Selective Medium for Isolation of Bacteroides gracilis

Transcription

1 JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 1990, p /90/ $02.00/0 Copyright 1990, American Society for Microbiology Vol. 28, No. 8 Selective Medium for Isolation of Bacteroides gracilis KYUNGWON LEE,' ELLEN JO BARON,2'3 PAULA SUMMANEN,2'3 AND SYDNEY M. FINEGOLD3.4.5* Department of Clinical Pathology, Wonju Medical College, Yonsei University, Wonju, Koreal; Clinical Anaerobic Bacteriology Research Laboratory2 and Research and Medical Services,3 Veterans Administration Medical Center, West Los Angeles, California 90073; and Department of Medicine4 and Department of Microbiology and Immunologys University of California at Los Angeles School of Medicine, Los Angeles, California Received 18 December 1989/Accepted 22 May 1990 A new medium selective for Bacteroides gracilis was developed. The medium is tryptic soy agar (Difco Laboratories, Detroit, Mich.) containing nalidixic acid, teicoplanin, sodium formate, sodium fumarate, and potassium nitrate. Ail 18 strains of B. gracilis tested grew with only minimal inhibition. Most of the other 214 organisms tested, including most Bacteroides species, other anaerobes, and a substantial number of facultative anaerobes, were significantly inhibited by the medium. In a diagnostic study of 49 clinical specimens (28 patients with intra-abdominal infection, mostly gangrenous or perforated appendicitis), four strains of B. gracilis were isolated (from 4 different patients) on B. gracilis selective agar but were not detected on standard media. Bacteroides gracilis is an obligately anaerobic gram-negative, asaccharolytic, nonmotile, agar-pitting bacillus found in oral and dental infections, brain abscesses, pulmonary infections, peritonitis, intra-abdominal abscesses, wound infections, and soft tissue abscesses (10, 14). Since B. gracilis has been recovered primarily from polymicrobial infections (9), the pathogenic potential of this organism itself remains uncertain. The frequency of association of this species with serious anaerobic infections, however, implies virulence; moreover, many I-lactam and other drugs have relatively poor in vitro activity against B. gracilis (10). The organism grows slowly and forms small, translucent colonies and therefore can easily be overlooked in mixed cultures; for this reason, development of a selective medium for specific recovery of B. gracilis would be of particular value. We have developed such a medium, designated B. gracilis selective agar (BGSA). It contains tryptic soy agar base, formate, fumarate, nitrate, and two selective agents, nalidixic acid and teicoplanin. MATERIALS AND METHODS Bacterial strains. Two hundred thirty-two strains of bacteria were used in this study (Tables 1 and 2). These included 18 strains of B. gracilis; 112 strains of other anaerobic, gram-negative rods; 11 strains of anaerobic, gram-negative cocci; 10 strains of anaerobic gram-positive cocci, 14 strains of anaerobic, gram-positive rods; and 67 strains of various aerobic or facultative bacteria. All anaerobic isolates, except the reference strain of B. gracilis (ATCC 33236), were from the Wadsworth Anaerobic Bacteriology Laboratory collection and were kept frozen in 20% skim milk at -70 C until use. These organisms had been identified by the methods outlined in the Virginia Polytechnic Institute and Wadsworth anaerobic laboratory manuals (7, 13). Aerobic and facultative isolates were from the hospital clinical microbiology laboratory at the Veterans Administration Wadsworth Medical Center. They were identified by using Vitek, API 20E (Analytab Products, Plainview, N.Y.), and conventional methods (1, 11) and were maintained by weekly subculture * Corresponding author. on commercially prepared brucella blood agar without cysteine (BBL Microbiology Systems, Cockeysville, Md.). Media. BGSA was prepared by rehydrating tryptic soy agar base (Difco Laboratories, Detroit, Mich.), autoclaving it for 15 min at 121 C, cooling it to 56 C, and adding nalidixic acid (Sterling Winthrop, Rensselaer, N.Y.) and teicoplanin (Merrell Dow, Cincinnati, Ohio) to final concentrations of 8 and 32 mg/liter, respectively. Filter-sterilized supplements were also