Sepsis in a renal transplant recipient due to Citrobacter braakii.

Abstract: Cellulitis is usually caused by organisms such as [beta]-hemolytic streptococci and Staphylococcus aureus. Citrobacter are gram-negative bacilli that can cause opportunistic infections in immunocompromised hosts. They are rarely implicated in skin or soft tissue infections. The genus Citrobacter has been respeciated according to genetic relatedness. Citrobacter braakii refers to the genomospecies 6 of the Citrobacter freundii complex. There are no detailed studies of infections caused by the newly formed specific genetic species. We report a case of C. braakii infection in a renal transplant patient receiving immunosuppressive therapy. The patient's lower extremity cellulitis did not respond to conventional antibiotic therapy. Blood cultures grew C. braakii. Sensitivity studies and treatment with appropriate antibiotics resulted in prompt recovery, Immunosuppressive therapy in renal transplant recipients predisposes to infection by unusual pathogens, and this should be suspected when lack of a clinical response to conventional antibiotics is observed. We believe this is the first reported case of C. braakii cellulitis and bacteremia in a renal transplant recipient.

**********

Key Points

* Citrobacter is a genus of gram-negative bacilli that can cause opportunistic infections in immunocompromised hosts.

* The genus Citrobacter has recently been respeciated according genetic relatedness and C. braakii refers to the genomospecies 6 of the Citrobacter freundii complex. Citrobacter has rarely been implicated in skin and soft tissue infections.

* Blood cultures should be obtained in all immunocompromised patients with cellulitis not responding to conventional antibiotics.

Case Report

A 48-year-old white woman presented with fever and shaking chills for 1 day. She also had had progressively worsening pain, swelling, and erythema of the left leg for 1 week and was unable to walk. She had been treated with oral cephalexin for the previous 5 days, with no clinical improvement. This drug was discontinued because of a skin rash that was attributed to cephalosporin sensitivity. There was no history of trauma, gardening, or exposure to pets. Her medical history consisted of insulin-dependent diabetes mellitus since the age of 8 years, hypertension, coronary artery disease, and end-stage renal failure due to diabetic nephropathy. She had received a cadaveric renal transplant in 1993 and since then bad received immunosuppressive therapy with cyclosporine and prednisone. She was allergic to penicillins and sulfa drugs.

On physical examination, the patient appeared ill. She had a temperature of 38.7[degrees]C, a pulse rate of 110 beats/rain, a respiratory rate of 22 breaths/min, and a blood pressure of 112/68 mm Hg. Examination of the skin showed an area of erythema and swelling over the anterior and lateral aspects of the left leg extending from below the knee to the ankle. Two skin breaks were visible, and the area was warm and tender to touch. Inguinal lymph nodes on the left side were found to be mildly enlarged and tender. Examination of the cardiovascular system was unremarkable, and the chest was clear to auscultation. The renal allograft was in the right lower quadrant of the abdomen, and it was not tender on palpation. The remainder of the abdominal examination was unremarkable. There were no neurologic deficits.

Laboratory values were as follows: hemoglobin, 9.8 g/dl; hematocrit, 28.8%; white blood cell count, 12,000/ mm3 (93% neutrophils, 4.6% lymphocytes, 2% monocytes); platelet count, 201,000/mm3; blood urea nitrogen, 78 mg/dl; and creatinine, 1.60 mg/dl. Serum electrolyte values and results of liver function tests were within normal limits. Chest radiography and Doppler study of the lower extremities, performed to exclude deep vein thrombosis, revealed no abnormality.

Empiric therapy with levofloxacin and clindamycin was started after blood samples were sent for microbiologic study. The blood was collected in blood culture bottles and plated on MacConkey and sheep blood agar. Identification was done by Vitek kit (automated identification with gram-negative identification card). Both sets of blood cultures drawn from two sites grew C. braakii. Urine culture showed no bacterial growth. After the results of the blood cultures, the antibiotic regimen was changed to a combination of aztreonam and levofloxacin. The cellulitis improved and the fever resolved within 72 hours. Antibiotic treatment was continued for 2 weeks. Repeat blood cultures were found to be negative, and the patient was discharged.

