Plague As a Biological Weapon Medical and Public Health Management

Home , Epidemiology of plague, Pathogenic bacteria

CONSENSUS STATEMENT

Plague as a Biological Weapon Medical and Public Health Management

Thomas V. Inglesby, MD Objective The Working Group on Civilian Biodefense has developed consensus- David T. Dennis, MD, MPH based recommendations for measures to be taken by medical and public health profes- Donald A. Henderson, MD, MPH sionals following the use of plague as a biological weapon against a civilian population. Participants The working group included 25 representatives from major academic John G. Bartlett, MD medical centers and research, government, military, public health, and emergency man- Michael S. Ascher, MD agement institutions and agencies. Edward Eitzen, MD, MPH Evidence MEDLINE databases were searched from January 1966 to June 1998 for the Medical Subject Headings plague, Yersinia pestis, biological weapon, biological ter- Anne D. Fine, MD rorism, biological warfare, and biowarfare. Review of the bibliographies of the refer- Arthur M. Friedlander, MD ences identified by this search led to subsequent identification of relevant references pub- Jerome Hauer, MPH lished prior to 1966. In addition, participants identified other unpublished references and sources. Additional MEDLINE searches were conducted through January 2000. John F. Koerner, MPH, CIH Consensus Process The first draft of the consensus statement was a synthesis of in- Marcelle Layton, MD formation obtained in the formal evidence-gathering process. The working group was Joseph McDade, PhD convened to review drafts of the document in October 1998 and May 1999. The final statement incorporates all relevant evidence obtained by the literature search in conjunc- Michael T. Osterholm, PhD, MPH tion with final consensus recommendations supported by all working group members. Tara O’Toole, MD, MPH Conclusions An aerosolized plague weapon could cause fever, cough, chest pain, Gerald Parker, PhD, DVM and hemoptysis with signs consistent with severe pneumonia 1 to 6 days after exposure. Rapid evolution of disease would occur in the 2 to 4 days after symptom onset Trish M. Perl, MD, MSc and would lead to septic shock with high mortality without early treatment. Early treat- Philip K. Russell, MD ment and prophylaxis with streptomycin or gentamicin or the tetracycline or fluoroquinolone classes of antimicrobials would be advised. Monica Schoch-Spana, PhD JAMA. 2000;283:2281-2290 www.jama.com Kevin Tonat, DrPH, MPH the availability of Y pestis around the biological terrorism, biological warfare, for the Working Group on Civilian Biodefense world, capacity for its mass produc- and biowarfare. Review of the bibliog- tion and aerosol dissemination, diffi- raphies of the references identified by HIS IS THE THIRD ARTICLE IN A culty in preventing such activities, high series entitled Medical and Pub- fatality rate of pneumonic plague, and Author Affiliations: Center for Civilian Biodefense Studies, Johns Hopkins University Schools of Medi- lic Health Management Follow- potential for secondary spread of cases cine (Drs Inglesby, Bartlett, and Perl) and Public Health ing the Use of a Biological during an epidemic, the potential use (Drs Henderson, O’Toole, Russell, and Schoch- Spana and Mr Koerner), Baltimore, Md; National Cen- TWeapon: Consensus Statements of the of plague as a biological weapon is of ter for Infectious Diseases, Centers for Disease Con- Working Group on Civilian Biode- great concern. trol and Prevention, Fort Collins, Colo (Dr Dennis), and Atlanta, Ga (Dr McDade); Viral and Rickettsial 1,2 fense. The working group has iden- Diseases Laboratory, California Department of Health tified a limited number of agents that, CONSENSUS METHODS Services, Berkeley (Dr Ascher); United States Army Medical Research Institute of Infectious Diseases, Fred- if used as weapons, could cause dis- The working group comprised 25 rep- erick, Md (Drs Eitzen, Friedlander, and Parker); Sci- ease and death in sufficient numbers to resentatives from major academic medi- ence Application International Corporation, McLean, Va (Mr Hauer); Office of Communicable Disease, New cripple a city or region. These agents cal centers and research, government, York City Health Department, New York, NY (Drs Fine and Layton); Office of Emergency Preparedness, De- also comprise the top of the list of military, public health, and emergency partment of Health and Human Services, Rockville, Md “Critical Biological Agents” recently de- management institutions and agencies. (Dr Tonat); and Infection Control Advisory Network Inc, Eden Prairie, Minn (Dr Osterholm). veloped by the Centers for Disease Con- MEDLINE databases were searched Corresponding Author and Reprints: Thomas V. trol and Prevention (CDC).3 Yersinia from January 1966 to June 1998 using Inglesby, MD, Johns Hopkins Center for Civilian Bio- defense Studies, Johns Hopkins University, Candler Bldg, pestis, the causative agent of plague, is the Medical Subject Headings (MeSH) Suite 850, 111 Market Place, Baltimore, MD 21202 one of the most serious of these. Given plague, Yersinia pestis, biological weapon, (e-mail: [email protected]).

