Chromobacterium violaceum Infection in Two Dogs
Chromobacterium violaceum is a saprophyte of soil and water in tropical and subtropical environments that is associated with rare but highly fatal infections in animals and humans. Systemic infection was diagnosed in two critically ill dogs from Florida. Fever was absent in both dogs. Both dogs were treated surgically and provided with intensive care, but only one survived. The identification of characteristic, violet-pigmented bacterial colonies on routine microbial cultures should alert microbiologists and clinicians to the likelihood of this dangerous pathogen. Because of the rapidly progressive nature of this infection, empirical antibiotic administration with fluoroquinolones should be employed pending susceptibility testing.
Introduction
Chromobacterium violaceum (C. violaceum) is generally recognized as a nonpathogenic saprophyte of soil and water in tropical and subtropical environments of the world.1,2 Infections of animals and humans are rare but are often fatal. Production of the pigment violacein gives C. violaceum colonies a characteristic violet color. This pigment and several other microbial metabolic products are under investigation for a variety of biomedical and industrial uses, including antibiotic, antitumoral, and environmental detoxification activities.3,4 However, not all strains of C. violaceum produce the pigment, which can complicate prompt diagnosis.5
Although infections of humans and animals appear to be rare, outcomes are frequently fatal. The first case in humans was reported in Malaysia in 1927, and the medical literature contains reports of approximately 150 human cases since then.6 In the United States, C. violaceum is found in the southeastern region, with most human cases arising from Florida.7–11 Most human cases are reported in children and immunosuppressed patients. Infections develop following exposure of damaged skin to water or soil or ingestion of contaminated water.7–12 Clinical signs appear 1 day to 2 months after exposure.10,12 The initial presentation in humans often includes local cellulitis, lymphadenitis, cutaneous lesions, pain at the site of inoculation, vomiting, fever, abdominal pain, and diarrhea.7,8,10,12–15 Septic shock develops rapidly, and involvement of liver, lungs, and spleen is common.7,9–12,15 Osteomyelitis, cerebral edema and infarction, and septic myocarditis have also been reported.7,9,15 In one review, only four of 12 patients survived, with survival depending on early detection and treatment with antibiotics.10 One patient experienced reinfection from insect bites received while fishing in a river 18 months after the original infection.14
The pathogenicity of C. violaceum in animals was first described in 1905 when it was identified as the cause of fatal infections in water buffaloes in the Philippines.16 The infection has since been reported in gibbons and a sun bear in Malaysia, a dog in Australia, a Barbary sheep in Spain, and pigs in Taiwan and Malaysia.1,17–20 The only previously reported animal cases in the United States involve an Assam macaque and a red panda from Florida and pigs in Georgia.21–23 In most animal cases, the route of infection is speculative. In the case of the gibbons, skin lacerations were noted in four out of 10 animals, and all drank from streams from which C. violaceum was cultured.1 Viable C. violaceum has been recovered from ixodid ticks, but whether ticks can transmit the bacterium to animals is unknown.24 In the only dog previously reported with C. violaceum, diagnosis was made on postmortem examination.18
This report describes life-threatening infections with C. violaceum in two dogs from Florida, suggesting that the infection may be more common than previously recognized.
Case Reports
Case No. 1
A 9-year-old, castrated male beagle weighing 17.3 kg was examined by the referring veterinarian for a distended abdomen and cough. The dog lived primarily indoors but was allowed to roam over 10 acres on the west coast of Florida. Congestive heart failure was suspected, and the dog was treated with furosemide, enalapril, and spironolactone. One liter of serosanguinous fluid was removed by abdominocentesis 22 days after the initial presentation. Thirty-eight days after initial presentation, the dog was reexamined by the referring veterinarian for signs of orthopnea, dyspnea, trembling, and distended abdomen. Additional serosanguineous fluid was removed by abdominocentesis, and thoracic radiographs revealed cardiomegaly and mild pleural effusion.
The dog was presented to the College of Veterinary Medicine 39 days after its first examination for abdominal distension with decreased appetite, decreased exercise tolerance, and increased thirst. On physical examination, the dog was quiet, alert, responsive, and mildly dehydrated (approximately 5%). Heart rate (108 beats per minute [bpm]), pulse quality, and temperature (37.7°C) were normal. Tachypnea (120 breaths per minute) with increased inspiratory effort, increased bronchovesicular sounds, bilateral muffled lung sounds ventrally, and muffled right heart sounds were ausculted, but no murmur was evident. The epaxial muscles were atrophied, and the abdomen was distended. Rectal examination revealed dark, tarry feces.
