Pseudomonas Aeruginosa-Associated Diarrheal Diseases in Children

ORIGINAL STUDIES

Pseudomonas aeruginosa-associated Diarrheal Diseases in Children

Chih-Hsien Chuang, MD, PhD,*†‡ Rajendra Prasad Janapatla, PhD,† Yi-Hsin Wang, MS,† Hsin-Ju Chang, BS,† Yhu-Chering Huang, MD, PhD,§ Tzou-Yien Lin, MD,§ and Cheng-Hsun Chiu, MD, PhD†§

sites such as respiratory, urinary and gastrointestinal (GI) tract. It Background: The gastrointestinal tract is not the common infection site of is a leading pathogen of hospital-acquired pneumonia and a major Pseudomonas aeruginosa. The role of P. aeruginosa as a causative agent for pathogen of catheter-related urinary tract infection.3,4 Although diarrhea in children without preexisting disease is controversial. P. aeruginosa can affect entire GI tract, the diseases are usually Methods: From 2003 to 2012, we reviewed the records of 259 diarrheal mild. The presence of P. aeruginosa in stool is usually considered patients less than 5 years of age whose stool culture grew P. aeruginosa. to be of no clinical significance. The role ofP. aeruginosa as a diar- Virulence phenotypes of bacterial isolates were determined in vitro, includ- rheal agent remains controversial. ing cytotoxicity, penetration and adherence to epithelial cells. Nevertheless, outbreaks of diarrheal diseases caused by Results: The presence of P. aeruginosa in children with diarrhea less than P. aeruginosa in neonatal intensive care units and pediatric wards 5 years old is 0.91%. P. aeruginosa-associated diarrheal diseases were clas- have been reported.5,6 Diarrheal disease caused by P. aeruginosa sified into 4 groups: Shanghai fever (enteric infection and sepsis) (5%), has been described to be associated with antibiotic exposure.7,8 On P. aeruginosa enterocolitis (15%), P. aeruginosa-related diarrhea (19%) the other hand, community-acquired P. aeruginosa sepsis has been and antibiotic-associated diarrhea (43%). The remaining patients had coin- reported from East Asian countries.9–13 Diarrhea is the most common fection with other pathogens (18%). Shanghai fever was the most severe initial symptom of the entity, namely Shanghai fever.10 Necrotizing enteric disease with invasive infection and complications. The clinical enteritis is the major complication.10 All patients with Shanghai fever features of P. aeruginosa enterocolitis were prolonged fever with bloody were previously healthy without underlying conditions. Shanghai or mucoid diarrhea mimicking bacterial enterocolitis. The clinical fea- fever is a distinct P. aeruginosa disease with clinical manifestations tures of P. aeruginosa-related diarrhea and antibiotic-associated diarrhea different from other enteric diseases caused by P. aeruginosa. Apart were similar to viral or toxin-mediated diarrhea. Compared with other from that, diarrheal diseases related to P. aeruginosa were seldom P. aeruginosa-associated diarrheal diseases, patients with Shanghai fever discussed in children without preexisting diseases. The present study were younger, usually infants, and the characteristic laboratory findings aimed to clarify P. aeruginosa-associated diarrheal diseases by clini- included leukopenia, thrombocytopenia, high C-reactive protein, hypona- cal analysis and to examine the role that microorganisms play. tremia and hyperglycemia. Except for Shanghai fever, antibiotic treatment is not recommended. Isolates from Shanghai fever were more cytotoxic and adherent than