Duration of Pain Is Correlated With Elevation in Liver Function Tests in Patients With Symptomatic Choledocholithiasis

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2010;8: Duration of Pain Is Correlated With Elevation in Liver Function Tests in Patients With Symptomatic Choledocholithiasis ALA I. SHARARA, NABIL M. MANSOUR, MUSTAFA EL HAKAM, OLA GHAITH, and MUSTAPHA EL HALABI Division of Gastroenterology, Department of Internal Medicine, American University of Beirut Medical Center, Beirut, Lebanon BACKGROUND & AIMS: We assessed the temporal relationship between abdominal pain and elevation in liver function tests (LFTs) in patients with acute symptomatic choledocholithiasis. METHODS: Retrospective study of patients that presented within 12 hours of pain onset and were subsequently found to have choledocholithiasis. RESULTS: We identified 40 patients with complete medical records. Levels of aspartate and alanine aminotransferases (AST and ALT) correlated with duration of pain (Pearson correlation, r and respectively, P.001 for both); the correlation was not as strong for -glutamyl transpeptidase (GGT) (r 0.326, P.046) and was not significant for alkaline phosphatase or bilirubin. This temporal association was stronger in patients that had undergone cholecystectomy versus those with intact gallbladders (for ALT, r vs r 0.311, respectively). Eighteen patients, evaluated within 6 hours of pain, had normal or minimal alterations in LFTs; transabdominal ultrasound was abnormal in 6 (sensitivity 33.3%). All had repeat LFTs within 24 hours (mean hours later) and large increases in ALT and aspartate aminotransferase levels (mean and 6.8-fold respectively; P.01 for both), intermediate increases in glutamyl transpeptidase levels, (mean 4-fold, P.05), and no changes in alkaline phosphatase levels. This significant increase in LFTs was the only indication of biliary pathology before endoscopy in 11/18 patients. CONCLUSIONS: Increasing duration of pain is associated with increasing LFTs (particularly transaminases) in patients with acute symptomatic choledocholithiasis. Patients with normal LFTs and ultrasound upon presentation should have repeat LFTs if biliary pain is suspected. The absence of significant biochemical abnormalities within the first 24 hours makes the diagnosis of symptomatic choledocholithiasis unlikely. Keywords: Common Bile Duct; Sphincter of Oddi; Biliary; Stones. Biliary pain or colic is characterized by acute onset of pain without warning, and is presumably the result of acute, and usually, transient obstruction of the biliary tree by stones or sludge. Anatomically, acute biliary pain can be divided into gallbladder-type pain when the obstruction is at the level of the cystic duct and choledochal-type pain when the obstruction is at the level of the common bile duct (CBD). These types are, however, clinically indistinguishable as far as quality of pain, and their differentiation therefore relies on imaging (commonly transabdominal ultrasound) and biochemical liver function tests (LFTs) to confirm or refute the diagnosis of acute symptomatic choledocholithiasis. These diagnostic tests, however, have limited accuracy. The sensitivity of transabdominal ultrasound for choledocholithiasis is around 50% 1-3 while the true predictive value of LFTs, other than bilirubin, is unknown although it is likely limited given the multiplicity of conditions associated with abnormal LFTs (eg, steatosis, drugs, alcohol, etc). In a retrospective study, Onken et al identified CBD diameter, maximum preoperative bilirubin, aspartate aminotransferase (AST), and alkaline phosphatase (AP) as independent predictors of choledocholithiasis in patients undergoing cholecystectomy, and a model containing these variables was found to be moderately accurate. 4 However, the proposed model likely underestimates the true prevalence of choledocholithiasis in patients with symptomatic cholelithiasis because of an inherent selection bias given that only those suspected a priori of having choledocholithiasis undergo definitive confirmatory testing (eg, endoscopic ultrasound, intraoperative or endoscopic retrograde cholangiography). Acute biliary pain due to choledocholithiasis is triggered by a stone traversing or obstructing the biliary component of the sphincter of Oddi (sphincter choledochus). This is supported by the fact that patients with nonobstructive choledocholithiasis remain asymptomatic as long as the stone(s) is(are) freely floating in the CBD. This concept is held up further by the complete, and often abrupt, resolution of symptoms and gradual normalization of LFTs once the pain aborts, conceptually coinciding with dislodgment of the stone from the sphincter choledochus. 