Serum Pancreatic Lipase Immunoreactivity Concentrations in Dogs Treated with Potassium Bromide and/or Phenobarbital*

Transcription

1 J. M. Steiner, P. G. Xenoulis, J. A. Anderson, A. C. Barr, and D. A. Williams Serum Pancreatic Lipase Immunoreactivity Concentrations in Dogs Treated with Potassium Bromide and/or Phenobarbital* Jörg M. Steiner, MedVet, DrMedVet, PhD, DACVIM, DECVIM-CA a Panagiotis G. Xenoulis, DVM a Joseph A. Anderson, BS a A. Catherine Barr, PhD b David A. Williams, MA, VetMB, PhD, DACVIM, DECVIM-CA a, a Gastrointestinal Laboratory Department of Small Animal Clinical Sciences College of Veterinary Medicine and Biomedical Sciences Texas A&M University College Station, TX b Texas Veterinary Medical Diagnostic Laboratory PO Drawer 3040 College Station, TX CLI NI CA L RE L E VA N C E Potassium bromide, phenobarbital, or a combination of both is commonly used in the treatment of canine epilepsy. Several cases of clinical pancreatitis have been reported in dogs after treatment with potassium bromide, but the risk of elevated serum canine pancreatic lipase immunoreactivity concentrations in dogs treated with potassium bromide and/or phenobarbital has not previously been evaluated in a large group of dogs.this study suggests an increased risk for elevated serum canine pancreatic lipase immunoreactivity concentrations and possibly pancreatitis in dogs treated with potassium bromide or phenobarbital alone or in combination. INTRODUCTION Epilepsy is a common problem in dogs, with a reported incidence of 0.6% to 2.3%. 1 For many years, phenobarbital was considered the treatment of choice for dogs with epilepsy. 1 However, side effects such as hepatotoxicity and, more important, refractory epilepsy led to a search for new antiepileptic drugs and the rediscovery of potassium bromide (KBr) for the treatment of epilepsy in dogs. KBr was first used in the treatment of epilepsy in humans by Locock in With the advent of newer antiepileptic drugs, KBr was used less frequently. Two decades ago, howev- *A portion of this study was supported by Idexx Laboratories, Westbrook, Maine. Presented in part at the 23 rd Annual Forum of the American College of Veterinary Internal Medicine, Baltimore, June Dr. Williams current address is Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL

2 Veterinary Therapeutics Vol. 9, No. 1, Spring 2008 er, KBr reemerged as an antiepileptic drug for human patients and soon thereafter was being used in dogs. 3 5 Successful use of KBr in 22 dogs with therapy-resistant epilepsy was reported in Since then, the pharmacokinetics of KBr have been studied in dogs and several reports of its clinical use in the management of canine epilepsy have been published. 1,5 7 For the most part, the recommended use of KBr was limited to combination therapy with other antiepileptic drugs, most commonly phenobarbital, but KBr has also been used for monotherapy. 8 In general, KBr is considered to be safe and effective. However, exocrine pancreatic disease and pancreatitis have been reported in dogs treated with this drug. In the study of 22 dogs with epilepsy treated with KBr mentioned above, 2 patients developed pancreatitis, 1 of creatitis in dogs with epilepsy may be related to their treatment with antiepileptic drugs. 9 Finally, in a recent study, 2 of 51 dogs (3.9%) treated with KBr and phenobarbital had histologically confirmed pancreatitis and others had a presumptive diagnosis of pancreatitis based on clinical signs and elevations of serum amylase and lipase activities. 8 Canine pancreatitis can have numerous etiologies or be associated with many risk factors, but it is often idiopathic. 10 A large variety of pharmaceutical compounds have been implicated in causing pancreatitis in humans and dogs; more than 50 drugs and drug classes have been implicated in humans, although definitive proof for a causal relationship is often lacking. 11 Suspected drugs commonly used in veterinary medicine include L-asparaginase, azathioprine, estrogen, furosemide, salicylates, Exocrine pancreatic disease and pancreatitis have been reported in dogs treated with potassium bromide. which subsequently developed exocrine pancreatic insufficiency; furthermore, of 7 patients from this study that underwent necropsy, 2 dogs were found to have localized areas of pancreatic fibrosis. 