Abstract: The study took placed during the summer time of 2007. A 20 years old university hospital with 600 beds equipped with central air conditioning. No special disinfection program was achieved for the hospital water supplies at the time of investigation. The hospital is supplied by city water and sewages organization and is treated with standard chlorination. To analysis the first nosocomial outbreak of Legionnaires` disease in a major university hospital of Iran. Seventy Broncho Alveolar Lavage specimens were obtained from patients with pneumonia. In addition 20 water samples of various hospital points were screened for the presence of Legionella species and free-living amoebae. Six nosocomial cases occurred over an 8 weeks period, between the first and last case detection. Legionella isolates from the patients matched the water sample isolates. L. pneumophila were grown up from only 3 out of 70 samples, while the bacteria mip gene were detected from additional three cases. L. pneumophila (serogrup 1) were isolated from two hospital sites. Since, Legionella positive patients had been admitted to the hospital at least 2 weeks prior to sampling, the cases could be assumed as hospital acquired Legionnaire`s disease, originated from hospital water supplies which should be treated for effective disinfection.
INTRODUCTION
Hospital-acquired Legionnairs` disease has been reported from many parts of the world since the first outbreak of Legionnares` disease in 1976 (Fields et al., 2002). Pneumonia caused by Legionella has a poor prognosis unless it is diagnosed early and treated with specific antibiotics. In the warm and cold water systems of large buildings such as hotels and hospitals, growth conditions are favorable for Legionellae. Among the bacteria responsible for atypical pneumonias, Legionellae are the main agents involved in severe forms of the disease (Brooks et al., 2001). Within members of the family Legionellaceae, L. pneumophila is involved in more than 95% of cases of severe atypical pneumonia (Balows and Duerden, 1998). Legionella may be widely spread through hospital water systems and may have been primarily isolated in potable hot water systems and cooling towers of hospitals and other institutions (Anaissie et al., 2002). These bacteria are present at the highest concentrations in biofilms within hot water systems and openings of water outlets (Meyer, 2005). Aerosol inhalation and ingestion or aspiration of potable water from cooling towers containing the organism has been more frequently associated with community outbreaks (Sala et al., 2007). However, this route has also been implicated in nosocomial cases (Kowska et al., 2003). Isolation of causative agent from bronchoalveolar lavage (BAL) fluid specimens is fastidious and time consuming. Although legionellosis occurs in persons with no underlying disease, the risk is significantly increased in people with illnesses such as chronic obstructive pulmonary disease, malignancy, immunisuppressive therapy, or organ transplantation (England and Fraser, 1981). Therefore, hospitals are points of congregation for persons at risk for infection, besides providing suitable conditions for amplification of Legionella. The goal of this study was to determine whether legionellosis was occurring in the hospital and if so, whether these were nosocomial infections. To answer this question, all respiratory tract secretions submitted to the clinical laboratory over an eight weeks period were tested for Legionella contamination. In addition, water samples from the warm and cold water system of the hospital building were cultured for Legionella. Isolates from patients were then compared with those from hospital water samples and typing.
MATERIALS AND METHODS
Setting
A twenty-year-old teaching hospital with 600-beds and three Intensive
Care Units (ICU) with 25 beds. Three cooling towers which located in open
area provide necessary water for hospital air condition system. There
is no especial program for water supplies disinfection at the time of
the outbreak.
Outbreak
Sample collection of suspected patients was taken placed during 8
weeks, between 7 July and 31 August 2007. Legionella pneumophila
genome was diagnosed from two patients during the first week of July,
followed by four other patients over seven week`s period. The samples
of positive patients were reconfirmed by PCR technique using mip gene
primers. After the first detection, 20 samples of hospital water supplies
from rooms and sections where patients stayed were tested and Legionella
pneumonia were isolated from two points. The risk factors for Legionella
infections such as alcohol consumption, smoking, chronic obstructive pulmonary
disease, aging, history of recent surgery and immunosuppression and also
chemical injuries were recorded for each case. The clinical characteristics
of the patients were shown in table. All cases were detected within the
impatient group who were admitted at least four weeks prior to BAL sampling.
