Intra-Abdominal Actinomycetoma in a Cat

Abstract

A 5-year-old, female Ragdoll cat was diagnosed with an intra-abdominal mycetoma involving the ileocaecal region. Diagnosis was obtained via histopathological examination following surgical resection of the mass and an ileocolic anastomosis. The initial surgery was complicated by lymphangiectasia, chylous abdominal effusion and mild bacterial leakage from the anastomosis site. A second, exploratory laparotomy was performed to augment the anastomosis with serosal patching and omentalisation and to investigate a cystic structure observed on follow-up abdominal ultrasound. Initial amoxycillin clavulanate (Clavulox; Pfizer Animal Health) therapy was ineffective, but clindamycin (Antirobe; Pfizer Animal Health) proved successful in resolving the infection. Abdominal actinomycetoma in the cat may be an under-diagnosed condition due to its close resemblance to neoplastic disease. Standard diagnostic and therapeutic regimens are commonly ineffective in Actinomyces species infections. Surgical resection along with adjunctive, long-term, selective antimicrobial therapy is effective and prognosis is good for localised lesions.

Actinomycetales are Gram-positive, branching, filamentous bacteria.^1,2 They grow in anaerobic/microaerophilic (Actinomyces species) or aerobic (Nocardia species, Dermatophilus congolensis, Streptomyces griseus) conditions. All genera may be responsible for chronic, pyogranulomatous infections that are frequently unresponsive to routine antimicrobial therapy. These organisms may also cause mycetoma formation, a granulomatous tumour-like lesion of the skin, tissues or occasionally bone. Mycetomas are characterised by a triad of clinical signs; chronic induration and swelling, draining sinuses, and discharge of granules or grains. These grains are colonies or aggregates of organisms that form distinctive granules in infected tissue. ^1–3

In cats individual case reports of pyothorax, cutaneous and/or subcutaneous infections caused by Actinomyces species have been reported previously.^1,2,4 Typically other feline oral cavity bacteria are also present and single organism infections are rarely reported. ⁵ Only a single report describing an intraperitoneal mass caused by actinomyces in this species could be found in the literature. ⁶ In contrast, intra-abdominal actinomycetomas are more frequently reported in humans, predominantly affecting the ileocaecal junction. ^7–11

An unusual presentation of an abdominal actinomycetoma in a cat is described in this report. The similarity of clinical presentation to intra-abdominal neoplasia and the difficulties in obtaining a definitive diagnosis are highlighted. Treatment of feline actinomycotic infections is also discussed.

A 5-year-old female neutered Ragdoll was presented for inappetence, weight loss, post-prandial vomiting and general malaise over the previous 3–4 weeks.

On examination the cat was bright, alert and responsive. Overall body condition was lean with a body condition score of 3.5–4/9. Thoracic and cardiovascular examination was unremarkable. Peripheral lymph nodes were normal in size and consistency. A large (estimated 10 cm), irregular, firm, mobile mass in the mid-abdomen was palpated. No other abnormalities were detected.

Significant abnormalities detected on haematological and serum biochemical analysis included a marked hyperproteinaemia (152 g/l; reference range (RR) 58–84 g/l) associated with a marked increase in globulins (135 g/l; RR 28–48 g/l) and a mild decrease in albumin (17 g/l; RR 23–35 g/l). Serum calcium concentration was decreased following correction for albumin (2.12 mmol/l; RR 2.2–2.66 mmol/l). Creatinine was normal (106 μmol/l; RR 78–120 μmol/l), urea was mildly decreased (4.3 mmol/l; RR 5.87–10.6 mmol/l), and urine specific gravity was 1.030. Additionally, feline immunodeficiency virus and feline leukaemia virus antigen ELISA testing was negative.

Thoracic radiographs were unremarkable. Lateral and ventrodorsal abdominal views confirmed a mid-abdominal, soft tissue mass that displaced intestines caudally. Kidneys were normal in size and position and hepatic margins were normal.

