Serratia marcescens is a common cause of nosocomial infections, such as urinary tract infections, pneumonia, and bacteremia, but it is an unusual cause of skin and soft tissue infections [1,2]. We report a case of S. marcescens necrotizing fasciitis in a bone marrow transplant patient.
A 35-year-old man with relapsed low-grade nonhodgkin’s lymphoma was admitted for autologous bone marrow transplant. He received BCNU, etoposide, cytarabine, and melphalan-conditioning therapy, followed by autologous peripheral blood stem cell transplantation. The post-transplant clinical course was complicated by delayed emesis and grade II mucositis that was treated with IV corticosteroids, analgesics, and anti-emetics. On the fourth day after transplantation, 48 hours after becoming neutropenic, the patient complained of moderate right axillary pain and became febrile to 39.1°C. On examination, the patient was in moderate distress with an area of erythema and tenderness approximately 3 cm in diameter on the medial aspect of the proximal right arm. The patient’s tunneled triple lumen right internal jugular catheter was without signs of infection. The remainder of his exam was unrevealing. Blood cultures were drawn and the patient was empirically started on vancomycin and cefepime. The area of erythema expanded at a rate of 1–2 cm/hour and the lesion developed a 1.5 cm diameter area of central necrosis with an adjacent hemorrhagic bulla. The patient became hypotensive and developed respiratory failure requiring mechanical ventilation and vasopressor support. Gentamicin, imipenem, and clindamycin were added to the antibiotic regimen and cefepime was discontinued. A CT scan of the right upper extremity showed stranding of the subcutaneous fat down to the muscle without involvement of the muscle or deep fascial compartment. The stranding extended from 5 cm above the medial epicondyle proximally to the axilla, extending posteriorly to the triceps. There was no free air or abscess. A surgery consult was obtained and the patient was taken to the operating room for emergent debridement where infection and necrosis were identified down to the level of the muscular fascia. The axilla, anterior chest and pectoralis muscle, and the biceps and triceps down to the level of the elbow were exposed before the margins were clear of infection. Histopathology of the lesion showed acute necrotizing inflammation and bacterial overgrowth with healthy tissue at the surgical margins. The wound gram stain revealed gram-negative organisms but no neutrophils. The initial blood cultures (central and peripheral) grew S. marcescens and K. pneumoniae. Three out of four wound cultures grew S. marcescens
Vancomycin and clindamycin were discontinued. The patient’s condition improved and he was successfully extubated and weaned off of pressors. Gentamicin was continued until patient had early neutrophil recovery. He completed 3 weeks of intravenous imipenem and was discharged home with a 7-day course of oral ciprofloxacin after a split thickness skin graft.
Necrotizing fasciitis (NF) is a rapidly progressive soft tissue infection associated with a high mortality (30%–60%) [3,4]. Rapid diagnosis and prompt initiation of aggressive medical and surgical management are necessary for survival [3–5]. Predisposing conditions include surgery or trauma, diabetes mellitus, peripheral vascular disease, and immunosupression. Patients typically present with fever, a soft tissue infection with pain, and systemic toxicity greater than would be expected from cellulitis alone [3–5]. Though classically it is felt NF presents with gas in the soft tissues or crepitation, a recent study noted these findings on presentation in only 39% of patients . Radiographic studies may assist in the diagnosis of NF, but NF is a clinical diagnosis and therapy should not be delayed to obtain these studies [4,5]. Computed tomography and MRI both are able to define the extent of the lesion, and CT scans may be slightly more sensitive because they are better at detecting free air . Computed tomography findings of NF and cellulitis are similar, and the former is diagnosed when free air or extensive fascial edema is present but is not excluded by the absence of these findings [6,7]. Once the diagnosis is made, initiation of broad-spectrum antibiotics and emergent wide debridement are necessary. A recent retrospective study showed patients with NF average 3.3 operative procedures (range 1–8), including procedures for covering the wound . In a prior retrospective study, 32% of patients required multiple débridements .
