In 2014 CDC published updated national guidelines on the clinical management of anthrax [1–3]. These practice guidelines outline the use of combination intravenous (IV) antimicrobials in the treatment of systemic anthrax. Treatment regimens are defined by clinical manifestations of disease and involve two sets of recommendations – one for treatment of individuals with confirmed, or suspected anthrax meningitis (or for instances where meningitis cannot be ruled out), and one for individuals in whom meningitis has been ruled out. Both recommendations call for intravenous treatment that includes at least one bactericidal agent combined with a protein synthesis inhibitor; in the setting of meningitis, the addition of a third, preferably bactericidal, antimicrobial agent is recommended [2]. Bactericidal agents with good blood-brain barrier penetration recommended for the treatment of suspected or confirmed anthrax meningitis include quinolones, carbapenems, and if the isolate is susceptible, β-lactams such as penicillin G and ampicillin [2].

Regarding protein synthesis inhibitors, options include linezolid as the preferred agent because of high central nervous system (CNS) concentrations while acceptable alternatives include clindamycin, rifampin and, if linezolid, clindamycin, and rifampin are unavailable, chloramphenicol. For non-meningitis, systemic anthrax (which includes inhalation anthrax and other forms of anthrax with systemic involvement), recommendations also include vancomycin as a potential bactericidal agent and doxycycline as an alternative protein synthesis inhibitor [2]. Combination intravenous antimicrobials are recommended for a minimum of two weeks for patients with systemic anthrax. This is consistent with findings from a systematic review of inhalation anthrax cases from the pre-antimicrobial era (1900) to 2005, where survivors were significantly more likely to have received a multi-drug antimicrobial regimen [4]. Cutaneous anthrax without systemic involvement historically is associated with a substantially lower mortality and thus oral monotherapy is considered adequate [2]. Combination antimicrobial therapy for cutaneous anthrax is recommended for cases with systemic signs or symptoms (e.g., systemic inflammatory response syndrome), or cutaneous disease involving the head/neck, or associated with significant edema [2]. These recommendations were developed as Best Practices recommendations for the treatment of an individual (or small number of) patient(s) [2] (herein referred to as Best Practices guidance). However, an anthrax mass casualty incident has the potential to produce hundreds of thousands of patients with systemic anthrax disease. Given that the current recommendations advise the use of multiple intravenous antimicrobials in the treatment of systemic anthrax, a mass casualty event involving large numbers of affected individuals could result in challenges in adequately meeting intravenous antimicrobial needs. As part of ongoing efforts to plan for large-scale bioterrorism events, CDC is developing clinical guidance for an anthrax mass casualty incident. To inform such guidance, a systematic review of antimicrobials utilized for the treatment of anthrax was conducted to summarize what is known about historical survival with combination antimicrobial treatment. This includes an assessment of the types and combinations of antimicrobials used to treat anthrax as well as information on the duration of anthrax treatment.

The systematic review was conducted to identify available data regarding: 1) combining a bactericidal antimicrobial with a protein synthesis inhibitor in the treatment of systemic anthrax disease, 2) the use of three antimicrobials for the treatment of anthrax meningitis, and 3) the duration of parenteral antimicrobial therapy for systemic anthrax.

One hundred forty-nine individual cases, identified in 98 articles (Figure 2; Appendix Table A), received an antimicrobial recommended in the Best Practices guidance [2]; had evidence of cutaneous disease (with systemic signs or symptoms and/or complicated by meningitis), GI anthrax, inhalation anthrax, injection anthrax, or primary anthrax meningitis (CNS disease without any other route of infection noted); and a defined outcome (survived or died). The cases were reported in the literature between 1945 and 2014. Over 80% of the cases were reported from the following 7 countries (from which ≥5 cases were reported): India (n=30), U.S. (n=27), Turkey (n=24), Iran (n=24), U.K. (n=9), Lebanon (n=6), and Zimbabwe (n=5). One hundred forty-five cases had recorded sex; 42 (29%) were female and 103 (71%) were male. Age was recorded in 145 cases; 111 (77%) were 18 years and older and 34 (23%) 17 years and younger. Appendix Table A presents the evidence table generated from this systematic review.

During an anthrax mass casualty incident, it is anticipated that individuals will present with inhalation anthrax from breathing in spores, cutaneous anthrax from physical contact with spores, and cases may be complicated by anthrax meningitis. Inhalation anthrax, anthrax meningitis, and cutaneous disease involving the head/neck or surrounded by significant edema have been reported to be associated with high mortality [4, 9, 10]. It is important to understand the role of combination antimicrobial regimens (in terms of number of agents and type of agents) for the treatment of systemic anthrax to ensure judicious and optimal use of antimicrobials during a mass casualty incident. On the other hand, for uncomplicated cutaneous anthrax, oral monotherapy should suffice and therefore this is less of a concern for mass event planning [2].

Before evaluating treatment regimens, we sought to understand the types of anthrax that have historically been most severe: inhalation anthrax and anthrax meningitis. In an earlier systematic review of inhalation anthrax [4], the overall mortality was 85%. Of note, that systematic review includes cases from 1900 through 2005, and thus includes inhalation anthrax cases from prior to the antimicrobial era. However, even among the 32 patients treated with antimicrobials, mortality remained high, at 75%. Consistent with these findings, our systematic review of anthrax cases treated with antimicrobials also found a high mortality: 70% among the 30 inhalation cases. Furthermore, in one review of 70 cases of anthrax meningoencephalitis, mortality was reported to be 94% [10]. In our systematic review, meningitis complicated 77 of 149 cases (52%) of systemic anthrax and 81% of these individuals died.

Combination bactericidal and protein synthesis inhibitor therapy may be appropriate in severe anthrax disease, particularly anthrax meningitis, as noted in the Best Practices guidelines [2]. In another systematic review focused on inhalation anthrax cases, including cases from the pre-antimicrobial era, previous investigators showed that receipt of a multi-drug regimen was associated with decreased mortality when compared to individuals that did not receive such treatment (which included single drug treatment as well as no antimicrobial treatment) [4]. In this review of antimicrobial therapy for systemic anthrax, our findings did not contradict current national guidelines, finding that among individuals treated with one or more Best Practices [2] antimicrobials for systemic anthrax, 45% of those receiving overlapping bactericidal-protein synthesis inhibitor therapy survived versus 28% of those without any overlapping bactericidal-protein synthesis inhibitor treatment. Much of anthrax pathogenesis is toxin mediated, thus the finding of an added benefit of protein synthesis inhibition is consistent with known mechanisms of disease. In vitro data confirm that a protein synthesis inhibitor (linezolid) is more effective than a bactericidal antimicrobial (ciprofloxacin) in attenuating B. anthracis toxin production [11]. In addition, the use of a protein synthesis inhibitor (clindamycin) has been shown to improve clinical outcomes of invasive group A streptococcal infections in humans, an infection that is known to be associated with bacterial toxin production [12].

While the Best Practices guidance [2] calls for combination triple antimicrobial therapy for meningitis, there was a paucity in our dataset of cases receiving 3 or more antimicrobials for meningitis. Among the subset of individuals with meningitis who received, for some period of time, combination bactericidal and protein synthesis inhibitor therapy, a higher percent survival was observed for those receiving 3 or more antimicrobials versus those receiving only 2 antimicrobials. These recommendations are consistent with a review of the management of anthrax meningoencephalitis by Sejvar and colleagues, in which they recommend the use of a fluoroquinolone coupled with 1 or 2 other antimicrobials with good CNS penetration [13]. Whether the improved outcomes associated with the use of three or more antimicrobials over the course of treatment relates to greater likelihood of at least one antimicrobial crossing the blood-brain barrier, versus a synergistic antimicrobial effect, or some other undefined factor, remains to be elucidated.

One limitation of these data relates to the fact the reports described in this manuscript were from the English literature only, though we did review nearly 6,700 titles and abstracts, conducted nearly 400 full text reviews, and the cases included in this analysis were from a very wide geographic distribution. In addition, several limitations in the available data impede our ability to know with certainty whether improved survival was due to combination antimicrobial therapy, advances in critical care medicine, or other confounding factors, such as shorter times to treatment or ancillary treatments such as intravenous fluid. As all the data was observational in nature, its quality could be considered, at best, low [5]. The reports we reviewed contained diverse and varying levels of information, requiring clinical interpretation. As noted, duration of treatment was not captured in a uniform manner, and this was the case with other treatment related variables, such as timing of antimicrobial therapy and route of administration. And, the definitions used for systemic anthrax and anthrax meningitis were based on interpretation of reported signs, symptoms, and epidemiologic, radiologic, pathologic and laboratory parameters. In addition, the data analyzed include cases spanning a 70-year period, during which antimicrobial options and intensive care have changed tremendously. The primary bactericidal antimicrobials recommended in the Best Practices guidance (fluoroquinolones) [2] were only used in 22% of the severe cases of anthrax. Another first-line antimicrobial class (carbapenems) was used in only 3% of cases. The most frequently used antimicrobials for severe illness were penicillins (88% of cases) and chloramphenicol (16% of cases), the latter of which is unlikely to be used in the current treatment of anthrax in the United States. In addition, these cases are heterogeneous, representing treatment in widely different locations, with differing health care capacities, and cases could have had widely varying severities of illness.

While data on total duration of parenteral therapy were sparse, they are consistent with the concept of a minimum of two weeks of intravenous therapy. Therapy beyond two weeks could not be evaluated.

The limitations and confounding factors described above must be taken into account when considering the results presented. And, as additional data become available regarding the use of newer, combination, and broader spectrum antimicrobial therapies, and in situations of severe antimicrobial shortages, treatment guideline developers may need to revisit whether fewer antimicrobials, perhaps used for a shorter time, could be effective for treating severe anthrax disease. These considerations are particularly relevant in the setting of a large-scale anthrax mass casualty incident when resource constraints may exist.