Cellulitis is a commonly seen clinical syndrome usually caused by [beta]-hemolytic streptococci and Staphylococcus aureus. (1) The risk factors associated with cellulitis include systemic conditions such as diabetes mellitus, alcohol abuse, and anatomically localized factors such as athlete's foot, (2) trauma, and venous and lymphatic compromise. Coagulase-negative staphylococci, gram-negative bacteria such as Escherichia coli, Serratia, Proteus, Klebsiella species and other Enterobacteriaceae, (3) Histoplasma, (4) Cryptococcus, (5) cutaneous mycobacteria, (6,7) and Legionella micdadei (8) have been reported to cause cellulitis in immunocompromised hosts, including renal transplant recipients. In a large majority of cases, however, an etiologic diagnosis is lacking. Citrobacter is a genus of aerobic enteric Gram-negative bacilli that have been infrequently implicated in nosocomial infections. Case reports of meningitis, (9,10) pyelonephritis, hepatic abscess, empyema, (11) osteomyelitis, (12) endocarditis, (13) and perinephric abscess (14) have been cited in the literature. In rare cases, Citrobacter has been implicated in skin and wound infections. (15) We report a case of C. braakii cellulitis and bacteremia in a renal transplant patient receiving immunosuppressive therapy.

Discussion

Citrobacter is a distinct group of gram-negative bacilli belonging to the family Enterobacteriaceae. They are commonly found in soil, water, food, and the intestinal tracts of animals and humans. Before 1993, the only three species recognized were C. freundii, C. koseri (C. diversus), and C. amalonaticus. In 1993, Brenner et al (16) classified the genus into 11 distinct species, depending on their DNA relatedness, including C. freundii, C. koseri, C. amalonaticus, C. farmeri, C. youngae, C. braakii, C. werkmanii, C. sedlaki, and three unnamed groups, genomospecies 9, 10, and 11. Janda et al (17) biochemically identified 235 strains of Citrobacter to the species level on the basis of the taxonomic changes by Brenner and found that 68% of the isolates were members of the C. freundii complex. Within this complex, C freundii predominated (37%), followed by C. youngae (24%), C. braakii (13%), and C. werkmanii (6%). Another study reported that Brenner's methods identified 111 of 122 clinical Citrobacter isolates derived from inpatient and outpatient sources, and nine of these were recognized as C. braakii. (18) O'Hara et al (19) evaluated the potential ability of some commercial identification systems to separate these Citrobacter species in the laboratory.

Citrobacter species were initially considered as harmless contaminants but are now being increasingly recognized as human pathogens. Stone et al, (3) in 1976, conducted a prospective study of 7,519 operations and found 243 surgical wound infections, with Citrobacter species in 5% of isolates. A 1-year survey reported by Lipsky et al (20) in 1980 yielded only four cases of Citrobacter infection of superficial wounds, three of them due to C. koseri and one due to C. freundii. Hicks and Chulay (21) described a case of C. freundii cellulitis associated with tub immersion in a patient with nephrotic syndrome. Shukla (22) described a child who had plantar cellulitis caused by C. freundii and Klebsiella oxvtoca. Bishara et al (15) reported a case of C. diversus cellulitis in a patient with multiple myeloma. From 1982 to 1994, Shih et al (23) found 45 patients with Citrobacter bacteremia. All 45 patients had at least one underlying disease, the most common being malignancy (48.9%) and hepatobiliary stones (22.2%). The mortality was 17.8%, and the poor prognostic features identified were pneumonia, altered mental status on presentation, oliguria, shock, hyperbilirubinemia, and azotemia. (23)

