Necrotizing fasciitis caused by Erysipelothrix rhusiopathiae.

Abstract: A woman with diabetes mellitus type 2 had a thigh infection that drained foul-smelling pus. Necrotizing fasciitis was diagnosed surgically and histopathologically, with Erysipelothrix rhusiopathiae being the predominant organism. A pet goldfish might have been the source.

**********

Erysipelothrix rhusiopathiae is better known as a veterinary pathogen than a human pathogen. The usual human disease is a localized, indolent skin lesion (erysipeloid). Hence, the paradox of isolating E. rhusiopathiae as the predominant pathogen in proven necrotizing fasciitis prompted this report and a review of the literature that showed no previously reported cases.

Case Report

A 70-year-old black woman was admitted after 5 days of progressive swelling, pain, and a foul-smelling discharge from the upper aspect of the left thigh. She also reported fever, chills, myalgia, and nausea. There was no history of trauma. She had poorly controlled diabetes mellitus type 2, for which she injected insulin into the abdominal wall. At admission, she was afebrile and had a heart rate of 107 beats/min, a blood pressure of 145/76 mm Hg, and a respiratory rate of 20 breaths/ min. Physical examination showed tenderness over the left lower abdomen and an 8 x 12-cm area of erythema, fluctuation, and crepitus in the posteromedial aspect of the left upper thigh from which oozed a small amount of foul-smelling serosanguinenus fluid. The leukocyte count was 16,200/ml, with 56% neutrophils, 25% band forms, 11% lymphocytes, 4% monocytes, 1% eosinophils, and 3% metamyelocytes. Laboratory values were blood urea nitrogen, 40 mg/dl (normal, 7-18 mg/dl); creatinine, 1.8 mg/dl (normal, 0.6 1.0 mg/dl); glucose, 486 mg/dl (normal, 65 115 rag/all); and hemoglobin A1C, 14.8 g/dl (normal, 4.4-6.4 g/dl). A chest radiograph showed bibasilar atelectasis.

The abscess was incised and drained in the emergency room. The Gram's stain showed rare white blood cells, many Gram-positive cocci in pairs and clusters, many Gram-positive bacilli, and many Gram-negative bacilli, but the culture result was delayed until Day 7. Intravenous ampicillin/sulbactam (1.5 g every 6 hours) and tobramycin (140 mg) were administered.

On hospital Day 2, wide incision and debridement were performed and were later repeated on Days 4, 9, 10, and 12. Histopathology showed extensive necrosis of the adipose tissue and fascia. Computed tomography of the abdomen and pelvis performed on Day 3 showed emphysema in the left anterior abdominal wall extending into the pelvis and left hip. This finding was thought to be related to the surgical procedure. On Day 9, computed tomography of the thigh showed evidence of air in the soft tissues between the hamstring muscles, in the inferior third of the thigh (Fig. 1).

On Day 7, cultures of pus drained from the thigh in the emergency department on Day 1 were reported growing a large amount of E. rhusiopathiae and a small amount of Prevotella buccae. The E. rhusiopathiae was identified as a catalase-negative, Gram-positive rod and as an H2S producer on triple sugar iron agar, the latter being a differentiating characteristic from other Gram-positive bacilli. Imipenem/cilastatin (500/500 mg intravenously every 6 hours) was administered, at which time ampicillin/ sulbactam therapy was stopped. The wound culture obtained at the time of the fourth debridement (Day 10) grew only a small amount of Pseudomonas aeruginosa (sensitive to amikacin), most likely a nosocomial colonization. On Day 12, recognition of extension of the infection into tile soft tissue of the abdominal wall resulted in surgical debridement of the left lower abdominal wall.

The hospital course was complicated by development of acute respiratory distress syndrome necessitating intubation, a left middle cerebral artery infarction, and myocardial infarction. Refractory hyperglycemia was controlled with high doses of intravenous insulin. The patient was transferred to a nursing home after 2 months of hospitalization.

Discussion

Necrotizing fasciitis is an uncommon, severe infection involving the fascia. Relative sparing of the skin and underlying muscle by a subcutaneous soft tissue infection and destruction of the superficial and often of the deep fascia are its characteristics. (1) The pathogen causing necrotizing fasciitis usually gains entrance into the subcutancous space through a disruption of the overlying skin, but lymphohematogenous spread from a distant site is also an acknowledged route of infection. (2)

Two types of necrotizing fasciitis are recognized. Type 1 involves at least one anaerobic species in combination with one or more facultative anaerobic species. Type 2 is caused by Group A [beta]-hemolytic streptococci either alone or in combination with other species. (1) Our case is Type 1.

Prompt diagnosis and treatment of fasciitis are essential for a better outcome, but this goal is difficult to accomplish because early necrotizing fasciitis may be difficult to distinguish from acute cellulitis. Clues to the diagnosis include underlying diabetes mellitus, systemic toxicity, necrosis, vesicles, and gas in the soft tissue. Whenever necrotizing fasciitis is suspected, a full-thickness biopsy, performed early, has been shown to correlate with improved outcome, (2) particularly if combined with more extensive surgery. The presentation in this patient was misinterpreted as being that of an abscess. Incision and drainage were performed initially, rather than immediate debridement and biopsy. The general surgeon can sometimes establish the diagnosis of necrotizing fasciitis by observing the lack of resistance of normally adherent fascia to blunt dissection.2 Diabetic patients, especially those having uncontrolled hyperglycemia, are far more likely to have severe bacterial infections, which, in turn, are underestimated by patients and physicians alike.

