anthrax as a biological weapon

7/27/2019 Anthrax as a Biological Weapon

http://slidepdf.com/reader/full/anthrax-as-a-biological-weapon 1/12

CONSENSUS STATEMENT

Anthrax as a Biological WeaponMedical and Public Health ManagementThomas V. Inglesby, MD

Donald A. Henderson, MD, MPH

John G. Bartlett, MD

Michael S. Ascher, MD

Edward Eitzen, MD, MPH

Arthur M. Friedlander, MD

Jerome Hauer, MPHJoseph McDade, PhD

Michael T. Osterholm, PhD, MPH

Tara O’Toole, MD, MPH

Gerald Parker, PhD, DVM

Trish M. Perl, MD, MSc

Philip K. Russell, MD

Kevin Tonat, PhD

for the Working Groupon Civilian Biodefense

OF THE NUMEROUS BIOLOGI-cal agents that may beusedas weapons, the WorkingGroup on Civilian Biode-

fense has identified a limited numberof organisms that could cause diseaseand deaths in sufficient numbers tocripple a city or region. Anthrax is oneof the most serious of these diseases.

High hopes were once vested in theBiological Weapons and Toxins Con-vention, whichprohibited offensive bio-logical weapons research or production

andwassignedby most countries.How-ever, Iraq and the former Soviet Union,bothsignatoriesof the convention, havesubsequently acknowledged having of-fensive biowarfare programs; a numberof other countries are believed to havesuch programs, as have some autono-mous terrorist groups. Thepossibility of a terrorist attack using bioweaponswould be especially difficult to predict,

detect, or prevent, and thus, it is amongthe most feared terrorist scenarios.1

Biological agents have seldom beendispersed in aerosol form, the expo-sure mode most likely to inflict wide-spread disease. Therefore,historical ex-perience provides little informationabout the potential impact of a biologi-cal attack or the possible efficacy of postattack measures such as vaccina-tion, antibiotic therapy, or quarantine.

Policies and strategies must therefore

rely on interpretation and extrapola-tion from an incomplete knowledgebase. The Working Group on CivilianBiodefense reviewed the available litera-ture and expertise and developed con-sensus recommendations for medicaland public health measures to be takenfollowing such an attack.

CONSENSUS METHODS

The working group comprised 21 rep-

resentativesfrom academic medical cen-

Author Affiliations: The Center for Civilian Biode-fense Studies (Drs Inglesby, Henderson, Bartlett,O’Toole, Perl, and Russell), and the Schools of Medi-cine(Drs Inglesby, Bartlett,and Perl) andPublicHealth(Drs Henderson, O’Toole, and Russell), Johns Hop-kins University, Baltimore, Md; Viral and Rickettsial Dis-eases, California Department of Health, Berkeley (Dr Ascher); US Army Medical Research Institute of In-fectious Diseases, Frederick, Md (Drs Eitzen, Fried-lander, and Parker); Office of Emergency Manage-ment, NewYork, NY (MrHauer); Centers forDisease

Control and Prevention, Atlanta, Ga (Dr McDade);Acute Disease Epidemiology, Minnesota Depart-ment of Health, Minneapolis (DrOsterholm);and theOffice of Emergency Preparedness, Departmentof Health and Human Services, Rockville, Md (Dr Tonat).Corresponding Author and Reprints: Thomas V.Inglesby, MD, Johns Hopkins Center for Civilian Bio-defense Studies, Johns Hopkins University, Candler Bldg, Suite 850, 111Market Pl,Baltimore, MD 21202(e-mail: [email protected]).

Objective To develop consensus-based recommendations for measures to be takenby medical and public health professionals following the use of anthrax as a biologicalweapon against a civilian population.

Participants The working group included 21 representatives from staff of major aca-demic medical centers and research, government, military, public health, and emer-gency management institutions and agencies.

Evidence MEDLINE databases were searched from January 1966 to April 1998, us-ing the Medical Subject Headings anthrax, Bacillus anthracis, biological weapon, bio-

logical terrorism, biological warfare, and biowarfare. Review of references identifiedby this search led to identification of relevant references published prior to 1966. Inaddition, participants identified other unpublished references and sources.

Consensus Process The first draft of the consensus statement was a synthesis of in-formation obtained in the formal evidence-gathering process. Members of the workinggroup provided formal written comments which were incorporated into the second draftof the statement. The working group reviewed the second draft on June 12, 1998. Nosignificant disagreements existed and comments were incorporated into a third draft.The fourth and final statement incorporates all relevant evidence obtained by the litera-ture search in conjunction with final consensus recommendations supported by all work-ing group members.

Conclusions Specific consensus recommendations are made regarding the diagno-sis of anthrax, indications for vaccination, therapy for those exposed, postexposureprophylaxis, decontamination of the environment, and additional research needs.

JAMA. 1999;281:1735-1745 www.jama.com

©1999 American Medical Association. All rights reserved. JAMA, May 12, 1999—Vol 281, No. 18 1735

wnloaded From: http://jama.jamanetwork.com/ by a Universidade Presbiteriana Mackenzie User on 10/15/2013



ters and research, government, mili-tary, public health, and emergencymanagement institutions and agencies.

MEDLINE databases were searchedfrom January 1966to April 1998for the

Medical Subject Headings anthrax ,Bacillus anthracis, biological weapon,biological terrorism, biological warfare, and biowarfare. Review of refer-ences led to identification of addi-tional relevant references publishedprior to 1966. In addition, experts inthe working group identified unpub-lished references and sources.

The first draft of the working group’sconsensus statement was the result of synthesis of informationobtainedin theformal evidence-gathering process.Members of the working group were

asked to make formal written com-ments on this first draft of the docu-ment in May 1998. Suggested revi-sions wereincorporatedintothe seconddraft of the statement. The workinggroup was convened to review the sec-ond draft of the statement on June 12,1998, at the Johns Hopkins Center forCivilian Biodefense Studies, Balti-more, Md. Consensus recommenda-tions were made; no significant dis-agreements existed at the conclusion of this meeting. The third draft incorpo-

rated changes suggested at the confer-ence and working group members hadan additional opportunity to review thedraft and suggest final revisions. The fi-nal statement incorporates all rel-evant evidence obtained by the litera-ture search in conjunction with finalconsensus recommendations sup-ported by all working group mem-bers. Funding for the development of the working group consensus state-ment was primarily provided by eachrepresentative’s institution or agency.

The Office of Emergency Prepared-ness, Department of Health and Hu-man Services (DHHS), provided travelfunds for 4 members of the group.

The assessment and recommenda-tions providedherein representthe bestprofessional judgment of the workinggroup based on data and expertise cur-rently available. The conclusions andrecommendations need to be regu-

larly reassessed as new information be-comes available.

HISTORY OF CURRENT THREAT

For centuries, anthrax has caused dis-

ease in animals and, uncommonly, se-rious illnessin humans throughout theworld.2 Research on anthrax as a bio-logical weapon began more than 80years ago.3 Today, at least 17 nationsare believed to have offensive biologi-cal weapons programs4; it is uncertainhow many are working with anthrax.Iraq has acknowledged producing andweaponizing anthrax.5

Most experts concur that the manu-facture of a lethal anthrax aerosol is be-yond the capacity of individuals orgroups without access to advanced bio-

technology. However, autonomousgroups with substantial funding andcontacts may be able to acquire the re-quired materials for a successful at-tack. One terrorist group, Aum Shin-rikyo,responsible for the releaseof sarinin a Tokyo, Japan, subway station in1995,6 dispersed aerosols of anthrax andbotulism throughout Tokyoon at least8 occasions. For unclear reasons, theattacks failed to produce illness.7

The accidental aerosolizedrelease of anthrax spores from a military micro-

biology facility in Sverdlovsk in theformer Soviet Union in 1979 resultedin at least 79 cases of anthrax infec-tion and 68 deaths and demonstratedthe lethal potential of anthrax aero-sols.8 An anthrax aerosol would beodorless and invisible following re-lease and would have the potential totravel many kilometers before dissemi-nating.9,10 Evidence suggests that fol-lowing an outdoor aerosol release, per-sons indoors could be exposed to asimilar threat as those outdoors.11

In 1970, a World Health Organiza-tion (WHO) expert committee esti-mated that casualties following thetheoretical aircraft release of 50 kg of anthrax over a developed urban popu-lation of 5 million would be 250 000,100 000 of whom would be expected todie without treatment.9 A 1993 reportby theUS CongressionalOffice of Tech-nology Assessment estimated that be-

tween 130 000 and 3 million deathscould follow the aerosolized release of 100 kg of anthrax spores upwind of the Wa sh in gt on , DC , ar ea —l et ha li tymatching or exceeding that of a hydro-

gen bomb.