The clinical utility of blood cultures in uncomplicated cellulitis has been questioned. Our patient, however, was febrile and immunocompromised and had failed to respond to 1 week of antibiotic therapy. The blood culture in our case not only identified an unusual pathogen, but also helped in the selection of appropriate antibiotics (Table 1). Because there are no comparative studies of antibiotic therapy for infection due to Citrobacter, treatment follows the principles of treatment of infection from other Enterobacteriaceae organisms. The various species of Citrobacter show different antimicrobial susceptibility profiles. Aminoglycosides, fluoroquinolones, carbapenems, and third- or fourth-generation cephalosporins have the highest in vitro antimicrobial activity against C. freundii. Shih et al (23) found that combination therapy with a [beta]-lactam agent and an aminoglycoside produced better results than a single agent. In patients allergic to [beta]-lactams, fluoroquinolones were found to be a good alternative. (23) Our patient was treated with a combination of aztreonam and levofloxacin. We chose to avoid aminoglycosides because of compromised renal function. The endpoints for monitoring therapy in our patient included improvement of pain and erythema in the lower extremities, resolution of fever, and negative blood cultures.

The genus Citrobacter has undergone frequent changes in nomenclature. Most studies of clinical cases are based on the previous classification of Citrobacter, making it difficult to correlate the types identified in previous reports with the present system of speciation. Also, there are no detailed studies of infections caused by the specific genetic species of Citrobacter identified by Brenner et al. (16) A MEDLINE search of articles published since 1993 failed to reveal any documentation of cellulitis or bacteremia caused by C. braakii. It is possible that some of the infections caused by C. freundii in previous reports might have been reclassified as C. braakii today, but data are not available. We were also unable to find any documentation of cellulitis due to Citrobacter in a renal transplant recipient. The presentation of our patient was similar to streptococcal cellulitis.

Conclusion

Citrobacter species are being increasingly involved in infections in immunocompromised individuals. We report a case of cellulitis and bacteremia caused by C. braakii in a renal transplant recipient. In immunocompromised individuals, cellulitis not responding to conventional antibiotics should prompt suspicion of an unusual pathogen. In such eases, blood cultures and sensitivity studies may help identify and treat these infections effectively.

Table 1. Antibiotic susceptibility of Citrobacter braakii

                         Minimal inhibitory
Antibiotic           concentration ([micro]g/ml)         Interpretation

Cefazolin          [greater than or equal to] 32.0         Resistant
Cefotaxime             [less than or equal to] 4.0         Sensitive
Aztreonam              [less than or equal to] 4.0         Sensitive
Gentamicin             [less than or equal to] 0.5         Sensitive
Levofloxacin           [less than or equal to] 1.0         Sensitive
Tobramycin             [less than or equal to] 0.5         Sensitive

References

(1.) Baddour LM. Cellulitis syndromes: An update, Int J Antimicrob Agents 2000;14:113-116.

(2.) Cox NH, Colver GB, Patterson WD. Management and mortality of cellulitis of the leg. J R Soc Med 1998;91:34-37.

(3.) Stone AM, Tucci VJ, Isenberg HD, Wise L. Wound infection: A prospective study of 7519 operations. Am Surg 1976;42:849-852.

(4.) Cooper PH, Walker AW, Beacham BE. Cellulitis caused by Histoplasma organisms in a renal transplant recipient. Arch Dermatol 1982;118:3-4.

(5.) Carlson KC, Mehlmaner M, Evans S, Chandrasoma P. Cryptococcal cellulitis in renal transplant recipients. J Am Acad Dermatol 1987;17: 469-472.

(6.) Farooqui MA, Berenson C, Lohr JW. Mycobacterium marinum infection in a renal transplant recipient. Transplantation 1999;67:1495-1496.

(7.) Rosen T. Cutaneous Mycobacterium kansasii infection presenting as cellulitis. Cutis 1983;31:87-89.

