Pneumocystis canis pneumonia in dogs

Pneumocystis canis is a potential cause of life-threatening interstitial fungal pneumonia in dogs. It is seen almost exclusively in two canine breeds, miniature Dachshunds and Cavalier King Charles Spaniels (CKCS)¹. Historically, Australian veterinarians had a key role in the documentation of this entity and its conspicuous breed associations^2–4. Affected Dachshunds and CKCS are likely to have an inherited immunodeficiency that predisposes them to infection with this commensal organism of the respiratory tract and pharynx^1,2,5,6. A high index of suspicion is required to make a timely diagnosis and save affected patients, as these dogs cope poorly with anaesthesia and other measures to procure the specimens required to make a definitive diagnosis. Possible co-infection with Bordetella bronchiseptica must be considered when determining antimicrobial strategies. Affected dogs occasionally have a previous or concurrent history of generalised demodicosis^5,7,8. With early intervention, affected dogs can be saved, although some require life-long therapy to prevent recurrence. The future challenge is to develop fast molecular techniques to diagnose P. canis^† pneumonia (PCP)^7,9 and to determine the underlying immune defect in over-represented breeds through the rapidly advancing field of canine genomics¹⁰.

Pneumocystis spp. are ubiquitous commensals of the respiratory tract of many mammalian species, including dogs¹. This group of organisms has the potential to cause life-threatening pneumonia in a wide range of mammals, including rats, pigs, horses and goats¹. P. jirovecii has the same pathogenic potential in people¹. Airborne droplet transmission from subclinically affected normal dogs, often transmitted from bitch to pup soon after birth, is suspected of being the means for transmission¹. PCP results from the effects of masses of organisms within the alveolar spaces, combined with the associated inflammatory response of the host^1–6.

Historically, most canine cases had been reported in young (less than one-year-old) miniature Dachshunds^1–3,6,11. The first documentation of this entity was provided in a paper from the University of Sydney (UoS) by Farrow and colleagues², although a brilliant immunoparasitologist at McMaster Laboratory (Robert J. Love) provided key immunologic insights.^‡ The disease in Dachshunds was characterised further by South African veterinarians, with Lobetti and colleagues adapting human treatment strategies to successfully manage their canine cases^1,11. The best radiological description of PCP was provided by Kirberger¹², although this has been complemented by recent advances in imaging such as digital radiology and computed tomography (CT). Although PCP continues to be seen in Dachshunds in South Africa and the USA, the great majority of cases encountered in Australia over the past 20 years have been in adult CKCS, the first detailed report being provided by Paul Canfield and colleagues from UoS in 1993⁴. This breed preponderance was then detected in the UK^5,13 and subsequently Europe, the USA and Japan⁷. ‘Cavaliers’ tend to get the disease as young adult dogs^4,5,13. There may be antecedent or concurrent footprints of immune deficiency, such as generalised demodicosis^5,8.

In people, the disease is best known as a complication of HIV/AIDS, although it is also seen in transplant recipients and other patients receiving immunosuppressive drug regimens. A very well characterised case cluster in Australia that closed a transplant ward at Westmead Hospital was reported by Sharon Chen, Wieland Meyer and their ANZMIG colleagues¹⁴. In Arabian horses and related breeds, foals with autosomal recessive severe combined immunodeficiency syndrome (analogous to the like-named SCID condition of men and mice) typically died of PCP in the neonatal period¹⁵, and this was common in the Camden district in the 1970s. A test was developed for this genetic disease of horses, and nowadays the condition is hardly encountered, a cogent example of genetic counselling and preventative veterinary medicine¹⁶.

In dogs, the nature of the immune defect in Dachshunds and Cavaliers has not been determined at the molecular level. Poor lymphocyte stimulation (despite normal lymphocyte numbers in peripheral blood) has been documented in miniature Dachshunds^1,2. IgA, IgM and IgG concentrations are subnormal in affected members of this breed⁶. Subsequent assessment of CKCS with pneumocystosis also reveals immunoglobulin deficiencies and decreased lymphocyte function^5,7. These immunodeficiencies persist after resolution of the infection with effective antimicrobial therapy. Studies of the involvement of adaptive cell-mediated immunity and innate immunity (including the potential involvement of Toll-like receptors) would be sensible but have not yet been performed.