isolates from uncomplicated diarrheal patients. MATERIALS AND METHODS Conclusions: P. aeruginosa could be an enteric pathogen even in healthy children. Young age and highly virulent bacterial strains were risk factors Participants and Definition of Terms for Shanghai fever. Chang Gung Children’s Hospital (CGCH) is a tertiary hospital with more than 400 beds specific for pediatric care. We iden- Key Words: Pseudomonas aeruginosa, diarrhea, enteric infection, chil- tified all stool cultures that were positive for P. aeruginosa from dren, shock January 2003 to December 2012 at CGCH. The clinical microbiology laboratory in CGCH reported only pure or predominant (Pediatr Infect Dis J 2017;36:1119–1123) growth of P. aeruginosa in stool cultures from xylose lysine deoxy- cholate (XLD) agar. Patients under 5 years of age were enrolled in this study. The medical records of patients with positive stool seudomonas aeruginosa is an important opportunistic gram- cultures were reviewed and analyzed. Information was obtained Pnegative pathogen that usually causes infection in patients from several categories including demographics, clinical data, who are immunocompromised, critically ill, having cystic fibrosis microbiologic data and treatment. Patients were classified into 4 or burn wounds.1,2 P. aeruginosa has been isolated from mucosal groups according to their clinical characteristics: group 1, diarrheal patients with growth of P. aeruginosa from blood culture, defined as Shanghai fever; group 2, diarrheal patients with clini- Accepted for publication January 6, 2017. cal manifestations mimicking bacterial enterocolitis (fever with From the *Department of Pediatrics, St. Paul’s Hospital, Taoyuan, Taiwan; †Molecular Infectious Diseases Research Center, Chang Gung Memorial mucoid or bloody diarrhea), defined as P. aeruginosa enterocolitis; Hospital, Chang Gung University College of Medicine, Taoyuan, Taiwan; group 3, diarrheal patients with clinical manifestations suggestive ‡School of Medicine, College of Medicine, Fu-Jen Catholic University, New of viral or toxin-mediated enteritis (watery diarrhea), defined as P. Taipei, Taiwan; and §Division of Pediatric Infectious Diseases, Chang Gung aeruginosa-related diarrhea and group 4, diarrheal patients with Children’s Hospital, Chang Gung University College of Medicine, Taoyuan, Taiwan. previous antibiotic treatment in recent 1 month, defined as antibi- This study was supported in part by Ministry of Science and Technology otic-associated diarrhea. Hospital-acquired infection was defined (103-2314-B-182A-089-MY3), Chang Gung Memorial Hospital when the stool culture was taken 3 days after hospitalization. The (CMRPG3B0071-3, and CMRPG3F1141) and Aim for the Top University Plan. The authors have no conflicts of interest to disclose. Institutional Review Board of the Chang Gung Memorial Hospital Address for correspondence: Cheng-Hsun Chiu, MD, PhD, Department of approved the study (102-1012B). Pediatrics, Division of Pediatric Infectious Diseases, Chang Gung Chil- dren’s Hospital, 5 Fu-Hsin Street, Kweishan 333, Taoyuan, Taiwan. E-mail: Bacterial Strains [email protected]. 10 Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. Seventeen blood isolates from patients with Shanghai fever ISSN: 0891-3668/17/3612-1119 were collected from 2003 to 2012. Only 75 P. aeruginosa from DOI: 10.1097/INF.0000000000001567 stool cultures, defined as GI isolates, were obtained from 2009 to

