Synonym: Pontiac fever
Legionnaires' disease (LD) is a severe, potentially fatal acute pneumonia acquired by droplet inhalation of water contaminated by bacteria of the genus Legionella. It was named in 1976 following a serious outbreak of pulmonary infection at a convention of the American Legion in Philadelphia, when 29 out of 182 affected individuals died. The Legion was meeting to commemorate the American 200th anniversary of independence. The causative organism was Legionella pneumophila, a bacterium which is only found in aquatic systems where it exists symbiotically with some amoebae.
Pontiac fever was named after an outbreak in Pontiac, Michigan in 1968 which was also caused by legionella organisms but which did not affect the lungs.
Most cases of LD are caused by L. pneumophila; however, there are many other species of the organism (most of which are also found in soil) and mild illnesses including Pontiac fever may be caused by these organisms. Legionella spp. are aerobic, Gram-negative bacilli. Serogroup 1 is the most common cause of disease but 18 serogroups are known.
Other than L. pneumophila, 34 species of legionella have been identified, mostly from immunocompromised patients with pneumonia.
L. pneumophila is found in natural water supplies and soil. It is also found in many recirculating and water supply systems, where it breeds rapidly in favourable conditions (optimal conditions being a temperature of 35°C, range 25-45, stagnant water and water containing debris which can act as foci for bacterial breeding). The bacteria are distributed widely in the environment and can breed in natural and artificial water sources. However, they only become a risk to health when conditions allow their rapid multiplication.
The bacterium can survive low temperatures and multiplies in water temperatures between 25°C and 45°C. A temperature of 60° is considered bactericidal, although the bacteria will eventually return when the temperature is lowered.
The risk of infection has prompted guidance on the precautions required to minimise this risk. This guidance can be found in the Approved Code of Practice (ACOP) for Legionnaires' disease on the Health and Safety Executive website. Guidance was updated in November 2013. It includes keeping water moving and clean, and maintaining temperatures below 20°C or above 60°C.
LD is an uncommon form of pneumonia. The disease has no particular clinical features that clearly distinguish it from other types of pneumonia. Laboratory investigations must therefore be carried out in order to obtain a diagnosis. Epidemiological definitions are given in the table below.
The Public Health England (PHE) data are derived from laboratory-confirmed reports and follow-up questionnaires that collect information on whether the case is community-acquired, hospital-acquired or travel-associated. Between 1980 and 2011 the Health Protection Agency received reports of, on average, 240 cases a year, although with a peak of over 500 cases in 2006. There were 239 cases in 2011. From these figures it can be derived that:
- LD is 2 to 5 times more common in men than in women.
- The death rate from diagnosed infections is 10-15% (representing about 10% in previously well individuals and up to 80% in the elderly).
- 40-50% of all cases are related to travel, usually abroad, where air-conditioning systems and stored water may be less well regulated.
- These numbers are probably an underestimate because of failure to diagnose, and hence report, less severe cases (Centers for Disease Control and Prevention (CDC) estimate that <10% of cases are reported).
- Older people, smokers and those with reduced immunity are more likely to be infected and are more likely to die.
- Case numbers are highest between June and September.
- Most cases are isolated - outbreaks are less common.
L. pneumophilia lives naturally in water. It causes disease when water droplets containing sufficient numbers of the bacterium enter the atmosphere as fine particles or in an aerosol that enables them to be inhaled deep into the lungs. There they invade lung epithelial cells and macrophages and replicate intracellularly. Cooling towers are often a source of infection, as they produce a very fine aerosol. Spa baths, misting machines and showers have also been identified as sources of infection in the past. So too have air-conditioning systems, especially if poorly maintained. The disease is not spread from person to person. It is found throughout the world.
50% of Pontiac fever outbreaks have been linked to whirlpool spas or hot tubs.
These are all factors that impair immunity and include:
- Malignancy (particularly lung cancer and haematological malignancy).
- End-stage chronic kidney disease.
- Chronic obstructive pulmonary disease (COPD).
- Advanced age.
- Alcohol abuse.
- Recent surgery.
Two distinct clinical syndromes can be caused by different species, 'pneumonic' (LD) and 'non-pneumonic' (primarily Pontiac fever).
Pneumonic - the more severe, also called Legionnaires' disease (LD)
- The incubation period is usually 2-10 days (average 6-7 days).
- There is a prodrome for a day or two of mild headache and myalgia, followed by high fever, chills, rigors and a cough.
- 90% present with cough. This is usually dry at first but may progress to being productive of green or bloodstained sputum.
- Up to a third of patients have dyspnoea, pleuritic chest pain and haemoptysis.
- Gastrointestinal symptoms are present in a third and include nausea, vomiting, diarrhoea and anorexia.
- Headache is common. Confusion, impaired cognition and ataxia may also occur.
- Patients often display a relative bradycardia despite the presence of a fever.
- Symptoms outside the chest predominate in the early stages but pulmonary symptoms predominate later.