(8.) Kilborn JA, Manz LA, O'Brien M, Douglass MC, Horst HM, Kupin W, et al. Necrotizing cellalitis caused by Legionella micdadei. Am J Med 1992;92:104-106.

(9.) Chuang YC, Chang WN, Lu CH. Adult Citrobacter freundii meningitis: Case report. Changgeng Yi Xue Za Zhi 1999;22:649-653.

(10.) Doran TI. The role of Citrobacter in clinical disease of children: Review. Clin Infect Dis 1999;28:384-394.

(11.) Chuang AW, Lewis JB, Golden EB, Bhaskar SK, Bronze MS. Citrobacter freundii empyema in a patient with occult pulmonary histoplasmosis. Am J Med Sci 1998;315:314-316.

(12.) Stricker T, Frohlich S, Nadal D. Osteomyelitis and septic arthritis due to Citrobacter freundii and Haemophilus influenzae type B. J Pediatr Child Health 1998;34:90-91.

(13.) Clemente Gonzalez C, Ruiz Aguirre J, Vilert Garrofa E, Garcia Bragado F. Citrobacter freundii endocarditis [in Spanish]. An Med Interna 1999; 16:363-364.

(14.) Williams RD, Simmons RL. Citrobacter perinephric abscess presenting as pneumoscrotum in transplant recipient. Urology 1974;4:478-480.

(15.) Bishara J, Gabay B, Samra Z, Hodak E, Pitlik S. Cellulitis caused by Citrobacter diversus in a patient with multiple myeloma. Cutis 1998: 61:158-159.

(16.) Brenner DJ, Grimont PA, Steigerwalt AG, Fanning GR, Ageron E, Riddle CF. Classification of citrobacteria by DNA hybridization: designation of Citrobacter farmeri sp. nov., Citrobacter youngae sp. nov., Citrobacter braakii sp. nov., Citrobacter werkmanii sp. nov., Citrobacter sedlakii sp. nov., and three unnamed Citrobacter genomospecies, Int J Syst Bacteriol 1993;43:645-658.

(17.) Janda JM, Abbott SL, Cheung WK, Hanson DF. Biochemical identification of citrobacteria in the clinical laboratory. J Clin Microbiol 1994; 32:1850-1854.

(18.) Manganello S, Tayara A, Perazzi B, Neira L, Famiglierci A, Pugliese L, et al. Characterization and distribution of Citrobacter species in a university hospital [in Spanish]. Enferm Infecc Microbiol Clin 2001;19:11-14.

(19.) O'Hara CM, Roman SB, Miller JM. Ability of commercial identification systems to identify newly recognized species of Citrobacter. J Clin Microbiol 1995;33:242-245.

(20.) Lipsky BA, Hook EW III, Smith AA, Plorde JJ. Citrobacter infections in humans: Experience at the Seattle Veterans Administration Medical Center and a review of the literature. Rev Infect Dis 1980;2:746-760.

(21.) Hicks CB, Chulay JD. Bacteremic Citrobacter freundii cellulitis associated with tub immersion in a patient with the nephrotic syndrome. Mil Med 1988;153:400-401.

(22.) Sbukla P. Plantar cellulitis. Pediatr Emerg Care 1994;10:23-25.

(23.) Shih CC, Chen YC, Chang SC, Luh KT, Hsieh WC. Bacteremia due to Citrobacter species: significance of primary intraabdominal infection. Clin Infect Dis 1996;23:543-549.

From the Divisions of Infectious Diseases and Nephrology, Department of Internal Medicine, and the Department of Ophthalmology, University of Texas Medical Branch, Galveston, TX.

Reprint requests to Mahendra L. Agraharkar, MD, Department of Internal Medicine, University of Texas Medical Branch, 4.200 John Scaly Annex, 301 University Blvd., Galveston, TX 77555-0562. Email: [e-mail]

Accepted May 28, 2002.