LETTERS

Empyema Thoracis During the study, 973 patients with Choleraesuis is more invasive than 2 a diagnosis of empyema thoracis were other Salmonella species. from Salmonella identified; 12 (1.23%) of these Among the S. Choleraesuis iso- Choleraesuis patients, including 9 men and 3 lates recovered from 7 patients, 2 women, were infected with were resistant to ampicillin and sul- To the Editor: The clinical pre- Salmonella species. The clinical char- famethoxazole-trimethoprim, 3 were sentations of nontyphoidal acteristics of the 12 patients are sum- resistant to , 1 was Salmonella are protean, marized in the online Appendix Table resistant to ciprofloxacin, and all were including gastroenteritis (most com- (available from http://www.cdc.gov/ susceptible to cefotaxime. All S. mon), bacteremia, septic arthritis, ncidod/eid/vol11no09/05-0030_ Typhimurium isolates were suscepti- osteomyelitis, and endovascular app.htm). The median age was 49 ble to sulfamethoxazole-trimetho- infection. (1). Despite the growing years; 1 patient was >65 years of age. prim, ciprofloxacin, and cefotaxime. number of patients with invasive Underlying diseases were present in Two of the 4 patients had isolates that infection due to nontyphoid all patients, including 7 with malig- were resistant to chloramphenicol, Salmonella, reports of thoracic nancy, 5 with gallstones, and 3 each and 2 other patients had isolates that empyema due to these organisms with diabetes mellitus and chronic were resistant to ampicillin. The remain rare (2–6). renal failure. Five patients had used group C2 salmonella isolate was We searched the computer data- antacids and 3 patients had received resistant to chloramphenicol only. base of our microbiology laboratory chemotherapy or steroids. Ten Among the 12 Salmonella isolates for patients with positive pleural effu- patients (83.3%) were immunocom- from patients with empyema thoracis, sion culture from June 1997 to promised and had a variety of illness- 9 were resistant to >1 commonly used February 2004. Patients were includ- es, including malignancy, liver cirrho- antimicrobial. Treatment and outcome ed if they met the following criteria: sis, and diabetes mellitus. Common information was available for 11 of 1) thoracentesis recovered purulent symptoms were dyspnea (83.3%), the 12 patients. All 11 patients pleural fluid; 2) microorganisms iden- (75%), and cough (50%). received antimicrobials drugs (medi- tified by microscopic examination; Analysis of pleural effusion showed a an duration 35 days); this therapy was and 3) a Salmonella species isolated median leukocyte count of 25,600/µL, appropriate in 9 of 11 patients. Six from a pleural effusion specimen. a lactate dehydrogenase level of 513 patients had thoracentesis, 2 had close Isolates of Salmonella were identi- U/L, and a glucose level of 88 mg/dL. tube thoracostomy, and 1 had open fied to the serotype level, according to Gram staining was conducted on 3 drainage. One of the 4 patients with S. the Kauffman and White scheme, patients’ pleural effusion but none of Typhimurium empyema who did not using somatic and flagellar antigens them showed positive results. receive appropriate antimicrobial (Denka Seiken Co., Ltd., Tokyo, Twenty-three Salmonella isolates drugs died. In contrast, 4 (57%) of the Japan) and also by conventional were recovered as the sole pathogen 7 patients with S. Choleraesuis infec- methods and the Phoenix System from various clinical specimens, tion, including 1 who did not receive (panel type, NMIC/ID4) (Becton including pleural effusion (15 isolates), appropriate antimicrobial therapy, Dickson, Sparks, MD, USA) (7). blood (6 isolates), ascites (1 isolate), died. Another factor related to out- Susceptibilities of Salmonella isolates and aortic wall (1 isolate). Among the come was drainage. One (20%) of the to ampicillin, cefotaxime, chloram- 12 patients with empyema thoracis, 4 5 patients who underwent tube thora- phenicol, ciprofloxacin, and trimetho- had serotype costomy or thoracoscopy died, while prim-sulfamethoxazole were deter- Typhimurium (S. Typhimurium) and 1 3 (50%) of the 6 patients who under- mined by the disk diffusion method. had group C2 Salmonella during went thoracentesis died. All 3 of these Organisms were categorized as sus- 1997–1999; 7 patients had Salmonella patients had S. Choleraesuis. ceptible or resistant to the antimicro- enterica serotype Choleraesuis Most (92%) of our patients were bial agents tested on the basis of (S. Choleraesuis) after 1998. All S. <65 years of age. These data indicate National Committee for Clinical Typhimurium and group C2 that Salmonella should be considered Laboratory Standards (NCCLS) Salmonella were isolated from pleural as a potential cause of thoracic empye- guidelines (8). Antimicrobial therapy effusion specimens, but S. ma, even in younger patients, especial- was considered to be appropriate Choleraesuis was isolated from multi- ly in the presence of malignancy or when the antimicrobial agent was ple extrapulmonary sites including hepatobiliary disease. More than half active in vitro by the disk diffusion blood, ascites, and aortic wall (Online of our patients had used antacids or susceptibility method against a Table). Although the number of study had suffered from gallstones. This Salmonella isolate. cases is limited, it may suggest that S. finding suggests that susceptibility to

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 11, No. 9, September 2005 1493 LETTERS

