Histiocytic ulcerative colitis (HUC) is an inflammatory

Transcription

1 J Vet Intern Med 2004;18: Antibiotic-Responsive Histiocytic Ulcerative Colitis in 9 Dogs Roger A. Hostutler, Brian J. Luria, Susan E. Johnson, Steven E. Weisbrode, Robert G. Sherding, Jordan Q. Jaeger, and W. Grant Guilford Canine histiocytic ulcerative colitis (HUC) is characterized by colonic inflammation with predominantly periodic acid-schiff (PAS)- positive macrophages. The inflammation results in colonic thickening, ulcerations, and distortion of normal glandular architecture. Resultant clinical signs consist of chronic large bowel diarrhea, tenesmus, and marked weight loss, and the disease frequently results in euthanasia. Conventional therapy consists of some combination of prednisone, azathioprine, sulfasalazine, and metronidazole. Nine dogs (8 Boxers and 1 English Bulldog) with histologic confirmation of HUC were treated with antibiotic therapy (either with enrofloxacin alone or in combination with metronidazole and amoxicillin). Clinical signs, physical examination findings, laboratory abnormalities, and the histologic severity of the disease were evaluated. Four of the 9 dogs had been treated previously with conventional therapy and had failed to respond favorably; then, these dogs were placed on antibiotic therapy (enrofloxacin, n 1; enrofloxacin, metronidazole, and amoxicillin, n 3) and had resolution of clinical signs within 3 12 days. Five dogs were treated solely with antibiotic therapy (enrofloxacin, n 1; enrofloxacin and metronidazole, n 1; enrofloxacin, metronidazole, and amoxicillin, n 3), and clinical signs resolved in 2 7 days. Repeated biopsy specimens were obtained from 5 dogs after treatment, and all showed marked histologic improvement. The increase in body weight after treatment was statistically significant (P.01). Three dogs currently are not on any treatment and have had resolution of clinical signs for up to 14 months. These observations suggest that an infectious agent responsive to antibiotics plays an integral role in the clinical manifestation of canine HUC, and they support the use of antibiotics in its treatment. Key words: Boxer colitis; Canine; Colon; Diarrhea; Granulomatous enteritis. Histiocytic ulcerative colitis (HUC) is an inflammatory bowel disease that causes tenesmus, hematochezia, and profound weight loss. The disease most commonly is described in young Boxer Dogs, 1,2 but it also has been reported in other breeds of dogs, including 1 Mastiff, 3 1 Alaskan Malamute, 3 1 Doberman Pinscher, 3 and 2 French Bulldogs. 4,5 One cat with HUC also has been described. 6 HUC differs from other forms of inflammatory bowel disease in dogs because it is characterized histologically by periodic acid-schiff (PAS)-positive macrophages, is more likely to be associated with mucosal ulcerations, is less responsive to therapy, and has a poorer long-term prognosis. 2 HUC in Boxer Dogs was 1st described by Van Kruiningen et al 7 in Since that time, the gross, histopathologic, and ultrastructural findings have been well characterized. 8,9 The pathognomonic lesion of HUC is the accumulation of distinctive, PAS-positive macrophages (indicative of glycoprotein within the macrophages) in the lamina propria and submucosa of the colon with loss of the associated epithelial surface 7,10 (Figs 1, 2). The PAS-positive material may be derived from remnants of bacterial cell wall glycoprotein, and accumulation of PAS-positive material in macrophages may occur because of abnormal ly- From the Departments of Veterinary Clinical Sciences (Hostutler, Johnson, Sherding) and Veterinary Biosciences (Weisbrode), College of Veterinary Medicine, The Ohio State University, Columbus, OH; the Department of Small Animal Clinical Sciences, University of Florida, Gainesville, FL (Luria); Carolina Veterinary Specialists, Charlotte, NC (Jaeger); and Massey University, Palmerston North, New Zealand (Guilford). Reprint requests: Roger A. Hostutler, DVM, Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Ohio State University, Columbus, OH 43210; hostutler.4@osu.edu. Submitted November 17, 2003; Revised December 30, 2003, and January 13, 2004; Accepted February 4, Copyright 2004 by the American College of Veterinary Internal Medicine /04/ /$3.00/0 sosomal activity, exhaustion of lysosomal activity, or inhibition of lysosomal activity by toxic substances. 