Results of a complete blood count (CBC) indicated a mild neutrophilia (13,000/μL; reference range 3000 to 11,000/μL), mild lymphopenia (910/μL; reference range 1000 to 4800/μL), thrombocytosis (544 K/L; reference range 160 to 430 K/μL), and a normal hematocrit (44%; reference range 37% to 54%). Biochemical abnormalities included increased alkaline phosphatase (306 U/L; reference range 6 to 126 U/L), hyperphosphatemia (6.2 mg/dL; reference range 2.1 to 6.0 mg/dL), increased creatinine (2.2 mg/dL; reference range 0.7 to 1.6 mg/dL), and increased urea nitrogen (83 mg/dL; reference range 6 to 27 mg/dL). Results of a urinalysis revealed a specific gravity of 1.014 and pH of 5.0. Test results for Dirofilaria immitus antigen, Ehrlichia canis antibodies, Rickettsia rickettsii antibodies, and antinuclear antibodies were negative.
Abdominal radiographs revealed decreased serosal detail and moderate generalized hepatomegaly. Ultrasonography of the abdomen confirmed peritoneal effusion and showed mild dilatation of the caudal vena cava and hepatic veins, and a moderately hyperechoic liver. A total of 700 mL of a serosanguineous fluid was removed by abdominocentesis. The cytological interpretation was a nonseptic exudate with a packed cell volume (PCV) of 5%, total solids of 5.0 g/dL, 550,000 red blood cells (RBCs)/μL, and 59,280 white blood cells (WBCs)/μL. The WBC population included nondegenerate neutrophils (81%), small and intermediate lymphocytes and lymphoblasts (3%), and reactive mononuclear phagocytes (16%). No infectious agents or neoplastic cells were identified. The abdominal fluid was submitted for aerobic and anaerobic bacterial culture.
Thoracic radiographs were consistent with a moderate volume of pleural effusion and moderate generalized cardiomegaly. An echocardiogram revealed a small amount of pericardial effusion without evidence of cardiac tamponade. The pericardium appeared thickened. Mild mitral regurgitation was noted. Electrocardiography was normal. Right-sided congestive heart failure was diagnosed, but there was no echocardiographic evidence of primary cardiac disease. Intracardiac pressures measured by right heart catheterization were consistent with effusive constrictive pericarditis (mean right atrial pressure 14 mm Hg, normal 0 to 5 mm Hg; right ventricular diastolic pressure 9 to 15 mm Hg, normal 0 to 5 mm Hg; right ventricular systolic pressure 25 mm Hg, normal 20 to 25 mm Hg). Thoracocentesis resulted in removal of 140 mL of a serosanguineous fluid with cytological characteristics of a nonseptic exudate with a PCV of 4%, total solids of 4.3 g/dL, 660,000 RBCs/μL, and 38,860 WBCs/μL. The WBC population was similar to the abdominal fluid. Pericardiocentesis was performed for both diagnostic and therapeutic purposes, and 11 mL of a dark red fluid was removed. The PCV of the fluid was 45%, and total solids were 7.3 g/dL. Cytological evaluation of the pericardial fluid was consistent with blood contamination at the time of collection or with peracute pericardial hemorrhage. Aerobic and anaerobic bacterial cultures of the pericardial effusion failed to yield any microbial growth.
Culture of the abdominal effusion showed a heavy growth of a violet-pigmented, aerobic, nonenteric, gram-negative rod [Figure 1]. Empirical treatment for septic peritonitis with ampicillina (22 mg/kg intravenously [IV] q 8 hours), enrofloxacinb (5 mg/kg IV q 12 hours), and famotidinec (0.6 mg/kg IV q 12 hours) was initiated. The furosemide, spironolactone, and enalapril were continued. Although exploratory laparotomy for septic peritonitis was recommended, the owners elected a conservative medical approach. The organism recovered from the abdominal effusion was identified as C. violaceum 3 days after admission. The isolate was susceptible to aminoglycosides, fluoroquinolones, chloramphenicol, tetracycline, imipenem/cilastatin, nitrofurantoin, and trimethoprim/sulfamethoxazole. The isolate was resistant to cephalosporins and penicillins, including those potentiated with beta-lactamase inhibitors, with the exception of piperacillin/tazobactam.
