT and positive SAP of only 38%. Nevertheless, these values are still significantly higher than the 5% (SSI) and 14% (SAP) positive results found in the APT− group. The main difference between this study and that of Jung et al. is the classi-

1614

fication of patients who do not report symptoms on their 24-hr esophageal pH studies. This study found that 17% of APT+ patients but 44% of APT− patients did not report symptoms during their studies. Although this technically means that SI, SSI, and SAP cannot be calculated, in most instances the lack of symptoms is likely to represent a “negative” study, and in contrast to Jung et al. (17) we have recorded all three indexes as zero. The effect of not doing this is to bias the data toward those with a degree of esophageal acid sensitivity, be they APT+ or APT−, and thus underestimate the ability of the APT to detect this acid sensitivity. This study regards the SAP as the most reliable parameter of symptom association. Compared to the SAP it was found that the APT had a good negative predictive value (86%), so patients with a negative APT are unlikely to have symptoms caused by acid reflux. Up to 30% of patients complaining of typical reflux symptoms have a normal endoscopy and esophageal acid exposure that falls within physiological values. If, however, they have a good correlation between their (physiological) reflux episodes and their symptoms (as determined by SI, SSI, or SAP), they may be considered to have an “acid-sensitive esophagus” (5, 19, 20). These patients have lower thresholds for perception of distension and discomfort in response to mechanical stimuli and, presumably, have a heightened perception of normal reflux events (21). This study shows that 29% of patients who have a positive APT, hence evidence of esophageal acid sensitivity, have physiological levels of acid secretion. Could these be patients with an acid-sensitive esophagus? Although only 13 of the 37 had a positive SI, it is possible that the APT is detecting at least some patients with an acid-sensitive esophagus that were not detected on pH monitoring. In a previous study 43% of patients complaining of heartburn had a normal acid contact time on 24-hour pH monitoring, but of these 89% had typical heartburn during an APT (20). In this study the APT was compared to indexes derived from 24-hr esophageal pH monitoring, considered by many to be the “gold standard” for evaluating the frequency and duration of acid reflux, however, pH monitoring is not without limitations. Volume of acid may play an important role in producing the patient’s symptoms, however, 24-hr pH monitoring gives no indication of the volume of endogenous acid present in the esophagus. Unlike the APT, 24-hr pH monitoring does not quantify the volume of (endogenous) acid needed to sensitize the esophagus. As far as acid volume is concerned the two tests are not directly comparable, and this may be one explanation for the poor relationship between the APT and the SI, SSI, and SAP. Digestive Diseases and Sciences, Vol. 50, No. 9 (September 2005)

ACID PERFUSION TEST

A further problem with 24-hr pH monitoring is that the patient’s symptoms may not occur during the test period. During a study patients are asked to avoid acid drinks and foods, and the patients may modify their diet and activities due to poor tolerability of the nasal probe (22). In addition, a pH study may miss reflux events because the pH probe is buried in a mucosal fold. These problems give rise to false-negative studies, and esophageal acid exposure may vary threefold on repeat testing (23). The Genval report of 1998 (24) noted that normal acid exposure values are recorded in up to a quarter of patients with otherwise typical reflux esophagitis. The report stated that neither 24-hr esophageal acid exposure nor symptom association was sufficiently sensitive or reliable to be used as a diagnostic gold standard. In this study 17% of APT-positive patients did not report any symptoms on their 24-hr esophageal pH study. It seems likely that some of these patients have genuine acid sensitivity with symptoms due to GERD or acid-sensitive esophagus but simply have not experienced their symptoms during the 24-hr pH test. Thus the APT may be useful in detecting acid sensitivity in those who do not report symptoms on 24-hr esophageal pH monitoring. In conclusion, this study shows that the APT is still a useful test for evaluating esophageal acid sensitivity. Patients who have a negative APT are highly unlikely to have symptoms due to esophageal acid sensitivity. This study did not show a strong correlation between a positive APT and a positive SI, SSI, or SAP, however, this may be as much a limitation of 24-hr esophageal pH monitoring as a limitation of the APT. It is possible that the APT may detect acid sensitivity in some patients with false-negative 24-hr pH studies. This study shows that the APT may be used as a complimentary test to the SI in evaluating patients with typical reflux symptoms. Finally, not all units offer 24-hr pH monitoring. Here the APT is a rather simpler test requiring less specialized equipment, and could potentially be offered as an alternative. REFERENCES 1. Locke GR 3rd, Talley NJ, Fett SL, Zinsmeister AR, Melton LJ 3rd: Prevalence and clinical spectrum of gastroesophageal reflux: a population based study in Olmsted County, Minnesota. Gastroenterology 112:1448–1456, 1997 2. Radu T, Castell DO: Management of gastroesophageal reflux disease. Am J Med Sci 326(5):309–318, 2003 3. Klauser AG, Schindlebeck NE, Muellar-Lissner SA: Symptoms of gastro-oesophageal reflux disease. Lancet 335:205–208, 1990 4. Chan CC, Lee CL, Wu CH: Twenty-four-hour ambulatory esophageal pH monitoring in patients with symptoms of gastroesophageal reflux. J Formos Med Assoc 96:874–878, 1997 5. Shi G, Bruley des Varannes S, Scarpignato C, Le Rhun M, Galmiche JP: Reflux related symptoms in patients with normal oesophageal exposure to acid. Gut 37(4):457–464, 1995 Digestive Diseases and Sciences, Vol. 50, No. 9 (September 2005)

