Rhodotorula septicemia: case report and minireview

CONCISE COMMUNICATIONS Modified Hodge and EDTA-disk synergy tests to screen metallo-blactamase-producing strains of Pseudomonas and Acinetobacter species K. Lee, Y. Chong, H. B. Shin, Y. A. Kim, D. Yong and J. H. Yum Department of Clinical Pathology and Research Institute of Bacterial Resistance, Yonsei University College of Medicine, C.P.O. Box 8044, Seoul, Korea 

Tel: ‡ 82 2361 5866 Fax: ‡ 82 2313 0908

E-mail: [email protected]

Accepted 30 November 2000

Many clinically relevant species of Gram-negative bacilli are often resistant to b-lactam antibiotics, including extendedspectrum cephalosporins, but rarely to carbapenems [1]. Carbapenem resistance was mainly due to decreased outer-membrane permeability [2]. However, IMP-1 metallo-b-lactamaseproducing Pseudomonas aeruginosa emerged in Japan [3], and then the resistance spread to other species. Recently, IMP-2producing Acinetobacter baumannii [4] and VIM-1- and VIM-2producing strains of P. aeruginosa have been reported in Europe [5±7]. The metallo-b-lactamase-producing strains of P. aeruginosa which we reported in 1998 [8] were subsequently identi®ed as VIM-2 b-lactamase producers. As metallo-b-lactamase genes reside in mobile integrons, rapid detection of metallo-blactamase-producing bacteria is necessary not only to augment our understanding of the resistance, but also to control the spread of the resistance. Hodge et al [9] developed a test to detect penicillinaseproducing Neisseria gonorrhoeae and other species of bacteria. A double disk synergy test using b-lactam and b-lactamaseinhibitor disks is a convenient method of detecting extendedspectrum b-lactamase-producing Gram-negative bacilli [10]. EDTA inhibition of b-lactamase activity is used to differentiate a metallo-b-lactamase from other b-lactamases [11]. We found that the tests of modi®ed Hodge and EDTA-disk synergy were simple to use to screen metallo-b-lactamase-producing strains [12]. The aim of this study was to evaluate the usefulness of the modi®ed Hodge and EDTA-disk synergy tests for the screening of metallo-b-lactamase-producing strains from a large number of imipenem-resistant clinical isolates of Pseudomonas spp. and Acinetobacter spp. Pseudomonas spp. and Acinetobacter spp. were isolated in 1995± 99 from urine, sputum and other clinical specimens and identi®ed by conventional methods and by the ID 32 GN system (bioMerieux Vitek, Marcy-l'Etoile, France). Isolates resistant to imipenem by the disk diffusion test [13] were kept at 76 8C in 20% skimmed milk until used in this study. P. aeruginosa strains producing IMP-1, VIM-1 and VIM-2 metallo-b-lactamases,

and a Serratia marcescens strain producing Sme-1 serine blactamase, were used as control strains. The test of Hodge et al [9] was modi®ed by substituting Escherichia coli ATCC 25922 for penicillin-susceptible Staphylococcus aureus ATCC 25923, and 10-mg imipenem disk for a 10U penicillin disk. The surface of a Mueller±Hinton agar plate was inoculated evenly using a cotton swab with an overnight culture suspension of E. coli, which was adjusted to one-tenth turbidity of the McFarland no. 0.5 tube. After brief drying, an imipenem disk was placed at the center of the plate, and imipenem-resistant test strains from the overnight culture plates were streaked heavily from the edge of the disk to the periphery of the plate. The presence of a distorted inhibition zone after overnight incubation was interpreted as modi®ed Hodge test positive. For the EDTA-disk synergy test, an overnight culture of the test strain was suspended to the turbidity of a McFarland no. 0.5 tube and used to swab inoculate a Mueller±Hinton agar plate. After drying, a 10-mg imipenem disk (BBL, Cockeysville, MD) and a blank ®lter paper disk were placed 10 mm apart from edge to edge, and 10 mL of 0.5 M EDTA solution was then applied to the blank disk, which resulted in approximately 1.5 mg/disk. After overnight incubation, the presence of an enlarged zone of inhibition was interpreted as EDTA-synergy test positive. Tests used to determine the characteristics of the EDTA-disk synergy-positive strains were agar dilution susceptibility [14] to imipenem (Merck Sharp & Dohme, West Point, PA, USA), and imipenem-hydrolyzing activity [15] of the crude cell sonicate with or without EDTA treatment at 30 8C by using a UV spectrophotometer (Shimadzu Corp., Tokyo, Japan). Isoelectric focusing of the b-lactamase [16] was performed using a polyacrylamide gel (pH 3±10) and a ThermoFlow ETC Unit (Novex Experimental Technology, San Diego, CA, USA), and visualization of the bands with a 1 mg/mL solution of nitroce®n (Unipath Ltd, Basingstoke, UK). The blaVIM-2 gene was detected by PCR using the following primers: forward 50 -ATT GGT CTA TTT GAC CGC GTC-30 (nucleotides

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases

Concise Communications

89

Table 1 Screening of metallo-b-lactamase-producing species of Pseudomonas and Acinetobacter by the modified Hodge test and the EDTA-disk synergy test

Modified Hodge test (no. tested) Positive (43) P. aeruginosa (28) P. putida (6) Acinetobacter spp. (9)b Equivocal (31) P. aeruginosa (24) P. putida (3)c A. baumannii (4) Negative (456) P. aeruginosa (441) Acinetobacter spp. (15) Total (530)

No. (%) of isolates with EDTA-disk synergya

No. (%) of isolates with imipenem hydrolysisa

Positive

Negative

Positive

28 (100) 6 (100) 9 (100)

0 (0) 0 (0) 0 (0)

28 (100) 6 (100) 9 (100)

0 (0) 0 (0) 0 (0)

15 (62.5) 3 (100) 4 (100)

9 (37.5) 0 (0) 0 (0)

15 (62.5) 3 (100) 4 (100)

9 (37.5) 0 (0) 0 (0)

0 (0) 0 (0) 65

441 (100) 15 (100) 465

0 (0) 0 (0) 65

Negative

39 (100) 15 (100) 63

a Sensitivities and specificities for the modified Hodge test were 100% and 88%, respectively, and for the EDTA-disk synergy test both 100%, as compared to imipenem hydrolysis. Only 54 of 456 Hodge and EDTA-disk synergy-negative isolates and nine Hodge-equivocal, but EDTA-disk synergy-negative isolates were tested for imipenem hydrolysis. bSeven isolates were A. baumannii and two were Acinetobacter spp. cEquivocal when incubated at 35 8C, but positive at 30 8C.

1342±1362), and reverse 50 -TGC TAC TCA ACG ACT GAG CG-30 (nucleotides 2102±2121) [7]. Template DNA was extracted by boiling for 5 min. The PCR conditions were one cycle of predenaturation at 94 8C for 5 min, 35 cycles of denaturation at 94 8C for 30 s, annealing at 56 8C for 30 s and extension at 72 8C for 1 min, and one cycle of ®nal extension at 72 8C for 7 min. Among a total of 530 imipenem-resistant isolates screened by the modi®ed Hodge test, 43 were positive and 31 were equivocal (Table 1, Figure 1). Six of nine imipenem-resistant P. putida isolates showed positive results at 35 8C, but the remaining three equivocal isolates became positive at 30 8C. Among the 28 isolates of Acinetobacter spp. tested, nine were positive and four were equivocal. The distortion patterns shown by all of the isolates were similar to those of the control strains producing IMP-1, VIM-1, VIM-2 and Sme-1. All of the 43 modi®ed Hodge test-positive isolates showed an EDTA-disk synergy-positive pattern (Figure 2A, C) similar to those shown by IMP-1- VIM-1- and VIM-2-producing control strains. Twenty-two of 31 equivocal strains were also synergy positive. The inhibition zones which were measured from the edge of the disk were 0±3 mm with imipenem disks alone, while those between EDTA and imipenem disks were 6± 14 mm. The inhibition zones were enlarged at least 6 mm by the presence of the EDTA disk. All of the 65 EDTA-disk synergy-positive strains hydrolyzed imipenem. Fifty-four of the 456 Hodge- and EDTA-disk synergy-negative isolates and nine Hodge-equivocal but EDTA-disk synergy-negative isolates were spectrophotometrically negative for imipenem hydrolysis (Table 1). Sensitivities

and speci®cities were for the modi®ed Hodge test 100% and 88%, respectively, and for the EDTA-disk synergy test both 100% as compared to imipenem hydrolysis. The hydrolysis of imipenem by all of the EDTA-disk synergy-positive strains was inhibited by 10 mM EDTA, but

Figure 1 Modified Hodge test. A Mueller^Hinton agar plate was inoculated with E. coli ATCC 25922. An imipenem disk was put in place and imipenemresistant test isolates were streaked from the edge of the disk to the periphery of the plate and incubated overnight. (A) and (C) are imipenem-hydrolyzing strains which distorted the inhibition zone. (B) and (D) are imipenemnon-hydrolyzing strains with no effect on the zone.

