Long-Term Evaluation of Canine Perianal Fistula Disease Treated With Exclusive Fish and Potato Diet and Surgical Excision

Introduction

Canine perianal fistula disease (PAF) is a chronic, debilitating, painful, and progressive disease of the anus, perianal skin, anal sacs, and adjacent tissues.^1–3 Perianal fistulas have been associated with chronic, granulomatous inflammatory bowel disease (IBD) in both humans and dogs.⁴ The disease affects medium- to large-breed dogs—predominantly German shepherd dogs and German shepherd mixed-breed dogs. In one study, German shepherd dogs comprised 84% of dogs evaluated; however, other breeds affected included Labrador retrievers, Irish setters, Old English sheepdogs, bulldogs, and collies.⁵

Perianal fistula disease is characterized by inflammation, ulceration, and draining tracts that affect the perianal region and occasionally the rectal lumen.² Clinical signs include tenesmus, hematochezia, constipation, self-mutilation, anal stenosis, and severe discomfort; these can lead to systemic signs of illness such as lethargy, anorexia, and weight loss.² The fistula size and location can vary from pinpoint and superficial to large, deep ulcerations extending into the perianal region, involving the anal sacs and deep rectal tissue. Histological lesions are characterized by infiltration of lymphoid cells, plasma cells, and eosinophils consistent with immunological activation.³ Human patients with Crohn’s disease of the perineum have similar histological lesions.³

Although PAF is a well-documented disease, the cause is poorly understood. Recent reports on the etiopathogenesis have attempted to link an immunopathological, bacterial, hormonal, endocrine, or anatomical basis for the disease; however, none have been conclusive.^1–3,6,8 Both medical and surgical treatment options have been reported with inconsistent results. Historically, medical management of canine PAF has been directed at altering the local environment of the perineum through use of tail braces, regular cleansings of the affected area, and controlling infection using a variety of topical and systemic antibiotics.^2,6–8 Because long-term results of medical management are poor, this treatment modality is considered palliative.^1,9–11 Surgical treatments such as excision, deroofing and fulguration, cryosurgery, and chemical cauterization are associated with varying recurrence rates and a high prevalence of complications.^6,8,11,12

Similar to canine PAF, severe Crohn’s disease of the perineum in human patients is noticeably difficult to treat; however, some success has been achieved utilizing immunosuppressive drugs such as cyclosporine, azathioprim, tacrolimus, and prednisone.^{1,3,9,10,13–15} Successful use of these drugs in dogs with PAF supports the theory that PAF has an immunological basis.^{1,3,9,10,13–15} However, treatment with immunosuppressive agents is associated with an increased risk of complications, significant costs, and frequent relapses when therapy is discontinued,3,9,16 making alternative treatment strategies desirable.

The purpose of this study is to report the results of feeding exclusively a specific, novel protein diet (i.e., Eukanuba brand Fish and Potato^a) in conjunction with surgical excision of fistulas and bilateral anal saculectomy in the treatment of PAF in a series of dogs.

Materials and Methods

The medical records of 33 consecutive dogs that were presented for PAF at Long Island Veterinary Specialists from September 1998 through April 2003 were reviewed. All dogs selected for the study were treated with the study protocol and had at least 1-year long-term follow-up evaluations. No dogs underwent any form of surgical treatment prior to presentation. Medical records were reviewed to determine degree and extent of fistulas, degree of concurrent anal sac disease, and outcomes at three follow-up times. The degree and extent of fistulas were classified according to a previously used grading system [Table 1].^2,8,17

All dogs were first treated with cephalexin^b (22 mg/kg per os [PO] q 8 hours), metronidazole^c (15 mg/kg PO q 12 hours), and sulfasalazine^d (15 mg/kg PO q 8 hours) in conjunction with a white fish and potato diet for 1 to 180 days prior to surgical excision, depending on owner compliance and surgeon’s discretion. Eukanuba Fish and Potato^a was fed exclusively to all dogs except one. In the one dog where Eukanuba Fish and Potato^a diet was not considered palatable, a salmon and rice^e exclusive diet was used. In addition, all dogs were taken off previous medications, supplements, or chew toys that contained flavoring or gelatin derived from pieces of cattle, swine, horses, or chicken. The family members were counseled and instructed to strictly follow these dietary guidelines.