Abdominal ultrasound characterised a soft tissue mass, 5–7 cm in diameter with multiple fluid-filled cavities, and a surrounding solid, irregular, soft tissue margin, resembling a capsule. The mass was not associated with any major organ but appeared closely adherent to a segment of bowel. Creamy fluid was aspirated with ultrasound guidance from a cavity within the mass, and submitted for cytological assessment, culture and sensitivity testing.

Cytological evaluation revealed degenerate neutrophils and a moderate number of mononuclear cells which were predominantly macrophages and plasma cells. Occasional eosinophils, spindle cells and giant cells were observed. No bacteria were observed. These findings were consistent with a focus of pyogranulomatous inflammation.

An exploratory laparotomy was conducted to obtain a definitive diagnosis, resect abnormal tissue, and assess the overall health of the abdominal structures. Peri-operative intravenous fluid therapy with Hartmanns (10 ml/kg/h) and empirical antibiotic therapy with amoxicillin clavulanate (Clavulox; Pfizer Animal Health) (20 mg/kg IM q12 h) was commenced.

A standard ventral midline laparotomy was performed. A firm, pale, encapsulated mass encompassed the ileocaecal region and associated mesentery with mild enlargement of adjacent lymph nodes. A wedge biopsy of an avascular region of the mass was taken for immediate impression smear cytology to rule out lymphoma. Cytological findings were consistent with pre-operative analysis. The ileocaecal region was resected (en bloc) along with the mass and submitted for histopathological analysis. A side to side ileocolic anastomosis was performed and omentalised prior to routine closure.

Within 24 h a decreased packed cell volume (PCV) (13%) and low albumin (17 g/l) were observed on haematological monitoring. A transfusion with type A blood from a donor cat was administered following blood typing. Ultrasonographic examination of the abdomen demonstrated a small volume effusion which was aspirated and revealed a sterile, serosanguinous transudate. Haemodilution and gastrointestinal tract blood losses were assumed to be the likely cause of the decrease in PCV. Improvement of demeanour was noted following the blood transfusion.

Over the following 24 h, the cat remained bright and alert, but continued to have a poor voluntary appetite and was intermittently pyrexic (38.6–40.4°C). Oral syringe feeding of Hill's Science prescription a/d was well tolerated. Moderate small bowel diarrhoea was present.

Fluid originally obtained from the mass by ultrasound guided aspiration was inoculated directly on to Columbia sheep blood agar (Oxoid Australia) and anaerobic agar (Oxoid Australia) and incubated for 7 days. In addition, the fluid was inoculated into a blood culture bottle (Signal Blood Culture System; Oxoid Australia) before subculture both aerobically and anaerobically. Escherichia coli was the only isolate grown from the blood culture bottle.

Histopathological results revealed a large inflammatory mass attached to the colon and its serosa by dense, fibrous granulation tissue. The mass consisted of intersecting bands of fibrovascular tissue with varying sized necrotic regions. Neutrophils, macrophages and epithelioid cells enclosed some necrotic regions with mononuclear cells and granulation tissue peripherally. Specific staining for routine bacteria, mycobacteria or fungal elements detected mixed Gram-positive bacilli and filamentous, branching microorganisms on examined sections ( ^{Fig 1} ). Filamentous microorganisms were occasionally found in small, dense tangled clusters and were non-acid-fast using the Fite-Faraco modification of the Ziehl–Neelsen technique. There was a mild increase in mononuclear inflammatory cells and submucosal lymphoid nodules in the adjacent colon. The associated paracolonic lymph node showed moderately dilated sinuses containing proteinaceous fluid and inflammatory cells, with an increase in fibrosis of the nodal parenchyma. The negative reaction to acid-fast stain and mixed population of bacteria present was more suggestive of an Actinomyces species than Nocardia species, though neither organism was isolated. A working diagnosis of an actinomycetoma was established. Amoxicillin clavulanate (Clavulox; Pfizer Animal Health) was discontinued and clindamycin (Antirobe; Pfizer Animal Health) commenced (20 mg/kg PO q12 h followed by a bolus of 5 ml water).