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

90 Clinical Microbiology and Infection, Volume 7 Number 2, February 2001

Figure 2 EDTA-disk synergy test. A Mueller^Hinton agar plate was inoculated with an imipenem-resistant test isolate, and a 10-mg imipenem disk (left) and a 5 mM EDTA disk (right) were put in place and incubated overnight. (A) A synergistic inhibition zone is shown between an imipenem disk (left) and an EDTA disk (right) by a VIM-2 metallo-b-lactamase-producing strain. (B) Only a partially inhibited zone around an EDTA disk is shown by a metallo-b-lactamase-non-producing strain. (C) A metallo-b-lactamaseproducing strain showing a small enhanced inhibition zone when the disks were placed 15 mm apart.

not by 50 mM clavulanic acid. The MIC range of imipenem for the isolates was 8±256 mg/L. Relative rates of hydrolysis of imipenem by representative isolates were 24±47% when compared to that of ampicillin. The blaVIM-2 gene was detected from all of the EDTA-synergy-positive strains by PCR. The pI of VIM-2 b-lactamase was approximately 5.3. In the original test of Hodge et al [9], a penicillin G disk and penicillin G-susceptible Staphylococcus aureus strain were used to differentiate penicillinase-producing gonococci. In this study, it was found that background-lawn formation by Staphylococcus aureus was inhibited by some test strains of P.

aeruginosa and that the inhibition zones with imipenem disks were too large. Our modi®ed Hodge test, using a commercially available 10-mg imipenem disk and an indicator E. coli strain at one-tenth the turbidity of a McFarland 0.5 tube, detected not only IMP-1, VIM-1 and VIM-2, but also an Sme-1 producer. The test was simple to use for screening carbapenem-hydrolyzing strains out of a large number of imipenem-resistant isolates. The distortion of the inhibition zone was less distinct compared to that shown by gonococci [9], but in most cases it was not dif®cult to suspect carbapenem hydrolysis (Figure 1). Three of nine P. putida isolates which were modi®ed Hodge test equivocal at 35 8C became positive at 30 8C, possibly because of the lower optimal growth temperature of the organism. Our preliminary results [12] showed that a double disk synergy test using a 10-mg imipenem disk was easier to read when the EDTA disk content was 1.5 mg than when it was 3 or 4.5 mg. An approximately 1.5-mg EDTA disk can be prepared by adding 10 mL of a readily available 0.5 M EDTA solution (pH 8.0) to a blank ®lter paper disk. The test results were similar with both dried and wet disks. Occasional imipenem-hydrolyzing strains did not show enhanced inhibition zones or showed only small zones when the two disks were placed 15 mm apart from edge to edge (Figure 2C). In general, the results were more distinct when the distance between the two disks was 10 mm. It was recently reported by Arakawa et al that a disk containing 2-mercaptopropionic acid gave the clearest synergy in screening IMP-1-producing strains [17]. However, 0.5 M EDTA solution (pH 8.0) is readily available in most laboratories. Arakawa et al used ceftazidime disks instead of imipenem disks to increase the sensitivity of the test, as some strains were resistant to low levels of imipenem only. However, ceftazidime should be less speci®c for the detection of metallo-b-lactamase producers. McLaughllin and Laconis [18] used a disk containing 20 mL of a 5 mM EDTA solution to screen for the presence of metallo-b-lactamase in exponentially growing cell sonicates. However, it takes time to prepare cell sonicates. Although the usefulness of our modi®ed Hodge and EDTAdisk synergy tests was evaluated using VIM-2-producing clinical isolates only, the results with control strains producing IMP-1, VIM-1, VIM-2 and Sme-1 b-lactamase suggested that the tests could also be used to screen for strains producing carbapenemases other than VIM-2. A limitation of this study is that although many VIM-2 metallo-b-lactamase-producing isolates were tested, PFGE patterns suggest that some of them belong to the same clone. In conclusion, the tests of modi®ed Hodge and EDTA-disk synergy are simple to perform and should be suitable to screen metallo-b-lactamase-producing clinical isolates of P. aeruginosa, P. putida and Acinetobacter spp., to characterize the b-lactamases and to control spread of the resistance.

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

Concise Communications

ACK NOW L EDGME N T S We thank Dr David M. Livermore, Central Public Health Laboratory, London, UK, for providing IMP-1 and VIM-2 b-lactamase-producing P. aeruginosa and Sme-1-producing Serratia marcescens, and Professor Gian M. Rossolini, Universita di Siena, Siena, Italy for VIM-1-producing P. aeruginosa. We thank Young Hee Seo for her technical assistance. R EFER E NCE S 1. Rasmussen BA, Bush K. Carbapenem-hydrolyzing b-lactamases. Antimicrob Agents Chemother 1997; 41: 223±32. 2. Buecher K, Cullmann W, Dick W, Opferkuch W. Imipenem resistance in Pseudomonas aeruginosa resulting from diminished expression of outer membrane protein. Antimicrob Agents Chemother 1987; 31: 703±8. 3. Watanabe M, Iyobe S, Inoue M, Mitsuhashi S. Transferable imipenem resistance in Pseudomonas aeruginosa. Antimicrob Agents Chemother 1991; 35: 147±51. 4. Riccio ML, Franceschini N, Boschi L et al. Characterization of the metallo-b-lactamase determinant of Acinetobacter baumannii AC-54/97 reveals the existence of blaIMP allelic variants carried by gene cassettes of different phylogeny. Antimicrob Agents Chemother 2000; 44: 1229±35. 5. Lauretti L, Riccio ML, Mazzariol A et al. Cloning and characterization of blaVIM, a new integron-borne metallo-blactamase gene from a Pseudomonas aeruginosa clinical isolate. Antimicrob Agents Chemother 1999; 43: 1584±90. 6. Tsakris A, Pournaras S, Woodford N et al. Outbreak of infections caused by Pseudomonas aeruginosa producing VIM-1 carbapenemase in Greece. J Clin Microbiol 2000; 38: 1290±2. 7. Poirel L, Nass T, Nicolas D et al. Characterization of VIM-2, a carbapenem-hydrolyzing metallo-b-lactamase and its plasmid- and integron-borne gene from a Pseudomonas aeruginosa clinical isolate in France. Antimicrob Agents Chemother 2000; 44: 891±7. 8. Lee K, Chong Y, Shin HB, Yong D. Rapid increase of imipenemhydrolyzing Pseudomonas aeruginosa in a Korean hospital [abstract

9. 10. 11. 12.

13.

14.

15. 16. 17. 18.

91

E-85]. In: Program and abstracts of the 38th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, CA. Washington, DC: American Society for Microbiology, 1998: 193. Hodge W, Ciak J, Tramont EC. Simple method for detection of penicillinase-producing Neisseria gonorrhoeae. J Clin Microbiol 1978; 7: 102±3. Livermore DM. b-lactamases in laboratory and clinical resistance. Clin Microbiol Rev 1995; 8: 567±84. Payne DJ, Cramp R, Bateson JH, Neal J, Knowles D. Rapid identification of metallo- and serine b-lactamases. Antimicrob Agents Chemother 1994; 38: 991±6. Chong Y, Lee K, Shin HB, Kim YA. Simple modified Hodge and EDTA disk synergy tests to differentiate carbapenem-hydrolyzing Pseudomonas aeruginosa [abstract D-40]. In: Program and abstracts of the 38th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, CA. Washington, DC: American Society for Microbiology, 1998: 139. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disk susceptibility tests. Approved standards M2-A6. Wayne PA: National Committee for Clinical Laboratory Standards, 1997. National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Approved standards M7±A4. Wayne PA: National Committee for Clinical Laboratory Standards, 1997. Bush K. Characterization of b-lactamases. Antimicrob Agents Chemother 1989; 33: 259±63. Matthew M, Harris AM, Marshall MJ, Ross GW. The use of analytical isoelectric focusing for detection and identification of blactamases. J Gen Microbiol 1973; 88: 169±78. Arakawa Y, Shibata N, Shibayama K et al. Convenient test for screening metallo-b-lactamase-producing Gram-negative bacteria by using thiol compounds. J Clin Microbiol 2000; 38: 40±3. McLaughllin DM, Iaconis JP. Bioassay development for detection of zinc-dependent metallo-b-lactamase (Mb)-producing bacteria [abstract C2-1488]. In: Program and abstracts of the 39th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA. Washington, DC: American Society for Microbiology, 1999: 170.

