Acute bacterial endocarditis following burns: case report and review

ELSEVIER

Burns

24 (1998)’ 369-373

Acute bacterial endocarditis following burns: case report and review R. C. Cartottoa*, D. B. Macdonald”, ‘Departmeat hDepartmenr

S. M. Wasanb

of Surgery, Queen’s Univelsi~, Kingston, Ontario, Canada of Pathology, Queen’s University, Kingston, Ontario, Canada

Accepted 18 December 1997

Abstract Acute bacterial endocarditis (ABE) is a rare but deadly complication following major thermal injury. Typically the presentation is silent, with persistent fever and positive blood cultures being th[e only consistent findings. Fibrin-platlet vegetations on the valvular endocardium are thought to be seeded during bacteremic episodes. Manipulation of the burn wound is probably the most likely source of bacteremia, with Stuphylococc~s ~UI%US and Gram-negative bacilli being the most commonly implicated bacteria. In addition to causing local damage to a valve or the myocardium, infected vegetations may dislodge septic emboli systemically. Diagnosis is most easily obtained by echocardiography. Treatment usually involves prolonged administration of intravenous antibiotics. In rare circumstances, valvular resection and replacement may be indicated. 0 1998 Eisevier Science Ltd for ISBI. All rights reserved. Keywords:

--

Burn infection; MRI; Septicaemia; Endocarditis

1. Case report

A previously healthy 39-year-old male was admitted with a self-inflicted 40 per cent body surface area (BSA) full thickness burn injury involving his anterior trunk, both Aanks and both upper extremeties. He was resuscitated using the Parkland Formula and enteral feeding was initiated at post-burn day (PBD) 2. He remained hae.modynamically stable, afebrile, and had a white blood cell count (WBC) in the normal range over the next 4 days. Gram-stain of routine swabs fro;m the burn wounds on day 3 showed Gram-positive bacilli resembling coryneform bacilli. On PBD 5, a fever of 38.5”C developed. On the same day, the patient was taken to the operating room for tangential excision of burns on his anterior trunk. Preoperatively a femoral venous line and a femoral arterial line were inserted. The excision of the abdomen and chest burns comprised approximately 15 per cent BSA. The excised areas were closed with split thickness skin grafts harvested from the buttocks and back. *Corresponding

author: Hotel Dieu Hospital, 166 Brock Street,

Kingston, Ontario. Canada fax: 001 613 544 9897.

K7L

5G2.

0305-41’79/98/$19.00 0 1998 Elsevier PII: SO305-4179(98)00027-8

Tel.:

Science

001 613 544 3400 x2300;

Ltd for ISBI.

All rights

reserved.

The patient tolerated surgery well but his fever persisted intermittently over the next 7 days, ranging from 37 to 39°C. Over the same time the WBC gradually rose from 4.9 x lo9 to 22.9 x IO” 1-l. During this time the patient had no cough, sputum production or dysuria. The heart sounds were normal during this period with no audible murmurs. There was no sign of infection at the venous and arterial line sites. On PBD 7, two days after its insertion, the femoral arterial line was removed and on the same day the femoral vein line was switched to the opposite groin. Chest X-rays showed subsegmental atelectasis post-operatively which resolved over 3 days. No further cultures of blood, urine or sputum were obtained. The patient went on to have full take of his skin grafts and: aside from his in.termittent fever, the patient remained stable and showed good progress. On PBD 12 the patient was taken to the operating room for further excision and grafting. Tangential excision was performed on the dorsum of the right hand, forearm and upper arm to the axilla, comprising approximately 7 per cent BSA. The wounds were closed with split thickness skin grafts harvested from the right anterior thigh. Post-operatively the patient was afebrile, haemodynamically stable, and had a WBC of 17.2 x lo9 1-l. Twelve hours post-operatively, the

R. C. Cartotto et d.JBums 24 (1998) 369-373

370

patient’s condition deteriorated rapidly. Severe shortness Iof breath and hypotension were followed by cardiac arrest with irreversible electromechanical dissociation. Resuscitation attempts were unsuccessful over 4.5 min of attempted treatment. The resuscitation was ended and the patient expired. At autopsy a 2 cm x3 cm x 1.5 cm vegetation was found on the anterior mitral valve leaflet, extending onto the mural endocardium of the left ventricle nea.r its outflow tract (Figs 1 and 2). The myocardium showed no areas of infarction but there was a 2 cm vegetation or! the !eft ventricle mural endocardial surface (Fig. 2). The coronary vessels were normal. The vegetation was estimated to be 7 to 10 days old based on the presence of proliferating fibroblasts. Cultures of the vegetation grew Staphylococcus aweus and E. coli. Evidence of recurrent embolic infarcts, of varying age, were found in the kidney (Fig. 3) and spleen (Fig. 4). There was no evidence of acute left heart failure. ‘The likely cause of death was endotoxic shock arising from Gram-negative septicemia.