5,6 Importantly, this clinical and biochemical improvement occurs once the obstruction resolves regardless of whether the stone passes distally into the duodenum or floats back up freely into the CBD. 7 The same principle of transient sphincteric noncalculous obstruction presumably secondary to spasm explains the elevation in LFTs seen in patients with pain associated with sphincter (choledochus) of Oddi dysfunction or dyskinesia (SOD). Symptomatic choledocholithiasis is generally not included in the differential diagnosis of patients presenting with recent onset upper abdominal pain, normal or near normal LFTs, and a nondilated CBD on transabdominal ultrasound. In a study involving 100 consecutive patients without cholecystitis or liver disease, Anciaux and colleagues noted at least 1 biochemical abnormality in 99% of patients with symptomatic choledocho- Abbreviations used in this paper: ALT, alanine aminotransferase; AP, alkaline phosphate; AST, aspartate aminotransferase; CBD, common bile duct; ED, emergency department; GGT, gamma-glutamyl transpeptidase; LFTs, liver function tests; SOD, sphincter (choledochus) of Oddi dysfunction or dyskinesia; TUS, transabdominal ultrasound; ULN, upper limit of normal by the AGA Institute /$36.00 doi: /j.cgh

2 1078 SHARARA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 12 Table 1. Baseline Characteristics of the 40 Study Patients Mean age Gender 20 F:20 M Prior cholecystectomy 7/40 (17.5%) Mean time from pain onset to LFT (2 12) testing in hours (range) Mean baseline LFTs (range): ALT (IU/L) 250 (7 1021) AST (IU/L) 266 (22 982) GGT (IU/L) 235 ( ) Alkaline phosphatase (IU/L) 147 (13 555) Total bilirubin (mg/dl) 1.8 ( ) ALT, alanine aminotransferase; AST, aspartate aminotransferase; F, female; GGT, glutamyl transpeptidase; LFT, liver function test; M, male. lithiasis within the first 3 days of presentation. 5 The most frequent biochemical abnormalities were high serum AP and gamma-glutamyl transpeptidase (GGT) in more than 90% of patients, followed by high serum transaminases in 71% 88% of patients. However, the exact time sequence leading to elevation in LFTs in choledochal-type biliary pain and the diagnostic accuracy of LFTs within the first hours of presentation, when the need for accurate and prompt diagnosis is paramount, remains unknown. The aims of this study were to assess the temporal relationship between upper abdominal pain duration and elevation of LFTs in biliary pain secondary to choledocholithiasis, and to examine the value of repeated LFT testing in the early diagnosis of symptomatic CBD stones prior to invasive endoscopic procedures. Methods Using the endoscopy database of the American University of Beirut Medical Center, patients with proven choledocholithiasis from 2004 to 2008 were identified and their medical records reviewed retrospectively. Those who presented to the emergency department (ED) within 12 hours of onset of acute upper abdominal pain subsequently confirmed to be secondary to choledocholithiasis by endoscopic ultrasound (EUS) and/or endoscopic retrograde cholangiography (ERC) were included. Patients with sepsis and those with known liver disease or biliary pathology other than stone disease were excluded. Only charts with complete ED records regarding time of pain onset and biochemical liver testing were included. The study was approved by the Institutional Review Board. Information regarding clinical presentation, exact time of pain onset, time of initial phlebotomy for LFTs, and radiology findings was collected and entered into a database using SPSS version 16.0 for Windows (Microsoft Corp, Redmond, Washington). Measured LFTs consisted of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (AP), and gamma-glutamyl transpeptidase (GGT), and, when available, serum total and direct bilirubin. These were plotted against time, measured in hours from onset of pain to phlebotomy, to ascertain whether there was a significant rise in LFTs over time. Pearson correlation was used with a P value of.05 for significance. For patients who had repeat LFTs, the means of the first and second set were compared using Student t test with P.05 for significance. Results From the endoscopy database, 148 patients with choledocholithiasis were identified and upon chart review, 40 were found to have presented to the ED within 12 hours of sudden pain onset and to have complete records. Table 1 shows the baseline characteristics of the study patients. Upon presentation, serum transaminases were elevated to more than 2 times the upper limit of normal (ULN) in 22 of the 40 patients (55%) while AP and GGT Figure 1. Percent distribution of ALT values (in IU/L) on presentation and upon repeat testing within the following 24 hours. Of the 40 patients, only those with minimally disturbed LFTs (n 18) had repeat testing within 24 hours.

