Pseudomonas putida

Pseudomonas putida Newly Recognized Pathogen in Patients with Cancer

ELIAS ANAISSIE, M.D. VICTOR FAINSTEIN, M.D. PEGGY MILLER, A.S.C.P.(M.) HASSANALI KASSAMALI, M.D. SILVIO PITLIK, M.D. GERALD P. BODEY, M.D. KENNETH ROLSTON, M.D. Houston,

Texas

Pseudomonas putida was recovered from blood culture specimens between 1980 and 1985 in 15 patients with cancer. No isolates were found in specimens obtained before 1980. Eight patients were considered to have septicemia (more than one positive blood culture result plus clinical signs of infection). Septicemia was monomicrobial in three of those eight patients and polymicrobial in five. Df these eight patients, one had pneumonia and three had phlebitis, cellulitis, or both at the site of the venous catheter. The infection appeared to be catheter-related in these three patients, with response to catheter removal in one patient, response to catheter removal and antibiotics in one patient, and response to antibiotics alone in one patient. P. putida was isolated from the site of insertion and the tip of the catheter in one of these three patients. Following therapy, all patients had a rapid recovery from their infection. In vitro susceptibility testing revealed that 90 percent of the isolates were susceptible to piperacillin, ceftazidime, imipenem, and ciprofloxacin. Pseudomonas putida is a ubiquitous environmental contaminant that is recovered occasionally from clinical specimens. It has rarely been isolated from blood specimens, and its role as a human pathogen is uncertain [ 11. The pathogenic@ of a closely related organism, P. fluorestens, in cold-blooded animals, and the association of both organisms with bacteremia transmitted by contaminated blood products stored at 4’C suggest a facilitating role of hypothermia in the pathogenicity of P. putida [2-41. All cases of P. putida bacteremia in a cancer hospital during a lo-year period constitute the basis of this report. The clinical presentation of infection, microbiologic characteristics, antimicrobial susceptibilities, and temperature growth patterns of P. putida are discussed.

PATIENTS AND METHODS

From the Department of Internal Medicine, Section of Infectious Diseases, University of Texas M. D. Anderson Hospital and Tumor Institute at Houston, Houston, Texas. Requests for reprints should be addressed to Dr. Elias Anaissie, M. D. Anderson Hospital (Box 47), 6723 Bertner Avenue, Houston, Texas 77030. Manuscript submitted October 8, 1986, and accepted December 23, 1986.

The records of the Clinical Microbiology Laboratory of the University of Texas M. D. Anderson Hospital and Tumor Institute at Houston from June 1975 through June 1985, were reviewed for cases of P. putida bacteremia. Subjects whose only positive culture finding was collected from a heart blood specimen at autopsy examination were excluded from clinical analysis. A patient was considered to have septicemia if P. putida was isolated from at least two blood culture specimens and if his or her clinical presentation was compatible with septicemia, i.e., if it was characterized by fever, chills, hypotension, shock, or a combination of these symptoms. Blood specimens were inoculated in tryptic soy broth, and processed by a radiometric blood culture system (BACTEC, Johnston Laboratory, Cockeysville, Maryland). Identification of Pseudomonas species was performed using the API 20E microtube recognition system.

June 1987

The American Journal of Medicine

Volume 82

1191

PSEUDOMONAS PUTIDA-ANAISSIE ET AL

TABLE

I

Clinical

Characteristics

of Patients

with Pseudomonas

putida Septicemia

Number of Age

Patient Number Group

I:

1

and Sex

Posltive Blood Culture Results

Underlying Disease

Monomicrobial 19F Acute lymphoblastic leukemia

2

45F

3

50M

Chronic myelogenous leukemia Lung cancer

4

2 2

Group II: Polymicrobial 4 72M Smoldering leukemia 5 36M Acute myelogenous leukemia 6 13M Osteosarcoma 7