On day 4, the dog was less responsive, normothermic (37.7°C), tachypneic, and tachycardic (168 bpm). Systolic blood pressure was normal (110 mm Hg; reference range 80 to 160 mm Hg). Treatment with ampicillin was discontinued when anaerobic cultures were reported as negative. Diuretics and enalapril were discontinued, as there was no evidence of volume overload. Imipenem/cilastatind (5 mg/kg IV q 8 hours) was initiated in the event that the in vivo response to enrofloxacin therapy was not predicted by the in vitro susceptibility findings. Thoracic ultrasonography revealed scant amounts of pleural and pericardial fluid. Abdominal ultrasonography revealed hepatic and pancreatic vascular distension consistent with chronic passive congestion, and a mildly enlarged lymph node in the medial iliac region. A liver aspirate was performed and interpreted as hepatocellular hyperplasia with moderate vacuolar degeneration. Because of the dog’s deterioration, an abdominal exploratory laparotomy was performed to search for a focus of infection. Anesthesia was complicated by hypotension (systolic blood pressure of 45 mm Hg; reference range 80 to 160 mm Hg) and hypoxemia (PaO2 of 64 mm Hg on 100% oxygen; reference range 400 to 500 mm Hg). No lesions other than the previously recognized hepatomegaly and a generalized cyanosis of the viscera were observed. A total of 600 mL of peritoneal fluid was removed, the peritoneum was lavaged, and a closed-suction peritoneal drain was placed. Aerobic and anaerobic bacterial cultures of the peritoneal fluid taken during surgery were negative.
The dog’s condition remained poor postoperatively with tachycardia (150 bpm), tachypnea, normotension, and normothermia. Thoracic ultrasonography on day 5 revealed a moderate amount of pleural fluid, a mild increase in pericardial fluid, and a mobile mass in the right atrium, which was suspected to be a thrombus. Pleural and peritoneal effusions continued to accumulate, and paracentesis of both cavities was performed on days 5 and 6. Fluid analyses were consistent with neutrophilic exudates, but no organisms were identified, and bacterial cultures were negative. A subtotal pericardectomy was performed via a right lateral thoracotomy on day 7. Hemothorax developed postoperatively and was accompanied by hypotension resistant to blood transfusion and IV fluid support. Cardiac arrest ensued and was unresponsive to open-chest cardiopulmonary resuscitation.
Necropsy confirmed constrictive, effusive pericarditis characterized by histiocytic and lymphoplasmacytic inflammation and fibrosis of the pericardium. Pulmonary, pericardial, epicardial, and myocardial fibrosis and chronic hepatic congestion were also noted. No mass was observed in the right atrium, and the source of the postoperative hemothorax was not identified. Aerobic culture of a peritoneal swab produced scant growth of Acinetobacter spp., which was considered a contaminant. Aerobic cultures of the thoracic cavity and pericardium were negative.
Case No. 2
A 9-month-old, male Labrador retriever weighing 28.9 kg underwent routine surgery for castration and repair of an inguinal hernia by the referring veterinarian. The dog was discharged the same day with ivermectin to treat facial demodicosis. The dog lived primarily indoors but was walked off-leash and allowed to swim at local parks, including the day prior to surgery. The dog was reexamined the day after surgery, because the owner perceived the dog to be uncomfortable; deracoxib was prescribed. The next day, the dog was afebrile but had severe edema and pain in the inguinal, scrotal, and right hind-limb areas. The dog was treated with buprenorphine, fentanyl, amoxicillin/clavulanic acid, cefazolin, prednisolone sodium succinate, diphen-hydramine, famotodine, and IV fluids.
On the fourth postoperative day, the dog was referred to the College of Veterinary Medicine. On presentation, the dog was depressed, responsive, and very painful. Physical examination revealed a normal temperature (38.5°C), pulse quality, and systolic blood pressure (100 mm Hg; reference range 80 to 160 mm Hg). The dog was 7% dehydrated and tachycardic (150 bpm). Marked cutaneous edema and ecchymosis were present along the right side of the ventrum, extending from the axilla to the scrotum and involving the right hind limb. Multiple 1- to 2-mm vesicles were present on the surface of the affected skin.