Surgical excision occurred from 1 to 180 days after presentation. All dogs were premedicated with a one-time dose of butorphanol tartrate (0.2 mg/kg subcutaneously [SC]) and atropine sulfate (0.02 mg/kg SC) prior to induction with tiletamine hydrogen chloride (0.2 mg/kg intravenously [IV]). After endotracheal intubation, anesthesia was maintained with isoflurane and oxygen. Epidural anesthesia was administered using morphine sulfate (0.1 mg/kg). Cefazolin sodium (22 mg/kg IV) was administered at the time of anesthetic induction.

En bloc excision of all visibly diseased tissue was performed with bilateral anal saculectomy. At the discretion of the surgeon, dogs with severe bilateral anal sac involvement had two unilateral anal saculectomies staged 4 weeks apart to preserve maximal external anal sphincter function. After tissue excision, the subcutaneous layer was apposed to the external anal sphincter or rectal wall using 3-0 polydioxanone suture in a simple interrupted fashion. The rectal mucosa and skin were reapposed using 3-0 polydioxanone suture in a simple inverted interrupted pattern. All excised tissue was submitted for histological evaluation.

All dogs received hydromorphone (0.05 mg/kg IV or SC) or butorphanol tartrate (0.2 mg/kg IV or SC) every 6 hours for 24 to 48 hours after surgery. Hydrotherapy of the surgical site was performed for 15 minutes every 6 hours for the first 48 to 72 hours postoperatively. An Elizabethan collar was fitted to prevent self-mutilation and was maintained for 2 weeks after surgery. Postoperative medical therapy consisted of metronidazole,^c cephalexin,^b and sulfasalazine^d for 30 to 60 days after surgery. Dietary management with Fish and Potato^a diet was continued for the duration of the study.

All dogs were grouped into four different categories: resolution of fistulas without clinical signs, resolution of fistulas with clinical signs, incomplete resolution of fistulas without clinical signs, and incomplete resolution of fistulas with clinical signs. Clinical signs consisted of licking, tenesmus, and straining. All dogs were evaluated at 3 weeks, 3 months, and 1 year postoperatively. Dogs with staged or additional procedures had a follow-up time of 1 year beginning with the second surgery.

Results

Thirty-three dogs were included in the study. Sixteen (48.5%) dogs were male, and 17 (51.5%) were female. Twenty-six (78.8%) were German shepherd dogs, two (6.1%) were Labrador retrievers, one (3.0%) was a beagle, one (3.0%) was a border collie, and three (9.1%) were mixed-breed dogs. Ages of dogs ranged from 2 to 13 years, with a mean of 7.4 years and a median of 7.0 years. The most common presenting clinical signs were tenesmus (35.5%), licking (19.4%), and hematochezia (19.4%) [Table 2]. Thirty-two (96.9%) dogs had concurrent anal sac disease, which included abscessation, impaction, and histological abnormalities. Four (12.9%) of 33 dogs had concurrent dermatopathy. The degree and extent of fistulation ranged from 180° to 360°, with a median of 270°. Treatments prior to presentation consisted of various systemic antibiotics, immunosuppressants, steroids, stool softeners, non-steroidals, local antiseptics, and dietary changes, including bulk fiber diets or bland commercial or cooked diets.

Medical and dietary management was initiated at the time of presentation and continued until surgical excision, which ranged from 1 to 180 days. Twenty-nine dogs were treated with bilateral anal saculectomy and complete surgical excision of all visibly diseased tissue. Two dogs with severe disease were treated with staged unilateral anal saculectomy. Two dogs with moderate disease were treated with complete surgical excision of diseased tissues and unilateral anal saculectomy. Recurrence of PAF on the untreated side was observed 3 months after surgery in one dog and 2 years after surgery in the other dog. The problem was resolved with subsequent anal saculectomy and excision of the associated fistula tracts in both dogs.

