No evidence of Legionella infection in general practice patients presenting with acute respiratory infections in The Netherlands.

The role of Legionella spp. in the aetiology of acute respiratory infections (ARIs) is largely unknown. In this case-control study, conducted in a general practitioner setting during 2000 and 2001, nose and throat samples from patients presenting with ARIs (n = 230) and controls (n = 200) were analysed for the presence of Legionella spp. by real-time PCR. Legionella DNA was not detected in any of the cases or controls. Thus, Legionella spp. do not seem to play a role in patients presenting with ARIs, nor were they present in patients who visited their general practitioner for complaints other than ARIs.

Legionella spp. are an important cause of community‐acquired and nosocomial pneumonia. Since the first description of Legionella pneumophila, more than 40 species of Legionella have been identified, approximately half of which have been isolated from patients [1]. Infection with Legionella spp. can present as a severe pneumonia, with or without multisystem disease and high mortality, but can also present as a self‐limiting influenza‐like illness. Many individuals who seroconvert to Legionella are entirely asymptomatic [1, 2].

Acute respiratory infections (ARIs) are a major cause of morbidity and mortality worldwide. Various infectious agents, especially viruses, have been associated with clinical syndromes ranging from mild disease, such as the common cold, to more severe conditions, such as pneumonia. However, no aetiological agent is found in a large percentage of cases [3]. Although ARIs are very common, there is only one report on the prevalence of Legionella spp. as a cause of community‐acquired infections in general practice [4]. This may be because the diagnosis is not considered, existing diagnostic tests are insensitive, or legionellosis is distributed unevenly across the world. In The Netherlands, c. 200 cases of severe legionellosis are reported annually to the health authorities [5].

The aim of this study was to assess the frequency of Legionella spp. in nose and throat samples of patients presenting with ARIs in general practice. In The Netherlands, more than 100 general practitioners (GPs) participate in nationwide surveillance of respiratory pathogens and influenza‐like illnesses, coordinated by the Netherlands Institute for Health Services Research (NIVEL, Utrecht, The Netherlands) [6]. The GPs register all patients with influenza‐like illness and other ARIs. ARIs without influenza‐like illness are defined as respiratory infections with acute onset (< 5 days) and coryza, sore throat and/or cough. Twenty GPs participated in the present study and were requested to collect a nose and throat sample from patients with ARIs without influenza‐like illness (case). A corresponding control was also sampled within 7 days, defined as an individual who consulted the GP for a complaint other than ARIs, who was in the same age category as the case (0–4, 5–14, 15–24, 25–44, 45–64 or > 65 years), was not a member of the same household as the case, and had not received antibiotics or antiviral therapy in the preceding 14 days. A nose and a throat swab were placed together in 4 mL of Hanks' balanced salt solution containing gelatin, lactalbumin, yeast and antibiotics (GLY medium), and sent to the central laboratory by regular mail. Swabs for bacterial culture were also taken.

The primers and probes of the Legionella PCR assay were based on the 16S rRNA gene as described previously [7]. In brief, nucleic acid isolation was performed with a proteinase K lysis protocol, and primers Leg1 (5′‐TACCTACCCTTGACATACAGTG‐3′) and Leg2 (5′‐CTTCCTCCGGTTTGTCAC‐3′) were used to obtain a 200‐bp amplicon. Real‐time detection of the fragment was done with a Legionella genus‐specific fluorescent probe conjugated to a minor groove binder (LSPP‐VIC: 5′‐GGTTGCGTCGTTACG‐3′). An L. pneumophila‐specific fluorescent minor groove binder probe was used on the complementary DNA strand (LPN‐FAM: 5′‐GAGTCCCCACCATCACATG‐3′). Inhibition was monitored in a separate assay by amplifying a phocine herpes virus, and negative controls were included to monitor processing and cross‐contamination. Sensitivity controls comprised ten‐fold dilutions of L. pneumophila DNA (1000 fg to 10 fg), while 1000 fg of Legionella bozemanii DNA served as a control for discrimination between L. pneumophila and other Legionella spp.