Herpes zoster in non-hospitalized children M. Socan Center for Communicable Diseases, Institute of Public Health, Trubarjeva 2, 1000 Ljubljana, Slovenia Fax: ‡ 386 61 323940 E-mail: [email protected] Accepted 13 November 2000

The varicella-zoster virus (VZV) causes two different clinical presentations: chickenpox and herpes zoster [1,2]. The majority of people have chickenpox in childhood. After chickenpox, the VZV remains dormant in the sensory ganglia for the person's entire life. The mechanism of its maintenance in the dormant

phase remains unclear, as does the cause of its reactivation [3]. After chickenpox, speci®c IgG antibodies remain in the serum, but they do not prevent the reactivation of the virus, i.e. herpes zoster. Cellular immunity is probably more important than humoral immunity in the maintenance of the virus in the

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

92 Clinical Microbiology and Infection, Volume 7 Number 2, February 2001

dormant phase. Diseases impairing this part of the immune system increase the risk of herpes zoster [4]. In 10±20% of the population, VZV is reactivated and spreads along the sensory nerves to the skin, where it causes lesions characteristic of herpes zoster [5]. In adults, herpes zoster most frequently appears in the abdominal or thoracic regions, and in the elderly also as ophthalmic herpes zoster [6]. The clinical presentation of herpes zoster is characteristic and usually requires no laboratory con®rmation. It recurs a second time in 4% of patients, and only in exceptional cases does it recur a third time [7]. Herpes zoster most often occurs without any complications. The most frequent complication is prolonged pain, which can last for several months or even years after the disappearance of skin lesions. Other complications are rare [4]. The incidence of herpes zoster increases with age. According to data from a US study, as many as 30% of patients with herpes zoster were older than 55, even though they accounted for only 8% of the entire population [7]. Childhood herpes zoster is rare and therefore only a small amount of data has been collected. This paper presents our data on reported cases of herpes zoster in childhood. In 1996, herpes zoster was added to the list of reportable communicable diseases in our country. The data were collected at the Center for Communicable Diseases. In the period from January 1996 to December 1998, 126 non-hospitalized cases of herpes zoster were reported in children up to 15 years of age. The physicians reporting these cases were requested to ®ll out a short questionnaire and provide additional data on these children. Additional data requested by physicians were: date of birth, the exact date of varicella and herpes zoster, the localization of herpes zoster, any possible complications, whether or not the child was treated with acyclovir and for how long, the occurrence of varicella during pregnancy, and whether there was a congenital abnormality or any chronic disease (how long the chronic disease was present, and the therapy prescribed). It is presumed that most of the diagnoses of herpes zoster were clinical. Information on whether the case of herpes zoster was laboratory-proven was not required. The response rate was 68.2%, so 86 (35 boys and 51 girls) children remained for the ®nal analysis of the data. The reasons for not complying with our request for more detailed information on reported herpes zoster were not investigated. Herpes zoster appeared throughout the year without a characteristic seasonal peak as is observed in chickenpox. The majority of ill children were aged between 14 and 15, and only a few were younger than 4 years (Figure 1). In one child who had herpes zoster at 14 months, it was found that the mother had had chickenpox during pregnancy. The data on the child's age during chickenpox were stated for 60 children, while for six children the medical ®le only contained a note that they

Figure 1 Age of children with herpes zoster.

had chickenpox, but the exact age was not stated. Ten of the children did not have a clinical history of chickenpox, and for a further 10 there were no data. The interval between chickenpox and the appearance of herpes zoster ranged from less than 1 year to 13 years (Figure 2). In the majority of children, herpes zoster was located in the thoracic region (64 children, or 74.4%); more rarely it was located in the cervical region (10 children, 11.6%) or lumbosacral region (10 children, 11.6%), and in only two (2.4%) children was it located in the facial region. There were no complications, and only 14 (12%) children were treated with acyclovir. The children who had herpes zoster were mainly healthy. Seven children were being treated for asthma (one child with inhaled steroids, others with bronchodilators) and one for toxoplasma panuveitis. One child had alopecia areata and received corticosteroids, and another was treated for rhabdomyosarcoma. Studies analyzing herpes zoster in childhood are extremely rare. They mainly describe the course of herpes zoster in children with malignant diseases who were treated in hospital [8,9]. Only two studies included children with herpes zoster who were mainly treated in outpatient facilities [10,11]. Data on the incidence of herpes zoster in childhood were collected in

Figure 2 Time elapsed between chickenpox and herpes zoster.

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

Concise Communications

the ®rst study. Guess [10] found that 20 in 100 000 children under the age of 5 fall ill, and 63 in 100 000 of persons between the ages of 15 and 19. Herpes zoster was signi®cantly more frequent in children with acute lymphocytic leukemia. According to the data from this study, the risk factor for childhood herpes zoster was chickenpox before the ®rst year of age. Another study [11], which was conducted a few years ago, reported an even higher incidence of herpes zoster; it affects 160 in 100 000 persons younger than 19. Herpes zoster most frequently appeared in the thoracic region. The course of the disease was mainly mild, and none of the patients had post-herpetic neuralgia. A signi®cantly higher number of complications were observed in adult patients [3]. The data we collected do not differ from those of the abovementioned studies. The number of patients was found to increase with age. The interval between chickenpox and herpes zoster was 5.6 years on average. In a study conducted on the population of Iceland, on average slightly more than 8 years passed from chickenpox to the appearance of herpes zoster, but adolescents with herpes zoster aged up to 19 were also included. In all of our patients, the course of the disease was extremely mild and there were no complications. It is therefore disputable whether treatment with acyclovir is justi®ed, apart from exceptional cases (e.g. in severely immunocompromised children). In the children from this study as well, herpes zoster was most often located in the thoracic area and only in one child in the ophthalmic area. In adults, however, thoracic herpes zoster is found in only half of the cases and cranial herpes zoster in over 10% of patients [5]. In children whose mothers had chickenpox during pregnancy, herpes zoster may appear as early as the ®rst year of life. In only one of 86 children did the mother have chickenpox during pregnancy, and herpes zoster appeared very soon after the ®rst year of life. The immune system of the fetus is undoubtedly immature, which, in the opinion of some, is the reason for the early appearance of the disease [2]. The correlation between

93

chickenpox before the ®rst year of age and herpes zoster in children was not analyzed, because the number of children with chickenpox at the age of less than 1 year was insuf®cient to form a sample. Even though herpes zoster is a disease of the elderly, it is not all that rare in children; however, the course of the disease is mainly mild and without sequelae.

R EFER E NCE S 1. Straus SE. Introduction to Herpesviridae. In: Mandell GL, Bennets JE, Dolin R, eds. Principles and Practice of Infectious Diseases, 4th edn, Vol. 2. New York: Churchill Livingstone, 1995: 1330±5. 2. Miller E, Marshall R, Vurdien J. Epidemiology, outcome and control of varicella-zoster infection. Rev Med Microbiol 1993; 4: 222±30. 3. Weller TH. Varicella and herpes zoster. Changing concepts of the natural history, control, and importance of not-so-benign virus. N Engl J Med 1983; 309: 1362±8. 4. Straus SE, Ostrove JM, Inchaupse G et al. Varicella-zoster infections. Biology, natural history, treatment, and prevention. Ann Intern Med 1988; 108: 221±37. 5. Ragozzino MW, Melton MJ, Kurland LT, Chu CP, Pery HO. Population-based study of herpes zoster and its sequelae. Medicine 1982; 61: 310±16. 6. Wutzler P, Doerr HW, von Essen J et al. Epidemiology of herpes zoster. Biotest Bull 1997; 5: 301±6. 7. Donahue JG, Choo PW, Manson JE, Platt R. The incidence of herpes zoster. Arch Intern Med 1995; 155: 1605±9. 8. Mazur MH, Dolin R. Herpes zoster at NIH: a 20-year experience. Am J Med 1978; 65: 738±4. 9. Latif R, Shope TC. Herpes zoster in normal and immunocompromised children. Am J Dis Child 1983; 137: 801±2. 10. Guess HA, Broughton DD, Melton LJ, Kurland LT. Epidemiology of herpes zoster in children and adolescents: a population-based study. Pediatrics 1985; 76: 512±17. 11. Petursson G, Helgason S, Gudmundsson S, Sigurdsson A. Herpes zoster in children and adolescents. Pediatr Infect Dis J 1998; 17: 905±8.