11 The cause of HUC has yet to be determined. Early studies proposed an infectious etiology on the basis of the presence of chlamydia-like organisms in macrophages on electron microscopy 12 and clinical improvement after chloramphenicol therapy. 7 In a subsequent ultrastructural study, organisms were not conclusively demonstrated. 11 Attempts to create the disease experimentally by mycoplasma infection failed. 13 Management of HUC consists of various combinations of the following: dietary modifications; antibiotics such as chloramphenicol, metronidazole, and tylosin; and anti-inflammatory or immunosuppressive drugs such as sulfasalazine, prednisone, and azathioprine. 1,2 Response to treatment generally is poor, frequently resulting in euthanasia of affected animals. 14 The purpose of this study was to evaluate the antibiotic responsiveness of histologically confirmed HUC in 9 dogs (8 Boxers and 1 English Bulldog) treated with enrofloxacin, either alone or in combination with metronidazole and amoxicillin. Clinical and histopathologic findings are described, including follow-up biopsy results in 5 of the dogs after long-term antibiotic therapy. Materials and Methods The medical records and colonic biopsy specimens of 9 dogs with HUC, confirmed by the presence of PAS-positive macrophages, were reviewed. The affected dogs were seen on referral by the authors at The Ohio State University Veterinary Teaching Hospital (OSU-VTH), Carolina Veterinary Specialists (CVS), or the University of Florida Veterinary Medical Teaching Hospital (UFVMTH). The data collected from the medical records included signalment, history, physical examination findings, laboratory findings, imaging results, gross colonoscopic abnormalities, histopathologic findings, and response to therapy. Colon biopsy specimens were reviewed by one of the authors (SW) in a randomized, blinded manner and assessed for the presence or absence of key histologic abnormalities. Biopsy specimens from 8 dogs were reviewed before treatment, and additional specimens from 5 dogs were evaluated after treatment.

2 500 Hostutler et al Fig 1. Hematoxylin-eosin stain of a colonic biopsy from a dog with histiocytic ulcerative colitis (HUC). There is marked infiltration of the lamina propria with numerous macrophages and lesser numbers of lymphocytes, plasma cells, and neutrophils (black arrow). The infiltrate expands the space between the base of the glands and the subjacent muscularis mucosa (white arrow) and distorts the glands. Note the hyperplasia of glandular epithelium and the loss of surface epithelium (magnification 16 ). Statistical Analysis Descriptive statistics were used to describe the affected dogs. The Wilcoxon rank sum test was used to evaluate differences in body weight before and after treatment. Statistical significance was defined as P.05. Results Signalment Dogs ranged in age from 6 to 42 months (median, 13 months). Eight of the dogs were Boxers, and 1 dog was an English Bulldog. Five of the dogs were spayed females (56%), 3 were castrated males (33%), and 1 (the English Bulldog) was an intact male (11%). All dogs were current on routine vaccinations. Clinical History All dogs presented for evaluation of chronic, large bowel diarrhea. Eight (89%) of the dogs had tenesmus, 7 (78%) of the dogs had a history of hematochezia, and all dogs had increased frequency of defecation. Duration of clinical signs ranged from 1.5 to 24 months (median, 8 months). Weight loss was reported in 7 (78%) of the dogs. These signs resolved in all patients after antibiotic therapy was initiated. Physical Examination Findings Initial physical examination findings included body condition scores that ranged from 1.5 to 3 of 5 (median, 2). After treatment, body condition scores ranged from 3 to 3.5