Non-pneumonic (Pontiac disease)
- This has a shorter incubation period (1-2 days) than the pneumonic form.
- Influenza-like illness with myalgia, fever, and headache is almost invariable.
- Only 50% have a cough.
- The disease is self-limiting.
- This similarity to other viral illnesses means many cases are overlooked and the disease is certainly relatively undiagnosed.
- Pyrexia and tachypnoea are common but for the level of temperature, there is often a relative bradycardia. This means that the pulse rate is not as high as would be expected for that level of temperature and not that it is slower than usual.
- Patients classically look unwell.
- There are no abnormalities in the upper respiratory tract but examination of the chest may show rales, rhonchi or consolidation.
- Rarer features include pericarditis, hepatomegaly and impaired mental state.
British Thoracic Society (BTS) guidelinesrecommend investigations for legionella infection for all patients with severe community-acquired pneumonia, for other patients with specific risk factors and for all patients with community-acquired pneumonia during outbreaks.
For the purpose of confirmation of a case, the PHE requires one of the following:
- The gold standard: isolation and culture of legionella species from clinical specimens, usually sputum. Cases of Pontiac fever are almost universally culture-negative.
- Seroconversion, meaning at least a four-fold increase in titre of indirect immunofluorescent antibody test (IFAT) incorporating a monovalent L. pneumophila serogroup 1 antigen by a validated technique.
- Confirmation of L. pneumophila urinary antigen, using validated reagents or kits. Culture of sputum and serology have been the standard tests for many years but the development of urine testing gives faster results, improves the rate of correct diagnosis and because the result is swifter, allows earlier correct treatment resulting in a lower mortality.
The HPA considers a case presumptive where there is either one of:
- A clinical diagnosis of pneumonia with a single high titre of 128, using IFAT, or a single titre of 64 in an outbreak.
- A positive direct fluorescence (DFA) on a clinical specimen, using validated monoclonal antibodies (also referred to as a positive result by direct immunofluoresence).
There are no commercial tests to confirm specific non-pneumophilic infections, and serological conversion rates are variable. The shift towards urine testing is likely to limit detection of non-pneumophilic cases further unless better tests for the full range of legionella species are developed.
PHE considers cases to be hospital-acquired if patients spent the entire 10 days prior to symptom onset in hospital. Cases are considered probably hospital-acquired if a part of this time was spent in hospital AND there are either other cases of legionella in the same establishment or matching organisms cultured from the water system.
Other investigations include:
- FBC shows leukocytosis and may show disseminated intravascular coagulation (DIC).
- U&E and creatinine, as there may be dehydration. Infection can occasionally result in inappropriate secretion of ADH, which produces hyponatraemia.
- Alkaline phosphatase may be low: along with abnormal LFTs this is an uncommon feature of pneumonia that may suggest the diagnosis of LD.
- Blood gases sre used to assess lung function and clinical status in pneumonia.
- CXR is not specific but may show patchy or bi-basal consolidation: up to half have pleural effusion which may take one to four months to resolve.
- With alteration in mental state, lumbar puncture and head CT scan may be required to exclude other causes. They are both normal in LD.
- Other causes of pneumonia, especially mycoplasma or viral.
- Left ventricular failure.
- Severe acute respiratory syndrome (SARS).
LD must be in the differential diagnosis for any community-acquired pneumonia. Even with a high degree of watchfulness it is thought that up to 90% of cases are missed.
Useful investigations to diagnose atypical pneumonias
|Cause of pneumonia:||Mycoplasma pneumoniae||Legionella pneumophila||Chlamydophila pneumoniae|
|Blood tests||There may be raised WCC or (rarely) evidence of haemolytic anaemia. ESR may be elevated. Serology titres and complement fixation tests/ELISA can help to confirm the diagnosis.||FBC may show left shift. Severe cases may have DIC evident on FBC/INR. Hyponatraemia may occur due to syndrome of inappropriate ADH secretion. Urea/creatinine can be raised if complicated by acute kidney failure or dehydration. LFTs often nonspecifically deranged. CK may be elevated in rhabdomyolysis. Serological tests on blood or urine may be used to confirm diagnosis.||Usually nonspecific and unhelpful. Serology titres or polymerase chain reaction tests may be used to confirm the diagnosis.|
|CXR||Usually single lower-lobe bronchopneumonia pattern with lobar consolidation rare. Other possible patterns include atelectasis, nodular infiltration akin to tuberculosis/sarcoidosis, hilar adenopathy and (rarely) pleural effusion.||50% have pleural effusion. Patchy alveolar infiltrates may be seen. CXR can take up to four months to return to normal and may initially progress despite therapy.||Usually lower-lobe single subsegmental infiltrate. Pleural effusion found in up to a quarter of cases. Can progress to acute respiratory distress syndrome. CXR changes may take up to three months to resolve.|
Pain and temperature may require drugs such as paracetamol, and oxygen or even assisted ventilation may be indicated if there is hypoxia.