Quality control parameters for cefditoren susceptibility tests P. C. Fuchs, A. L. Barry, S. D. Brown and M. M. Traczewski The Clinical Microbiology Institute, 9725 SW Commerce Circle, Wilsonville, OR 97070, USA 

Tel: ‡1 503 682 3232

Fax: ‡1 503 682 2065

E-mail: [email protected]

Accepted 30 November 2000

Cefditoren is a new expanded-spectrum orally administered cephalosporin currently undergoing clinical trials in the USA

[1]. Its antimicrobial spectrum includes both Gram-positive and Gram-negative species, but it is noteworthy for its particular

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

94 Clinical Microbiology and Infection, Volume 7 Number 2, February 2001

Table 1 Distribution of MICs for six QC strainsa Broth microdilution MICs (mg/L) Organism (ATCC No.)

No. of tests

0.004

0.008

0.016

0.03

0.06

Escherichia coli (25922) Staphylococcus aureus (29213) Streptococcus pneumoniae (49619) H. influenzae (49247) H. influenzae (49766)

594 600 500 300 499

^ ^ ^ ^ 45

^ ^ ^ ^ 424

^ ^ 5 ^ 28

^ ^ 279 ^ 2

^ ^ 245 5 ^

0.12 2 ^ 1 236 ^

0.25

0.5

1.0

2.0

285 11 ^ 59 ^

305 377 ^ ^ ^

2 212 ^ ^ ^

^ 0 ^ ^ ^

a

Number of times each MIC was reported; values in bold are within the proposed QC limits for each strain.

potency against the major respiratory pathogens Haemophilus in¯uenzae and Streptococcus pneumoniae [1,2]. In order to ensure accurate in vitro susceptibility tests with this drug, quality control (QC) parameters are necessary for the standard QC organisms. In the course of our in vitro studies of cefditoren, we have conducted a multilaboratory study to determine the QC parameters for cefditoren when tested by the National Committee for Clinical Laboratory Standards (NCCLS) methods for broth microdilution [3] and disk diffusion [5] susceptibility tests. The design of these studies exceeded the minimum requirements of the NCCLS [4]. Ten North American medical centers participated in each study. Each laboratory performed the broth microdilution and disk diffusion tests as outlined by the NCCLS [3,5]. Broth microdilution trays were prepared to contain serial dilutions of cefditoren with concentrations ranging from 0.004 to 32 mg/L in each of six different lots of cation-adjusted Mueller±Hinton broth (CAMHB) from four different manufacturers for testing non-fastidious strains, and ®ve lots of CAMHB with 3% lysed horse blood for Streptococcus pneumoniae, ®ve lots with Haemophilus Test Medium (HTM) supplements

for H. in¯uenzae ATCC 49766, and three lots for H. in¯uenzae ATCC 49247. These media were freshly prepared and stored at  60 8C until used (up to 14 days). Serial dilutions of cefotaxime or cefuroxime were prepared in one lot of broth as control drugs. Three lots of Mueller±Hinton agar (MHA) from three different manufacturers were used for disk diffusion tests, along with two lots of 5-mg cefditoren disks prepared by two different manufacturers. The MHA was supplemented with 5% sheep blood or HTM supplements for testing S. pneumoniae or H. in¯uenzae. HTM agar was freshly prepared and used within 14 days. Cefotaxime 30-mg disks or ce®xime 5-mg disks served as controls. On each of 10 separate days, all participants inoculated one microdilution tray and three different MHA agar plates for each organism tested. This resulted in a total of 600 disk diffusion zone diameter measurements for each QC strain tested, 600 MIC determinations for each non-fastidious QC strain, 500 MIC determinations for the S. pneumoniae QC strain and H. in¯uenzae ATCC 49766, and 300 MIC determinations for H. in¯uenzae ATCC 49247. Whenever the control drug results were outside the NCCLS range, the corresponding

Table 2 Distribution of zone diameters for six QC strainsa Disk diffusion zone diameters (mm) Organism (ATCC No.)

No. of tests

20

21

22

23

24

25

26

27

EC (25992) SA (25923) SP (49619) HI (49247)

590

^

^

14

76

175

195

89

34

7

10

34

108

188

149

51

23

17

5

590

5

28

29

30

31

32

33

34

35

36

37

^

^

^

^

^

^

^

^

^

^

^

^

^

^

^

^

^

597

^

^

^

^

^

3

5

21

45

77

120

154

98

44

19

11

^

^

576

^

^

^

^

^

11

21

75

106

110

122

66

40

8

8

3

4

2

a

Number of times each zone diameter was reported; values in bold are within the proposed QC limits for each strain. EC, Escherichia coli; SA, Staphylococcus aureus; SP, Streptococcus pneumoniae; HI, Haemophilus influenzae.

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

Concise Communications

95

Table 3 Recommended QC ranges for cefditorena and percentage of results in range MICs (mg/L)

Zone diameters (mm)

Organism (ATCC No.)

Range

% in range

Range

% in range

Escherichia coli (25922) Staphylococcus aureus (29213) Staphylococcus aureus (25923) Streptococcus pneumoniae (49619) Haemophilus influenzae (49247) Haemophilus influenzae (49766)

0.12^1.0 0.25^2.0 NA 0.016^0.12 0.06^0.25 0.004^0.016

100 100 ^ 100 100 99.6

22^28 NA 20^28 27^35 25^34 NR

100 ^ 98.3 98.7 98.3 ^

a

These QC ranges were approved by the NCCLS in June, 1999. NA, not applicable; NR, no recommendations (results off-scale or zones too large).

cefditoren results were omitted from the calculations even though these would not have in¯uenced the cefditoren proposed ranges. Colony counts were performed on the growth control suspension on each test day to ensure appropriate inoculum density. The distribution of MICs and zone diameter results with these control organisms are summarized in Tables 1 and 2. There were no appreciable differences in these distributions when the results were strati®ed by media lot or laboratories. The QC organisms tested, the proposed ranges and the percentage of results falling within these ranges are summarized in Table 3. A four-dilution MIC range was proposed for most strains because the mode fell between two adjacent concentrations that were tested. One laboratory also performed cefditoren agar dilution tests with each QC strain on 15 separate days, and all MIC results fell within these proposed ranges. The MIC QC range for H. in¯uenzae ATCC 49247 was approved by the NCCLS in June 2000; all other QC ranges listed in Tables 1 and 2 for both disk diffusion and dilution tests were approved by the NCCLS subcommittee on antimicrobial susceptibility tests in June 1999. ACK NOW L EDGME N T S This study was supported by a grant from TAP Pharmaceuticals. Portions of this material have been previously presented at the 39th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, USA, 26±29 September 1999, as Abstract 2324. Participants included: M. Bauman, Providence St Vincent Medical Center, Portland, OR, USA; B. Brogden-Torres, University of Massachusetts Medical

Center, Worcester, MA, USA; M. J. Ferraro, Massachusetts General Hospital, Boston, MA, USA; D. Goldman, Children's Hospital and Medical Center, Boston, MA, USA; D. Hardy, University of Rochester Medical Center, Rochester, NY, USA; J. Hindler, UCLA Medical Center, Los Angeles, CA, USA; S. Jenkins, Carolinas Medical Center, Charlotte, NC, USA; C. Knapp, Trek Diagnostic Systems, Westlake, OH, USA; G. Overturf, University of New Mexico Medical Center, Albuquerque, NM, USA; M. Pfaller, University of Iowa, Iowa City, IA, USA; R. Rennie, University of Alberta Hospital, Edmonton, Alberta, Canada; M. Saubolle, Good Samaritan Hospital, Phoenix AZ, USA; and J. Washington, The Cleveland Clinic, Cleveland, OH, USA. R EFER E NCE S 1. Felmingham D, Robbins MJ, Ghosh G et al. An in vitro characterization of cefditoren, a new oral cephalosporin. Drugs Exp Clin Res 1994; 20: 127±47. 2. Jones RN, Biedenbach DJ, Croco MAT, Barrett MS. In vitro evaluation of a novel orally administered cephalosporin (cefditoren) tested against 1249 recent clinical isolates of Haemophilus influenzae, Moraxella catarrhalis, and Streptococcus pneumoniae. Diagn Microbiol Infect Dis 1998; 31: 573±8. 3. National Committee for Clinical Laboratory Standards. Methods of dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 4th edn. Approved Standard M7-A4. Wayne, PA: NCCLS, 1997. 4. National Committee for Clinical Laboratory Standards. Development of in vitro susceptibility testing criteria and quality control parameters. Tentative Guideline, M23-T3. Wayne, PA: NCCLS, 1998. 5. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disk susceptibility tests, 6th edn. Approved Standard M2-A6. Wayne, PA: NCCLS, 1997.