Real‐time PCR was performed with an ABI Prism 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA, USA).

Nose and throat samples from 230 cases and 200 controls were collected between October 2000 and November 2001. The median age of the cases was 35 years (range 0–87 years), and 104 (45%) of them were male. The median age of the control patients was 36 years (range 0–86 years), and 76 (38%) of them were male. Evidence of at least one infectious agent was found by standard techniques in 146 (63%) of the cases, compared to 44 (22%) of the controls (Table 1). Real‐time PCR failed to detect Legionella DNA in any of the samples from either the patients with ARI or the control group. Inhibition of the PCR occurred with one sample.

Table 1

Frequency of detection of specific infectious agents in 230 cases of acute respiratory tract infection and 200 control patients

Pathogen	Cases n (%)	Controls n (%)
Influenza virus A	25 (10)	2 (1)
Influenza virus B	7 (3)	0 (0)
Parainfluenza virus 1	3 (1)	0 (0)
Parainfluenza virus 2	0 (0)	0 (0)
Parainfluenza virus 3	2 (1)	0 (0)
Rhinovirus	61 (25)	23 (11)
Adenovirus	2 (1)	0 (0)
Coronavirus	20 (8)	11 (5)
Respiratory syncytial virus	6 (3)	2 (1)
Human metapneumovirus	5 (2)	0 (0)
Enterovirus	7 (3)	3 (1)
Chlamydia pneumoniae	2 (1)	0 (0)
Mycoplasma pneumoniae	3 (1)	1 (0)
β‐Haemolytic streptococci	20 (9)	4 (2)
One or more of the above	146 (63)	44 (22)
Negative	84 (37)	156 (78)
Total	230 (100)	200 (100)

ARIs are the most common diseases seen in the community and a frequent reason to consult a GP. Knowledge of the causes of respiratory infections continues to evolve, with recognition of new pathogens [8], but no aetiological diagnosis is established in a large percentage of cases [3]. It might be expected that Legionella spp. could play a role in ARIs, as they are ubiquitous in the aquatic environment and artificial water systems, and can cause a variety of diseases, ranging from asymptomatic cases to severe life‐threatening pneumonia [1, 9]. Upper respiratory signs and symptoms, such as a dry cough and sore throat, have been reported in cases of Legionnaires' disease [10], and a previous study [4] found serological evidence of infection with Legionella spp. in 11.2% of patients who presented with a febrile respiratory tract infection. However, observations based solely on antibody seroconversion should be viewed with scepticism, as the sensitivity and specificity of serology for Legionella spp. other than L. pneumophila have not been validated, and cross‐reactions with other organisms may occur [1, 11, 12].

Culture on buffered charcoal yeast extract (BCYE) plates is the standard method for the laboratory diagnosis of Legionella infections, but has its limitations. Isolates are identified by a combination of colony and Gram's stain morphology, and reactions with specific antibodies. However, because of the fastidious nature of the organism and problems inherent in specimen collection, culture of clinical specimens may have a sensitivity as low as 10%[11]. Moreover, several non‐pneumophila Legionella spp. tend to grow poorly on BCYE agar [13]. Nucleic acid amplification techniques are attractive tools for detection of Legionella spp. in clinical samples, as they are able to detect all legionellae and provide results rapidly. Several studies have found that PCR methods have a higher rate of detection than culture‐based methods [14, 15, 16]. In the present study, the assay sensitivity was 10 fg for all Legionella spp., equivalent to two bacterial cells/reaction.

Throat swabs have been used successfully to detect respiratory pathogens such as Mycoplasma pneumoniae[17], Chlamydia pneumonia[17, 18] and Legionella spp. [16, 17] by PCR. In contrast to a previous report based on serological observations [4], Legionella spp. were not detected in the nose or throat of patients with ARIs in general practice, and no evidence was found for asymptomatic carriage of Legionella spp. in the upper respiratory tract.

Acknowledgements

Phocine herpes virus was kindly provided by H. Niesters, Erasmus Medical Center, Rotterdam, The Netherlands.