A CBC indicated neutropenia (2100/μL; reference range 3000 to 11,500/μL) with a left shift (bands 1500/μL; reference range 0 to 300/μL, metamyelocytes 280/μL, myelocytes 170/μL), and lymphopenia (670/μL; reference range 1000 to 4800/μL) resulting in panleukopenia (WBCs 5580/μL; reference range 6000 to 17,000/μL). Marked toxic changes were present in most neutrophils. A coagulation profile revealed thrombocytopenia (47,000/μL; reference range 160,000 to 430,000/μL), normal prothrombin time (7.8 seconds; control 7.9 seconds), and slightly increased partial thromboplastin time (12.3 seconds; control 8.9 seconds). Fibrin split products were normal (<5 μg/mL; reference range <5 μg/mL), and D-dimers were mildly elevated (250 to 500 ng/mL; reference range <250 ng/mL). A biochemical profile revealed a low anion gap (-4; reference range 11 to 26), hyperphosphatemia (6.5 mg/dL; reference range 1.8 to 5.9 mg/dL), hyperbilirubinemia (0.5 mg/dL; reference range 0 to 0.4 mg/dL), hypoproteinemia (4.6 g/dL; reference range 5.8 to 7.8 g/dL), hypoalbuminemia (2.0 g/dL; reference range 2.8 to 3.8 g/dL), increased alkaline phosphatase (310 U/L; reference range 14 to 126 U/L), and increased aspartate transaminase (116 U/L; reference range 11 to 44 U/L). Serology results for Dirofilaria immitis antigen, Ehrlichia canis, and Borrelia burgdorferi antibodies were negative.e
Thoracic radiography revealed a moderate amount of soft-tissue swelling along the ventral thoracic and cranial abdominal wall, more severely on the right side. On ultrasonographic examination, diffuse infiltrates of fluid with punctate, hyperechoic foci were seen within the scrotum and dissecting throughout the soft tissues of the inguinal and umbilical regions. The intraabdominal organs were within normal limits. Upon cytological examination, the fluid aspirated from the soft tissues was characterized as a hemorrhagic effusion with degenerate neutrophils without visible organisms. An aspirate of one of the cutaneous vesicles revealed highly karyolytic neutrophils and reactive macrophages. Small bacterial rods were observed in the cytoplasm of some cells. Three venous blood samples were collected for aerobic and anaerobic blood cultures.
The dog was treated with hydromorphonef (0.1 mg/kg IV q 3 hours) and a ketamineg continuous-rate infusion (0.3 mg/kg per hour IV) for pain. An IV crystalloid solution, enrofloxacin (5 mg/kg IV q 12 hours), and ampicillin/sul-bactamh (15 mg/kg IV q 8 hours) were also started. The thrombocytopenia and mild partial thromboplastin time elevation suggested the possibility of early disseminated intravascular coagulation, so one unit of fresh-frozen plasma (250 mL IV) was administered for clotting factor replacement.
On day 2 of hospitalization, the edema had spread down the right forelimb, and the dog remained very painful. Platelet numbers (<5000/μL; reference range 163,000 to 430,000/μL) and albumin levels (1.7 g/dL; reference range 2.8 to 3.8 g/dL) worsened, and the partial thromboplastin time was increased (21.8 seconds; control 14 seconds). Another unit of fresh-frozen plasma (250 mL) was administered, and heparini (100 units/kg subcutaneously [SC] q 8 hours) was initiated to treat possible consumptive coagulopathy.
Blood cultures were positive for C. violaceum that was susceptible to aminoglycosides, fluoroquinolones, chloramphenicol, tetracycline, imipenem/cilastatin, nitrofurantoin, and trimethoprim/sulfamethoxazole. The isolate was resistant to cephalosporins and penicillins, including those potentiated with beta-lactamase inhibitors. Treatment with enrofloxacin was continued for the C. violaceum. Ampicillin was continued to protect against possible anaerobes or other organisms that might not have grown on routine cultures.
