- Email: [email protected]
One-year surveillance of legionellosis in burned patients and Legionella environmental monitoring
Burns 31 (2005) 50–54 www.elsevier.com/locate/burns
One-year surveillance of legionellosis in burned patients and Legionella environmental monitoring L. Franzina,*, M. Stellab, T. Zaccariaa, D. Cabodia, M. Castellani Pastorisc a
Infectious Diseases Unit, University of Turin, Corso Svizzera 164, 10149 Turin, Italy b Burn Centre, CTO Hospital, Via Zuretti 29, 10126 Turin, Italy c Department of Bacteriology and Medical Mycology, Istituto Superiore di Sanita`, Viale Regina Elena 299, 00161 Rome, Italy Accepted 16 June 2004
Abstract Burned patients have a theoretically high risk of Legionella infection because burns produce a compromised immune system. Cutaneous surfaces are without protective barriers, and bathing tank water is frequently used for washing and caring. A one-year surveillance study was performed on 65 burned patients by antibody determination and by culture of bronchial aspirates. Environmental culturing for Legionella was done in the patients’ care areas every four months during the same period. Low titers ranging from 8 to 32 were found in 30 (46.1%) subjects against 18 antigens including several Legionella species. No increase in antibody titers was shown in 193 patients’ sera. Cultures of respiratory samples were negative. L. pneumophila serogroups 4, 5, 6 and 8 and L. rubrilucens were isolated from 55.5% of water samples. Despite no evidence of Legionella infection among patients included in this study, the authors believe it to be advisable to improve control measures in hospital water supplies, used by burned patients, to minimise the risk of legionellosis. # 2004 Elsevier Ltd and ISBI. All rights reserved. Keywords: Legionella; Burns; Culture; Environment.
1. Introduction Nosocomial legionellosis is more frequent in immunocompromised patients (immunosuppressed or with previous diseases and transplant recipients) [1,2]. An association between Legionella infection and exposure to contaminated water has been described [3]. Legionnaires’ disease is thought to be acquired by inhalation of aerosols containing Legionella or by microaspiration of contaminated water [4]. Extrapulmonary infections through topical contact, without overt pulmonary involvement, as well as wound infections following contact with contaminated water, are rarely reported [5]. Burned patients have an increased risk of nosocomial infection because of lack of protective barriers, a * Corresponding author. Tel.: +390114393908; fax: +390114393908. E-mail address: [email protected] (L. Franzin). 0305-4179/$30.00 # 2004 Elsevier Ltd and ISBI. All rights reserved. doi:10.1016/j.burns.2004.06.009
compromised immune system [6] and often a long hospitalization period. Respiratory complications due to pulmonary infection are considered the major cause of death for these patients [7,8]. Because bathing tanks are used periodically for washing and caring, if the hospital water supply is contaminated by Legionella, burned patients may be at additonal risk for legionellosis during bathing by two types of exposition: inhalation of aerosols and water contact by cutaneous lesions. However, studies concerning legionellosis in these patients are lacking. The aims of the study were: (1) to recognize a possible Legionella infection of nosocomial origin in burned patients hospitalized at the Burn Centre of Turin, by antibody determination in sequential sera and by culture from clinical respiratory samples during a one-year study period; (2) an environmental microbiological investigation for Legionella detection from hot tap water and from bathing tank water at the Burn Centre during the same period.
L. Franzin et al. / Burns 31 (2005) 50–54
2. Materials and methods 2.1. Burned patients Sixty five patients (41 males and 24 females aged 15 to 85 years, mean age 46 years) affected by extensive burns were enrolled in the study. The hospital stay period ranged from 5 to 163 days. The total burned surface area (TBSA) ranged from 10 to 90% (mean value 33%). Full thickness burns ranged from 5 to 60% (mean value 15%). Burns were produced by thermal injury in 30 patients. The fatality rate was 18.5% (24.4% for males and 8.3% for females). The age of patients with fatal outcome ranged from 24 to 85 years (mean age 61 years); most of them had more than 60 years. Half of the patients under study had respiratory tract burns, suspected because of the burns’ site and type. Some of them had compromised respiratory function and required assisted ventilation. Twenty seven patients had pulmonary complication confirmed by chest radiograph showing pulmonary infiltrates. All the subjects who died, included in this study, showed pulmonary infective complications, and severe burns of the respiratory tract. Sera of 65 patients were collected at hospitalization, during the hospital stay and before discharge. A total of 193 sera (at least two for patients) were examined. Eighty bronchial aspirates were collected from 15 patients with pulmonary infection confirmed by chest radiograph.