5 In another study, 2 of 23 dogs treated with KBr and phenobarbital were reported to have pancreatitis. 1 The authors speculated that the dogs showed polyphagia secondary to KBr therapy, which resulted in pancreatitis due to dietary indiscretion. 1 In a recent study of risk factors for fatal pancreatitis in dogs, 6 of 70 dogs (8.6%) with fatal pancreatitis but only 2 of 104 control dogs (1.9%) were reported to have epilepsy. 9 Although the increased risk for pancreatitis was not statistically significant in dogs with epilepsy (95% CI for odds ratio, 0.9 to 24.2), the authors speculated that the trend toward an increased risk for pan- sulfonamides, tetracyclines, thiazide diuretics, and vinca alkaloids. 10,11 Corticosteroids are no longer believed to cause pancreatitis in humans, 11 and there is little evidence that glucocorticoid administration causes pancreatitis in dogs. A new diagnostic test, canine pancreatic lipase immunoreactivity (cpli) concentration, has recently been developed and analytically validated. 12 Serum cpli concentration has been shown to be specific for being of exocrine pancreatic origin, and preliminary studies have suggested that it is also highly sensitive for the diagnosis of pancreatitis The goal of this study was to measure serum cpli concentrations in a large group of dogs treated with either KBr alone, phenobarbital alone, or a combination of KBr and phenobarbital and to 38

3 J. M. Steiner, P. G. Xenoulis, J. A. Anderson, A. C. Barr, and D. A. Williams compare concentrations with those measured in healthy control dogs. MATERIALS AND METHODS Remnants of 337 serum samples submitted to the Texas Veterinary Medical Diagnostic Laboratory from October to December 2004 and from November 2005 to January 2006 for measurement of serum KBr and/or phenobarbital concentrations were collected. Serum phenobarbital and/or KBr concentrations are often tested repeatedly in dogs treated with antiepileptic medications, but only one sample per dog was entered into the study; repeat samples from the same patient were excluded from the study. All samples were stored at 80 C until analysis. The samples had been submitted by veterinary clinics throughout Texas and, thus, were thought to reflect the general population of dogs treated with antiepileptic drugs. KBr and phenobarbital concentrations were measured in serum using spectrophotometric methods (Shimadzu UV-160A, Shimadzu Corporation, Nakagyo-ku, Kyoto, Japan, and Hitachi 911 Clinical Chemistry Analyzer, Roche Diagnostics Corporation, Indianapolis, IN, respectively). Of the 337 samples, 219 were submitted for measurement of KBr concentration. Because KBr is most commonly used in combination with phenobarbital, these 219 serum samples were subsequently analyzed for phenobarbital concentration. For the remaining 118 samples, only measurement of serum phenobarbital concentration had been requested by the submitting veterinarian. The 337 samples were thus divided into three groups: Group 1 (n = 98): Samples from dogs classified as having been treated only with KBr based on undetectable serum phenobarbital concentrations (<5 ppm) Group 2 (n = 118): Samples from dogs classified as having been treated only with phenobarbital based on the submitting veterinarian s request to assess only serum phenobarbital concentration Group 3 (n = 121): Samples from dogs classified as having been treated with both KBr and phenobarbital based on measurable serum concentrations of both analytes An undetectable serum concentration of KBr could not be ascertained in all dogs in Group 2 because of the additional cost of assaying serum KBr concentrations in all serum samples used for this study. However, serum KBr concentrations were measured in all samples from Group 2 in which an increased serum cpli concentration was detected, and all such samples analyzed had an undetectable serum KBr concentration. In addition, serum samples from 74 healthy dogs previously used to calculate the reference range for the serum cpli assay 12 were used as control samples. The control dogs comprised 45 dogs belonging to several research colonies and 29 pet dogs; they were of different breeds and sexes, and all were adults (>1 year of age). None of the dogs had any clinical signs reported by the owner or animal caretaker, and all had an unremarkable physical examination. Serum cpli concentrations were measured using an in-house ELISA within 3 months after initial sample submission. 