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

96 Clinical Microbiology and Infection, Volume 7 Number 2, February 2001

Campylobacter lari bacteremia 

M. Martinot1,2 , B. Jaulhac2, R. Moog3, S. De Martino2, P. Kehrli4, H. Monteil2 and Y. Piemont2 1

Service de MeÂdecine Interne et des Maladies Infectieuses et Tropicales, HoÃpitaux Universitaires de Strasbourg, 1 place de l'hoÃpital, 67091 Strasbourg, France; 2Institut de BacteÂriologie de la Faculte de MeÂdecine, 3Service de Chirurgie Infantile des HoÃpitaux Universitaires de Strasbourg, and 4Service de Neurochirurgie, HoÃpitaux Universitaires de Strasbourg, Strasbourg, France 

Tel: ‡33 388 53 51

Fax: ‡33 388 64 64

E-mail: [email protected]

Accepted 25 December 2000

The genus Campylobacter includes slender spirally curved Gramnegative rods. Campylobacter spp. are the etiologic agents of worldwide zoonoses and are found in the gastrointestinal tracts of many animals and in the environment, e.g. surface water. Culture of Campylobacter necessitates speci®c conditions for optimal recovery, i.e blood agar incubated in a microaerobic atmosphere containing approximately 5% O2, 10% CO2 and 85% N2 for several days at 42 8C. Among the four species of clinical relevance, Campylobacter jejuni appears to be by far the most important pathogen in humans. It is responsible for diarrhea, sometimes associated with blood in stools, abdominal pain, fever and occasionally vomiting. C. coli causes the same disease but with a lower frequency. C. fetus subsp. fetus is known to cause septicemia and focal infections in immunocompromised patients. C. lari was ®rst isolated from mammalian and avian species, especially seagulls. In 1984, the ®rst case of human disease related to C. lari was reported: fatal bacteremia in an immunocompromised patient with multiple myeloma [1]. Since then, sporadic cases of enteric infection [2] as well as a water-borne outbreak of C. lari enteric infection [3] have been reported. Bacteremia caused by C. lari is exceptional. We report here a case of C. lari bacteremia occurring in a 10year-old girl who underwent an occipital cavernoma excision. The patient was admitted to the neurosurgery department because of a post-radiotherapy occipital cavernoma. Her medical story was marked by the discovery in 1993 of a large cell anaplastic lymphoma, which was treated with chemotherapy and cerebral radiotherapy. The child had been in remission since 1995. In 1999, she experienced occipital partial seizures. Radiologic examinations led to the discovery of an occipital cavernoma. She was treated with valproic acid, and an excision of the cavernoma was performed. The patient received 0.25 mg of tetracosactide (corticotropin), and antibiotic prophylaxis with cloxacillin and amikacin in the immediate postoperative period. On the following day, fever was noted without any diarrhea or clinically evident infection. Hematological results showed 15  109 leukocytes/L with 87% polymorphonuclear

neutrophils, and a C-reactive protein level of 104 mg/L. One of two blood cultures (Bactec plus Aerobic/F, Becton Dickinson, Sparks Maryland, MD) was positive within 48 h for C. lari. Subculture was possible within 24 h. The strain was sensitive to amoxycillin/clavulanate, ticarcillin/clavulanate, gentamicin, fosfomycin, cipro¯oxacin, erythromycin and imipenem, but was resistant to pe¯oxacin, sulfamethoxazole±trimethoprim, piperacillin, piperacillin/tazobactam, amoxycillin, ticarcillin and cephalosporins. Intravenous therapy with amoxicillin and clavulanic acid was then begun in association with amikacin. Evolution was favorable and marked by fever regression within 48 h. One stool culture, carried out several hours after the beginning of therapy, was negative. There was no history of exposure to animals or birds, or recent travel abroad. C. lari infection is an infrequently detected event. In our laboratory (Strasbourg University Hospital), from 1 January 1995 to 31 May 1999, 319 stool cultures were found to be positive for Campylobacter spp. out of a total of 37, 432 stool cultures performed during this period. Among these 319 stool cultures, 292 were positive for C. jejuni (92%), 11 for C. coli (3%), nine for C. lari (3%) and seven for Campylobacter spp. (2%) (C. fetus is not searched for routinely in stools in our laboratory). During the same period, 16 blood cultures were found to be positive for Campylobacter spp.: 10 C. jejuni (63%), ®ve C. fetus (31%) and one C. lari (6%). Bacteremia due to Campylobacter spp. remains infrequent. It seems to be favored by older age, immunocompromised state and serious underlying disease [4±6]. Our experience is in accordance with this, as among the 16 patients with campylobacter bacteremia, 11 were immunocompromised (one leukemia, one lymphoma, two myeloma, one breast cancer, one systemic lupus erythematosus, one rheumatoid arthritis, one cirrhosis, one recent ischemic stroke, two serious postoperative status), and 14 were over 40. Digestive symptoms may be absent or mild. To our knowledge, in addition to our patient, seven other cases of bacteremia due to C. lari have been published (Table 1).

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

Concise Communications

97

Table 1 Clinical characteristics of patients with C. lari bacteremia

Sex Age Medical status

Digestive symptoms C. lari stool culture Antibiotherapy

Evolution

Case 1 (1)

Case 2 (7)

Case 3 Case 4 Case 5 (4) (4) (8)

Case 6 (9)

Case 7 (10)

M 71years Multiple myeloma Chronic renal failure No NA Vancomycin Gentamicin Moxalactam Penicillin G Death

M 22 years AIDS

NA ^ Yesa

NA ^ Yesa

Neonate 25 days None

M 42 years None

F 83 years Pacemaker Coronary artery bypass

Yes No NA

Yes NA NA

No NA NA

Yes Yes Erythromycin Netilmicin

Yes No Erythromycin

Death

NA

NA

Good

Good

Case 8 This report

F 10 years Lymphoma Occipital cavernoma Corticotropin No No No No Ciprofloxacin/Vancomycin Amoxicillin^clavulanic Then acid Amikacin Imipenem^gentamicin Pacemaker removal Good Good

a

Underlying diseases mentioned by the authors without further description. ND, no data provided.

Four patients had an immunocompromised state: one patient with myeloma [1], one patient with AIDS [7] and two patients with underlying diseases not speci®ed [4]. Two other cases occurred in healthy patients in association with gastroenteritis: a 25-day-old neonate [8] and an adult [9]. The last case occurred in a previously healthy woman who experienced recurrent C. lari bacteremia with pacemaker infection [10]. A few case reports of other extraintestinal infections due to C. lari, such as urinary [11] or pleural [12] infections, have also been described. We report here the eighth case of C. lari bacteremia. Invasion of blood from an intestinal focus has been suspected, although the patient had no diarrhea before bacteremia. The patient was not immunosuppressed, but she had undergone surgical cavernoma excision and received a single dose of corticotropin before the beginning of infection. C. lari is infrequently found in humans but may occasionally lead to intestinal disease and exceptionally to bacteremia, particularly in immunocompromised patients. R EFER E NCE S 1. Namamkin I, Stowell C, Skalina D, Jones AM, Hoop RM, Smibert M. Campylobacter laridis causing bacteremia in an immunocompromised host. Ann Intern Med 1984; 101: 55±7. 2. Tauxe RV, Patton CM, Edmonds P, Barret TJ, Brenner DJ, Blake PA. Illness associated with Campylobacter laridis, a new recognised Campylobacter species. J Clin Microbiol 1985; 21: 222±5.