51
micdadei and L. dumoffii; V: L.gormanii, L. jordanis and L. longbeacheae. Polivalent antigens were prepared at the Istituto Superiore di Sanita`, Rome, by Dr M. Castellani Pastoris [9]. A four-fold increase of antibody titer to at least 128 was considered evidence of confirmed infection of Legionella. Respiratory tract samples were diluited in Nutrient Broth and homogenized. Microscopic examination was performed by Gram and Gimenez stain. Aliquots (0.1 mL) of the untreated, heat-treated and acid washed suspensions [10] were plated onto BCYE a, BMPA a, and MWY agar media. The plates were incubated at 37 8C in humidified atmosphere and examined daily for 15 days [11,12]. Environmental samples were processed as previously reported [12]. Briefly, all samples were concentrated by filtration through cellulose acetate membrane filters (Sartorious, 0.2 mm pore size) and resuspended into 10 mL of the same sample by swabbing the filters and vortexing. Aliquots (0.1 mL) of untreated, heat-treated and acid-treated were plated on the same media as used for clinical samples. Colonies with typical morphology, gram-negative staining and requiring L-cysteine for the growth, were further identified serologically by latex agglutination (Oxoid, Italy), by agglutination with polyclonal rabbit antisera (Biogenetics, Italy) and by direct immunofluorescence assay (MarDx, U.S.A.).
2.2. Environmental sampling 3. Results Nine patients’ rooms, two bathing rooms with tank and an operating theatre are present at the Burn Centre. The most severly compromised patient is hospitalized in one of the nine room (isolation room) in which there is a bathing tank. The ward is provided by air filters to reduce environmental microbial contamination. Hot water basins are distant about three meters from patients beds. The taps water temperature is controlled by thermostatic system that mix hot and cold water before emission. Five-liter samples of hot tap water from two bathing tanks were collected every four months (January, May, September). Hot water samples were also taken from shower-heads of the same bathing tanks by swabbing the heads and resuspending the swab in 100 mL of hot water from showerhead. Hot tap water (5 l) of surgical room and of four patients’ rooms was also sampled. Water temperature and free chlorine content were determined after collection. 2.3. Laboratory investigations Antibodies were detected in sequential sera by the indirect immunofluorescence test (IFA) using L. pneumophila serogroup 1 antigen, L. rubrilucens antigen, and polyvalent antigens containing I: L. pneumophila serogroups 1–3; II: L. pneumophila serogroups 4–6; III: L. pneumophila serogroups 7–10; IV: L. bozemanii serogroup 1 and 2, L.
3.1. Burned patients Antibody titers <8 were found in 108 sera of burned patients for all the antigens tested. In the other 85 sera antibody titers ranging from 8 to 32 were observed, as shown in Table 1. Sequential sera belonging to 35 patients showed antibody titers <8 against all the antigens used. The titers against L. pneumophila serogroup 1 were always <8 in all
Table 1 Results of antibody titers against Legionella in sera of burned patients by indirect immunofluorescence assay Antibody titer
Legionella antigen Lp 1a
<8 8 16 32 a
193
Lrb
Ic
II
III
IV
V
185 8
179 12 2
190 3
182 9 1 1
144 30 16 3
L. pneumophila serogroup 1. L. rubrilucens. c I: antigens including L. pneumophila serogroups 1, 2, 3; II: antigens L. pneumophila serogroups 4, 5, 6; III: antigens including L. pneumophila serogroups 7, 8, 9, 10; IV: antigens L. bozemanii serogroup 1 and 2, L. micdadei, L. dumoffii; V: antigens L. gormanii, L. jordanis and L. longbeachae. b
52
L. Franzin et al. / Burns 31 (2005) 50–54
inhalation of aerosols containing Legionella or sometimes by microaspiration of contaminated water [4]. In extrapulmonary disease, the infection is likely due to bacteremia. Isolation of L. pneumophila from an abscess at a haemodialysis fistula site [15] and of L. micdadei from a skin abscess on the leg of an immunosuppressed patient [16] are described, probably caused by bacteraemic seeding via the lungs [4]. Legionella infections not associated with pneumonia have been reported. The mode of infection may be the direct inoculation of Legionella bacteria into wounds by contaminated water used for their irrigating or washing [4,5]. Brabender et al. [17] reported about a patient in whom a hip wound became infected by L. pneumophila and P. aeruginosa following bathing in a contaminated whirlpool spa and Arnow et al. [18] described a perirectal abscess due to Legionella after multiple tap water enemas. Three patients with sternal wound infection by water contact after scrubbing the chest area have been reported [19]. Burned patients have an increased risk of infection because the cutaneous surface is without protective barriers and burns often produce a compromission of the immune system [6]. The frequent use of bathing tanks for washing and caring and a long hospital stay are risk factors for burned patients. In the present study, clinical and microbiological surveillance and environmental monitoring for Legionella species were therefore performed for one-year at the Burn Centre. The serological survey using antigens of different Legionella species and serogroups did not show significant increase in antibody titers in burned patients, even if low titers ranging from 8 to 32 against several antigens found in 30 (46.1%) subjects (Table 1). However, cross-reactive antibody titers are occasionally reported in the literature for patients with non-Legionella infections, especially Pseudomonas, when antibody titers against Legionella antigens other than L. pneumophila serogroup 1 are measured [20]. For these reasons the interpretative criteria for sera reactivity against antigens other than L. pneumophila serogroup 1 and other species are not clearly established. In the present study
the sera of the 65 patients, while antibody titers from 8 to 32 were found in 30 (46.1%) subjects for polivalent antigens I, 7, 7, 2, 6, and 29 patients respectively. Dividing these 30 patients in five groups by their period of hospitalization (<1 month, 1–2 months, 2–3 months, 4–5 months, >5 months) no significant increase in antibody titers was observed in subjects with long hospital stay. Legionella was not isolated from 80 bronchial aspirates of 15 subjects. The hospitalization period of these patients ranged from 30 to 150 days, with a mean stay of 75 days. Despite the addition of, selective supplement to Legionella culture media, microrganisms (stafilococci, Pseudomonas and Candida) often responsible for nosocomial infections, grew on the plates, possibly leading to an inhibitory effect on Legionella growth.