added (final concentrations of 3 g of sodium formate per liter, 3 g of sodium fumarate per liter, and 2 g of potassium nitrate per liter). Qualitative studies. Bacterial suspensions equivalent in turbidity to a McFarland turbidity standard of 0.5 were prepared in thioglycolate medium with no indicator (135C; BBL) supplemented with vitamin K (0.1 mg/liter) and hemin (5 mg/liter). A Steers-Foltz replicator was used to inoculate test strains onto BGSA and unsupplemented brucella blood agar (BBL), which served as a control. Plates were incubated in GasPak jars (BBL) at 35 C for 48 h. Quantitative and morphologic studies. Bacterial suspensions were prepared in thioglycolate medium as described above; 0.1 ml of a 106 dilution in thioglycolate medium was spread over the surface of unsupplemented brucella blood agar and BGSA with glass rods. Plates were incubated in GasPak jars at 35 C for 48 h, and colonies were counted. Colonies of different morphologies (as visualized under x50 magnification) were Gram stained and described. Isolation of B. gracilis from simulated mixed oral flora. Three strains of B. gracilis and species of common oral flora (Fusobacterium nucleatum, B. oralis, pigmented bacteroides, porphyromonads, peptostreptococci, and alpha-hemolytic streptococci) were diluted individually in thioglycolate medium to give final concentrations of 104 CFU of B. gracilis per ml and 108 CFU of each of the other species per ml. One milliliter of each of the diluted oral flora cultures was transferred to each of three separate test tubes. One milliliter of one of the B. gracilis strains was added to each mixture, making three mixed suspensions (7 ml each). After mixing, 0.1 ml of each suspension was placed onto BGSA and onto brucella agar with 5% sheep blood, 3 g each of formate and fumarate per liter, and 2 g of nitrate per liter (nonselective supplemented agar prepared in house); the 0.1 ml was then Downloaded from on March 7, 2019 by guest 1747

2 1748 LEE ET AL. TABLE 1. Growth of anaerobic organisms on BGSA No. of No. of strains Organism(s) strains inhibited( Bacteroides gracilis 18 0 (0) Bacteroides ureolyticus il 4 (36) Wolinella sp (20) Bacteroides fragilis group (100) Pigmented Bacteroides sp. and 30 30,(100) Porphyromonas sp.c Nonpigmented Bacteroides sp.d il il (100) Fusobacterium sp.e 19 4f (21) Vèillonella sp. il 3 (27) Peptostreptococcus sp.y (100) Clostridium sp.' il il (100) Nonsporeforming gram-positive rods' 3 3 (100) a No growth detectable. b B. caccae, 3;,. distasonis, 5; B. eggerthii, 3; B. fragilis, 5; B. ovatus, 4; B. stercoris, 2; B. thetaiotaomicron, 2; B. uniforms, 5; and B. vulgatus, 2. c B. bivius, 5; B. corporis, 3; B. denticola, 4; B. intermedjus, 6; B. melaninogenicus, 3; other pigmented Bacteroides sp., 1; Porphyromonas asaccharolytica, 1; P. endodontalis, 2; P. gingivalis, 4; and Porphyromonas Sp., 1. d B. buccae, 1; B. oralis, 6; B. oris, 2; B. splanchnicus, 2. e F. mortiferum, 3; F. necrophorum, 1; F. nucleatum, 10; F. varium, 4; and other Fusobacterium sp., 1. f Four strains of F. nucleatum. g Peptostreptococcus anaerobius, 1; P. asaccharolytica, 1; P. magnus, 1; P. micros, 1; P. prevotii, 3; and P. tetradius, 3. h Ciostridium butyricum, 1; C. cadaveris, 1; C. clostridijforme, 1; C. difficile, 1; C. perfringens, 1; C. ramosum, 1; C. septicum, 1; C. sordellii, 1; C. sporogenes, 1; and C. tertium, 2. 'Eubacterium sp., 1; Lactobacillus sp., 1; and Propionibacterium sp., 1. streaked in the usual five-sector manner to obtain isolated colonies. Plates were incubated in GasPak jars (BBL) at 35 C for 48 h. (BBL states that 90 min after GasPak envelopes are activated by addition of water there is a residual hydrogen concentration of 25% in an anaerobic jar set up in this manner. Since hydrogen may serve as an electron donor, formate and fumarate may be unnecessary in media incubated in such jars.) The amount of growth of each species was determined semiquantitatively by examining the streaked plates. (Growth in only the first [inoculum]istreak area was graded 1+; growth throughout all five sectors was graded 5+.) Isolation of B. gracilis from saliva. Seven unstimulated TABLE 2. Growth of aerobic and facultative organisms on BGSA No. of No. of strains Organism(s) strains inhibited (%) Alpha-hemolytic streptococci (100) Enterococcus sp. 3 3 (100) Staphylococcus aureus 3 3 (100) Acinetobacter calcoaceticus subsp. 