The antibiotic of choice is erythromycin although gastrointestinal upset from the disease and the tendency of this drug to do the same may make it less attractive. Azithromycin, clarithromycin, levofloxacin, ciprofloxacin and doxycycline are excellent alternatives. Tetracycline, ciprofloxacin, and other fluoroquinolone and macrolide drugs may be indicated. The course should be 10-21 days and is usually intravenous, at least at first. In critically ill patients the newer macrolides and fluoquinones should be used as first-line therapy.
Most patients with LD are admitted to hospital, as intensive treatment is usually needed.
Respiratory insufficiency, dehydration, multiple organ failure, sepsis, and acute kidney injury.
About 10-15% die overall, representing about 10% of those who are treated and not immunocompromised but rising to 80% in those who are untreated and immunocompromised. The death rate in nosocomial infection is around 50% but figures on death rates are very much complicated by other pathology. Early detection and treatment improve outcome
Travel-associated LD is linked to stays in holiday accommodation where intermittent use of facilities, variable temperatures and seasonal fluctuations of water flow may increase the risk of bacterial multiplication and thus infection. Prevention needs to be a collaborative effort between health professionals, governments and the tourist industry and enormous progress has been made since the European Working Group for Legionella Infections (EWGLI) began in 1986 to coordinate such activities in Europe - as a result, a legionellosis surveillance network now connects 36 countries that report travel-associated cases to a computerised database in London.
Travellers in the highest-risk groups must seek medical attention as soon as possible if they suspect they might have LD. The increasing number of whirlpool spas and hot tubs in hotels and cruise ships, and the difficulty in tracking outbreaks of Pontiac fever, emphasise the importance of adequate filtration and disinfection measures, coupled with strict application of bather capacity. If doctors are suspicious of chest infections acquired abroad then erythromycin or doxycycline are preferable to amoxicillin.
Legionella spp. are ubiquitous in water supplies and the role of hospital water supplies in the spread of nosocomial infection is unclear. Regular maintenance and testing of water systems, even in the absence of known infection, are generally recommended, especially in places like transplant units where patients are very vulnerable. Public opinion about hospital-acquired infection is also creating pressure for a proactive approach.
Good surveillance also aids prevention. Detection of outbreaks and detection of clusters are important and secondary preventative measures can then be undertaken. A high level of awareness on the part of medical staff increases the chances of detection of individual or related cases
Further reading and references
Guidelines for the management of community acquired pneumonia in children; British Thoracic Society (2011), Thorax Vol 66 Sup 2
Chest infections - adult; NICE CKS, November 2015 (UK access only)
Legionnaires' disease: guidance, data and analysis; Public Health England
Legionnaires' disease. The control of legionella bacteria in water systems; Approved Code of Practice and Guidance, Health and Safety Executive
Guidelines for the management of community acquired pneumonia in adults; British Thoracic Society (2009), Thorax Vol 64 Sup III
Den Boer JW, Yzerman EP; Diagnosis of Legionella infection in Legionnaires' disease. Eur J Clin Microbiol Infect Dis. 2004 Dec23(12):871-8.
File TM Jr, Garau J, Blasi F, et al; Guidelines for empiric antimicrobial prescribing in community-acquired pneumonia. Chest. 2004 May125(5):1888-901.
Mills GD, Oehley MR, Arrol B; Effectiveness of beta lactam antibiotics compared with antibiotics active against atypical pathogens in non-severe community acquired pneumonia: meta-analysis. BMJ. 2005 Feb 26330(7489):456. Epub 2005 Jan 31.
Amsden GW; Treatment of Legionnaires' disease. Drugs. 200565(5):605-14.
Roig J, Rello J; Legionnaires' disease: a rational approach to therapy. J Antimicrob Chemother. 2003 May51(5):1119-29. Epub 2003 Mar 28.
Lettinga KD, Verbon A, Weverling GJ, et al; Legionnaires' disease at a Dutch flower show: prognostic factors and impact of therapy. Emerg Infect Dis. 2002 Dec8(12):1448-54.
European Guidelines for Control and Prevention of Travel Associated Legionnaires' Disease; European Surveillance Scheme for Travel Associated Legionnaires' Disease and the European Working Group for Legionella Infections, 2005 (archived content)
O'Neill E, Humphreys H; Surveillance of hospital water and primary prevention of nosocomial legionellosis: what is the evidence? J Hosp Infect. 2005 Apr59(4):273-9.
Stout JE, Yu VL; Hospital-acquired Legionnaires' disease: new developments. Curr Opin Infect Dis. 2003 Aug16(4):337-41.
Den Boer JW, Verhoef L, Bencini MA, et al; Outbreak detection and secondary prevention of Legionnaires' disease: a national approach. Int J Hyg Environ Health. 2007 Jan210(1):1-7. Epub 2006 Sep 7.