SMJ 2001: 46(6); 167-170
Microbial threats; a perspective
The World Health Organisation estimate that over 13 million deaths a year occur as a result of an infectious disease.1 Many of these infections could be preventable or controllable if appropriate resources were available. Three infections; malaria, tuberculosis and HIV cause illness in more than 300 million and deaths in five million annually. The considerable economic and social consequences of infection are borne by the world’s poorest nations in Africa and Asia.
The developed nations however have significant but scaled down infection related problems; In Scotland the seasonal influenza outbreaks threaten many thousands of vulnerable, elderly patients but vaccination uptake by the at-risk population remains poor. E coli 0157 outbreaks in rural areas may occur because of hand hygiene failure after visits to the countryside (up to one quarter of cattle herds in Scotland carry this organism2). Methicillin-resistant Staphylococcus aureus (MRSA) blood stream infection and other hospital acquired infections which affect one in ten hospital inpatients could be controlled by simple alcohol based hand disinfection and prudent antimicrobial use.3 Every year, travellers to malaria endemic regions fail to protect themselves from Plasmodium falciparum malaria and hospital admissions and deaths result. Individuals and institutions therefore accept unnecessary risks with infection on a daily basis.
The deliberate release of Anthrax in the USA poses a risk that was not previously widely recognised and one which is beyond any individual or institutions control. The fear that has been engendered throughout the developed world is not least because other potentially aerosolised agents eg small pox,4 primary pneumonic plague,5 modified influenza or viral haemmorhagic fever agents are now threats which cannot be easily predicted or guarded against. It is essential that clinicians familiarise themselves with some of these previously unusual or unlikely agents.6 Herein we discuss Bacillus anthracis as an agent of biological warfare.
Anthrax
Anthrax
has long been recognised as a potential weapon7
and there are at least 17 countries with biological weapons programs.8
The largest recorded outbreak resulting from aerosoled release of manufactured anthrax spores occurred 1979 in the former Soviet Union. Seventy-nine cases and at least 66 deaths occurred four kilometers downwind of a military microbiology laboratory in Sverdlovsk. It was not until 1992 that the Soviet Union admitted that the laboratory was an offensive biological warfare laboratory and that the outbreak was due to the accidental release of anthrax spores.9 Autopsy findings have been published and these with animal studies provide most of the current knowledge regarding pathology of inhalational anthrax.
In Scotland during World War II bombs containing anthrax droplets were dropped on the uninhabited island of Gruinard. The island remained contaminated for 45 years.10 The low visibility, high potency, dormancy and hardiness of anthrax spores (resistant to UV light, drying, heat radiation and many disinfectants) allow them to be used successfully as weapons. It has been estimated that the aircraft release of 50kg of anthrax over a city of five million would result in 250,000 casualties, 100,000 deaths and would cost $26.2 billion per 100,000 people exposed.11 There is now concern that genetically modified antibiotic or vaccine resistant strains can be produced12 and with the advances in genomics this may pose an even more serious threat.13 In 1999 a working group published a consensus statement on anthrax as a biological weapon.14
Pathogenesis
Bacillus
anthracis is a gram positive, aerobic spore-forming zoonosis infecting mainly
herbivores. Infection in animals occurs by ingestion of soil-borne spores.
Distribution is worldwide. Humans are generally resistant to infection and
despite significant exposure in some industries there are relatively few cases.
Spores enter human hosts through skin abrasions, ingestion or inhalation. Human
to human transmission has not been documented. Infectious doses have not been
established for man but it probably does not take many spores to initiate
cutaneous infection. For inhalational anthrax the dose required to kill 50% of
exposed is estimated at 8000-10000 spores.9
Anthrax spores are phagocytosed by macrophages and then produce the vegetative forms, which multiply within the lymphatics and are released into the bloodstream.15 There are two recognised virulence factors associated with B. anthracis – the poly-D-glutamyl capsule and the toxin complex. The capsule inhibits phagocytosis of vegetative anthrax bacilli. The toxin complex consists of three synergistically acting proteins:
1. Protective Antigen (PA)– the binding protein which
allows entry of the toxin into the host cell
2. Oedema factor (OF) – a calmodulin dependent
adenylate cyclase
3. Lethal factor (LF) – a zinc and calcium dependent
metalloprotease
PA
binds to receptors on the host cell surface and also binds to OF or LF forming
Oedema Toxin (OT) and Lethal Toxin (LT) respectively.16
OT impairs neutrophil function and increases intracellular levels of cyclic-AMP, altering water homeostasis and leading to massive oedema.17 LT inactivates host mitogen activated protein kinase, disrupting the signalling pathways responsible for cell growth and maturation.18 The major target for LT is macrophages and the initial response is the release of tumour necrosis factor a, nitric oxide (NO) and interleukin-1 cytokines. It is this release that mediates septic shock in systemic anthrax.
The other major target for LT is the endothelial cell lining of the capillary network, and necrosis of lymphatic and blood vessel walls is responsible for systemic release of bacilli and haemorrhage.19 The virulence factors are encoded on two plasmids pXO1 and pXO2 and their expression is regulated by a transcriptional activator – AtxA.20 AtxA activity is dependent on host specific factors such as temperature and carbon dioxide concentration. Temperatures over 37 degrees and carbon dioxide concentrations over 5% favour expression of the virulence factors.21 Attenuated strains of anthrax for vaccine production have been made by producing bacteria lacking one of the plasmids (both are required for virulence).22
A peptide that binds to PA and prevents its interaction with OF and LF has already been described 23 and may be a potential future adjunct therapy.
Clinical features of anthrax
Cutaneous anthrax
typically occurs two to seven days after exposure but may be considerably
longer. In the recent USA outbreak there have been 11 cases to date (seven
confirmed) and the time from exposure to clinical presentation was one to 12
days. After entry of the infecting bacillus through a skin lesion, a small
papule appears which gradually becomes surrounded by vesicles and oedema over
the next four days. The painless lesion ulcerates by day seven to form the
characteristic eschar, which takes up to six weeks to resolve. A small
proportion of cases develop systemic anthrax if untreated and regional
lymphadenopathy is not uncommon. The lesions are most often on the head, neck or
limbs.
Mortality is usually low if antibiotic therapy is given, although untreated mortality may be 20%.
Gastro-intestinal anthrax has to date not been noted in any patients during the USA outbreak. It usually occurs about one week after the ingestion of contaminated meat. Intestinal symptoms begin with fever and abdominal pain and progress to constipation or diarrhoea, malaena, frank rectal bleeding or haematemesis. Ascites may develop on day three to four. Most cases recover in 10-14 days. Differential diagnosis includes food poisoning, gastroenteritis and any cause of acute abdomen. Alternatively patients may present with oropharyngeal anthrax. Symptoms are of pharyngitis, fever, dysphagia and regional lymphadenopathy in the neck. Mortality can be up to 50% even with treatment. Differential diagnosis includes streptococcal pharyngitis, Vincent’s angina, Ludwig’s angina and parapharyngeal abscess. Both gastro-intestinal forms may progress to toxaemia, septic shock and death.
Inhalational anthrax usually
follows an incubation period of about 10 days but may occur weeks following the
inhalation of spores. In the Sverdlovsk outbreak two cases were diagnosed six
weeks after the accidental release.9 In the USA the time from exposure
to presentation was four to six days.24
A mild initial ‘flu-like illness with headache, fever, myalgia and
fatigue is typical followed on day four of the illness with nausea and vomiting,
sepsis syndrome, cough, substernal chest pain, dyspnoea and cyanosis.
Rhinorrhoea has been noted to be uncommon (only one of the USA inhalational
anthrax cases to date). Profound sweating was common and abdominal pain has been
present in three of the cases. Many patients progress rapidly to coma and death
during the hyper-acute phase. Mediastinal and peribronchial lymphadenopathy
occur with haemorrhage and necrosis within lymph nodes, causing haemorrhagic
mediastinitis.25
Chest radiograph characteristically shows widening of the mediastinum.
In the USA outbreak all 10 patients had abnormal chest radiographs; seven had pulmonary infiltrates, eight had pleural effusions (two further cases later developed effusions) and seven had mediastinal widening (including paratracheal fullness and hilar fullness.24 Typically these radiograph changes were present 48 hours after the onset of symptoms.
An
eleventh fatal case has recently been described in Connecticut and this patient
developed a pleural infiltrate and effusion in the absence of mediastinal
widening.26 Pleural effusions accumulate rapidly, requiring drainage and
are heavily blood stained. Peripheral white cell counts have typically risen
rapidly following hospitalisation in the hyper-acute phase. Mortality in
inhalational anthrax in the USA prior to recent events has been reported as very
high (85%)24 but more recent experience has suggested
mortality of about 40% with full supportive care and dual or triple antibiotic
combinations.24 Meningitis is a potential complication of disseminated
anthrax and is usually haemmorhagic and rapidly fatal. In the Sverdlovsk
outbreak up to 50% of autopsies revealed haemmorhagic meningitis.25 Fatal anthrax meningitis has been observed in
one of the 11 inhalational cases in the USA to date.24
Diagnosis
In the context of a winter epidemic of respiratory tract infections inhalational anthrax may be difficult to spot. The recent USA cases have been differentiated from viral respiratory tract infections by lack of rhinorrhoea and frequency of gastro-intestinal symptoms.24 Cutaneous anthrax has a wide differential diagnosis including insect or spider bite, orf, syphilitic chancre, staphylococcal or streptococcal skin sepsis, echthyma gangrenosum, plague, cutaneous tuberculosis, cutaneous leishmaniasis and leprosy.
The differential diagnosis of inhalational anthrax is wide. It includes bacterial pneumonia or mediastinitis, atypical pneumonia, viral pneumonia, sarcoidosis, tularaemia and pneumonic plague.
Anthrax is usually diagnosed by the isolation of B. anthracis from any clinically suspicious site although expression of fluid from eschars is not recommended because of risk of dissemination. Swabs from an ulcer base, vesicle fluid or punch biopsies are recommended.
Anthrax is distinguishable from other members of the B. cereus group (the toxin of which is a common cause of self limiting food poisoning) by its grey, nonhaemolytic, tenacious colonies which, on inoculation into carbon dioxide containing nutrient agar, will form a characteristic capsule. Normal sheep’s blood- agar in use in all UK microbiology laboratories will pick up colonies of anthrax under normal conditions. In the USA outbreak all eight patients with inhalational anthrax who had not had prior antibiotics have all yielded positive blood cultures within 24 hours.24, 26 In addition B. anthracis has been isolated from pleural fluid and cerebrospinal fluid.
Non-culture methods have been useful in the USA outbreak and have established the diagnosis in three patients without positive cultures. B. anthracis capsule and cell wall antigen specific immunohistochemical staining of pleural fluid and transbronchial biopsy specimens were positive in all cases where this was attempted.24 A Polymerase Chain Reaction (PCR) method using primers to the Protective Antigen gene and to an enzyme, which mediates capsule formation has previously been used in epidemiological investigations27 but may be used as a rapid diagnostic test in the acute illness.27 In the USA outbreak it has given rapid results in patients with inhalational and cutaneous anthrax when applied to pleural fluid and skin biopsies within three hours.24 Specific enzyme linked Immunosorbent Assay (ELISA) will detect a fourfold rise in antibody titres to the capsule but is not useful in the diagnosis of acute disease. In the US outbreak immunoglobulin (Ig) G to B. anthracis protective antigen has been employed by the Centre for Disease Control. A delayed type hypersensitivity reaction to Anthaxin may be useful in retrospective diagnoses. Ninety-nine percent of cases have been shown to have a positive skin test at four weeks.29
Management of anthrax in deliberate
release setting
Penicillin
is the treatment of choice for naturally occurring cutaneous anthrax when the
organism’s sensitivity is known (resistance is very rare). In addition high
dose intravenous benzyl penicillin (eg 2.4g
4-hourly for 14 days) should always be used if there is a suspicion of
meningeal involvement as it reliably penetrates the inflamed blood brain
barrier. Rifampicin should also be considered in such circumstances.
Since the only large autopsy series demonstrated meningeal involvement in
up to 50% of fatal cases25 it would seem prudent to the
authors to include penicillin (+/- rifampicin) in all inhalational anthrax
regimes. This however is not a recommendation in the latest CDC guidelines where
penicillin is only one of a choice of second additional agents (http://www.bt.cdc.gov).
Although penicillin is clearly very important it is not recommended as a single first line agent therapy for a number of reasons;
(1) There is concern over the production of inducible
beta lactamases which could inactivate penicillin
(2) The large number of bacilli in disseminated anthrax
may form a critical mass which leads to “switching off “
of penicillin binding protein production (similar to the
so called “Eagle effect” which is well described in deep
seated Streptococcus pyogenes infections30)
(3) Anecdotal experience from the recent anthrax cases
in the USA has suggests benefit in dual or triple
antibiotic regimes.24
Anthrax is intrinsically resistant to third generation cephalosporins due to the production of cephalosporinases so these agents must not be used.
The CDC recommendation for all cases of bioterrorism related anthrax is ciprofloxacin because it is suspected that a multi-resistant quasi species of B anthracis may have been engineered. The dose in uncomplicated cutaneous anthrax is 500mg 12 hourly for 60 days. In severe forms treatment should be via the intravenous route (eg 400mg 12-hourly) with oral switch when the clinical condition improves. An alternative to ciprofloxacin is doxycycline (100mg 12-hourly). To date all of the B anthracis isolates in the USA have been sensitive to ciprofloxacin, levofloxacin, doxycycline, penicillin, clindamycin, macrolides, aminoglycosides, rifampicin, vancomycin, chloramphenicol, ampicillin, imipenem and meropenem.
Although not substantiated by any experimental data there may be rational for using clindamycin as a second or third agent in systemic anthrax because of its possible useful activity during the static growth phase of anthrax when penicillin-binding protein may be produced. Clindamycin also may have inhibitory effects on toxin production as is seen in vitro with the exotoxin of Streptococcus pyogenes31.
Current recommendations are that children and pregnant women should also be treated with ciprofloxacin since the risks from infection outweigh risks of potential adverse effects. Antibiotics in cutaneous anthrax do not alter progression of the skin lesion but do prevent systemic involvement and render the lesion sterile in 24 hours. If the organism is shown to be penicillin-sensitive, treatment in cutaneous anthrax can be changed to oral amoxycillin.
When exposure to anthrax spores are suspected on epidemiological grounds then post exposure prophylaxis should be given. Oral ciprofloxacin 500mg 12-hourly or doxycycline (100mg 12-hourly) for 60 days after exposure will cover the longest interval associated with germination of the spores. In addition to antibiotic therapy, post exposure immunisation may be given; five doses of vaccine at 0,3,6 weeks, six months and one year after exposure.32 If vaccination is given, post exposure antibiotic therapy can probably be reduced to four weeks.
There is no evidence of person-to-person spread in anthrax and therefore there is no indication to treat or vaccinate contacts of cases (unless they share the same exposure).
Corticosteroids may be of benefit if there is severe oedema, particularly in the head and neck or in meningitis. Early supportive treatment for septic shock in the form of intensive care involvement, inotropic support and ventilation is essential. In cases of meningitis early institution of anti-cerebral oedema measures such as intravenous mannitol is recommended. In the USA repeated pleural drainage has been required in many of the inhalational cases.
Although anti-anthrax serum is no longer used for treatment, plasmaphoresis and administration of specific human gamma globulin from vaccinated individuals may be life saving in the critically ill.
The US Food and Drug Administration (FDA) is using animal efficacy data to “ fast track” newer vaccines since current vaccines are not highly purified and require several doses to confer protection. The vaccine currently available in the UK is an alum precipitated cell-free culture filtrate of strain 34F2 (three doses at intervals of three weeks then six months, then annually). The vaccine is 92.5% effective33 although 35% of recipients experience minor local reactions and 0.5% experience systemic reactions. Usually only people with high occupational risk are vaccinated. In the USA this has included the military since 1998.
The high mortality of inhalational and gastro-intestinal anthrax despite antibiotic therapy, and the potential emergence of antibiotic or vaccine resistant strains have prompted research into new approaches to treatment, in particular anti-toxin treatments.34-36 The sequencing of the B. anthracis genome is currently underway and may offer other ways of tackling anthrax.
The future
Future possible biological releases could be more transmissible, less recognisable and even more difficult to manage. Future episodes may utilise more familiar organisms as has previously been seen in the USA when a religious cult deliberately contaminated 10 restaurants with Salmonella typhimurium causing infection in 751 people.37 It is likely that future infection outbreaks, even with endemic pathogens, will need to be scrutinised by public health clinicians for the possibility of a bio-terrorist connection. One should also consider the possibility of biological attacks against livestock (agro-terrorism) as these could lead to potentially destabilising political and economic effects. Such effects have been well demonstrated in the recent UK outbreak of foot and mouth disease, which is caused by a highly contagious picornavirus. The source of this outbreak remains unknown but bioterrorism has not yet been ruled out.
At the time of writing the anthrax outbreak in the USA has claimed five lives, a tiny number compared to traditional heavy weight infections. In contrast concerns over antimicrobial misuse and growing microbial resistance are heightened as rumours of indiscriminate use of fluoroquinolones outside normal indications for prophylaxis are circulated. It is likely this will result in a significant impact on antimicrobial resistance patterns in urinary tract and faecal isolates.38 We are also concerned over the potential inappropriate use of ciprofloxacin in other soft tissue infections caused by Staphylococci or pulmonary infections caused by Pneumococci where such therapy may result in treatment failure. The other public health consequence of “preparedness programs” in the USA is the diversion of resources from primary prevention of war and bioterrorism, and further reduction in funding for existing public health programs.39 To date, the economic and political impact of the anthrax attacks and the potential consequences of antibiotic misuse have far outweighed the clinical consequences of anthrax. None-the-less it is essential that clinicians are well informed about the clinical aspects of potential bioterrorism agents and the procedures recommended in the event of a bioterrorism attack.40,41
Updating guidelines
Regularly
updated guidelines for the management of an outbreak of anthrax due to
deliberate release of spores including environmental decontamination, protection
of workers, dealing with packages suspected of containing anthrax, sampling and
transportation procedures are described on the Public Health Laboratory Services
website (http://www.phls.co.uk/facts/deliberate releases.htm and the
Centres for Disease Control and Prevention (CDC) website, at http://www.bt.cdc.gov.
N Bodasing, RA Seaton
Department of Infectious Diseases and Tropical Medicine,
Brownlee Centre,
Gartnavel General Hospital,
1053 Great Western Road,
Glasgow G12 OYN