2 2 (100) anitratus Citrobacter sp.b 3 0 (0) Enterobacter sp.c 8 2 (25) Escherichia coli (93) Hafnia alvei 1 1 (100) Klebsiella sp.d 10 5 (50) Proteus mirabilis 3 1 (33) Pseudomonas aeruginosa 3 0 (0) Serratia marcescens 3 3 (100) No growth detectable. b C. diversus, 2; C. freundii, 1. C Enterobacter aerogenes, 1; E. cloacae, 7. d Klebsiella oxytoca, 1; K. pneumoniae, 9. J. CLIN. MICROBIOL. human saliva specimens were diluted in prereduced sterile distilled water containing 0.05% yeast extract; 0.1 ml of each dilution (100, 10', 102, and 103) was inoculated onto BGSA and onto nonselective supplemented agar. Plates were incubated in GasPak jars at 35 C for 7 days. The plates were carefully observed after 2 and 7 days. AUl isolates on BGSA and those resembling B. gracilis on nonselective agar were identified by growth in formate- and fumarate-supplemented thioglycolate broth, nitrate reduction, lack of saccharolytic activity' in prereduced, anaerobically sterilized biochemicals, negative urease, tryptophanase (negative indole test), and lack of motility in a hanging-drop preparation (13). Isolation of B. gracilis from clinical specimens. BGSA, along with media used routinely in our laboratory, was used for 49 clinical specimens (28 patients) in ongoing studies (2) in the Wadsworth Anaerobic Bacteriology Laboratory from July through October, The media used included Bacteroides bile esculin agar, kanamycin-vancomycin lakedblood agar, phenylethyl alcohol agar, and brucella blood agar for anaerobes and MacConkey agar, phenylethyl alcohol agar,'chocolate agar, and brucella blood agar for aerobes; thioglycolate broth was used for nonselective enrichment. Specimens were processed, and anaerobic bacteria were identified by the methods outlined in the Virginia Polytechnic Institute and Wadsworth anaerobic laboratory manuals (i, 13). Aerobic gram-negative bacilli were identified by using the API 20E system. Aerobic gram-positive organisms and nonfermenting gram-negative organisms were identified by using methods outlined by Baron and Finegold (1). RESULTS In preliminary studies, we found that a few strains of B. gracilis were inhibited by low concentrations of many antimicrobial agents: amoxicillin-clavulanic acid, ampicillinsulbactam, aztreonam, cefotetan, cefoxitin, ceftazidime, ceftizoxime, chloramphenicol, ciprofloxacin, clindamycin, colistin, cycloserine, enoxacin, gentamicin, imipenem, josamycin, kanamycin, lomefloxacin, metronidazole, neomycin, norfloxacin, ofloxacin, paromomycin, penicillin G, piperacillin, rifampin, sulfamethoxazole-trimethoprim, ticarcillin-clavulanic acid, and tobramycin. Therefore, thése drugs were not suitable as selective agents. Many' chemical reagents and dyes have been used to achieve selective bacterial inhibition and to aid in differentiation of bacteria. We evaluated several reagents and dyes, i.e., phenylethyl alcohol, sodium azide, sodium deoxycholate, acid fuchsin, azure Il, basic fuchsin, brilliant green, crystal violet, ethyl violet, janus green, methylene blue, and neutral red. None of these wa's useful for selective inhibition of organisms in our system. Qualitative studies. All of the B. gracilis strains tested grew on both BGSA and brucella blood agar. Most bacteroides species, including B. fragilis group (the specific species studied are listed in Table 1, footnote b) strains and pigmented bacteroides and porphyromonads, were inhibited. Gram-positive anaerobes were also inhibited, but B. ureolyticus and some fusobacteria, wolinellas, and veillonellas were not inhibited by BGSA (Table 1). Among the aerobic and facultative organisms tested, gram-positive cocci did not grow on BGSA. The serratias, acinetobacters, and hafnias and most of the strains of Escherichia coli tested did not grow either; however, the Citrobacter diversus, C. freundii, and Pseudomonas aeruginosa strains tested consistently grew on BGSA (Table 2). Downloaded from on March 7, 2019 by guest

