Pasteurella multocida I n f e c t i o n i n T o t a l K n e e
Arthroplasty Case Report and Literature Review Tammy
L. G u i o n , M D , a n d T h o m a s P. S c u l c o , M D
Abstract: Pasteurella mu[tocida, a small gram-negative bacterium, is part of the normal mouth flora of many animals, including domestic cats and dogs. While commonly associated with infections in animals, it is a rare cause of human disease. The majority of Pasteurella infections in humans occur with percutaneous inoculation of the organism following a bite by a cat or dog, although disease without antecedent animal exposure or with causal animal contact does occur. The spectrum of disease produced ranges from localized, including abscess, cellulitis, lymphadenopathy, and osteomyelitis, to systemic, with septicemia, septic arthritis, respiratory, and central nervous system involvement. Altered host defenses and underlying chronic disease, such as rheumatoid arthritis, corficosteroid therapy, and severe hepatic or renal disease, may predispose to more serious systemic manifestations of infection. The authors report a case of P. multocida infection in a total knee arthroplasty as a result of a dog scratch and review the literature reporting P. multocida infections in total knee arthroplasty. Key words: Pasteurella, infection, total knee arthroplasty.
She was seen 1 week later and reported some improvement in her symptoms. Moderate swelling of the left knee was noted, but anteroposterior and lateral radiographs showed the prosthetic c o m p o n e n t s to be in excellent position. A diagnosis of hemarthrosis was presumed, a conservative course of rest and ice was prescribed, and she continued to make gradual improvement. Household pets included a rabbit and two cats, with w h o m she denies contact, and a dog, with w h o m she was in constant contact. The dog frequently licked her feet, on w h i c h she often had blisters, but there had been no change in the nature of their contact until late November 1988, w h e n she received a deep 3-cm-long scratch from the dog on her left lower leg, 15 cm distal to the tibial tubercle. Her knee pain became significantly worse and within 1 week she developed an erythematous nodule lateral to the left patella that began to drain purulent material. This area was contiguous with the knee joint and 10 ml of purulent material was aspirated,
A 45-year-old w o m a n with a 20-year history of severe seropositive rheumatoid arthritis and an 8year history of corticosteroid therapy u n d e r w e n t left total knee arthroplasty in February 1986 for increasing pain and disability. She had excellent left knee function postoperatively until late September 1988, at which time she developed pain and swelling following a 3-hour period of squatting, while housepainting, with the left knee in full extension. There was no history of sharp or blunt trauma. She saw her rheumatologist, w h o attempted unsuccessfully to aspirate the joint, obtaining only a small a m o u n t of sanguineous fluid that was not sent for culture or Gram stain.
From the Department of Orthopedic Surgery, The Hospitalfor Special Surgery, New York, New York.
Reprint requests: Thomas P. Sculco, MD, The Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021.
The Journal of Arthroplasty Vol. 7 No. 2 June 1992
cultures of which revealed heavy growth of a pure isolate of Pasteurella multocida. She was admitted to the Hospital for Special Surgery and u n d e r w e n t removal of the infected total knee arthroplasty. At surgery the prosthesis and all remaining methyl methacrylate cement, as well as necrotic tissue, were removed and the sinus tract excised. Cultures were positive for P. multocida. A Broviac subclavian catheter was placed and, in that she was allergic to penicillin, a 6-week course of parenteral cefotaxime 500 mg every 6 hours was begun. The patient was reimplanted with an Insall/Burnstein II posterior stabilized cemented prosthesis after 6 weeks; tissue cultures at that time were negative. Thirty months later she again has excellent left knee function without pain, and there has been no evidence of prosthetic loosening or recurrence of infection.