On day 4, a 2 cm-diameter section of skin sloughed, leaving an open wound in the inguinal area. Cutaneous edema progressed to involve the ventral neck and the entire right forelimb. The ecchymotic areas on the ventral abdomen turned from red to brown/black [Figure 2]. The dog was febrile (42.5°C), tachycardic (150 bpm), nor-motensive (systolic blood pressure 110 mm Hg; reference range 80 to 160 mm Hg), and had pale mucous membranes. Platelet numbers (<5000/μL; reference range 163,000 to 430,000/μL), PCV (28%; reference range 37% to 54%), total protein (4.4 g/dL; reference range 5.8 to 7.8 g/dL), and albumin (1.8 g/dL; reference range 2.8 to 3.8 g/dL) were decreased, and the partial thromboplastin time remained elevated (17.3 seconds; control 12.9 seconds). Following administration of cross-matched, fresh, whole blood (500 mL), the dog was anesthetized for debridement of the right ventrum from the axilla to the proximal right hind leg and scrotum. A 7-cm × 20-cm area of skin and associated SC tissue and fat was resected [Figure 3]. The wound was left open, and continuous drainage was applied using a sponge-vacuum suction device.25,26 Medical management with hydromorphone, a ketamine continuous-rate infusion, crystalloid solutions, enrofloxacin, ampicillin, and heparin was continued (as before) with the addition of fresh-frozen plasma (500 mL IV). Morphinej (0.3 mg/kg per hour IV) was added 5 hours after surgery for additional pain control.
Recovery was uneventful. The dog became more responsive and less painful with a normal heart rate, respiratory rate, blood pressure, and temperature postoperatively. Three days after surgery, the WBC population was normal (6460/μL; reference range 6000 to 17,000/μL) but was characterized by neutropenia (1800/μL; reference range 3000 to 11,500/μL) with a left shift (bands 1800/μL; reference range 0 to 300/μL, metamyelocytes 170/μL) and lymphopenia (850/μL; reference range 1000 to 4800/μL). A coagulation profile revealed thrombocytopenia (35,000/μL; reference range 160,000 to 430,000/μL), normal prothrombin time (6.2 seconds; control 7.0 seconds), and slightly prolonged partial thromboplastin time (13 seconds; control 10.4 seconds). A biochemical profile revealed improvement in some parameters. Changes included hyperphosphatemia (7.3 mg/dL; reference range 1.8 to 5.9 mg/dL), hyperbilirubinemia (0.5 mg/dL; reference range 0 to 0.4 mg/dL), hypoproteinemia (4.9 g/dL; reference range 5.8 to 7.8 g/dL), hypoalbuminemia (2.0 g/dL; reference range 2.8 to 3.8 g/dL), increased alkaline phosphatase (286 U/L; reference range 14 to 126 U/L), and increased aspartate transaminase (83 U/L; reference range 11 to 44 U/L).
Four additional surgeries were performed to further debride the granulating inguinal wound during the following 11 days. The dog was discharged on the 16th day of hospitalization with wet-to-dry bandages in place. During the following 2 weeks, bandages were changed every other day, and enrofloxacin (5.5 mg/kg per os [PO] q 12 hours), amoxicillin/clavulanic acidk (15 mg/kg PO q 12 hours), and carprofenl (2 mg/kg PO q 12 hours) were given. After 2 weeks, a 3 × 2-cm area on the right medial thigh was left to close by second intention. By 112 days after the initial admission to the College of Veterinary Medicine, the wound had completely healed.
Discussion
In both of these cases, C. violaceum was isolated by primary culture on 5% sheep blood agar and MacConkey agar at 37°C. After 24 hours of incubation, there were 1 to 2 mm-diameter, shiny, smooth, violet-pigmented, convex colonies on both media [Figure 1]. Pigmentation was also evident on MacConkey agar. Microscopically, the colonies consisted of gram-negative, rod-shaped, motile bacteria. They were not acid-fast, had no definite capsule or metachromatic granules, and were nonsporulating as defined by Sneath.26 The following were biochemical reactions: reduction of nitrate to nitrite, L-tryptophane (indole) negative, glucose fermentation positive, arginine dihydrolase positive, urease negative, esculin hydrolysis negative, gelatin hydrolysis positive, 4-nitrophenyl-BD-galactopyranoside negative, and oxidase positive.m The following sugars were assimilated: glucose, N-acetyl-glucosamine, potassium gluconate, capric acid, and malic acid. D-Mannose was assimilated in case no. 2 but not in case no. 1. In both cases, the isolates recovered were susceptible to aminoglycosides, fluoroquinolones, sulfonamides, tetracycline, and chloramphenicol, but they were resistant to penicillins and cephalosporins, which is common for C. violaceum.27–29
In a previous report, experimental intraperitoneal inoculation of the organism in mice, guinea pigs, and rabbits resulted in infection and death. Oral inoculation was less pathogenic.2 Experimental oral inoculation in two pigs resulted in no clinical signs, except reduced weight gain. At 12 weeks postinoculation, the pigs were sacrificed and visceral abscesses were identified.2
Historically, in most infected animals, the time between the onset of clinical signs and death ranged from 1 to 10 days.1,16–23 Clinical signs include vomiting, diarrhea, anorexia, depression, dyspnea, and fever. In one case, a macaque showed no clinical signs and died within 4 days of being transferred from Florida.22 Chromobacterium violaceum infection in animals often results in hepatic abscesses.1,18,22,30 Acute pleuropneumonia has been reported in sheep and pigs.19,21 In the only previously reported canine infection, multifocal pinpoint abscesses in the lungs and liver were present.18 The jejunal and ileal mucosa was extensively replaced by granulation tissue containing irregular foci of hyperplastic crypt epithelium resembling enteritis due to parvovirus. Although parvovirus was not isolated, the authors hypothesized that the C. violaceum septicemia occurred after a primary infection with parvovirus.
The source of infections in the dogs reported here was not confirmed. Case no. 1 had no history of skin lesions, but the dog had free access to rural property in Florida. Invasive procedures (e.g., abdominocentesis) were performed twice within the 2 weeks prior to referral. The fluid was not examined cytologically or cultured, so it was unknown whether the infection was already present at the time of the procedures or resulted from contamination during the procedures. The dog’s primary condition was likely constrictive, effusive pericarditis that was secondarily complicated by C. violaceum infection.
Infection in case no. 2 followed a routine surgical procedure in a dog that swam in a local park the night prior to surgery. The dog presented with severe ventral edema, ecchymosis, and pain arising from the site of a recent surgical incision. Chromobacterium violaceum was cultured from the blood, and a suspected consumptive coagulopathy arose secondary to septicemia.
Conclusion
Two dogs in Florida were diagnosed with C. violaceum infections, and isolates from both cases were resistant to several commonly used antibiotics. Both dogs were normothermic at the time of referral and were initially treated medically with antibiotics. Surgical debridement was delayed for several days, which may have impeded the response to treatment. The identification of characteristic, violet-pigmented bacterial colonies on routine microbial cultures should alert microbiologists and clinicians to the likelihood of this dangerous infection. Because of the rapidly progressive nature of C. violaceum infection, early surgical intervention accompanied by empirical administration of fluoroquinolones should be employed pending susceptibility testing.
Ampicillin for Injection, USP; American Pharmaceutical Partners, Inc., Schaumburg, IL 60173
Baytril; Bayer, Shawnee Mission, KS 66201
Famotidine Injection; Bedford Laboratories, Bedford, OH 44146
Primaxin IV; Merck, Inc., Whitehouse Station, NJ 08889
Canine SNAP 3Dx Test; IDEXX Laboratories, Westbrook, ME 04093
Hydromorphone HCl; Baxter Healthcare Corporation, Deerfield, IL 60015
VetaKet; Lloyd Laboratories, Shenandoah, IO 51601
Unasyn; Pfizer, Inc., New York, NY 10017
Heparin Sodium Injection, USP; Baxter Healthcare Corporation, Deerfield, IL 60015
Morphine Sulfate Injection, USP; Baxter Healthcare Corporation, Deerfield, IL 60015
Clavamox; Pfizer Animal Health, Exton, PA 19341
Rimadyl; Pfizer Animal Health, Exton, PA 19341
API 20NE; bioMerieux, Inc., Durham, NC 27712
Citation: Journal of the American Animal Hospital Association 42, 2; 10.5326/0420154
Citation: Journal of the American Animal Hospital Association 42, 2; 10.5326/0420154
Citation: Journal of the American Animal Hospital Association 42, 2; 10.5326/0420154
Chromobacterium violaceum (from case no. 1) develops distinctive violet colonies on most culture media (5% Columbia blood agar).
Extensive necrotizing fasciitis and necrosis of inguinal skin associated with infection with Chromobacterium violaceum in a 9-month-old, male Labrador retriever (case no. 2).
Contributor Notes