3 VOL. 28, 1990 TABLE 3. Quantitative growth of B. gracilis on brucella blood agar and BGSA B. gracilis Mean log1, no. of organisms on: strain no. Brucella blood agar BGSA il Mean + SD Quantitative and morphologic studies. No major differences were observed between unsupplemented brucella blood agar and BGSA in the numbers of colonies of 18 strains of B. gracilis grown (Table 3), but minimal inhibition on BGSA was noted. One B. gracilis strain (strain 4) was moderately inhibited by BGSA, although it grew in the qualitative studies. The colonies of 8 of the 18 strains on BGSA were larger than those on brucella blood agar (mean difference, 0.58 mm; range, 0.3 to 0.9 mm). Colonies were yellow to brown, translucent, entire, circular, raised, and generally 1.0 to 2.5 mm in diameter. B. gracilis did not pit or corrode BGSA agar. Denser, more distinct, and more even stain reactions and some long bacilli were observed in Gram stains from colonies on BGSA. Isolation of B. gracilis from simulated mixed oral flora. Different results were observed for different strains of B. gracilis. In the mixed suspension of B. gracilis (strain 3), B. gracilis was isolated only on BGSA (2+). In the mixed suspension of B. gracilis (strain 4), B. gracilis was not isolated on either medium. In the mixed suspension of B. gracilis (strain 18), B. gracilis was isolated on both media (5+). F. nucleatum was isolated on BGSA (2+) and on nonselective agar (4+). The other oral flora tested were not detected on BGSA but were isolated on nonselective agar (4+ to 5+). Isolation of B. gracilis from saliva. The following were isolated on BGSA from seven saliva specimens: 1 B. gracilis strain; 6 strains of Veillonella sp.; 10 strains of Wolinella sp. or Campylobacter concisus (we did not differentiate them further); and 2 strains of F. nucleatum. On nonselective agar, B. gracilis was not detected, but Veillonella sp. (4 strains), Wolinella sp., or C. concisus (3 strains), F. nucleatum (1 strain), and many other anaerobes, as well as aerobes that were not identified, were isolated. No significant difference in recovery was observed among dilutions. Isolation of B. gracilis from clinical specimens. Four strains of B. gracilis were isolated on BGSA from either abdominal wound or abscess specimens obtained from four patients; none was detected from among the mixed flora recovered on B. GRACILIS SELECTIVE MEDIUM 1749 standard media. Other isolates recovered on BGSA from these specimens included an occasional Fusobacterium species, B. ureolyticus, other Bacteroides species, P. aeruginosa, and C. freundii. DISCUSSION B. gracilis was first described as a species in 1981 (14). Johnson et al. (9) subsequently reported that B. gracilis was recovered from patients with serious visceral or head and neck infections, whereas a similar organism, B. ureolyticus, was isolated primarily from superficial skin and soft tissue infections (survey of specimens studied over a period of 11 years). B. gracilis showed a relatively high incidence of antibiotic resistance (10). Tryptic soy agar was chosen as the base medium for BGSA on the basis of a comparison with other media, including Mueller-Hinton and brain heart infusion agars (data not shown). The medium does not contain sheep blood, which is not required for growth of B. gracilis, but does enhance the growth of most other anaerobes. Growth of all strains of B. gracilis requires formate and fumarate, and some of them use nitrate as an electron acceptor (14). Several other investigators have used crystal violet, bacitracin, vancomycin, and teicoplanin for inhibition of grampositive organisms (5, 8, 12, 15). Recently, Greenwood et al. (6) described good antibacterial activity of teicoplanin against many Bacteroides species and Eley et al. (3) developed a medium selective for B. ureolyticus that incorporates teicoplanin (20 mg/liter). All strains of B. gracilis were resistant to bacitracin (16 mg/liter), nalidixic acid (16 mg/ liter), teicoplanin (64 mg/liter), and vancomycin (64 mg/liter). Teicoplanin at 32 mg/liter inhibited most bacteroides species and gram-positive bacteria (Tables 1 and 2) but showed little activity against B. gracilis. This effect was increased by using media lacking blood. The inhibitory activity of teicoplanin was stronger than those of bacitracin and vancomycin. Nalidixic