25M

a

16F

RelevantClinical and MicrobiologicDala

Yes

Moxalactam; catheter removal

Yes

Pneumonia; P. putida cultured from sputum

Yes

CeftriaxoneIamikacin; catheter removal Cefoperazone/mezlocillin

Yes

Piperacillin/ceftazidime

Phlebitis at catheter site; septic shock Septic shock

Yes

Pus at catheter site

Yes

Moxalactam/ticarcillin/ tobramycin Catheter removal; ticarcillin/tobramycin Catheter removal

No

Piperacillin

-

June

1987

The American

Journal

of Medicine

Volume

Yes

RESULTS

It has been shown recently that the API 20E microtube recognition system correctly identifies only 67 percent of fluorescent pseudomonads, and that additional tests are required for accurate classification [5]. The failure to grow at 42OC differentiates P. putida from P. aeruginosa, and the lack of gelatin hydrolysis and egg yolk reaction distinguishes P. putida from P. fluorescens [I]. All strains diagnosed by the microbiology laboratory as P. putida were subjected to these three tests, and only strains fulfilling the aforementioned criteria for identification of P. putida were included in this analysis. Of 26 isolates classified as P. putida using the API 20E system, only 15 were reconfirmed by additional procedures. Eight of the 11 misidentified isolates were P. fluorescens, and all eight had been obtained from heart blood culture specimens at autopsy examination. Of the remaining three, two were subsequently identified as P. stutzeri and one as P. aeruginosa. Antibiotic susceptibility testing was performed by a previously described microtiter broth dilution method [6]. The range of antibiotic concentrations utilized varied from 0.004 to 128 pg/ml. Mueller-Hinton broth was used as the test medium. The broth was supplemented with cations (Ca+, Mg+) when amikacin was studied. Organisms ( IO5 ml) were inoculated into broth cultures and incubated at 37°C for 18 hours. The minimal inhibitory concentration was defined as the lowest concentration of antimicrobial agent resulting in the inhibition of bacterial growth after 18 to 24 hours of incubation at 37%. Overnight cultures in Mueller-Hinton broth of 15 clinical isolates of P. putida were plated on blood agar. Following 24 and 48 hours of incubation at different temperatures (4O, 24’, 30°, and 37’C), colony size and number were determined. Culture specimens were maintained at 4’C for four weeks.

1192

Therapy

Phlebitis at catheter site; P. putida cultured from catheter tip and site -

-

Acute myelogenous leukemia Acute myelogenous leukemia

CentralVenous Catheter

No cases of P. putida septicemia were detected in specimens obtained before 1980. After 1980, there was a substantial increase in the use of indwelling central venous catheters at our institution. In 1977, 53 of these catheters were utilized, but this number increased to 2,350 by 1982. P. putida was isolated from blood culture specimens collected from 15 patients, but only eight of these patients had more than one positive blood culture result and signs of infection. Their clinical features are summarized in Table I. Three additional patients had clinical signs and symptoms of septicemia and could have been infected with P. putida. They are not, however, included in the analysis, as this organism was isolated from only one of two or one of three blood specimens obtained. The remaining four patients had no clinical signs of infection. It is important to mention that the cases of septicemia were not associated in time or space. On the basis of epidemiologic studies conducted over the 1980 to 1985 period, P. putida was rarely isolated in our hospital environment, and a common source of isolation could never be found. A silicone elastomer (Centrasil, Travenol Laboratories, Inc., Deerfield, Illinois) central venous catheter was in place in seven of the eight patients. In three of these seven patients (Patients 1, 5, and 7), the clinical picture was consistent with that of catheter-related sepsis, i.e., presence of severe phlebitis (Patients 1 and 5), purulent material at the catheter site (Patients 1 and 7), defervescence upon removal of the central venous catheter (Pa-