Fig 1.

Gram stain of a section of the excised inflammatory mass demonstrating necrotic debris and mixed bacteria including Gram-positive, filamentous, branching rods.

Despite clinical improvement progressive abdominal distension was noted. Ultrasound examination demonstrated a moderate abdominal effusion together with a large cystic structure. Fluid analysis demonstrated a septic chylous effusion characterised by a triglyceride level 32-fold higher than that in blood (blood 0.46 mmol/l; abdominal fluid 14.75 mmol/l), elevated protein and total nucleated cell count (46 g/l; RR<20 g/l and 16.3×10⁹/l; RR<5×10⁹/l, respectively) and a mixed population of pleomorphic rods and cocci both intra- and extra-cellularly. A repeat exploratory laparotomy was conducted to locate and eliminate the source of bacterial contamination, to assess the viability of the previous ileocolic anastomosis and adjacent intestine, and to investigate the source of chylous effusion.

Approximately 150 ml of chylous effusion was drained upon entering the abdomen. The previous anastomosis site was mildly inflamed and considered a potential site of slow bacterial leakage. A serosal patch was sutured over this site using adjacent loops of small intestine, and omentalised. A 5 cm cystic structure containing milky effusion, similar in appearance to the chylous abdominal effusion was visualised within the mesentery of the small intestine. Lymphangiectasia and lymphadema of the small intestine were noted along with numerous white plaques over the surface of the intestine and mesentery. Disruption of lymphatics following the previous surgery was felt to be the cause of these changes and the chylous effusion. The cyst was opened, drained and omentalised. The cat recovered more rapidly and was discharged 4 days later on clindamycin (Antirobe; Pfizer Animal Health) (20 mg/kg PO q12 h) for 4 weeks. Health checks 2 and 4 weeks postoperatively revealed stable clinical parameters and no obvious recurrence of disease.

Differentiating between lesions caused by Actinomyces species as opposed to Nocardia species can be challenging, but is important due to prognosis and treatment differences. Cytological and histopathological evaluation in combination with culture and sensitivity testing is recommended. For both organisms, histological sections stained with Gram stain reveal Gram-positive, branching, filamentous rods. Acid-fast staining (using the Fite-Faraco modification of the Ziehl–Neelsen technique) can be of assistance as Nocardia species are more commonly acid-fast whilst Actinomyces species and other anaerobic actinomycetes are not. However, culture of the organism is still required for definitive diagnosis.^1,3 Isolation of Actinomyces species may take several weeks and usually requires strictly anaerobic conditions. Cultures may be negative for Actinomyces species, but frequently contain other bacteria. In contrast, Nocardia species are strictly aerobic, are usually cultured and are frequently the sole isolate obtained from the sample. ⁵ In the case described, definitive diagnosis via bacterial culture was not obtained. However, the presence of long, non-acid-fast, filamentous bacteria that were present as microscopic dense mats made a diagnosis of infection with an Actinomyces species more likely.

Actinomycetes are endogenous saprophytes on mucous membranes of the oral cavity and gastrointestinal tract of mammals.^1,12 Actinomyces viscosus, Actinomyces odontolyticus, Actinomyces israelii, Actinomyces naeslundii and Actinomyces bovis have been cultured from dental plaque of dogs and Actinomyces hordeovulneris and Actinomyces denticolens have been cultured from normal feline gingiva. ¹³ These opportunistic pathogens depend on mechanical disruption of normal skin and mucosal barriers by disease, perforation or trauma. ⁵ Lesions are not considered contagious.^1,3,14 Bite wounds and penetrating foreign bodies are the most common routes of infection in both dogs and cats.^1,14 Various species have been isolated from infections in dogs and cats with A viscosus and A hordeovulneris being the most common species isolated from dogs and A viscosus, Actinomyces meyeri, Actinomyces pyogenes and Actinomyces bowdenii isolated from cats. ^4,12,15–17 Most isolates are not identified beyond genus and are often found in combination with other oral bacteria.^1,14

Actinomycosis most commonly occurs in young to middle aged, large breed, male dogs.^7,11,18 Roaming male cats are more likely to be inoculated with Actinomyces through fighting, although there is not sufficient evidence to confirm this. ¹ No predisposing factors were identified in the current case.