Recent reviews of the microbiology of NF are consistent with prior studies [3,8]. Necrotizing fasciitis can be classified into type I (polymicrobial) and type II (hemolytic streptococcal gangrene) NF . Most commonly, NF is polymicrobial with less than 10% being monomicrobial. Type I is typically a mixed aerobic-anaerobic infection. The most commonly involved aerobes include S. aureus, Enterobacteriaceae, and group A streptococcus. The most common anaerobes include Peptostreptococcus spp., Bacteroides spp., Prevotella spp., and Clostridium spp. The microbiology of type I NF usually parallels the microbiology of the local flora of the infected site; e.g., oral anaerobes are seen in NF of the head and neck, and enteric anaerobes are seen below the diaphragm. Type II NF is caused by Streptococcus pyogenes alone or in combination with S. aureus. S. pyogenes M types 1, 3, 12, and 28 are most commonly involved and they usually elaborate pyogenic exotoxin A. Type II NF often is associated with streptococcal toxic shock syndrome. Though S. aureus is typically found in association with type I or II NF, it has been described as a single etiologic pathogen of NF .
S. marcescens is a gram negative enteric organism of the Enterobacteriaceae family that is motile, has type 1 and 3 fimbriae, an extracellular lipase and proteinase, and an inducible chromosomal class I cephalosporinase; all of which may contribute to its virulence [11–14]. Initially it was felt this organism was non-pathogenic, but now there are known predisposing factors to S. marcescens infection. These include indwelling catheters (vascular or urinary), chronic debilitating disease, diabetes mellitus, corticosteroid use, prior antibiotic exposure, mechanical ventilation, surgery, and carcinoma [1,2,11]. An outbreak of S. marcescens in a bone marrow transplant unit has recently been described. A feature of the outbreak was prolonged colonization, which frequently led to invasive infection during periods of neutropenia .
S. marcescens necrotizing fasciitis is a rare clinical entity that has been reported in only five patients [16–19]. In addition, cases of cellulitis caused by S. marcescens have been reported only seven times in the literature [20–23]. All of the patients with S. marcescens necrotizing fasciitis (including the patient in this report) had a known predisposing factor to S. marcescens (end stage renal disease, diabetes mellitus, intravascular catheter, lower extremity edema, and corticosteroid use). Our patient had neutropenia as well, which is a known risk factor for gram-negative sepsis.
There are three possibilities for the primary focus of infection in this patient. The first is a polymicrobial bacteremia originating from a gastrointestinal source related to the grade II mucositis and neutropenia, with subsequent seeding of the axilla by S. marcescens. Alternatively, there may have been a direct invasion of the axilla through a microscopic break in the skin with ensuing bacteremia. S. marcescens and K. pneumoniae are known to colonize the human axilla . Our patient’s presentation included axillary pain simultaneous with the onset of fevers. Early empiric antibiotic therapy may have limited recovery of causative microbes from the wound. A third possibility would be a line infection with bacteremia and subsequent seeding of the axilla.
In conclusion, this case, as well as those reviewed in the literature, demonstrates that S. marcescens should be considered as the cause of necrotizing fasciitis in patients at risk for infections caused by S. marcescens, including bone marrow transplant patients. For patients that have compromised granulocyte function due to chemotherapy, diabetes, renal failure, or steroids, initial empiric therapy for necrotizing fasciitis should cover gram-negative bacteria that carry inducible cephalosporinases . Carbapenems or cefepime combined with clindamycin and an aminoglycoside are reasonable empiric regimens pending identification of causative microbe(s). In this clinical setting, standard antibiotic regimens for necrotizing fasciitis such as high dose penicillin G plus clindamycin with an aminoglycoside may prove to be inadequate therapy.
1. Crowder JG, Gilkey GH, White AC. Serratia marcescens Bacteremia. Arch Intern Med 1971; 128:247–53.
2. Yu VL. Serratia marcescens: Historical perspective and clinical review. N Engl J Med 1979; 300:887–93.
3. Brook I, Frazier EH. Clinical and microbiological features of necrotizing fasciitis. J Clin Microbiol 1995; 33 (9):2382–7.
4. Bilton BD, Zibari GB, McMillan RW, et al. Aggressive surgical management of necrotizing fasciitis serves to decrease mortality: a retrospective study. American Surgeon 1998; 64 (5):397–400.
5. Andreasen TJ, Green S, Childers BJ. Massive infectious soft-tissue injury: diagnosis and management of necrotizing fasciitis and purpura fulminans. Plastic and Reconstructive Surgery 2001; 107 (4):10025–35.
6. Ma LD, Frasica FJ, Bluemke DA, et al. CT and MRI evaluation of musculoskeletal infection. Critical Reviews in diagnostic Imaging 1997; 36 (4):535–68.
7. Beauchamp Jr, NJ Scott WW, Gottlieb LM, et al. CT evaluation of soft tissue and muscle infection and inflammation: a systematic compartmental approach. Skeletal Radiol 1995; 24 (5):317–24.
8. Elliot D, Kufera JA, Myers RAM. The microbiology of necrotizing soft tissue infections. American J Surg 2000; 179 (5):361–6.
9. Urschel JD. Necrotizing soft tissue infections. Postgrad Med J 1999; 75:645–49.
10. Regev A, Weinberger M, Fishman M, et al. Necrotizing fasciitis caused by Staphylococcus aureus. Eur J Clin Microbiol Infect Dis 1998; 17:101–3.
11. Wilkowske CJ, Washington II, JA Martin WJ, et al. Biochemical characteristics, antibiotic susceptibility patterns, and clinical significance. JAMA 1970; 214:2157–62.
12. Conroy MC, Bander NH, Lepow IH. Effects in the rat of intradermal injection of purified proteinases from Streptococcus and Serratia marcescens. Proc Soc Exp Biol Med 1975; 150:801–6
13. Auken HM, Pitt TL. Antibiotic resistance and putative virulence factors of Serratia marcescens with respect to O and K serotypes. J Med Microbiol 1998; 47:1105–13.
14. Franczek SP, Williams RP, Hull SI. A survey of potential virulence factors in clinical and environmental isolates of Serratia marcescens. J Med Microbiol 1986; 22:151–56.
15. Knowles S, Herra C, O’Brien A, et al. An outbreak of multiply resistant Serratia marcescens: the importance of persistent carriage. Bone Marrow Transplantation 1999; 25 (8):873–7.
16. Huang JW, Fang CT, Hung KY, et al. Necrotizing fasciitis caused by Serratia marcescens in two patients receiving corticosteroid therapy. J Formos Med Assoc 1999; 98 (12):851–4.
17. Zipper RP, Bustamante MA, Khatib R. Serratia marcescens: A single pathogen in necrotizing fasciitis. Clin Infect Dis 1996; 23:648–9.
18. Bornstein PF, Ditto AM, Noskin GA. Serratia marcescens cellulitis in a patient on hemodialysis. Am J Nephrol 1992; 12:374–6.
19. Rimailho A, Riou B, Richard G, et al. Fulminant necrotizing fasciitis and nonsteroidal anti-inflammatory drugs. J Infect Dis 1987; 155:143–5.
20. Hsieh S, Bab FE. Serratia marcescens cellulitis following an iguana bite. Clin Infect Dis 1999; 28:1181–2.
21. Bonner MJ, Meharg JG. Primary cellulitis due to Serratia marcescens. JAMA 1983; 250:2348–9.
22. Bogaert MA, Hogan DJ, Miller Jr. AE Serratia cellulitis and secondary infection of leg ulcers by Serratia. J Amer Acad Dermat 1991; 25 (3):565.
23. Brenner DE, Lookingbill DP. Serratia marcescens cellulitis. Arch Dermatol 1977; 113:1599–1600.
24. Kloos WE, Musselwhite MS. Distribution and persistence of Staphylococcus and Micrococcus species and other aerobic bacteria on human skin. Applied Microbiology 1975; 30 (3):381–5.
25. Jones, RN. Important and emerging betalactamase-mediated resistances in hospital-based pathogens: the Amp C enzymes. Diagn Microbiol Infect Dis 1998; 31 (3):461–6.