Most dogs with PCP present with respiratory signs, including dyspnoea, tachypnoea, increased breath sounds on thoracic auscultation or cyanotic mucous membranes. The presence of a cough is variable, and this sign is frequently absent. Duration of signs prior to presentation for veterinary attention can be as long as four weeks^1–4. Haematologic and serum biochemical abnormalities are non-specific¹. Thoracic radiographs reveal diffuse, bilaterally symmetrically increased radiodensity of the pulmonary parenchyma, classically described as a miliary-interstitial to alveolar pattern (Figure 1)^1,12,17. There is often radiological evidence of right-sided heart enlargement and pulmonary hypertension, and this can be confirmed echocardiographically⁴. Solitary opacities, cavitary lesions or a non-symmetric radiographic changes are less commonly observed^1,12. Severe long-standing cases may develop emphysema¹. The CT findings in canine PCP are dramatic (Figure 2), providing a better appreciation of the lung pathology and disclosing regional lymphadenomegaly.

Figure 1. (a) Dorsoventral thoracic radiograph from a one-year-old Cavalier King Charles spaniel with pneumocystosis, revealing a patchy, diffuse, interstitial pattern in all lung fields. (b) Right lateral thoracic radiograph of the same dog reveals a diffuse interstitial to alveolar pattern, especially in the dorsocaudal lung fields.

Figure 2. Thoracic computer tomography scan in the same dog as Figure 1 reveals a marked, patchy, increased pulmonary opacity throughout all lung lobes with some peripheral consolidation. The ground-glass appearance of the lung parenchyma is quite characteristic of advanced PCP in the dog.

Diagnosis can be challenging in a veterinary context, with identification of P. canis ‘cysts’ in bronchoalveolar lavage fluid (BAL) or transthoracic fine needle aspirates of lung generally being required. Intact cysts are distinctive, being 5–10 µm in diameter and containing 4–8 dark-staining intracystic bodies (2–3 µm in diameter; Figure 3). PCR testing of BAL fluid or lung aspirates is uncommon in veterinary laboratories, but should probably be attempted more often^7,9, as the requirement to perform open lung biopsies to secure a diagnosis is unnecessarily invasive given the characteristic nature of the imaging findings and the strong breed associations. If BAL fluid specimens are obtained, they should be subjected to routine bacterial culture as well as microscopy and PCR, as Bordetella bronchiseptica can be an important co-morbidity in canine PCP pneumonia. If a BAL is not done and treatment is based on a presumptive diagnosis, then doxycycline should be given as well as trimethoprim sulfonamide to cover this possibility. Research is currently focused on development of qPCR assays for canine P. carinii strains, as assays developed for P. jirovecii may not detect P. canis associated with infections in dogs¹. Panfungal PCR with sequence analysis provides an interim molecular diagnostic strategy, as does genus-specific Pneumocystis PCR assays^7–9.

Figure 3. A single cyst-like structure (red arrow) suspicious of Pneumocystis spp. ‘cyst’ seen on cytology from a one-year-old Cavalier King Charles Spaniel with pneumocystosis (Diff Quik^® stain, oil immersion microscopy).

Treatment includes appropriate antimicrobial therapy, using intravenous trimethoprim sulfonamide combinations, and less commonly nebulised pentamidine, along with appropriate supportive care (e.g. supplementary oxygen, nebulisation, chest physiotherapy). Long-term ventilation of affected dogs is challenging and beyond the financial limits of most owners. There is no zoonotic potential for humans in close contact with a Pneumocystis infected dogs, as the organisms in humans and dogs are distinct and no evidence of cross-species transmission has been documented.

As thoughtful and responsible veterinary physicians, we should be banking DNA from all dogs with confirmed and presumptive PCP. The most likely way to eliminate this condition is by harnessing the power of canine genomics to detect any underlying genetic defect, utilising either a genome-wide association study or whole genome sequencing of ‘Trios’ of affected and closely related dogs¹⁰. It is much better to try and prevent this severe mycosis by screening the breeding population using a molecular genetic test, rather than to have to treat a severely affected individual. This is where vets have an advantage over medics, we can use eugenics!