Pneumonic Plague in a Dog and Widespread Potential Human Exposure in a Veterinary Hospital, United States

Paula A. Schaffer; Stephanie A. Brault; Connor Hershkowitz; Lauren Harris; Kristy Dowers; Jennifer House; Tawfik A. Aboellail; Paul S. Morley; Joshua B. Daniels


Emerging Infectious Diseases. 2019;25(4):800-803. 

In This Article


Several factors delayed the diagnosis of pneumonic plague, resulting in many potential exposures. Pneumonic plague is uncommon in dogs; most dogs with plague have bubonic or septicemic plague and signs of fever, lethargy, and peripheral lymphadenopathy.[1] The occurrence of this case during December was outside the predominant period of plague transmission in the Northern Hemisphere (April–October).[5] Of 89 animals with plague reported to CDPHE during 2008–2017, only 1 case (in a wild lynx) occurred in December. The mild winter in Colorado during 2017 might have prolonged activity of flea vectors, consistent with climate models that predict altered plague seasonality.[6] In humans, radiographic abnormalities typically include bilateral lobar changes;[7] in this dog, the accessory lung lobe was primarily affected on initial imaging, and this finding was interpreted as aspiration pneumonia. Histologically, pneumonic plague usually results in acute necrosuppurative, hemorrhagic pneumonia with obvious colonies of bacteria.[8] In this case, antibiosis might have resulted in the histologic absence of bacteria.

Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry identified the isolate as Y. pseudotuberculosis with a high confidence score, but the species-level identification was considered suspicious. Y. pestis has been previously misidentified as Y. pseudotuberculosis, Pseudomonas luteola, and Acinetobacter lwoffii by automated systems[9] even when information for Y. pestis is present in databases. This isolate produced a strong peak at 3,061 m/z, which was present in spectra of the 5 Y. pestis isolates in the database and absent from Y. pseudotuberculosisspectra, likely corresponding to a Y. pestis biomarker previously described at 3,065 m/z.[10]

At the time of this incident, CSU-VTH infection control standard operating procedures stressed suspicion of plague in cats and minimized the presence of this disease in dogs. Criteria for designating patients as high-risk plague suspects were opportunity for exposure (including geography and season), fever, and enlarged peripheral lymph nodes. Although the dog was febrile and had interacted with a prairie dog, the species, lack of peripheral lymphadenopathy, and nonseasonal presentation resulted in low-risk designation. Standard operating procedures are being updated to emphasize that dogs might be affected with plague year round and that lymphadenopathy might not be present. Isolation and plague testing for all dogs with pneumonia is not feasible, but increased suspicion of patients with hemoptysis might be appropriate. A relatively rare finding in dogs,[11] hemoptysis was present in this dog, as well as in 2 other previously reported cases of canine pneumonic plague.[12,13]

This unique case highlights the public health response challenges in a large teaching institution. Veterinary workers are at increased risk for infection with zoonotic diseases,[14] and exposure to infectious agents is an occupational stressor with potential emotional toll.[15] Veterinary hospital administrations should educate staff about zoonotic hazards, mitigate exposures, and communicate rapidly to personnel when potential exposures occur. The communication process in place at the CSU-VTH for zoonotic exposures has historically been used in small-scale events. The extensive potential exposures in this instance highlighted the shortcomings of the process for reaching large numbers of persons. These problems are being addressed through development of frequently updated email listservs and telephone lists. Computerized logs may also be useful for documenting contact with patients with historical and syndromic factors consistent with potential zoonoses.

In summary, pneumonic plague, although rare, should be considered in dogs that have fever and respiratory signs with potential exposure in disease-endemic areas, regardless of season and lobar distribution. Efficient zoonotic disease communication and response plans should be prepared for large-scale events.