©2000 American Medical Association. All rights reserved. JAMA, May 3, 2000—Vol 283, No. 17 2281 MANAGEMENT OF PLAGUE USED AS A BIOLOGICAL WEAPON this search led to subsequent identifi- rats and humans or more quickly from a microbiologist with suspect motives cation of relevant references published country to country by ships. The pan- was arrested after fraudulently acquir- prior to 1966. In addition, participants demic lasted more than 130 years and ing Y pestis by mail.10 New antiterrorism identified other unpublished refer- had major political, cultural, and reli- legislation was introduced in reaction. ences and sources in their fields of ex- gious ramifications. The third pan- pertise. Additional MEDLINE searches demic began in China in 1855, spread EPIDEMIOLOGY were conducted through January 2000 to all inhabited continents, and ulti- Naturally Occurring Plague during the review and revisions of the mately killed more than 12 million Human plague most commonly occurs statement. people in India and China alone.4 Small when plague-infected fleas bite hu- The first draft of the consensus state- outbreaks of plague continue to occur mans who then develop bubonic plague. ment was a synthesis of information ob- throughout the world.4,5 As a prelude to human epidemics, rats tained in the initial formal evidence- Advances in living conditions, pub- frequently die in large numbers, precipi- gathering process. Members of the lic health, and antibiotic therapy make tating the movement of the flea popula- working group were asked to make for- future pandemics improbable. How- tion from its natural rat reservoir to hu- mal written comments on this first draft ever, plague outbreaks following use of mans. Although most persons infected of the document in September 1998. The a biological weapon are a plausible threat. by this route develop bubonic plague, a document was revised incorporating In World War II, a secret branch of the small minority will develop sepsis with changes suggested by members of the Japanese army, Unit 731, is reported to no bubo, a form of plague termed pri- working group, which was convened to have dropped plague-infected fleas over mary septicemic plague. Neither bu- review the second draft of the docu- populated areas of China, thereby caus- bonic nor septicemic plague spreads di- ment on October 30, 1998. Following ing outbreaks of plague.6 In the ensu- rectly from person to person. A small this meeting and a second meeting of the ing years, the biological weapons pro- percentage of patients with bubonic or working group on May 24, 1999, a third grams of the United States and the Soviet septicemic plague develop secondary draft of the document was completed, Union developed techniques to aerosol- pneumonic plague and can then spread reviewed, and revised. Working group ize plague directly, eliminating depen- the disease by respiratory droplet. Per- members had a final opportunity to re- dence on the unpredictable flea vector. sons contracting the disease by this route view the document and suggest revi- In 1970, the World Health Organiza- develop primary pneumonic plague.11 sions. The final document incorpo- tion (WHO) reported that, in a worst- Plague remains an enzootic infection rates all relevant evidence obtained by case scenario, if 50 kg of Y pestis were re- of rats, ground squirrels, prairie dogs, and the literature search in conjunction with leased as an aerosol over a city of 5 other rodents on every populated con- consensus recommendations sup- million, pneumonic plague could oc- tinent except Australia.4 Worldwide, on ported by all working group members. cur in as many as 150000 persons, 36000 average in the last 50 years, 1700 cases The assessment and recommenda- of whom would be expected to die.7 The have been reported annually.4 In the tions provided herein represent the best plague bacilli would remain viable as an United States, 390 cases of plague were professional judgment of the working aerosol for 1 hour for a distance of up reported from 1947 to 1996, 84% of group based on data and expertise cur- to 10 km. Significant numbers of city in- which were bubonic, 13% septicemic, rently available. The conclusions and habitants might attempt to flee, further and 2% pneumonic. Concomitant case recommendations need to be regu- spreading the disease.7 fatality rates were 14%, 22%, and 57%, larly reassessed as new information be- While US scientists had not suc- respectively.12 Most US cases were in comes available. ceeded in making quantities of plague New Mexico, Arizona, Colorado, and organisms sufficient to use as an effec- California. Of the 15 cases following ex- HISTORY AND POTENTIAL tive weapon by the time the US offen- posure to domestic cats with plague, 4 AS A BIOTERRORIST AGENT sive program was terminated in 1970, were primary pneumonic plague.13 In the In AD 541, the first recorded plague pan- Soviet scientists were able to manufac- United States, the last case of human-to- demic began in Egypt and swept across ture large quantities of the agent suit- human transmission of plague oc- Europe with attributable population able for placing into weapons.8 More curred in Los Angeles in 1924.14,15 losses of between 50% and 60% in North than 10 institutes and thousands of sci- Although pneumonic plague has Africa, Europe, and central and south- entists were reported to have worked rarely been the dominant manifesta- ern Asia.4 The second plague pan- with plague in the former Soviet Union.8 tion of the disease, large outbreaks of demic, also known as the black death or In contrast, few scientists in the United pneumonic plague have occurred.16 In great pestilence, began in 1346 and even- States study this disease.9 an outbreak in Manchuria in 1910- tually killed 20 to 30 million people in There is little published information 1911, as many as 60000 persons devel- Europe, one third of the European popu- indicating actions of autonomous groups oped pneumonic plague; a second large lation.5 Plague spread slowly and inexo- or individuals seeking to develop plague Manchurian pneumonic plague out- rably from village to village by infected as a weapon. However, in 1995 in Ohio, break occurred in 1920-1921.16,17 As

2282 JAMA, May 3, 2000—Vol 283, No. 17 ©2000 American Medical Association. All rights reserved. MANAGEMENT OF PLAGUE USED AS A BIOLOGICAL WEAPON would be anticipated in the preantibi- son stain (FIGURE 1).20 Y pestis is a lac- Figure 1. Peripheral Blood Smear From otic era, nearly 100% of these cases were tose nonfermenter, urease and indole Patient With Septicemic Plague reported to be fatal.16,17 Reports from the negative, and a member of the Entero- Manchurian outbreaks suggested that in- bacteriaceae family.21 It grows opti- door contacts of affected patients were mally at 28°C on blood agar or Mac- at higher risk than outdoor contacts and Conkey agar, typically requiring 48 that cold temperature, increased humid- hours for observable growth, but colo- ity, and crowding contributed to in- nies are initially much smaller than creased spread.14,15 In northern India, other Enterobacteriaceae and may be there was an epidemic of pneumonic overlooked. Y pestis has a number of plague with 1400 deaths reported at virulence factors that enable it to sur- about the same time.15 While epidem- vive in humans by facilitating use of ics of pneumonic plague of this scale host nutrients, causing damage to host have not occurred since, smaller epi- cells, and subverting phagocytosis and demics of pneumonic plague have oc- other host defense mechanisms.4,11,21,22 Smear shows characteristic bipolar staining of Yersinia pestis bacilli (Wright-Giemsa stain; magnification, curred recently. In 1997 in Madagas- ϫ1000). Figure from Centers for Disease Control and car, 1 patient with bubonic plague and PATHOGENESIS AND Prevention, Division of Vector-Borne Infectious Dis- secondary pneumonic infection trans- CLINICAL MANIFESTATIONS eases, Fort Collins, Colo. mitted pneumonic plague to 18 per- Naturally Occurring Plague sons, 8 of whom died.18 In most cases of naturally occurring tis septicemia without a discernable bubo, plague, the bite by a plague-infected flea the form of disease termed primary sep- Plague Following Use leads to the inoculation of up to thou- ticemic plague.23 Septicemia can also arise of a Biological Weapon sands of organisms into a patient’s skin. secondary to bubonic plague.21 Septice- The epidemiology of plague following The bacteria migrate through cutane- mic plague may lead to disseminated its use as a biological weapon would dif- ous lymphatics to regional lymph nodes intravascular coagulation, necrosis of fer substantially from that of naturally where they are phagocytosed but re- small vessels, and purpuric skin lesions occurring infection. Intentional dis- sist destruction. They rapidly multi- (Figure 2, B). Gangrene of acral regions semination of plague would most prob- ply, causing destruction and necrosis such as the digits and nose may also occur ably occur via an aerosol of Y pestis,a of lymph node architecture with sub- in advanced disease, a process believed mechanism that has been shown to pro- sequent bacteremia, septicemia, and en- responsible for the name black death in duce disease in nonhuman primates.19 dotoxemia that can lead quickly to the second plague pandemic (Figure 2, A pneumonic plague outbreak would shock, disseminated intravascular co- C).21 However, the finding of gangrene result with symptoms initially resem- agulation, and coma.21 would not be expected to be helpful in bling those of other severe respiratory Patients typically develop symptoms diagnosing the disease in the early stages illnesses. The size of the outbreak would of bubonic plague 2 to 8 days after being of illness when early antibiotic treat- depend on factors including the quan- bitten by an infected flea. There is sud- ment could be lifesaving. tity of biological agent used, character- den onset of fever, chills, and weakness Secondary pneumonic plague devel- istics of the strain, environmental con- and the development of an acutely swol- ops in a minority of patients with bu- ditions, and methods of aerosolization. len tender lymph node, or bubo, up to bonic or primary septicemic plague— Symptoms would begin to occur 1 to 1 day later.23 The bubo most typically approximately 12% of total cases in the 6 days following exposure, and people develops in the groin, axilla, or cervical United States over the last 50 years.4 This would die quickly following onset of region (FIGURE 2, A) and is often so pain- process, termed secondary pneumonic symptoms.16 Indications that plague had ful that it prevents patients from mov- plague, develops via hematogenous been artificially disseminated would be ing the affected area of the body. Buboes spread of plague bacilli to the lungs. Pa- the occurrence of cases in locations not are 1 to 10 cm in diameter, and the over- tients commonly have symptoms of se- known to have enzootic infection, in lying skin is erythematous.21 They are vere bronchopneumonia, chest pain, persons without known risk factors, and extremely tender, nonfluctuant, and dyspnea, cough, and hemoptysis.16,21 in the absence of prior rodent deaths. warm and are often associated with con- Primary pneumonic plague result- siderable surrounding edema, but sel- ing from the inhalation of plague ba- MICROBIOLOGY AND dom lymphangitis. Rarely, buboes cilli occurs rarely in the United States.12 VIRULENCE FACTORS become fluctuant and suppurate. In addi- Reports of 2 recent cases of primary Y pestis is a nonmotile, gram-negative tion, pustules or skin ulcerations may pneumonic plague, contracted after han- bacillus, sometimes coccobacillus, that occur at the site of the flea bite in a minor- dling cats with pneumonic plague, re- shows bipolar (also termed safety pin) ity of patients. A small minority of veal that both patients had pneumonic staining with Wright, Giemsa, or Way- patients infected by fleas develop Y pes- symptoms as well as prominent gastro-

Figure 2. Patients With Naturally Occurring Plague

A B C

A, Cervical bubo in patient with bubonic plague; B, petechial and ecchymotic bleeding into the skin in patient with septicemic plague; and C, gangrene of the digits during the recovery phase of illness of patient shown in B. In plague following the use of a biological weapon, presence of cervical bubo is rare; purpuric skin lesions and necrotic digits occur only in advanced disease and would not be helpful in diagnosing the disease in the early stages of illness when antibiotic treatment can be lifesaving. Figures from Centers for Disease Control and Prevention, Division of Vector-Borne Infectious Diseases, Fort Collins, Colo. cal attack would be notably different agulation abnormalities, aminotrans- Figure 3. Chest Radiograph of Patient With Primary Pneumonic Plague than naturally occurring plague. In- ferase elevations, azotemia, and other haled aerosolized Y pestis bacilli would evidence of multiorgan failure. All are cause primary pneumonic plague. The nonspecific findings associated with time from exposure to aerosolized sepsis and systemic inflammatory re- plague bacilli until development of first sponse syndrome.11,21 symptoms in humans and nonhuman The time from respiratory exposure to primates has been found to be 1 to 6 death in humans is reported to have been days and most often, 2 to 4 days.12,16,19,26 between 2 to 6 days in epidemics dur- The first sign of illness would be ex- ing the preantibiotic era, with a mean of pected to be fever with cough and dys- 2 to 4 days in most epidemics.16 pnea, sometimes with the production of bloody, watery, or less commonly, DIAGNOSIS Radiograph shows extensive lobar consolidation in left purulent sputum.16,19,27 Prominent gas- Given the rarity of plague infection and lower and left middle lung fields. Figure from Cen- ters for Disease Control and Prevention, Division of trointestinal symptoms, including nau- the possibility that early cases are a har- Vector-Borne Infectious Diseases, Fort Collins, Colo. sea, vomiting, abdominal pain, and di- binger of a larger epidemic, the first clini- arrhea, might be present.24,25 cal or laboratory suspicion of plague must intestinal symptoms including nausea, The ensuing clinical findings of pri- lead to immediate notification of the hos- vomiting, abdominal pain, and diar- mary pneumonic plague are similar to pital epidemiologist or infection con- rhea. Diagnosis and treatment were de- those of any severe rapidly progres- trol practitioner, health department, and layed more than 24 hours after symp- sive pneumonia and are quite similar the local or state health laboratory. De- tom onset in both patients, both of to those of secondary pneumonic finitive tests can thereby be arranged rap- whom died.24,25 plague. Clinicopathological features idly through a state reference labora- Less common plague syndromes in- may help distinguish primary from sec- tory or, as necessary, the Diagnostic and clude plague meningitis and plague ondary pneumonic plague.11 In con- Reference Laboratory of the CDC and pharyngitis. Plague meningitis follows trast to secondary pneumonic plague, early interventions instituted. the hematogenous seeding of bacilli into features of primary pneumonic plague The early diagnosis of plague re- the meninges and is associated with fe- would include absence of buboes (ex- quires a high index of suspicion in natu- ver and meningismus. Plague pharyn- cept, rarely, cervical buboes) and, on rally occurring cases and even more so gitis follows inhalation or ingestion of pathologic examination, pulmonary dis- following the use of a biological weapon. plague bacilli and is associated with cer- ease with areas of profound lobular exu- There are no effective environmental vical lymphadenopathy.21 dation and bacillary aggregation.11 warning systems to detect an aerosol of Chest radiographic findings are vari- plague bacilli.28 Plague Following Use able but bilateral infiltrates or consoli- The first indication of a clandestine of a Biological Weapon dation are common (FIGURE 3).22 terrorist attack with plague would most The pathogenesis and clinical manifes- Laboratory studies may reveal leu- likely be a sudden outbreak of illness tations of plague following a biologi- kocytosis with toxic granulations, co- presenting as severe pneumonia and

2284 JAMA, May 3, 2000—Vol 283, No. 17 ©2000 American Medical Association. All rights reserved. MANAGEMENT OF PLAGUE USED AS A BIOLOGICAL WEAPON sepsis. If there are only small numbers Table 1. Diagnosis of Pneumonic Plague Infection Following Use of a Biological Weapon of cases, the possibility of them being Epidemiology Sudden appearance of many persons with fever, cough, shortness of breath, plague may be at first overlooked given and symptoms hemoptysis, and chest pain the clinical similarity to other bacte- Gastrointestinal symptoms common (eg, nausea, vomiting, rial or viral pneumonias and that few abdominal pain, and diarrhea) Western physicians have ever seen a Patients have fulminant course and high mortality case of pneumonic plague. However, Clinical signs Tachypnea, dyspnea, and cyanosis the sudden appearance of a large num- Pneumonic consolidation on chest examination ber of previously healthy patients with Sepsis, shock, and organ failure fever, cough, shortness of breath, chest Infrequent presence of cervical bubo pain, and a fulminant course leading to (Purpuric skin lesions and necrotic digits only in advanced disease) death should immediately suggest the Laboratory studies Sputum, blood, or lymph node aspirate Gram-negative bacilli with bipolar (safety pin) staining on Wright, Giemsa, or possibility of pneumonic plague or in- Wayson stain 1 halational anthrax. The presence of he- Rapid diagnostic tests available only at some health departments, the Centers moptysis in this setting would strongly for Disease Control and Prevention, and military laboratories suggest plague (TABLE 1).22 Pulmonary infiltrates or consolidation on chest radiograph There are no widely available rapid Pathology Lobular exudation, bacillary aggregation, and areas of necrosis in pulmonary diagnostic tests for plague.28 Tests that parenchyma would be used to confirm a suspected diagnosis—antigen detection, IgM en- identification, and there is some chance rate the development of primary pneu- zyme immunoassay, immunostain- that the diagnosis may be missed en- monic plague.19,31 It was used in special ing, and polymerase chain reaction— tirely. Approaches for biochemical char- circumstances for individuals deemed to are available only at some state health acterization of Y pestis are described in be at high risk of developing plague, such departments, the CDC, and military detail elsewhere.20 as military personnel working in plague laboratories.21 The routinely used pas- If a laboratory using automated or endemic areas, microbiologists work- sive hemagglutination antibody detec- nonautomated techniques is notified that ing with Y pestis in the laboratory, or tion assay is typically only of retrospec- plague is suspected, it should split the researchers working with plague- tive value since several days to weeks culture: 1 culture incubated at 28°C for infected rats or fleas. Research is ongo- usually pass after disease onset before rapid growth and the second culture ing in the pursuit of a vaccine that pro- antibodies develop. incubated at 37°C for identification of tects against primary pneumonic 22,32 Microbiologic studies are important the diagnostic capsular (F1) antigen. plague. in the diagnosis of pneumonic plague. Using these methods, up to 72 hours may A Gram stain of sputum or blood may be required following specimen pro- THERAPY reveal gram-negative bacilli or cocco- curement to make the identification Recommendations for the use of anti- bacilli.4,21,29 A Wright, Giemsa, or Way- (May Chu, PhD, CDC, Fort Collins, biotics following a plague biological son stain will often show bipolar stain- Colo, written communication, April 9, weapon exposure are conditioned by the ing (Figure 1), and direct fluorescent 1999). Antibiotic susceptibility testing lack of published trials in treating plague antibody testing, if available, may be should be performed at a reference labo- in humans, limited number of studies in positive. In the unlikely event that a cer- ratory because of the lack of standard- animals, and possible requirement to vical bubo is present in pneumonic ized susceptibility testing procedures for treat large numbers of persons. A num- plague, an aspirate (obtained with a 20- Y pestis. A process establishing criteria ber of possible therapeutic regimens for gauge needle and a 10-mL syringe con- and training measures for laboratory treating plague have yet to be ad- taining 1-2 mL of sterile saline for in- diagnosis of this disease is being under- equately studied or submitted for ap- fusing the node) may be cultured and taken jointly by the Association of Pub- proval to the Food and Drug Adminis- similarly stained (Table 1).22 lic Health Laboratories and the CDC. tration (FDA). For these reasons, the Cultures of sputum, blood, or lymph working group offers consensus recom- node aspirate should demonstrate VACCINATION mendations based on the best available growth approximately 24 to 48 hours af- The US-licensed formaldehyde-killed evidence. The recommendations do not ter inoculation. Most microbiology labo- whole bacilli vaccine was discontinued necessarily represent uses currently ap- ratories use either automated or semi- by its manufacturers in 1999 and is no proved by the FDA or an official posi- automated bacterial identification longer available. Plans for future licen- tion on the part of any of the federal systems. Some of these systems may mi- sure and production are unclear. This agencies whose scientists participated in sidentify Y pestis.12,30 In laboratories with- killed vaccine demonstrated efficacy in these discussions. Recommendations out automated bacterial identification, preventing or ameliorating bubonic dis- will need to be revised as further rel- as many as 6 days may be required for ease, but it does not prevent or amelio- evant information becomes available.

In the United States during the last turally occurring plague strains. Ex- trials of fluoroquinolones in human 50 years, 4 of the 7 reported primary perimental murine infection with plague, and they are not FDA approved 12 pneumonic plague patients died. Fa- F1-deficient variants of Y pestis have for this indication. tality rates depend on various factors shown decreased efficacy of doxycy- Chloramphenicol has been used to 47,48 including time to initiation of antibi- cline, but only 1 human case of F1- treat plague infection and has been rec- otics, access to advanced supportive deficient plague infection has been ommended for treatment of plague men- care, and the dose of inhaled bacilli. The reported.49 Russell and colleagues50 ingitis because of its ability to cross the fatality rate of patients with pneu- reported poor efficacy of doxycycline blood-brain barrier.21,34 However, human monic plague when treatment is de- against plague-infected mice, but the dos- clinical trials demonstrating the superi- layed more than 24 hours after symp- ing schedules used in this experiment ority of chloramphenicol in the therapy tom onset is extremely high.14,24,25,33 would have failed to maintain drug lev- of classic plague infection or plague men- Historically, the preferred treatment els above the minimum inhibitory con- ingitis have not been performed. It has for plague infection has been strepto- centration due to the short half-life of been associated with dose dependent mycin, an FDA-approved treatment for doxycycline in mice. In another study, hematologic abnormalities and with rare plague.21,34,35 Administered early dur- doxycycline failed to prevent death in idiosyncratic fatal aplastic anemia.35 ing the disease, streptomycin has re- mice intraperitoneally infected with 29 A number of different sulfonamides duced overall plague mortality to the 5% to 290000 times the median lethal have been used successfully in the treat- to 14% range.12,21,34 However, strepto- inocula of Y pestis.51 ment of human plague infection: sulfa- mycin is infrequently used in the United There are no controlled clinical trials thiazole,56 sulfadiazine, sulfamerazine, States and only modest supplies are comparing either tetracycline or doxy- and trimethoprim-sulfamethoxa- available.35 Gentamicin is not FDA ap- cycline to aminoglycoside in the treat- zole.57,58 The 1970 WHO analysis re- proved for the treatment of plague but ment of plague, but anecdotal case series ported that sulfadiazine reduced mor- has been used successfully36-39 and is rec- and a number of medical authorities sup- tality for bubonic plague but was ommended as an acceptable alterna- port use of this class of antimicrobials ineffective against pneumonic plague tive by experts.23,40 In 1 case series, 8 pa- for prophylaxis and for therapy in the and was less effective than tetracycline tients with plague were treated with event that streptomycin or gentamicin overall.59 In a study comparing trimeth- gentamicin with morbidity or mortal- cannot be administered.23,27,38-40,52-54 Based oprim-sulfamethoxazole with strepto- ity equivalent to that of patients treated on evidence from in vitro studies, ani- mycin, patients treated with trimetho- with streptomycin (Lucy Boulanger, mal studies, and uncontrolled human prim-sulfamethoxazole had a longer MD, Indian Health Services, Crown data, the working group recommends median duration of fever and a higher Point, NM, written communication, July that the tetracycline class of antibiotics incidence of complications.58 Authori- 20, 1999). In vitro studies and an in vivo be used to treat pneumonic plague if ties have generally considered trimeth- study in mice show equal or improved aminoglycoside therapy cannot be oprim-sulfamethoxazole a second-tier activity of gentamicin against many administered. This might be the case in choice.21,23,34 Some have recommended strains of Y pestis when compared with a mass casualty scenario when paren- sulfonamides only in the setting of pe- streptomycin.41,42 In addition, gentami- teral therapy was either unavailable or diatric prophylaxis.22 No sulfonamides cin is widely available, inexpensive, and impractical. Doxycycline would be con- have been FDA approved for the treat- can be given once daily.35 sidered pharmacologically superior to ment of plague. Tetracycline and doxycycline also other antibiotics in the tetracycline class Antimicrobials that have been shown have been used in the treatment and for this indication, because it is well to have poor or only modest efficacy in prophylaxis of plague; both are FDA absorbed without food interactions, is animal studies have included rifampin, approved for these purposes. In vitro well distributed with good tissue pen- aztreonam, ceftazidime, cefotetan, and studies have shown that Y pestis suscep- etration, and has a long half-life.35 cefazolin; these antibiotics should not be tibility to tetracycline43 and doxycy- The fluoroquinolone family of anti- used.42 cline41,44 is equivalent to that of the microbials has demonstrated efficacy in Antibiotic resistance patterns must aminoglycosides. In another investiga- animal studies. Ciprofloxacin has been also be considered in making treat- tion, 13% of Y pestis strains in Madagas- demonstrated to be at least as efficacious ment recommendations. Naturally oc- car were found to have some in vitro as aminoglycosides and tetracyclines in curring antibiotic resistance to the tet- resistance to tetracycline.45 Experimen- studies of mice with experimentally in- racycline class of drugs has occurred tal murine models of Y pestis infection duced pneumonic plague.44,50,51 In vitro rarely.4 Recently, a plasmid-mediated have yielded data that are difficult to studies also suggest equivalent or greater multidrug-resistant strain was isolated extrapolate to humans. Some mouse activity of ciprofloxacin, levofloxacin, in Madagascar.60 A report published by studies have shown doxycycline to be and ofloxacin againstY pestis when com- Russian scientists cited quinolone- a highly efficacious treatment of in- pared with aminoglycosides or tetracy- resistant Y pestis.61 There have been as- fection44,46 or prophylaxis47 against na- clines.41,55 However, there have been no sertions that Russian scientists have en-

2286 JAMA, May 3, 2000—Vol 283, No. 17 ©2000 American Medical Association. All rights reserved. MANAGEMENT OF PLAGUE USED AS A BIOLOGICAL WEAPON gineered multidrug-resistant strains of 8 Table 2. Working Group Recommendations for Treatment of Patients With Pneumonic Y pestis, although there is as yet no sci- Plague in the Contained and Mass Casualty Settings and for Postexposure Prophylaxis* entific publication confirming this. Patient Category Recommended Therapy Contained Casualty Setting Recommendations Adults Preferred choices for Antibiotic Therapy Streptomycin,1gIMtwice daily The working group treatment recom- Gentamicin, 5 mg/kg IM or IV once daily or 2 mg/kg loading dose followed mendations are based on literature re- by 1.7 mg/kg IM or IV 3 times daily† ports on treatment of human disease, Alternative choices Doxycycline, 100 mg IV twice daily or 200 mg IV once daily reports of studies in animal models, reports on in vitro susceptibility testing, Ciprofloxacin, 400 mg IV twice daily‡ and antibiotic safety. Should antibi- Chloramphenicol, 25 mg/kg IV 4 times daily§ Children࿣ Preferred choices otic susceptibility testing reveal resis- Streptomycin, 15 mg/kg IM twice daily (maximum daily dose, 2 g) tance, proper antibiotic substitution Gentamicin, 2.5 mg/kg IM or IV 3 times daily† would need to be made. Alternative choices In a contained casualty setting, a situ- Doxycycline, ation in which a modest number of pa- If Ն45 kg, give adult dosage tients require treatment, the working If Ͻ45 kg, give 2.2 mg/kg IV twice daily (maximum, 200 mg/d) group recommends parenteral antibi- Ciprofloxacin, 15 mg/kg IV twice daily‡ otic therapy (TABLE 2). Preferred par- Chloramphenicol, 25 mg/kg IV 4 times daily§ enteral forms of the antimicrobials Pregnant women¶ Preferred choice streptomycin or gentamicin are recom- Gentamicin, 5 mg/kg IM or IV once daily or 2 mg/kg loading dose followed mended. However, in a mass casualty by 1.7 mg/kg IM or IV 3 times daily† Alternative choices setting, intravenous or intramuscular Doxycycline, 100 mg IV twice daily or 200 mg IV once daily therapy may not be possible for rea- Ciprofloxacin, 400 mg IV twice daily‡ sons of patient care logistics and/or ex- Mass Casualty Setting and Postexposure Prophylaxis# haustion of equipment and antibiotic Adults Preferred choices supplies, and parenteral therapy will Doxycycline, 100 mg orally twice daily†† need to be supplanted by oral therapy. Ciprofloxacin, 500 mg orally twice daily‡ In a mass casualty setting, the work- Alternative choice ing group recommends oral therapy, Chloramphenicol, 25 mg/kg orally 4 times daily§** preferably with doxycycline (or tetra- Children࿣ Preferred choice cycline) or ciprofloxacin (Table 2). Doxycycline,†† If Ն45 kg, give adult dosage Patients with pneumonic plague will require substantial advanced medical If Ͻ45 kg, then give 2.2 mg/kg orally twice daily supportive care in addition to antimi- Ciprofloxacin, 20 mg/kg orally twice daily Alternative choices crobial therapy. Complications of gram- Chloramphenicol, 25 mg/kg orally 4 times daily§** negative sepsis would be expected, in- Pregnant women¶ Preferred choices cluding adult respiratory distress Doxycycline, 100 mg orally twice daily†† syndrome, disseminated intravascular Ciprofloxacin, 500 mg orally twice daily coagulation, shock, and multiorgan Alternative choices failure.23 Chloramphenicol, 25 mg/kg orally 4 times daily§** Once it was known or strongly sus- *These are consensus recommendations of the Working Group on Civilian Biodefense and are not necessarily approved by the Food and Drug Administration. See “Therapy” section for explanations. One antimicrobial agent should pected that pneumonic plague cases be selected. Therapy should be continued for 10 days. Oral therapy should be substituted when patient’s condition improves. IM indicates intramuscularly; IV, intravenously. were occurring, anyone with fever or †Aminoglycosides must be adjusted according to renal function. Evidence suggests that gentamicin, 5 mg/kg IM or IV cough in the presumed area of expo- once daily, would be efficacious in children, although this is not yet widely accepted in clinical practice. Neonates up to 1 week of age and premature infants should receive gentamicin, 2.5 mg/kg IV twice daily. sure should be immediately treated with ‡Other fluoroquinolones can be substituted at doses appropriate for age. Ciprofloxacin dosage should not exceed 1 antimicrobials for presumptive pneu- g/d in children. §Concentration should be maintained between 5 and 20 µg/mL. Concentrations greater than 25 µg/mL can cause monic plague. Delaying therapy until reversible bone marrow suppression.35,62 ࿣Refer to “Management of Special Groups” for details. In children, ciprofloxacin dose should not exceed 1 g/d, chlor- confirmatory testing is performed would amphenicol should not exceed 4 g/d. Children younger than 2 years should not receive chloramphenicol. greatly decrease survival.59 Clinical de- ¶Refer to “Management of Special Groups” for details and for discussion of breastfeeding women. In neonates, gentamicin loading dose of 4 mg/kg should be given initially.63 terioration of patients despite early ini- #Duration of treatment of plague in mass casualty setting is 10 days. Duration of postexposure prophylaxis to prevent plague infection is 7 days. tiation of empiric therapy could signal **Children younger than 2 years should not receive chloramphenicol. Oral formulation available only outside the United antimicrobial resistance and should be States. ††Tetracycline could be substituted for doxycycline. promptly evaluated.

Management of Special Groups the contained casualty setting receive not been studied in human or animal Consensus recommendations for spe- gentamicin (Table 2). Since streptomy- models of pneumonic plague infection. cial groups as set forth in the following cin has been associated with rare re- Therefore, the consensus recommenda- reflect the clinical and evidence-based ports of irreversible deafness in chil- tion is to administer antibiotics accord- judgments of the working group and do dren following fetal exposure, this ing to the guidelines developed for im- not necessarily correspond to FDA ap- medication should be avoided if pos- munocompetent adults and children. proved use, indications, or labeling. sible.35 The tetracycline class of antibi- Children. The treatment of choice for otics has been associated with fetal POSTEXPOSURE PROPHYLAXIS plague in children has been streptomy- toxicity including retarded skeletal RECOMMENDATIONS cin or gentamicin.21,40 If aminoglyco- growth,35 although a large case-control The working group recommends that sides are not available or cannot be study of doxycycline use in pregnancy in a community experiencing a pneu- used, recommendations for alterna- showed no significant increase in tera- monic plague epidemic, all persons de- tive antimicrobial treatment with effi- togenic risk to the fetus.66 Liver toxic- veloping a temperature of 38.5°C or cacy against plague are conditioned by ity has been reported in pregnant women higher or new cough should promptly balancing risks associated with treat- following large doses of intravenous tet- begin parenteral antibiotic treatment. ment against those posed by pneu- racycline (no longer sold in the United If the resources required to adminis- monic plague. Children aged 8 years States), but it has also been reported fol- ter parenteral antibiotics are unavail- and older can be treated with tetracy- lowing oral administration of tetracy- able, oral antibiotics should be used cline antibiotics safely.35,40 However, in cline to nonpregnant individuals.35 Bal- according to the mass casualty recom- children younger than 8 years, tetra- ancing the risks of pneumonic plague mendations (Table 2). For infants in cycline antibiotics may cause discol- infection with those associated with this setting, tachypnea would also be ored teeth, and rare instances of re- doxycycline use in pregnancy, the work- an additional indication for immedi- tarded skeletal growth have been ing group recommends that doxycy- ate treatment.29 Special measures would reported in infants.35 Chlorampheni- cline be used to treat pregnant women need to be initiated for treatment or pro- col is considered safe in children ex- with pneumonic plague if gentamicin is phylaxis of those who are either un- cept for children younger than 2 years not available. aware of the outbreak or require spe- who are at risk of “gray baby syn- Of the oral antibiotics historically cial assistance, such as the homeless or drome.”35,40 Some concern exists that used to treat plague, only trimethoprim- mentally handicapped persons. Con- fluoroquinolone use in children may sulfamethoxazole has a category C preg- tinuing surveillance of patients would cause arthropathy,35 although fluoro- nancy classification65; however, many be needed to identify individuals and quinolones have been used to treat se- experts do not recommend trimetho- communities at risk requiring postex- rious infections in children.64 No com- prim-sulfamethoxazole for treatment of posure prophylaxis. parative studies assessing efficacy or pneumonic plague. Therefore, the Asymptomatic persons having house- safety of alternative treatment strate- working group recommends that preg- hold, hospital, or other close contact gies for plague in children has or can nant women receive oral doxycycline with persons with untreated pneu- be performed. for mass casualty treatment or postex- monic plague should receive postex- Given these considerations, the work- posure prophylaxis. If the patient is un- posure antibiotic prophylaxis for 7 ing group recommends that children in able to take doxycycline or the medi- days29 and watch for fever and cough. the contained casualty setting receive cation is unavailable, ciprofloxacin or Close contact is defined as contact with streptomycin or gentamicin. In a mass other fluoroquinolones would be rec- a patient at less than 2 meters.16,31 Tet- casualty setting or for postexposure pro- ommended in the mass casualty set- racycline, doxycycline, sulfonamides, phylaxis, we recommend that doxycy- ting (Table 2). and chloramphenicol have each been cline be used. Alternatives are listed for The working group recommendation used or recommended as postexpo- both settings (Table 2). The working for treatment of breastfeeding women sure prophylaxis in this setting.16,22,29,31,59 group assessment is that the potential is to provide the mother and infant with Fluoroquinolones could also be used benefits of these antimicrobials in the the same antibiotic based on what is most based on studies in mice.51 treating of pneumonic plague infection safe and effective for the infant: genta- The working group recommends the substantially outweigh the risks. micin in the contained casualty setting use of doxycycline as the first choice Pregnant Women. It has been rec- and doxycycline in the mass casualty set- antibiotic for postexposure prophy- ommended that aminoglycosides be ting. Fluoroquinolones would be the rec- laxis; other recommended antibiotics avoided in pregnancy unless severe ill- ommended alternative (Table 2). are noted (Table 2). Contacts who de- ness warrants,35,65 but there is no more Immunosuppressed Persons. The velop fever or cough while receiving efficacious treatment for pneumonic antibiotic treatment or postexposure pro- prophylaxis should seek prompt medi- plague. Therefore, the working group phylaxis for pneumonic plague among cal attention and begin antibiotic treat- recommends that pregnant women in those who are immunosuppressed has ment as described in Table 2.

INFECTION CONTROL tients make individual isolation impos- ing bacteria such as Bacillus anthracis. Previous public health guidelines have sible, patients with pneumonic plague Moreover, Y pestis is very sensitive to the advised strict isolation for all close con- may be cohorted while undergoing an- action of sunlight and heating and does tacts of patients with pneumonic plague tibiotic therapy. Patients being trans- not survive long outside the host.72 Al- who refuse prophylaxis.29 In the mod- ported should also wear surgical masks. though some reports suggest that the ern setting, however, pneumonic plague Hospital rooms of patients with pneu- bacterium may survive in the soil for has not spread widely or rapidly in a monic plague should receive terminal some time,72 there is no evidence to sug- community,4,14,24 and therefore isola- cleaning in a manner consistent with gest environmental risk to humans in this tion of close contacts refusing antibi- standard precautions, and clothing or lin- setting and thus no need for environ- otic prophylaxis is not recommended ens contaminated with body fluids of pa- mental decontamination of an area ex- by the working group. Instead, per- tients infected with plague should be dis- posed to an aerosol of plague. In the sons refusing prophylaxis should be infected as per hospital protocol.29 WHO analysis, in a worst case scenario, carefully watched for the develop- Microbiology laboratory personnel a plague aerosol was estimated to be ef- ment of fever or cough during the first should be alerted when Y pestis is sus- fective and infectious for as long as 1 7 days after exposure and treated im- pected. Four laboratory-acquired cases hour.7 In the setting of a clandestine re- mediately should either occur. of plague have been reported in the lease of plague bacilli, the aerosol would Modern experience with person-to- United States.68 Simple clinical mate- have dissipated long before the first case person spread of pneumonic plague is rials and cultures should be processed of pneumonic plague occurred. limited; few data are available to make in biosafety level 2 conditions.31,69 Only specific recommendations regarding ap- during activities involving high poten- ADDITIONAL RESEARCH propriate infection control measures. The tial for aerosol or droplet production Improving the medical and public available evidence indicates that person- (eg, centrifuging, grinding, vigorous health response to an outbreak of to-person transmission of pneumonic shaking, and animal studies) are bio- plague following the use of a biologi- plague occurs via respiratory droplets; safety level 3 conditions necessary.69 cal weapon will require additional transmission by droplet nuclei has not Bodies of patients who have died fol- knowledge of the organism, its genet- been demonstrated.14-17 In large pneu- lowing infection with plague should be ics, and pathogenesis. In addition, im- monic plague epidemics earlier this cen- handled with routine strict precau- proved rapid diagnostic and standard tury, pneumonic plague transmission tions.29 Contact with the remains should laboratory microbiology techniques are was prevented in close contacts by wear- be limited to trained personnel, and the necessary. An improved understand- ing masks.14,16,17 Commensurate with safety precautions for transporting ing of prophylactic and therapeutic an- this, existing national infection control corpses for burial should be the same as tibiotic regimens would be of benefit in guidelines recommend the use of dis- those when transporting ill patients.70 defining optimal antibiotic strategy. posable surgical masks to prevent the Aerosol-generating procedures, such as Ex officio participants in the Working Group on Civil- 29,67 transmission of pneumonic plague. bone-sawing associated with surgery or ian Biodefense: George Counts, MD, National Insti- Given the available evidence, the postmortem examinations, would be tutes of Health, Margaret Hamburg, MD, Assistant Sec- retary for Planning and Evaluation, Robert Knouss, MD, working group recommends that, in ad- associated with special risks of trans- Office of Emergency Preparedness, Brian Malkin, Food dition to beginning antibiotic prophy- mission and are not recommended. If and Drug Administration, Stuart Nightingale, MD, Food and Drug Administration, and William Raub, PhD, Of- laxis, persons living or working in close such aerosol-generating procedures are fice of Assistant Secretary for Planning and Evaluation, contact with patients with confirmed or necessary, then high-efficiency particu- Department of Health and Human Services. Funding/Support: Funding for the development of this suspect pneumonic plague that have had late air filtered masks and negative- working group document was primarily provided by each less than 48 hours of antimicrobial treat- pressure rooms should be used as would representative’s individual institution or agency; the Johns ment should follow respiratory droplet be customary in cases in which conta- Hopkins Center for Civilian Biodefense Studies provided travel funds for 5 members of the group (Drs precautions and wear a surgical mask. gious biological aerosols, such as Myco- Ascher, Fine, Layton, and Osterholm and Mr Hauer). Further,theworkinggrouprecommends bacterium tuberculosis, are deemed a pos- Acknowledgment: We thank Christopher Davis, OBE, 71 MD, PhD, ORAQ Consultancy, Marlborough, En- avoidance of unnecessary close contact sible risk. gland; Edward B. Hayes, MD, Centers for Disease Con- withpatientswithpneumonicplagueun- trol and Prevention (CDC), Atlanta, Ga; May Chu, PhD, ENVIRONMENTAL CDC, Fort Collins, Colo; Timothy Townsend, MD, Johns til at least 48 hours of antibiotic therapy Hopkins University, Baltimore, Md; Jane Wong, MS, and clinical improvement has taken DECONTAMINATION California Department of Health, Berkeley; and Paul A. Pham, PharmD, Johns Hopkins University, for their place. Other standard respiratory drop- There is no evidence to suggest that re- review of the manuscript and Molly D’Esopo for ad- let precautions (gown, gloves, and eye sidual plague bacilli pose an environ- ministrative support. protection) should be used as well.29,31 mental threat to the population follow- The patient should remain isolated ing the dissolution of the primary aerosol. REFERENCES during the first 48 hours of antibiotic There is no spore form in the Y pestis life 1. Inglesby TV, Henderson DA, Bartlett JG, et al. An- thrax as a biological weapon: medical and public health therapy and until clinical improvement cycle, so it is far more susceptible to en- management. JAMA. 1999;281:1735-1745. occurs.29,31,59 If large numbers of pa- vironmental conditions than sporulat- 2. Henderson DA, Inglesby TV, Bartlett JG, et al. Small-

©2000 American Medical Association. All rights reserved. JAMA, May 3, 2000—Vol 283, No. 17 2289 MANAGEMENT OF PLAGUE USED AS A BIOLOGICAL WEAPON pox as a biological weapon: medical and public health 27. Poland JD, Dennis DT. Plague. In: Evans AS, Brach- PL, Friedlander A. Pathology of experimental pneu- management. JAMA. 1999;281:2127-2137. man PS, eds. Bacterial Infections of Humans: Epide- monic plague produced by fraction-1 positive and frac- 3. Centers for Disease Control and Prevention. Criti- miology and Control. New York, NY: Plenum Medi- tion-1 negative Yersinia pestis in African Green Mon- cal Biological Agents for Public Health Prepared- cal Book Co; 1998:545-558. keys. Arch Pathol Lab Med. 1996;120:156-163. ness: Summary of Selection Process and Recommen- 28. Institute of Medicine National Research Council. 50. Russell P, Eley SM, Green M, et al. Efficacy of doxy- dations. October 16, 1999. Unpublished report. Detection and measurement of biological agents. In: cycline and ciprofloxacin against experimental Yersinia 4. Perry RD, Fetherston JD. Yersinia pestis— Chemical and Biological Terrorism: Research and De- pestis infection. J Antimicrob Chemother. 1998;41: etiologic agent of plague. Clin Microbiol Rev. 1997; velopment to Improve Civilian Medical Response. 301-305. 10:35-66. Washington, DC: National Academy Press; 1999:95. 51. Russell P, Eley SM, Bell DL, Manchee RJ, Titball 5. Slack P. The black death past and present. Trans R 29. American Public Health Association. Plague. In: RW. Doxycycline or ciprofloxacin prophylaxis and Soci Trop Med Hyg. 1989;83:461-463. Benenson AS, ed. Control of Communicable Dis- therapy against experimental Y. pestis infection in mice. 6. Harris SH. Factories of Death. New York, NY: Rout- eases Manual. Washington, DC: American Public J Antimicrob Chemother. 1996;37:769-774. ledge; 1994:78, 96. Health Association; 1995:353-358. 52. Butler T. Plague. In: Strickland GT, ed. Tropical 7. Health Aspects of Chemical and Biological Weap- 30. Wilmoth BA, Chu MC, Quan TC. Identification of Medicine. Philadelphia, Pa: WB Saunders Co; 1991: ons. Geneva, Switzerland: World Health Organiza- Yersinia pestis by BBL crystal enteric/nonfermenter iden- 408-416. tion; 1970:98-109 tification system. J Clin Microbiol. 1996;34:2829- 53. Expert Committee on Plague. Geneva, Switzer- 8. Alibek K, Handelman S. Biohazard. New York, NY: 2830. land: World Health Organization; 1959. Technical Re- Random House; 1999. 31. Centers for Disease Control and Prevention. Pre- port Series 165. 9. Hughes J. Nation’s Public Health Infrastructure Re- vention of plague: recommendations of the Advisory 54. Burkle FM. Plague as seen in South Vietnamese garding Epidemics and Bioterrorism [congressional tes- Committee on Immunization Practice (ACIP). MMWR children. Clin Pediatr. 1973;12:291-298. timony]. Washington, DC: Appropriations Commit- Morb Mortal Wkly Rep. 1996;45(RR-14):1-15. 55. Frean JA, Arntzen L, Capper T, Bryskier A, Klug- tee, US Senate; June 2, 1998. 32. Titball RW, Eley S, Williamson ED, Dennis DT. man KP. In vitro activities of 14 antibiotics against 100 10. Carus WS. Bioterrorism and Biocrimes: The Il- Plague. In: Plotkin S, Mortimer EA, eds. Vaccines. Phila- human isolates of Yersinia pestis from a Southern Af- licit Use of Biological Agents in the 20th Century. delphia, Pa: WB Saunders; 1999:734-742. rican plague focus. Antimicrob Agents Chemother. Washington, DC: Center for Counterproliferation Re- 33. McCrumb FR, Mercier S, Robic J, et al. Chloram- 1996;40:2646-2647. search, National Defense University; 1998. phenicol and terramycin in the treatment of pneu- 56. Brygoo ER, Gonon M. Une epidemie de peste pul- 11. Dennis D, Meier F. Plague. In: Horsburgh CR, Nel- monic plague. Am J Med. 1953;14:284-293. monaire dans le Nor-Est de Madagascar. Bull Soc son AM, eds. Pathology of Emerging Infections. Wash- 34. Barnes AM, Quan TJ. Plague. In: Gorbach SL, Bart- Pathol Exot. 1958;51:47-66. ington, DC: ASM Press; 1997:21-47. lett JG, Blacklow NR, eds. Infectious Diseases. Phila- 57. Nguyen VI, Nguyen DH, Pham VD, Nguyen VL. 12. Centers for Disease Control and Prevention. Fa- delphia, Pa: WB Saunders Co; 1992:1285-1291. Peste bubonique et septicemique traite´ e avec succes tal human plague. MMWR Morb Mortal Wkly Rep. 35. American Hospital Formulary Service. AHFS Drug par du trimethoprime-sulfamethoxazole. Bull Soc 1997;278:380-382. Information. Bethesda, Md: American Society of Health Pathol Exot. 1972;769-779. 13. Centers for Disease Control and Prevention. Hu- System Pharmacists; 2000. 58. Butler TJ, Levin J, Linh NN, Chau DM, Adickman man plague—United States, 1993-1994. MMWR 36. Wong TW. Plague in a pregnant patient. Trop M, Arnold K. Yersinia pestis infection in Vietnam. Morb Mortal Wkly Rep. 1994;43:242-246. Doct. 1986;16:187-188. J Infect Dis. 1976;133:493-499. Rocky 14. Meyer K. Pneumonic plague. Bacteriol Rev. 1961; 37. Lewiecki EM. Primary plague septicemia. 59. WHO Expert Committee on Plague: Third Re- Mt Med J. 1978;75:201-202. 25:249-261. port. Geneva, Switzerland: World Health Organiza- 15. Kellogg WH. An epidemic of pneumonic plague. 38. Welty TK, Grabman J, Kompare E, et al. Nine- tion; 1970:1-25. Technical Report Series 447. Am J Public Health. 1920;10:599-605. teen cases of plague in Arizona. West J Med. 1985; 60. Galimand M, Guiyoule A, Gerbaud G, et al. Mul- 16. Wu L-T. A Treatise on Pneumonic Plague. Geneva, 142:641-646. tidrug resistance in Yersinia pestis mediated by a trans- Switzerland: League of Nations Health Organization; 39. Crook LD, Tempest B. Plague: a clinical review of ferable plasmid. N Engl J Med. 1997;337:677-680. 1926. 27 cases. Arch Intern Med. 1992;152:1253-1256. 61. Ryzhko IV, Shcherbaniuk AI, Samokhodkina ED, 17. Chernin E. Richard Pearson Strong and the Man- 40. Committee on Infectious Diseases. Plague. In: Pe- churian epidemic of pneumonic plague, 1910-1911. ter G, ed. 1997 Redbook. Elk Grove Village, Ill: Ameri- et al. Virulence of rifampicin and quinolone resistant J Hist Med Allied Sci. 1989;44:296-319. can Academy of Pediatrics; 1997:408-410. mutants of strains of plague microbe with Fra+ 18. Ratsitorahina M, Chanteau S, Rahalison L, Rati- 41. Smith MD, Vinh SX, Hoa NT, Wain J, Thung D, and Fra− phenotypes. Antibiot Khimioter. 1994;39: sofasoamanana L, Boisier P. Epidemiological and di- White NJ. In vitro antimicrobial susceptibilities of strains 32-36. agnostic aspects of the outbreak of pneumonic plague of Yersinia pestis. Antimicrob Agents Chemother. 62. Scott JL, Finegold SM, Belkin GA, et al. A con- in Madagascar. Lancet. 2000;355:111-113. 1995;39:2153-2154. trolled double blind study of the hematologic toxicity 19. Speck RS, Wolochow H. Studies on the experi- 42. Byrne WR, Welkos SL, Pitt ML, et al. Antibiotic of chloramphenicol. N Engl J Med. 1965;272: mental epidemiology of respiratory infections: experi- treatment of experimental pneumonic plague in mice. 113-142. mental pneumonic plague in Macaccus rhesus. J In- Antimicrob Agents Chemother. 1998;42:675-681. 63. Watterberg KL, Kelly HW, Angelus P, Backstrom fect Dis. 1957;100:58-69. 43. Lyamuya EF, Nyanda P, Mohammedali H, Mhalu C. The need for a loading dose of gentamicin in neo- 20. Aleksic S, Bockemuhl J. Yersinia and other en- FS. Laboratory studies on Yersinia pestis during the nates. Ther Drug Monit. 1989;11:16-20. terobacteriaceae. In: Murray P, ed. Manual of Clini- 1991 outbreak of plague in Lushoto, Tanzania. J Trop 64. Consensus Report of the International Society of cal Microbiology. Washington, DC: American Soci- Med Hyg. 1992;95:335-338. Chemotherapy Commission: use of fluoroquinolones ety for Microbiology; 1999:483-496. 44. Bonacorsi SP, Scavizzi MR, Guiyoule A, Amour- in pediatrics. Pediatr Infect Dis J. 1995;14:1-9. 21. Butler T. Yersinia species (including plague). In: oux JH, Carniel E. Assessment of a fluoroquinolone, 65. Sakala E. Obstetrics and Gynecology. Baltimore, Mandell GL, Bennett JE, Dolin R, eds. Principles and three ␤-lactams, two aminoglycosides, and a cycline Md: Williams & Wilkins; 1997:945. Practice of Infectious Diseases. New York, NY: Churchill in the treatment of murine Yersinia pestis infection. 66. Cziel A, Rockenbauer M. Teratogenic study of Livingstone; 1995:2070-2078. Antimicrob Agents Chemother. 1994;38:481-486. doxycycline. Obstet Gynecol. 1997;89:524-528. 22. McGovern TW, Friedlander A. Plague. In: Zajtchuk 45. Rasoamanana B, Coulanges P, Michel P, Rasolo- 67. Garner JS. Guidelines for isolation precautions in R, Bellamy RF, eds. Medical Aspects of Chemical and fonirina N. Sensitivity of Yersinia pestis to antibiotics: hospitals: Hospital Infection Control Practices Advi- Biological Warfare. Bethesda, Md: Office of the Sur- 277 strains isolated in Madagascar between 1926 sory Committee. Infect Control Hosp Epidemiol. 1996; geon General; 1997:479-502. and 1989. Arch Inst Pasteur Madagascar. 1989;56: 17:53-80. 23. Campbell GL, Dennis DT. Plague and other 37-53. 68. Burmeister RW, Tigertt WD, Overholt EL. Labo- Yersinia infections. In: Fauci AS, Braunwald E, Issel- 46. Makarovskaia LN, Shcherbaniuk AI, Ryzhkova VV, ratory-acquired pneumonic plague. Ann Intern Med. bacher KJ, et al, eds. Harrison’s Principles of Internal Sorokina TB. Effectiveness of doxycycline in experi- 1962;56:789-800. Medicine. New York, NY: McGraw-Hill; 1998: mental plague. Antibiot Khimioter. 1993;38:48-50. 69. Morse S, McDade J. Recommendations for work- 975-983. 47. Samokhodkina ED, Ryzhko IV, Shcherbaniuk AI, ing with pathogenic bacteria. Methods Enzymol. 1994; 24. Centers for Disease Control and Prevention. Pneu- Kasatkina IV, Tsuraeva RI, Zhigalova TA. Doxycy- 235:1-26. monic plague—Arizona. MMWR Morb Mortal Wkly cline in the prevention of experimental plague in- 70. Safety Measures for Use in Outbreaks in Com- Rep. 1992;41:737-739. duced by plague microbe variants. Antibiot Khimioter. municable Disease Outbreaks. Geneva, Switzerland: 25. Werner SB, Weidmer CE, Nelson BC, Nygaard GS, 1992;37:26-28. World Health Organization; 1986. Goethals RM, Poland JD. Primary plague pneumonia 48. Ryzhko IV, Samokhodkina ED, Tsuraeva RI, 71. Gershon RR, Vlahov D, Cejudo JA, et al. Tuber- contracted from a domestic cat in South Lake Tahoe, Shcherbaniuk AI, Tsetskhladze NS. Characteristics of culosis risk in funeral home employees. J Occup En- California. JAMA. 1984;251:929-931. etiotropic therapy of plague infection induced by atypi- viron Med. 1998;40:497-503. 26. Finegold MJ, Petery JJ, Berendt RF, Adams HR. cal strains of F1-phenotype plague microbe. Antibiot 72. Freeman BA. Yersinia; Pasturella; Francisella; Ac- Studies on the pathogenesis of plague. J Infect Dis. Khimioter. 1998;43:24-28. tinobacillus. In: Textbook of Microbiology. Philadel- 1968;53:99-114. 49. Davis KJ, Fritz DL, Pitt ML, Welkos SL, Worsham phia, Pa: WB Saunders Co; 1985:513-530.