An autopsy of the index patient supported the diagnosis of inhalational anthrax. Autopsy findings included markedly enlarged hemorrhagic mediastinal lymph nodes on gross examination and laboratory detection of B. anthracis by immunohistochemical tests in mediastinal lymph nodes, spleen, liver sinusoids, and phagocytic cells.
The patient had no reported exposure typically associated with naturally occurring anthrax, including exposure to animals or animal products potentially harboring B. anthracis spores. He worked as a photo editor for a national media company that produces tabloid newspapers and other publications. He bicycled and fished for recreation, and his only travel in the 60 days before symptom onset on September 30 was a 5-day automobile trip to North Carolina. No typical naturally occurring anthrax sources were seen at any location inspected, and no B. anthracis contamination was detected among 44 samples from nonworkplace specimens. B. anthracis was identified in 2 of 12 specimens obtained on October 5: from the index patient’s computer keyboard and his mailbox in the company mailroom.
Workplace interviews regarding mail exposure showed that the index patient rarely handled or opened workplace mail, but co-workers recalled that he had examined a piece of stationery containing a fine, white, talc-like powder on September 19. The patient was observed holding the stationery close to his face as he looked at it over his computer keyboard.
Case-Finding and Surveillance
No anthrax cases were detected in Palm Beach County ICU patients, although six patients underwent extensive follow-up from >500 medical charts reviewed through October 31, 2001. No anthrax cases were reported through surveillance by medical examiners. An autopsy was performed to rule out anthrax in one case reported through surveillance of medical examiners and Palm Beach County ICUs, and the patient was determined not to have anthrax. Through 2001, FDOH laboratories reported no B. anthracis isolations among 293 clinical isolates received to rule out anthrax. No reports of veterinary anthrax were received through the Florida Department of Agriculture and Consumer Services. No anthrax cases were reported through nearby county case-finding efforts among persons not exposed in the workplace.
Surveillance among Potentially Exposed Groups
Among six workplace-exposed persons who were extensively evaluated after medical providers reported their illness, one was also identified among Palm Beach ICU patients, and inhalational anthrax was confirmed in another, a Miami-Dade County resident. This second case-patient was a 73-year-old mail distributor and co-worker of the index patient, who was reported by his medical provider on October 4. His illness began on September 28, and he was admitted to a hospital in Miami-Dade County on October 1, 2001. A nasal swab culture obtained on October 5 showed B. anthracis, but cultures from blood, bronchial washings, and pleural fluid, obtained after initiation of antibiotics, were negative. Two specimens of pleural fluid obtained on October 5 and 12 were tested by PCR and were positive for B. anthracis. Immunohistochemical staining of B. anthracis capsule and cell-wall antigens from pleural fluid cytology preparations and from transbronchial and pleural biopsy tissues obtained on October 5 and 12 were positive. Serial serum samples, obtained on October 7, 10, 11, and 17, indicated a serum IgG antibody response to the PA component of the anthrax toxin consistent with acute B. anthracis infection. The patient was treated with antibiotics and was discharged from the hospital on October 17 (10
Of 1,076 nasal cultures obtained from workplace-exposed persons, two yielded B. anthracis. The first was the second case-patient, and the second was from an asymptomatic mail sorter in the same workplace. Nasal swab cultures obtained from two workers that handled workplace trash did not yield B. anthracis.
Interviews with employees regarding suspicious mail showed that the two workplace-exposed persons with nasal cultures positive for B. anthracis had extensive mail exposure. One, the second case-patient, was the workplace mail distributor; he did not generally open mail and did not recall handling or seeing any mail containing powder or described as unusual or as hate mail. He picked up 10,000–15,000 pieces of mail from the post office each weekday in the company mail van and distributed it at the workplace. The other co-worker, a 36-year-old woman, sorted mail and opened mail addressed to a periodical different from the one to which the index patient contributed. She recalled opening an envelope that released powder in her office on or about September 25. Afterwards, she discarded it in the trash without reading it. The letter most likely had arrived during the previous 2 weeks while she was on vacation. No other workplace mail likely to contain B. anthracis was suggested through further interviews.
Workplace information about exposure to suspicious mail indicated that the incubation period for both Florida case-patients was <12 days (). The index patient had onset of illness 11 days after handling suspicious mail on September 19. The second case-patient had illness onset September 28, 9 days after the index-patient viewed suspicious mail on September 19 and 3 days after his co-worker opened a letter with powder in it on September 25.