The Pediatric Infectious Disease Journal • Volume 36, Number 12, December 2017 www.pidj.com | 1119

2012. These strains were preserved in Lysogeny broth containing 20% glycerol at −80°C. They were revived in Lysogeny broth agar before experiments. Seventy-five GI isolates were used for analysis of type III secretion system (TTSS) encoding toxin genes: exoS and exoU. Seventeen blood isolates from patients with Shanghai fever and 21 GI isolates obtained from diarrheal patients without Shanghai fever were used to compare the virulence profile between Shanghai fever isolates and GI isolates. Selection of 21 GI isolates for virulence testing was according to the distribution of exoS- and exoU encoding isolates among 75 GI isolates. The cytotoxic P. aeruginosa laboratory strains PA14 and PA103 and the noncytotoxic but invasive strain PAO1 were used as phenotypic controls.14 Sal- monella enteric serovar Typhimurium SL1344 was the positive control for the Madin-Darby Canine Kidney (MDCK) cell monolayer penetration assay.14 Noninvasive rabbit enterotoxigenic Escherichia coli strain RDEC-1 was used as a negative control.14

Detection of TTSS-encoding Toxin Genes The polymerase chain reaction and sequencing were used to FIGURE 1. Diseases associated with Pseudomonas detect the presence of TTSS encoding toxin genes. The primers and aeruginosa that grew in stool culture. methods used for detection of exoU and exoS were as described in a previous study.15 TABLE 1. Frequency of Pseudomonas aeruginosa From Cytotoxicity, Penetration and Adhesion Assays Diarrheal Patients Stratified by Age In vitro cytotoxicity of bacterial isolates to MDCK cells and evaluation of invasive phenotype of the isolated strains were exam- No. of Stool No. of Isolated Frequency ined by methods described previously.10,14 The method of Caco-2 Age (mo) Cultures Pseudomonas aeruginosa (%) 10 cell adhesion assay was described in a previous report. 0–11 9549 106 1.11 12–23 9215 70 0.76 Statistical Analysis 24–35 4952 39 0.79 Continuous variables were tested with the use of Student’s 36–47 2754 23 0.84 t-test or Mann–Whitney U test. Categorical data were analyzed 48–59 2083 21 1.01 Total 28,553 259 0.91 using χ2 or Fisher’s exact test. If a significant difference was identified by analysis of variance, pairwise comparison was conducted with Scheffe’s method. Statistical significance was defined as No significant difference in age distribution, proportion of 2-sided P < 0.05. infant, sex ratio and percentage of underlying disease were identified between groups 2 and 3. The clinical manifestations were RESULTS different between groups 2 and 3. More patients in group 2 had bloody or mucoid diarrhea than in group 3 (P < 0.001). Patients Clinical Data in group 2 had higher proportion of fever (P = 0.001) and longer A total of 28,553 stool cultures were performed in patients fever duration than patients in group 3 (95% CI: 1.31–3.94; under 5 years of age during the study period (Fig. 1). Pure or predom- P < 0.001). Thirty-five patients (92%) in group 2 had fever; 33 of inant growth of P. aeruginosa was reported in 259 patients (0.91%). them (87%) had a fever duration ≥3 days. In contrast, 19 patients The prevalence of P. aeruginosa in stool cultures ranged from 0.76% (39%) in group 3 did not have fever. Among patients with fever in to 1.11% in different age groups (Table 1). The highest rate was in group 3, 39 (81%) had a fever duration ≤3 days. No patients in groups infants. About 186 patients (72%) had community-acquired P. aer- 2 and 3 had complications. Only 1 patient in group 3 had febrile sei- uginosa infection, others had hospital-acquired infection. zure. Twenty patients (53%) in group 2 were treated with ceftriaxone There were significant differences in age distribution, pro- initially because Salmonella enteritis was suspected. Nine patients portion of infant, fever duration and stool pattern between group 1 (24%) in group 2 received antipseudomonal antibiotic treatment; their (Shanghai fever), group 2 (P. aeruginosa enterocolitis) and group 3 fever duration was significantly longer than patients without antipseu- (P. aeruginosa-related diarrhea; Table 2). Patients in group 1 were domonal antibiotic treatment (95% CI: 0.52–4.01; P = 0.013). Only significantly younger than the other 2 groups [95% confidence 1 patient (2%) in group 3 was treated with ceftriaxone. Five patients interval (CI): −24.50 to −1.08; P = 0.03 in comparison between (10%) in group 3 received antipseudomonal antibiotic treatment. groups 1 and 2; 95% CI: −24.50 to −1.75; P = 0.02 between groups Group 4 (antibiotic-associated diarrhea) was the most com- 1 and 3]. The fever duration was longer in group 1 than the other mon diarrheal disease associated with P. aeruginosa. Among them, two groups (95% CI: 3.21–7.11; P < 0.001 in comparison between 45 patients (40%) had underlying diseases. Fifty-two patients (46%) groups 1 and 2; 95% CI: 5.89–9.69; P < 0.001 between groups 1 were infants. Bloody and mucoid stool were seen in 3 patients (3%). and 3). All the patients in group 1 had fever and fever duration was Others were watery or loose stool without bloody tinge or mucus. more than 3 days. The stool pattern was similar between groups Twenty-one patients (18%) received antipseudomonal antibiotic 1 and 2. All patients in group 1 had complications. Necrotizing treatment. None had invasive infection. All the patients recovered enteritis is the most common complication. All patients in group 1 uneventfully. Among patients co-infected with other enteric patho- received antipseudomonal antibiotic treatment and all of them sur- gens, rotavirus is the most common co-infective pathogen. vived. All of the patients with seizure in group 1 had hyponatremia; There were significant differences in leukocyte count, platelet 2 of them had meningitis. count, level of hemoglobin, C-reactive protein (CRP) concentration,

Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. The Pediatric Infectious Disease Journal • Volume 36, Number 12, December 2017 Enteric P. aeruginosa Infection

TABLE 2. Demographic Data and Clinical Features of Patients With Pseudomonas aeruginosa-associated Diarrheal Diseases