3 December 2010 LFTS AND SYMPTOMATIC CHOLEDOCHOLITHIASIS 1079 were elevated to 2 ULN in 8/40 (20%) and 26/40 (65%), respectively. Figure 1 shows the ALT values on presentation and following repeat testing within the first 24 hours. When plotted against time from onset of pain to testing in all 40 patients, there was a strong and significant trend for higher AST (Pearson correlation, r 0.633, P.001) and ALT (r 0.622, P.001) levels over time (Figure 2). The correlation was weaker for GGT (r 0.326, P.046) and was not significant for AP (r 0.259, P.111) (Figure 2) or bilirubin (r 0.165; P.374). Seven out of 40 patients (17.5%) had prior cholecystectomy. The mean LFTs on presentation of these 7 patients were not significantly different from the other 33 Figure 2. The temporal relationship between onset of upper abdominal pain and LFTs (n 40) for (A) alkaline phosphatase, (B) ALT, and (C) total bilirubin (n 32). patients with intact gallbladder. However, when plotted over time from pain onset including repeat values within the first 24-hour interval, there was a stronger correlation between time and elevation in transaminases (r for ALT and for AST) and to a lesser extent GGT (r 0.299) but not AP (r 0.119), in cholecystectomized patients compared with those with an intact gallbladder (Figure 3). Of the 40 study patients, 18 (45%) presented with normal or minimally disturbed LFTs ( 2 ULN in at least 3 of the 4 tests: AST, ALT, AP, and GGT) within 6 hours of pain and in those, the diagnosis of symptomatic choledocholithiasis was initially considered biochemically unsubstantiated. Of those 18 patients, 3 had prior cholecystectomy and 15 had an intact gallbladder. In the 3 cholecystectomized patients, transabdominal ultrasound (TUS) failed to confirm choledocholithiasis while in the 15 patients with intact gallbladder, TUS showed abnormality only in 6: CBD dilatation ( 5 mm) in all 6 and choledocholithiasis in 1 with an overall sensitivity of TUS in this group of 33.3%. All 18 had repeat LFTs within the first day after presentation with a mean time between the first set in the ED and the second set of hours (range 4 24). There was a pronounced rise in ALT and AST (mean and fold rise, respectively; P.01 for both), and GGT (mean fold; P.05), but an insignificant rise in AP ( fold) (Figure 4). A rise of 50% of the initial value on presentation was seen in 18/18 patients for AST and ALT (100%) (Figure 5) and 14/18 for GGT (83.3%) versus 6/18 (33.3%) for AP (P.01 for both against AP). When taking into account the results of the initial biochemical liver tests and those of the transabdominal ultrasound, this rise in LFTs was the only early clinical indication of biliary pathology prior to endoscopy (endoscopic ultrasound and/or endoscopic retrograde cholangiography) in 11 of the 18 patients (61.1%). Discussion To our knowledge, this is the first study addressing the temporal relationship between acute biliary pain secondary to choledocholithiasis and biochemical liver tests used routinely to support this diagnosis. Our findings show a direct relationship between duration of pain (from pain onset to LFT testing) and elevation of LFTs and suggest a role for repeat testing, in particular aminotransferases (and to a lesser extent GGT), in support of the clinical diagnosis in patients presenting early after pain onset with minimally disturbed LFTs. As discussed above, the diagnosis of symptomatic choledocholithiasis may be difficult early during the acute event because of the limited sensitivity of transabdominal ultrasound 1-3 and the often polymorphic nature of the clinical and biochemical features of choledocholithiasis. In the largest study to date involving 100 patients with symptomatic choledocholithiasis, Anciaux and colleagues 5 concluded that the absence of biochemical abnormalities in the first 3 days following the onset of symptoms makes the diagnosis of choledocholithiasis unlikely. Albeit useful, this conclusion is of limited utility in the management of patients presenting with acute pain where there is a pressing need for rapid diagnosis and management. The so-called cholestatic liver enzymes, namely AP and GGT, are elevated in the setting of biliary obstruction. Determination of cholestatic liver enzymes has therefore been considered superior to that of transaminases for the diagnosis of biliary obstruction. Whether this applies to transient sphincter of

4 1080 SHARARA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 12 Figure 3. Serum ALT values over time in patients with intact gallbladder vs those with prior cholecystectomy. All values obtained during the initial 24 hours were included. Oddi obstruction by a stone or possibly in the absence of one, like in SOD is largely unsupported. In the study by Anciaux et al, 5 serum alkaline phosphatase and GGT were elevated within the first 3 days in 100% of patients while transaminases were elevated in up to 88% of patients with a mean of IU/L for AST and ALT, respectively. These mean transaminase values are hardly specific for the diagnosis of choledocholithiasis given the wealth of conditions associated with mildly to moderately elevated transaminases (hepatic steatosis, drugs, alcohol, etc). A more marked elevation in transaminases has, however, been recognized as a presentation of choledocholithiasis but remains more widely confused as acute hepatitis in clinical practice. 6,8 In a retrospective study by Nathwani and colleagues, 6 transaminases levels 100 IU/L were noted in 81.2% of 467 patients with unspecified gallstone disease but levels greater than 250, 500, and 1000 IU/L were seen in 55.8%, 28%, and 9.6% of patients, respectively. However, no details about patient presentation, findings on endoscopy, time and duration of pain, or the time to peak transaminases were provided in that report. Our study specifically addresses all these points within the first 24 hours after onset of symptoms and confirms that transaminases are more sensitive than AP or GGT in this early phase of the disease coinciding anatomically with transient CBD obstruction. The exact mechanism leading to hepatocytolysis in symptomatic choledocholithiasis, evidenced by significant elevation in serum transaminases (72.5% in this study had ALT 250 IU/L), is unclear. The general absence of this phenomenon in patients with malignant biliary obstruction or following iatrogenic clipping of the CBD during laparoscopic cholecystectomy, as well Figure 4. Repeat LFTs in patients with normal or minimally disturbed values on presentation (n 18). The figure depicts the rise in LFTs represented as the average fold-increase for each specific test between the initial blood draw in the Emergency Department and the repeat blood draw within 24 hours (*P.05; **P.01).