12

An economic model de-velopedby theCenters for Disease Con-trol and Prevention (CDC) suggesteda cost of $26.2 billion per 100 000 per-sons exposed.13

EPIDEMIOLOGY

Naturally occurring anthrax is a dis-ease acquired following contactwith an-thrax-infected animals or anthrax-contaminated animal products. Thedisease most commonly occurs in her-bivores, which are infected by ingest-ing spores from the soil. Large an-

thrax epizooticsin herbivoreshavebeenreported; during a 1945 outbreak inIran, 1 million sheep died.14 Animalvac-cination programs have reduced dras-tically the animal mortality from thedisease.15 However, anthrax spores con-tinue to be documented in soil samplesfrom throughout the world.16-18

In humans, 3 types of anthrax infec-tionoccur: inhalational, cutaneous, andgastrointestinal.Naturally occurringin-halational anthrax is now a rare causeof human disease. Historically, wool

sorters at industrial mills were at high-est risk. Only 18 cases were reportedin theUnited States from 1900 to 1978,with the majority occurring in special-risk groups, including goat hair mill orgoatskin workers and wool or tanneryworkers. Twoof the18 cases were labo-ratory associated.19

Cutaneous anthrax is the most com-mon naturally occurring form, with anestimated 2000 cases reported annu-ally.18 Disease typically follows expo-sure to anthrax-infected animals. In the

United States, 224 cases of cutaneousanthrax were reported between 1944and 1994.20 The largest reported epi-demic occurred in Zimbabwe between1979 and1985, when more than 10 000human cases of anthrax were re-ported, nearly all of them cutaneous.21

Gastrointestinal anthrax is uncom-monly reported.18,22,23 However, gastro-intestinal outbreaks have beenreported

MANAGEMENT OF ANTHRAX USED AS A BIOLOGICAL WEAPON

1736 JAMA, May 12, 1999—Vol 281, No. 18 ©1999 American Medical Association. All rights reserved.




in Africa andAsia.24 Gastrointestinal an-thrax follows ingestion of insufficientlycooked contaminated meat and in-cludes 2 distinct syndromes, oral-pharyngealandabdominal.22,24-27 In1982,

there were 24 cases of oral-pharyngealanthrax ina rural northernThailandout-break following the consumptionof con-taminated buffalomeat.24 In 1987, therewere 14 cases of gastrointestinal an-thraxreportedin northernThailandwithbothoral-pharyngeal andabdominal dis-ease occurring.25

No case of inhalational anthrax hasbeen reported in theUnited Statessince1978,19,20 making even a single case acause for alarm today. As was demon-strated at Sverdlovsk in 1979, inhala-tional anthrax is expected to account

for most morbidity and essentially allmortality following the use of anthraxas an aerosolized biological weapon.8,28

In the setting of an anthrax outbreakresulting from an aerosolized release,cutaneous anthrax would be less com-mon than inhalational anthrax, easierto recognize, simpler to treat, and as-sociated with a much lower mortality.In the Sverdlovsk experience, therewere no deaths in patients developingcutaneous anthrax.8 There is little in-formation available about the risks of

direct contamination of food or waterwith anthrax spores. Although humaninfections have been reported, experi-mental efforts to infect primates by di-rect gastrointestinal instillation of an-thrax sporeshave notbeen successful.29

MICROBIOLOGY

Bacillus anthracis derives from theGreek word for coal, anthrakis, be-cause the disease causes black, coal-like skin lesions. Bacillus anthracis isan aerobic, gram-positive, spore-

forming, nonmotile Bacillus species.The nonflagellated vegetative cell islarge (1-8 µm in length, 1-1.5 µm inbreadth). Spore size is approximately1 µm. Spores grow readily on all ordi-nary laboratory media at 37°C, with a“jointed bamboo-rod” cellular appear-ance and a unique “curled-hair” colo-nial appearance, and display no hemo-lysis on sheep agar (FIGURE 1). This

cellular and colonial morphology theo-retically should make its identifica-tion by an experienced microbiologiststraightforward, although few practic-ing microbiologists outside the veteri-

nary community have seen anthraxcolonies other than in textbooks.30

Anthrax spores germinate when theyenter an environment rich in amino ac-ids, nucleosides, and glucose, such asthat found in the blood or tissues of ananimal or human host.The rapidly mul-tiplying vegetative anthrax bacilli,on thecontrary, will only form spores after lo-cal nutrients are exhausted, such aswhen anthrax-infected body fluids areexposed to ambient air.16,17 Full viru-lence requires the presence of both anantiphagocytic capsule and3 toxin com-

ponents (ie, protective antigen, lethalfactor, and edema factor).30 Vegetativebacteria have poor survival outsideof ananimal or humanhost; colony countsde-cline to undetectable within 24 hoursfollowing inoculation into water.17 Thiscontrasts with the environmentallyhardy properties of theB anthracisspore,which can survive for decades.30

PATHOGENESIS ANDCLINICAL MANIFESTATIONS

Inhalational Anthrax

Inhalational anthrax follows deposi-tion of spore-bearing particles of 1 to5 µm into alveolar spaces.31,32 Macro-phages ingestthespores, some of whichundergo lysis and destruction. Surviv-ing spores are transportedvia lymphat-ics to mediastinal lymph nodes, wheregermination may occur up to 60 dayslater.28,29,33 The process responsible forthe delayed transformation of spores tovegetativecells is poorly understood butwell documented. In Sverdlovsk, casesoccurred from 2 to 43 days after expo-

sure.8

In experimental monkeys, fataldisease occurred up to 58 days28 and 98days34 after exposure. Viablesporeshavebeen demonstrated in the mediastinallymph nodes of monkeys 100 days af-ter exposure.35

Oncegermination occurs,disease fol-lows rapidly. Replicating bacteria re-lease toxins leading to hemorrhage,edema, and necrosis.23,36 In experimen-

tal animals, once toxin production hasreached critical threshold, death oc-curs even if sterility of the blood-stream is achieved with antibiotics.19

Based on primate data, it has been es-timated that for humans the LD 50 (le-thal dose sufficient to kill 50% of per-sons exposed to it) is 2500 to 55 000inhaled anthrax spores.37

The term inhalational anthrax re-flectsthenature of acquisition of the dis-ease. Theterm anthrax pneumonia is mis-leading. Typical bronchopneumoniadoes not occur. Postmortem pathologi-cal study of patients who died becauseof inhalational anthrax in Sverdlovskshowed hemorrhagic thoracic lym-

phadenitis and hemorrhagic mediasti-nitis in all patients. In up to half of thepatients, hemorrhagic meningitis alsowas seen. No patients who underwentautopsy had evidence of a bronchoal-veolar pneumonic process, although 11of 42 patients undergoing autopsy hadevidence of a focal, hemorrhagic, nec-rotizingpneumonic lesion analogous tothe Ghon complex associated with tu-berculosis.38 These findings are consis-tent with other human case series andexperimentally inducedinhalational an-

thrax in animals.33,39,40

Early diagnosis of inhalational an-thrax would be difficult and would re-quire a high index of suspicion. Clini-cal information is available from onlysome of the 18 cases reported in theUnited States in this century and fromthe limited available information fromSverdlovsk. Theclinical presentation hasbeen described as a 2-stage illness. Pa-

Figure 1. Gram Stain of Bacillus anthracis

Gram-positive anthrax bacilli in a peripheral bloodsmear from a rhesus monkey that died of inhala-tional anthrax. Reprintedwith permissionfrom Zajtchukand Bellamy.23






tients first developeda spectrum of non-specific symptoms,including fever, dys-pnea, cough, headache, vomiting, chills,weakness, abdominal pain, and chestpain.8,19 Signs of illness and laboratory

studies were nonspecific. This stage of illness lasted from hours to a few days.In some patients, a brief period of ap-parent recovery followed. Other pa-tients progressed directly to the sec-ond, fulminant stage of illness.2,19,41

Thissecond stagedevelopedabruptly,with sudden fever, dyspnea, diaphore-sis, and shock. Massive lymphadenopa-thy and expansion of the mediastinumled to stridor in some cases.42,43 A chestradiograph most often showed a wid-ened mediastinum consistent withlymphadenopathy (FIGURE 2).42 Up to

half of patients developed hemorrhagicmeningitis withconcomitant meningis-mus, delirium,andobtundation.In thissecondstageof illness, cyanosis andhy-potension progress rapidly; death some-

times occurs within hours.2,19,41

The mortality rate of occupationallyacquired cases in the United States is89%, but the majority of cases occurredbefore the development of critical careunits and, in some cases, before the ad-vent of antibiotics.19 At Sverdlovsk, it isreported that 68 of the 79 patients withinhalational anthrax died, although thereliability of the diagnosis in the survi-vors is questionable.8 Patients who hadonset of disease 30 or more days after re-lease of organisms had a higher re-ported survivalratecompared withthosewith earlierdiseaseonset.Antibiotics, an-tianthrax globulin,andvaccine were usedtotreatsomeresidentsintheaffectedareasome time after exposure, but which pa-tients received these interventions andwhen isnot known. In fatal cases, the in-terval between onset of symptoms anddeath averaged 3 days. This is similar tothe disease course and case fatality ratein untreated experimental monkeys,which have developed rapidly fataldisease even after a latency as long as58 days.28

Modern mortality rates in the set-ting of contemporary medical and sup-portive therapy might be lower thanthose reported historically. However,the 1979 Sverdlovsk experience is notinstructive. Although antibiotics, anti-

anthrax globulin, corticosteroids, andmechanical ventilation were used, in-dividual clinical records have not beenmade public.8 It is also uncertain if theB anthracis strain to which patientswere

exposed was susceptible to the pre-dominant antibiotics that were usedduring the outbreak.