Clinical forms in dogs involve the cervicofacial region, thorax, abdomen, retroperitoneal space and subcutaneous tissues, whilst in the cat predominantly involve the thorax and subcutaneous bite wounds.^1,3,5,19 Bone and the central nervous system may rarely be involved.^3,14 Infections of subcutaneous tissues develop following bite wounds, perforation of the oropharynx by a foreign body, or chronic gingivitis–periodontitis, whereas pulmonary infections develop after aspiration of oropharyngeal material or inhalation of grass awns.^1,3 Abdominal actinomycosis, as observed in the current case, is rarely reported in the cat though is well documented in humans and accounts for approximately 20% of human actinomycosis with the appendix and ileocaecal region the most commonly involved sites (65%).^6,7,9 This form usually develops following penetration of the gastrointestinal mucosa by plant material or any other foreign body, which carries the bacteria into the mucosal wall and/or abdominal cavity. Whilst a foreign body was not identified in the current case, perforation of the gastrointestinal by a foreign body, with subsequent formation of an inflammatory pyogranulomatous mass is a reasonable scenario.

Other organisms may cause pseudomycetomas, which appear similar to the lesions observed in this case, and must, therefore, be differentiated. Methicillin-resistant staphylococci have been reported to cause cervicofacial, extremity and abdominal mesenteric lesions which are characterised by pyogranulomatous inflammation and yellow-white granules. ⁵ Dermatophytic pseudomycetomas have also been observed in Persian cats most frequently as cutaneous nodules, with one reported case of an intra-abdominal mass. ²⁰ These organisms may be readily differentiated from actinomycetes upon culture, cytology or histology.^5,20

In cats, infections with Actinomyces species are almost always mixed, with two to five other oral cavity or intestinal pathogens frequently isolated.^5,16,17 Associated bacteria help to produce, and maintain, an anaerobic tissue environment preferred by the Actinomyces species and can also increase pathogenicity. In the current case, E coli was isolated from the fluid aspirated from the original mass and may represent a co-colonising bacteria. Actinomyces species with fimbriae can bind to specific cell surface receptors on other bacteria and this co-aggregation inhibits neutrophil phagocytosis of, and bactericidal activity on the bacterial complex. ²¹ Actinomyces species induce neutrophil chemotaxis, activate macrophages and stimulate B-lymphocyte hyperplasia. ⁵ This results in the characteristic lesions of dense mats of Actinomyces species and associated organisms surrounded by neutrophils, macrophages and plasma cells. ⁵ Granulomatous inflammation may surround the abscess and form a mycetoma, though in cats this is not always present. ³ In addition, abscesses frequently have a malodorous, yellow serosanguinous exudate containing clumps of bacteria resembling grains, commonly referred to as ‘sulphur granules’ which may be visible macroscopically. ¹

Whilst positive culture is required for definitive diagnosis, cultures are frequently positive only for associated bacteria as was the case in the current report.^3,4 Actinomyces species are either facultatively anaerobes or strict anaerobes.^1,12 Although they are usually cultured within 5–7 days, growth may require up to 4 weeks. Associated bacteria may overgrow the more slowly growing Actinomyces species during this time. ⁵ Resident flora of the oral cavity or intestinal tract are commonly isolated including E coli, Bacteroides species, Corynebacterium species, Eubacterium species, Fusobacterium species, Pasteurella multocida, Peptococcus species, Staphylococcus aureus, and Streptococcus species. ⁵ Diagnosis of actinomycosis is, therefore, often based on cytological or histological identification of the organism in specimens from animals with appropriate clinical signs.

Diagnosis of abdominal actinomycosis is difficult and less than 10% of human cases of abdominal actinomycosis are diagnosed pre-operatively. ⁷ Abdominal ultrasound and radiography may be augmented with computed tomography analysis to characterise the extent of the lesion and degree of infiltration. However, exploratory laparotomy is usually required to differentiate from foreign body perforation or neoplasia or in emergencies such as obstructive gastrointestinal lesions or severe peritonitis. In addition, surgery to excise solitary lesions, assist debridement or remove inflammatory lesions, may reduce the duration of antimicrobial therapy.^1,3,8

If abdominal actinomycosis is suspected based on cytological evaluation, initial antimicrobial therapy should be directed at a variety of facultatively anaerobic and strictly anaerobic bacteria as the majority of these infections are mixed. Selection is aided by definitive diagnosis of Actinomyces species infection after appropriate culture and sensitivity testing. However, discrepancies between effectiveness of drugs in vitro on the antibiogram and effectiveness in vivo have been documented, ²² and therefore, empirical therapy based on therapeutic response may be required.

Long-term, high-dose penicillin is the treatment of choice for actinomycosis. ³ No in vitro resistance to easily attainable serum concentrations of penicillin, nor acquired resistance in vivo has been documented. ⁵ However, dense granulomatous tissue, the production of calcified sulphur granules and L-phase or cell-wall variants of some species (eg, A hordeovulneris) may limit the activity of penicillins in the lesions, reducing efficacy. ²³ Treatment for weeks to months beyond the resolution of measurable disease is recommended. ⁵ Cats with pyothorax or subcutaneous abscesses may be cured with drainage and a shorter duration of antibiotic treatment. Sulphonamides have been recommended for the treatment of L-phase variants, but are not effective in mixed anaerobic infections and may not be tolerated in cats. Alternate antibiotics include those with a broad spectrum of activity against anaerobes and L-phase variants and include clindamycin, erythromycin, ampicillin, doxycycline, imipenem, first generation cephalosporins, ceftriaxone and chloramphenicol.^3,5,23 Clindamycin (Antirobe; Pfizer Animal Health) was clinically efficacious in the current case, though the positive influence of surgical excision of the lesion must also be taken into account. Aminoglycosides, metronidazole, oral cephalexin, oxacillin and dicloxacillin are contraindicated. ³

Poor response may be attributable to poor surgical drainage and failure to eliminate associated bacteria. ⁵ Although surgery has a controversial role in the treatment of actinomycosis, drainage of abscesses and effusions (thoracic, abdominal and pericardial) should always be used as an adjunct to antimicrobial treatment. In dogs with solitary masses involving the thoracic and abdominal walls, radical surgical excision can have a high cure rate, although repeat surgeries may be necessary. ⁵ In the current case, surgical excision together with appropriate antimicrobial therapy was successful. Adjunctive therapy, may have included a fat free diet to reduce intestinal lymphatic flow and lymphatic loss, particularly if the chylous effusion or protein loss through lymphangiectasia had persisted. ²⁴

Prognosis for complete resolution is good for localised actinomycosis, and in the dog cure rates of >90% have been reported in dogs post-surgery and long-term antimicrobial therapy. ⁵ Relapse rates of 15–45% are reported for more disseminated infections and some patients may require life-long maintenance of antimicrobial therapy.^3,4

In conclusion, actinomycosis remains a diagnostic challenge due to the similarity of lesions with neoplastic or other infectious diseases and specific requirements for bacterial isolation. Abdominal actinomycosis is a rare form of the disease in cats, but may be under-diagnosed due to inappropriate culture techniques or failure to include histopathology on diagnostic testing. In addition, abdominal actinomycosis may mimic more common neoplastic conditions such as lymphoma resulting in a poor prognosis and possible euthanasia without further diagnostic tests being undertaken. Therefore, the incidence of abdominal actinomycosis in cats may be higher and requires further investigation. Surgery should be reserved for complicated or obstructive cases, or where neoplasia cannot be otherwise ruled out. Medical therapy, using an appropriate choice and duration of antimicrobials, is effective and prognosis is good for cases with localised disease.