3 Antibiotic-Responsive Histiocytic Ulcerative Colitis 501 Fig 2. Periodic acid-schiff (PAS) stain of a sample from the colonic biopsy from the same patient as shown in Figure 1. Note the PAS-positive macrophages, particularly between the base of the glands and the muscularis mucosa (black arrow) (magnification 10 ). of 5 (median, 3). Body weight at presentation ranged from 16.3 to 23.8 kg (median, 19.6 kg). After antibiotic treatment, body weight ranged from 22 to 33.3 kg (median, 25.8 kg). The difference in body weight before and after treatment was statistically significant (P.001). When the 2 young dogs (aged 6 and 8 months) were excluded to rule out the effect of growth, the differences were still significant (P.01). Other abnormalities noted on physical examination included the presence of fresh blood within or coming from the rectum (5 of 9 dogs, 56%) and a palpable rectal mass that was later determined to be a rectal abscess (1 dog, 11%). These physical examination abnormalities were no longer present after antibiotic therapy. Laboratory Abnormalities A CBC, serum biochemistry, fecal flotation, and urinalysis were performed on all dogs at initial presentation. The only significant laboratory abnormality identified was hypoalbuminemia in 4 of 9 dogs (44%), ranging from 2.3 to 2.7 g/dl in these 4 dogs (normal, g/dl). This abnormality resolved in all 4 dogs after antibiotic therapy. One dog had rare Ancylostoma sp. eggs seen on fecal flotation and was treated with fenbendazole. Abdominal Ultrasound Findings Seven of the 9 dogs had abdominal ultrasonographic examinations performed at the time of the initial evaluation. Of these dogs, 4 of 7 (57%) had evidence of a thickened colonic wall (3.7 6 mm), and 6 of 7 (86%) had evidence of enlarged abdominal lymph nodes. Cytologic evaluation of an ultrasound-guided aspirate of the enlarged lymph nodes in 3 dogs indicated reactive lymphadenopathy, and histiocytes were present in 2 of these aspirates. Abdominal ultrasound examination was not repeated after therapy in any of the dogs. Colonoscopic Findings Seven of the dogs had a flexible colonoscopy performed at initial presentation, 1 had a rigid proctoscopy performed, and 1 had blind colonic biopsies performed. Of the 7 dogs that had flexible colonoscopies performed, 6 were observed to have gross mucosal ulcerations (87%). Five dogs underwent another biopsy after treatment (1 by flexible colonoscopy and 4 by rigid proctoscopy), and none of these dogs had visible mucosal ulcerations. Response to Therapy Four of the 9 dogs (44%) of this study had been treated for HUC before evaluation by the authors. Various combinations of drugs were used, including metronidazole (4 dogs), sulfasalazine (3 dogs), corticosteroids (5 dogs), and azathioprine (3 dogs). The duration of previous therapy ranged from 1 to 20 weeks, with minimal (1 dog) to no (3 dogs) response seen. All of these dogs eventually were switched to antibiotic therapy, which included enrofloxacin (1 dog) or a combination of amoxicillin, metronidazole, and enrofloxacin (3 dogs) at standard recommended dosages, resulting in a resolution of the diarrhea within 3 12 days. Before starting antibiotic therapy, all other medications were discontinued, except for metronidazole. Anthelmintic

4 502 Hostutler et al Table 1. Treatments used in 9 dogs with HUC that responded to antibiotics. Treatment Conventional treatment, a followed by antibiotics (4 dogs) Enrofloxacin Enrofloxacin, metronidazole, and amoxicillin Antibiotics initially without conventional treatment a (5 dogs) Enrofloxacin Enrofloxacin with metronidazole Enrofloxacin, metronidazole, and amoxicillin No. Dogs HUC, histiocytic ulcerative colitis. a Conventional treatment consisted of azathioprine (3 dogs), prednisone (4 dogs), sulfasalazine (3 dogs), metronidazole (4 dogs), and dexamethasone (1 dog) after stopping prednisone. therapy (fenbendazole 50 mg/kg PO q24h) for 3 days was attempted in 6 of 9 dogs without resolution of clinical signs. All dogs in the study had fecal examinations performed, and only 1 had evidence of concurrent parasitic infection (rare Ancylostoma sp. ova seen). In 5 of 9 (56%) dogs, antibiotics were initiated as primary therapy after the diagnosis of HUC was made. One of these dogs received only enrofloxacin and had normal feces within 7 days of starting treatment. Three of the 5 dogs were started on a combination of enrofloxacin, metronidazole, and amoxicillin