The paradox of this case is the recovery from the first-day culture of the distinctly uncommon E. rhusiopathiae as the dominant organism. It is a Gram-positive aerobic or facultatively anaerobic rod, isolated first by Robert Koch from mice and later by Louis Pasteur from swine. (3) Rosenbach isolated it from a patient with localized skin lesions and coined the term erysipeloid, implying a forme fruste of erysipelas. The human disease can manifest itself as a localized skin lesion (erysipcloid), a diffuse cutaneous eruption with systemic symptoms, or bacteremia sometimes associated with endocarditis. (3) The route of transmission of E. rhusiopathiae to humans is usually by direct contact between contaminated fish or fish products, animals or animal products, or soil and a break in the skin. (3) Our patient's pet goldfish is a likely source of infection through her touching the fish or the fish tank and then scratching her inner thigh. In a study in Sweden, E. rhusiopathiae was isolated from 60% of the cod and 30% of the herring tested, (4) but no studies were performed in exotic or pet fish.

E. rhusiopathiae is susceptible to penicillins, cephalosporins, erythromycin, and clindamycin, but it is often resistant to other antibiotics including vancomycin, a drug frequently used in empiric therapy for infections with Gram-positive organisms. (5) The drug of choice for treating E. rhusiopathiae infection is penicillin, but a fluoroquinolone may be considered when [beta]-lactam administration is contraindicated. (6) In our case, antibiotic susceptibility testing was not performed, but the prompt response to ampicillin was emphasized by the absence of viable, recoverable E. rhusiopathiae organisms from the debrided tissue after 24 hours of treatment.

The microbiology laboratory runs a risk of dismissing E. rhusiopathiae as a nonpathogenic Gram-positive bacillus, such as a diphtheroid skin contaminant, or as Bacillus or Clostridium species if identification is incomplete. The isolate might even be misidentified as a Gram-positive coccus. (7) Formerly, it was known as E. insidiosa. Because vancomycin is so often used for Gram-positive infections of unknown bacteriology or when there is a suspicion of methicillin resistance or penicillin allergy, the possibility of infection with this organism should be explored when there is no response to therapy with vancomycin. Of course, if there is a suggestive exposure, it should be considered as well.

Conclusion

Although diabetes mellitus is not a risk factor for infection with E. rhusiopathiae, diabetic patients are more susceptible to necrotizing fasciitis. There has been a dramatic resurgence of necrotizing fasciitis caused by group A streptococcal disease, but other pathogens should not be ignored. (8) Clinical suspicion of necrotizing fasciitis should be high because early, correct treatment, consisting of wide excision and debridement, cell wall active antimicrobials other than vancomycin, and supportive measures are essential. (9) A diabetic with necrotizing fasciitis is not a surprise, but finding E. rhusiopathiae to be the predominant microorganism is a significant and new observation. In our patient's polymicrobie neerotizing fasciitis, E. rhusiopathiae was the predominant pathogen. We believe this is the first case to be reported.

Key Points

* Erysipelothrix rhusiopathiae can be misidentified as a Gram-positive coccus or a nonpathogenic Gram-positive bacillus, misleading therapeutic decisions.

* Necrotizing fasciitis can be difficult to distinguish from abscess or cellulitis.

* Diabetes mellitus is a predisposing factor for necrotizing fasciitis but not for E. rhusiopathiae infections.

* Early diagnosis, extensive surgical debridement, and appropriate antibiotics are paramount in managing all types of necrotizing fasciitis.

References

(1.) Swartz MN. Necrotizing fasciitis, in Mandell GL, Bennett JE, Dolin R (eds): Mandell, Douglas, and Bennett's Principles and Practice of Infectious Diseases. Philadelphia, Churchill Livingstone, 2000. ed 5, pp 1053-1054.

(2.) Green ill, Dafoe DC, Raffin TA. Necrotizing fasciitis. Chest 1996;110: 219-229.

(3.) Reboli AC, Farrar EW. Erysipelothrix rhusiopathiae, in Mandell GL, Bennett JE, Dolin R (eds): Mandell, Douglas, and Bennett's Principles and Practice of Infectious Diseases. Philadelphia, Churchill Livingstone, 2000, ed 5, pp 2226-2227.

(4.) Stenstrom IM, Norrung V, Ternstrom A, Molin G. Occurrence of different serotypes of Erysipelothrix rhusiopathiae in retail pork and fish. Acta Vet Scand 1992;33:169-173.

(5.) Reboli AC, Farrar WE. Erysipelothrix rhustopathiae: An occupational pathogen. Clin Microbiol Rev 1989;2:354-359.

(6.) Venditti M, Gelfusa V, Tarasi A, Brandimarte C, Serra P. Antimicrobial susceptibilities of Erysipelothrix rhusioputhiae. Antimicrob Agents Chemother 1990;34:2038-2040.

(7.) Schuster MG, Brennan PJ, Edelstein P. Persistent bacteremia with Erysipelothrix rhusiopathiae in a hospitalized patient. Clin Infect Dis 1993; 17:783-784.

(8.) Haywood CT, McGeer A, Low DE. Clinical experience with 20 cases of group. A streptococcus necrotizing fasciitis and myonecrosis: 1995 to 1997. Plast Reconstr Surg 1999;103;1567-1573.

(9.) Briton BD, Zibari GB, McMillan RW, Aultman DF, Dunn G, McDonald JC. Aggressive surgical management of necrotizing fasciitis serves to decrease mortality: A retrospective study. Am Surg 1998:64:397-401.

From the Department of Medicine, Huron Hospital, Cleveland, OH.

Supported by the Department of Medicine, Huron Hospital, Cleveland, OH.

Reprint requests to Ramona Simioneseu, MD, Infectious Diseases PC, 2325 Dougherty Ferry Road. Suite 206, St. Louis, MO 63122.

Accepted February 26, 2002.

Copyright [c] 2003 by The Southern Medical Association 0038-4348/03/9609-0937