All patients had a history of either using ventilator or taking shower
connected to the hospital water supplies.
BAL Sampling
A retrospective study of 70 patients suffering of pneumonia was conducted.
All patients were adults with pneumonia diagnosed by temperature of 38°C
or higher and pulmonary infiltrates on chest X-ray who admitted to hospital
at least 2 weeks prior to study. Patients were eligible for the study
if bronchoscopy was to be performed and if the patients agreed to it as
a part of the endoscopic procedure. Patients without respiratory infection
and pulmonary infiltrates were excluded from the study. Patients with
mentioned criteria were subjected to BAL sampling irrespective of empiric
therapy for infection pneumonia prior to study. The standard Broncho Alveolar
Lavage (BAL) fluid sampling procedures were achieved in hospital bronchoscopy
section, by using an Olympus BF10 or P10 fiber-optic bronchoscope. The
bronchoscope apparatus were washed twice with soap solution and clean
water in disinfected in Sidex or Korosolex for 20 min before next use.
The BAL samples were then transferred to laboratory immediately. Each
BAL sample was homogenized, divided into two portions of 2 mL each. One
portion of each sample kept frozen at -80°C until it was used for
DNA amplification and the other portion processed for standard culturing
and identification.
Water Sampling
Water sample were obtained from shower heads (hot and cold), outlets and
ventilator equipments of hospital sections and rooms where patients stayed.
All water samples were prepared from main supplies, according to methods previously
described by Doust et al. (2008b). The temperature
of cold water samples were between 8-15°C and of hot water samples between
40-55°C. The hospital was provided with potable water connected to city water
network. Each sample of 1 L water were collected from each point aseptically.
Samples were concentrated aseptically by membrane filtration, using a three-
piece PVC manifold (Millipore) and cellulose type HA membranes with a pore size
of 0.45 μm (Millipore Corporation) (Hajia et al.,
2004). The concentrated samples were removed by cutting the membranes into
pieces, placing them in sterile containers containing 10 mL of original sample
and shaking vigorously for 15 min until the membranes appeared to be clean.
Culturing Procedures
Standard protocols Buffered Charcoal Yeast Extract (BCYE) agar formulated
in laboratory and enriched with L-cystein and iron salts, polymaxin B, anisomycin
and vancomycin (PAV) were then added (Doust et al.,
2002). The samples then were subjected to heat treatment procedure which
is believed to be more effective, before inoculated onto BCYE plates. All inoculated
plates were incubated at 35-37°C for up to 14 days and examined daily for
evidence of growth. Colonies resembling Legionella were subcultured on
BCYE with and without L- cystein. Suspicious Legionella colonies that
failed to grow in the absence of L-cystein were typed by direct immuno-fluorescence
with a commercial monoclonal antibody test. A commercially available latex agglutination
test kit (Oxoid, UK) was used for the final confirmation of presumptive Legionella
colonies.
Molecular Detection
A thermic-lysis procedure was used to release the bacterial DNA extraction
(Patterson et al., 1997). Briefly, a total of
400 μL of Legionella suspension was boiled for 5 min in a 1.5 mL
μ centrifuge tube and chilled on ice for 5 min; this procedure was repeated
once. After 2 min of centrifugation at 1100 x g, 20 μL of the supernatant
was tested for specific DNA amplification. A set of primers and detection probe
were chosen from the mip gene of L. pneumophila which was described previously
by Doust et al. (2008b) and Julhac
et al. (1992). The mip gene, which codes for a virulence protein,
seemed to be a promising target for the diagnosis of legionellosis. Two 20-base
oligonuclotides (L.pm-1 = 5′-GGTGACTGCGGCTGTTATGG-3′ , L.pm-2= 5 ′-GGCCAATAGGTCCGCCAACG-3′
). A 25-base internal probe, (L.pm-3 =5′-CAGCAATGGCTGCAACCGATGCCAC-3 ′) located
at nucleotides 885 to 912 from the coding strand, was labeled with [γ-32P]
ATP as described previously and was used as detection probe.