and had normal feces within 2 7 days. One dog was started on enrofloxacin and metronidazole and had normal feces in 7 days. All drugs were given at standard recommended dosages (Table 1). Histopathologic Findings Colonic biopsy specimens from 8 of the 9 dogs were available for evaluation in a blinded fashion. One of the biopsy specimens had PAS-positive macrophages but could not be properly evaluated because of the plane at which the specimen was cut, and this dog was not further evaluated histologically. Of the remaining 7 dogs, histologic lesions present before treatment included PAS-positive macrophages (7 of 7; 100%), increased separation between or distortion of glands by inflammatory cells in the lamina propria (7 of 7; 100%), abnormal glandular epithelium (ie, hyperplasia or reduction in numbers of goblet cells) (7 of 7; 100%), and atrophy or loss of luminal surface epithelium (ie, ulcerations) (5 of 7; 71%). Five of these dogs had repeated biopsies performed after treatment, and the histologic lesions resolved, except for the persistence of PAS-positive macrophages in 3 of 5 dogs and changes in the numbers of goblet cells in 1 dog. Of the dogs in which PASpositive macrophages were still present, the numbers of these cells were subjectively lower than before treatment. Clinical Outcome All of the dogs of this study were currently free of clinical signs up to 21 months after treatment. Three of the dogs (33%) treated with a combination of enrofloxacin, metronidazole, and amoxicillin for 4 weeks (1 dog) or 6 weeks (2 dogs) are no longer on antibiotic therapy and have been clinically normal for 7 14 months after discontinuation of treatment. The remaining 6 dogs still are receiving enrofloxacin at 68 mg PO q24 h (1 dog), 68 mg PO q48h (4 dogs), or 68 mg PO q72h (1 dog). Duration of therapy in these dogs ranged from 2 to 21 months. In one of these dogs, enrofloxacin therapy was stopped after 4 months, and clinical signs recurred 12 days later. The dog in question responded within 5 days of resuming enrofloxacin therapy. One dog that was started on enrofloxacin at 6 months of age subsequently has developed carpal lameness after approximately 1 year of therapy. Whether this lameness is a consequence of therapy cannot be definitively determined. Cartilage abnormalities have been described in young dogs receiving enrofloxacin. 15 This dog continued to receive enrofloxacin 68 mg PO q72h for 6 months after diagnosis of lameness. No other complications secondary to long-term therapy have been noted. Discussion This study provides evidence that HUC is an antibioticresponsive disease and suggests the possibility that HUC is caused by an infectious agent. Clinical signs including diarrhea resolved in all 9 dogs within 12 days of beginning enrofloxacin alone or in combination with either metronidazole or amoxicillin and metronidazole. Four of these dogs had a previous history of minimal or no response to previous treatment with anti-inflammatory or immunosuppressive drugs and dietary modification. In addition to resolution of clinical signs, colonic biopsy specimens showed marked improvement after antibiotic treatment compared to pretreatment biopsy specimens from all 5 dogs for which adequate follow-up biopsy specimens were available. All 9 dogs are free of clinical signs up to 21 months after beginning treatment, and 3 of them have been off all medication without recurrence for 7 14 months. These results suggest that an infectious agent susceptible to enrofloxacin is the cause of canine HUC. Since the disease was 1st described in 1965, the etiology of HUC has been a matter of debate. Suggested causes have included infectious agents, 7,11,12 immune-mediated mechanisms, genetic predisposition, or a combination of these. 7 The high prevalence of the disease in Boxers suggests that a genetic predisposition plays an important role in the development of the disease. In the 1st report of HUC, most of the affected dogs could be traced back to a single breeding pair of Boxers. 7 Current recommendations for treatment of HUC found in veterinary textbooks include immunosuppressive therapy with prednisone, alone or in combination with azathioprine, and the addition of sulfasalazine and metronidazole as needed. 1,2,16 Various dietary modifications also are used. In dogs treated in this manner, clinical improvement usually is minimal, and a poor-to-guarded prognosis is warranted. 2 The use of antibiotics for HUC 1st was advocated in the initial description of the disease in In that report, both tetracycline and chloramphenicol were used, with the most dramatic results seen with chloramphenicol. The use of chloramphenicol and tylosin in various combinations with prednisone, sulfasalazine, metronidazole, and azathi-