Shanghai P. aeruginosa P. aeruginosa- Fever Enterocolitis related Diarrhea

Group 1 Group 2 Group 3 Variable (N = 13) (N = 38) (N = 49) P

Age (mo) 9.1 ± 6.1 21.9 ± 15.3 22.2 ± 15.7 0.014* Infant, no. (%) 10 (77) 16 (42) 17 (35) 0.024* Sex, male no. (%) 11 (85) 19 (50) 30 (61) 0.086 Underlying disease, no. (%) 0 (0) 2 (5) 1 (2) 0.542 Fever duration (d) 9.8 ± 3.7 4.6 ± 2.4 2.0 ± 2.0 0.000* Stool pattern, no. (%) Bloody 6 (46) 20 (53) 2 (4) 0.000* Mucoid 9 (69) 19 (50) 1 (2) 0.000* Greenish 5 (38) 5 (13) 4 (8) 0.020* Vomiting, no. (%) 7 (54) 8 (21) 18 (37) 0.070 Complications, no. (%) Necrotizing enteritis 13 (100) 0 (0) 0 (0) Ecthyma gangrenosum 8 (62) 0 (0) 0 (0) Seizure 6 (46) 0 (0) 1 (2) Antipseudomonal 13 (100) 9 (24) 5 (10) treatment, no. (%) Mortality, no. (%) 0 (0) 0 (0) 0 (0) *P < 0.05

TABLE 3. Laboratory Findings of Patients With Pseudomonas aeruginosa-associated Diarrheal Diseases

Shanghai P. aeruginosa P. aeruginosa- Fever Enterocolitis related Diarrhea

Laboratory Test Group 1 Group 2 Group 3 P

Leukocyte count† (per mm3) 5330 ± 5018 9791 ± 5222 9538 ± 5183 0.024* Leukocytosis (>15000), 1/13 (8) 4/35 (11) 4/39 (10) 0.931 no./total no. (%) Leukopenia (<4000), 7/13 (54) 4/35 (11) 4/39 (10) 0.006* no./total no. (%) Hemoglobin† (per g/dl) 10.4 ± 1.5 11.0 ± 1.4 12.0 ± 1.3 0.001* Anemia (<10), no./total no. (%) 4/13 (31) 5/35 (14) 4/39 (10) 0.197 Platelet count† (×103 per mm3) 95 ± 91 235 ± 109 273 ± 111 0.000* Thrombocytopenia (<105), 10/13 (77) 5/35 (14) 1/39 (2.6) 0.000* no./total no. (%) C-reactive protein† (mg/L) 242 ± 127 120 ± 54 20 ± 32 0.000* High CRP (>150), no./total 10/13 (77) 8/35 (23) 0/39 (0) 0.000* no. (%) Hyponatremia (<130 mEq/L), 7/13 (54) 2/24 (8) 0/29 (0) 0.000* no./total no. (%) Hypokalemia (<3.0 mEq/L), 2/12 (17) 1/20 (5) 1/27 (4) 0.307 no./total no. (%) Hyperglycemia (>150 mg/L), 6/12 (50) 0/15 (0) 0/20 (8) 0.000* no./total no. (%)

*Numbers tested was 13 for group 1, 35 for group 2 and 39 for group 3. P < 0.05 † and percentages of leukopenia, thromobocytopenia, high CRP, gene (P < 0.001). Seven (41%) of Shanghai fever isolates harbored hyponatremia, and hyperglycemia among groups 1, 2 and 3 (Table 3). exoS gene and 10 (59%) the exoU gene. There was a significant More patients in group 1 had leukopenia, thrombocytopenia, high difference in the distribution of exoS and exoU between Shanghai CRP, hyponatremia and hyperglycemia than patients in the other 2 fever isolates and GI isolates (P = 0.021). groups (P < 0.05; Table 3). Laboratory findings were similar between In cytotoxicity assay, isolates from Shanghai fever were sig- groups 2 and 3, except for the level of CRP which was higher in nificantly more cytotoxic than GI isolates (P = 0.001; Fig. 2A). In group 2 than in group 3 (95% CI: 62.93–135.69; P < 0.001). penetration assay, 11 Shanghai fever isolates (65%) and 16 GI isolates (76%) were detected in the basolateral media 1 hour after inoc- Microbiologic Examinations ulation (Fig. 2B). All Shanghai fever isolates and GI isolates, except Among 75 GI isolates, 53 isolates (71%) harbored exoS 1 GI isolate, were detected in the media 3 hours after inoculation. gene and 22 isolates (29%) the exoU gene. Isolates carrying exoU There was no difference in penetration between Shanghai fever and gene were significantly more cytotoxic than isolates carrying exoS GI isolates (P = 0.438). The number of bacteria that penetrated the