5 December 2010 LFTS AND SYMPTOMATIC CHOLEDOCHOLITHIASIS 1081 Figure 5. Baseline (}) and repeat ( ) serum ALT levels in the 18 patients with minimally disturbed LFTs on presentation. as in bile duct-ligated animals, suggests that the mechanism is related to the acute rise in hydrostatic pressure in the biliary tree as a result of high amplitude retrograde phasic contractions of the obstructed sphincter choledochus. The normal hydrostatic pressure in the CBD is cm H 2 O ( mm Hg) and increases to cm H 2 O (18 30 mm Hg) with nonsphincteric biliary obstruction in laboratory animals. 9,10 Under normal conditions, the human sphincter of Oddi exhibits spontaneous phasic contractions that migrate distally 11,12 with a mean amplitude in healthy volunteers of mm Hg. 13 Elevation of sphincteric basal pressure and in the frequency and amplitude of phasic waves (greater than 350 mm Hg) occurs in patients with SOD 14 and plausibly in those with acute sphincteric obstruction by a calculus. In fact, it has been shown that patients with CBD stones have an increase in CBD pressure 15 and in the frequency of retrograde propagation of phasic waves. 16 Such retrograde-propagating high amplitude waves may lead to a rapid increase in biliary hydrostatic pressure which, in turn, results in distortion of ductal epithelial junctional complexes increasing their permeability 17 and to hepatocytolysis in zone 1 (periductular zone) of the liver acinus as a result of transcellular bile regurgitation 18 and may explain the significant and timerelated increase in serum transaminases. This pressure-related phenomenon is subject to Boyle s law regulating the pressure-volume relationship of a fluid confined in a container where pressure created by the application of a force with a solid piston is equal to the force applied divided by the area of the piston (Figure 6A). This model conceptually and physically applies to acute sphincteric obstruction where the stone-sphincter complex acts as the force-generating piston on the fluid (bile) confined within the container (biliary tree) (Figure 6B). This suggests that when all else is equal, particularly the duration of obstruction and time to LFT testing, the elevation in serum aminotransferases over time is expected to be higher in cholecystectomized patients compared with those with an intact gallbladder and lower in those with a dilated biliary system or CBD. The rationale is that the gallbladder helps dissipate the acute rise in pressure by providing a larger total biliary volume, 19 and CBD dilatation diffuses intrabiliary pressure resulting in a less pronounced rise in LFTs. The latter was in fact suggested in a Korean study showing an inverse correlation between CBD diameter and transaminase levels in patients with symptomatic CBD stones. 20 Our findings are perfectly in line with the above-described pathophysiologic mechanism adding credence to the suggested physical principle regulating this phenomenon by taking into account the most important variable, namely time, in the interpretation of LFTs in symptomatic choledocholithiasis. In conclusion, this study clarifies the time sequence for elevation of liver biochemical tests in acute biliary pain secondary to choledocholithiasis and provides further insight into the mechanisms underlying these biochemical changes. Understandably, this temporal relationship is most robust in the early hours and would be expected to decline and, in fact, reverse with resolution of the biliary sphincteric obstruction over time. Based on the above, we conclude that, in the absence of biliary dilatation, the lack of significant biochemical abnormalities, particularly in serum transaminases and GGT, on 2 occasions within the first 24 hours makes the diagnosis of choledocholithiasis highly unlikely and may circumvent the need for invasive biliary endoscopy in such patients. Given the low overall prevalence of choledocholithiasis in patients with gallstones, this clinical approach should be considered when the suspicion of biliary colic secondary to choledocholithiasis

