Valvular endocarditis and septic thrombosis associated with a radial fracture in a red-tailed hawk (Buteo jamaicensis)

Abstract

A free-ranging adult female red-tailed hawk died suddenly after 3 weeks in rehabilitation for a radial fracture. Cause of death was septic thrombosis from a chronic bacterial valvular endocarditis, probably associated with injury at the fracture site. The challenge of clinical diagnosis of sepsis in wild birds is emphasized.

The decision to rehabilitate an injured wild animal should be based, partly, on a thorough clinical examination with the aid of basic diagnostic techniques including radiography and hematology. Even when standard practices are followed, wild animals may harbor hidden diseases that interfere with the success of rehabilitation. This case report describes the occurrence of bacterial valvular endocarditis and subsequent bacterial sepsis in a red-tailed hawk that had sustained a traumatic radial fracture approximately 7 wk prior to death; the importance of thorough evaluation of systemic health status during rehabilitation is emphasized.

Case description

A free-ranging adult female red-tailed hawk was found injured and unable to fly in Falmouth, Nova Scotia (NS), Canada, in April 2009. The bird was transported to a local veterinary hospital by a conservation officer from the Nova Scotia Department of Natural Resources. On initial examination, the bird was bright, alert, and responsive, and was in good body condition. The left wing was held in a dropped position. There was swelling at the level of the elbow, and the bird experienced pain on manipulation of this joint. There were no open wounds. A ventrodorsal radiograph was taken (Figure 1). On the basis of this radiograph and the clinical examination, a diagnosis of a closed transverse radial fracture was made. The fracture fragments remained in good apposition and alignment.

Figure 1.

Ventrodorsal radiograph of the left wing of the red-tailed hawk. a) white rectangle outlines a proximal radial fracture with evidence of healing and periosteal response; b) enlargement of the healing fracture.

The radiograph was consistent with ongoing repair of the fracture. A mixture of new bone production and lysis in the space between the fractured ends of the radius and a periosteal response were evident. These findings were consistent with bone remodeling and formation of a callus, dating the fracture to at least 1 mo prior to examination. Given that the lesion was located in close proximity to the joint, and that the fracture was already beginning to heal, internal fixation was not considered as a treatment option. Therefore, the wing was stabilized with a figure-eight bandage to immobilize the elbow. Prognosis was guarded due to the location of the lesion. However, because the bird appeared to be in good condition, it was transported to a wildlife rehabilitation center in NS for supportive care, and plans were made to repeat radiographs 3 to 4 wk later in order to assess healing.

The hawk was treated empirically with 0.2 mg of ivermectin (Ivomec; Mérial, Duluth, Georgia, USA) intramuscularly for internal parasitism, and was given 1.1 g of oral vitamin supplement (Vitahawk Maintenance, D.B. Scientific, Oakley, California, USA) daily. Attitude and appetite were closely monitored every day. The bird was eating well, was in good body condition, and attitude was considered to be normal until it died suddenly, approximately 3 wk after arrival at the rehabilitation center. The carcass was frozen and submitted to the Canadian Cooperative Wildlife Health Centre (CCWHC), at the Atlantic Veterinary College, Charlottetown, Prince Edward Island, for postmortem examination.

The bird weighed 1.21 kg and was in good body condition. There were abundant subcutaneous, coronary, and abdominal adipose tissue reserves, and well-developed pectoral musculature. The left elbow could not be fully extended, and there was marked bruising of the overlying skin. Dissection of the joint revealed a raised, rough-surfaced callus, approximately 1 cm in diameter on the proximal radius. The flexor tendon immediately ventral to the callus was firm and gritty with dark hemorrhagic streaks running longitudinally. No abnormalities were noted within the elbow joint itself. Internally, the capsular surface of the spleen had a sparsely pitted appearance. In the heart, there was a pale 3-mm diameter nodule in one of the cusps of the aortic valve. No samples were taken for bacterial culture. Samples of all organs were placed in 10% buffered formalin and prepared for routine histological examination.

Histology of the soft tissues ventral to the affected elbow revealed poorly cellular fibrous tissue, forming adhesions between the tendon structures and the overlying skin as well as the underlying periosteum. Collagen degeneration and colonies of coccoid bacteria associated with mixed inflammation were also observed in this region. These microscopic findings supported the radiographic diagnosis of repair and callus formation, and reflected chronicity of the bone lesion. The affected cusp of the aortic valve was severely thickened by an area of necrosis, inflammation, and abundant bacterial proliferation. The bacteria were coccoid, and were postulated to be of the genus Staphylococcus based on colony morphology. The surface of the affected valve was ulcerated and had prominent fibrinous attachments to the adjacent endocardium. The inflammation was primarily composed of multinucleated giant cells and degenerate heterophils, with fewer macrophages, lymphocytes, and plasma cells. Large septic thrombi with numerous bacteria were present in the coronary arteries. Several small bacterial emboli along with mononuclear inflammation were present in the myocardium itself. Septic fibrin thrombi were found in the encephalon, leptomeninges, kidneys, and gastric mucosa and submucosa (Figure 2). Areas of vascular fibrinoid necrosis and focally extensive heterophilic or granulomatous inflammation were associated with most of these lesions. In the spleen, there were several microgranulomas consisting of rings of multinucleated giant cells surrounding degenerate heterophils and colonies of bacteria as previously described. A diagnosis of bacterial aortic valvular endocarditis with associated disseminated septic vascular thrombosis was reached.

Figure 2.

Gastric artery of the red-tailed hawk with intraluminal septic fibrin thrombus. A white star indicates the bacterial embolus within the thrombus. Note the necrosis of the wall and associated inflammation extending into the adjacent connective tissue. Bar = 50 μm.

