Cross-Contamination of Clinical Specimens with Bacillus anthracis During a Laboratory Proficiency Test --- Idaho, 2006

On July 18, 2006, the Utah Department of Health notified epidemiologists at the Idaho Department of Health and Welfare that Bacillus anthracis, the causative agent for anthrax, had been isolated from a patient. On the same day, the Idaho epidemiologists were notified by the Idaho Bureau of Laboratories of a specimen from a second patient received for anthrax testing. The two reports resulted briefly in alerts to the Federal Bureau of Investigation (FBI) and precautionary treatment of one of the patients for anthrax. Subsequent investigation revealed that, during July 2006, the Idaho Bureau of Laboratories had been conducting a sentinel laboratory proficiency testing exercise among Idaho's hospital laboratories. The exercise included specimens with the Sterne strain of B. anthracis, a nonvirulent strain. Subsequent laboratory testing of the two patient isolates detected the Sterne strain of B. anthracis; neither patient had signs or symptoms consistent with B. anthracis infection. Further investigation revealed that the Idaho hospital laboratories that tested the two specimens had been conducting the laboratory proficiency testing exercise simultaneously, but the Idaho epidemiologists were not aware of the exercise. The two specimens had become cross-contaminated with B. anthracis in the laboratories. The findings in this report underscore the need to follow proper laboratory practices to minimize cross-contamination. In addition, to guard against false reports of anthrax, public health epidemiologists who monitor reportable diseases should be notified of upcoming proficiency testing of high-priority bioterrorism agents.

Case Reports

Case 1. On the afternoon of July 18, the Utah Department of Health notified epidemiologists at the Idaho Department of Health and Welfare that B. anthracis had been isolated from a wound culture from an Idaho patient. The specimen had been forwarded by an Idaho hospital laboratory to a commercial reference laboratory in Utah, where initial morphologic testing results prompted transfer of the isolate for additional analysis to the Utah Public Health Laboratories, in accordance with Laboratory Response Network (LRN) sentinel laboratory protocols.*

Following LRN protocols, Utah Public Health Laboratories demonstrated that the isolate 1) was a gram-positive, spore-forming bacillus; 2) formed tenacious, nonhemolytic colonies on blood agar; 3) was nonmotile; and 4) was positive for lysis by gamma phage and positive for B. anthracis cell wall by direct fluorescent antibody (DFA). However, DFA for B. anthracis capsule was negative, an indication of nonvirulence. Polymerase chain reaction (PCR) results were inconclusive for wild-type B. anthracis DNA and were more consistent with a plasmid-cured strain (two of three signatures present), such as the Sterne strain, a nonvirulent strain used for animal vaccinations. Per existing laboratory protocols for biologic threat agent isolates from sentinel laboratories, the FBI was notified of the result. On July 19, Utah Public Health Laboratories consulted with CDC's Bioterrorism Rapid Response and Advanced Technology laboratory, which asked to receive the isolate for additional analysis.

On July 18, the day of the initial identification of B. anthracis, the local Idaho public health district initiated a public health investigation, which revealed the specimen was taken from a dog-bite wound on the face of a male health-care worker aged 36 years in southern Idaho. The patient and health-care provider were notified of the anthrax test result by investigators the same day.

The patient had been prescribed amoxicillin/clavulanate potassium to treat the dog-bite wound, which was unremarkable and healing. The patient reported no risk factors for cutaneous anthrax infection. Although B. anthracis can be carried by dogs (1), nothing indicated that the dog, a household pet living in an urban area, had been exposed to anthrax. The health-care provider noted no clinical indications of cutaneous anthrax or any other infection in the patient during initial examination. On July 24, the isolate was confirmed by CDC as a Sterne strain of B. anthracis by multiple-locus variable-number tandem repeat analysis.

Case 2. In the late afternoon of July 18, the same day case 1 was reported, the Idaho Bureau of Laboratories notified epidemiologists at the Department of Public Health and Welfare of a specimen received for anthrax rule-out testing using LRN protocols from an Idaho hospital laboratory in a geographic area different from the area where the hospital laboratory in case 1 was located. Because two possible anthrax cases had been reported on the same day, the possibility of a bioterrorist act was considered briefly; however, the PCR results of case 1 suggesting a Sterne strain made this possibility seem less likely.

Investigation by the local Idaho public health district revealed the specimen was from a male sculptor aged 45 years living in northern Idaho. The specimen was taken from an incision made during removal of wire implanted in the patient's hand in March 2006 to repair a table-saw wound (the incision had become inflamed). Although anthrax spores can survive for years in soil and the patient had contact with clay while sculpting, no clinical indications of cutaneous anthrax were detected. The patient's surgeon was notified of the positive anthrax test result by the local Idaho public health district and referred the patient to an infectious-disease specialist, who initiated treatment for anthrax as a precaution. The Idaho Department of Health and Welfare informed the FBI of the circumstances of this second possible case.

The isolate was recovered at the Idaho hospital laboratory after incubation of the specimen by broth enrichment. Incubation of the original specimen on solid media had yielded a coagulase-negative Staphylococcus species. The isolate received by the Idaho Bureau of Laboratories on July 18 was phenotypically consistent with B. anthracis. DFA and PCR results were identical to those reported for the isolate in case 1. On July 19, the Idaho Bureau of Laboratories conferred with CDC, and a decision was made that forwarding the isolate to CDC was not necessary.

Follow-Up Investigation

On July 12, the Idaho Bureau of Laboratories had sent proficiency test samples containing the Sterne strain to the two Idaho source hospital laboratories and other Idaho sentinel laboratories. On July 19, the Idaho Bureau of Laboratories conducted telephone interviews with laboratorians at the two Idaho hospital laboratories. For case 1, the laboratory manager used laboratory information system data to construct a partial timeline that indicated the patient specimen arrived and was set up for culture on the final day of work on the proficiency testing sample. No additional details were available. For case 2, the hospital laboratorian who worked on the patient specimen and proficiency sample indicated that both were set up for culture in the same biosafety cabinet within minutes of each other.

The precise mechanism of cross-contamination could not be ascertained for either case. The lyophilized proficiency sample might have become aerosolized during processing, or materials used in setting up the patient specimen might have become contaminated, possibly through incomplete sterilization, within the biosafety cabinet. In both cases 1) the hospital laboratories received proficiency samples containing the Sterne strain; 2) the Sterne strain was isolated from the patient specimens only after broth enrichment of wound specimens, a practice generally not considered appropriate for nonsterile sites such as wounds; and 3) both patient specimens and proficiency samples were in the same laboratory area at the same time, including, in case 2, in the same biosafety cabinet.

Reported by: J Bartschi, MHE, C Hahn, MD, L Tengelsen, PhD, DVM, W DeLong, MS, Idaho Dept of Health and Welfare; J Lee, Panhandle District Health, Coeur d'Alene, Idaho. S Mottice, PhD, J Coombs, Utah Dept of Health. R Meyer, PhD, CDC.

Editorial Note:

This report is the first to describe cross-contamination of clinical specimens with B. anthracis during laboratory proficiency sample testing. The Idaho Bureau of Laboratories regularly conducts proficiency testing surveys with participating sentinel laboratories. Proficiency testing is intended to improve the ability of a sentinel laboratory to either rule out the presence of potential category A agents^† or refer the isolates to the state laboratory for confirmation. The Idaho sentinel laboratory proficiency testing exercise described in this report included the Sterne strain of B. anthracis, which is used widely as a live veterinary vaccine and by research laboratories to produce crude toxins (3,4). The Sterne strain lacks a 60 megadalton plasmid, pX02, which mediates the formation of a capsule, rendering the strain relatively avirulent, although cases of vaccine-related illness have been reported in cattle (5). The negative capsule DFA was an indicator that the strain was not virulent and was crucial to ruling out the more virulent form of B. anthracis in the two Idaho cases.

Cross-contamination of specimens and cultures is not a rare event in clinical laboratories (6--8). However, because of heightened awareness of the potential significance of gram-positive, spore-forming bacilli in recent years, clinical laboratories are less likely to ignore such isolates. This report underscores the need to use good laboratory practices to minimize cross-contamination of specimens during set up and analysis, not only when dealing with proficiency samples, but during daily operation with patient specimens. Recommended practices include opening one sample at a time, carefully handling samples to avoid splashing or aerosolization, changing gloves between samples, immediately cleaning up spills, disinfecting the work area often, and properly using biosafety cabinets. To reduce the likelihood of cross-contamination, laboratory workers should avoid practices such as inappropriate use of mixing devices (i.e., vortexers, blenders, and homogenizers) or fixing slides or sterilizing inoculating loops containing infectious material over an open flame (9).

Many public health laboratories designated as LRN reference laboratories have developed programs to evaluate the readiness of sentinel laboratories in their jurisdictions to rule out bioterrorism agents, including the practice of sending proficiency samples. Recently, the College of American Pathologists enhanced its Laboratory Preparedness Survey to include select agent-exempt strains of category A and B agents, including the Sterne strain of B. anthracis. Reference laboratories, whether or not they are LRN members, and epidemiologists should be aware that these strains are being distributed.

Public health epidemiologists who monitor reportable diseases in the jurisdictions where laboratory proficiency testing of high-priority bioterrorism agents will take place should be notified when testing is scheduled so they can be alert for potential cross-contamination. However, vigilance for biologic threats must always be maintained; public health responders should never assume that laboratory reports of positive test results are linked to proficiency testing events.

References

1. Langston C. Postexposure management and treatment of anthrax in dogs---executive councils of the American Academy of Veterinary Pharmacology and Therapeutics and the American College of Veterinary Clinical Pharmacology. AAPS J 2005;7:E272--3.
2. CDC. Emergency preparedness and response: bioterrorism agents/diseases. Atlanta, GA: US Department of Health and Human Services, CDC; 2008. Available at http://www.bt.cdc.gov/agent/agentlist-category.asp.
3. Hambleton P, Carman JA, Melling J. Anthrax: the disease in relation to vaccines. Vaccine 1984;2:125--32.
4. Turnbull PC, Broster MG, Carman JA, Manchee RJ, Melling J. Development of antibodies to protective antigen and lethal factor components of anthrax toxin in humans and guinea pigs and their relevance to protective immunity. Infect Immun 1986;52:356--63.
5. Brossier F, Mock M, Sirard JC. Antigen delivery by attenuated Bacillus anthracis: new prospects in veterinary vaccines. J Applied Microbiol 1999;87:298--302.
6. Segal-Maurer S, Kreiswirth BN, Burns JM, et al. Mycobacterium tuberculosis specimen contamination revisited: the role of laboratory environmental control in a pseudo-outbreak. Infect Control Hosp Epidemiol 1998;19:101--5.
7. Jasmer RM, Roemer M, Hamilton J, et al. A prospective, multicenter study of laboratory cross-contamination of Mycobacterium tuberculosis cultures. Emerg Infect Dis 2002;8:1260--3.
8. Katz KC, McGeer A, Low DE, Willey BM. Laboratory contamination of specimens with quality control strains of vancomycin-resistant enterococci in Ontario. J Clin Microbiol 2002;40:2686--8.
9. Carlberg DM, Yeaman MR. Biosafety in the teaching laboratory. In: Fleming DO, Hunt DL, eds. Biological safety principles and practices. Washington, DC: ASM Press; 2006.

* Established in 1999, LRN is a network of local, national, and international reference and sentinel laboratories equipped to respond rapidly to acts of terrorism (biologic or chemical), emerging infectious diseases, and other public health emergencies. Sentinel laboratories (e.g., private clinical or hospital-based laboratories with the capacity and training to recognize potential agents of bioterrorism and rule them out), using American Society of Microbiology protocols, perform presumptive identification of possible biologic terrorism agents and submit isolates to reference laboratories for confirmatory testing. Additional information is available at http://www.bt.cdc.gov/lrn.

^† Category A agents are easily disseminated or transmitted agents that can result in high mortality rates, have potential for major public health impact, can cause social disruption, and require special preparedness actions. Diseases caused by category A agents include anthrax, botulism, plague, smallpox, tularemia, and viral hemorrhagic fevers (2).

All MMWR HTML versions of articles are electronic conversions from typeset documents. This conversion might result in character translation or format errors in the HTML version. Users are referred to the electronic PDF version (http://www.cdc.gov/mmwr) and/or the original MMWR paper copy for printable versions of official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.

Date last reviewed: 9/10/2008