3.2. Environmental culture Legionella was isolated from 15 (55.5%) out of 27 water samples. All the hot water samples (mean water temperature 39.6 8C) collected every 4 months during the one-year study, from taps and the shower-heads (swab and water) of the bathing tank at the isolation room, were positive. The hot tap water of the bathing room yielded Legionella in two samples, while shower-head was positive in one sample and negative in two (<1 cfu/mL). Water samples (mean water temperature 23.5 8C) from two out of four rooms were always negative (<20 cfu/L). Positive hot tap water samples yielded 40 to 7000 cfu/L of Legionella, shower-heads water 1–90 cfu/mL. The results of Legionella cultures are shown in Table 2. The strains were typed as L. pneumophila of serogroup 4, 5, 6, 8 and L. rubrilucens.
4. Discussion Nosocomial legionellosis is often unrecognized unless specialized laboratory tests for Legionella are introduced for routine use [13,14]. Legionnaires’ disease is acquired by
Table 2 Results of Legionella isolation from water samples of Burn Centre Source
Sample site
Hot Hot Hot Hot Hot Tap Tap Tap Tap
Bathing tank of Bathing tank of Bathing tank of Bathing tank of Surgical room Room 1 Room 2 Room 3 Room 4
tap water shower–head water tap water shower–head water tap water water water water water
Water temperature (8C) of sample
bathing room bathing room isolation room isolation room
Free chlorine (ppm) of sample
Legionella (cfu) of sample
N.1
N.2
N.3
N.1
N.2
N.3
N.1
N.2
N.3
38 36 42 42 30 20 15 25 32
37 35 40 31 37 29 21 29 37
35 35 42 41 33 28 19 32 42
0.1 0.1 0.2 0.1 0.1 0.1 0.2 0.1 0.1
0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1
0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.1
2400/L 90/mL 7000/L 10/mL 60/L 80/L – – –
1000/L – 40/L 20/mL 400/L 800/L – – 1800/L
– – 1800/L 1/mL – – – – 2000/L
L. Franzin et al. / Burns 31 (2005) 50–54
the prolonged hospital stay seems not to affect the antibody titers. Since L. rubrilucens was isolated from hospital water, patients’ sera were also checked for reactivity against this microorganism, not included in polivalent antigens IVand V. No positivity was detected in the sera tested. Legionella was not isolated from repeated respiratory samples examined for 15 patients, but antibiotic therapy may have influenced the culture results. Moreover, microrganisms that may have inhibited Legionella growth, and particularly Pseudomonas [21], were isolated on the nonselective and selective culture plates. Urinary antigen detection was not done, because the test was not commercially available when the study was performed. Environmental samples showed contamination of the hot water supply by five different L. pneumophila serogroups and by L. rubrilucens. L. pneumophila serogroup 1, that is usually considered the most virulent, was not isolated in this study, but was found in a previous control at the Burn Centre. Burned patients enrolled in this one-year study have an increased risk of infection. Nevertheless, despite the above environmental results, nosocomial legionellosis was not observed. Infections should however, not be excluded in a longer period of observation, because this population, as previously discussed, have an increased risk of infection; moreover, Legionella contamination in hospital water was up to 7000 cfu/L and two different species and four serogroups of L. pneumophila were found. Even if the human infective dose has not been established, levels of 102– 104 cfu/L are considered able to determine one case of infection in a year [22]. Pulmonary infections in burned patients are the major cause of death [7,8] and P. aeruginosa is often the most commonly recovered organism [23,24]. Co-infections by Legionella and Pseudomonas are also reported in not burned patients [4]. Dual infection cannot be excluded, when samples negative for Legionella yielded Pseudomonas growth, because of the inhibitory effect of Pseudomonas against Legionella [21]. As the extrapulmonary Legionella infections reported in the literature often result from direct topical exposure of susceptible tissue to contaminated tap water, an additional cutaneous risk for burned patients should be considered, also because of the immune system’s compromise caused by the burns themselves. To our knowledge, this is the first prospective study concerning legionellosis surveillance in burned patients. On the basis of the results obtained Legionella does not appear an important cause of illness for burned patients, at least in the considered Burn Centre. However, due to the compromission of the immune system of these patients and that different species or serogroups of Legionella, all potentially pathogens for humans, can be present in the water supply, we believe strongly advisable to carefully monitor tap water used for burned patients, particularly in dressing changes and bathing, as suggested for other surgical patients [5]. Efforts should be made to improve the control measures
53
(disinfection, correct maintenance and procedures, etc.) and possibly to reach a system with no Legionella contamination in high risk patient wards in order to minimise the risk of legionellosis at the hospital.
Acknowledgements We thank Dr Robert Benson (CDC, Atlanta, Georgia, U.S.A.) for typing L. rubrilucens and Specchio dei Tempi, La Stampa Foundation, Turin, Italy for financial support.
References [1] Marshall W, Foster Jr RS, Winn W. Legionnaires’ disease in renal transplant recipients. Am J Surg 1981;141:423–39. [2] Chow J, Yu VL. Legionella: a major opportunistic pathogen in transplant recipients. Sem Resp Infect 1998;13:132–9. [3] Baskerville A, Fitzgeorge RB, Broster M, Hambleton P, Dennis PJ. Experimental transmission of Legionnaires’disease by exposure to aerosols of Legionella pneumophila. Lancet 1981;ii:1389–90. [4] Edelstein PH. Legionnaires’disease. Clin Infect Dis 1993;16: 741–9. [5] Lowry PW, Tompkins LS. Nosocomial legionellosis: a review of pulmonary and extrapulmonary syndromes. Am J Infect Control 1993;21:21–7. [6] Youn Y-K, Lalonde C, Demling R. The role of mediators in the response to thermal injury. World J Surg 1992;16:30–6. [7] Herndon DN, Barrow RE, Linares HA. Inhalation injury in burned patients: effects and treatment. Burns 1988;14:349–56. [8] Shook CD, MacMillan BG, Altemeier WA. Pulmonary complications of the burn patient. Arch Surg 1968;97:215–24. [9] Castellani Pastoris M, Ciarrocchi S, Di Capua A, Temperanza AM. Comparison of phenol- and heat-killed antigens in the indirect immunofluorescence test for serodiagnosis of Legionella pneumophila serogroup 1 infections. J Clin Microbiol 1984;20:780–3. [10] Bopp CA, Summer JW, Morris GK, Wells JG. Isolation of Legionella spp. from environmental water samples by low-pH treatment and use of a selective medium. J Clin Microbiol 1981;13:714–49. [11] Wilkinson HW. Hospital-laboratory diagnosis of Legionella infections. Atlanta: Centers for Disease Control and Prevention, 1988. p. 43. [12] Franzin L, Castellani Pastoris M, Gioannini P, Villani G. Endemicity of Legionella pneumophila serogroup 3 in a hospital water supply. J Hosp Infect 1989;13:281–8. [13] Stout JE, Yu VL. Legionellosis. N Engl J Med 1997;337:682–7. [14] Yu VL. Nosocomial legionellosis. Curr Opin Infect Dis 2000;13: 385–8. [15] Kalweit WH, Winn Jr WC, Rocco Jr TA, Girod JC. Hemodialysis fistula infections caused by Legionella pneumophila. Ann Intern Med 1982;96:173–5. [16] Ampel NM, Ruben FL, Norden CW. Cutaneous abscess caused by Legionella micdadei in an immunosuppressed patient. Ann Intern Med 1985;102:630–2. [17] Brabender W, Hinthorn DR, Asher M, Lindsey NJ, Liu C. Legionella pneumophila wound infection. J Am Med Assoc 1983;250:3091–2. [18] Arnow PM, Boyko EJ, Friedman EL. Perirectal abscess caused by Legionella pneumophila and mixed anaerobic bacteria. Ann Intern Med 1983;98:184–5. [19] Lowry PH, Blankenship RJ, Gridley W, Tompkins LS. A cluster of Legionella sternal-wound infections due to postoperative topical exposure to contaminated tap water. N Engl Med 1991;324: 109–13.
54
L. Franzin et al. / Burns 31 (2005) 50–54
[20] Edelstein PH. Detection of antibodies to Legionella spp. In: Rose NR, De Macario ED, Folds JD, Lane HC, Nakamura RM., editors. Manual of Clinical Laboratory Immunology. 5th ed. Washington DC: American Society for Microbiology; 1997. p. 502–9. [21] Gomez-Lus R, Lomba E, Gomez-Lus P, et al. In vitro antagonistic activity of Pseudomonas aeruginosa, Klebsiella pneumoniae, and Aeromonas spp. against Legionella spp. In: Barbaree JM, Breiman RF, Dufour AP, editors. Legionella Current Status and Emerging Perspectives. Washington DC: American Society for Microbiology; 1993. p. 265–7.
[22] Ezzedine H, Van Ossel C, Delmee M, Wauters C. Legionella spp. in a hospital hot water system: effect of control measures. J Hosp Infect 1989;13:121–31. [23] Still J, Newton T, Friedman B, Furhman S, Law E, Dawson J. Experience with pneumonia in acutely burned patients requiring ventilator support. Am Surg 2000;66:206–9. [24] Douglas MW, Mulholland K, Denyer VV, Gottlieb T. Multi-drug resistant Pseudomonas aeruginosa outbreak in a burns unit–an infection control study. Burns 2001;27:131–5.
One-year surveillance of legionellosis in burned patients and Legionella environmental monitoring L. Franzina,*, M. Stellab, T. Zaccariaa, D. Cabodia, M. Castellani Pastorisc a
Infectious Diseases Unit, University of Turin, Corso Svizzera 164, 10149 Turin, Italy b Burn Centre, CTO Hospital, Via Zuretti 29, 10126 Turin, Italy c Department of Bacteriology and Medical Mycology, Istituto Superiore di Sanita`, Viale Regina Elena 299, 00161 Rome, Italy Accepted 16 June 2004
Abstract Burned patients have a theoretically high risk of Legionella infection because burns produce a compromised immune system. Cutaneous surfaces are without protective barriers, and bathing tank water is frequently used for washing and caring. A one-year surveillance study was performed on 65 burned patients by antibody determination and by culture of bronchial aspirates. Environmental culturing for Legionella was done in the patients’ care areas every four months during the same period. Low titers ranging from 8 to 32 were found in 30 (46.1%) subjects against 18 antigens including several Legionella species. No increase in antibody titers was shown in 193 patients’ sera. Cultures of respiratory samples were negative. L. pneumophila serogroups 4, 5, 6 and 8 and L. rubrilucens were isolated from 55.5% of water samples. Despite no evidence of Legionella infection among patients included in this study, the authors believe it to be advisable to improve control measures in hospital water supplies, used by burned patients, to minimise the risk of legionellosis. # 2004 Elsevier Ltd and ISBI. All rights reserved. Keywords: Legionella; Burns; Culture; Environment.
1. Introduction Nosocomial legionellosis is more frequent in immunocompromised patients (immunosuppressed or with previous diseases and transplant recipients) [1,2]. An association between Legionella infection and exposure to contaminated water has been described [3]. Legionnaires’ disease is thought to be acquired by inhalation of aerosols containing Legionella or by microaspiration of contaminated water [4]. Extrapulmonary infections through topical contact, without overt pulmonary involvement, as well as wound infections following contact with contaminated water, are rarely reported [5]. Burned patients have an increased risk of nosocomial infection because of lack of protective barriers, a * Corresponding author. Tel.: +390114393908; fax: +390114393908. E-mail address: [email protected] (L. Franzin). 0305-4179/$30.00 # 2004 Elsevier Ltd and ISBI. All rights reserved. doi:10.1016/j.burns.2004.06.009
compromised immune system [6] and often a long hospitalization period. Respiratory complications due to pulmonary infection are considered the major cause of death for these patients [7,8]. Because bathing tanks are used periodically for washing and caring, if the hospital water supply is contaminated by Legionella, burned patients may be at additonal risk for legionellosis during bathing by two types of exposition: inhalation of aerosols and water contact by cutaneous lesions. However, studies concerning legionellosis in these patients are lacking. The aims of the study were: (1) to recognize a possible Legionella infection of nosocomial origin in burned patients hospitalized at the Burn Centre of Turin, by antibody determination in sequential sera and by culture from clinical respiratory samples during a one-year study period; (2) an environmental microbiological investigation for Legionella detection from hot tap water and from bathing tank water at the Burn Centre during the same period.
L. Franzin et al. / Burns 31 (2005) 50–54
2. Materials and methods 2.1. Burned patients Sixty five patients (41 males and 24 females aged 15 to 85 years, mean age 46 years) affected by extensive burns were enrolled in the study. The hospital stay period ranged from 5 to 163 days. The total burned surface area (TBSA) ranged from 10 to 90% (mean value 33%). Full thickness burns ranged from 5 to 60% (mean value 15%). Burns were produced by thermal injury in 30 patients. The fatality rate was 18.5% (24.4% for males and 8.3% for females). The age of patients with fatal outcome ranged from 24 to 85 years (mean age 61 years); most of them had more than 60 years. Half of the patients under study had respiratory tract burns, suspected because of the burns’ site and type. Some of them had compromised respiratory function and required assisted ventilation. Twenty seven patients had pulmonary complication confirmed by chest radiograph showing pulmonary infiltrates. All the subjects who died, included in this study, showed pulmonary infective complications, and severe burns of the respiratory tract. Sera of 65 patients were collected at hospitalization, during the hospital stay and before discharge. A total of 193 sera (at least two for patients) were examined. Eighty bronchial aspirates were collected from 15 patients with pulmonary infection confirmed by chest radiograph.