The histopathological examinations revealed moderate, lymphoplasmacytic, fibrosing anal saculitis in 18 dogs; severe, lymphoplasmacytic, fibrosing anal saculitis in three dogs; moderate, pyogranulomatous anal saculitis in eight dogs; and severe, pyogranulomatous anal saculitis in three dogs. Normal anal sacs were found in one dog.

At the 3-week postoperative evaluation, 18 dogs had no clinical signs and no evidence of fistula disease. Nine dogs had intermittent clinical signs consisting of licking, diarrhea, dyschezia, tenesmus, hematochezia, and intermittent incontinence with no evidence of fistula disease. Four dogs had a pinpoint fistula tract, with three of these dogs showing intermittent bloody discharge, tenesmus, diarrhea, and decreased anal tone. Two dogs had fistulas approximating 90°. One of these dogs had intermittent diarrhea and licking; the other dog had intermittent tenesmus.

At the 3-month postoperative evaluation, 23 dogs had no clinical signs and no evidence of fistula disease. Four dogs that previously had no evidence of fistula disease developed pinpoint fistula tracts, but only one of these dogs had clinical signs of mild intermittent diarrhea. Two of these dogs were off medications at the time of evaluation. Of the four dogs with pinpoint fistula disease present at the 3-week examination, one dog had complete resolution of disease, while the other three dogs continued to have pinpoint fistula tracts but showed improvement in clinical signs. Of the two dogs with 90° fistula involvement, one continued to have fistula tracts approximating 90° involvement of the anal sphincter but with an improvement in clinical signs, while the other had complete resolution of fistula disease and clinical signs.

At the 1-year postoperative evaluation, 29 (87.9%) dogs had complete resolution of visible fistula disease. Twenty-three (79.3%) dogs were completely free of clinical signs, while six (20.7%) dogs continued to have intermittent licking, diarrhea, constipation, tenesmus, or decreased anal tone on digital examination. Fecal incontinence was not reported for any dog. One dog with mild tenesmus was not tolerating the Eukanuba Fish and Potato^a diet and was switched to Hill’s Salmon and Rice^e several months prior to the 1-year evaluation. Four dogs presented with approximately 90° fistula disease; three of these dogs had no clinical signs, and one dog had mild tenesmus and intermittent diarrhea. Of the four dogs that had evidence of macroscopic disease, all had some form of a concurrent dermatopathy or autoimmune disease, which included severe chronic-active allergic dermatitis, systemic lupus erythematosus, and ulcerative erosive vaginitis.

Discussion

The gender, age, and breed distributions in the authors’ study are consistent with previous reports.^8,11,17 Several factors have been reported to explain the overrepresentation of German shepherd dogs with PAF. Anatomical variations like an increased number of apocrine sweat glands in the anocutaneous zone of the German shepherd dog^11,18 and low tail carriage⁷ have been purported to play a role in the etiopathogenesis; however, neither has been clearly identified as a contributing factor. Immune dysregulation associated with IBD and bacterial overgrowth in German shepherd dogs has been reported to result in increased concentrations of interleukin (IL)-2, IL-5, IL-12p40, tumor necrosis factor alpha, and tumor growth factor beta in intestinal tissue.^19,20 Histological evaluation of PAF tissue samples supports the theory of immune dysregulation in the etiopathogenesis of PAF in German shepherd dogs.^{1,2,5,11,18–20}

The numerous treatments for PAF have had mixed results.^6,8 Surgical techniques described include chemical cauterization, deroofing and fulguration, cryosurgery, laser excision, surgical excision, and tail amputation.^{6–8,11,12,17} Chemical cauterization and excision of superficial fistulas, followed by application of Lugol’s solution to deep fistula tracts, has a reported success rate as high as 87%; however, fecal incontinence occurred in 20% of dogs, and the recurrence rate was 17%.⁸ Using silver nitrate as the cautery agent resulted in a similar recurrence rate of 16.5%.⁸ Deroofing and fulguration have been recommended for dogs with mild to moderate disease,¹² but these were unsuccessful when treating dogs with a 270° to 360° perianal involvement.⁶ Additionally, wounds are left open to heal by second intention and require constant cleansing for up to 3 to 4 weeks postoperatively.⁶ A recurrence rate of 70% was recorded, with 30% of dogs euthanized due to poor long-term outcome.¹² Deroofing and fulguration do not alleviate stricture formation and may contribute to scar formation if disease is extensive.¹² Cryosurgery using liquid nitrogen or nitrous oxide was reported to have success rates ranging from 48% to 97%.⁶ However, recurrence and stricture formation were seen in 45% and 47% of the dogs, respectively.¹⁷