6 1082 SHARARA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 12 Figure 6. (A) Boyle s law regulating pressure-volume relationship of a fluid in a closed cylinder. The pressure created in the cylinder equals the force applied divided by the area of the piston (P F/A) and (B) pressure dynamic model of a stone obstructing the biliary outflow at the sphincter choledochus, resulting in retrograde high amplitude phasic contractions. remains high despite a negative ultrasound and normal, or near normal, LFTs upon presentation. References 1. Dittrick G, Lamont JP, Kuhn JA, et al. Usefulness of endoscopic ultrasound in patients at high risk of choledocholithiasis. Proc (Bayl University Med Cent) 2005;18: O Connor HJ, Hamilton I, Ellis WR, et al. Ultrasound detection of choledocholithiasis: prospective comparison with ERCP in the postcholecystectomy patient. Gastrointest Radiol 1986;11: Varghese JC, Liddell RP, Farrell MA, et al. Diagnostic accuracy of magnetic resonance cholangiopancreatography and ultrasound compared with direct cholangiography in the detection of choledocholithiasis. Clin Radiol 2000;55: Onken JE, Brazer SR, Eisen GM, et al. Predicting the presence of choledocholithiasis in patients with symptomatic cholelithiasis. Am J Gastroenterol 1996;91: Anciaux ML, Pelletier G, Attali P, et al. Prospective study of clinical and biochemical features of symptomatic choledocholithiasis. Dig Dis Sci 1986;31: Nathwani RA, Kumar SR, Reynolds TB, et al. Marked elevation in serum transaminases: an atypical presentation of choledocholithiasis. Am J Gastroenterol 2005;100: Roston AD, Jacobson IM. Evaluation of the pattern of liver tests and yield of cholangiography in symptomatic choledocholithiasis: a prospective study. Gastrointest Endosc 1997;45: Abbruzzese A, Jeffery RL. Marked elevations of serum glutamic oxalacetic transaminase and lactic dehydrogenase activity in chronic extraheptic biliary disease. Am J Dig Dis 1969;14: Strasberg SM, Dorn BC, Small DM, et al. The effect of biliary tract pressure on bile flow, bile salt secretion, and bile salt synthesis in the primate. Surgery 1971;70: Strasberg SM, Redinger RN, Dorn BC, et al. Effects of alteration of biliary pressure on bile composition--a method for study: primate biliary physiology. V. Gastroenterology 1971;61: Geenen JE, Hogan WJ, Dodds WJ, et al. Intraluminal pressure recording from the human sphincter of Oddi. Gastroenterology 1980;78: Meshkinpour H, Mollot M, Eckerling GB, et al. Bile duct dyskinesia. Clinical and manometric study. Gastroenterology 1984;87: Funch-Jensen P, Kruse A, Ravnsbaek J. Endoscopic sphincter of Oddi manometry in healthy volunteers. Scand J Gastroenterol 1987;22: Rosenblatt ML, Catalano MF, Alcocer E, et al. Comparison of sphincter of Oddi manometry, fatty meal sonography, and hepatobiliary scintigraphy in the diagnosis of sphincter of Oddi dysfunction. Gastrointest Endosc 2001;54: Csendes A, Sepulveda A, Burdiles P, et al. Common bile duct pressure in patients with common bile duct stones with or without acute suppurative cholangitis. Arch Surg 1988;123: Toouli J, Geenen JE, Hogan WJ, et al. Sphincter of Oddi motor activity: a comparison between patients with common bile duct stones and controls. Gastroenterology 1982;82: Toyota N, Miyai K, Hardison WG. Effect of biliary pressure versus high bile acid flux on the permeability of hepatocellular tight junction. Lab Invest 1984;50: Watanabe N, Kojima S, Takashimizu S, et al. Initial site of bile regurgitation following extrahepatic biliary obstruction in living rats. J Gastroenterol Hepatol 2007;22: Torsoli A, Corazziari E, Habib FI, et al. Pressure relationships within the human bile tract. Normal and abnormal physiology. Scand J Gastroenterol Suppl 1990;175: Jeon WJ, Han JH, Seo JC, et al. Clinical features of patients with choledocholithiasis showing high levels of aminotransferases [in Korean]. Korean J Gastroenterol 2006;47: Reprint requests Address requests for reprints to: Ala I. Sharara, MD, FACG, AGAF, Division of Gastroenterology, American University of Beirut Medical Center, P.O. Box /16-B, Beirut, Lebanon. as08@aub.edu.lb; fax: Conflicts of interest The authors disclose no conflicts.