5 Antibiotic-Responsive Histiocytic Ulcerative Colitis 503 oprine may have accounted for the better prognosis for boxers with HUC in a more recent report. 14 Enrofloxacin, the common treatment used in all of the dogs of this report, is a bactericidal antibiotic that has a broad gram-negative spectrum. Specific infectious agents such as Mycobacterium, Mycoplasma, Chlamydia, and rickettsiae have been suggested to play a role in the development of HUC, 7,11,12 and these organisms also are susceptible to enrofloxacin. Other agents, however, cannot be discounted on this observation alone. The authors are unaware of any immunologic effects of enrofloxacin. In some dogs, enrofloxacin was combined with metronidazole and amoxicillin, which provides broad-spectrum coverage against gram-positive, gram-negative, aerobic, and anaerobic bacteria. In addition to its bactericidal anaerobic spectrum, metronidazole is trichomonacidal and amebicidal and has inhibitive actions on cell-mediated immunity. 17,18 The severity of the histologic lesions decreased dramatically in response to treatment in the 5 dogs that underwent follow-up biopsies. Two dogs had no PAS-positive macrophages detectable at 4 and 6 months after starting enrofloxacin therapy. This suggests the cure of the disease histologically as well as clinically. The other dogs that still had PAS-positive macrophages in the biopsy specimens had subjectively fewer macrophages than before treatment. With time, perhaps all histologic lesions, including PASpositive macrophages, would revert to normal. A similar finding is seen in humans with Whipple s disease in whom it takes an average of 14 months for microscopic lesions to subside. 19 HUC often has been compared to other granulomatous enteritides, including Whipple s disease (Tropheryma whipplei) and Crohn s disease in humans and Johne s disease in ruminants (Mycobacterium paratuberculosis). 7,10,20 Whipple s disease is a rare disease that primarily affects white males (8 : 1 male : female ratio), with a mean age of onset of 50 years. 21 Whipple s disease differs from HUC in that it affects the small intestine and other organ systems, whereas HUC is confined to the colon, rectum, and cecum. Both Whipple s disease and HUC are characterized by severe inflammation and distortion of the lamina propria of the affected segment of bowel. Although differences exist, both diseases are granulomatous and appear to be responsive to antibiotics. Whipple s disease was uniformly fatal before the causative agent was identified and found to be susceptible to a variety of antibiotics, including penicillin, penicillin with streptomycin, erythromycin, ampicillin, chloramphenicol, tetracycline, and trimethoprim-sulfamethoxazole. 21 Whipple s disease, however, appears to be resistant to the fluoroquinolones. 22 As hypothesized in dogs with HUC, immunologic and genetic factors may be important in the pathogenesis of Whipple s disease, as evidenced by isolation of the causative organism from normal individuals and reports of the familial occurrence of Whipple s disease. 21 Johne s disease is a systemic granulomatous disease of ruminants caused by M paratuberculosis that affects the small intestine, cecum, and colon. The disease is progressive and results in chronic wasting of the infected animal. Histopathologically, Johne s disease consists of macrophages that contain acid-fast bacteria and distend the mucosa and submucosa of the affected region of the gastrointestinal tract. 20 This finding differs from HUC, in which acid-fast bacteria have not been demonstrated, making mycobacteria less likely as a causative organism. 