3. Broczy A, Thomson S, Smith D, Lior H. Water-borne outbreak of Campylobacter laridis-associated gastroenteritis. Lancet 1987; 1: 164±5. 4. Skirrow MB, Jones DM, Sutcliffe E, Benjamin J. Campylobacter bacteremia in England and Wales, 1981±1991. Epidemiol Infect 1993; 110: 567±73. 5. Schonheyder HC, Sogaard P, Frederiksen W. A survey of Campylobacter bacteremia in three Danish counties, 1989 to 1994. Scand J Infect Dis 1995; 27: 145±8. 6. Font C, Cruceta A, Moreno A et al. A study of 30 patients with bacteremia due to Campylobacter. spp. Med Clin 1997; 108: 336±40. 7. Dionisio D, Milo D, Mazzotta D, Cecile P. Campylobacter laridis bacteremia in an AIDS patient. Boll Ist Sieroter Milan 1989; 68: 199±200. 8. Cheng-hsun C, Chii-Yuh KK, Jonathan T. Chronic diarrhea and bacteremia caused by Campylobacter lari in a neonate. Clin Infect Dis 1995; 21: 700±1. 9. Soderstrom C, Schalen C, Walder M. Septicemia caused by an unusual Campylobacter species (C. laridis and C. mucosalis). Scand J Infect Dis 1991; 23: 369±71. 10. Morris CN, Scully B, Garvey GJ. Campylobacter lari associated with permanent pacemaker infection and bacteremia. Clin Infect Dis 1998; 27: 220±1. 11. Bezian MC, Ribou G, Barberis-Giletti C, Megraud F. Isolation of a urease positive thermophilic variant of Campylobacter lari from a patient with urinary tract infection. Eur J Clin Microbiol Infect Dis 1990; 9: 895±7. 12. Bruneau B, Burc L, Bizet C, Lambert-Zechovsky N, Branger NC. Purulent pleurisy caused by Campylobacter lari. Eur J Clin Microbiol Infect Dis 1998; 17: 185±8.

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

98 Clinical Microbiology and Infection, Volume 7 Number 2, February 2001

Evolution of TEM b-lactamase genes identified by PCR with newly designed primers in Korean clinical isolates S. H. Lee1, S. H. Jeong2 and K. J. Lee3 1

Department of Genetic Engineering, Youngdong University, Chungbuk 370-701, 2Department of Internal Medicine and Clinical Pathology, College of Medicine, Kosin University, Pusan, and 3Department of Microbiology, College of Natural Sciences, Research Center for Molecular Microbiology, Seoul National University, Seoul, South Korea 

Tel: ‡ 82 43 740 1112

Fax: ‡ 82 43 740 1109

E-mail: [email protected]

Accepted 25 October 2000

Among Gram-negative pathogens in Korea, the incidence of resistance to extended-spectrum b-lactam antibiotics is becoming an ever-increasing problem in nosocomial infections [1]. The single most prevalent mechanism responsible for resistance to b-lactam antibiotics among clinical isolates of the family Enterobacteriaceae is the production of extended-spectrum b-lactamases (ESBLs). We designed 16 primer pairs from most b-lactamase genes belonging to four molecular classes [2] (classes A, B, C and D) of ESBL to differentiate the genes encoding ESBLs and to detect a new ESBL gene. The BLASTN program at the National Center for Biotechnology Information was used for database searches of ESBL genes, and the Clustal W program was used to align multiple nucleotide sequences. Nucleotide sequence accession numbers taken from GenBank, EMBL or DDBJ databases, and reference reports, are available on our website: http://www.inticity.com/ ulg07100. The primers (Table 1) were designed by selecting identical sequences in multiple nucleotide alignments of ESBL genes and using the Primer Calculator program (Williamstone Enterprises, Waltham, MA, USA). The electrophoretic analysis of polymerase chain reaction (PCR) products of reference genes (blaTEM-1a, blaSHV-12, blaOXA-2, blaMOX-1, blaCMY-2, blaMIR-1, blaIMP-1, blaIMI-1, and blaToho-1) showed that the fragment sizes of PCR products were identical to those indicated in Table 1. Using the criteria for the reduced susceptibility or resistance to oxyimino-cephalosporins (cefotaxime and ceftazidime) and monobactams (aztreonam) and con®rmation by double disk test, we detected ®ve Escherichia coli and two Klebsiella pneumoniae clinical isolates producing ESBLs from January 1999 to March 1999 in Kosin Medical Center, South Korea. The clinical isolates were recovered from urine (four strains), exudates (two strains), or sputum samples (one strain). Genotypes of ESBLs determined by PCR using plasmid DNA as a template with 16 primer pairs included seven TEM-derived and seven SHV-derived ESBLs. The seven clinical isolates yielded positive

results with TEM- and SHV-speci®c primers. Nucleotide sequence analysis (automatic sequencer 373A; Applied Biosystems, Weiterstadt, Germany) of PCR products revealed that blaTEM-1b and blaSHV-12 were the dominant types of b-lactamase gene. In addition, we also identi®ed blaTEM-52, blaSHV-5, and a new ESBL gene. The blaTEM-17 gene was only identi®ed in Capnocytophaga ochracea CIP 105321, a capnophilic gramnegative fusiform rod with gliding motility [4]. Because a new ESBL gene ®rst detected in E. coli K992740-1 is different from blaTEM-17 at seven positions (silent point mutations in the coding region: C226 ! T, C436 ! T, A469 ! T, G604 ! T, C682 ! T, T863 ! C, and T985 ! C), we named it blaTEM-17b. The nucleotide sequence of blaTEM-17b has been submitted to GenBank and assigned accession number AF264753. PCR of blaTEM-17b containing promoter and whole coding regions was carried out with T10 primer (50 -CAATAACCCTGGTAAATGCT-30 ) and T20 primer (50 -TTACCAATGTTTGGTAAGGGA-30 ). Nucleotide sequence analysis of the PCR product showed that four mutations at positions (according to the numbering scheme of Sutcliffe [5]) 175, 226, 436 and 604 were the same as those of blaTEM-1b (Tn-2 type) [6]. The results suggest that blaTEM-17b is harbored in a Tn-2 type transposon. The strain containing blaTEM-17b was resistant to ceftazidime (MIC 32 mg/L), aztreonam (MIC 64 mg/L), and amoxycillin (MIC 512 mg/L). The strain had intermediate resistance to cefotaxime (MIC 16 mg/L) and was susceptible to cefotetan (MIC 1 mg/L), cefoxitin (MIC 16 mg/ L), and imipenem (MIC 0.25 mg/L). The MIC pro®le of E. coli K992740-1 was similar to that of C. ochracea CIP 105321 [4]. The blaTEM-17b gene presents the single Glu104 ! Lys amino acid substitution. The blaTEM-52 gene presents Glu104 ! Lys, Met182 ! Thr and Gly238 ! Ser amino acid substitutions. Therefore, the detection of blaTEM-17b suggests that the evolution of blaTEM-52 from those genes identi®ed from Korean clinical isolates would seem to be

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

Concise Communications

99

Table 1 Nucleotide sequences of the oligonucleotides used for amplification Gene and primer namea blaTEM T1 T2 T3 T4 blaSHV S1 S2 S3 S4 blaOXA O(10)1 O(10)2 O(2)1 O(2)2 blaCMY-1, blaFOX, and blaMOX C1 C2 C3 C4 blaCMY, blaAmpC[C], blaLAT, and blaBIL CA1 CA2 CA3 CA4 blaMIR, blaAmpC[E], and blaACT M1 M2 M3 M4 blaIMP I1 I2 blaIMI, blaNMC, and blaSme IN1 IN2 blaCTX-M and blaToho CT1 CT2 blaPER P1 P2

Expected size (bp) of PCR product (primer pair)