6. Bernstein LM, Baker LA: A clinical test for esophagitis. Gastroenterology 34:760–781, 1958 7. DeMeester TR, Johnson LF, Joseph GJ, Toscano MS, Hall AW, Skinner DB: Patterns of gastroesophageal reflux in health and disease. Ann Surg 184(4):459–470, 1976 8. Richter JE, Bradley LA, DeMeester TR, Wu WC: Normal 24hr ambulatory oesophageal pH values: influence of study centre, pH electrode, age and gender. Dig Dis Sci 37:849–856, 1992 9. Johnson LF, DeMeester TR: Development of the 24-hour intraesophageal pH monitoring composite scoring system. J Clin Gastroenterol 8 (Suppl 1):52–58, 1986 10. Wiener GJ, Richter JE, Copper JB, Wu WC, Castell DO: The symptom index: a clinically important parameter of ambulatory 24-hour esophageal pH monitoring. Am J Gastroenterol 83:358–361, 1988 11. Breurnelhof R, Smout AJPM: The symptom sensitivity index: a valuable additional parameter in 24-hour esophageal pH recording. Am J Gastroenterol 86:160–164, 1991 12. Weusten BL, Roelofs JM, Akkermans LM, Van Berge-Henegouwen GP, Smout AJ: The symptom-association probability: an improved method for symptom analysis of 24-hour esophageal pH data. Gastroenterology 107:1742–1745, 1994 13. Anderson SHC, Anggiansah A, Arastu MH, Anggiansah R, Owen WJ: Heartburn in patients with untreated achalasia. Gastroenterology 122(4):A213, 2002 14. Singh S, Richter JE, Bradley LA, Haile JM: The symptom index: differential usefulness in suspected acid-related complaints of heartburn and chest pain. Dig Dis Sci 38:1402–1408, 1993 15. Shi G, Tatum RP, Joehl RJ, Kahrilas PJ: Esophageal sensitivity and symptom perception in gastroesophageal reflux disease. Curr Gastroenterol Rep 1:214–219, 1999 16. Richter JE: Disorders of esophageal function. In Gastrointestinal Motility Disorders for the Clinician: practical Guidelines for Patient Care. McCallum RW, Phillips SF, Reynolds JC (eds). New York, Academy Professional Information Services, 1998, pp 5.1–28 17. Jung B, Steinbach J, Beaumont C, Mittal RK: Lack of association between esophageal acid sensitivity detected by prolonged pH monitoring and Bernstein testing. Am J Gastroenterol 99(3):416–418, 2004 18. Howard PJ, Maher L, Pryde A, Heading RC: Symptomatic gastrooesophageal reflux, abnormal oesophageal acid exposure, and mucosal acid sensitivity are three separate, though related, aspects of gastro-oesophageal reflux disease. Gut 32:128–132, 1991 19. Trimble KC, Douglas S, Pryde A, Heading RC: Clinical characteristics and natural history of symptomatic but not excess gastroesophageal reflux. Dig Dis Sci. 40(5):1098–1104, 1995 20. Rodriguez-Stanley S, Robinson M, Earnest DL, Greenwood-Van Meerveld B, Miner PB: Esophageal hypersensitivity may be a major cause of heartburn. Am J Gastroenterol 94(3):628–631, 1999 21. Trimble KC, Pryde A, Heading RC: Lowered oesophageal sensory thresholds in patients with symptomatic but not excess gastrooesophageal reflux: evidence for a spectrum of visceral sensitivity in GORD. Gut 37:7–12, 1994 22. Richter JE: Diagnostic tests for gastroesophageal reflux disease. Am J Med Sci 326(5):300–308, 2003 23. Wiener GJ, Morgan TM, Copper JB, Wu WC, Castell DO, Sinclair JW, Richter JE: Ambulatory 24-hour esophageal pH monitoring. Reproducibility and variability of pH parameters. Dig Dis Sci 33(9):1127–1133, 1988 24. Dent J, Brun J, Fendrick AM, Fennerty MB, Janssens J, Kahrilas PJ, Lauritsen K, Reynolds JC, Shaw M, Talley NJ: An evidencebased appraisal of reflux disease management—the Genval Workshop Report. Gut 44 (Suppl 2):S1–S16, 1998

1615