Discussion History P. multocida, "killer of m a n y species," is a pathogenic as well as commensal organism for humans 3"6"~°'18 and animals. A long-recognized cause of veterinary disease, the organism was first observed during the 1877 fowl cholera epidemic in the blood of affected birds, and was identified and characterized by Pasteur in 1880. It is responsible for epidemic cholera in fowl and hemorrhagic septicemia (shipping fever) and p n e u m o n i a in cattle, horses, and other domestic and wild animals, v'8
Infection in Humans As a pathogen in humans, P. multocida is responsible for a wide variety of local and systemic disease. Most c o m m o n l y infection follows animal trauma, especially cat or dog bites or cat scratches. Pain, swelling, erythema, and gray serous or sanguinopurulent discharge from the puncture wounds appear within hours to days of a bite or scratch. ~9 This may progress to regional adenopathy, with systemic toxicity, including chills, fever, and on rare occasions septicemia. Osteomyelitis and tenosynovitis may follow cat bites, especially at sites where tendons or bones are close to the surface, because the sharp teeth deposit a direct innoculum under the periosteum, or in relatively avascular areas of synovial sheath. 23 Cases of systemic involvement, including septicemia, meningitis, and septic arthritis, generally occur in the i m m u n o c o m p r o m i s e d population: patients with underlying chronic diseases such as rheumatoid
arthritis or severe hepatic and renal disease, and those who use steroids chronically. 7 Rarely, P. multocida infections occur without obvious animal contact. These have included respiratory tract infections, septicemia, polyarthritis, endocarditis, and peritonitis. Even in the absence of a clear history of animal exposure, an animal reservoir of P. multocida is believed to be the primary source of h u m a n infection. 7 The organism remains viable in water for 7 to 25 days, and in soil for up to 3 weeks. 25 Person-to-person transmission m a y also be possible where the organism occurs commensally in h u m a n respiratory tracts. 1°'I 1,2o
Case Report We report the first case of P. multocida infection following a dog scratch. Furie et al., in 1980, stated that no infections following dog scratches have been reported, which may reflect either a significantly lower incidence of dog scratches as opposed to cat scratches, or that the habitual grooming by cats transfers bacteria from m o u t h to paw, making infection more likely following a cat scratch. 7
Review In reviewing the literature reporting P. multocida infections we find eight other cases involving prosthetic joints, all in prosthetic knees and all in w o m e n (Table 1). Ages ranged from 33 to 74 years; two cases presented with infections in bilateral knee prostheses. Five of the nine total patients had underlying rheumatoid arthritis with chronic steroid use. In three of these patients, including one with bilateral infected prostheses, salvage of the prostheses was apparently achieved with intravenous antibiotic therapy and open irrigation. Follow-up period, however, ranged from "some m o n t h s " to 1 year, and in one case was not listed. The remaining two patients required removal of the prostheses in addition to intravenous antibiotic therapy for control of the infection. The other four patients had long-standing degenerative joint disease. Two prostheses were salvaged with intravenous antibiotic therapy, open irrigation, and surgical degridement, although length of followup was 12 and 16 months, respectively, and one patient was kept on oral penicillin suppressive therapy for 1 year. Two patients, including one with bilateral infected prostheses, required removal of the prostheses as well as antibiotic therapy. In six patients, infection arose following a cat bite or scratch; in one patient infection followed licking of the surgical w o u n d by a pet dog, and in our patient
Pasteurella multocida I n f e c t i o n
infection occurred as a result of a deep scratch by a pet dog. One patient with frequent contact with a dog but no reported trauma subsequently developed P. multocida infection in one of bilateral knee prostheses. Although there is no clear explanation for the fact that all reported cases occurred in w o m e n , one possibility m a y be that more w o m e n than m e n are primary caretakers of pets and are therefore at greater risk for traumatic contact by virtue of increased exposure. Our data also show all cases to have occurred in prosthetic knees as opposed to prosthetic hips. Given the local contiguous nature of spread of this organism, one m a y postulate that local host defenses succeed in containing infection from distal extremity w o u n d s (the most c o m m o n site of animal trauma) in patients with prosthetic hips. In a review of series reporting infectious arthritis, underlying rheumatoid arthritis was present in rates as high as 1 5 - 5 0 % . 21 Indeed, in our o w n review over half of the cases occurred in patients with rheumatoid arthritis and chronic steroid use. Susceptibility to P. multocida infection in patients with rheumatoid arthritis m a y represent a systemic abnormality of granulocyte chemotaxis resulting in a predisposition to bacterial infection. 2"17 Additionally, steroid therapy m a y interfere with granulocyte function. 4'24 The presence of a prosthetic joint, with disruption of normal mechanical barriers to joint infection, m a y also be a factor in predisposition of this population to infection. Although some of the cases reviewed here apparently had resolution of infection with antibiotic therapy and open irrigation or surgical debridement, follow-up of these patients was limited in most cases to less than 1 year. All of the remaining cases required removal of the prostheses following failed attempts at salvage, in one case 19 resulting in fusion of one knee and pseudoarthrosis of the other. We believe that standard therapy for these infections should include debridement of involved tissues and removal of the prosthetic components and cement 5"13 in a two-stage procedure, 12 in addition to appropriate intravenous antibiotic therapy. In conclusion, P. multocida infection represents a slight but serious threat to patients with prosthetic joints. This risk of infection could be reduced by avoiding contact with household pets, especially cats and dogs. In the event that traumatic contact does occur, early prophylaxis with penicillin, to which the organism is quite sensitive, 21 should be considered, especially in the i m m u n o c o m p r o m i s e d patient. Physician and patient awareness of the need for caution in avoiding animal trauma and early prophylaxis and treatment of traumatic contact m a y reduce the se-
Guion and Sculco
rious consequences of prosthetic infection in this susceptible population.