The samples were submitted to 40 cycles of amplification in a 50 μL volume containing 1 U of Taq polymerase 0.1 m (each) primer, 0.2 mM (each) the four deoxynucleotides, 10 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl2 and 0.2 mg of gelatin per mL. After an initial denaturation step at 95°C for 5 min, each amplification cycle was performed at follows; 2 min for annealing at 62°C, 1.5 min for primer extension at 72°C and 1.5 min for denaturation at 94°C. The sample preparation, PCR amplification and electrophoresis were performed in three different rooms. Ten microliter of each sample containing amplified DNA was electrophorsed through a 1.5% agarose gel in 0.5x TEB buffer (44.5 mM Tris-HCl, 44.5 mM boric acid, 1.25 mM disodium EDTA [pH 8.3] ), after staining with ethidium bromide photographed under UV light. DNA extracted from L. pneumophila Philadelphia strain was used as positive control and double distilled water as the negative one. The restriction analysis was used to check product specificity according the method previously described by Doust et al. (2007). The kpn1 restriction enzyme was used for control of 630 bp PCR products specificity. Two bands approximately 400 and 230 bp were produced after 12 h incubation of PCR product with kpn1 restriction enzyme.
RESULTS
Using the primer set and the detection probe described here we were able to detect the Legionnaire`s disease bacterium in 6 (17%) who were admitted to the hospital during two to four weeks prior to BAL sampling. All 70 patients had pneumonia and were suspected having Legionnaires` disease. Over 36% of patients` natraemia were over 130 and 63% bellow 130 mM L-1. Male/Female rate of the patients under investigation were 64/36%. The most dominant clinical findings were observed as; 67% dry cough, 17% productive cough, 70% chills and fever and 16% abdominal pains. Clinical and laboratory backgrounds of patients with Legionnaires` disease bacterium within BAL specimens are shown in Table 1. The case No. 1 had bronchiectasis, breath shortness, pulmonary infections, case No. 2 complained of chest pain and breath shortness, case No. 3 had homeostasis, chest pain, breath shortness and cases No. 4-6 suffered from breath shortness on admission. All patients` age was between 40 and 65 years old, 66% male and 34% female. The blood natremia of the patients with proved Legionnaires ` disease was between 120 and 125 mM L-1 (Table 1). Sub clinical findings of the patients are detailed in Table 2. All patients were treated with erythromycin protocol. The sensitivity of this assay has been evaluated for L. pneumophila serogroup 1-14, L. micdadei and L. bozemani serogroup 1. Legionella pneumophila of serogroups 1 were isolated from three cases. The amplification product of 600 bp was observed in all 6 patients (Fig. 1). The pattern of restriction enzyme showed specificity of PCR products (Fig. 2). Amplification of Legionella specific sequences by using primers from the mip gene sequences permitted the detection of DNA all L. pneumophila strains of serogroups 1-14. Non L. pneumophila serogroup 1 or L. micdadei and L. bozemani might be included in PCR positive and culture negative cases. Since, L. micdadei and L. bozemani serogroup 1 had been also detected by the primer and probe system. DNA from gram positive and gram negative species tested and also from human leukocyte were not amplified by this procedure. Using the primers from the mip gene sequence, also reported the detection of DNA from L. pneumophila serogroup 1-14 strains but not those from some other Legionella species (Julhac et al., 1992). L. micdadei, a non L. pneumophila species is known to be responsible for Legionellosis and use of DNA amplification to detect this species in biological fluids is therefore a great interest. Considering the data, we decided to use the primers constructed of the mip gene and reported here, to assess the feasibility of Legionella detection in BAL fluid specimens.