MDCK monolayer was not significantly different between Shanghai Antibiotic exposure has been identified as the main risk factor fever and GI isolates 3 hours after inoculation (P = 0.352). In adhe- for P. aeruginosa in stool.18–20 Antibiotic-associated diarrhea has sion assay, Shanghai fever isolates were significantly more adherent been excluded from comparison in this study because 40% of to Caco-2 cells than GI isolates (P = 0.024; Fig. 2C). patients had underlying diseases which could confound its clinical characteristics. The presence of P. aeruginosa in stool culture is usually DISCUSSION regarded as of no clinical significance. However, severe GI disease Carriage of P. aeruginosa in stool is uncommon in healthy 16 17 such as necrotizing enterocolitis has been reported to be associated persons. The carriage rate is about 2% in healthy children. with P. aeruginosa infection.6,21 Patients with P. aeruginosa in stool The presence of P. aeruginosa in stool in hospitalized patients 18 18 had high incidence of diarrhea. Patients with P. aeruginosa colo- is approximately 1%. Our study revealed that the presence of nized in GI tract were associated with abdominal symptoms like P. aeruginosa in children with diarrhea is about 1%, which is sim- 19 18 diarrhea, ileus and abdominal distension. Enteric disease associ- ilar to previous studies. Although prevalence of P. aeruginosa ated with sepsis caused by P. aeruginosa has been reported in as in children with diarrhea is about 1%, P. aeruginosa accounted early as 1918 with the depiction of Shanghai fever. Community- for 6.4% (259/4059) of all positive stool cultures among diar- acquired diarrheal disease complicated with sepsis has been docu- rheal children. The prevalence of P. aeruginosa in diarrheal mented in several Asian countries, especially in Chinese ancestry.9–13 children lower than normal carriage might be because of dif- Our previous study has clear delineation of Shanghai fever, which is ferent selection agar (XLD agar and Pseudomonas isolation really a severe enteric disease with sepsis caused by P. aeruginosa.10 agar) used and the positive criteria (only pure or predominant The organism could be an enteric pathogen even in healthy children. growth from XLD agar was reported in our study) applied. Based on our clinical analysis, diarrheal diseases caused by P. aeruginosa isolated from stool is usually acquired at the time 18,19 P. aeruginosa could be broadly classified into 2 types. One is the of more than 3 weeks of hospital stay. The longer the hos- clinical manifestation that mimics bacterial enterocolitis. The major- pital stay, the higher carriage rate of P. aeruginosa in stool is 18 ity of patients in this type had blood or mucus in stool and prolonged observed. However, most of our patients were community- fever. Laboratory finding was high CRP level. Patients in group 1 acquired P. aeruginosa infection. Among hospital acquired and group 2 belonged to this type. More than half of patients in P. aeruginosa in stool culture, 68 patients (93%) were antibiotic- group 2 were suspected to have salmonellosis before their culture associated diarrhea. Antibiotic-associated diarrhea is the most reports were known and were treated with ceftriaxone. The other common P. aeruginosa-associated diarrheal disease in this study. is the clinical manifestation that mimics viral or toxin-mediated

FIGURE 2. Cytotoxicity assay (A), penetration assay (B) and adhesion assay (C). Shanghai fever isolates were significantly more cytotoxic than GI isolates (P = 0.001). There were no difference in terms of bacterial penetration through MDCK monolayer between Shanghai fever isolates and GI isolates (P = 0.438). Shanghai fever isolates were more adherent to Caco-2 cells than were GI isolates (P = 0.024). Horizontal lines represent the means. CFU indicates colony forming units.