Physiological sequelae of severe an-thrax infection in animal models havebeen described as hypocalcemia, pro-found hypoglycemia, hyperkalemia, de-pressionand paralysisof respiratory cen-ter, hypotension, anoxia, respiratoryalkalosis, and terminal acidosis.44,45

Those animal studies suggest that in ad-dition to therapid administration of an-tibiotics, survival might improve withvigilant correction of electrolyte distur-

bances and acid-base imbalance, glu-cose infusion,andearly mechanical ven-tilationand vasopressor administration.

Cutaneous Anthrax

Cutaneous anthrax occurs followingthedeposition of the organism into skinwith previous cuts or abrasions espe-cially susceptible to infection.21,46 Ar-eas of exposed skin, such as arms,hands, face, and neck, are the most fre-quently affected. There are no data tosuggest the possibility of a prolonged

latency period in cutaneousanthrax. InSverdlovsk, cutaneous cases occurredonly as late as 12 days after the origi-nal aerosol release.8 After thespore ger-minates in skin tissues, toxin produc-tion results in local edema (FIGURE 3).An initially pruritic macule or papuleenlarges into a round ulcer by the sec-ond day. Subsequently, 1- to 3-mmvesicles may appear, which dischargeclear or serosanguinous fluid contain-ingnumerous organismson Gram stain.As shown in Figure 3, development of

a painless, depressed, black eschar fol-lows, often associated withextensive lo-cal edema. The eschar dries, loosens,and falls off in the next 1 to 2 weeks,most often leaving no permanent scar.Lymphangitis and painful lymphad-enopathy can occur withassociated sys-temic symptoms. Although antibiotictherapy does not appear to change thecourse of eschar formation and heal-

Figure 2. Chest Radiograph of a PatientWith Inhalational Anthrax

Chest radiograph of a 51-year-old laborer with oc-cupationalexposureto airborne anthrax spores takenon day 2 of illness. Lobulated mediastinal widening(arrowheads) is present, consistent with lymphad-enopathy, with a small parenchymal infiltrate at theleft lungbase. Reprinted with permission from Zajtchukand Bellamy.23

Figure 3. Cutaneous Anthrax

Left, Forearm lesion on day 7 of illness shows vesiculation and ulceration of the initial macular or papular an-thrax skin lesion. Right, Eschar of the neck on day 15 of illness is typical of the last stage of the lesion beforeit resolves over 1 to 2 weeks.Reprinted with permission from Binford CH,Connor DH,eds. Pathology of Tropi-cal and Extraordinary Diseases. Vol 1. Washington, DC: Armed Forces Institute of Pathology; 1976:119. AFIPnegative 71-1290-2.






ing, it does decrease the likelihood of systemic disease. Without antibiotictherapy, the mortality rate has been re-ported to be as high as 20%; with an-tibiotics, death due to cutaneous an-

thrax is rare.

2

Gastrointestinal Anthrax

Gastrointestinal anthrax occurs follow-ing deposition and subsequent germi-nation of spores in the upper or lowergastrointestinal tract. The former re-sults in the oral-pharyngeal form of dis-ease.24-26 An oral or esophageal ulcerleads to development of regional lym-phadenopathy, edema, and sepsis.24-26

The latter results in primary intestinallesions occurring predominantly in theterminal ileum or cecum,38 presenting

initially with nausea,vomiting,andmal-aise and progressing rapidly to bloodydiarrhea, acute abdomen, or sepsis.22

Massive ascites has occurred in somecases of gastrointestinal anthrax.27 Ad-vanced infection may appear similar tothe sepsis syndrome occurring in ei-ther inhalational or cutaneous an-thrax.2 Some authors suggest that ag-gressive medical intervention such aswould be recommended for inhala-tional anthrax may reduce mortality, al-though, given the difficulty of early di-

agnosis, mortality almost inevitablywould be high.2,22

DIAGNOSIS

Given therarity of anthrax infection andthepossibility that early cases are a har-binger of a larger epidemic, thefirst sus-picion of an anthrax illness must leadto immediate notification of the localor state health department, local hos-pital epidemiologist, and local or statehealth laboratory. By this mechanism,definitive tests can be arranged rap-

idly through a referencelaboratory and,as necessary, the US Army Medical Re-search Institute of Infectious Diseases(USAMRIID), Fort Detrick, Md.

Thefirst evidence of a clandestine re-lease of anthrax as a biological weaponmost likely will be patients seekingmedical treatment for symptoms of in-halationalanthrax. The sudden appear-ance of a large number of patients in a

city or region with an acute-onset flu-like illness andcase fatality rates of 80%or more, with nearly half of all deathsoccurring within 24 to 48 hours, ishighly likely to be anthrax or pneu-monic plague (TABLE 1). Currently,there areno effective atmospheric warn-ing systems to detect an aerosol cloudof anthrax spores.47

Rapid diagnostic tests for diagnosinganthrax, such as enzyme-linked immu-nosorbent assay for protective antigenandpolymerase chain reaction,areavail-ableonlyat national reference laborato-ries. Giventhelimitedavailability ofthesetests and the time required to dispatchspecimens and perform assays,rapid di-agnostic testing would be primarily forconfirmation of diagnosisand determin-ing in vitro susceptibility to antibiotics.Inaddition, these tests willbe usedintheinvestigation and management of an-

thrax hoaxes, such as the series occur-ring in late 1998.48 They would also beof value should suspicious materials inthe possession of a terrorist be identi-fied as possibly containing anthrax.

If only small numbers of cases pre-sent contemporaneously, the clinicalsimilarity of early inhalational anthraxto other acute respiratory tract infec-tions maydelay initialdiagnosis for somedays. However, diagnosis of anthraxcould soon becomeapparentthroughtheastute recognition of an unusual radio-

logical finding, identification in the mi-crobiology laboratory, or recognition of specific pathologic findings. A widenedmediastinum on chest radiograph (Fig-ure 2) in a previously healthy patientwith evidence of overwhelming flulikeillness is essentially pathognomonic of advanced inhalational anthrax andshould prompt immediateaction.23,42 Al-thoughtreatment at this stage wouldbe

unlikely to alter the outcome of illnessin the patient concerned, it might leadto earlier diagnosis in others.

Microbiologic studies can also dem-onstrate B anthracis and may be themeans for initial detection of an out-break. The bacterial burden may be sogreat in advanced infection that bacilliare visible on Gram stain of unspunpe-

ripheral blood, as has been demon-strated in primate studies (Figure 1). While this is a remarkable finding thatwould permit an astute clinician or mi-crobiologist to make the diagnosis, thewidespread use of automated cell-counter technology in diagnostic labo-ratories makes this unlikely.41

The most useful microbiologic test isthe standard blood culture, whichshould show growth in 6 to 24 hours.If the laboratory has been alerted to thepossibility of anthrax, biochemical test-

ing and review of colonial morphologyshould provide a preliminary diagnosis12 to 24 hours later. Definitive diagno-sis would require an additional 1 to 2days of testing in all but a few nationalreference laboratories. It should benoted,however,that ifthelaboratory hasnot been alerted to the possibility of an-thrax, B anthracis may not be correctlyidentified. Routine laboratory proce-durescustomarily identify a Bacillus spe-cies from a blood culture approxi-mately 24 hours after growth, but most

laboratories do not further identify Ba-cillus species unless specifically re-quested to do so. In the United States,the isolationof Bacillus species most of-ten represents growth of Bacillus ce-reus. The laboratory and clinician mustdetermine whether its isolation repre-sents specimen contamination.49 Therehave been no B anthracis bloodstreaminfections reported for more than 20

Table 1. Diagnosis of Inhalational Anthrax Infection

Epidemiology Sudden appearance of multiple cases of severe flulike i llnesswith fulminant course and high mortality

Diagnostic studies Chest radiograph: widened mediastinum

Peripheral blood smear: gram-positive bacilli on unspun smear

Microbiology Blood culture growth of large gram-positive bacil li with preliminaryidentification of Bacillus species

Pathology Hemorrhagic mediastinitis, hemorrhagic thoracic lymphadenitis,hemorrhagic meningitis






years. However, given the possibility of anthrax being used as a weapon and theimportance of early diagnosis, it wouldbe prudent for laboratory procedures tobe modified so that B anthracis is ex-

cluded after identification of a Bacillusspecies bacteremia.Sputum culture and Gram stain are

unlikely to be diagnostic, given the lackof a pneumonic process.30 If cutane-ous anthrax is suspected, a Gram stainand culture of vesicular fluid will con-firm the diagnosis.

A diagnosis of inhalational anthraxalso might occur at postmortemexami-nation following a rapid, unexplainedterminal illness. Thoracic hemor-rhagic necrotizing lymphadenitis andhemorrhagic necrotizing mediastini-

tis in a previously healthy adult areessentially pathognomonic of inhala-tional anthrax.38,43 Hemorrhagic men-ingitis should also raise strong suspi-cion ofanthraxinfection.23,38,43,50 Despitepathognomonic features of anthrax ongrosspostmortem examination, the rar-ity of anthrax makes it unlikely that apathologist would immediately recog-nize these findings. If thecase were notdiagnosed at gross examination, addi-tional days would likely pass beforemicroscopic slides would be available

to suggest the disease etiology.

VACCINATION

The US anthrax vaccine, an inacti-vated cell-free product, was licensed in1970 and is produced by Bioport Corp,Lansing, Mich (formerly called theMichigan Biologic Products Insti-tute). Thevaccineis licensed to be givenin a 6-dose series andhas recentlybeenmandated for all US military active- andreserve-duty personnel.51 The vaccineis made from the cell-free filtrate of a

nonencapsulated attenuated strain of B anthracis.52 The principal antigen re-sponsible for inducing immunity is theprotective antigen.18,23 A similar vac-cine hasbeen shown in 1 small placebo-controlled human trial to be effica-cious against cutaneous anthrax.53 Asof March 1, 1999, approximately590000 doses of anthrax vaccine havebeen administered to US Armed Forces

(Gary Strawder, Department of De-fense, Falls Church, Va, oral commu-nication, April 1999); no serious ad-verse events have been causally related(Miles Braun, Food and Drug Admin-

istration, Rockville, Md, written com-munication, April 1999). In a study of experimental monkeys, inoculationwith this vaccine at 0 and 2 weeks wascompletely protective against an aero-sol challenge at 8 and38 weeksand 88%effective at 100 weeks.54

A human live attenuated vaccine isproduced and used in countries of theformer Soviet Union.55 In the Westernworld, live attenuated vaccines havebeen considered unsuitable for use inhumans.55

Current vaccine supplies are limited

andthe US production capacity is mod-est. It will be yearsbefore increasedpro-duction efforts can make available suf-ficient quantities of vaccine for civilianuse. However, even if vaccine were avail-able,populationwide vaccinationwouldnotbe recommended at this time, giventhe costs and logistics of a large-scalevaccination program and the unlikelyoccurrenceof a bioterrorist attack in anygiven community. Vaccination of someessential service personnel should beconsidered if vaccine becomes avail-

able. Postexposure vaccination follow-ing a biological attack with anthraxwould be recommended with antibi-otic administration to protect against re-sidual retained spores, if vaccine wereavailable.

THERAPY

Recommendations regarding antibi-otic and vaccine use in the setting of abiological anthrax attack are condi-tioned by a limited number of studiesin experimental animals, current un-

derstanding of antibiotic resistancepat-terns, and the possible requirement totreat large numbers of casualties. Anumber of possible therapeutic strate-gies have yet to be fully explored ex-perimentally or submitted for ap-proval to the FDA. For these reasons,the working group offers consensus rec-ommendations based on the best avail-able evidence. The recommendations

do not represent uses currently ap-proved by the FDA or an official posi-tion on the part of any of the federalagencies whose scientists participatedin these discussions and will need to be

revised as further relevant informa-tion becomes available.Given the rapid course of symptom-

atic inhalational anthrax, early antibi-otic administration is essential. A de-lay of antibiotic treatment for patientswith anthrax infection even by hoursmay substantially lessen chances forsurvival.56,57 Given the difficulty inachieving rapid microbiologic diagno-sis of anthrax, all persons with fever orevidence of systemic disease in an areawhere anthrax cases are occurringshould be treated for anthrax until the

disease is excluded.There are no clinical studies of the

treatmentof inhalationalanthrax in hu-mans. Thus, antibiotic regimens com-monly recommended for empiricaltreatment of sepsis have not been stud-iedin this setting. In fact, natural strainsof B anthracis are resistant to many of the antibiotics used in these empiricalregimens, such as those of theextended-spectrum cephalosporins.58,59 Mostnaturally occurring anthrax strains aresensitiveto penicillin, andpenicillin his-

torically has been the preferred therapyfor the treatment of anthrax. Penicil-lin is approved by the FDA for this in-dication,41,56,57 as is doxycycline.60 Instudies of small numbers of monkeysinfectedwithsusceptiblestrains of Ban-thracis, oral doxycycline hasproved ef-ficacious.41

Doxycycline is the preferred optionfrom the tetracycline class of antibiot-icsbecause of itsprovenefficacy in mon-key studies and its ease of administra-tion. Other members of this class of

antibiotics are suitable alternatives. Al-though treatment of anthrax infectionwith ciprofloxacin has not been stud-ied in humans, animal models suggestexcellent efficacy.28,41,61 Invitrodatasug-gest that other fluoroquinolone antibi-otics would have equivalent efficacy intreating anthrax infection, although noanimal data exist for fluoroquinolonesother than ciprofloxacin.59






Reports have been published of a Banthracis vaccine strain that has beenengineered by Russian scientists to re-sist the tetracycline and penicillinclasses of antibiotics.62 Although the en-

gineering of quinolone-resistant B an-thracis may also be possible, to datethere have been no published ac-counts of this.

Balancing considerations of efficacywith concerns regarding resistance, theworking grouprecommends that cipro-floxacin or other fluoroquinolonetherapy be initiated in adults with pre-sumed inhalational anthrax infection.Antibiotic resistance to penicillin- andtetracycline-class antibiotics should beassumed following a terrorist attack un-til laboratory testing demonstrates oth-

erwise. Once the antibiotic suscepti-bility of the B anthracis strain of theindex case has been determined, themost widely available, efficacious, andleast toxic antibiotic should be admin-istered to patients and persons requir-ing postexposure prophylaxis.

In a contained casualty setting (a situ-ation in which a modest number of pa-tients require therapy), the workinggroup recommends intravenous antibi-otic therapy, as shown in TABLE 2.Ifthenumber of persons requiring therapy is

sufficiently high (ie, a mass casualty set-ting), the working group recognizesthatintravenous therapy will no longer bepossible forreasonsof logistics and/or ex-haustion of equipment and antibioticsupplies, and oral therapy will need tobeused (TABLE 3). The threshold num-

ber of cases at which parenteral therapybecomes impossible depends on a vari-ety of factors, including local and re-gional health care resources.

In experimental animals, antibiotic

therapy during anthrax infection hasprevented development of an immuneresponse.28,62 This suggests that even if the antibiotic-treated patient survivesanthrax infection, risk for recurrenceremains for at least 60 days because of the possibility of delayed germinationof spores. Therefore, the working grouprecommends that antibiotic therapy becontinued for 60 days, with oral therapyreplacing intravenous therapy as soonas a patient’s clinical condition im-proves. If vaccine were to becomewidely available, postexposure vacci-

nation in patients being treated for an-thrax infection might permit the dura-tion of antibiotic administration to beshortened to 30 to 45 days, with con-comitant administration of 3 doses of anthrax vaccine at 0, 2, and 4 weeks.

The treatment of cutaneous anthraxhistorically has been with oral penicil-lin. Theworking grouprecommendsthatoral fluoroquinolone or tetracycline an-tibioticsas well asamoxicillin in theadultdosage schedules described in Tables 2and3 wouldbe suitablealternativesifan-

tibiotic susceptibility is proved. Al-though previous guidelines have sug-gested treating cutaneous anthrax for 7to 10 days,23,49 the working group rec-ommends treatment for 60 days in thesetting of bioterrorism, given the pre-sumed exposure to the primary aero-

sol. Treatment of cutaneous anthrax gen-erally prevents progression to systemicdisease, althoughit does not prevent theformation and evolution of the eschar.Topical therapy is not useful.2

Other antibiotics effective against Banthracis in vitro include chlorampheni-col, erythromycin, clindamycin, ex-tended-spectrum penicillins, macro-lides, aminoglycosides, vancomycinhydrochloride, cefazolin,and other first-generation cephalosporins.58,59,64 Theef-ficacy of these antibiotics has not beentestedin humansor animal studies. Theworking group recommends the use of these antibiotics only if the previouslycited antibiotics areunavailable or if thestrain is otherwise antibiotic resistant.Natural resistance of B anthracis strains

exists againstsulfamethoxazole, trimeth-oprim, cefuroxime, cefotaxime so-dium, aztreonam, andceftazidime.58,59,64

Therefore, these antibiotics should notbe used in the treatment or prophy-laxis of anthrax infection.

Postexposure Prophylaxis

Guidelines regarding which popula-tions would require postexposure pro-phylaxis following the release of an-thrax as a biological weapon wouldneed to be developed quickly by state

and local health departments in con-sultation with national experts. Thesedecisions require estimates of the tim-ingand location of theexposure and therelevant weather conditions in an out-door release.65 Ongoing monitoring of cases would be needed to define the

Table 2. Working Group Recommendations for Medical Therapy for Patients With Clinically Evident Inhalational Anthrax Infectionin the Contained Casualty Setting28,41,60,62,63*

Initial Therapy† Optimal Therapy if Strain Is Pr ov en Susceptible Duration, d‡

Adults Ciprofloxacin, 400 mg i ntravenouslyevery 12 h

Penicillin G, 4 million U intravenously every 4 hDoxycycline, 100 mg intravenously every 12 h§

60

Children Ciprofloxacin, 20-30 mg/kg per dayintravenously divided into 2 dailydoses, not to exceed 1 g/d

Age Ͻ12 y: penicillin G, 50 000 U/kg intravenously every 6 h Age Ն12 y: penicillin G, 4 million U intravenously every 4 h

60

Pregnant women¶ Same as for nonpregnant adults

Immunos uppress ed persons Same as for nonimmunos uppress ed adults and chi ldren

*Most recommendationsare based onanimal studies or in vitro studies andare notapproved by theUS Food andDrugAdministration(FDA). These recommendationsare notFDA approved but were reached by consensus of the working group. See text for explanations and alternatives.

†In vitro studies suggest ofloxacin, 400 mg intravenously every 12 hours, or levofloxacin, 500 mg intravenously every 24 hours, could be substituted for ciprofloxacin.‡Oral antibiotics should be substituted for intravenous antibiotics as soon as clinical condition improves.§In vitro studies suggest tetracycline could be substituted for doxycycline.Doxycycline could also be used. For children heavier than 45 kg, use adult dosage. For children 45 kg or lighter, use 2.5 mg/kg doxycycline intravenously every 12 hours. Refer to

management of pediatric population in text for details.¶Refer to section on management of pregnant women in text for details.






high-risk areas, direct follow-up, andguide the addition or deletion of groupsto receive postexposure prophylaxis.

There are no FDA-approved postex-posure antibiotic regimens following

exposuretoananthraxaerosol.Forpost-exposure prophylaxis, the workinggroup recommends the same antibi-otic regimen as that recommended fortreatment of mass casualties; prophy-laxis should be continued for 60 days(Table 3).

Management of Special Groups

Consensus recommendations for spe-cial groupsas set forth herein reflect theclinical and evidence-based judgmentsofthe working group and at thistime donot necessarily correspond with FDA-

approved use, indications, or labeling.Children. It has been recommended

that ciprofloxacin and other fluoroqui-nolones should not be used in childrenyounger than 16 to 18 years because of a link to permanent arthropathy in ado-lescent animals and transient arthropa-thyin a small number ofchildren.60 How-ever, balancing these risks against therisks of anthrax caused by an engi-neered antibiotic-resistant strain, theworking group recommends that cipro-floxacin be used in the pediatric popu-

lation for initial therapy or postexpo-sure prophylaxis following an anthraxattack (Table 2). If antibiotic suscepti-bility testingallows, penicillin shouldbesubstituted for the fluoroquinolone.

As a third alternative, doxycyclinecould be used. The American Acad-

emy of Pediatrics has recommendedthat doxycycline not be used in chil-dren younger than 9 years because thedrug has resulted in retarded skeletalgrowth in infants and discolored teeth

in infants and children.

60

However, theseriousrisk of infection following an an-thrax attack supports the consensusrec-ommendation that doxycycline be usedin children if antibiotic susceptibilitytesting, exhaustion of drug supplies, orallergic reaction preclude use of peni-cillin and ciprofloxacin.

In a contained casualty setting, theworking group recommends that chil-dren receive intravenous antibiotics(Table2). In a mass casualty setting andas postexposure prophylaxis, the work-ing group recommends thatchildrenre-

ceive oral antibiotics (Table 3).The US vaccine is licensed for use

only in persons aged 18 to 65 years be-cause studies to date have been con-ducted exclusively in this group.52 Nodata exist for children, but based on ex-perience with other inactivated vac-cines, it is likely that thevaccinewouldbe safe and effective.

Pregnant Women. Fluoroquino-lones are not generally recommendedduring pregnancy because of theirknown association with arthropathy in

adolescent animals and small num-bers of children. Animal studies havediscovered no evidence of teratogenic-ity related to ciprofloxacin, but no con-trolled studies of ciprofloxacin in preg-nant women have been conducted.Balancing these possible risks against

the concerns of anthrax due to engi-neered antibiotic-resistant strains, theworking group recommends that cipro-floxacin be used in pregnant women fortherapy and postexposure prophy-

laxis following an anthrax attack(Tables 2 and 3). No adequate con-trolled trials of penicillin or amoxicil-linadministration during pregnancy ex-ist. However, the CDC recommendspenicillin for the treatment of syphilisduring pregnancy and amoxicillin as atreatment alternativefor chlamydial in-fections during pregnancy.60

The working group recommendsthatpregnant women receive fluoroquino-lones in the usual adult dosages. If sus-ceptibility testing allows, intravenouspenicillin in the usual adult dosages

should be substituted for fluoroquino-lones. As a third alternative, intrave-nous doxycycline could be used. Thetetracycline class of antibiotics has beenassociated with both toxic effects in theliver in pregnant women and fetal toxiceffects, including retarded skeletalgrowth.60 Balancing the risks of an-thrax infection with those associatedwith doxycycline use in pregnancy, theworking grouprecommends that doxy-cycline be used in pregnant women fortherapy and postexposure prophy-

laxis if antibiotic susceptibility test-ing, exhaustion of drug supplies, or al-lergic sensitivity preclude the use of penicillin and ciprofloxacin. If doxy-cycline is used in pregnant women, pe-riodic liver function testing should beperformed if possible.

Table 3. Working Group Recommendations for Medical Therapy for Patients With Clinically Evident Anthrax Infection in the Mass CasualtySetting or for Postexposure Prophylaxis41*

Initial Ther apy† Optimal Therapy if Strain Is Pr ov en Susceptible Dur ation, d

Adults Ciprofloxacin, 500 mg by mouth every 12 h Amoxicillin, 500 mg every 8 hDoxycycline, 100 mg by mouth every 12 h‡

60

Children§ Ciprofloxaci n, 20-30 mg/kg per day by mouthdivided into 2 daily doses, not to exceed1 g/d

Weight Ն20 kg: amoxicillin, 500 mg by mouth every 8 hWeight Ͻ20 kg: amoxicillin, 40 mg/kg divided into

3 doses to be taken every 8 h

60

Pregnant women Ciprofloxaci n, 500 mg by mouth every 12 h Amoxicillin, 500 mg by mouth every 8 h 60

Immunos uppress ed persons Same as for nonimmunosuppres sed adults and chi ldren

*Most recommendationsare based onanimal studies or in vitro studies andare notapproved by theUS Food andDrugAdministration(FDA). These recommendationsare notFDA approved but were reached by consensus of the working group. See text for explanations and alternatives.

†In vitro studies suggest ofloxacin, 400 mg by mouth every 12 hours, or levofloxacin, 500 mg by mouth every 24 hours, could be substituted for ciprofloxacin.‡In vitro studies suggest tetracycline, 500 mg by mouth every 6 hours, could be substituted for doxycycline.§Doxycycline could also be used. For children heavier than 45 kg, use adult dosage. For children 45 kg or lighter, use 2.5 mg/kg doxycycline by mouth every 12 hours. Refer to

management of pediatric population in text for details.Refer to management of pregnant population in text for details.






Ciprofloxacin (and other fluoroqui-nolones), penicillin, and doxycycline(and other tetracyclines) are each ex-creted in breast milk. Therefore, abreast-feeding woman shouldbe treated

or given prophylaxis with the same an-tibiotic as her infant based on what ismost safe and effective for the infant(see pediatric guidelines herein) tominimize risk to the infant.

Immunosuppressed Persons. Theantibiotic treatment or postexposureprophylaxis for anthrax among thosewho are immunosuppressed has notbeen studied in human or animal mod-els of anthrax infection. Therefore, theworking group consensus recommen-dationis to administer antibiotics as forimmunocompetent adults and chil-

dren (Tables 2 and 3).

INFECTION CONTROL

There are no data to suggest patient-to-patient transmission of anthrax oc-curs.8,46 Thus, standard barrier isola-tion precautions are recommended forhospitalized patients with all forms of anthrax infection, but the use of high-efficiency particulate air filter masks orother measures for airborne protec-tion are not indicated.66 There is noneed to immunize or provide prophy-

laxis to patient contacts (eg, house-hold contacts, friends, coworkers) un-less a determination is made that they,like the patient, were exposed to theaerosol at the time of the attack.

In addition to immediate notifica-tion of the hospital epidemiologist andstate health department, the local hos-pital microbiology laboratories shouldbe notified at the first indication of an-thrax so that safe specimen processingunder biosafety level 2 conditions canbe undertaken.41,67 A number of disin-

fectants used for standard hospital in-fection control, such as hypochlorite, areeffective in cleaning environmentalsur-faces contaminatedwithinfectedbodilyfluids.17,66

Proper burial or cremation of hu-mans and animals who have died be-cause of anthrax infection is impor-tant in preventing further transmissionof the disease. Serious consideration

should be given to cremation. Embalm-ing of bodies could be associated withspecial risks.66 If autopsies are per-formed, all related instruments andma-terials should be autoclaved or incin-

erated.66 Animal transmission mightoccur if infected animal remains are notcremated or buried.16,21

DECONTAMINATION

Recommendations regardingdecontami-nationin theeventof anintentionalaero-solization of anthrax sporesare based onevidence concerning aerosolization,an-thrax spore survival, and environmen-tal exposures at Sverdlovsk and amonggoat hair mill workers. Thegreatest riskto human health following an inten-

tional aerosolization of anthrax sporesoccurs during the period in which an-thrax spores remain airborne, called pri-mary aerosolization. The duration forwhich spores remain airborne and thedistance spores travel before they be-come noninfectious or fall to the groundis dependent on meteorological condi-tions and aerobiological propertiesof thedispersed aerosol.8,65 Under circum-stances of maximum survival and per-sistence, the aerosol would be fully dis-persed within hoursto 1 dayat most, well

beforethefirst symptomatic cases wouldbe seen. Following the discovery that abioweaponhasbeen used, anthraxsporesmay be detected on environmental sur-facesusing rapidassaykitsor culture,butthey provide no indication as to the riskof reaerosolization.

The risk that anthrax spores mightpose to public health after the period of primary aerosolization can be inferred

from the Sverdlovsk experience, inves-tigations in animal hair processingplants, and modeling analyses by theUSArmy. At Sverdlovsk, new cases of in-halational anthrax developed as late as

43 days after the presumed date of re-lease, but none occurred during themonthsandyears afterward.68 Some havequestioned whether any of those caseswith onset of disease beyond 7 daysmight have represented illness follow-ing resuspension of spores from theground or other surfaces, a process thathas been called secondary aerosoliza-tion. While it is impossible to state withcertainty that secondary aerosoliza-tions did not occur, it appears un-likely. It should be noted that few ef-

forts were made to decontaminate theenvironment after the accident and only47 000 of the city’s 1 million inhabit-ants were vaccinated.8 The epidemiccurve (FIGURE 4) is typical for a com-mon-source epidemic, and it is pos-sible to account for virtually all pa-tients havingbeen withinthearea of theplume on thedayof theaccident. More-over, if secondary aerosolization hadbeen important, new cases almost cer-tainly would have continued for a pe-riod well beyond the observed 43 days.

Although persons working with ani-mal hair or hides are known to be at in-creased risk of developing inhalationalor cutaneous anthrax, surprisingly fewof those exposed in the United Stateshavedeveloped disease.During the firsthalf of this century, a significant num-ber of goat hair mill workers were likelyexposed to aerosolized spores. Manda-tory vaccination became a requirement

Figure 4. Day of Onset of Inhalational Anthrax Following Sverdlovsk Accident

8

7

6

5

3

4

2

1

00 2 4 6 8 10 12 14 1 6 18 2 0 22 2 4 2 6 28 3 0 32 3 4 36 38 4 0 42 4 4

No. of Days After Accident

N o . o f N e w C

a s e s

Figure is based on data from Guillermin. 68






for workingin goat hair mills only in the1960s. Meanwhile, many unvacci-nated person-years of high-risk expo-sure had occurred, but only 13 cases of inhalational anthrax were reported.19,44

One study of environmental exposurewas conducted at a Pennsylvania goathair mill at which workers were shownto inhale up to 510 B anthracis par-ticlesofat least 5 µmin diameter per per-son per 8-hour shift. These concentra-tions of spores were constantly presentin the environment during the time of this study,44 but no cases of inhala-tional anthrax occurred.

Modeling analyses have been carriedoutby US Army scientists seeking to de-termine the risk of secondary aerosol-ization. Onestudy concluded that there

was no significantthreatto personnel inareas contaminated by 1 million sporesper square meter either from traffic onasphalt-paved roads or from a runwayused by helicopters or jet aircraft.69 Aseparate study showed that in areas of ground contaminated with 20 millionBacillus subtilis spores per square me-ter, a soldier exercising actively for a3-hour period would inhale between1000 and 15 000 spores.70

Much has been written about thetechnical difficulty of decontaminating

an environment contaminated with an-thrax spores. A classic case is the expe-rience at Gruinard Island in the UnitedKingdom. During World War II, Brit-ish military undertook explosives test-ing with anthrax spores on this islandoff the Scottish coast. Spores persistedand remained viable for 36 years fol-lowing the conclusion of testing. De-contamination of the islandoccurred instages, beginning in 1979 andending in1987, when the island was finally de-clared fully decontaminated. The total

cost is unpublished, but materials re-quired included 280 tons of formalde-hyde and 2000 tons of seawater.17,71

If an environmental surface is provedto be heavily contaminated with an-thrax spores in the immediate area of a spill or close proximity to the pointof release of an anthrax aerosol, decon-tamination ofthat area may decrease theslight risk of acquiring anthrax by sec-

ondary aerosolization. However, de-contamination of large urban areas oreven a building following an exposureto an anthrax aerosol would be ex-tremely difficult and is not indicated.

Although the risk of disease caused bysecondary aerosolization would be ex-tremely low, it would be difficult to of-fer absolute assurance that there wasnotrisk whatsoever. Postexposure vacci-nation, if vaccine were available, mightbe a possible intervention that couldfurther lower the risk of anthrax infec-tion in this setting.

In the setting of an announced al-leged anthrax release, such as the se-ries ofanthrax hoaxesoccurring in manyareas of the United States in 1998,48 anyperson coming in direct physical con-

tact with a substance alleged to be an-thrax should perform thorough wash-ing of the exposed skin and articles of clothing with soap and water.72 Fur-ther decontamination of directly ex-posed individuals or of others is not in-dicated. In addition,any personin directphysical contact with the alleged sub-stance should receive postexposure an-tibiotic prophylaxis until the sub-stance is proved not to be anthrax.If thealleged substance is proved to be an-thrax, immediate consultation with ex-

pertsat theCDC and USAMRIIDshouldbe obtained.

ADDITIONAL RESEARCH

To develop a maximally effective re-sponse to a bioterrorist incident involv-ing anthrax,the medicalcommunitywillrequire new knowledge of the organ-ism, its genetics and pathogenesis, im-proved rapid diagnostic techniques, im-proved prophylactic and therapeuticregimens, and an improved second-generation vaccine.47 A recently pub-

lishedRussian study indicatesthat genestransferred fromthe relatedB cereuscanactto enableB anthracis to evadethe pro-tective effect of the live attenuated Rus-sian vaccine in a rodent model.73 Re-search is needed to determine the roleof these genes with respect to virulenceandability to evade vaccine-inducedim-munity. Furthermore, the relevance of this finding for the US vaccine needs to

beestablished. An acceleratedvaccine de-velopment effort is needed to allow themanufacture of an improved second-generation product that requires fewerdoses. Finally, an expanded knowledge

base is neededregarding possible maxi-mum incubation times after inhalationof spore-containing aerosols and opti-mal postexposure antibiotic regimens.

Ex Officio Participants in the Working Group on Ci-vilian Biodefense: George Curlin, MD, National In-stitutes of Health, Bethesda, Md; Margaret Ham-burg,MD, andWilliamRaub,PhD,Office of AssistantSecretary for Planning and Evaluation,DHHS, Wash-ington, DC; Robert Knouss, MD, Office of Emer-gency Preparedness, DHHS, Rockville, Md; MarcelleLayton, MD, Office of Communicable Disease, NewYork CityHealth Department, NewYork,NY; andBrianMalkin and Stuart Nightingale, MD, FDA, Rockville.Funding/Support: Funding forthis study primarily wasprovided by each participant’s institution or agency.The Office of Emergency Preparedness, DHHS, pro-

vided travel funds for 4 members of the group.Disclaimers: In many cases, the indication and dos-ages andotherinformationare notconsistent withcur-rent approved labeling by theUS Food andDrug Ad-ministration(FDA). The recommendations on the useof drugs andvaccine foruses notapprovedby theFDAdo not represent the official views of the FDA or ofany of the federal agencies whose scientists partici-pated in these discussions. Unlabeled usesof theprod-ucts recommended are noted in the sections of thisarticle in which these products are discussed. Whereunlabeleduses areindicated, information used as thebasis for the recommendation is discussed.

The views, opinions, assertions, and findings con-tainedherein arethoseof theauthors andshould notbe construedas official US Department of Defense or US Department of Army positions, policies, or deci-sions unless so designated by other documentation.Additional Articles: This article is 1 in a series entitled

Medical and Public Health Management FollowingtheUseof a Biological Weapon: ConsensusStatements of the Working Group on Civilian Biodefense.Acknowledgment: Theworkinggroup wishesto thankJeanne Guillermin, PhD, professor of sociology, Bos-ton College, Boston, Mass, for her comments on themanuscript. Starting in 1992, Dr Guillermin directedthe interview project to verify onset, hospital, anddeathdata for the 1979 Sverdlovsk victims, which will bedetailed in Anthrax, A Book of Names, from Califor-nia Press. We also thank Matthew Meselson, Timo-thyTownsend, MD, Martin Hugh-Jones,MA, VetMB,MPH, PhD, and Philip Brachman, MD, for their re-view and commentary of the manuscript.

REFERENCES

1. Carter A, Deutsch J, Zelicow P. Catastrophic ter-rorism. Foreign Aff . 1998;77:80-95.2. Lew D. Bacillus anthracis(anthrax).In: MandellGL,Bennett JE, Dolin R, eds. Principles and Practices of Infectious Disease. New York, NY: Churchill Living-stone Inc; 1995:1885-1889.3. Christopher GW, Cieslak TJ, Pavlin JA, Eitzen EM.Biologicalwarfare: a historicalperspective. JAMA. 1997;278:412-417.4. ColeLA. Thespecter of biological weapons. Sci Am.December 1996:60-65.5. ZilinskasRA. Iraq’sbiological weapons: thepast asfuture? JAMA. 1997;278:418-424.6. Public Health Service Office of Emergency Pre-paredness. Proceedings of the Seminar on Respond-ing to the Consequences of Chemical and Biological






Terrorism. Washington, DC: US Dept of Health andHuman Services; 1995.7. WuDunnS, Miller J,Broad W.How Japan germter-roralerted world. NewYorkTimes.May 26, 1998:1-6.8. MeselsonM, GuilleminJ, Hugh-Jones M,et al.TheSverdlovsk anthrax outbreak of 1979. Science. 1994;266:1202-1208.9. World Health Organization. Health Aspects of

Chemical and BiologicalWeapons.Geneva, Switzer-land: World Health Organization; 1970:98-99.10. SimonJD. Biologicalterrorism: preparing to meetthe threat. JAMA. 1997;278:428-430.11. Cristy GA, Chester CV. Emergency ProtectionAgainst Aerosols. Oak Ridge, Tenn: Oak Ridge Na-tional Laboratory; 1981. Publication ORNL-5519.12. Office of Technology Assessment, US Congress.Proliferation of Weaponsof MassDestruction. Wash-ington, DC: US Government Printing Office; 1993:53-55. Publication OTA-ISC-559.13. Kaufmann AF, Meltzer MI, Schmid GP. The eco-nomic impact of a bioterroristattack.Emerg Infect Dis.1997;3:83-94.14. Kohout E, Sehat A, Ashraf M. Anthrax: a con-tinuous problem in south west Iran. Am J Med Sci.1964;247:565.15. Pienaar UV.Epidemiology of anthrax in wildani-mals andthecontrolon anthrax epizootics inthe Kruger

National Park, South Africa. Fed Proc . 1967;26:1496-1591.16. Dragon DC, Rennie RP. The ecology of anthraxspores. Can Vet J. 1995;36:295-301.17. Titball RW, Turnbull PC, Hutson RA. The moni-toring and detection of Bacillus anthracis in the en-vironment. J Appl Bacteriol. 1991;70(suppl):9S-18S.18. Brachman PS,FriedlanderA. Anthrax. In:PlotkinSA, Orenstein WA, eds. Vaccines. 3rd ed. Philadel-phia, Pa: WB Saunders Co; 1999:629-637.19. Brachman PS. Inhalation anthrax. Ann N Y Acad Sci. 1980;353:83-93.20. Centersfor Disease Control andPrevention. Sum-mary of notifiable diseases, 1945-1994. MMWRMorb Mortal Wkly Rep. 1994;43:70-78.21. Myenye KS, Siziya S, Peterson D. Factors associ-atedwith human anthraxoutbreakin theChikupo andNgandu villages of Murewa district in MashonalandEast Province, Zimbabwe. Cent Afr J Med. 1996;42:

312-315.22. Tekin A, Bulut N, Unal T. Acute abdomen due toanthrax. Br J Surg . 1997;84:813.23. Friedlander A. Anthrax. In: Zajtchuk R, BellamyRF, eds. Textbook of Military Medicine: Medical As- pects of Chemical and Biological Warfare. Washing-ton, DC: Office of the Surgeon General, US Dept ofthe Army; 1997:467-478.24. SirisanthanaT, NelsonKE, EzzellJW, Abshire TG.Serological studiesof patients withcutaneous andoral-pharyngeal anthrax from northern Thailand. AmJTrop Med Hyg . 1988;39:575-581.25. KunanusontC, Limpakarnjanarat K, Foy HM. Out-break of anthrax in Thailand. Ann Trop Med Parasi-tol. 1989;84:507-512.26. Sirisanthana T, Navachareon N, Tharavichitkul P,Sirisanthana V, Brown AE. Outbreak of oral-pharyngeal anthrax. Am J Trop Med Hyg . 1984;33:144-150.

27. Dutz W, Saidi F, Kouhout E. Gastricanthrax withmassive ascites. Gut . 1970;11:352-354.28. Friedlander A, Welkos SL, Pitt ML, et al. Postex-posureprophylaxisagainst experimental inhalation an-thrax. J Infect Dis. 1993;167:1239-1242.29. LincolnRE, HodgesDR, Klein F,et al. Roleof thelymphatics in thepathogenesis of anthrax. J Infect Dis.1965;115:481-494.30. Williams RP. Bacillus anthracis and other spore

forming bacilli. In: Braude AI, Davis LE, Fierer J, eds.Infectious Disease and Medical Microbiology. Phila-delphia, Pa: WB Saunders Co; 1986:270-278.31. Druett HA, Henderson DW, Packman L, Pea-cock S. Studies on respiratory infection. J Hyg . 1953;51:359-371.32. Hatch TF. Distribution and deposition of inhaledparticles in respiratory tract. Bacteriol Rev. 1961;25:

237-240.33. Ross JM. The pathogenesis of anthrax followingthe administration of spores by the respiratory route. J Pathol Bacteriol. 1957;73:485-495.34. Glassman HN. Industrialinhalationanthrax.Bac-teriol Rev. 1966;30:657-659.35. HendersonDW, Peacock S, BeltonFC. Observa-tions on the prophylaxis of experimental pulmonaryanthrax in the monkey. J Hyg. 1956;54:28-36.36. Smith H, Keppie J. Observations on experimen-tal anthrax. Nature. 1954;173:869-870.37. Defense Intelligence Agency. Soviet BiologicalWarfare Threat . Washington, DC: US Dept of De-fense; 1986. Publication DST-161OF-057-86.38. Amramova FA, Grinberg LM, Yampolskaya O,Walker DH. Pathology of inhalational anthrax in 42cases fromthe Sverdlovsk outbreakin 1979. ProcNatlAcad Sci U S A. 1993;90:2291-2294.39. Dalldorf F, Kaufmann AF, Brachman PS. Wool-

sorters’ disease. Arch Pathol. 1971;92:418-426.40. Gleiser CA, Berdjis CC, Harman HA, Gochenour WS. Pathology of experimental respiratory anthrax inMacaca Mulatta. Br J Exp Pathol. 1963;44:416-426.41. Franz DR, Jahrling PB, Friedlander A, et al. Clini-cal recognition and management of patients ex-posedto biologicalwarfare agents. JAMA. 1997;278:399-411.42. VessalK, Yeganehdoust J, Dutz W, KohoutE. Ra-diologic changes in inhalation anthrax. Clin Radiol.1975;26:471-474.43. Albrink WS, Brooks SM, Biron RE, Kopel M. Hu-man inhalation anthrax. Am J Pathol. 1960;36:457-471.44. Dahlgren CM, Buchanan LM, Decker HM, et al.Bacillusanthracisaerosols in goat hairprocessing mills.Am J Hyg . 1960;72:24-31.45. Walker JS, Lincoln RE, Klein F. Pathophysiologi-cal andbiochemical changesin anthrax.FedProc . 1967;

26:1539-1544.46. Pile JC,Malone JD,Eitzen EM,FriedlanderA. An-thrax as a potentialbiological warfare agent. Arch In-tern Med . 1998;158:429-434.47. Institute of Medicine National Research Council.Improving Civilian Medical Response to Chemical and Biological Terrorist Incidents. Washington, DC: Na-tional Academy Press; 1998:1-70.48. Centers for Disease Control and Prevention. Bio-terrorism alleging use of anthrax and interim guide-linesfor management—UnitedStates, 1998. MMWR Morb Mortal Wkly Rep. 1999;48:69-74.49. Penn C, Klotz SA. Anthrax. In: Gorbach SL, Bar-tlett JG, Blacklow NR, eds. Infectious Diseases. Phila-delphia, Pa: WB Saunders Co; 1998:1575-1578.50. Brachman PS. Anthrax. In: Hoeprich PD, JordanMC, Ronald AR, eds. Infectious Diseases. Philadel-phia, Pa: JB Lippincott; 1994:1003-1008.51. Anthrax vaccine, military use in Persian Gulf re-

gion [pressrelease]. Washington, DC:US Dept ofDe-fense; September 8, 1998.52. Michigan Department of Public Health. AnthraxVaccine Absorbed . Lansing: Michigan Dept of PublicHealth; 1978.53. Brachman PS,Gold H, PlotkinSA, FeketyFR, Wer-rin M, Ingraham NR. Field evaluation of humananthrax vaccine. Am J Public Health. 1962;52:632-645.

54. IvinsBE, Fellows P, Pitt ML,et al.Efficacyof stan-dard human anthrax vaccine against Bacillus anthra-cis aerosol spore challenge in rhesus monkeys. Salis-bury Med Bull. 1996;87:125-126.55. Turnbull PC. Anthrax vaccines: past,present andfuture. Vaccine. 1991;9:533-539.56. BarnesJM. Penicillinand B anthracis. J PatholBac-teriol. 1947;194:113-125.

57. Lincoln RE, Klein F, Walker JS, et al. Successfultreatment of monkeys for septicemic anthrax. In: An-timicrobialAgents and Chemotherapy—1964. Wash-ington,DC: AmericanSocietyfor Microbiology; 1965:759-763.58. Odendaal MW, Peterson PM, de Vos V, BothaAD. The antibiotic sensitivity patterns of Bacillus an-thracis isolatedfrom theKruger NationalPark. Onder- stepoort J Vet Res. 1991;58:17-19.59. Doganay M, Aydin N. Antimicrobial susceptibil-ity of Bacillus anthracis. Scand J Infect Dis. 1991;23:333-335.60. AmericanHospital Formulary Service. AHFS Drug Information. Bethesda,Md: American Society of HealthSystem Pharmacists; 1996.61. Kelly D, ChulayJD, Mikesell P, Friedlander A. Se-rum concentrationsof penicillin,doxycycline,and cipro-floxacin during prolonged therapy in rhesus mon-keys. J Infect Dis. 1992;166:1184-1187.

62. StepanovAV, MarininLI, Pomerantsev AP, Starit-sin NA. Development of novel vaccines against an-thrax in man. J Biotechnol. 1996;44:155-160.63. Schaad UB, Abdus Salam M, Aujard Y, et al. Useof fluoroquinolones in pediatrics. Pediatr InfectDis J.1995;14:1-9.64. Lightfoot NF, Scott RJ, Turnbull PC. Antimicro-bialsusceptibilityof Bacillus anthracis:proceedingsoftheinternational workshopon anthrax. SalisburyMed Bull. 1990;68:95-98.65. Perkins WA. Public health implications of air-borne infection. Bacteriol Rev. 1961;25:347-355.66. American PublicHealth Association.Anthrax. In:Benenson AS, ed. Control of Communicable Dis-eases Manual. Washington, DC: American PublicHealth Association; 1995:18-22.67. Morse S, McDade J. Recommendations for work-ing with pathogenicbacteria. Methods Enzymol. 1994;235:1-26.

68. Guillermin J. Anthrax: The Investigation of a Le-thal Outbreak. Berkeley: University of California Press.In press.69. Chinn KS. Reaerosolization Hazard Assessment for Biological Agent-Contaminated Hardstand Ar-eas. Life Sciences Division, Dugway Proving Ground,Utah: US Dept of the Army; 1996:1-40. PublicationDPG/JCP-96/012.70. Resnick IG, Martin DD, Larsen LD. Evaluation of Need forDetection of Surface Biological Agent Con-tamination. Dugway Proving Ground, Life SciencesDivision, US Dept of the Army; 1990:1-35. Publica-tion DPG-FR-90-711.71. Manchee RJ, Stewart WD. The decontaminationof Gruinard Island. Chem Br. July 1988;690-691.72. US Army Medical Research Institute of Infec-tious Diseases, Centers for Disease Control and Pre-vention, andUS Food andDrug Administration. Medi-cal Response to Biological Warfare and Terrorism.

Gaithersburg, Md: US Army Medical Research Insti-tute of Infectious Diseases, Centers for Disease Con-trol and Prevention, and US Food and Drug Admin-istration; 1998.73. Pomerantsev AP, Staritsin NA, Mockov YV, Ma-rinin LI. Expression of cereolysine AB genes in Bacil-lus anthracisvaccine strain ensures protectionagainstexperimental hemolytic anthrax infection. Vaccine.1997;15:1846-1850.






RESEARCH LETTER

NIH Research Grants: Funding and Re-funding

To the Editor: National Institutes of Health (NIH) grants playan important role in the careers of the research faculty in medi-

cal schools. Given the importance of NIH funding in an aca-demic career, we sought to determine several aspects of NIH

funding that have, to the best of our knowledge, not been ex-

amined in detail. These are (1) the median length of time of NIH funding during the academic life of an individual, (2) the

predictive value, for long-term funding, of receiving NIH fund-ing at any given time, and (3) the effect of length of the un-

funded period on the likelihoodthat an investigator will be sub-

sequently funded.

Methods. We obtained the funding histories of individu-als who were awarded any NIH grants in the index years1972 and 1982, and tracked individual funding histories for25 years and 15 years, respectively. Data consisted of the

number of grants each individual received annually follow-ingthe index year.A totalof 1707 individuals received awardsin 1972 and 1639 in 1982. Because of the nature of recordkeeping at the NIH, the precise length of each award wasnotavailable. Basedon thehistorical experience of NIHfund-ing, we assumed that the average length of an NIH award is4 years.

Results. The mean number of awards obtained by indi-viduals throughout their careers was 2.5 for the 1972 and3.3 for the 1982 cohort; the median for both groups was 2.0.About 40% of individuals who obtain an NIH grant neverreceived another for the rest of their careers.

The best predictor of future funding appears to be thenumber of grants garnered during a 10-year period. Thus,

for both the 1972 and 1982 cohorts, we computed the pro-portion of individuals obtaining l, 2, or more than 2awards over a 10-year period from 1972 to 1981 or 1982to 1991, respectively. The distribution of the individualsinto these 3 categories was similar in the 2 cohorts: 39.0%vs 35.0%, 18.3% vs 22.0%, and 42.4% vs 42.0%, respec-tively. For individuals who only obtained 1 grant for the10-year period (1972-1981 or 1982-1991), 92.5% and93%, respectively, did not obtain any funding for the sub-sequent period until 1997. Of those who received 1 addi-

tional grant (or a total of 2 grants for the 10-year period),72% did not obtain any subsequent funding until 1997.These 2 cohorts were in contrast with those who obtainedmore than 2 grants for the 10-year interval. Only 27% of these individuals failed to obtain a grant for the subse-quent period.

Comment. The cumulative period of NIH funding ap-pears to be rather brief for most individuals. Neither the pos-session of an NIH grant nor the apparent ability to acquireone based on prior research training or history predictswhether an individual will receive future grants.

T. V. Rajan, MD, PhD Jonathan Clive, PhDUniversity of Connecticut Health CenterFarmington

CORRECTIONS

Incorrect Equation: In the Original Contribution entitled “Empirical Evidence ofDesign-Related Bias in Studies of Diagnostic Tests” published in the September

15, 1999, issue of THE JOURNAL (1999;282:1061-1066), the equation was incor-rect. On page 1063, the equation should have appeared:

DOR =Sensitivity

(1 − Sensitivity)Ϭ

(1 − Specificity)

Specificity

Incorrect Color Reproduction: In the Consensus Statement entitled “Anthrax asa Biological Weapon: Medical and Public Health Management” published in theMay 12, 1999, issueof THE JOURNAL (1999;281:1735-1745), the colorof the pho-tomicrograph in Figure 1 on page 1727 was incorrectly reproduced. The correctimage showing gram-positive anthrax bacilli in a peripheral blood smear from arhesus monkey that died of inhalational anthrax is shown below, left. A secondphotomicrograph (lower magnification) of a gram stain of the peripheral blood ofa rhesus monkey that died of inhalational anthrax is also shown below, right.

Figure. Gram Stains of Bacillus anthracis

Left, Reprinted with permission from Zajtchuk R, Bellamy RF, eds. Textbook of Military Medicine: Medical Aspects of Chemical and Biological Warfare. Wash-ington, DC: Office of the Surgeon General, US Dept of the Army; 1997.

LETTERS

©2000 American Medical Association. All rights reserved. JAMA, April 19, 2000—Vol 283, No. 15 1963

anthrax as a biological weapon

Documents