acid is commonly used to inhibit aerobic and facultatively anaerobic, gram-negative bacteria for selective media (3, 4). Nalidixic acid at 8 mg/liter in BGSA did not completely inhibit facultative organisms but did suppress most of those strains (Table 2) while permitting growth of B. gracilis. Most of the microorganisms in the samples of saliva cultured did not grow on BGSA; however, some Fusobacterium, Veillonella, and B. ureolyticus group strains grew on BGSA; B. gracilis was difficult to distinguish on these streak plates because colonies of all of the organisms able to grow appeared similar, except for the "bread crumb" colonies of F. nucleatum. Gram stain and some simple biochemical tests (spot indole, Kovacs oxidase, requirement for formate and fumarate, nitrate reduction disk test, urease, and motility) (13), however, permitted presumptive identification. Limited stability studies using 28 strains of eight different species of anaerobes, including B. gracilis, showed no deterioration of the medium after 3 weeks of storage in an anaerobic atmosphere (GasPak pouches and jars) at 4 C. We conclude that BGSA is useful for isolating B. gracilis from clinical material containing mixed populations of microorganisms and may aid in clinical and epidemiologic studies to assess further the role of B. gracilis in human disease. ACKNOWLEDGMENTS This work was supported in part by Roerig (a division of Pfizer Inc., New York, N.Y.) and by Veterans Administration merit review funds. Downloaded from on March 7, 2019 by guest

4 1750 LEE ET AL. LITERATURE CITED 1. Baron, E. J., and S. M. Finegold Bailey and Scott's diagnostic microbiology, 8th ed. C. V. Mosby Co., St. Louis. 2. Bennion, R. S., E. J. Baron, J. E. Thompson, Jr., J. Downes, P. Summanen, and S. M. Finegold The bacteriology of gangrenous and perforated appendicitis-revisited. Ann. Surg. 211: Eley, A., T. Clarry, and K. W. Bennett Selective and 4. Ellner, P. D., C. J. Stoessel, E. Drakeford, and F. Vasi A new culture medium for medical bacteriology. Am. J. Clin. Pathol. 45: Greenwood, D Microbiological properties of teicoplanin. J. Antimicrob. Chemother. 21(Suppl. A): Greenwood, D., J. Palfreyman, A. Eley, and T. Clarry Activity of teicoplanin against gram-negative anaerobes. J. Antimicrob. Chemother. 21: Holdeman, L. V., E. P. Cato, and W. E. C. Moore (ed.) Anaerobe laboratory manual, 4th ed. Virginia Polytechnic Institute and State University, Blacksburg. 8. Hunt, D. E., J. V. Jones, and V. R. Dowell, Jr Selective medium for the isolation of Bacteroides gingivalis. J. Clin. Microbiol. 23: Johnson, C. C., J. F. Reinhardt, M. A. C. Edelstein, M. E. Muiigan, W. L. George, and S. M. Finegold Bacteroides gracilis, an important anaerobic bacterial pathogen. J. Clin. J. CLIN. MICROBIOL. Microbiol. 22: Johnson, C. C., J. F. Reinhardt, M. E. Mulligan, W. L. George, and S. M. Finegold In vitro activities of 17 antimicrobial agents against the formate/fumarate-requiring, anaerobic gramnegative bacilli. Diagn. Microbiol. Infect. Dis. 5: Lennette, E. H., A. Balows, W. J. Hausler, Jr., and H. J. Shadomy (ed.) Manual of clinical microbiology, 4th ed. American Society for Microbiology, Washington, D.C. 12. Morgenstein, A. A., D. M. Citron, and S. M. Finegold New medium selective for Fusobacterium species and differential for Fusobacterium necrophorum. J. Clin. Microbiol. 13: Sutter, V. L., D. M. Citron, M. A. C. Edelstein, and S. M. Finegold Wadsworth anaerobic bacteriology manual, 4th ed. Star Publishing Co., Belmont, Calif. 14. Tanner, A. C. R., S. Badger, C.-H. Lai, M. A. Listgarten, R. A. Visconti, and S. S. Socransky Wolinella gen. nov., Wolinella succinogenes (Vibrio succinogenes Wolin et al.) comb. nov., and description of Bacteroides gracilis sp. nov., Wolinella recta sp. nov., Campylobacter concisus sp. nov., and Eikenella corrodens from humans with periodontal disease. Int. J. Syst. Bacteriol. 31: Walker, C. B., D. Ratliff, D. Muller, R. Mandell, and S. S. Socransky Medium for selective isolation of Fusobacterium nucleatum from human periodontal pockets. J. Clin. Microbiol. 10: Downloaded from on March 7, 2019 by guest