enteritis. Watery or loose stool was characteristic. Most of the 3. Quartin AA, Scerpella EG, Puttagunta S, et al. A comparison of micro- patients in this type did not have prolonged fever. Laboratory find- biology and demographics among patients with healthcare-associated, hospital-acquired, and ventilator-associated pneumonia: a retrospective ings were nonspecific. Patients in group 3 and group 4 belonged to analysis of 1184 patients from a large, international study. BMC Infect Dis. this type. Majority of them did not receive antibiotic treatment. 2013;13:561. Is it necessary to give antipseudomonal antibiotic to patient 4. Ortega M, Marco F, Soriano A, et al. Epidemiology and prognostic deter- with P. aeruginosa in stool? Somekh et al20 reported that antipseu- minants of bacteraemic catheter-acquired urinary tract infection in a single domonal antibiotic treatment did not shorten the time of P. aeruginosa institution from 1991 to 2010. J Infect. 2013;67:282–287. clearance in stool or eliminate more rapidly GI symptoms that were 5. Falcão DP, Mendonça CP, Scrassolo A, et al. Nursery outbreak of severe present. Ohara et al did not identify risk factors for development of diarrhoea due to multiple strains of Pseudomonas aeruginosa. Lancet. 1972;2:38–40. invasive infection in patients with P. aeruginosa GI colonization.19 They did not recommend antipseudomonal antibiotic treatment.19 In 6. Florman AL, Schifrin N. Observations on a small outbreak of infan- tile diarrhea associated with Pseudomonas aeruginosa. J Pediatr. this study, all patients in group 1, except 1, received antipseudomonal 1950;36:758–766. antibiotic treatment within 24 hours of admission. This might be the 7. Adlard PA, Kirov SM, Sanderson K, et al. Pseudomonas aeruginosa as a reason that no mortality occurred in current series. Immediate initia- cause of infectious diarrhoea. Epidemiol Infect. 1998;121:237–241. tion of appropriate antibiotic treatment has been reported to reduce 8. Kim SW, Peck KR, Jung SI, et al. Pseudomonas aeruginosa as a poten- mortality.9,10 In group 2, the duration of fever in patients who received tial cause of antibiotic-associated diarrhea. J Korean Med Sci. 2001;16: antipseudomonal antibiotic treatment was significantly longer than 742–744. those without antipseudomonal antibiotic treatment. Persistent fever 9. Huang YC, Lin TY, Wang CH. Community-acquired Pseudomonas aer- might be the reason for shifting to antipseudomonal antibiotic treat- uginosa sepsis in previously healthy infants and children: analysis of forty- ment after stool culture report was known. Nevertheless, most patients three episodes. Pediatr Infect Dis J. 2002;21:1049–1052. (76%) in group 2 received inappropriate antibiotic or did not receive 10. Chuang CH, Wang YH, Chang HJ, et al. Shanghai fever: a distinct Pseudomonas aeruginosa enteric disease. Gut. 2014;63:736–743. antibiotic treatment. None of them had invasive infection or compli- 11. Yeung CK, Lee KH. Community acquired fulminant Pseudomonas infection cation. All of them recovered without sequalae. Majority of patients of the gastrointestinal tract in previously healthy infants. J Paediatr Child (84%) in groups 3 and 4 did not receive antipseudomonal antibiotic Health. 1998;34:584–587. treatment, none of them had invasive infection or complication. We did 12. Zhang Q, Smith JC, Zhu Q, et al. A five-year review of Pseudomonas aer- not recommend antipseudomonal antibiotic treatment for patients with uginosa bacteremia in children hospitalized at a single center in southern P. aeruginosa-associated diarrheal diseases. Only when patients were China. Int J Infect Dis. 2012;16:e628–e632. suspected to be complicated with systemic infection, Shanghai fever, 13. Chan YH, Chong CY, Puthucheary J, et al. Ecthyma gangrenosum: a mani- antipseudomonal antibiotic treatment should be administered as soon festation of Pseudomonas sepsis in three paediatric patients. Singapore Med as possible. Shanghai fever and other P. aeruginosa-associated diar- J. 2006;47:1080–1083. rheal diseases could be distinguished by clinical features and labora- 14. Hirakata Y, Finlay BB, Simpson DA, et al. Penetration of clinical isolates of Pseudomonas aeruginosa through MDCK epithelial cell monolayers. tory findings. The clinical features of Shanghai fever are younger age, J Infect Dis. 2000;181:765–769. especially infant, with bloody, mucoid or greenish stool pattern. The 15. Stepińska M, Trafny EA. Diverse type III secretion phenotypes among characteristic laboratory findings are leukopenia, thrombocytopenia, Pseudomonas aeruginosa strains upon infection of murine macrophage-like high CRP, hyponatremia and hyperglycemia. and endothelial cell lines. Microb Pathog. 