Antimicrobial Susceptibility Patterns (Antibiograms) as ...

Antimicrobial Susceptibility Patterns (Antibiograms) as an Aid in Differentiating Citrobacter Species PAUL M. SOUTHERN, JR., M.D., AND MARY K. BAGBY, B.S.

Southern, Paul M., Jr., and Bagby, Mary K.: Antimicrobial susceptibility patterns (antibiograms) as an aid in differentiating Citrobacter species. Am J Clin Pathol 67: 187-189, 1977. The hydrogen sulfide-negative Citrobacter group represents a taxonomic problem. Various investigators have proposed such designations as Padlevvskia, Levinea, atypical Enterohacter cloacae, H2S-negative variants of Citrobacter, Citrobacter koseri and Citrobacter diversus. This problem has been investigated with emphasis on antibiograms as a means of discrimination. Clinical isolates fitting the designation Citrobacter were studied and, using the criteria of Ewing and Davis, separated into two groups: C. diversus (40 strains) or C. freundii (25 strains). Susceptibilities to ampicillin, carbenicillin, cefazolin, cephaloridine and cephalothin were determined by the agar-dilution method. C. diversus strains were resistant to 8 /ug/ml ampicillin (97.5%) and 32 /ig/ml carbenicillin (87.5%), and were susceptible to 8 /ag/ml cephalosporins (>90%). C. freundii strains were moderately susceptible to 8 /ug/ml ampicillin (25%) and susceptible to 8 /Ug/ml carbenicillin (92%), and were resistant to 8 /xg/ml cephalosporins (>92%). Using these data one can separate C. diversus from C. freundii with 90% accuracy. (Key words: Citrobacter species: Citrobacter diversus; Citrobacter freundii; Antimicrobial susceptibility; Antibiograms.)

From the Department of Pathology, The University of Texas Health Science Center at Dallas, and Parkland Memorial Hospital, Dallas, Texas 75235

Because of this area of disagreement upon classification, it is difficult for the average clinical laboratory to handle this issue satisfactorily. In the process of examining this problem in our laboratory we became aware of differences in antimicrobial susceptibility patterns manifested by various strains that we had classified as members of the genus Citrobacter. Several months earlier, working in the same laboratory, Jones and associates 5 reported data relating to a series of isolates designated C. diversus. They reported relative ineffectiveness of ampicillin and susceptibility to cephalothin in their isolates. We then reviewed the antibiograms of the strains designated C. diversus (54 strains) and C. freundii (28 strains) isolated in the clinical microbiology laboratory of Parkland Memorial Hospital in a later three-month period during 1974 (unpublished data). While only 2% of C. diversus strains were inhibited by ampicillin at 20 /ug/ml and none at 5 ptg/ml, more than half of the C. freundii strains were inhibited at 20 jug/ml. The difference was more obvious with cephalothin. All strains of C. diversus were inhibited by 20 ^ug/ml, and 93% were inhibited by 5 ptg/ml. On the other hand, only 58% of C. freundii strains were inhibited by 20 /u-g/ml, and only a few at lower concentrations. Under appropriate conditions antibiograms may serve as a taxonomic aid. For this reason we proposed to study a group of Citrobacter strains, placing major emphasis on antimicrobial susceptibility patterns. For purposes of arbitrarily separating the strains into groups for comparison, we used the criteria of Ewing and Davis. 2 This separates these organisms into the species C. freundii and C. diversus. Important determinants of that separa-

THE HYDROGEN sulfide-negative Citrobacter group has been a taxonomic problem for several years. Macierewicz 7 proposed the genus designation Padlewskia for this group, Fredricksen 3 suggested the name Citrobacter koseri for a group of indolenegative organisms, Washington and associates 11,12 studied a group of atypical Enterohacter cloacae and hydrogen sulfide-negative variants of Citrobacter, Young and co-workers 14 proposed a new genus, Levinea, while Ewing and Davis2 advocated using the designation Citrobacter diversus. More recently Markel and colleagues 8 ' 9 presented additional data in support of the use of the designation Levinea. Received March 2, 1976; received revised manuscript April 7, 1976; accepted for publication April 7, 1976. Address reprint requests to Dr. Southern: Department of Pathology, University of Texas Health Science Center at Dallas, 5323 Harry Hines Blvd., Dallas, Texas 75235.

187

Downloaded from https://academic.oup.com/ajcp/article-abstract/67/2/187/1771531 by guest on 02 April 2018

A.J.C.I'. • February 1977

SOUTHERN, JR. AND BAGBY

188

Table 1. Sources of Clinical Isolates of Citrobacter species Number of Isolates

Urine Respiratory secretions Surgical infections Blood Other*

C. diversus (40)

C. freiindii (25)

13 10 8 6 3

10 8 4 3 0

* One isolate each from synovial fluid, ear drainage, and an infected hematoma.

tion include: growth on KCN, adonitol fermentation, production of H2S on triple sugar iron agar (TSI) or Kligler's iron agar (KI) and indole production. Materials and Methods Strains of Citrobacter isolated from clinical specimens submitted to the diagnostic microbiology laboratory at Parkland Memorial Hospital were examined (Table 1). They were classified as either C. diversus or C. freiindii using the criteria of Ewing and Davis.2 These strains were collected over a period of approximately six weeks during 1974. Forty isolates were classified as C. diversus and 25 as C. freiindii. They were studied for antimicrobial susceptibility by the agar-dilution method as described by Washington.13 Standard reference antimicrobial agents obtained from the respective manufacturers of the drugs were employed. Antimicrobial agents tested were ampicillin, carbenicillin, cefazolin, cephaloridine and cephalothin. Results No significant separation between C. diversus and C.freundii was evident with respect to the sources of the respective isolates (Table 1). Essentially similar percentages of the two species were isolated from the same clinical sources. Table 2 shows the susceptibilities of C. diversus strains to ampicillin and the three cephalosporin analogs. Only 2.5% were inhibited by 8 /xg/ml ampicillin, and a majority required >32 /Ag/ml for inhibition. On the other hand, at least 80% of strains were inhibited by 2 /xg/ml of the cephalosporins, and at least 90% by 8 /xg/ml. All strains were inhibited by cephalothin and cefazolin at