Invariably, ABE appears only after significant thermal injury. Of the cases of ABE recorded in the literature, the average associated body surface area (BSA) burn was between 40 and 90 per cent [l-6].

2. Discussion

Acute bacterial endocarditis (ABE) is a rare but deadly complication of burn injury. The largest experience with ABE after burns comes from the Brooke Army Medical Centre Burn Unit [I., 21. The incidence of ABE between 1953 and 1970, during which time over 3000 burr: victims were admitted, was 0.6 per cent [I]. Between 1969 and 1974, ABE occurred in only 1.3 per cent of 1731 burn patients [2]. In the experience of the Brooke Institute, the mortahty rate from ABE was 95 per cent (1, 21.

Fig. 3. Vegetation

rhe chordae

tendinae

the left atrium, and papillary muscle..

spanning

mitral

valve

and left ventricle.

Fig. 1. Large endocardium.

mitral

The vegetation

involves

valve

vegetation

the anterior

mitral

extending

leailet

into

the

with extension

mural

into

R. C. Cartotto

Fig. 3. Classicai

wedge-shaped

embolic

infarction

et ui.lBums

Oi! sple en she )wing embolic

infarcts

369-373

371

in kidney

ABE is best characterized as a ‘silent’ source of sepsis in the burn patient [2]. The diagnosis is usually made post-mortem. In the reports from the Brooke Institute, correct antemortem diagnosis was made in 28 per cent [l] and 13 per cent [2] of cases of ABE. The only consistent clinical features are persistent fever, and positive blood cultures [l-3, 5, 61. A cardiac murmur is a helpful dia.gnostic finding when present, but absence of a murmur is usually the case. Munster et al. [l] and Baskin et al. [2] reported cardiac murmurs in only 25 and 9 per cent of cases, respectively. Other less common clinical features include pericardial rub, acute heart failure, arrhythmia, cerebral infarction, purpura, splinter haemorrhages and splenomegaly [l-.5].

Set

24 (1998)

of varying

ages (older

The pathogenesis of ABE remains incompletely understood. In the vast majority of cases there is no history of rheumatic or other valvular disease and the vegetations appear to form on healthy normal valves 11-61. In Baskin’s series [2], only one patient had a history of rheumatic fever as a child, but at autopsy, no residual rheumatic cardiac lesions were found. The patient reported in our paper had no clinical or pathological evidence of pre-existing valvular disease. Presumably, thrombotic fibrin--platelet vegetations (non-bacterial thrombotic endocarditis) develop on endocardium as a result of the transient hypercoaguability seen after thermal injury [7, S]. Additionally, the presence of an indwelling catheter on or near the

on left side, more

recent

on right

side).

372

R. C. Cartotto

et al.JBurns

endocardium can easily cause Local trauma, contributing to the formation of a vegetation [2, 7, 9, lo]. In fact, ABE has been reported in burn patients with pulmonary artery (Swan Ganz) catheters by both Koshal et al. 13-1and Ehrie et al. [7]. Once a thrombotic vegetation forms, it may be seeded during any transient bacteremia, resulting in the development of bacterial endocarditis. Surgical and non-surgical manipulation of the burn wound can result in bacteremia [ll, 121 and this is believed to be the most likely source of bacteremia related to the development of ABE [l]. Of interest-, at thle Brooke Institute, the incidence of cardiac infection declined from 3.3 to 1.1 per cent after the introduction of topical burn wound antimicrobials in 1964 [I]. The presumed source of sepsis was the burn wound in 64 per cent of cases in the earlier period, compared to 45 per cent i.n the later. Other possible source of bacteremia would include suppurative intravenous thrombophlebitis, catheters, pneumonia, urinary tract infections, and translocation from the gut. The ‘silent’ nature of ABE makes the exact temporal relationship between surgical manipulation of the burn wound and the onset of ABE difficult, if not impossible, to define. In several case reports [3-51 ABE was detected after multiple excision and grafting procedures. In the case reported here, the patient developed a fever following the first excision and grafting procedure. His fever persisted intermittently for over a week until he died. At autopsy, the vegetation was estimated to be at least a week old, suggesting that the first surgical procedure was the likely source of the bacteremia that ultimately caused the endocarditis. Bacterial endocarditis is most commonly seen on the mitral and aortic valves, but vegetations can develop on any of the valves as well as on the mural endocardium [l-4, 71. In Baskin’s [2] series, vegetations formed on the right side of the heart in 46 per cent of cases, on the left side in 27 per cent, and on both sides in 27 per cent. Bacterial vegetations on the right side may send infected emboli to the lungs, resulting in pulmonary sepsis [2, 61, while lesions on the left side may release septic emboli to the coronary arteries, brain, spleen and kidneys (21. In our patient, at autopsy, there was clear evidence of recurrent emboli from the cardiac vegetation resulting in multiple splenic and renal infarctions (Figs 3 and 4). In addition to the embolic potential, local invasion of the vegetation may occur, resulting in endocardial spread, myocardial abscesses, pericarditis and even ventricular septal defects. Historically, the bacteriology of ABE has usually been determined from blood cultures, since the infected valves were not routinely cultured at autopsy. Baskin ef aE. [2] reported that the most common organism grown in the blood from patients with ABE was Staphylococcus azmw either by itself (41 per cent)