2008;44:448–458. Studies of P. aeruginosa population revealed that the preva- 16. Stoodley BJ, Thom BT. Observations on the intestinal carriage of lence of TTSS encoding toxin genes were 72% of P. aeruginosa Pseudomonas aeruginosa. J Med Microbiol. 1970;3:367–375. isolates carrying exoS gene and 28% exoU gene.22,23 Carriage of 17. Speert DP, Campbell ME, Davidson AG, et al. Pseudomonas aeruginosa exoS gene and exoU gene were mutually exclusive. The distribution colonization of the gastrointestinal tract in patients with cystic fibrosis. of TTSS encoding toxin genes in GI isolates was similar to that of J Infect Dis. 1993;167:226–229. general population. However, nearly 60% of Shanghai fever isolates 18. Steinbrückner B, Fehrenbach J, Philippczik G, et al. Clinical significance of pure or predominant growth of Pseudomonas aeruginosa in faecal speci- hold exoU gene. In the transition of P. aeruginosa from the soil to mens of medical patients. J Hosp Infect. 1999;43:164–165. 24 certain clinical settings, the loss of ExoS expression is favored. 19. Ohara T, Itoh K. Significance of Pseudomonas aeruginosa colonization of The acquisition of exoU genomic island may enhance colonization the gastrointestinal tract. Intern Med. 2003;42:1072–1076. 25 and survival in different host environments. Expression of exoU 20. Somekh E, Abishai V, Hanani M, et al. The clinical significance of is a marker for highly virulent clinical isolates.26 The role of TTSS Pseudomonas aeruginosa isolation from stool of neonates. Arch Pediatr encoding toxins, such as exoS and exoU involved in diarrhea is still Adolesc Med. 1996;150:108–109. unknown. However, there was no difference in clinical manifesta- 21. Stone HH, Kolb LD, Geheber CE. Bacteriologic considerations in perfo- tions of Shanghai fever between exoS and exoU strains.10 rated necrotizing enterocolitis. South Med J. 1979;72:1540–1544. Adhesion to host tissue is the first step for a pathogen to 22. Feltman H, Schulert G, Khan S, et al. Prevalence of type III secretion establish colonization. Shanghai fever isolates were more adher- genes in clinical and environmental isolates of Pseudomonas aeruginosa. Microbiology. 2001;147(Pt 10):2659–2669. ent to Caco-2 cells than GI isolates. Although the penetrative abil- 23. Pirnay JP, Bilocq F, Pot B, et al. Pseudomonas aeruginosa population struc- ity was similar between shanghai fever isolates and GI isolates, ture revisited. PLoS One. 2009;4:e7740. Shanghai fever isolates were more cytotoxic than GI isolates. The 27 24. Ferguson MW, Maxwell JA, Vincent TS, et al. Comparison of the exoS gene virulence of P. aeruginosa is multifactorial and combinatorial. and protein expression in soil and clinical isolates of Pseudomonas aerugi- Shanghai fever is caused by highly virulent strains of P. aeruginosa. nosa. Infect Immun. 2001;69:2198–2210. 25. Wolfgang MC, Kulasekara BR, Liang X, et al. Conservation of genome REFERENCES content and virulence determinants among clinical and environmen- 1. Richards MJ, Edwards JR, Culver DH, et al. Nosocomial infections in tal isolates of Pseudomonas aeruginosa. Proc Natl Acad Sci U S A. combined medical-surgical intensive care units in the United States. Infect 2003;100:8484–8489. Control Hosp Epidemiol. 2000;21:510–515. 26. Schulert GS, Feltman H, Rabin SD, et al. Secretion of the toxin ExoU is a 2. Pirnay JP, De Vos D, Cochez C, et al. Molecular epidemiology of marker for highly virulent Pseudomonas aeruginosa isolates obtained from Pseudomonas aeruginosa colonization in a burn unit: persistence of a patients with hospital-acquired pneumonia. J Infect Dis. 2003;188:1695–1706. multidrug-resistant clone and a silver sulfadiazine-resistant clone. J Clin 27. Lee DG, Urbach JM, Wu G, et al. Genomic analysis reveals that Pseudomonas Microbiol. 2003;41:1192–1202. aeruginosa virulence is combinatorial. Genome Biol. 2006;7:R90.