Sequenceb

Nucleotide positionc

50 -AGAGTATGAGTATTCAACATT-30 50 -ATCTCAGCGATCTGTCTAT-30 50 -AGTCACAGAAAAGCATCTTA-30 50 -ATAGTTTGCGCAACGTTGT-30

204 1041 523 782

50 -GGGTTATTCTTATTTGTCGCT-30 50 -TAGCGTTGCCAGTGCTCG-30 50 -AGATCCACTATCGCCAGCA-30 50 -TCATTCAGTTCCGTTTCCCA-30

58 987 393 624

50 -GTTCAATTACAGAAAATACGT-30 50 -TTAGCCACCAATGATGCC-30 50 -GTTTTCCGATGGGACGG-30 50 -ACCCATCCTACCCACCA-30

180 912 980 1361

50 -GAGCAGACCCTGTTCGAGAT-30 50 -GATTGGCCAGCATGACGATG-30 50 -TACTCCAACCCCAGCATAGG-30 50 -CCACATAGGCGCCAAAGCC-30

570 1416 852 1371

50 -TGCTGCTGACAGCCTCTTTC-30 50 -TTTCAAGAATGCGCCAGGCC-30 50 -GCGATCCGGTCACGAAATAC-30 50 -ATAACGCTGGATTTCACGCCA-30

71 1177 359 785

50 -CTATAAGTAAAACCTTCACCGG-30 50 -TATGCCGCCTCAACGCGTG-30 50 -TGCGCTTTTATCAAAACTGGCA-30 50 -GCCACGTAGCTGCCAAACC-30

1178 2048 1382 1967

585 (M3 and M4)

50 -CTACCGCAGCAGAGTCTTTG-30 50 -ACAACCAGTTTTGCCTTACCAT-30

494 1082

588 (I1and I2)

50 -ATGTCATTAGGTGATATGGCT-30 50 -GCATAATCATTTGCCGTACC-30

492 889

397 (IN1and IN2)

50 -ATCTGACGCTGGGTAAAGC-30 50 -ATATCGTTGGTGGTGCCATA-30

646 808

162 (CT1and CT2)

50 -GATTTGTTATTTGAACTGGT-30 50 -TGAACCTAATAACTGCATAA-30

729 1326

597 (P1and P2)

837 (T1and T2) 259 (T3 and T4)

929 (S1and S2) 231 (S3 and S4)

732 (O(10)1and O(10)2) 381 (O(2)1and O(2)2)

846 (C1and C2) 519 (C3 and C4)

1106 (CA1and CA2) 426 (CA3 and CA4)

870 (M1and M2)

a Primers are consensus sequences of the bla genes taken from GenBank, EMBL, DDBJ database, or reference report (in the text). bPrimersT1,T3, S1, S3, O(10)1, O(2)1, C1, C3, CA1, CA3, M1, M3, I1, IN1, CT1and P1are identical to the leading strand; primersT2,T4, S2, S4, O(10)2, O(2)2, C2, C4, CA2, CA4, M2, M4, I2, IN2, CT2 and P2 are identical to the lagging strand. cNumbers correspond to the position of the first 50 base of each oligonucleotide according to the numbering of the nucleotide sequences of V01119 (blaTEM-1a) forT1,T2,T3 and T4; M59181 (blaSHV-1) for S1, S2, S3 and S4; U37105 (blaOXA-10) for O(10)1and O(10)2; X07260 (blaOXA-2) for O(2)1and O(2)2; X92508 (blaCMY-1) for C1, C2, C3 and C4; blaAmpC[C] of Citrobacter freundii [3] for CA1, CA2, CA3 and CA4; M37839 (blaMIR-1) for M1, M2, M3 and M4; S71932 (blaIMP-1) for I1and I2; U50278 (blaIMI-1) for IN1and IN2; X92506 (blaCTX-M-1) for CT1and CT2; Z21957 (blaPER-1) for P1and P2.

blaTEM-1b ! blaTEM-17b ! a novel b-lactamase gene (Glu104 ! Lys plus Met182 ! Thr or Met182 ! Thr plus Gly238 ! Ser) ! blaTEM-52, although the novel b-lactamase gene has not been identi®ed (http://www.lahey.org/studies/

webt.htm). Furthermore, blaTEM-1b (E. coli K992806-2), blaTEM-17b (E. coli K992740-1) and blaTEM-52 (K. pneumoniae K992806-3) were detected in the different samples of one patient (1-year-old baby). E. coli K992806-2 and E. coli

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

100 Clinical Microbiology and Infection, Volume 7 Number 2, February 2001

K992740-1 had intermediate resistance to cefotaxime (MIC 16 mg/L), whereas K. pneumoniae K992806-3 was resistant to cefotaxime (MIC 64 mg/L). Three strains were resistant to ceftazidime and aztreonam, but the MICs were higher in E. coli K992806-2, E. coli K992740-1, and K. pneumoniae K992806-3, in that order. Sequence analysis revealed that blaTEM-17b was more similar to blaTEM-1b than to blaTEM-1a, and four of seven silent mutations of blaTEM-17b were the same as those of blaTEM-52 (at positions 469, 682, 863 and 985). Our ®ndings from the same patient appear to represent the in vivo evolution of b-lactamase genes (from blaTEM-1b to blaTEM-17b and from blaTEM-17b to blaTEM-52) under the selective pressure of antimicrobial therapy (especially cefotaxime), as was the case with blaSHV-8 [7]. ACK NOW L EDGME N T S This work was supported by a research grant of the Korea Science and Engineering Foundation (1999-2-211-002-5).We would like to thank J. G. Sutcliffe, H. HaÈchler, T. Horii, G. A. Jacoby and Y. Ishii for the generous gift of reference b-lactamase genes.

R EFER E NCE S 1. Kim J, Kwon Y, Pai H, Kim J, Cho D. Survey of Klebsiella pneumoniae strains producing extended-spectrum b-lactamases: prevalence of SHV-12 and SHV-2a in Korea. J Clin Microbiol 1998; 36: 1446±9. 2. Bush K, Jacoby GA, Medeiros AA. A functional classification scheme for b-lactamases and its correlation with molecular structure. Antimicrob Agents Chemother 1995; 39: 1211±33. 3. Lindberg F, Normark S. Sequence of the Citrobacter freundii OS60 chromosomal ampC b-lactamase gene. Eur J Biochem 1986; 156: 441±5. 4. Rosenau A, Cattier B, Gousset N, Harriau P, Philippon A, Quentin R. Capnocytophaga ochracea; characterization of a plasmidencoded extended-spectrum TEM-17 b-lactamase in the phylum Flavobacter-Bacteroides. Antimicrob Agents Chemother 2000; 44: 760±2. 5. Sutcliffe JG. Nucleotide sequence of the ampicillin resistance gene of Escherichia coli plasmid pBR322. Proc Natl Acad Sci USA 1978; 75: 3737±41. 6. Goussard S, Courvalin P. Sequence of the genes blaT-1B and blaT-2. Gene 1991; 102: 71±3. 7. Rasheed JK, Jay C, Metchock B. Evolution of extended-spectrum b-lactam resistance (SHV-8) in a strain of Escherichia coli during multiple episodes of bacteremia. Antimicrob Agents Chemother 1997; 41: 647±53.

Rhodotorula septicemia: case report and minireview 

V. Petrocheilou-Paschou1,2 , H. Prifti2, E. Kostis1, C. Papadimitriou1, M. A. Dimopoulos1 and S. Stamatelopoulos1 1

Department of Clinical Therapeutics, University of Athens School of Medicine, `Alexandra' Hospital, 80 V. So®as Avenue, 11528 Athens, and 2Department of Microbiology, `Alexandra' Hospital, Athens, Greece 

Tel: ‡ 301 7488941

Fax: ‡ 301 81313

E-mail: [email protected]

Accepted 26 November 2000

Rhodotorula, a yeast member of the family Cryptococcaceae, was only recently recognized as a human pathogen especially affecting immunocompromised patients [1]. To our knowledge, over the last 40 years, no more than 40 cases of rhodotorula infections have been reported in patients with central indwelling vascular catheters. We isolated Rhodotorula rubra from both the catheter and the peripheral venous blood of a patient on azole prophylaxis. On this occasion, we reviewed the literature and the limited experience with rhodotorula hospital infections in Greece. A 21-year-old female patient, with non-Hodgkin's lymphoma in second remission, underwent autologous peripheral blood progenitor cell transplantation. A large peripheral double lumen catheter was placed in the subclavian vein for the collection and reinfusion of stem cells. The conditioning regimen was given on days 5 to 1. The patient received prophylactic oral cipro¯oxacin 500 mg twice daily

and ¯uconazole 200 mg/day. Because of severe vomiting and gastrointestinal mucositis, total parenteral nutrition was mandatory. One day after stem cell reinfusion (day 1), the patient developed grade 4 neutropenia on the WHO toxicity scale (PMN count <100/mL), and 3 days later (9 days after insertion of the catheter) she became febrile. Blood cultures were obtained from both the catheter sites as well as from the antecubital vein. Cipro¯oxacin was discontinued, and intravenous ceftazidime, vancomycin and amikacin were started. Microscopic examination of all blood cultures revealed budding yeasts. Blood cultures were performed with the continuously monitored non-invasive system BACTEC 9240 (Becton Dickinson Medical Systems, Rutherford, NJ, USA). Blood specimens were inoculated in BACTEC plus F aerobic and anaerobic bottles (Becton Dickinson Microbiology Systems, Sporks, Maryland, USA) and were incubated for a standard 7-day

