Surgical and Medical Treatment of Pyloric and Duodenal Pythiosis in a Dog

Introduction

Pythiosis, an aquatic oomycete caused by the agent Pythium insidiosum, is a cause of gastrointestinal or cutaneous disease in dogs. This organism affects animals living in temperate, tropical, and subtropical climates throughout the world. In the United States, it is most often in the Gulf Coast states, but has been documented in New Jersey, Virginia, North Carolina, southern Illinois, southern Indiana, Kentucky, Oklahoma, Missouri, Kansas, and California.¹ Since the cell wall lacks chitin and contains cellulose and β-glucan, the organism resembles algae rather than a fungus.¹ Gastrointestinal pythiosis is difficult to manage with a poor to grave prognosis. In general, the gastrointestinal form is characterized by severe transmural, segmental thickening, most commonly involving the gastric outflow tract or the ileocecal-colic junction.^2–4 A combination of medical and surgical therapy is warranted as only a small percentage of dogs (<20%) will respond to medical treatment alone.¹ Currently recommended medical management consists of a combination of itraconazole and terbinafine.¹ Surgical excision for gastrointestinal pythiosis with radical surgical excision of 3–4 cm margins is the current recommendation.^1,2–4 Survival has only been documented in three dogs with gastrointestinal pythiosis: one dog treated with a partial gastrectomy; one dog treated medically with itraconazole, terbinafine, and mefanoxam; and one dog treated with a subtotal colectomy, itraconazole, terbinafine, and immunotherapy.^2,5,6 This case report describes survival in a dog with pyloric and duodenal pythiosis treated with radical surgical excision, itraconazole, and terbinafine.

Case Report

A 5 yr old male neutered mixed breed dog weighing 13.5 kg presented to his primary veterinarian for persistent vomiting and weight loss. A barium series was performed where a gastric outflow obstruction was diagnosed. Exploratory surgery at the referring veterinarian revealed a diffusely thickened pylorus and an incisional biopsy of the pylorus was performed. Biopsy samples of the abnormal tissue were obtained and submitted for histopathological analysis. Histopathologic analysis revealed eosinophilic and mixed inflammation with necrosis and fibrosis with intralesional hyphae. A Gomori's methenamine sliver (GMS) stain revealed rare hyphal elements with wide, irregular braches and poorly visible septations. Pythiosis was considered the primary differential. Approximately 2 wk after the incisional biopsy of the pylorus, the persistent vomiting recurred and the patient was referred to the small animal teaching hospital at Mississippi State University College of Veterinary Medicine for further diagnostics and treatment.

Upon presentation, the dog was depressed but alert and responsive. Mild tachycardia was noted with a heart rate of 160 beats per min. He was severely emaciated with a body condition score of 1/5. The remainder of the physical examination was unremarkable.

A complete blood count and serum biochemistry revealed a mild hypoproteinemia (5.8 g/dl: 6.0–8.0 g/dL), mild hypocalcemia (8.3 mg/dL: 8.8–11.2 mg/dL), mild hypophosphatemia (2.3 mg/dL: 2.5–5.0 mg/dL), and a mild hypocholesterolemia (137 mg/dl: 140–360 mg/dL). A urinalysis via cystocentesis was unremarkable. Thoracic radiographs were unremarkable. Abdominal radiographs revealed a loss of serosal detail presumed to be due to the nutritional state of the patient. No free air or gas was noted in the abdomen. The remainder of the abdominal radiographs was unremarkable. Abdominal ultrasound revealed a gas-distended stomach with a circumferentially thickened (up to 1 cm thick) and hypoechoic pyloric wall. The pancreas was mildly enlarged and hypoechoic. Multiple mesenteric lymph nodes were hypoechoic but normal in size. Ultrasound findings of the pylorus and mesenteric lymph nodes were consistent with the presumptive diagnosis of pythiosis; however, other granulomatous disease or neoplasia could not be excluded.

An exploratory celiotomy was performed that revealed multifocal intramural masses within the pylorus and proximal duodenum. The abnormal tissue was clearly demarcated beginning at the pyloric antrum and continuing through the proximal 2.0 cm of the duodenum. The only other abnormal findings on abdominal exploration were enlarged and tortuous renal arteries (predominantly the left) and enlarged, slightly irregular adrenal glands. The significance of these findings are unknown, but they are most likely clinically insignificant. A partial gastrectomy with gastrojejunal anastomosis (Billroth II) was completed. Briefly, a 6.0 cm margin was determined from the mass demarcation aboral to the lesion and non-crushing forceps were placed across the duodenum. A thoracic-abdominal (TA) 30 (3.5 mm) stapler^a was placed across the duodenum oral to the forceps and fired. The duodenum was transected oral to the staple line and the remaining duodenal stump was over sewn with 3-0 Polydioxanone^b. Next, the vasculature and common bile duct were isolated and ligated from the diseased duodenum. Branches of the right gastroepiploic and right gastric arteries and veins were ligated from the diseased portion of the pylorus with a ligate-divide-stapler^a . The omentum and smaller vasculature were cauterized with monopolar electrocautery. In the area between the pylorus and the body of the stomach, a TA 90 (4.8 mm) stapler^a was placed 5.0 cm oral to the pylorus and fired. The stomach was transected aboral to the staple line and the staple line was over sewn using a Cushing pattern of 3-0 polydioxanone^b. The body of the pancreas was isolated, ligated, and transected near the pyloric antrum. The entire diseased portion of the pylorus, duodenum, and right limb of the pancreas was then removed from the abdomen and submitted for histopathologic analysis (Figure 1). The gallbladder was freed from its hepatic fossa and a cholecystoduodenostomy was performed (ensuring a stoma of 3–4 cm) at the level of the remaining duodenal stump using 3-0 polydioxanone^b. Using a side-to-side anastomosis, a gastrojejunostomy was completed. Briefly, a portion of the proximal jejunum was brought to the greater curvature of the stomach. The serosa of the antimesenteric border of the proximal jejunum was sutured lengthwise with 3-0 polydioxanone^b in a simple, continuous pattern to the serosa of the stomach. A stab incision was made into the stomach and along the antimesenteric jejunal border. A gastrointestinal anastomosis stapler^a and a TA 55 (2.5 mm) stapler^a were used to complete the anastomosis. Several interrupted sutures of 3-0 polydioxanone^b were placed between the stomach and the jejunum at the beginning and end of this stoma to protect the anastomosis. Prior to closing the abdomen, a gastrostomy tube^c was placed and held in place with a box lock suture pattern of 3-0 polydioxanone^b. A mesenteric lymph node was biopsied and submitted for histopathology. Abdominal closure was routine and recovery from anesthesia in the intensive care unit was unremarkable.

View Figure

• Download as Powerpoint
• Download Figure

Opioid pain management was used to provide analgesia. Initially, the dog was on a fentanyl (5 mcg/kg/hr) and lidocaine (50 mcg/kg/min) constant rate infusion Once the dog was able to tolerate oral pain medication, tramadol (5 mg/kg every 8 hr) was given. During hospitalization, the dog's caloric needs were calculated and provided via the gastrostomy tube^c until the dog was eating orally. Medical management was initiated with oral itraconazole at 10 mg/kg per day and oral terbinafine at 10 mg/kg per day. Because of the high incidence of gastric irritation and ulceration associated with the Billroth II procedure, famotidine at 1 mg/kg per day, omeprazole at 1 mg/kg per day, and sucralfate at 1.0 g every 8 hr were started. Pancrelipase oral powder^d was initiated at 0.5 teaspoons with each meal since the major/minor duodenal papillae and associated pancreatic ducts were excised during surgery. While awaiting histopathologic and serology results, the dog was discharged.

Histopathology revealed ulcerative and granulomatous gastritis with intralesional “fungal”-like organisms (Figures 2 and 3); the pancreas was unremarkable, and the mesenteric lymph node was considered reactive with no signs of organisms. No organisms were detected at the surgical margins; however, inflammation did extend to the within 1.0 mm of the glandular stomach. At the time of specimen submission, our lab used an outside lab for pythiosis enzyme-linked immunosorbent assay (ELISA) testing^e, which revealed a positive ELISA, but it gave no quantitative percentage to the positive value. At this time, a definitive diagnosis of gastrointestinal pythiosis localized to the pylorus and proximal duodenum was made.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

At the first recheck appointment 2 wk postoperatively, the dog was bright, alert, and responsive. The dog was eating on his own and the gastrostomy tube was removed. All medications were continued with a recheck appointment scheduled for 2 wk later. Approximately 5 wk after surgery, the dog presented again for anorexia and intermittent vomiting. A complete blood count and serum biochemistry were completed revealing severe elevations of his alkaline phosphatase (ALP) and alanine aminotransferase (ALT) of 1885 U/L (11–140 U/L), and 2148 U/L (10–90 U/L), respectively. Thoracic and abdominal radiographs were taken revealing gas distention of the stomach suggestive of an outflow obstruction. Abdominal ultrasound revealed a hypoechoic liver and fine needle aspirates of the liver were obtained. The aspirates were of uniform cellular cytology suggesting no hepatic disease. An upper gastrointestinal series with barium was completed revealing a decreased gastric emptying time. An acute, drug-induced hepatotoxicity was suspected and the dog was hospitalized. Both the itraconazole and terbinafine were discontinued with rechecking of liver values every 2 days to monitor progression. Recheck of serum biochemistry 48 hr later revealed an ALP of 2497 U/L and an ALT of 904 U/L. The dog was discharged with instructions to have blood work rechecked with the referring veterinarian in 2 wk and to discontinue the terbinafine and itraconazole. Because of the elevation in the liver values, the dog was started on a hepatoprotective supplement^f as well as metoclopramide at 0.4 mg/kg orally every 8 hr to help with gastrointestinal motility. Blood work obtained by the referring veterinarian 2 wk post-discharge (7 wk post-surgery) revealed an ALP of 2810 U/L and an ALT of 735 U/L. Because the elevations of liver values were waxing and waning over a 4 wk period and were increased longer than would be expected following acute drug hepatotoxicity, a recheck exam was scheduled. At this time (8 wk post-surgery) the ALP and ALT were 903 U/L and 914 U/L, respectively. An ammonia tolerance test revealed a normal result of <1.0 umol/L for both the pre and post ammonia levels. Blood was submitted for recheck serology testing to a different lab^g, which gave a negative ELISA serology result with a percent positivity of 11%. The terbinafine was once again started orally, but at a dose of 5 mg/kg per day. The dog's liver values were monitored every 2 wk with the referring veterinarian where the values were noted to be waxing and waning but consistently elevated. At 12 and 24 wk post-surgery, blood samples were submitted for ELISA serology^e revealing consistent negative ELISA tests with a percent positivity of 12%. At the 24 wk postoperative recheck, an abdominal ultrasound was completed showing no significant change in the liver aside from mild dilation of intrahepatic bile ducts, especially in the cranial-most region of the liver. No mural abnormalities of any portion of the gastrointestinal tract were noted.

Over the next 4 mo, the dog's liver values were monitored with waxing and waning elevated values. A presumed cholangiohepatits was suspected secondary to the cholecystoduodenostomy procedure and the patient was treated empirically with antibiotics. The dog was first treated with a 21 day course of amoxicillin/clavulanic acid orally at 13.75 mg/kg every 8 hr, then switched to a 30 day course of ciprofloxacin orally at 10.5 mg/kg every 12 hr. At approximately 10 mo postoperatively, the dog presented again with intermittent episodes of vomiting and lack of appetite. Repeat serum chemistry revealed an ALP of 3090 U/L and an ALT of 1455 U/L as well as an elevated pre ammonia tolerance test value of 89 umol/L (1–32 umol/L). The dog was continued on the ciprofloxacin and was started on neomycin and lactulose orally at 22 mg/kg every 8 hr and 15 mL every 8 hr, respectively. Ursodiol was also started orally at 15 mg/kg every 8 hr. Because of the elevated ammonia level, the dog underwent a laparoscopic abdominal explore with hepatic biopsies. Surgical findings were within normal limits except for the liver, which was small and had a brown mottled appearance. The previous surgical sites could not be completely evaluated during the laparoscopic procedure due to adhesion formation and omentum coverage. Samples were submitted for aerobic and anaerobic culture as well as histopathology. The cultures were negative for any bacterial growth and the histopathology revealed hepatic cirrhosis with fibrosis, bile duct hyperplasia, and chronic inflammation. Antibiotics were discontinued and the dog was discharged to remain on all previous medications. Colchicine was initiated orally at 0.03 mg/kg every 24 hr to help with current hepatic fibrosis and to prevent continued fibrotic development.

Fifteen months postoperatively, the dog developed unexplained hypoglycemia of 69 mg/dL (75–125 mg/dL) that was corrected after hospitalization with dextrose supplementation. At that time, the ALP was 2569 U/L and the ALT was 669 U/L and it was presumed the episode of hypoglycemia was associated with liver dysfunction. After hospitalization, his glucose normalized and he began regulating his glucose levels normally. At 17 mo postoperatively and at the time of writing, the dog is doing well with normal activity at home and no evidence of hypoglycemic episodes or recurrence of pythiosis. He is currently on the hepatoprotective supplements^d, all previous gastro-protectants, pancrelipase powder, and metoclopramide. The colchicine has since been discontinued due to cost. Because his body condition score has not improved, his diet has been adjusted for optimum weight gain. The last two rechecks of blood work revealed an ALP of 753 and 178 U/L and an ALT of 363 and 270 U/L, respectively. This is the lowest the liver values have been in over 15 mo postoperatively. At the time of writing, the dog is clinically doing well and gaining weight. Because of the improvement in condition, further diagnostics and follow up have not been pursued.

Discussion

Currently, there are few cases that describe long-term survival in dogs diagnosed with gastrointestinal pythiosis.^2,5–6 This case report describes long-term survival with radical surgical excision (Billroth II with cholecystoduodenostomy and partial pancreatectomy). Additionally, limited information is available in the literature that describes the long-term consequences of Billroth II with biliary rerouting in the dog. The prognosis has been described as poor with any disease condition requiring a Billroth II procedure.⁷ One study involving a pylorectomy with gastroduodenal anastomosis (Billroth I) involved 24 dogs of which 18 survived for 14 days and 10 of the dogs died by 3 mo with a mean survival time of 528 days. The decreased survival time was correlated with preoperative weight loss and malignant neoplasia.⁸ However, none of the cases required biliary rerouting in that study. There is limited information in the literature in terms of long-term biochemical changes associated with a Billroth II and biliary rerouting. This dog has had chronic waxing and waning elevation of his ALP and ALT for over 1 yr. The initial thought was that this was due to an acute hepatotoxicity due to drug administration. However, the true incidence of hepatotoxicity associated with itraconazole and terbinafine combined is not known. Toxicity due to these two drugs appears to be rare and one would expect the values to return to normal once discontinued. This dog's liver values remained elevated in the face of discontinuing the drugs. Laparoscopic liver biopsies revealed no growth in culture, decreasing the likelihood of ascending cholangiohepatits associated with the surgical procedure. One limitation to our culture results was that the dog was on ciprofloxacin at the time of liver biopsy, which may have influenced the negative result. Because the liver values never responded in a positive manner to any antibiotics, we suspect that the culture result was accurate despite the presence of antibiotics. Another differential to consider was potential infiltration of the liver with pythium organisms. In this particular case, we did not feel this was the cause of the liver changes seen since laparoscopic liver biopsies did not reveal evidence of the organism and the dog's pythium titers had returned to a normal level.

Due to the changes in biliary anatomy following cholecystoduodenostomy, a chronic insult to the liver may have resulted in some of the changes seen in our dog's liver values. Elevations of ALP and ALT have been documented in an abstract and case report for approximately 6 mo after cholecystoduodenostomy alone and may return to normal after 1–2 yr.^9,10 One proposed rationale behind these elevations is that they are due to duodenal secretions refluxing into the gallbladder and/or subclinical chronic periportal inflammation.^10,11 This can occur readily by the retrograde flow of secretions through a large stoma from the cholecystoduodenostomy. Overtime, this retrograde flow of secretions through the biliary system can lead to long-term chronic inflammatory insults to the liver.⁹ Since our dog's liver values appear to be decreasing, this may be what is occurring. However, it does not explain the failed ammonia tolerance test along with end-stage hepatic cirrhosis and fibrosis as was seen on our histopathology.

To achieve wide margins, the proximal duodenum had to be resected, which removed both the major and minor duodenal papilla. This results in the need for not only biliary rerouting but also the need for a partial pancreatectomy. Technically, removal of the proximal duodenum is possible, but there is the possibility of damaging the pancreatic blood supply to the right limb of the pancreas. Because of this risk, the authors chose to perform a partial pancreatectomy and remove the right limb of the pancreas. By leaving the left limb and body of the pancreas, the chances of causing a diabetes mellitus are limited. Because the pancreatic ducts were removed, lifelong pancreatic enzyme supplementation is required postoperatively. This is a major part of the long-term management of this procedure, which should be discussed with the owners prior to surgery.

As previously mentioned, the oomycete, Pythium insidiosum, is the causative agent of pythiosis and is mostly reported in dogs, horses, and humans.¹² Typically, pythiosis is not found outside of tropical or subtropical areas, but recently it has been documented in Arizona and California.³ Members in the genus Pythium are water- or soil-dwelling organisms that are important plant pathogens, while P. insidiosum is the only mammalian pathogen noted.¹ What sets this organism apart from fungi is the lack of chitin in the cell wall. Instead, pythium organisms contain predominately cellulose and β-glucan.¹³ Oomycetes differ from fungi in that ergosterol is not a principal sterol. Furthermore, members in the genus Pythium are sterol auxotrophs in which they incorporate sterols from their environment.¹

From a clinical standpoint, pythiosis presents with either a cutaneous or gastrointestinal form. In contrast to horses, dogs acquire the gastrointestinal form more commonly than the cutaneous form. It is most often diagnosed in young (<3 years old), male, large-breed dogs, with Labrador retrievers and German shepherd dogs appearing to be overrepresented.^14–16 The gastrointestinal form is characterized by severe segmental transmural thickening of the stomach; small intestine; colon; rectum; and, rarely, the esophagus or pharyngeal area.¹ Enlargement of the mesenteric lymph nodes can be a common finding, as was noted in our patient. Surprisingly, mesenteric lymphadenopathy can be observed without gastrointestinal tract lesions.¹ The most common locations pythiosis is diagnosed include the gastric outflow area, as was seen in our patient, and the ileocolic junction.¹ Inflammation is typically centered on the submucosa with variable mucosal ulcerations and occasional extension of disease through serosal surfaces.¹ Extension of disease can occur, as was documented in a prostatic abscess in a dog caused by P. insidiosum.¹⁷

Clinical signs are typically vague, but they can include vomiting, weight loss, intermittent diarrhea, and/or palpable masses in the abdomen.¹⁸ Physical examination typically reveals a patient with a thin body condition, as was noted in this case. Diagnostics, such as a complete blood count, serum biochemistry, urinalysis, and imaging, such as abdominal ultrasound, should be conducted.¹⁹

A more definitive diagnosis requires histopathology along with serologic testing. Histopathologic findings are characterized by eosinophilic granulomatous to pyogranulomatous inflammation with fibrosis. Affected tissue can also exhibit areas of necrosis surrounded by inflammatory cells. The organisms are usually found within the areas of necrosis or at the center of granulomas.²⁰ P. insidiosum can be difficult to see on hematoxylin and eosin stain, but they are more easily visualized on tissue stained with GMS. They appear as broad, rarely septate, and occasionally branching.^20,21

Culture of pythiosis can be attempted, but has historically been problematic. This is mainly because of inappropriate handling and rapid bacterial growth from specimens.¹ Because of this, other diagnostic tests have been developed, such as a polymerase chain reaction assay. Grooters et al. have developed and evaluated a polymerase-chain-reaction-based assay for P. insidiosum, which can be applied to DNA extracted from cultured isolates or from preserved infected tissue samples.²² It is a labor-intensive process, but can be used in cases where paraffin-embedded tissues are the only samples available.^22–23 To help facilitate the diagnosis of pythiosis, serologic evaluation has been developed using a soluble mycelial antigen-based ELISA. In 43 dogs in which the ELISA test was used, it was found to be highly sensitive and specific when evaluating serum.²⁴ Furthermore, the ELISA test is helpful in monitoring response to therapy, as was done in our case. In the original study by Grooters et al., a dramatic decrease in antibody levels approaching the range of healthy animals was detected 2–3 mo following surgical resection.²⁴ In animals that go on to develop clinical recurrence, the antibody levels remain high.²⁴ In our case, the first ELISA test^d was considered positive for pythium, but no predictive positivity was noted to help quantify the disease. Further ELISA tests^f were submitted at 8, 12, and 24 wk postoperatively where the predicative positivity noted was 11, 12, and 12%, respectively. In dogs with active pythiosis, the predicative positivity is 40–100%, while in healthy dogs (those never having pythium) the predicative positivity is 5–15%.²⁴ Based on a decrease in the predicative positivity and continued low values in the normal range, we can conclude that our dog was surgically cured of gastrointestinal pythiosis.

The treatment of choice for gastrointestinal pythiosis is aggressive surgical resection since this will provide the best opportunity for long-term cure. It has been suggested that achieving margins of 3–4 cm of healthy tissue should be attempted whenever possible; however, with diffuse disease this may not be achievable.¹ Schmiedt et al. was able to achieve an unusually positive response with marginal (0.6 cm) excision on the aboral margin.² In our case, we removed 6.0 cm of tissue aboral to the lesion and 5.0 cm oral to the lesion. Histopathology revealed margins free of organisms; however, inflammatory cells still came very close (1.0 mm) to the margin on the oral side.

Because of a high reoccurrence rate after surgery, medical management with itraconazole (10mg/kg per day) and terbinafine (5–10 mg/kg per day) is recommended for 2–3 mo.¹ Our dog was started on this therapy, but with the elevations in ALP and ALT, we chose to discontinue use. The terbinafine was restarted, but a true course of medical management was not achieved. Therefore, we suspect the cure was secondary to extensive surgical resection with minimal contributions from medical therapy. To monitor for recurrence, the ELISA serology test was repeated every 4 wk then once again at 6 mo with no change in the titer values.

Another potential option for a multimodal approach for pythium is immunotherapy, as was used by Schmiedt et al.² The current vaccine that is available targets both exoantigens and the cytoplasmic antigens of P. insidiosum.²⁵ The addition of the cytoplasmic antigen has increased the efficacy in horses, but unlike horses, dogs typically do not have a strong clinical response to the vaccine.²⁵ Furthermore, timing of the vaccine is important because poor efficacy is noted in chronic disease (>2 mo) with maximal benefit in acute disease.²⁵ Mendoza et al. reported only 2/6 dogs responding to immunotherapy versus 13/18 horses.²⁵ The reason for the poor response in dogs is unknown. The vaccine may cause immunomodulation and even gain efficacy by shifting the host–immune response from a T-helper type II response to a T-helper type I response.²⁵ This shift may cause a reduction in IgE titers and shifts in serum cytokines. Humans and horses respond well to the immunotherapy, and declining IgG and IgE titers have been documented as well as a corresponding reduction in clinical signs of disease.²⁵ Currently, this decline in IgE titers has not been found in dogs and, thus, may explain the decreased efficacy.

Conclusion

This is only the fourth documented case of survival in a dog with gastrointestinal pythiosis and the first describing survival in a dog with pythiosis of the pylorus and proximal duodenum. Most cases of gastrointestinal pythiosis present late in the stage of disease where disease is diffuse or not able to be resected.¹ We chose to pursue a Billroth II and partial pancreatectomy with biliary rerouting, which allowed us to remove as much tissue as possible. Achieving radical margins (3–4 cm) is one goal in the surgical treatment of pythiosis. The Billroth II, partial pancreatectomy, and biliary rerouting is well described, but the prognosis for long-term survival following surgery is grave. To our knowledge, there are limited recommendations or explanations of what is to be expected long-term if the patient survives surgery.^9,10 Based on our experience and the sparse literature available, chronic inflammatory changes to the liver may occur.^9,10 The underlying cause of our dog's liver values was not determined, but could have resulted from chronic insult to the liver as previously mentioned. In conjunction with surgical therapy, medical management is recommended for pythiosis along with frequent monitoring for reoccurrence. Because of the poor prognosis with pythiosis along with the need following lifelong medical therapy for the Billroth II, partial pancreatectomy, and biliary rerouting procedure, owners should be made well aware of the complications and postoperative care needed after this type of surgery.