ELEVATED GAMMA-GLUTAMYLTRANSFERASE CONCENTRATIONS IN PATIENTS WITH ACUTE KAWASAKI DISEASE

Abstract

Acute Kawasaki disease (KD) is associated with gallbladder hydrops, intrahepatic bile duct damage and hepatobiliary dysfunction.1-4 However, jaundice or elevated bilirubin is uncommon in KD patients.4 The enzyme gamma-glutamyl-transferase (GGT) is localized to cells lining the biliary tract and is a sensitive marker for epithelial injury in the biliary system. To better define the frequency of biliary system involvement in acute KD patients and in pediatric patients with other rash and fever syndromes, we determined the prevalence of elevated GGT in these patient populations. Methods.Patients. A retrospective chart review was performed on all patients with KD admitted to Children's Hospital, San Diego, during a 5-year period from May, 1991, to May, 1996. All patients met the clinical criteria for acute KD. Patients were excluded if a chemistry panel was not performed or was performed after the tenth day of fever. A computer data base search identified pediatric patients ages 0.25 to 6 years discharged from the same hospital between 1992 and 1995 with 8 rash/fever syndromes: (1) bacterial meningitis; (2) staphylococcal infection; (3) streptococcal infection; (4) Epstein-Barr virus infection; (5) sepsis; (6) fever, unspecified; (7) rheumatologic disease including juvenile rheumatoid arthritis, polyarthropathies and Felty's syndrome; and (8) drug reactions. Data. Age, gender, day of illness at the time of laboratory evaluation, coronary artery outcome and laboratory data [GGT, alanine aminotransferase (ALT) and total bilirubin] before administration of intravenous gamma-globulin were recorded for KD patients and entered into a computerized data base. Coronary arteries with an internal diameter >3 mm on echocardiogram were assessed as abnormal. Serum GGT, ALT and total bilirubin concentrations were recorded for all patients with other designated rash/fever syndromes. For all groups, if multiple laboratory evaluations were available only the first data set after hospitalization was analyzed. Statistical analysis. Values for continuous variables were expressed as the mean ± SD. Because of the skewed distribution of laboratory values, central tendency was described using geometric mean titers and their calculated 95% confidence intervals. Serum GGT and ALT values for KD patients and other groups were compared by analysis of variance (both raw data and after log transformation) with pairwise comparisons evaluated by Tukey's test. For continuous variables (age and day of illness), correlations with laboratory values (both raw and log transformed) were investigated. The effect of gender and coronary artery status on GGT was evaluated using Student t tests for independent comparisons of log transformed data. For categorical variables chi square analysis was performed. All analyses were performed using SAS Version 6.10 (SAS, Carey, NC). Results. During the study period 91 patients were diagnosed with Kawasaki disease. Laboratory values were available for 79 of these patients and 64 were evaluated within the first 10 days of illness. These 64 patients form the basis of our analysis. The mean age of these 64 KD patients was 2.9 ± 2.0 years (range, 0.25 to 8.5 years) and 36 (56%) were male. The mean day of illness at laboratory evaluation (±SD) was 5.1 ± 1.8 days (range, 1 to 10 days). Post-intravenous immunoglobulin echocardiograms revealed abnormal coronary arteries in 7 of 56 (13%) KD patients. Of the 497 patients with other rash/fever illnesses, 227 (46%) patients had serum GGT, ALT and total bilirubin values available. Because of the small numbers in each group, patients with Epstein-Barr virus infection (n = 15), sepsis (n = 13), fever unspecified (n = 16), drug reactions (n = 6) and rheumatologic disease (n = 12) were analyzed together under the category of other rash/fever syndromes. Serum GGT and ALT concentrations were elevated in 67 and 53% of the 64 KD patients, respectively, and 42% had a serum GGT of >100 units/l. In contrast serum GGT values were elevated in only 25% (14 of 57), 47% (26 of 55), 48% (24 of 50) and 40% (25 of 62) of patients with bacterial meningitis, staphylococcal infection, streptococcal infection and other rash/fever syndromes, respectively. The proportion of patients with an elevated GGT value was greater in KD patients than in non-KD patients (P = 0.003). Serum GGT values in KD patients were significantly higher than in all groups of pediatric patients with rash/fever illnesses (P < 0.05). In contrast serum ALT values in KD patients were significantly higher than in only two other patient groups, bacterial meningitis and other