Surgical excision and anal saculectomy resulted in severe complications, including recurring fistulas in 45% of dogs, anal stenosis in 9.5% of dogs, incontinence in 75% of dogs, and diarrhea in 71% of dogs.¹¹ When bilateral anal saculectomy was included with surgical excision, postoperative complications were similar and included a recurrence rate of 56.1%, with 14.6% of dogs developing anal stricture and 26.8% of dogs developing incontinence.¹⁷ Surgical debridement of visible disease, as well as bilateral anal saculectomy, are important components to successful therapy because of the high prevalence of concurrent anal sac disease.¹¹ Failure to remove this nidus for reinfection may influence the high recurrence rates seen in treatment modalities that do not include removal of fistulas and both anal sacs. Early reports of tail amputation to the level of the second or third coccygeal vertebra described success rates up to 80%, but with recurrence rates as high as 40%. Superficial tracts resolved, but deep perianal fistulas and anal stenosis did not heal.^6,7

Although all reported surgical options have some level of success, all were associated with higher recurrence rates as well as more complications than the dogs in this study.^6,8 Only six dogs experienced intermittent diarrhea, licking, constipation, tenesmus, or decreased anal tone on digital examination 1 year after surgery. No evidence of fecal incontinence was seen in the dogs of this study. These complications are considerably less in number and severity when compared to previous reports of various surgical or medical treatments.^6,8,11,12,17 This difference may be due to the benefits of starting medical and dietary therapy before surgery to reduce the diseased area, thus decreasing the extent of the surgical procedure. Additionally, enhanced knowledge of the surgical implications of the regional anatomy may have contributed to less iatrogenic trauma.

Crohn’s disease in humans is a form of severe IBD resulting in similar clinical manifestations as seen in dogs with PAF.^4,21–23 Previous reports have suggested a link between PAF and colitis, further lending support to the theory that canine PAF has an immune-mediated etiology.^9,11,14 Comparisons between canine PAF disease and fistula development associated with Crohn’s disease in humans stimulated research into the use of immunosuppressive therapy in dogs.^9,13–16,24 As a result, most recent medical recommendations for treating PAF include the use of immunosuppressive drugs such as prednisone, cyclosporine, and azathioprine—each with varying degrees of clinical efficacy, costs, and complications.¹⁶ Clinical resolution was seen in 33% of dogs treated with immunosuppressive doses of prednisone; improvement was seen in 66%.^14–16 Clinical improvement was seen in 83% to 100% of dogs treated with oral cyclosporine.^13,16 When five dogs were treated with a combination of azathioprine and metronidazole, all dogs showed reduction in severity and extent of perianal fistula lesions.^3,16 Although treatment with immunosuppressive agents results in clinical improvement, recurrence and complication rates still remain a considerable challenge.^17,19