51
micdadei and L. dumoffii; V: L.gormanii, L. jordanis and L. longbeacheae. Polivalent antigens were prepared at the Istituto Superiore di Sanita`, Rome, by Dr M. Castellani Pastoris [9]. A four-fold increase of antibody titer to at least 128 was considered evidence of confirmed infection of Legionella. Respiratory tract samples were diluited in Nutrient Broth and homogenized. Microscopic examination was performed by Gram and Gimenez stain. Aliquots (0.1 mL) of the untreated, heat-treated and acid washed suspensions [10] were plated onto BCYE a, BMPA a, and MWY agar media. The plates were incubated at 37 8C in humidified atmosphere and examined daily for 15 days [11,12]. Environmental samples were processed as previously reported [12]. Briefly, all samples were concentrated by filtration through cellulose acetate membrane filters (Sartorious, 0.2 mm pore size) and resuspended into 10 mL of the same sample by swabbing the filters and vortexing. Aliquots (0.1 mL) of untreated, heat-treated and acid-treated were plated on the same media as used for clinical samples. Colonies with typical morphology, gram-negative staining and requiring L-cysteine for the growth, were further identified serologically by latex agglutination (Oxoid, Italy), by agglutination with polyclonal rabbit antisera (Biogenetics, Italy) and by direct immunofluorescence assay (MarDx, U.S.A.).
2.2. Environmental sampling 3. Results Nine patients’ rooms, two bathing rooms with tank and an operating theatre are present at the Burn Centre. The most severly compromised patient is hospitalized in one of the nine room (isolation room) in which there is a bathing tank. The ward is provided by air filters to reduce environmental microbial contamination. Hot water basins are distant about three meters from patients beds. The taps water temperature is controlled by thermostatic system that mix hot and cold water before emission. Five-liter samples of hot tap water from two bathing tanks were collected every four months (January, May, September). Hot water samples were also taken from shower-heads of the same bathing tanks by swabbing the heads and resuspending the swab in 100 mL of hot water from showerhead. Hot tap water (5 l) of surgical room and of four patients’ rooms was also sampled. Water temperature and free chlorine content were determined after collection. 2.3. Laboratory investigations Antibodies were detected in sequential sera by the indirect immunofluorescence test (IFA) using L. pneumophila serogroup 1 antigen, L. rubrilucens antigen, and polyvalent antigens containing I: L. pneumophila serogroups 1–3; II: L. pneumophila serogroups 4–6; III: L. pneumophila serogroups 7–10; IV: L. bozemanii serogroup 1 and 2, L.
3.1. Burned patients Antibody titers <8 were found in 108 sera of burned patients for all the antigens tested. In the other 85 sera antibody titers ranging from 8 to 32 were observed, as shown in Table 1. Sequential sera belonging to 35 patients showed antibody titers <8 against all the antigens used. The titers against L. pneumophila serogroup 1 were always <8 in all
Table 1 Results of antibody titers against Legionella in sera of burned patients by indirect immunofluorescence assay Antibody titer
Legionella antigen Lp 1a
<8 8 16 32 a
193
Lrb
Ic
II
III
IV
V
185 8
179 12 2
190 3
182 9 1 1
144 30 16 3
L. pneumophila serogroup 1. L. rubrilucens. c I: antigens including L. pneumophila serogroups 1, 2, 3; II: antigens L. pneumophila serogroups 4, 5, 6; III: antigens including L. pneumophila serogroups 7, 8, 9, 10; IV: antigens L. bozemanii serogroup 1 and 2, L. micdadei, L. dumoffii; V: antigens L. gormanii, L. jordanis and L. longbeachae. b
52
L. Franzin et al. / Burns 31 (2005) 50–54
inhalation of aerosols containing Legionella or sometimes by microaspiration of contaminated water [4]. In extrapulmonary disease, the infection is likely due to bacteremia. Isolation of L. pneumophila from an abscess at a haemodialysis fistula site [15] and of L. micdadei from a skin abscess on the leg of an immunosuppressed patient [16] are described, probably caused by bacteraemic seeding via the lungs [4]. Legionella infections not associated with pneumonia have been reported. The mode of infection may be the direct inoculation of Legionella bacteria into wounds by contaminated water used for their irrigating or washing [4,5]. Brabender et al. [17] reported about a patient in whom a hip wound became infected by L. pneumophila and P. aeruginosa following bathing in a contaminated whirlpool spa and Arnow et al. [18] described a perirectal abscess due to Legionella after multiple tap water enemas. Three patients with sternal wound infection by water contact after scrubbing the chest area have been reported [19]. Burned patients have an increased risk of infection because the cutaneous surface is without protective barriers and burns often produce a compromission of the immune system [6]. The frequent use of bathing tanks for washing and caring and a long hospital stay are risk factors for burned patients. In the present study, clinical and microbiological surveillance and environmental monitoring for Legionella species were therefore performed for one-year at the Burn Centre. The serological survey using antigens of different Legionella species and serogroups did not show significant increase in antibody titers in burned patients, even if low titers ranging from 8 to 32 against several antigens found in 30 (46.1%) subjects (Table 1). However, cross-reactive antibody titers are occasionally reported in the literature for patients with non-Legionella infections, especially Pseudomonas, when antibody titers against Legionella antigens other than L. pneumophila serogroup 1 are measured [20]. For these reasons the interpretative criteria for sera reactivity against antigens other than L. pneumophila serogroup 1 and other species are not clearly established. In the present study
the sera of the 65 patients, while antibody titers from 8 to 32 were found in 30 (46.1%) subjects for polivalent antigens I, 7, 7, 2, 6, and 29 patients respectively. Dividing these 30 patients in five groups by their period of hospitalization (<1 month, 1–2 months, 2–3 months, 4–5 months, >5 months) no significant increase in antibody titers was observed in subjects with long hospital stay. Legionella was not isolated from 80 bronchial aspirates of 15 subjects. The hospitalization period of these patients ranged from 30 to 150 days, with a mean stay of 75 days. Despite the addition of, selective supplement to Legionella culture media, microrganisms (stafilococci, Pseudomonas and Candida) often responsible for nosocomial infections, grew on the plates, possibly leading to an inhibitory effect on Legionella growth.