Dates of onset of symptoms of inhalational anthrax cases in Florida, and timeline of related events, September 16–October 16, 2001.
Serial or paired serum tests for IgG antibody response to the PA component of the anthrax toxins were performed on serum of 436 workplace-exposed persons. No serum indicated a reaction consistent with acute B. anthracis infection except for that of the second case-patient. For most of the serologic tests, specimens were collected on October 10 and 17.
Among 32 postal workers who potentially handled workplace mail at two county postal facilities, 31 nasal cultures were obtained; none yielded B. anthracis. No anthrax cases were detected among 3,263 postal workers working at the 51 Palm Beach County postal facilities through the county postal worker surveillance system, which reported 226 illnesses and 7 hospitalizations during October 25–November 9, 2001.
Of 136 investigation-directed environmental samples obtained during October 8–10 from the workplace and company mail van, 20 were positive, including 10 of 20 from the mailroom, 1 of 2 from the company mail van, 5 of 6 from the office of the asymptomatic mail-sorter who had a positive nasal culture and had opened a letter containing powder, 2 of 21 from the index patient’s work area (at an incoming-mail desk near his workspace and a repeat sample from his computer keyboard), 1 of 9 in the text library, and 1 from the single basement ventilation filter sample. No B. anthracis contamination was detected from 8 trash receptacles or 2 roof ventilation filters, 28 bulk items removed from the building containing security camera information, 18 samples from a construction area, or 21 other samples from other work areas and the entrance lobby. Five samples from the third-floor HVAC ducts (three from the index patient’s office and one from another office), and three samples from the first floor HVAC ducts (from the mailroom, an office where an envelope with powder was opened, and the text library) were negative.
Eighty-four of 460 workplace samples obtained during October 25–November 8 yielded B. anthracis (). Isolates of B. anthracis were obtained from 66 of 247, 10 of 95, and 8 of 112 samples from the first, second, and third floors, respectively; none of 6 specimens were positive from the parking garage or roof vents. The northeast quadrant of the third floor, which contained executive office suites, a conference room, and storage areas, was the only quadrant of any floor without detected contamination. The index-patient’s office was located on the third floor of the building. The mailroom (the work area of the second case-patient) and the office near the mailroom where a powder-containing letter was opened are both on the first floor.
Figure 2 Environmental sample locations of specimens tested for Bacillus anthracis obtained October 25–November 8, 2001, on the three floors of the media company building where patients were employed, Palm Beach County, Florida. Sample locations of 59 (more ...)
No mail containing B. anthracis spores was recovered. Because workplace refuse is incinerated and waste receptacles did not show contamination, no environmental specimens were obtained from waste sites.
B. anthracis contamination was detected at six of seven postal facilities tested, from routes serving the current workplace headquarters and a former office. Contamination was not detected at a facility that receives mail addressed to the post office box of the former workplace office, last used 13 months earlier.
Molecular subtyping analysis (MLVA) was performed on one B. anthracis isolate recovered from a postal facility that processed workplace mail, 18 isolates recovered from the workplace, cerebrospinal fluid and blood culture isolates from the index patient, and two nasal swab isolates from workplace-exposed persons. All B. anthracis isolates tested were indistinguishable by MLVA.
Prophylaxis and Control Measures
Beginning October 8, we recommended 60-day antibiotic postexposure prophylaxis (2
) to 1,114 workplace-exposed persons identified through employers and responses to public service announcements. We provided medication refills on October 17–19 and November 1 at a workplace branch office and as needed through the Palm Beach County Health Department. Beginning October 24, we attempted telephone contact with persons who did not refill medications and advised them about our recommendations and how to obtain medications. Adjunct anthrax vaccination, available beginning December 22, was accepted by three workplace-exposed persons.
When the postal system risk assessment was initiated on October 12, antibiotic prophylaxis was offered to 32 postal workers who were most likely to have handled workplace mail at two local postal facilities. After we determined that at least 24 days had passed since contamination most likely took place in postal facilities, we did not recommend prophylactic antibiotics to Florida postal workers since more than two of the typical 1- to 7-day incubation periods for inhalational anthrax had passed, or two of the up to 12-day incubation periods estimated for the two Florida cases.