12 The total number of dogs with a serum cpli concentration greater than either the upper limit of the reference range (102.1 µg/l) 12 or the suggested diagnostic cutoff value for pancreatitis (199.9 µg/l) 13,17,18 was tabulated, and the number of dogs with concentrations above these limits was recorded for all three treatment groups. Statistical Analysis Serum cpli and KBr concentrations were analyzed for possible correlation in Groups 1 39

4 Veterinary Therapeutics Vol. 9, No. 1, Spring 2008 and 3 individually and in all dogs treated with KBr. Similarly, serum cpli and phenobarbital concentrations were analyzed for possible correlation in Groups 2 and 3 individually and in all dogs treated with phenobarbital. Serum cpli concentrations in the 74 healthy dogs and all three treatment groups were tested for normal distribution using the Kolmogorov Smirnov test. Median serum cpli concentrations were compared between the 74 healthy dogs previously described 12 and the three treatment groups using a Kruskal Wallis test. Median serum cpli concentrations of each of the three treatment groups were individually compared with the median serum cpli concentration of the healthy control group using a Mann Whitney test. Additionally, the proportion of dogs in each group with serum cpli concentrations above the upper limit of the reference range or above the diagnostic cutoff value for pancreatitis was compared with the proportion of control dogs with serum cpli concentrations meeting the same criterion by use of Fisher s exact test. Odds ratios, with their 95% CIs, for having serum cpli concentrations either >102.1 µg/l or >199.9 µg/l were calculated for all dogs treated with an antiepileptic agent and for dogs belonging to Group 1, 2, or 3 compared with the controls. All statistical analyses were performed using a statistical software package (Prism4, GraphPad, San Diego, CA), and a P value of <.05 was considered statistically significant for all analyses. RESULTS Serum KBr, phenobarbital, and cpli concentrations were measured in 219 samples originally submitted for assessment of serum KBr concentration. In 98 of these samples, serum phenobarbital concentrations were undetectable (<5.0 ppm; therapeutic range, 15 to 40 ppm), and the dogs from which these samples were collected were considered to have been treated with KBr alone (Group 1). In the remaining 121 of these 219 samples, serum phenobarbital concentrations were >5.0 ppm, and the dogs from which these samples were collected were considered to have been treated with a combination of KBr and phenobarbital (Group 3). Serum phenobarbital and cpli concentrations were measured in the 118 samples for which only phenobarbital assessment had been requested by the submitting veterinarian (Group 2); in those samples in which serum cpli concentration exceeded the upper limit of the reference range (n = 17), serum KBr concentrations were measured and found to be undetectable. Serum cpli concentration was above the upper limit of the reference range (102.1 µg/l) in 46 (13.6%) of 337 dogs (all dogs) and in 15 (15.3%) dogs in Group 1, 17 (14.4%) dogs in Group 2, and 14 (11.6%) dogs in Group 3 (Figure 1). Serum cpli concentration was above the suggested diagnostic cutoff value for pancreatitis (199.9 µg/l) in 23 (6.8%) of 337 dogs and in 7 (7.1%) dogs in Group 1, 9 (7.6%) dogs in Group 2, and 7 (5.8%) dogs in Group 3 (Figure 1). All four data sets (healthy dogs [controls], Group 1, Group 2, and Group 3) failed normality testing, and nonparametric methods were used for further analysis. There was no statistical difference in the median serum cpli concentrations among the three treatment groups and the healthy control dogs using a Kruskal Wallis test (P =.068). When using a Mann Whitney test to compare the medians for two groups each, median serum cpli concentration was not significantly different between 74 healthy control dogs (16.3 µg/l) and all 337 dogs evaluated (24.7 µg/l; P =.06) or dogs treated with phenobarbital alone (17.7 µg/l; P =.61). However, median serum cpli concentration was significantly higher in dogs treated with KBr alone (