Discussion

This red-tailed hawk died as result of acute bacterial sepsis, presumably disseminated from a vegetative lesion that had originally sequestered bacteria in the aortic valve of the heart. It is likely that the bacteria gained access to the bloodstream during a traumatic event which also caused fracture of the left radius. As bacteria proliferated within the aortic valve, the valve’s capacity to sequester the infection was overcome, resulting in dissemination of small pieces of the lesion containing bacterial colonies throughout the circulation, resulting in thrombosis and vasculitis in several organs including the brain, kidney, spleen, and stomach. The sudden death of this hawk was likely caused by the acute showering of bacteria to various organs, most importantly the heart and brain, and the subsequent vascular and tissue damage.

Bacterial valvular endocarditis is not an uncommon diagnosis in domestic mammals, but the condition is not often reported in free-ranging wildlife and rarely reported in avian species. It occurs uncommonly in the commercial poultry industry, with an incidence of < 0.5% (1). Although a small number of cases have been reported in exotic avian species (1–9), bacterial endocarditis has only been reported twice in free-ranging wild birds (3,7). In 1 report, a bald eagle (Haliaeetus leucocephalus) was found dead under its nest in California. Necropsy revealed bacterial colonization of the atrio-ventricular (AV) valves, and signs of a systemic bacterial infection (3). In the other instance, 3 mallard ducks (Anas platyrhynchos) found dead in a Saskatchewan river had vegetative lesions on the mitral valve and foci of infection in several other organs (7). The most common valves to be affected by bacterial endocarditis in birds, including the present one, are the mitral and aortic valves (6,8).

Most cases of bacterial endocarditis result from exposure of the heart to blood-borne bacteria (10). The route by which these bacteria gain access to the bloodstream may vary. For example, there may be extension of bacteria from a chronic focus of infection such as hepatitis or salpingitis (10). Alternatively, the bacteremia may result from an acute traumatic event such as a skin laceration or frostbite injury (2,7). Immunosuppression leading to translocation of bacteria from the gastrointestinal tract into the blood has also been suggested as a possible mechanism of bacterial dissemination in birds (3,9). In the present case, the hawk had been a free-ranging animal, but was held in captivity for approximately 7 wk following a traumatic injury to its wing. Given that the bird was being cared for daily after its admission to the rehabilitation facility, where it suffered no obvious open wounds, and the fact that the fracture had been bandaged, it is likely that the infection occurred prior to the bird having been found, possibly as a skin wound at the time of the fracture that had healed by the time the bird was examined. The bacteria that cause endocarditis in birds include Enterobacter, Streptococcus, Staphylococcus, Pasteurella, Erysiopelothrix, and Lactobacillus which have been isolated from culture of blood and organs such as the liver and kidneys in septic birds (1–3,5–8). In the present case, the causative bacterium was probably of the genus Staphylococcus, based on morphology. Staphylococcus aureus, in particular, is a likely pathogen for the proposed route of infection. This bacterium is a common inhabitant of the skin of birds, and may cause bacteremia if the skin barrier is broken and the bacteria gain access to the bloodstream (11).

Antemortem signs of bacterial endocarditis are highly variable. In some cases, captive or pet birds have been admitted to hospital based on clinical signs such as lethargy, anorexia, weakness, and inappetence (1,6,9). However, in most cases, including the present one, diagnosis of the lesion is made on postmortem examination after sudden death of the bird (2,5,8). Successful antemortem diagnosis of this condition in a natural case has only been reported twice. These diagnoses were based on physical, clinical, and laboratory examinations including cardiac auscultation, echocardiography, and positive blood culture (1,6). Blood culture has also been successfully used in the antemortem detection of bacteremia in experimentally induced infection in chickens and turkeys (12). However, this is not an ideal modality for diagnosing sepsis due to the low concentration of pathogens that may be circulating at any given time. This means that false negative results are common, and that diagnosis usually requires several serial cultures before a positive result is obtained. Though definitive diagnosis of bacterial endocarditis may be challenging in a live bird, the use of basic diagnostics such as complete blood cell count may be very useful for alerting the clinician to the presence of an inflammatory or infectious process occurring in the bird.

Treatment of this condition in non-poultry birds has not been described. In chickens, however, treatment of bacterial endocarditis is usually unrewarding. Antibiotics rarely reach high enough concentrations within the tissue of the affected valve, and even if infection is resolved, the residual fibrosis of the valve may have severe consequences for the hemodynamics of the heart (11,12). Most reported cases of avian bacterial endocarditis, including the present one, have resulted in showering of organs with bacteria released from vegetations on the cardiac valves. The final cause of death in most birds with this lesion is septic thrombosis, shown histologically by multi-focal areas of necrosis associated with bacterial colonies and heterophilic inflammation in various organs (5,6,8–10). Less commonly, valvular insufficiency caused by the vegetation may result in lesions of congestive heart failure (10). In 1 case, an emu showed hind limb paresis due to thromboembolism of the iliac and ischiatic arteries with fibrin clots derived from the cardiac valve lesion (6).

The lack of information concerning antemortem diagnosis of bacteremia and endocarditis in avian species necessitates thoroughness in initial evaluation and treatment of injured birds. This is especially important in wildlife rehabilitation, because of the tendency of wild animals to conceal injury and illness. This case demonstrates the challenge to recognition of severe systemic illness in birds recovering from traumatic injury. Though close observation of attitude, appetite, and body condition are important in the rehabilitation period, this alone is not sufficient for monitoring overall health of the animal. Complete clinical evaluation including diagnostic imaging, complete blood cell count, and serum biochemistry analysis should be considered for all injured wild animals at the time of initial presentation, and periodically throughout the period of rehabilitation.