Table 1: | Brief description of patients with Legionnaires` disease |
*Mycobacterium tuberculosis was also detected along with Legionella. Fev: Fever, Chi: Chills, Npc: Non productive coughs, Pc: Productive coughs, Wl: Weight losses, Ap: Abdominal pains, Pol: Polyarthritis, W: Weakness |
Table 2: | Para-clinical findings of patients |
Fig. 1: | Agarose gel of PCR products from patients BAL specimens. No. 1-6 represent BAL fluids from six patients, No. 7; non L. pneumophila, No. 8; Control negative, No. 9; L. pneumophila 1, No. 10; DNA ladder |
Fig. 2: | Specificity of PCR products of patient`s sample by restriction enzyme kpn1 |
DISCUSSION
This study reports the first nosocomial outbreak of Legionnaire`s disease within a university hospital of Iran. This is the first outbreak of Legionnaires` disease reported from Iran. The clinical aspects of the patients and also some of the risk factors (age the history of Sulphur Mustard exposure history) are unique in this outbreak. Ozerol et al. (2005) also reported an outbreak of Legionnaires` from Turkey. Legionella contamination have been reported within clinical and environment samples of Iran and also from some nearby countries (Aslani et al., 1997; Doust et al., 2008b). We detected Legionnaires` disease (ld) bacterium within 9% of nosocomial pneumonia cases during study period by standard culture method and PCR (Julhac et al., 1998) . No ld case was detected in second group also with pneumonia, which were selected randomly from the outpatient cases. The largest outbreak of nosocomial Legionnaires ` disease occurred in Los Angeles with at least 218 confirmed cases from 1977-1982. Since then, more than 300 reports of nosocomial Legionnaires` disease have appeared in public health reports (Fields et al., 2002). Present report complete the Legionnaires` disease related events across the world especially middle-east region, were usually are visited by considerable number of foreign tourists. The relation of at least 39 species and more than 60 serogroup of Legionella have been described within human infections. Legionnaires` disease, a sever form of nosocomial and community-acquired pneumonia, is cause mainly by L. pneumophila. Other species, mainly L. micdadei account for approximately 15% of cases of Legionella pneumonia and are more often reported in cases of Pontiac fever. L. pneumophila serogroups 1 is responsible for 80% of reported human cases of Legionella infections and is usually the only species isolated during routine testing (Brundrett, 1992).
Water samples of the mentioned hospital were collected after the first case of Legionnaires` disease and L. pneumophila was diagnosed (Doust et al., 2008b). The multi-points of hospital water samples were checked for Legionella and samples of two points proved to be contaminated with L. pneumophila by standard culture and PCR methods. Interestingly, predominant samples of the hospital water supplies were contaminated with different species of free-living amoeba which coexist with Legionella (Doust et al., 2008a) During present investigation we examined water supplies of the 20 year-old hospital yielded results indicating some samples demonstrated the presence of Legionellae with evidence of exposure in 3-6 individuals out of 70 tested. We managed to reduce the Legionella colonization of hospital sites by flushing water outlets for 30 min and superheating water at 65°C. To prevent contamination in future, hyperchlorization and ultraviolet radiation systems were introduced into the hospital infection committee. Hospital water system is frequently colonized with Legionella species, once present in such systems, they are able to survive and multiply, particularly when hot water temperatures are kept low to minimize the risk of scalding for patients (Qasem et al., 2008). The presence of different species of free-living amoebae within the hospital water distribution system, indicated the risk of Legionellae release in the water source of hospital despite of routine chlorination, this data are supported by report of Singh and Coogan (2005) who detected Legionella-laden amoebae in dental unit waterlines. The hospital water system was connected to the municipal water networks. An association between Legionella infection and exposure to contaminated water has been described previously by Garcia-Nunez et al. (2007) and Spalekova and Bazovska, (2003). Legionnaires` disease is thought to be acquired by inhalation of aerosols containing Legionella or by microaspiration of contaminated water that are rarely reported. Nosocomial Legionellosis is more frequent in immunocompromized patients, the patients with previous disease and transplant recipients. One or more Legionnaires` disease infection risks were noticed in majority of our nosocomial pneumonia cases even without Legionnaire`s disease bacterium. Interestingly enough, the only new risk factor for one patient, was war chemical agent inhalation, during last 15 years with present respiratory injuries.
ACKNOWLEDGMENTS
This study was partially supported by joined research Grant of Baqiyatollah University of Medical Sciences (Research Center of Molecular Biology) and Tarbiat Modares University.