5 Letter to the Editor Selective Medium for Bacteroides gracilis It was interesting to read a description by Lee et al. (3) of a new selective medium for Bacteroides gracilis which contained nalidixic acid (8 mg/liter) and teicoplanin (32 mg/liter) as selective agents. These investigators made reference to recently published findings on a differential and selective medium for Bacteroides ureolyticus (1), an organism which Lee et al. (3) describe as being similar to B. gracilis. We would like to make it clear to your readers that the medium developed for B. ureolyticus contained nalidixic acid (10 mg/liter), despite the fact that Lee et al. (3) only refer to its containing teicoplanin (20 mg/liter). Although we were pleased to see that these selective agents were not inhibitory to the growth of 18 strains of B. gracilis, the results were hardly surprising to us, as B. gracilis ATCC was used as a control in the routine assessment of differential activity of our medium, which also included urea. Rather than choosing tryptic soy agar as a basal medium, we used fastidious anaerobe agar, which has been shown to be excellent for the growth of fastidious anaerobic bacteria (2). Fastidious anaerobe agar is much more complex than tryptic soy agar, even though it does not incorporate potassium nitrate, a supplement added by Lee et al. (3). However, to date there have been no studies which show enhancement of growth by nitrate in the presence of formate and fumarate. We also supplemented fastidious anaerobe agar with an additional amount of agar, making a total of 22 g/liter (compared with 15 g/liter in tryptic soy agar), which can be useful in reducing the swarming of Proteus spp., bacteria known to grow on both selective media. A problem with the medium described by Lee et al. (3) was the difficulty in distinguishing between B. gracilis and other organisms isolated from clinical specimens. This difficulty might be partly overcome by the incorporation of, for example, urea, which would make it possible to identify urease-producing organisms. This would mean that time and reagents would not be wasted on the identification of isolates which are obviously not B. gracilis, which is urease negative. Such a modification would add a dimension to the selective medium for B. gracilis and ultimately make it of greater use in diagnostic microbiology. REFERENCES 1. Eley, A., T. Clarry, and K. W. Bennett Selective and 2. Heginbotham, M., T. C. Fitzgerald, and W. G. Wade Comparison of solid media for cultivation of anaerobes. J. Clin. Pathol. 43: Lee, K., E. J. Baron, P. Summanen, and S. M. Finegold Selective medium for isolation of Bacteroides gracilis. J. Clin. Microbiol. 28: Adrian Eley Department of Expe;imental and Clinical Microbiology University of Sheffield Medical School Beech Hill Road Sheffield SJO 2RX United Kingdom Kevin Bennett Department of Bacteriology Royal Hallamshire Hospital Glossop Road Sheffield S1O 2JF United Kingdom Author's Reply We appreciate Eley and Bennett's interest in our article describing a selective medium for Bacteroides gracilis (2). They state that in referring to their selective medium for Bacteroides ureolyticus (1) we note only its teicoplanin content and do not mention its nalidixic acid content; in fact, in the fifth sentence after our mention of teicoplanin (p. 1749), we mention the use of nalidixic acid in selective media and reference their article (plus one other). Tryptic soy agar was chosen as our basal medium because a number of anaerobes other than B. gracilis do not grow on it, and it is relatively low priced compared with some richer media. The suggestions that an increased agar content might reduce swarming of Proteus spp. and that incorporation of urea would permit rapid exclusion of B. ureolyticus are reasonable but need to be assessed in our selective medium before adoption. REFERENCES 1. Eley, A., T. Clarry, and K. W. Bennett Selective and 2. Lee, K., E. J. Baron, P. Summanen, and S. M. Finegold Selective medium for isolation of Bacteroides gracilis. J. Clin. Microbiol. 28: Kyungwon Lee Department of Clinical Pathology Wonju Christian Hospital 162 Ilsan-Dong Kangwon-Do, Korea Ellen Jo Baron Paula Summanen Sydney M. Finegold Research Service VA Medical Center West Los Angeles Wilshire and Sawtelle Boulevards Los Angeles, California