3.2. Environmental culture Legionella was isolated from 15 (55.5%) out of 27 water samples. All the hot water samples (mean water temperature 39.6 8C) collected every 4 months during the one-year study, from taps and the shower-heads (swab and water) of the bathing tank at the isolation room, were positive. The hot tap water of the bathing room yielded Legionella in two samples, while shower-head was positive in one sample and negative in two (<1 cfu/mL). Water samples (mean water temperature 23.5 8C) from two out of four rooms were always negative (<20 cfu/L). Positive hot tap water samples yielded 40 to 7000 cfu/L of Legionella, shower-heads water 1–90 cfu/mL. The results of Legionella cultures are shown in Table 2. The strains were typed as L. pneumophila of serogroup 4, 5, 6, 8 and L. rubrilucens.
4. Discussion Nosocomial legionellosis is often unrecognized unless specialized laboratory tests for Legionella are introduced for routine use [13,14]. Legionnaires’ disease is acquired by
Table 2 Results of Legionella isolation from water samples of Burn Centre Source
Sample site
Hot Hot Hot Hot Hot Tap Tap Tap Tap
Bathing tank of Bathing tank of Bathing tank of Bathing tank of Surgical room Room 1 Room 2 Room 3 Room 4
tap water shower–head water tap water shower–head water tap water water water water water
Water temperature (8C) of sample
bathing room bathing room isolation room isolation room
Free chlorine (ppm) of sample
Legionella (cfu) of sample
N.1
N.2
N.3
N.1
N.2
N.3
N.1
N.2
N.3
38 36 42 42 30 20 15 25 32
37 35 40 31 37 29 21 29 37
35 35 42 41 33 28 19 32 42
0.1 0.1 0.2 0.1 0.1 0.1 0.2 0.1 0.1
0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1
0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.1
2400/L 90/mL 7000/L 10/mL 60/L 80/L – – –
1000/L – 40/L 20/mL 400/L 800/L – – 1800/L
– – 1800/L 1/mL – – – – 2000/L
L. Franzin et al. / Burns 31 (2005) 50–54
the prolonged hospital stay seems not to affect the antibody titers. Since L. rubrilucens was isolated from hospital water, patients’ sera were also checked for reactivity against this microorganism, not included in polivalent antigens IVand V. No positivity was detected in the sera tested. Legionella was not isolated from repeated respiratory samples examined for 15 patients, but antibiotic therapy may have influenced the culture results. Moreover, microrganisms that may have inhibited Legionella growth, and particularly Pseudomonas [21], were isolated on the nonselective and selective culture plates. Urinary antigen detection was not done, because the test was not commercially available when the study was performed. Environmental samples showed contamination of the hot water supply by five different L. pneumophila serogroups and by L. rubrilucens. L. pneumophila serogroup 1, that is usually considered the most virulent, was not isolated in this study, but was found in a previous control at the Burn Centre. Burned patients enrolled in this one-year study have an increased risk of infection. Nevertheless, despite the above environmental results, nosocomial legionellosis was not observed. Infections should however, not be excluded in a longer period of observation, because this population, as previously discussed, have an increased risk of infection; moreover, Legionella contamination in hospital water was up to 7000 cfu/L and two different species and four serogroups of L. pneumophila were found. Even if the human infective dose has not been established, levels of 102– 104 cfu/L are considered able to determine one case of infection in a year [22]. Pulmonary infections in burned patients are the major cause of death [7,8] and P. aeruginosa is often the most commonly recovered organism [23,24]. Co-infections by Legionella and Pseudomonas are also reported in not burned patients [4]. Dual infection cannot be excluded, when samples negative for Legionella yielded Pseudomonas growth, because of the inhibitory effect of Pseudomonas against Legionella [21]. As the extrapulmonary Legionella infections reported in the literature often result from direct topical exposure of susceptible tissue to contaminated tap water, an additional cutaneous risk for burned patients should be considered, also because of the immune system’s compromise caused by the burns themselves. To our knowledge, this is the first prospective study concerning legionellosis surveillance in burned patients. On the basis of the results obtained Legionella does not appear an important cause of illness for burned patients, at least in the considered Burn Centre. However, due to the compromission of the immune system of these patients and that different species or serogroups of Legionella, all potentially pathogens for humans, can be present in the water supply, we believe strongly advisable to carefully monitor tap water used for burned patients, particularly in dressing changes and bathing, as suggested for other surgical patients [5]. Efforts should be made to improve the control measures
53
(disinfection, correct maintenance and procedures, etc.) and possibly to reach a system with no Legionella contamination in high risk patient wards in order to minimise the risk of legionellosis at the hospital.