5 J. M. Steiner, P. G. Xenoulis, J. A. Anderson, A. C. Barr, and D. A. Williams Serum cpli (µg/l) Healthy Dogs KBr PB KBr + PB Serum cpli (µg/l) Serum KBr (mg/ml) Figure 1. Serum cpli concentrations in healthy dogs and dogs treated with antiepileptic agents. This figure displays the serum cpli concentrations in 74 healthy dogs, 98 dogs treated with KBr alone (KBr), 118 dogs treated with phenobarbital alone (PB), and 121 dogs treated with a combination of KBr and phenobarbital (KBr + PB). The lines for each group depict that group s median. The dashed line shows the upper limit of the reference range for serum cpli concentration (102.1 µg/l), and the solid line shows the suggested cutoff value for serum cpli concentration for pancreatitis (199.9 µg/l). µg/l; P =.03) and those treated with a combination of phenobarbital and KBr (26.2 µg/l; P =.02) than in healthy controls. There was no correlation between serum concentrations of cpli and KBr in dogs treated with KBr alone (Group 1; Spearman r = 0.035; P =.73) or all dogs treated with KBr (Groups 1 and 3; Spearman r = ; P =.07). However, there was a weak positive correlation of serum cpli and KBr concentrations in dogs treated with KBr and phenobarbital (Group 3; Spearman r = ; P =.02; Figure 2). There was no correlation between serum concentrations of cpli and phenobarbital for dogs treated with phenobarbital alone (Group 2; Spearman r = 0.036; P =.70), for dogs treated with KBr and phenobarbital (Group 3; Spearman r = 0.045; P =.63), or for all dogs treated with phenobarbital (Groups 2 and 3; Spearman r = 0.059; P =.37). Odds ratios showed an increased risk for an Figure 2. Correlation of serum KBr and cpli concentrations in dogs treated with KBr and phenobarbital. This figure shows the correlation of serum KBr and cpli concentrations in 121 dogs treated with both KBr and phenobarbital (Group 3). There was a weak, but significant, correlation between the two parameters (Spearman r = ; P =.02). elevated serum cpli concentration above the upper limit of the reference range of µg/l, which was statistically significant for all 337 dogs treated and for Groups 1 and 2 but not for Group 3 (Table 1). Odds ratios also showed an increased risk for an elevated serum cpli concentration >199.9 µg/l in all three treatment groups compared with the healthy controls (Table 1). However, this increased risk did not meet the criterion for statistical significance. DISCUSSION Of the 337 dogs treated with KBr and/or phenobarbital, 23 (6.8%) showed a serum cpli concentration above the suggested cutoff value of µg/l for a diagnosis of pancreatitis. Unfortunately, because the samples were run blindly, further investigation as to whether the animals exhibited clinical signs of pancreatitis was not possible. We thus were unable to determine whether any of these dogs had clinical pancreatitis. It has previously been speculated that KBr in combination with phenobarbital is a risk factor for pancreatitis in dogs. 8 An increased risk for 41

