Sentinel Recognition and Referral of Select Agents

Sentinel clinical laboratories detect and refer suspected bioterrorism agents. They do not complete identification or susceptibility testing. As soon as one of these organisms is suspected, further work is performed in a certified Class II biological safety cabinet using BSL-3 practices and appropriate PPE. Automated identification systems, including MALDI-TOF, are not used for suspected isolates because they generate aerosols and have misidentified these organisms. Full identification, aerosol-generating confirmatory work, and any occupational-vaccination decision belong to the Laboratory Response Network (LRN) reference or public-health laboratory.^1,2

Bacillus anthracis

Bacillus anthracis is a large, spore-forming, gram-positive/gram-variable, boxcar-shaped, non-motile rod (nonmotility distinguishes it from nearly all other Bacillus spp.). Its glutamyl polypeptide capsule blocks phagocytosis, and a tripartite toxin (protective antigen, edema factor, lethal factor) produces the clinical manifestations; cell-wall peptidoglycan itself drives a severe intravascular inflammatory response that can progress to shock.

Because B. anthracis spores are a recognized bioterrorism agent, all suspected material is handled in a biological safety cabinet. Surface decontamination uses a fresh 1:10 (10%) household-bleach solution, about 5,000 ppm available chlorine. Organic load inactivates bleach, so visible soil is removed first and wet contact time is required. Phenolics are not sporicidal at usual working dilutions and are not used as the primary surface disinfectant. Routine clinical laboratories do not attempt a full workup of a suspected isolate. Colonies on blood agar are flat, irregular (“Medusa head”), off-white, ground-glass, tenacious when probed (standing up like beaten egg white), and typically non-hemolytic. Nonmotility on a hanging-drop or semisolid-media test distinguishes B. anthracis phenotypically from the otherwise similar-appearing B. cereus. Sentinel rule-out uses Gram stain, catalase, hemolysis, and motility; an isolate that remains non-hemolytic, catalase-positive, and nonmotile is referred.²

B. cereus causes two distinct food-poisoning syndromes: an emetic, preformed-toxin illness classically linked to reheated fried rice and a diarrheal illness caused by a heat-labile enterotoxin produced after ingestion. It also causes opportunistic infections, including bacteremia, endocarditis, and meningitis, particularly in people who use intravenous drugs or have indwelling catheters. Diagnosis of B. cereus food poisoning requires quantitative culture of the implicated food (≥10⁵ CFU/g) rather than stool culture because the organism is also a normal low-level gut inhabitant.

Brucella species

Brucella spp. are small, non-motile, aerobic, catalase/oxidase-positive gram-negative coccobacilli. These intracellular pathogens of macrophages and neutrophils evade phagosomal killing and disseminate through the reticuloendothelial system, forming granulomas in the liver, spleen, bone marrow, genitourinary tract, and lung. The principal human pathogens are B. melitensis (sheep/goats/camels), B. abortus (cattle), B. suis (swine), and B. canis (dogs). They are acquired through inhalation, direct contact with infected animal material, or consumption of unpasteurized dairy products. In the United States, imported unpasteurized goat cheese remains an ongoing route. Brucellosis produces a nonspecific, often diurnal or undulant fever syndrome with an incubation period of roughly 1-4 weeks. It is among the most frequently reported laboratory-acquired infections, which drives strict handling precautions.

Culture of blood, bone marrow, or abscess material is slow. Colonies may be barely visible after 48 hours even in a CO₂-enriched atmosphere, and automated blood-culture systems should be held 10 days with consideration of terminal subculture (ASM/APHL 2025); most positives flag within 5-7 days. All manipulation of suspected Brucella isolates must occur in a biosafety cabinet. Species-level identification based on CO₂ requirement, oxidative metabolic profile, urease kinetics, dye sensitivity, and phage lysis should be referred to an LRN reference laboratory. It should not be attempted with MALDI-TOF or automated identification systems because both risk generating an infectious aerosol. HHS/USDA removed B. abortus, B. melitensis, and B. suis from the select-agent list effective January 16, 2025. ASM/APHL 2025 sentinel guidance still requires rule-out testing and LRN referral because the biosafety risk is unchanged. Serologic tests, including serum agglutination, indirect Coombs, and ELISA, supplement culture, particularly in chronic or complicated presentations. Each has recognized false-negative and false-positive limitations and cross-reactivity with Francisella tularensis.

Routine susceptibility testing is not recommended; treatment failures are not generally attributed to acquired antimicrobial resistance.^3,4

Francisella tularensis

F. tularensis is a strictly aerobic, tiny, cysteine-requiring gram-negative coccobacillus with an extremely low infectious dose. It is found in North American rodents and rabbits and their tick vectors. Transmission occurs through tick or deerfly bite, direct contact with infected animals, inhalation, or ingestion of undercooked contaminated meat. Depending on the route of entry, infection produces ulceroglandular, glandular, oculoglandular, oropharyngeal, pneumonic, or typhoidal disease after a 1-21-day incubation period. It is a Category A bioterrorism agent.

Because of the extreme infectivity risk, specimens from suspected cases must be handled in a biosafety cabinet, and clinicians must notify the laboratory in advance. Most clinical laboratories act only as sentinel laboratories, making a presumptive identification and forwarding the isolate to a public-health/reference laboratory or the CDC for confirmation. The organism grows slowly over 2-4 days on cysteine-supplemented media such as chocolate, Thayer-Martin, or buffered charcoal yeast extract agar. It produces small, blue-gray to white, mucoid colonies with a small zone of α-hemolysis on blood-containing agar and is weakly catalase-positive, oxidase-negative, and biochemically inert. MALDI-TOF and automated identification systems are specifically discouraged for suspected isolates because of aerosol and misidentification risk. Serology supports diagnosis with a tube agglutination titer ≥1:160, a microagglutination titer ≥1:128, or a fourfold rise between acute and convalescent sera. Lower titers have recognized cross-reactivity with Brucella antibodies.

All isolates are β-lactamase-positive. Because of handling risk and the absence of reported acquired resistance, routine susceptibility testing is not recommended even though a CLSI reference broth microdilution method exists.⁵

Yersinia pestis

Yersinia pestis is a nonmotile, slow-growing facultative gram-negative coccobacillus in the order Enterobacterales and the agent of plague. Direct smears may show plump rods, 0.5 × 1-2 µm, singly or in pairs. Wright or Giemsa stains of blood or tissue can reveal bipolar (“safety-pin”) staining that a Gram stain often does not. Sentinel laboratories rule out the organism; they do not complete identification or susceptibility testing. The current ASM/APHL sentinel protocol for Y. pestis remains the March 2016 revision; anthrax and Brucella protocols have later dates.

On sheep-blood agar, colonies are usually too small to see as individuals at 24 hours (pinpoint or no growth at 35 °C). At 48 hours they are 1-2 mm, gray-white to slightly yellow, opaque, and nonhemolytic; older cultures may take on a fried-egg appearance. Growth on MacConkey or EMB is slow and non-lactose-fermenting and may be invisible at 24 hours. Optimal growth is at 25-28 °C; growth at 35-37 °C is slower.

An isolate from blood, sputum, or lymph-node aspirate that is catalase-positive and oxidase-, indole-, and urease-negative, with the colony morphology above, cannot be ruled out as Y. pestis. Urease separates it from Y. enterocolitica and Y. pseudotuberculosis, which are urease-positive. Automated identification systems and MALDI-TOF must not be used: they have misidentified Y. pestis as Y. pseudotuberculosis, Shigella, H₂S-negative Salmonella, Acinetobacter, or Pseudomonas.

When Y. pestis cannot be ruled out, the laboratory reports that the isolate is referred, notifies the designated LRN reference laboratory immediately, and transfers the isolate (and original specimens) by prior arrangement. Infection prevention is notified inside the hospital. If the rule-out criteria are not met, identification proceeds by established methods for other Yersinia species.⁶