20 Crohn s disease is a condition in humans that affects the mucosa and submucosa of the ileum and colon. This disease also produces fissures and microscopic sinuses that penetrate the muscular layers of the intestines. 20 On the basis of bacterial culture, serology, immunocytochemistry, pathology, and response to antibiotics, 23 Crohn s disease may be linked to an underlying bacterial infection. Many infectious organisms have the potential to cause granulomatous inflammation of the intestines, including fungi, bacteria, protozoa, helminths, and bacteria. Recently, Bartonella sp. has been reported to cause various granulomatous diseases in dogs, including granulomatous rhinitis, 24 lymphadenitis, 24 and hepatitis 25,26 as well as other problems Although intestinal disease has not been recognized with Bartonella sp., this enrofloxacin-susceptible infectious agent should be investigated as a potential cause of HUC in dogs. None of the dogs of this study were screened for Bartonella infection. The laboratory methods necessary to isolate and identify the suspected infectious organism in HUC may be extrapolated from the experience that has been gained from working with other unculturable organisms such as T whipplei. As in HUC, an infectious cause of Whipple s disease was suspected on the basis of the visualization of bacteria-like substances on light and electron microscopy. 30,31 Similar findings have been reported in Boxers with HUC. The earliest report on the ultrastructure of HUC found that all 13 dogs examined had macrophages with residual bodies. Many of the residual bodies contained membranes, parallel pairs of membranes, and electron-dense particles ranging in size from 100 to 500 nm. The report suggested that these particles were chlamydia, mycoplasma organisms, or noninfectious agents. 12 The causative agent of Whipple s disease subsequently has been identified by means of consensus sequence polymerase chain reaction technology. 32 The possibility of secondary pathogens in the pathogenesis of HUC has previously been mentioned. 14 This suggestion seems unlikely, given the response to treatment in the patients of our study. If bacterial pathogens were a secondary phenomenon leading to inflammation and HUC, a better response would have been expected in the patients that received conventional treatment with corticosteroids and antibiotics. Only 1 of 4 of the dogs of our study had a minimal response, and all dogs responded rapidly to antibiotic therapy once it was instituted. It also would have been very unlikely for the dogs that received only antibiotics to respond so dramatically and not to have relapsed clinically when antibiotics were discontinued. One of the dogs of this study has been off antibiotics for 14 months without recurrence of clinical signs. The dramatic clinical and histopathologic responses to antibiotics provide indirect evidence that a currently unidentified infectious agent plays a critical role in the development of HUC. The dogs of this report experienced a dramatic improvement in histologic lesions and a statistically significant increase in body weight (P.001) after treatment. Previous reports have suggested that the disease

6 504 Hostutler et al is self-limiting in mildly affected dogs. 7,14 This outcome seems unlikely given the findings in the patients of our study. All of the dogs of our study (even those in which conventional therapy did not work) had complete resolution of clinical signs within 12 days of starting antibiotic therapy. Spontaneous resolution also seems unlikely given the number of animals in the study and the duration of their clinical signs ( months) before antibiotic treatment. On the basis of our experience, enrofloxacin (5 mg/kg PO q24h) should rapidly lead to resolution of clinical signs in dogs with HUC. The optimal duration of treatment requires further investigation. One dog on prolonged enrofloxacin therapy relapsed within 12 days of discontinuing treatment. Three other dogs have been off enrofloxacin therapy for 14 months and are free of clinical signs, although they do retain PAS-positive macrophages. References 1. Guilford WG. Idiopathic inflammatory bowel diseases. In: Guilford WG, ed. Strombeck