Salmonella infection may be increased In conclusion, thoracic empyema 9. Cohen JI, Bartlett JA, Corey GR. Extra- by alterations in the gastrointestinal is a rare complication of nontyphoid intestinal manifestations of Salmonella infection. Medicine (Baltimore). tract, including decreased gastric acid- Salmonella infection and is closely 1987;66:349–88. ity and chronic gastrointestinal dis- associated with an immunocompro- ease. Leukocytosis was noted in 25% mised condition, even in patients <65 Address for correspondence: Po-Ren Hsueh, of patients. In fact, two thirds of the years of age. Higher rates of resist- Departments of Laboratory Medicine and patients had a normal leukocyte count ance and death were noted in patients Internal Medicine, National Taiwan University, with immature leukocytes, which may with empyema thoracic caused by S. No 7, Chung-Shan South Road, 100 Taipei, be attributable to their relatively Choleraesuis. Early diagnosis, appro- Taiwan, Republic of China; fax: 886-2- impaired cell-mediated immunity. priate antimicrobial drug therapy, and 23224263; e-mail: [email protected] The predominant organism in this aggressive drainage are necessary to series was S. Choleraesuis, followed improve the outcome of patients with by S. Typhimurium. In Taiwan, the thoracic empyema due to S. rate of resistance of S. Choleraesuis to Choleraesuis. ampicillin, chloramphenicol, or sul- famethoxazole-trimethoprim increased Chih-Cheng Lai,* Li-Na Lee,* to approximately 90% for all 3 drugs Po-Ren Hsueh,* Chong-Jen Yu,* and the rate of resistance to and Pan-Chyr Yang* Asymptomatic ciprofloxacin was from 7.7% to 59% *National Taiwan University Hospital, pestis (5–7). The resistance rate of S. Taipei, Taiwan, Republic of China Infection, China Choleraesuis to ciprofloxacin in this study was similar to our previous References To the Editor: Plague is one of the report (7). 1. Hohmann EL. Nontyphoid . oldest identifiable diseases. Modern Nine of the 11 patients who com- Clin Infect Dis. 2001;32:263–9. measures and effective pleted follow-up information received 2. Chen KY, Hsueh PR, Liaw YS, Yang PC, antimicrobial treatments have led to a Luh KT. A 10-year experience with bacteri- decrease in plague cases worldwide. appropriate antimicrobial drugs with ology of acute thoracic empyema: emphasis drainage; however, 4 died. These 4 on in patients with However, plague remains endemic in deaths (57%) were due to S. diabetes mellitus. Chest. 2000;117:1685–9. many natural foci. Since the early Choleraesuis-related empyema; 3 3. Samonis G, Maraki S, Kouroussis C, 1990s, the World Health Organization Mavroudis D, Georgoulias V. Salmonella (WHO) has reported a steadily increas- patients had underlying malignancy. enterica in a patient with Although appropriate antimicrobial cancer. J Clin Microbiol. 2003;41:5820–2. ing trend in human plague cases, drugs were used, our data suggest that 4. Biendo M, Laurans G, Thomas D, Dechepy which has resulted in the recognition more aggressive treatment with open O, Hamdad-Daoudi F, Canarelli B, et al. of plague as a reemerging disease (1). Regional dissemination of Salmonella The emergence of antimicrobial drug– drainage of the pleural effusion might enterica serovar Enteritidis is season have contributed to a better outcome dependent. Clin Microbiol Infect. resistant strains of , than closed tube thoracostomy or sim- 2003;9:360–9. along with an increasing number of ple thoracocentesis alone. In contrast 5. Chiu CH, Wu TL, Su LH, Chu C, Chia JH, plague cases, remind us that plague Kuo AJ, et al. The emergence in Taiwan of still poses a serious public health threat to S. Choleraesuis-related infection, fluoroquinolone resistance in Salmonella all 4 patients with non–S. enterica serotype Choleraesuis. N Engl J (2,3). In China, human cases of plague Choleraesuis-related thoracic empye- Med. 2002;346:413–9. have been reported to WHO nearly ma survived. One of these patients did 6. Chiu CH, Su LH, Chu C. Salmonella enter- every year from 1989 to 2003; these ica serotype Choleraesuis: epidemiology, account for 9.5% of cases and 15.5% not receive appropriate antimicrobial pathogenesis, clinical disease, and treat- drug treatment, but did have adequate ment. Clin Microbiol Rev. 2004;17:311–22. of deaths from this disease in Asia (1). drainage with simple thoracocentesis. 7. Hsueh PR, Teng LJ, Tseng SP, Chang CF, Human cases of plague in China are This suggests adequate and aggres- Wan JH, Yan JJ, et al. Ciprofloxacin-resist- usually caused by contact with plague- ant Salmonella enterica Typhimurium and infected . Here, we report the sive drainage of pleural effusion may Choleraesuis from pigs to humans, Taiwan. be as important as appropriate antimi- Emerg Infect Dis. 2004;10:60–8. results of a serologic survey by using 3 crobial drugs. However, the overall 8. National Committee for Clinical methods (passive hemagglutination death rate (36%) in this study was still Laboratory Standards. 2003. Performance assay, Western blot, and protein standards for antimicrobial disk susceptibil- microarray analysis) in marmot higher than that of other reports (9). ity test; approved standards-eighth edition. This might have been due to the high M2-A8. Wayne (PA): The Committee. hunters in Qinghai Province, China. number of immunocompromised One hundred twenty serum sam- patients in this study. ples were collected in 2 villages in

1494 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 11, No. 9, September 2005