82

PSEUDOMONAS

TABLE II

tients 1 and 7), and/or isolation of the same organism from both the insertion site and the tip of the central venous catheter when culture specimens were taken (Patient 1). Culture samples from both the insertion site and the tip of the central venous catheter were obtained in two patients. In Patient 1, resolution of bacteremia and signs and symptoms of infection coincided with removal of the offending catheter four days after initiation of moxalactam therapy. In Patient 7, removal of the catheter was the only therapeutic maneuver required for resolution of the bacteremia. Antimicrobial agents were administered to seven patients, all of whom received at least one antibiotic with good in vitro activity against P. putida. Antineoplastic chemotherapy and neutropenia were not observed to be major predisposing factors for the development of P. putida infection. Five of the eight patients had an adequate neutrophil count at the onset of their infection, and four of these five patients were not receiving chemotherapy. In all patients, fever and chills were present and recovery was complete and rapid. Patient 3 had associated pneumonia. Gram stain of the sputum revealed many white blood cells and many gramnegative rods, and P. putida was the only organism isolated. All blood culture specimens drawn from the five patients with polymicrobial septicemia (Patients 4, 5, 6, 7, and 8) supported the growth of co-pathogenic organisms (P. maltophilia in Patient 4, alpha-hemolytic Streptococcus and Staphylococcus epidermidis in Patient 5, Acinetobacter calcoaceticus and P. aeruginosa in Patient 6, Enterobacter cloacae and Achromobacter zyloxidans in Patient 7, and A. calcoaceticus, Klebsiella oxytoca, and Citrobacter diversus in Patient 8). All these co-pathogens were isolated from the same number of blood culture specimens, as was P. putida, except for A. zyloxidans (Patient 7), which was isolated from two of the three blood culture samples that showed positive findings. Table II summarizes the results of susceptibility testing for the 15 clinical isolates. Most strains were susceptible to ceftazidime, imipenem, and ciprofloxacin. Aztreonam, piperacillin, cefoperazone, and amikacin were moderately active against P. putida, whereas trimethoprim/sulfamethoxazole showed very poor activity. All P. putida isolates grew at 24’, 30°, and 37OC. We could not detect any significant difference in the number and/or size of colonies at these three different temperatures. Only one P. putida isolate grew at 4OC after two weeks of incubation.

PUTIDA-ANAISSIE

ET AL

In Vitro Activity of Nine Antimicrobial Agents against 15 Clinical Isolates of P. putida Minimal Inhibitory Concentration (pg/ml) Range MlCso MlC50

Amikacin Piperacillin Cefoperazone Ceftazidime lmipenem Aztreonam Trimethoprimi sulfamethoxazole Ciprofloxacin

0.5->I28 OS>128 2-64 1-128 0.25->128 2-128 0.125/2.37-161304 0.008-8

4 8 16 2 0.5 16 81152

>128 32 32 8 1 64 >16/304

0.06

0.25

MICSO, MIC& = the lowest concentration of antibiotic resulting in inhibition of bacterial growth in 50 percent or 90 percent, respectively, of tested isolates.

COMMENTS

P. putida is a common inhabitant of soil, water, and plants, and is frequently isolated from the hospital environment. It is also part of the normal oropharyngeal flora [l]. It is usually regarded as an environmental contaminant, but unanimity as to its pathogenic significance has been lacking [I]. As shown in our study, P. putida can be

June

pathogenic for humans, and its clinical spectrum comprises septicemia with pure growth of the organism, and polymicrobial septicemia. One case of septicemia with probable pneumonia is also presented. Although a positive result of a blood culture for P. putida should not be disregarded as evidence of a contaminant in the proper clinical setting, the significance of the isolation of P. putida from a single blood specimen is difficult to assess (contamination of the venipuncture site versus low-grade infection). It is important to remember, however, that the other fluorescent Pseudomonas species should always be considered in the differential diagnosis. This study and a previously reported one have shown that the API 20E microtube recognition system is inadequate for differentiating P. putida from other fluorescent Pseudomonas species, and that supplemental methodologies are required to complete the identification [5]. The role of P. putida in producing infection in patients with cancer is increasing, as suggested by the absence of cases of P. putida bacteremia at our institution prior to 1980 and their increasing occurrence since then. The increase at our institution during the last five years has paralleled a similar increase in the number of indwelling central venous catheters in our patients. Indeed, some of the cases of septicemia we are reporting appeared to be catheter-related. In such cases, removal of the catheter in addition to antimicrobial therapy might be required to control the infection. Among the currently available antibiotics tested, imipenem and ceftazidime were the most active against P. putida, and ciprofloxacin appears to be a promising agent on the basis of in vitro results. Finally, hypothermia does not seem to facilitate the pathogenicity of P. putida. Some observations on a closely related microorganism (P. fluorescens) demonstrate that hypothermia is important for its pathogenic@. P. fluorescens has optimal growth below 30°C, and most strains are able to multiply at temperatures below 4’C