24 (1998)

369-373

or in combination with Gram-negative organisms (36 per cent). Gram-negative organisms alone were grown from the blood in the remaining 23 per cent of patients. Correlation with Gram stain of the infected valves was not uniformly consistent however. In our laboratory we routinely culture the vegetations at autopsy, in order to clarify the relationship between blood cultures and the organisms causing ABE. Some uncommon organisms implicated in ABE following burn injury include fungi [l], Acinetobacfer calcoaceticus [S] and Streptococcus faecalis [6]. Diagnosis of ABE requires a high index of suspicion. Persistent fever and or consistently positive blood cultures, in the absence of a more obvious source, should raise the possibility of ABE. Frequent careful auscultation of the heart should be carried out and echocardiography should be arranged with any suspicion of ABE [13]. In our case, insufficient suspicion during the febrile period from post-burn day 5-12, along with hampered cardiac auscultation due to the chest dressings, probably contributed to the missed diagnosis. Treatment consists of initiation of appropriate intravenous antibiotics which should be continued for 6 weeks [13]. Additionally, potential sources of the bacteremia which caused the ABE .must be treated. These might include suppurative thrombophlebitis, infected central lines or cannulation sites, and foci of suppuration in the eschar. In selected cases, excision and replacement of the infected valve have been undertaken [2, 3, 51. This high risk procedure is usually reserved for patients with severe valvular insufficiency, usually with persistently positive bloo’cl cultures despite maximum antibiotic therapy. In summary, ABE ‘is a rare but lethal complication following major burn injury. The presentation is usually silent and most often the only clinical findings are persistent fever and or positive blood cultures. A high index of suspicion for ABE should be maintained at all times, especially in the situation whLere fever and or positive blood cultures persist despite no obvious source of infection. Echocardiography will usually provide a definitive diagnosis. While ABE is associated with a high mortality rate, successful treatment may be realized with the early institution and continuation of intravenous antibiotics.

References [l]

Munster AM, DiVincenti FC, Foley FD, Pruitt BA. Cardiac infection burns. Am J Surg 1971;122:524-7. [2] Baskin TW, Rosenthal A, Pruitt BA. Acute bacterial endocarditis: a silent source of sepsis in the burn patient. Ann Surg 1975;184:618-21. [3] Koshal A, Wynne J, Westerman R, Cohn IX. Bilateral infective endocarditis. Clin Cardiol 1981;4:243-6.

R. C. Cartotto et al.lBums 24 (1998) 369-373 [4] Gowda TKV, Sriprasad S, Korath MP, Jagadeesank ?. Right ventricular mural infective endocarditis in a patient with burns. J Assoc Phys India 1992;40:52-4. [S] Cumberland NS, Jones KP. Hospital acquired native valve endocarditis caused by acinobacter calcoaceticus and treated with imipenen/cilastin. J R Army Med Corps 1987;133:156-8. [6j Turn&y PA, Yardley JH. Clinicopathological conference case presentation. Johns Hopkins Med J 1967;2:107-18. [7] Ehrie M; Morgan AP, Moore FD, O’Connor NE. Endocarditis with the indwelling balloon tipped pulmonary artery catheter in burn patients. J Trauma 1978;18:664-6. [7] Moncr-ief 5.4. Burns N Engl J Med 1973;288: 444-454. [9] Greene JF, Cummings KC. Aseptic thrombotic endocarditis

373

vegetations: a complication of indwelling pulmonary artery catheters. J.A.M.A. 1973;225:1525-6. IlO] Pace NL, Horton WL. Indwelling pulmonary artery catheters: the relationship to a septic thrombotic endocardial vegetation. J.A.M.A. 1975;233:893-4. [ll] Beard CH, Ribeiro CD, Jones DM. The bacteremia associated with burns surgery. Br J Surg 1975;62:638-41. [12] Sasaki TM, Welch GW, Herndon DN, Kaplen JZ, Lindberg RB, Pruitt BA. Burn wound manipulation induced bacteremia. J Trauma 1979;19: 46-8. [13] Heggers J, Linares HA, Edgar P, Villarreal C, Herndon DN. Treatment of infection in burns. In: Herndon DN. editor. Total Burn Care. 1st edition. London: Saunders: 1996. 113.