6 Veterinary Therapeutics Vol. 9, No. 1, Spring 2008 TABLE 1. Odds Ratios for Elevated Serum cpli Concentrations in Dogs Treated with Antiepileptic Drugs* Treatment Group Odds Ratio 95% CI Statistically Significant? Serum cpli Concentration >102.1 µg/l All dogs treated Yes Group 1: KBr only Yes Group 2: Phenobarbital only Yes Group 3: KBr and phenobarbital No Serum cpli Concentration >199.9 µg/l All dogs treated No Group 1: KBr only No Group 2: Phenobarbital only No Group 3: KBr and phenobarbital No *This table shows the odds ratios and 95% CIs for elevated serum cpli concentrations above the upper limit of the reference range (102.1 µg/l) and above the suggested cutoff value for pancreatitis (199.9 µg/l). pancreatitis in patients treated with phenobarbital alone has not been described previously. Therefore, it was interesting to note that the rates of elevated serum cpli concentrations above the upper limit of the reference range and above the suggested cutoff value for pancreatitis were similar for the three treatment groups. Odds ratios for an elevated serum cpli concentration were between 4.28 and 5.66 for the three treatment groups and the two levels of serum cpli concentration, although statistical significance was reached only for serum cpli concentrations above the upper limit of the reference range for all dogs treated and for dogs treated with KBr or phenobarbital alone. These findings would suggest an increased risk for elevated serum cpli concentrations in patients treated with KBr and/or phenobarbital. The median serum cpli concentration was higher in dogs treated with KBr, phenobarbital, or a combination of KBr and phenobarbital than in healthy dogs, but this increase reached statistical significance only for the dogs treated with KBr alone or in combination with phenobarbital. This can most likely be explained by an insufficient number of dogs to show a significant difference. However, it is also possible that there is no direct effect of phenobarbital therapy on serum cpli concentration and that the elevated concentrations were caused by chance. It should be noted that phenobarbital is known to induce the expression of some enzymes (e.g., UDP-glucuronosyltransferase or cytochrome P-450), and it is therefore possible that an elevation of serum cpli concentration in dogs treated with phenobarbital and/or KBr could be related to such an induction of gene expression. 19,20 However, we do not consider this possibility very likely because clinically relevant pancreatitis has previously been described in dogs that had been treated with KBr or a combination of KBr and phenobarbital. 1,5,8 Further studies are necessary to exclude the in- 42

7 J. M. Steiner, P. G. Xenoulis, J. A. Anderson, A. C. Barr, and D. A. Williams serum cpli concentrations in dogs that are being treated with KBr and/or phenobarbital to correlate clinical signs with elevations in serum cpli concentrations. CONCLUSION In the present study, there was an increased risk for an elevated serum cpli concentration in patients treated with either KBr alone, phenobarbital alone, or a combination of KBr and phenobarbital. The reason for this cannot be determined based on this study, but it might be related to an effect (direct or indirect) of A large variety of pharmaceutical compounds have been implicated in causing pancreatitis in humans and dogs. duction of gene expression of pancreatic lipase rather than pancreatic inflammation as a cause of elevations of serum cpli concentration shown here. Stability studies in our laboratory have shown that cpli concentrations in serum samples stored at room temperature, refrigerated, or frozen at 20 C or 80 C are stable for at least 21 days. 21 Based on the finding that serum cpli concentrations were stable even at room temperature for at least 21 days, we believe that it is safe to assume that storage at 80 C should not have affected the serum cpli concentrations for the period ( 3 months) the samples were stored before measurement of cpli concentrations. Also, reevaluation of several sets of serum samples more than 1 year after initial collection did not show any significant changes in serum cpli concentrations. 