1987

The American

Journal

of Medicine

Volume

82

1193

PSEUDOMONAS

PUTIDA-ANAISSIE

ET AL

[I]. Although it rarely infects humans, P. fluorescens is commonly found as a pathogen in cold-blooded animals [2]. In addition, it has been found to contaminate blood stored at low temperatures and to produce human disease after transfusion of contaminated blood products [3]. In one animal experiment, contamination of the burned tails of rats with either P. aeruginosa or P. fluorescens led to local infection in all instances, whereas bacteremia and metastatic infection occurred only with P. aeruginosa, a species with optimal growth at body temperature. The low temperature at the rat’s tail allowed local infection by P. fluorescens, a psychrophilic organism [7]. Previous reports of P. putida sepsis transmitted by blood transfusions suggest a role for low temperature in facilitating P. putida

infection [4,8]. However, fever was a common denominator in our patients with P. putida septicemia as well as in one case of P. putida arthritis reported previously [9]. Also, we were unable to demonstrate that lower temperatures facilitated the in vitro growth of this organism. In conclusion, P. putida has emerged as an occasional bacterial pathogen in immunocompromised patients and is probably related to the increased use of central venous catheters. Physicians caring for such patients should not disregard positive results of blood cultures for P. putida as evidence of contaminants, and should perform supplemental bacteriologic tests for complete identification of this organism if more than one blood culture finding is positive.

REFERENCES 1.

2.

3.

4.

5.

1194

Gilardi G: Pseudomonas. In: Lennette EH, Balows A, Hausler WJ, Shadomy HJ, eds. Manual of clinical microbiology. Washington: American Society for Microbiology, 1985; 350-372. Bullock GL: The identification and separation of Aeromonas liquefaciens from Pseudomonas fluorescens and related organisms occurring in diseased fish. Appl Microbial 1961; 9: 587-590. Khabbaz RF, Arnow PM, Highsmith AK, et al: Pseudomonas fluorescens bacteremia from blood transfusion. Am J Med 1984; 76: 62-68. Tabor E, Gerety RJ: Five cases of Pseudomonas sepsis transmitted by blood transfusions (letter). Lancet 1984; I: 1403. Oberhofer TR: Use of the API 20E, Oxi/Ferm, and Minitek

June

1987

The American

Journal

of Medicine

Volume

6.

7.

8.

9.

82

systems to identify nonfermentative and oxidase-positive fermentative bacteria: seven years of experience. Diagn Microbial Infect Dis 1983; 1: 241-256. Bodey GP, Ho DH, LeBlanc B: In vitro studies of BMY-28142, a new broad-spectrum cephalosporin. Antimicrob Agents Chemother 1985; 27: 265-269. Liu PV: Pathogenicity of Pseudomonas fluorescens and related pseudomonads to warm-blooded animals. Am J Clin Pathol 1964; 41: 150-153. Taylor M, Keane CT, Falkiner FR: Pseudomonas putida in transfused blood (letter). Lancet 1984; II: 107. Madhavan T, Fisher EJ, Cox F, Quinn EL: Pseudomonas putida and septic arthritis. Ann Intern Med 1973; 78: 971-972.