s Small Animal Gastroenterology, 3rd ed. Philadelphia, PA: WB Saunders; 1996: Sherding RG. Diseases of the large intestine. In: Tams TR, ed. Handbook of Small Animal Gastroenterology, 2nd ed. St Louis, MO: Saunders/Elsevier; 2003: Stokes JE, Kruger JM, Mullaney T, et al. Histiocytic ulcerative colitis in three non-boxer Dogs. J Am Anim Hosp Assoc 2001;37: Van derngaag I, Van Toorenburg JIK, Voorhout G, et al. Histiocytic ulcerative colitis in a French Bulldog. J Small Anim Pract 1978; 19: Tanaka H, Nakayama M, Takase K. Histiocytic ulcerative colitis in a French bulldog. J Vet Med Sci 2003;65: Van Kruiningen HJ, Dobbins WO. Feline histiocytic colitis, a case report with electron microscopy. Vet Pathol 1979;16: Van Kruiningen HJ, Montali RJ, Strandberg JD, et al. A granulomatous colitis of dogs with histologic resemblance to Whipple s disease. Pathol Vet 1965;2: Russell SW, Gomez JA, Trowbridge JO. Canine histiocytic ulcerative colitis. The early lesion and its progression to ulceration. Lab Invest 1971;25: Cockrell BY, Krehbiel JD. Ultrastructural changes in histiocytic ulcerative colitis in a Boxer. Am J Vet Res 1972;33: Kennedy PC, Cello RM. Colitis of Boxer dogs. Gastroenterology 1966;51: Gomez JA, Russell SW, Trowbridge JO, Lee J. Canine histiocytic ulcerative colitis: An ultrastructural study of the early mucosal lesions. Am J Dig Dis 1977;22: Van Kruiningen HJ. The ultrastructure of macrophages in granulomatous colitis of Boxer Dogs. Vet Pathol 1975;12: Bowe PS, Van Kruiningen HJ, Rosendal S. Attempts to produce granulomatous colitis in Boxer Dogs with a mycoplasma. Can J Comp Med 1982;46: Churcher RK, Watson ADJ. Canine histiocytic ulcerative colitis. Aust Vet J 1997;75: Ihrke PJ, Papich MG, Demanuelle TC. The use of fluoroquinolones in veterinary dermatology. Vet Dermatol 1999;10: Jergens AE, Willard MD. Diseases of the large intestine. In: Ettinger SJ, ed. Textbook of Veterinary Internal Medicine, 5th ed. Philadelphia, PA: WB Saunders; 2000: Jergens AE. Feline idiopathic inflammatory bowel disease. Compend Cont Educ Pract Vet 1992;14: Willard MD. Inflammatory bowel disease: Perspectives on therapy. J Am Anim Hosp Assoc 1992;28: Wroe SJ, Pires M, Harding B, et al. Whipple s disease confined to the CNS presenting with multiple intracerebral mass lesions. J Neurol Neurosurg Psychiatry 1992;429: Van Kruiningen HJ. Granulomatous colitis of Boxer Dogs: Comparative aspects. Gastroenterology 1967;53: Dutly F, Altwegg M. Whipple s disease and Tropheryma whippelii. Clin Microbiol 2001;14: Masselot F, Boulos A, Maurin M, et al. Molecular evaluation of antibiotic susceptibility: Tropheryma whipplei paradigm. Antimicrob Agents Chemother 2003;47: Van Kruiningen HJ. On the use of antibiotics in Crohn s disease. J Clin Gastroenterol 1997;20: Pappalardo BL, Brown T, Gookin JL, et al. Granulomatous disease associated with Bartonella infection in 2 dogs. J Vet Intern Med 2000;14: Kitchell BE, Fan TM, Kordick D, et al. Peliosis hepatitis in a dog infected with Bartonella henselae. J Am Vet Med Assoc 2000; 216: Gillespie TN, Washabau RJ, Goldschmidt MH, et al. Detection of Bartonella henselae and Bartonella clarridgeiae DNA in hepatic specimens from two dogs with hepatic disease. J Am Vet Med Assoc 2003;222: Breitschwerdt EB, Kordick DL, Malarkey DE, et al. Endocarditis in a dog due to infection with a novel Bartonella subspecies. J Clin Microbiol 1995;33: Breitschwerdt EB, Atkins CB, Brown TT, et al. Bartonella vinsonii subspecies and related members of the alpha subdivision of Proteobacteria in dogs with cardiac arrythmias, endocarditis, or myocarditis. J Clin Microbiol 1999;37: Chomel BB, Mac Donald KA, Chang CC, et al. Aortic valve endocarditis in a dog due to Bartonella calrridgeiae. J Clin Microbiol 2001;39: Cohen AS, Schimmel EM, Holt PR, et al. Ultrastructural abnormalities in Whipple s disease Proc Soc Exp Biol Med 1960; 105: Chears WC, Ashworth CT. Electron microscopy study of the intestinal mucosa in Whipple s disease: Demonstration of encapsulated bacilliform bodies in the lesion. Gastroenterology 1961;41: Relman DA, Schmidt TM, MacDermott RP, et al. Identification of the uncultured bacillus of Whipple s disease. N Engl J Med 1992; 327: