Pneumatosis Coli in a Dog

Introduction

Pneumatosis is the presence of air or gas in an abnormal location in the body.¹ Pneumatosis is further described by the anatomical location (e.g., pneumatosis coli, pneumatosis intestinalis). Pneumatosis coli is a rare condition in dogs that is characterized by subserosal or submucosal cystic intramural emphysema of the colon.^1,2 Pneumatosis coli has been given many names, including pneumatosis cystoides intestinalis, intramural gas, cystic lymphopneumatosis, intestinal emphysema, intestinal gas cysts, interstitial lymphasema of the colon, bullous emphysema of the intestine, peritoneal lymphopneumatosis, and pseudolipomatosis.^3–5

Radiographic findings of pneumatosis coli are typically diagnostic and include scattered, localized, or diffuse intramural collections of radiolucent gas in the colon or rectum. The gas may be cystic, linear, or curvilinear in appearance.⁶ Colonoscopy, biopsy, and histopathology may also be used for diagnosis.⁷ Only five reports of pneumatosis coli in dogs appear to exist in the English literature.^1,2,8–10 Emphysematous gastritis has been reported in a young, collie-type dog with pancreatitis and in four dogs with gastric dilatation-volvulus.^11,12 Pneumatosis intestinalis has been reported in foals with jejunal volvulus, colonic torsion, and intestinal clostridiosis.¹³

In humans, pneumatosis coli is uncommon and associated with a variety of gastrointestinal diseases/procedures, bronchopulmonary diseases, psychiatric disorders, immune-mediated diseases, vasculitides, organ transplantation, acquired immunodeficiency syndrome, hemodialysis, cytomegalovirus infection, corticosteroid therapy, chemotherapy, nitric oxide anesthesia, and lactulose administration.^14–21 The stomach, mesentery, and omentum may rarely be involved.⁷ Despite the extensive list of associated conditions, approximately 15% of people have primary, idiopathic pneumatosis coli.²²

Pneumatosis coli has been reported previously in a 1-year-old, female Maltese.¹ Pneumatosis coli has also been reported in poodles, an English cocker spaniel, and a Yorkshire terrier.^2,8–10 Each dog had clinical signs of large intestinal disease, such as diarrhea, hematochezia, tenesmus, and ribbon-shaped stools. One dog also had signs of depression, anorexia, and vomiting.¹ The purpose of this report is to describe the presentation, diagnosis, and treatment of a dog with pneumatosis coli.

Case Report

A 17-year-old, 4.2-kg, castrated male Maltese was presented to the University of Melbourne Veterinary Clinic and Hospital, with chronic polyphagia and a 2-week history of tenesmus, diarrhea, hematochezia, weight loss, and ribbonlike feces. Prophylactic anthelmintics had not been administered for at least 6 months.

Physical examination findings included a bright demeanor, thin body condition score (3/9), pendulous abdomen, ventral abdominal alopecia, diffuse dermal thinning with prominent vasculature, blindness from bilateral mature cataracts, and bilateral anterior uveitis. Digital rectal examination induced marked tenesmus and revealed excessive soft tissue associated with the ventral aspect of the colorectal wall that extended cranially beyond the pelvic brim. The intra-pelvic soft tissue was spongy in texture. It was not possible to determine whether the protuberant soft tissue was intramural or extramural.

Hematological abnormalities included mild anemia (hematocrit 32%, reference range 37% to 55%; hemoglobin 11.3 g/dL, reference range 12.0 to 18.0 g/dL; red blood cell count 4300 × 10³ cells/μL, reference range 5500 to 8500 × 10³ cells/μL) and mild lymphopenia (600 cells/μL, reference range 1000 to 4800 cells/μL). A biochemical profile revealed mild azotemia (blood urea nitrogen 12.7 mmol/L, reference range 3.0 to 7.5; creatinine 0.10 mmol/L, reference range 0.08 to 0.17 mmol/L), mild hypercholesterolemia (9.6 mmol/L, reference range 3.6 to 9.0 mmol/L), moderately increased serum alkaline phosphatase (ALP) (657 U/L, reference range 40 to 100 U/L), and hyperlipasemia (6317 U/L, reference range <200 U/L). Urine was isosthenuric (specific gravity 1.013). The urinary cortisolto- creatinine ratio was increased (80.3, reference range <17.3), which was compatible with hyperadrenocorticism. Parasites were not detected on fecal flotation and cytology. Fecal culture yielded a mixed growth consistent with normal flora. Culture did not yield any anaerobes (e.g., Clostridia spp.) or growth of Salmonella spp. or Campylobacter spp.

Abdominal radiography [Figures 1A, 1B] revealed hepatomegaly and linear radiolucency throughout the wall of the large intestine. The latter was consistent with intramural gas and pneumatosis coli (pneumatosis cystoides intestinalis). The quantity of intramural gas varied from mild to moderate in the cecum and ascending and descending colon, to severe in the distal descending colon and rectum. The remaining abdominal organs were unremarkable. Thoracic radiography revealed diffuse mineralization of the tracheobronchial structures.

On abdominal ultrasonography, heterogeneous (mixed) echogenicity of the liver was found, and it consisted of diffuse hyperechogenicity and irregular multifocal hypoechoic regions. Equivocal bilateral adrenomegaly was present. The left and right adrenal glands were 18.7 mm × 6.9 mm and 15 mm × 7.9 mm in the long axis, respectively. The colon was difficult to examine sonographically because of the large amount of intramural and intraluminal gas. The remainder of the abdominal viscera was normal. Cytological examination of fine-needle aspirates of the liver revealed a mild, vacuolar hepatopathy.

The history of polyphagia combined with the dermatological findings, pendulous abdominal conformation, lymphopenia, hypercholesterolemia, increased urinary cortisol-to-creatinine ratio, equivocal bilateral adrenomegaly, vacuolar hepatopathy, and tracheobronchial mineralization were suggestive of hyperadrenocorticism. Further diagnostic investigations, including dynamic hypothalamic- pituitary-adrenal axis function testing, colonoscopy, colonic biopsy, histopathology, tissue culture, and thyroid function testing were recommended. The owner declined further investigations, and the dog was discharged on a highly digestible, low-residue diet^a and metronidazole^b (50 mg per os q 12 hours for 7 days).

Follow-up telephone communication revealed that the clinical signs of colorectal disease had resolved within 2 days of initiating the low-residue diet and metronidazole, and they had not recurred. The dog was euthanized by the referring veterinarian approximately 10 months after initial evaluation for chronic progressive weakness of undetermined etiology. Follow-up radiography and postmortem examination were not performed.

Discussion

The radiographic findings in this case were diagnostic of pneumatosis coli; however, the etiopathogenesis of the condition in this dog remains unknown. Histopathology from dogs with pneumatosis coli has been poorly described and may have provided an insight into the etiopathogenesis in this case. Unfortunately, colonoscopy and biopsy were not performed.

Three general mechanisms have been proposed for the etiopathogenesis of pneumatosis coli.⁷ The mechanical theory suggests that gas from ruptured pulmonary blebs or from the bowel lumen adjacent to an obstructing gastrointestinal ulcer or surgical site dissects under pressure along tissue planes to the intestinal wall, mesentery, or omentum.⁶ Not supportive of this theory, however, are the supra-atmospheric pressure of gas in the cysts and the absence of connection between the lumen and the cysts.⁷ Similarly, this theory is less likely because of the absence of a gas trail between the mediastinum and gastrointestinal tract in people with chronic obstructive pulmonary disease and frequent localization of pneumatosis coli to small areas of the colon.²³

The bacterial theory suggests that mural invasion by gasforming bacteria results in intestinal emphysema.³ This theory is supported clinically in people with pneumatosis coli and experimentally by the high hydrogen content of the cysts, the elevated breath hydrogen concentrations of affected patients, and resolution of the gas in some patients with antibiotic therapy and enemas.^24–29 Similarly, the dog in this case report responded favorably to antibiotic and dietary therapy, which suggested that enteric bacterial microflora may have played a role in the etiopathogenesis.

The third theory hypothesizes that cysts are formed by excessive amounts of gas created by intraluminal microbial fermentation of carbohydrate, with unusually high partial pressures of hydrogen.^7,23 This theory is supported by experimental evidence that helium from the colon and nitrous oxide delivered during anesthesia may diffuse through tissues to pneumatosis cysts.³⁰ Additionally, people with pneumatosis coli excrete excessive hydrogen in their breath, because they lack methanogenic and sulfate-reducing bacteria that normally use colonic hydrogen.³¹

A combined theory has also been proposed, whereby mechanical factors (e.g., abnormal mucosal permeability or integrity) together with elevated transmural pressure gradients lead to cyst initiation.¹⁴ Altered bacterial gas production from reduced numbers of hydrogen-consuming colonic bacteria was hypothesized to be important in maintaining the cysts.¹⁴

The demonstration of gas in the Peyer’s patches of immunocompromised patients with pneumatosis coli led to the hypothesis that stress or immunosuppressive drugs may deplete lymphocytes and cause loss of structural integrity of the patches, with increased bowel permeability and dissection of gas into the noninflamed intestinal wall.^32–34 Although not confirmed, the clinical findings in the case reported here were suggestive of hyperadrenocorticism that may or may not have caused immunosuppression. Further evaluation of the hypothalamic-pituitary-adrenal-axis and colonic histopathology would have been helpful to achieve a definitive diagnosis and to investigate the possible role of corticosteroids in the pathogenesis of the pneumatosis coli in this dog.

The majority of people with idiopathic pneumatosis coli do not require any specific treatment other than that for the underlying disease. The gas accumulations usually resolve spontaneously.⁷ Diarrhea associated with pneumatosis coli in a poodle resolved after 7 days of gentamicin sulfate therapy; however, radiographic resolution of the gas was not noted for several weeks following therapy.⁸ In a prior report of pneumatosis coli in a young Maltese, spontaneous resolution occurred without treatment 7 days after diagnosis.¹ Hence, without identification of an underlying cause, a contributing factor, or an associated complication, specific treatment cannot be recommended. If a heavy growth of gasforming bacteria is obtained from feces or a tissue biopsy, antibiotic therapy may be considered based on sensitivity testing. Diets with high carbohydrate content and polysaccharide-containing cathartics should be avoided, because they may increase bacterial hydrogen production and cystic hydrogen tensions.^27,29

To optimize proximal small intestinal digestion and absorption and to reduce the amount of large bowel gas formation, a low-residue diet with high digestibility and low insoluble fiber content was selected. Highly digestible, elemental diets have been shown to promote cyst resolution in people with pneumatosis coli and increased breath hydrogen excretion.²⁹

Conclusion

Pneumatosis coli was diagnosed in a Maltese with clinical signs of colorectal disease. Definitive diagnosis was based on radiographic findings, and an underlying cause was not elicited. While the long-term prognosis for dogs with pneumatosis coli is unknown and likely depends on the underlying etiology, this case suggests that the condition may resolve with dietary and antibiotic therapy. Histopathology and identification of underlying predisposing factors in dogs with pneumatosis coli may be helpful in understanding the etiopathogenesis.

Acknowledgments

The authors extend their gratitude to Geoff Griffiths for his technical assistance and to Dr. Caroline Mansfield for reviewing the manuscript.