22 One limitation of this study was the fact that samples could not be traced to the patient from which they originated. Because of this, the true disease status of the patients (i.e., whether they were appropriately treated for epilepsy) or any concurrent conditions or medications administered could not be excluded as factors contributing to the serum cpli concentrations observed. It is important to note, however, that glucocorticoids, which are commonly administered to canine patients with epilepsy, have been shown not to affect serum cpli concentrations in dogs. 23 Further studies are needed to assess serum cpli concentrations in dogs with epilepsy that have not yet been treated with either KBr or phenobarbital. Also, further longitudinal studies are needed and are in progress to evaluate KBr and/or phenobarbital on the pancreas or associated with other factors (such as epilepsy itself). These findings would suggest that dogs treated with KBr and/or phenobarbital should be evaluated for pancreatitis, especially when compatible clinical signs are present. REFERENCES 1. Podell M, Fenner WR. Bromide therapy in refractory canine idiopathic epilepsy. J Vet Intern Med 1993;7: Locock C. Discussion of paper by Sieveking: analysis of 52 cases of epilepsy observed by the author. Med Times Gazette 1857;1: Boenigk HE, Lorenz JH, Jürgens U. [Bromides today still useful as antiepileptic substances?]. Nervenarzt 1985;56: Ernst JP, Doose H, Baier WK. Bromides were effective in intractable epilepsy with generalized tonic-clonic seizures and onset in early childhood. Brain Dev 1988;10: Schwartz-Porsche D, Jürgens U. Wirksamkeit von Bromid bei den therapieresistenten Epilepsien des Hundes. Tierärztl Prax 1991;19: Dowling PM. Management of canine epilepsy with phenobarbital and potassium bromide. Can Vet J 1994;35: March PA, Podell M, Sams RA. Pharmacokinetics and 43

8 Veterinary Therapeutics Vol. 9, No. 1, Spring 2008 toxicity of bromide following high-dose oral potassium bromide administration in healthy beagles. J Vet Pharmacol Ther 2002;25: Gaskill CL, Cribb AE. Pancreatitis associated with potassium bromide/phenobarbital combination therapy in epileptic dogs. Can Vet J 2000;41: Hess RS, Kass PH, Shofer FS, et al. Evaluation of risk factors for fatal acute pancreatitis in dogs. JAVMA 1999;214: Williams DA, Steiner JM. Canine pancreatic disease. In: Ettinger SJ, Feldman EC, eds. Textbook of Veterinary Internal Medicine. 6 th ed. St. Louis: Elsevier Saunders; 2005: Frick TW, Speiser DE, Bimmler D, et al. Drug-induced acute pancreatitis: further criticism. Dig Dis 1993;11: Steiner JM, Teague SR, Williams DA. Development and analytic validation of an enzyme-linked immunosorbent assay for the measurement of canine pancreatic lipase immunoreactivity in serum. Can J Vet Res 2003;67: Steiner JM, Broussard J, Mansfield CS, et al. Serum canine pancreatic lipase immunoreactivity (cpli) concentrations in dogs with spontaneous pancreatitis [abstract]. J Vet Intern Med 2001;15: Steiner JM, Newman SJ, Xenoulis PG, et al. Comparison of sensitivity of serum markers in dogs with macroscopic evidence of pancreatitis [abstract]. J Vet Intern Med 2007;21: Steiner JM, Berridge BR, Wojcieszyn J, et al. Cellular immunolocalization of gastric and pancreatic lipase in various tissues obtained from dogs. Am J Vet Res 2002;63: Steiner JM, Rutz GM, Williams DA. Serum lipase activities and pancreatic lipase immunoreactivity concentrations in dogs with exocrine pancreatic insufficiency. Am J Vet Res 2006;67: Porterpan B, Zoran D, Steiner JM. Serial serum pancreatic lipase immunoreactivity concentrations in a dog with histologically confirmed pancreatitis. Vet Med 2006;101: Steiner JM. Is it pancreatitis? Vet Med 2006;101: Oguri K, Kurogi A, Yamabe K, et al. Purification of a phenobarbital-inducible UDP-glucuronosyltransferase isoform from dog liver which catalyzes morphine and testosterone glucuronidation. Arch Biochem Biophys 1996;325: Ciaccio PJ, Halpert JR. Characterization of a phenobarbital-inducible dog liver cytochrome P450 structurally related to rat and human enzymes of the P450IIIA (steroid-inducible) gene subfamily. Arch Biochem Biophys 1989;271: Gumminger SR, Steiner JM, Ruaux CG, et al. Stability of serum canine pancreatic lipase (cpli) concentration and comparison of cpli concentrations in serum and plasma [abstract]. J Vet Intern Med 2002; 16: Steiner JM, et al: Unpublished data, College Station, TX, Steiner JM, Lees GE, Willard MD, et al. Serum canine pancreatic lipase immunoreactivity (cpli) concentration is not altered by oral prednisone administration [abstract]. J Vet Intern Med 2003;17: