The initial evaluation of patients with any form of anthrax after a biological weapon exposure should be based on careful clinical examination and laboratory testing of sites of presumed infection. A Gram stain is likely to be the only rapid method for identifying B anthracis readily available in a clinical laboratory. The organism is characterized as a gram-positive bacillus that forms long chains of vegetative cells with a “jointed bamboo-rod” appearance in culture or short chains of 2 to 4 cells in direct clinical samples. The organism can be isolated from specimen cultures of blood; skin; respiratory secretions; vesicular, pleural, or ascetic fluid; or stool, as well as from CSF (Appendix 7). Given the high rates of meningoencephalitis in children who have signs and symptoms of systemic anthrax, lumbar puncture should be performed as resources permit, unless clinically contraindicated (such as suspected increased intracranial pressure).²⁷ Gram-positive bacilli with typical morphology found on unspun peripheral blood smears or in vesicular fluid or CSF are highly suggestive of anthrax.¹⁵

Although most clinical laboratories can provide preliminary culture results, they may not be able to perform sophisticated confirmatory tests, such as polymerase chain reaction (PCR) assay on patient specimens (blood, serum, lesion swabs), lethal factor detection analysis, and immunohistochemical staining of tissue, which are available through local or state public health laboratories or the CDC. Public health laboratories in the Laboratory Response Network (www.bt.cdc.gov/lrn/), a national network of local, state, and federal public health laboratories, can identify B anthracis by standard culture methodology in addition to other methods and can perform antibacterial drug susceptibility testing. For current information on the availability and types of traditional confirmatory tests, as well as newly developed molecular tests, that can be performed by public health authorities, particularly during a biological weapon exposure, visit the CDC Anthrax Web site. (www.cdc.gov/anthrax/labs/labresearch.html).

The production of toxins and the frequent occurrence of meningoencephalitis, as well as the presence of latent spores, must be taken into account when selecting antimicrobial agents and the duration of treatment of anthrax. Most of the data used to make these recommendations are based on historical information collected before availability of many of the antimicrobial agents discussed, in vitro studies, and limited nonhuman primate studies. Treatment data that are available almost uniformly come from adult patients. For many of the antimicrobial agents discussed herein, limited pharmacokinetic data are available to make dosing recommendations for anthrax, particularly for young infants and neonates. However, systemic anthrax can be a rapidly lethal disease in humans, so in the absence of compelling, well-documented adverse safety data in children, the most effective antimicrobial agents used in the adult population also should be made available for use in children. Children may require larger doses per kilogram of body weight and more frequent dosing for certain antimicrobial agents to achieve a therapeutic exposure equivalent to that documented to be effective for adults. The recommended doses must be sufficient to ensure a therapeutic antimicrobial exposure required for life-threatening infection (ie, doses that achieve adequate exposure in more than 98% of all children treated). Antimicrobial dosing regimens, both empirical and definitive, for all anthrax clinical presentations in children and neonates are provided in Appendix 2 (cutaneous anthrax without systemic illness), Appendix 3 (systemic anthrax to include inhalation and gastrointestinal infection, without meningitis), Appendix 4 (systemic anthrax to include inhalation and gastrointestinal infection, with meningitis), and Appendix 6 (dosing in preterm and term neonates 32 to 44 weeks’ postmenstrual age).

After an exposure to B anthracis aerosolized spores, patients treated for any form of anthrax are at risk for lateoccurring inhalation anthrax from residual ungerminated spores that may remain in the lungs for several weeks or months after an inhalation event. Depending on the patient’s immune status and age, as well as the type of antimicrobial agent used, some patients recovering from severe systemic disease will develop an immune response that would provide protection against germination of any remaining spores. However, if antimicrobial therapy is initiated early in the illness, the immune response may be blunted,³⁸ thus requiring continuation of antimicrobial therapy over several weeks to clear organisms as they emerge from the spore form. There is evidence that symptomatic nonhuman primates that are bacteremic with B anthracis and treated with antimicrobial agents for at least 10 days develop an immune response and do not develop clinical disease attributable to retained spores that germinate after discontinuation of antimicrobial therapy.⁴⁸ However, no data exist in either adults or children to identify which infected people may develop protective immunity and at what time after clinical infection. No commercially available assay currently exists that can determine adequate immunity to anthrax.

Children with active infection, either cutaneous or systemic, whose original exposure source was aerosolized spores, should complete initial antimicrobial treatment, then transition to oral PEP to prevent relapse from surviving B anthracis spores within the lung that may subsequently germinate. The duration of antimicrobial therapy to prevent relapse is not well established. On the basis of incubation times of up to 43 days in humans and 58 days in animals exposed to aerosol challenge, a full 60-day course of an oral antimicrobial agent is recommended to complete treatment and prophylaxis.^8,38,49 Therefore, once the treatment course for the active infection is completed, patients should continue with antimicrobial prophylaxis so that they receive antimicrobial treatment for a total of 60 days.

Naturally occurring B anthracis is very likely to be susceptible to penicillin; although genetic material that codes for the presence of a penicillinase is present, regulation of these genes may be deficient.^50,51 Most strains are resistant to cephalosporins.⁵² Further, the naturally occurring strains that are penicillin-resistant remain susceptible to amoxicillin/clavulanate. Given the relative safety and tolerability of the penicillin-class agents in children, most pediatric experts prefer these antimicrobial agents for treatment and antimicrobial prophylaxis if pathogens are documented to be susceptible. Other antimicrobial classes, such as fluoroquinolones or tetracyclines, are likely to exhibit poorer tolerance and increased toxicity, especially during prolonged treatment.

A theoretical concern exists that the use of penicillins as single antimicrobial agents could induce β-lactam resistance, especially if the number of organisms present is high, as can occur with systemic disease. In addition, infection with multidrug-resistant B anthracis, including bioengineered strains, is a concern.⁵³ Therefore, public health authorities will monitor for the development of penicillin resistance on an ongoing basis even if strains are initially susceptible.

Given the potential severity of the disease, the fluoroquinolone class of antimicrobial agents and doxycycline (a tetracycline) are acceptable alternatives if amoxicillin and penicillin are not options, either because B anthracis susceptibility is unknown to either of these drugs or because there are other contraindications for their use in treatment. Levofloxacin and ciprofloxacin have been prospectively studied in children, with adequate pharmacokinetic data for levofloxacin use in children older than 6 months and for ciprofloxacin use in children older than 1 year, but less robust data are available for younger infants or neonates. Studies have shown a small but statistically significant increase in arthralgia in children but no documented increase in long-term bone or joint toxicity.⁵⁴ Tetracycline-class antimicrobial agents, in general, are not recommended for first-line therapy of acute bacterial infection in children younger than 8 years (with the notable exceptions of rickettsial, Ehrlichia, and Anaplasma infections) because of deposition of these compounds in teeth and bones. Although prospective, controlled data are not available, limited retrospective data suggest that children who receive up to 5 standard treatment courses of doxycycline are not likely to have clinically detectable tooth staining.^40,55

Early diagnosis and initiation of combination antimicrobial therapy is considered essential for treatment of systemic B anthracis infection. Although it is always better to definitively identify the pathogen before treatment is administered, the rapid progression of anthrax may not support this approach. Samples for laboratory testing should be collected before treatment. The AAP and CDC Pediatric Anthrax Guidance Writing Group believe that severe systemic infection survival can improve with combination therapy on the basis of theoretical considerations of different mechanisms of antimicrobial activity, drug penetration characteristics, synergy with combinations, increased likelihood of effective drug activity in case of single or multidrug resistance, and decreased likelihood of emergence of resistance. Combinations recommended include both a bactericidal antimicrobial agent and a bacterial protein synthesis inhibitor (see the section “Treatment: Considerations for Combination Antimicrobial Therapy for Anthrax” later in this article). Compared with bacteriostatic agents, bactericidal antimicrobial agents are important for killing vegetative forms emerging from spores and are especially effective, in general, in clinical outcomes when used to treat bacterial meningitis.

The high morbidity and mortality observed with anthrax is attributable primarily to 2 exotoxins (lethal toxin and edema toxin) produced by B anthracis. Protein synthesis inhibitors, such as clindamycin, may decrease toxin production, as suggested with group A streptococcal toxic shock syndrome,^56–58 which may provide an added benefit; however, this potential benefit has not been demonstrated in animal infection models or human cases of anthrax. Additional benefits may be provided by immunologic agents (such as immunoglobulin or antitoxin) that bind to the protective antigen of B anthracis to prevent attachment to host cells, neutralize toxin activity, or promote clearance of toxin complexes.

For children with overwhelming, severe disease who are not likely to survive, pediatric palliative care teams, if available, may be of value to consult with treating physicians or to provide support to the family and child.

If cutaneous anthrax is suspected, swabs of vesicular fluid, eschar tissue, or punch biopsy can be used to establish the diagnosis. Specimens should be submitted for Gram stain and culture. PCR assay and immunohistochemical staining of biopsies may be available through state or local public health laboratories or the CDC. The CDC provides a complete list of recommended specimens, including those listed previously and additionally blood for culture (or lethal toxin testing or antiprotective antigen serology) and swabs from beneath the eschar (www.cdc.gov/anthrax/labs/recommended_specimen.html).

Localized or uncomplicated cutaneous anthrax is a less-severe disease. Historically, naturally acquired infection has been successfully treated in 7 to 10 days with a single oral antimicrobial agent, such as either amoxicillin for susceptible strains or ciprofloxacin, before susceptibility testing or for penicillinresistant strains (Appendix 2). Many patients with uncomplicated cutaneous anthrax can be treated as outpatients. However, hospitalization is necessary for patients with symptoms or signs of systemic involvement, particularly those with lesions of the head or neck that may rapidly progress to include edema that can compromise the airway.

The use of surgical procedures, such as incision, in the treatment of cutaneous anthrax is usually not required, because the lesions do not contain pus. Use of surgical procedures may be associated with poor outcomes.^59,60 Treatment should focus instead on early initiation of antimicrobial agents and on local wound care. Exceptions to the general surgical contraindication include interventions to relieve airway obstruction or compartment syndrome, particularly for large or circumferential lesions of the extremities.^61,62

As a child is started on therapy, cutaneous anthrax can evolve into disease with systemic involvement, requiring the use of intravenous combination therapy (Appendices 3 and 4). Children who present with cutaneous anthrax during a biological weapon–related event also may have inhaled spores; after treatment, they should receive additional antimicrobial therapy to complete a total of 60 days of therapy.

For suspected inhalation anthrax, a chest radiograph should be performed to assess for a widened mediastinum, pleural effusions, and/or pulmonary infiltrates. In a review of 82 cases of inhalation anthrax, abnormal radiologic findings were reported in all 26 patients for whom chest radiographs were obtained; findings included pleural effusion in 69% of these patients and widened mediastinum in 54%.²⁵ Radiologic findings in children have been reported in only 2 children with inhalation anthrax, but typical findings, such as pleural effusions and widened mediastinum, were reported in both.²⁷ If there is strong suspicion of anthrax on the basis of exposure history and a questionable chest radiograph, a computed tomography (CT) scan of the chest may provide additional information to support the presumptive diagnosis.⁶³ However, during a biological weapon event, access to diagnostic radiology services may become difficult; therefore, assessing risk of anthrax exposure or clinical signs may be used initially to determine suspicion of inhalation anthrax and guide management accordingly.

Extrapolating from adult data, inhalation anthrax can have a biphasic presentation with initial improvement, followed by precipitous hemodynamic deterioration.^7,24 Because of this potential for sudden decompensation, patients who are hospitalized and treated for suspected anthrax should have careful hemodynamic monitoring, including continuous pulse oximetry and continuous cardiorespiratory monitoring for at least 24 to 48 hours, even if initial findings are reassuring.

A review of inhalation anthrax cases from 1990 through 2005 showed a significant association between pleural fluid drainage and survival.²⁵ High concentrations of lethal toxin have been detected in pleural fluid and ascites; therefore, prompt and continuous drainage by chest and/or peritoneal tube is critical. Ultrasonography of the chest may be a useful way to detect and monitor pleural fluid. Thoracotomy may be required for loculated or gelatinous effusions.

Patients with anthrax may require mechanical ventilation because of respiratory distress or imminent shock. In addition, some patients with anthrax may require respiratory support for severe edema that may occur with cutaneous lesions involving the head, neck, oropharynx, or thorax. Intubation or tracheostomy may be required to maintain the airway. Although there are significant differences between the pathophysiology of anthrax and respiratory distress syndrome, standard principles of ventilator management apply in both situations. Standard pediatric sepsis guidelines for fluids, vasopressors, blood products, and hemodynamic monitoring should also be followed (Appendix 7).

The antimicrobial treatment of inhalation anthrax should follow the guidance for severe, systemic anthrax, both for those children for whom a lumbar puncture can be performed and meningitis is unlikely (Appendix 3) and those in whom meningitis cannot be ruled out (Appendix 4). Patients presenting with signs or symptoms suggestive of pneumonia or systemic anthrax should be started on combination antimicrobial therapy as soon as possible while awaiting results of confirmatory tests. Before recognition of a B anthracis biological weapon event, the differential diagnosis for a severe systemic bacterial infection in a child usually warranted broad-spectrum antimicrobial coverage; however, cephalosporin therapy, often used empirically for community-acquired infections, is not active against B anthracis. Once the diagnosis of anthrax is confirmed, or in the midst of a biological weapon exposure, individuals presenting with illness consistent with anthrax would warrant therapy targeted more specifically to B anthracis. Given the unique drug disposition in preterm and term neonates during the first month of life, specific doses for this age group have been provided in Appendix 6.

Antimicrobial agents must penetrate into appropriate tissue sites, particularly into the CNS for children with systemic disease, especially if meningitis cannot be excluded. Meningitis and hemorrhagic parenchymal brain infection resulting from hematogenous spread must be considered in all children presenting with severe systemic anthrax.

For therapy of severe systemic anthrax in which anthrax meningitis is a consideration, treatment regimens for meningitis that include 2 bactericidal antimicrobial agents and 1 proteinsynthesis inhibitor should be followed (Appendix 4). Before susceptibility testing, and for penicillin-resistant strains, ciprofloxacin remains the preferred bactericidal antimicrobial agent, in addition to meropenem. If those antimicrobial agents are not available, levofloxacin or imipenem should be used. For susceptible strains, penicillin G (intravenous) or ampicillin are recommended, in addition to ciprofloxacin, as first-line antimicrobial agents, avoiding the unnecessary broad-spectrum activity of meropenem. Linezolid is the preferred protein-synthesis inhibitor for CNS infections on the basis of tissue penetration characteristics (Appendix 4).

However, if meningitis has been ruled out by CSF examination and/or imaging, treatment may be reduced to a combination of 2, rather than 3, antimicrobial agents with activity against B anthracis: 1 with bactericidal activity and 1 protein synthesis inhibitor (Appendix 3). If it is not possible to examine the CSF or perform imaging studies, but the clinical examination 2 to 3 days into therapy suggests that meningitis was not present by a normal neurologic examination, initial triple antimicrobial therapy can be stepped down to dual antimicrobial therapy (Appendix 3).

When meningitis is not a concern, clindamycin should be used as a protein synthesis inhibitor, rather than linezolid, primarily on the basis of concerns for safety. Doxycycline is an alternative protein synthesis inhibitor choice for treatment in all pediatric age groups for non-CNS anthrax. Rifampin is an additional option with the potential to provide synergistic antibacterial activity when used in combination with other agents if the preferred antimicrobial agents are not available or not tolerated.

After completing initial combination therapy for severe disease, all children with systemic illness, excluding meningitis, may switch to oral follow-up therapy (assuming that children can tolerate oral therapy and families are adherent to providing therapy) to complete a 14-day or longer treatment course (as dictated by clinical improvement), followed by prophylaxis, totaling 60 days (Appendix 5). Although data on the efficacy of oral follow-up therapy after short-course parenteral therapy of systemic anthrax are lacking, the recommended antimicrobial agents all have good oral absorption characteristics and have been used as oral therapy of other infections in children. Children who appear to be well, with no ongoing signs or symptoms of active infection, may be transitioned from parenteral therapy to monotherapy for the remainder of their 14 days of treatment. Those for whom there is some degree of concern about persistent deep infection or who are slower to recover may be transitioned to oral therapy that includes both a bactericidal antimicrobial agent and a protein synthesis inhibitor for the remainder of their 14 days of treatment (Appendix 5). Ciprofloxacin is the preferred antimicrobial agent for penicillin-resistant strains. Doxycycline, clindamycin, and levofloxacin should all provide adequate singledrug therapy. For penicillin-susceptible strains, amoxicillin is preferred therapy, although treating physicians should be aware of theoretical concerns for emergence of penicillin-resistance during monotherapy and be alert to the possibility of clinical deterioration as a result of resistance. In these situations, cultures should be obtained, if possible, to determine whether the organism has become resistant to amoxicillin, and the class of prescribed antimicrobial agent should be changed to those active against penicillin-resistant strains, as noted previously.

When gastrointestinal anthrax is suspected, blood cultures should be performed before starting antimicrobial therapy, and culture and PCR of ascites fluid, stool or rectal swabs, or oropharyngeal lesions, if present, are recommended. Please see www.cdc.gov/anthrax/labs/recommended_specimen.html for all recommended diagnostic specimens for gastrointestinal anthrax.

Antimicrobial therapy should follow the same guidance as that for systemic inhalation anthrax: 3 antimicrobial agents for children with possible or documented associated meningitis (Appendix 4) and 2 for those in whom meningitis can be excluded (Appendix 3).

Evaluation of CSF in the stable patient can provide laboratory evidence of meningoencephalitis as well as microbiologic confirmation of the etiology. For children too unstable to undergo lumbar puncture, CNS imaging by CT with contrast or magnetic resonance imaging with contrast should be able to document both meningeal enhancement characteristics of infection in addition to identification of hemorrhagic parenchymal lesions characteristic of anthrax meningoencephalitis. If imaging is not possible in the child with altered mental status or seizure activity, then clinical examination findings suggestive of meningitis (such as nuchal rigidity) should be presumed to be caused by meningoencephalitis and treated accordingly. If meningitis cannot be ruled out by laboratory, imaging, or physical examination, the child should be treated as though meningoencephalitis is present.

Patients with confirmed or presumptive anthrax meningoencephalitis should be treated with 3 intravenous antimicrobial agents as noted previously in “Treatment: Considerations for Combination Antimicrobial Therapy for Anthrax.” All antimicrobial agents used should have good CNS penetration, at least 2 should have bactericidal activity, and at least 1 should inhibit protein synthesis (Appendix 4). It is assumed that during a biological weapon event, limited availability of some antimicrobial agents will exist; therefore, many alternative antimicrobial agents also have been listed, some lacking documented evidence of efficacy in CNS infections.

Intravenous ciprofloxacin is recommended as the primary bactericidal antimicrobial agent for meningoencephalitis on the basis of efficacy in nonhuman primate infection models and use in anthrax cases since 2001, unless ciprofloxacin use is contra-indicated. Levofloxacin and moxifloxacin are considered equivalent alternatives to ciprofloxacin, although less experience is available with these antimicrobial agents in children. These fluoroquinolone antimicrobial agents have been shown to have adequate CNS penetration.^64,65 No reports exist to document naturally occurring fluoroquinolone-resistant strains. However, in vitro resistance can be induced, which has implications for bioengineered strains.

An antimicrobial agent from the β-lactam class with activity against B anthracis is recommended in combination with fluoroquinolones to treat systemic anthrax cases in which meningitis may be present. If antimicrobial susceptibilities are unknown or if the B anthracis strain is resistant to penicillin, meropenem is the preferred antimicrobial agent, because carbapenems are stable to the β-lactamases of B anthracis, and meropenem is approved by the FDA for use in pediatric meningitis; however, meningoencephalitis caused by B anthracis has not been prospectively studied. If meropenem is not available, imipenem/cilastatin is considered equivalent in antibacterial activity. However, imipenem/cilastatin is associated with an increased risk of seizures when used in the treatment of meningitis, presumably by lowering the seizure threshold,^66,67 and should be used with caution in patients with suspected meningitis. Pharmacokinetic data are available for use in children. Limited data are available for doripenem pharmacokinetics or clinical efficacy in children, but if meropenem and imipenem are not available, data from experimental animal models of meningitis suggest similar efficacy.⁶⁸ If the isolated B anthracis strain is susceptible to penicillin, penicillin G or ampicillin are preferred antimicrobial agents given their long history of use, safety profile, and narrower antimicrobial spectrum. However, as noted earlier, most strains contain a β-lactamase that is usually not associated with clinically relevant resistance but represents a potential for development of resistance to penicillin or ampicillin while on therapy. Emergence of resistance to penicillin would require treatment with meropenem or another carbapenem (Appendix 4).

Vancomycin has demonstrated in vitro activity against naturally occurring strains but has limited human clinical treatment data in anthrax. Vancomycin is bactericidal, achieves adequate CNS concentrations, has been used extensively in pediatric meningitis, and may be used as an alternative antimicrobial agent if β-lactam antimicrobial agents are not available (Appendix 4).⁶⁹

At least 1 antimicrobial agent that inhibits protein synthesis is recommended for inclusion in the antimicrobial regimen to reduce production of exotoxins. Linezolid is recommended as the first-line protein synthesis inhibitor. It is recommended over clindamycin when meningitis is suspected or cannot be excluded, as it is likely to provide better CNS penetrance, although prospective, controlled data on the treatment of CNS infections with either antimicrobial agent do not exist.^70–73 Linezolid toxicity must be taken into consideration. Peripheral and optic neuropathy have been reported in both adults and children receiving prolonged courses of linezolid.^74–76 Myelosuppression is a potential adverse effect more commonly seen when linezolid is used for more than 14 days, but it is reversible and can be managed with marrow-stimulating agents.⁷⁷ Linezolid should be used with caution in patients with preexisting myelosuppression. If patients have contraindications to linezolid use or it is not available, clindamycin is an acceptable alternative. Rifampin, although not a protein synthesis inhibitor, has been widely used for antibacterial synergy with a primary drug and could be used in this capacity if neither linezolid nor clindamycin are available.

Chloramphenicol historically has been used to successfully treat anthrax⁷⁸; it is a protein synthesis inhibitor with excellent CNS penetration. The intravenous form is not widely available in the United States, and the oral formulation is no longer available. Intravenous chloramphenicol would be an acceptable alternative if linezolid, clindamycin, and rifampin are not available for use. Doxycycline should not be used as first-line therapy if meningitis is suspected because of variable CNS penetration compared with fluoroquinolones and β-lactams, but it may be effective if other preferred antimicrobial agents are not available or are not tolerated.

Duration of initial intravenous combination therapy is determined by meningitis risk and clinical response to treatment. With mortality rates for meningoencephalitis approaching 100%, no prospective data exist on which to base recommendations for an effective treatment duration.⁶⁹ For children with meningoencephalitis or in children for whom meningitis cannot be ruled out, combination therapy is recommended for a minimum of 2 weeks, depending on the severity of disease and presence of brain parenchymal hemorrhagic lesions. Intravenous therapy should continue until the patient is clinically stable, with all clinical signs and symptoms and laboratory and imaging data documenting resolution of inflammation, even if required for 3 to 6 weeks. As with other forms of anthrax secondary to a biological weapon exposure, ongoing oral PEP therapy should continue until antimicrobial therapy has been administered for a total of 60 days.