Antiinflammatory therapy has numerous complications and unwanted side effects. High-dose prednisone therapy can cause deep pyoderma, iatrogenic hyperadrenocorticism, nonseptic suppurative inflammation of the stifle, and pyogranulomatous cellulites in treated dogs.^14,15 Additionally, side effects such as polyuria, polydipsia, polyphagia, muscle wasting, pancreatitis, and pulmonary thromboembolism may be unacceptable to owners.³ Azathioprine suppresses both antibody-mediated and cell-mediated immunity, which may result in thrombocytopenia and myelosuppression. Other side effects of azathioprine include gastrointestinal upset, pancreatitis, and hepatotoxicosis.³ Cyclosporine is associated with vomiting, anorexia, lethargy, infection, gingival hyperplasia, and hypertrichosis.^13,15,16 Pyometra and septic arthritis of the scapulohumeral joint secondary to cyclosporine treatment has been reported.^15,16 Other disadvantages to using cyclosporine include an erratic rate of absorption and metabolism, which necessitates monitoring to determine if plasma concentrations are within therapeutic range.¹⁵ Also, drug interactions are unpredictable and can result in potential toxicity or treatment failure.¹⁶ The cost of using cyclosporine and the required monitoring during lifelong treatment can be prohibitive.^3,15 However, adding ketoconazole to cyclosporine in order to decrease its hepatic elimination decreases the efficacious dose of cyclosporine, thus decreasing costs substantially.^16,19,24

The medical regimen the authors implemented as an adjunct to dietary management consisted of cephalexin^b (22 mg/kg PO q 8 hours), metronidazole^c (15 mg/kg PO q 12 hours), and sulfasalazine^d (15 mg/kg PO q 8 hours). Cephalexin^b was chosen because of its excellent efficacy against most surgical wound pathogens, its low toxicity, its ability to penetrate most tissues and body fluids, its reasonable cost, and minimal side effects.²⁹ Sulfasalazine,^d a combination of sulfapyridine and 5-aminosalicylic acid joined by an azo bond, was selected because of its effectiveness in the management of canine colitis.²⁸ The exact mechanism of action in relation to canine colitis is unknown, but it appears to be unrelated to its antimicrobial activity.²⁸ Speculation is that the liberated sulfapyridine is absorbed and the 5-aminosalicylic acid acts topically within the colon to reduce inflammation.²⁶ The proposed mechanism of action of the 5-aminosalicylic acid is secondary to inhibition of prostaglandin synthesis, alteration of prostaglandin metabolism, and inhibition of leukotriene production.²⁶ Metronidazole^c was selected not only for its effectiveness against Giardia species and anaerobic bacteria, but also for its ability to suppress cell-mediated immune reactions responsible for colitis.²⁶

The novel protein dietary therapy used in this study may have contributed to the low recurrence rate. An association between anal furunculosis and colitis has already been documented.^2,3 Most theories on the exact etiology of canine colitis implicate an immunemediated inflammatory reaction specific to an antigen that gained access to the submucosa and lamina propria of the colon.^26–28 In addition, response of canine PAF to immunosuppressive drugs such as cyclosporine and tacrolimus supports an immunological basis as a possible underlying etiology. Although the underlying antigen could not be identified in this study, management directed at allergen avoidance may be a safer and more effective component of the management of the immune response compared to immunosuppressive drugs.

By feeding only Eukanuba Fish and Potato,^d the authors also increased the amount of fish oils (specifically omega-6 [OFA6] and omega-3 [OFA3] fatty acids) in the diet. A recent review regarding the nutritional management of Crohn’s disease summarized the results of various dietary therapies, including the use of OFA6 and OFA3.³⁰ Incorporation of fish oils in the diet of Crohn’s disease patients reduced the severity of colitis by 56%.⁴ Diets rich in omega fatty acids reduce the level of thromboxanes in the chronic stages of inflammation and shorten the course of colonic disease.³⁰ Thus, the clinical response the authors observed possibly was not due to the elimination of a dietary antigen, but rather to the beneficial effects of OFA6 and OFA3 in the diet. Additionally, recent reports emphasize the importance of maintaining a safe and effective OFA6 to OFA3 ratio of 5:1 when supplementing the diet.^31–35 The Eukanuba Fish and Potato^a diet meets this recommended ratio.³⁶

Conclusion

Dogs with PAF can be successfully treated with en bloc surgical excision of diseased tissue, bilateral anal saculectomy, and perioperative dietary management with an exclusive white fish and potato diet. The authors attribute the success rate in this study to such factors as presurgical dietary and medical therapies to reduce diseased tissue, bilateral anal saculectomy to remove the potential nidus for fistula recurrence, and long-term dietary management with a white fish and potato diet enriched with OFA3 and OFA6.