Acknowledgements We thank Dr Robert Benson (CDC, Atlanta, Georgia, U.S.A.) for typing L. rubrilucens and Specchio dei Tempi, La Stampa Foundation, Turin, Italy for financial support.
References [1] Marshall W, Foster Jr RS, Winn W. Legionnaires’ disease in renal transplant recipients. Am J Surg 1981;141:423–39. [2] Chow J, Yu VL. Legionella: a major opportunistic pathogen in transplant recipients. Sem Resp Infect 1998;13:132–9. [3] Baskerville A, Fitzgeorge RB, Broster M, Hambleton P, Dennis PJ. Experimental transmission of Legionnaires’disease by exposure to aerosols of Legionella pneumophila. Lancet 1981;ii:1389–90. [4] Edelstein PH. Legionnaires’disease. Clin Infect Dis 1993;16: 741–9. [5] Lowry PW, Tompkins LS. Nosocomial legionellosis: a review of pulmonary and extrapulmonary syndromes. Am J Infect Control 1993;21:21–7. [6] Youn Y-K, Lalonde C, Demling R. The role of mediators in the response to thermal injury. World J Surg 1992;16:30–6. [7] Herndon DN, Barrow RE, Linares HA. Inhalation injury in burned patients: effects and treatment. Burns 1988;14:349–56. [8] Shook CD, MacMillan BG, Altemeier WA. Pulmonary complications of the burn patient. Arch Surg 1968;97:215–24. [9] Castellani Pastoris M, Ciarrocchi S, Di Capua A, Temperanza AM. Comparison of phenol- and heat-killed antigens in the indirect immunofluorescence test for serodiagnosis of Legionella pneumophila serogroup 1 infections. J Clin Microbiol 1984;20:780–3. [10] Bopp CA, Summer JW, Morris GK, Wells JG. Isolation of Legionella spp. from environmental water samples by low-pH treatment and use of a selective medium. J Clin Microbiol 1981;13:714–49. [11] Wilkinson HW. Hospital-laboratory diagnosis of Legionella infections. Atlanta: Centers for Disease Control and Prevention, 1988. p. 43. [12] Franzin L, Castellani Pastoris M, Gioannini P, Villani G. Endemicity of Legionella pneumophila serogroup 3 in a hospital water supply. J Hosp Infect 1989;13:281–8. [13] Stout JE, Yu VL. Legionellosis. N Engl J Med 1997;337:682–7. [14] Yu VL. Nosocomial legionellosis. Curr Opin Infect Dis 2000;13: 385–8. [15] Kalweit WH, Winn Jr WC, Rocco Jr TA, Girod JC. Hemodialysis fistula infections caused by Legionella pneumophila. Ann Intern Med 1982;96:173–5. [16] Ampel NM, Ruben FL, Norden CW. Cutaneous abscess caused by Legionella micdadei in an immunosuppressed patient. Ann Intern Med 1985;102:630–2. [17] Brabender W, Hinthorn DR, Asher M, Lindsey NJ, Liu C. Legionella pneumophila wound infection. J Am Med Assoc 1983;250:3091–2. [18] Arnow PM, Boyko EJ, Friedman EL. Perirectal abscess caused by Legionella pneumophila and mixed anaerobic bacteria. Ann Intern Med 1983;98:184–5. [19] Lowry PH, Blankenship RJ, Gridley W, Tompkins LS. A cluster of Legionella sternal-wound infections due to postoperative topical exposure to contaminated tap water. N Engl Med 1991;324: 109–13.
54
L. Franzin et al. / Burns 31 (2005) 50–54
[20] Edelstein PH. Detection of antibodies to Legionella spp. In: Rose NR, De Macario ED, Folds JD, Lane HC, Nakamura RM., editors. Manual of Clinical Laboratory Immunology. 5th ed. Washington DC: American Society for Microbiology; 1997. p. 502–9. [21] Gomez-Lus R, Lomba E, Gomez-Lus P, et al. In vitro antagonistic activity of Pseudomonas aeruginosa, Klebsiella pneumoniae, and Aeromonas spp. against Legionella spp. In: Barbaree JM, Breiman RF, Dufour AP, editors. Legionella Current Status and Emerging Perspectives. Washington DC: American Society for Microbiology; 1993. p. 265–7.
[22] Ezzedine H, Van Ossel C, Delmee M, Wauters C. Legionella spp. in a hospital hot water system: effect of control measures. J Hosp Infect 1989;13:121–31. [23] Still J, Newton T, Friedman B, Furhman S, Law E, Dawson J. Experience with pneumonia in acutely burned patients requiring ventilator support. Am Surg 2000;66:206–9. [24] Douglas MW, Mulholland K, Denyer VV, Gottlieb T. Multi-drug resistant Pseudomonas aeruginosa outbreak in a burns unit–an infection control study. Burns 2001;27:131–5.