The Hirwa group consisted of 12 animals: 1 adult male, 6 adult females, 3 juveniles, and 2 infants. Moderate to severe respiratory disease (>2 characteristic signs) developed in 11 of 12 animals. Five (3 juvenile males and 2 adult females) received antimicrobial drug therapy (ceftriaxone, 50 mg/kg for adults, 100 mg/kg for infants), 4 by remote delivery and 1 while chemically immobilized. Two untreated animals (1 adult female and 1 male infant born to a symptomatic mother) died. On June 30, the adult female was first observed coughing and lethargic but still feeding. On July 3, she left her night nest in the morning but did not join her group; she exhibited severe clinical signs and was found dead on July 4 at ≈1:00 pm. The infant was 3 days old when it died on July 23. Clinical signs of respiratory illness had not been observed, although its mother showed severe clinical signs for 2–3 days before and after delivery; before delivery, she had received antimicrobial drugs by remote delivery (neither she nor her infant were handled by humans).

Gross postmortem examinations revealed bronchopneumonia in the adult and unilateral pulmonary congestion and an empty stomach in the infant. Formalin-fixed (10%) postmortem tissue samples from the adult and infant were prepared in 6 μm sections for histologic studies, stained with hematoxylin and eosin according to standard methods, and examined by light microscopy. Histologically, the respiratory tract of the adult was characterized by moderate mononuclear tracheitis, laryngitis, and air sacculitis; severe pulmonary alveolar histiocytosis; multifocal severe suppurative pneumonia; and multifocal pulmonary thrombosis and hemorrhage. One section of lung from the infant showed pulmonary atelectasis, congestion, mild alveolar hemorrhage, and histiocytosis. The infant also had moderate neutrophil and macrophage infiltration of the umbilicus at the body wall; neutrophilic inflammation in the media and adventitia of 1 umbilical artery at the level of the bladder; and mild, unilateral, focal, segmental, neutrophilic glomerulitis and tubulointerstitial nephritis. Coronary groove and mesenteric fat were absent.

Figure

Figure. Bayesian analysis of the G gene of human metapneumovirus (HMPV) isolated from an adult female mountain gorilla that died during an outbreak of respiratory disease, Rwanda. Bayesian phylogenetic analyses of sequence...

Multiplex PCR analysis for respiratory pathogens indicated sequences of human metapneumovirus (HMPV) in serum; lung tissue; and throat, nose, anus, and vagina swabs from the adult gorilla, and in lung tissue from the infant (Table). Streptococcus pneumoniae was detected in lung tissue and in throat and nose swabs of the adult gorilla but not in the infant. Klebsiella pneumoniae was also detected in all specimens from the adult gorilla. Microbial loads were determined by quantitative PCR (Table). The sample with the highest viral load, a throat swab from the adult female (6.2 × 10⁵ genome copies/µL), was pyrosequenced, yielding 607,484 reads comprising 3.8 kb of sequence (27.5% of the genome). Simple pairwise analysis indicated that the strain belonged to lineage B2 of HMPV (4). Bayesian analysis revealed close relationship of the gorilla virus to human isolates from South Africa (Figure). Sequence information for the reported HMPV has been deposited with GenBank under accession no. HM197719. Detailed methods are available in the Technical Appendix.

Experimental infections of cynomologus macaques with HMPV have suggested that pure infection with this virus causes minimal to mild lesions in conducting airways and increased macrophages in alveoli (5). However, paramyxoviruses, including HMPV, can predispose animals to bacterial pneumonia (6–8), as appeared to be the case in the adult female mountain gorilla reported here. That HMPV can be fatal for gorillas is supported by a report of a respiratory outbreak in wild, human-habituated chimpanzees in which several chimpanzees died (2,9).

We report conclusive evidence for association of a human virus with death in mountain gorillas (2,3). Viral RNA in multiple tissue samples from the adult female indicates that she was infected by an HMPV strain at the time of her death. The upper respiratory lesions were suggestive of a viral infection (9). The pulmonary lesions indicated a bacterial bronchopneumonia as the proximate cause of death, compatible with an etiologic agent such as S. pneumoniae and K. pneumoniae, the 2 organisms detected by PCR. Although the cause of death of the infant was likely inanition and acute dissemination of an umbilical infection to a kidney, detection of HMPV as the sole pathogen in the infant tissues supports the presence of this agent in the gorilla group during the respiratory disease outbreak.

The source of the virus is unknown; the strain was most recently described in South Africa. The 2 HMPV-positive animals were not handled by veterinarians or park personnel during the course of their illness. Although HMPV transmission as a result of human intervention to treat sick animals in the group is possible, it does not explain HMPV in the adult female, which died early in the outbreak before any clinical interventions were conducted. Although human proximity to mountain gorillas is essential for their conservation, also crucial is minimizing the risk for human-to–great ape transmission of respiratory pathogens.