References 1. Arvan GD, Goldberg V: A case report of total knee arthroplasty infected by Pasteurella multocida. Clin Orthop 132:167, 1978 2. Beeuwkes H, Bijlsma A: Reduced chemotaxis of polymorphonuclear leucocytes from patients with rheumatoid arthritis. Antonie van Leeuwenhoek 40:233, 1974 3. Bezjak V, Mimica M: Two human infections caused by Pasteurella multocida. Br Med J 2:757, 1952 4. Boggs DR, Athens JW, Cartwright GE: The effect of adrenal glucocorticosteroids upon the cellular composition of inflammatory exudates. Am J Pathol 44:763, 1964 5. Brause BD: Infected total knee replacement: diagnostic, therapeutic, and prophylactic considerations. Orthop Clin North Am 13(1):245, 1982 6. Cawson RA, Talbot JM: The occurrence of Pasteurella septica (syn. multocida) in bronchiectasis. J Clin Pathol 8:49, 1955 7. Furie RA, Cohen RP, Hartman BJ, Roberts RB: Pasteurella multocida infection: report in urban setting and review of spectrum of human disease. NY State J Med 80:1957, 1980 8. Gomez-Reino J J, Shah M, Gorevic P, Lusskin R: Pasteurella multocida arthritis. J Bone Joint Surg 62A: 1212, 1980 9. Griffin AJ, Barber HM: Joint infection by Pasteurella multocida. Lancet 1:1347, 1975 10. Henriksen SD, Jyssum K: A study of some Pasteurella strains from the human respiratory tract. Acta Pathol Microbiol Scand 51:354, 1961 11. Hubbert WT, Rosen MN: Pasteurella multocida infection in man unrelated to animal bite. Am J Public Health 60:1109, 1970 12. Insall JN, Thompson FM, Brause BD: Two-stage reimplantation for the salvage of infected total knee arthroplasty. J Bone Joint Surg 65A:1087, 1983 13. Insall JN: Infection of total knee arthroplasty, p. 669. In Insall JN (ed): Surgery of the Knee. Churchill Livingstone, New York, 1984 14. Jones FL Jr, Smull CE: Infections in man due to Pasteurella multocida. Pennsylvania Med 76:41, 64, 1973 15. Maurer KH, Hasselbacher P, Schumacher HR: Joint infection by Pasteurella multocida. Lancet 2:409, 1975 16. Mellors JW, Schoen RT: Pasteurella multocida septic arthritis. Connecticut Med 48:22 l, 1984 17. Mowat AG, Baum J: Chemotaxis of polymorphonuclear leucocytes from patients with rheumatoid arthritis. J Clin Invest 50:2541, 1971 18. Olsen AM, Needham GM: Pasteurella multocida in suppurative diseases of the respiratory tract. Am J Med Sci 224:77, 1952 i9. Orton DW, Fulcher WH: Pasteurella multocida: bilateral septic knee joint prostheses from a distant cat bite. Ann Emerg Med 13:1065, 1984
The Journal of Arthroplasty Vol. 7 No. 2 June 1992
20. Smith JE: Studies on Pasteurella septica. III. Strains from h u m a n beings. J Comp Pathol Ther 69:231, i959 21. Spagnuolo PJ: Pasteurella multocida infectious arthritis. Am J Med Sci 275:359, 1978 22. Sugarma n M, Quismorio FP, Patzakis M J: Joint infection by Pasteurella multocida (letter). Lancet 2:1267, 1975
23. Tindall JP, Harrison CM: Pasteurella multocida infections following animal injuries, especially cat bites. Arch Dermatol 105:412, 1972 24. Ward PA: The chemosuppression of chemotaxis. J Exp Med 124:209, 1966 25. Weber DJ, Wolfson JS, Swartz MN, Hooper DC: Pasteurella multocida infections: report of 34 cases and review of the literature. Medicine (Balt) 63:133, 1984