rash/fever syndromes (P < 0.05) (Fig. 1). Serum total bilirubin values were normal for all patient groups. Therefore GGT was the most consistently elevated marker for hepatobiliary dysfunction among KD patients and pediatric patients with rash/fever syndromes. There were no significant correlations between age and day of illness with serum GGT values in KD patients. There were also no significant differences in GGT concentration based on gender or coronary artery status.Fig. 1: Serum GGT and ALT concentrations in patients with acute KD and other rash/fever syndromes. Bars, 95% confidence limits of the geometric mean. Upper limits of normal for serum GGT and ALT are 37 and 40 units/l, respectively. *P < 0.05 vs. Kawasaki disease.Published data on Japanese children with acute KD are presented for comparison in Table 1. In the Japanese studies approximately one-third to one-half of the KD patients had elevated serum GGT concentrations. Of the patients with hydrops of the gallbladder documented by ultrasound, 71% had elevated serum GGT values.TABLE 1: Comparison of published GGT values among patients with acute KD Discussion. Of the laboratory tests reviewed in this study, serum GGT concentrations were the most commonly elevated marker for hepatobiliary dysfunction in patients with acute KD and were significantly higher for KD patients than for other pediatric patients with rash/fever syndromes. A serum chemistry panel that includes GGT and ALT is ordered as standard practice on all patients with suspected KD who are admitted to our institution. Thus these laboratory values should be representative of this patient population. In contrast serum chemistry panels were ordered in only 46% of the non-KD patients and may not represent the true frequency of abnormal serum GGT and ALT values in these populations. The limitations of this retrospective study design and the use of ICD.9 codes rather than clinical criteria to define the patients with rash/fever syndromes preclude any analysis of the potential utility of elevated GGT as a useful laboratory marker for acute KD. Tissue sources for GGT include kidney, pancreas, intestine and the hepatobiliary system and 11 different isoenzymes of GGT have been identified by isoelectric focusing techniques.7, 8 To identify the tissue source for serum GGT in acute KD, serum samples from 2 patients with acute KD and elevated serum GGT values were analyzed by 2-dimensional gel electrophoresis and the intensity of the bands was quantitated by densitometry,7 (data not shown). Both patients showed a greater predominance of the hepatobiliary isoenzyme band, suggesting that the elevation in total serum GGT was associated with the hepatobiliary isoenzyme in these two patients. The role of isoenzymes from other tissues in the genesis of the elevated serum GGT in acute KD should be evaluated in a larger cohort of patients. Pancreatitis associated with acute KD, although rare, has been reported,9 and elevated GGT from a pancreatic source was not specifically excluded in our patients. There is evidence of both intrahepatic and extrahepatic biliary tree pathology associated with acute KD. Hydrops of the gallbladder has been reported in 14 and 30% of American and Japanese KD patients, respectively.1, 5 Limited information is available regarding intrahepatic bile duct damage, which may include bile duct epithelial cell injury and necrosis as well as an inflammatory infiltrate localized around and within the lumina of the bile ducts.2 The etiology of this inflammatory process in acute KD is unknown. Although laboratory values are not strictly part of the KD case definition, clinicians use laboratory tests to exclude other diagnoses and to confirm the presence of systemic inflammation associated with the vasculitis of acute KD.4 We conclude that serum GGT is commonly elevated in acute KD and can be added to the list of laboratory markers suggesting inflammation in varied tissues during the acute phase of this disease. Acknowledgments. This work supported in part by NIH Grant HLO7491 (awarded to ECT) and the Kawasaki Disease Research Program, UCSD School of Medicine, San Diego. We thank Dr. Paul L. Wolf for the GGT isoenzyme measurements. This work was presented in part at the AFCR Western Regional Meeting, Carmel, CA, in February, 1997. Emily C. Ting, B.A. Edmund V. Capparelli, Pharm.D. Glenn F. Billman, M.D. Joel E. Lavine, M.D., Ph.D. Tomoyo Matsubara, M.D. Jane C. Burns, M.D. Departments of Pediatrics and Pharmacology; University of California, San Diego; School of Medicine, La Jolla, CA (ECT, EVC, JEL, JCB) Department of Pathology; Children's Hospital; San Diego, CA (GFB) Department of Pediatrics; Juntendo University; Tokyo, Japan (TM)