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

Concise Communications

Figure 1 Culture of blood, on Sabouraud^dextrose agar, that yielded Rhodotorula rubra.

protocol or until the instrument gave a positive signal. The yeasts recovered from blood cultures were subcultured on Sabouraud±dextrose agar and incubated at 35 8C for 48 h; they were identi®ed as Rhodotorula rubra according to colony morphology and coral red pigment production (Figure 1), microscopic characteristics on Gram stain, biochemical reactions obtained by API ID 32C (BioMerieux, I'Etoile, France), as well as urea production, carbon assimilation and fermentation. Antifungal susceptibility testing was performed using the broth microdilution system `Yeast one' sensititre (AccuMed International Limited, East Grinstead, West Sussex, UK). The MIC values (mg/L) of the following antifungal drugs against the Rhodotorula isolate were: amphotericin B 0.25, ¯uconazole >256, ketoconazole 0.25, 5-¯ucytosine 0.06, and itraconazole 1. Candida parapsilosis ATCC 90018 was also tested as a control, and MICs of the above antifungal agents (mg/L) were 0.25, 1, 0.016, 0.06 and 0.06, respectively. Fluconazole was discontinued and intravenous liposomal amphotericin B was started with a daily dose of 3 mg/kg of body weight. However, 3 days later, the patient remained febrile, and blood cultures were again positive for the same fungus, despite a PMN count of more than 1000/mL. Eventually, the central catheter was removed, and the following day the patient became afebrile. Rhodotorula was again isolated from the tip of the removed catheter. Following the removal of the catheter, treatment with liposomal amphotericin B was continued for a total of 7 days and the patient was discharged from the hospital, after complete hematologic recovery, without further complications.

101

Rhodotorula was traditionally considered to be a contaminant, but within the last decade it has been progressively recognized that this yeast can cause infections, especially in immunocompromised patients with central venous catheters [1,2]. Rhodotorula rubra has been regarded as having lower virulence (about 15% mortality) in comparison with Candida spp. and Trichosporon spp., as has been reported by Krcmery et al in a recent review [3]. It is thus an infrequent cause of infection, with only 12 culture-proven rhodotorula fungemia cases reported prior to 1984 [2]. More speci®cally, rhodotorula infection appears to represent a rare complication of bone marrow transplantation, with only one case reported in series of 1186 bone marrow recipients between 1974 and 1989 [4]. In 1992, 23 catheterrelated cases of sepsis caused by this yeast were described in cancer patients from one center [2], and since then another ®ve patients with fungemia have been reported in the Englishlanguage literature [5±9]. In our country, cases of sepsis due to Rhodotorula are even rarer. This is the only case documented in our hospital in the last 15 years. A case of fungemia following surgery for colon carcinoma was recently reported from Greece [10], and another one was included in a surveillance study conducted among several hospitals in Athens during the last 2 years [11]. In the latter study, Rhodotorula was isolated from a quadriplegic child. Most of the reported infections were documented in cancer patients with indwelling vascular catheters [2,5]; four of them were neutropenic [2,5,9] and only two had positive blood cultures from both catheter and peripheral vein [2,8]. Our patient had a lymphoma, was severely neutropenic and had a central venous catheter placed for stem cell collection and reinfusion, as well as for total parenteral nutrition. Our patient also had positive blood cultures obtained from the catheter, as well as from a peripheral vein. As has been stressed by Kiehn et al [2], Rhodotorula usually remains in the lumen of the catheter, and blood cultures from peripheral veins are usually negative. Severe neutropenia is the most likely cause of this dissemination, as Fanci et al [7] suggested in their case report. Central vein catheter has been recognized as a major risk factor for rhodotorula fungemia. In the majority of the reported cases, indwelling catheters had been placed 1±22 months prior to fungemia. Our patient developed rhodotorula sepsis very early, only 9 days after the catheter insertion. Although catheter insertion is a speci®c risk factor for Rhodotorula fungemia, severe neutropenia and total parenteral nutrition combined possibly played a supplementary role in the development of the septicemia [3]. The treatment of catheter-related Rhodotorula fungemia remains controversial. Some authors believe that catheter removal is not necessary for cure [1,2]. On the other hand, in a small number of cases, catheter removal without antifungal therapy was also curative. Our patient was febrile, and the cultures still yielded Rhodotorula, 3 days after amphotericin B introduction, while she became afebrile 1 day after removal of

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102

102 Clinical Microbiology and Infection, Volume 7 Number 2, February 2001

the catheter. Treatment with amphotericin B was continued for a further 7-day course, as Kiehn et al [2] recommended in their report. Our case strengthens the point of view that catheter removal and administration of antifungal therapy appears to be the optimal treatment. Liposomal amphotericin B was chosen, because it is generally better tolerated than the amphotericin B deoxycholate. Our patient became fungemic despite ¯uconazole prophylaxis, and in vitro susceptibility testing performed with antifungal drugs revealed that the Rhodotorula isolate was highly resistant to ¯uconazole, as has also been reported by others [10,12]. Regarding prophylactic treatment of immunocompromised patients with ¯uconazole, it should be remembered that widespread antifungal prophylaxis, although providing protection against sensitive organisms, also may allow the emergence of more resistant isolates such as Rhodotorula spp. R EFER E NCE S 1. Vartivarian SE, Annaissie EJ, Bodey GP. Emerging pathogens in immunocompromised patients: classification, diagnosis and management. Clin Infect Dis 1993; 17(suppl 2): S487±91. 2. Kiehn TE, Gorey E, Brown AE, Edwards FF, Armstrong D. Sepsis due to Rhodotorula related to use of indwelling central venous catheters. Clin Infect Dis 1992; 14: 841±6. 3. Krcmery V, Krupova I, Denning DW. Invasive yeast infections other than Candida spp. in acute leukemia. J Hosp Infect 1999; 41: 181±94.

4. Morrison VA, Haake RJ, Weisdorf DJ. The spectrum of noncandida fungal infections following bone marrow-transplantation. Medicine 1993; 72: 78±89. 5. Sheu MJ, Wang CC, Wang CC, Shi WJ, Chu ML. Rhodotorula septicemia: report of a case. J Formos Med Assoc 1994; 93: 645±7. 6. Goldani LZ, Craven DE, Sugar AM. Central venous catheter infection with Rhodotorula minuta in a patient with AIDS taking suppressive doses of fluconazole. J Med Vet Mycol 1995; 33: 267±70. 7. Fanci R, Pecile P, Martinez RL, Fabbri A, Nicoletti P. Amphotericin B treatment of fungemia due to unusual pathogens in neutropenic patients: report of two cases. J Chemother 1997; 9: 427±30. 8. Marinova I, Szabadosova V, Brandeburova O, Krcmery VJ. Rhodotorula spp fungemia in an immunocompromised boy after neurosurgery successfully treated with miconazole and 5flucytosine: case report and review of the literature. Chemotherapy 1994; 40: 287±9. 9. Lui AY, Turett GS, Karter DL, Bellman PC, Kislak JW. Amphotericin B lipid complex in an AIDS patient with Rhodotorula rubra fungemia. Clin Infect Dis 1998; 27: 892±3. 10. Papadogeorgakis H, Frangoulis E, Papaefstathiou C, Katsambas A. Rhodotorula rubra fungemia in an immunosuppressed patient. J Eur Acad Dermatol Venereol 1999; 12: 169±70. 11. Paniara O, Frangoulis E, Stylianea-Foundoulaki A, Danielidis V, Velegraki A, the Hellenic group for the study of fungal infections. Epidemiological survey on fungal infection in Europe. Acta Microb Hell 1999; 44: 573±81. 12. Galan-Sanchez F, Garcia-Martos P, Rodriguez-Ramos C, MarinCasanova P, Mira-Gutierrez J. Microbiological characteristics and susceptibility patterns of strains of Rhodotorula isolated from clinical samples. Mycopathologia 1999; 145: 109±12.

ß 2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 7, 88±102