From December 16, 2001, to December 15, 2002, 1,048 respiratory specimens from 851 children were tested by RT-PCR for HCoV-HKU1. Specimens from 9 of these children (1%) tested positive for HCoV-HKU1. Specimens from these children tested negative for RSV, parainfluenza viruses (types 1–3), influenza A and B viruses, and adenovirus by direct immunofluorescence assay as well as human metapneumovirus and HCoV-NH by RT-PCR. Two children had 2 specimens that tested positive for HCoV-HKU1. For each of these 2 children, the positive specimens were collected <10 days apart. Children whose specimens tested positive for HCoV-HKU1 infection had clinical evidence of either upper or lower respiratory tract infection or both (Table). The most common clinical findings were rhinorrhea (100%), cough (67%), fever (67%), and abnormal breath sounds on auscultation (44%). Hypoxia (oxygen saturation of <90%) was observed in only 1 patient. Chest radiographs were obtained for 4 patients, all of whom had abnormal findings that included peribronchial cuffing, atelectasis, hyperinflation, or infiltrates. One patient (patient 3) had respiratory decompensation requiring ventilatory support and was admitted to the pediatric intensive care unit. This patient had no history of underlying illness, had not been premature, and was 1 month of age at the time of specimen collection.

Two patients had evidence of disease beyond the respiratory tract. One patient (patient 1) was hospitalized for new-onset seizures. Workup for a central nervous system infection, including a lumbar puncture and head magnetic resonance imaging, was unrevealing. Although a febrile seizure remains a possible diagnosis, no evidence of fever was reported by the mother or noted during the hospital stay. A second patient (patient 7) was hospitalized with hepatitis. This patient had undergone liver transplantation 3 months before admission. Immunosuppressive medications included tacrolimus and prednisolone. The patient was also receiving ganciclovir for cytomegalovirus prophylaxis. The onset of abnormal liver enzyme levels occurred several days after the onset of respiratory symptoms and after collection of the respiratory specimen that tested positive for HCoV-HKU1. No evidence of abnormal liver function was detected (both prothrombin time and partial thromboplastin time were within normal ranges). Serologic assays for hepatitis viruses A, B, and C were negative. A liver biopsy specimen did not show evidence of rejection. Levels of the serum liver enzymes slowly decreased during hospitalization. No interventions (e.g., changes in immunosuppressive therapy) were performed.

Figure

Figure. Weekly distribution of human coronavirus (HCoV)-HKU1 infection in children <5 years of age, December 16, 2001, to December 15, 2002, New Haven, Connecticut. The weekly distributions of HCoV-HKU1 isolates are shown...

All HCoV-HKU1 infections occurred during a 7-week period from December 2001 to February 2002 (Figure). HCoV-HKU1–positive samples accounted for 5% of samples screened during that period. No HCoV-HKU1–positive isolates were detected in specimens collected in the remainder of the study period.

The RT-PCR amplicon from each positive specimen was sequenced. Nucleotide and amino acid identity between replicase 1B region of the original HCoV-HKU1 isolate and the New Haven isolates were both >95%. Rare polymorphisms (<1% of sequence) were noted in the HCoV-HKU1 sequences of the New Haven isolates (data not shown), which suggests that a single strain was circulating in the community during the study period.

We report the first identification of HCoV-HKU1 in the Western Hemisphere. These findings suggest that HCoV-HKU1 may have a worldwide distribution. We detected this coronavirus in 1% of children screened. All HCoV-HKU1–positive samples tested negative for RSV, influenza virus, parainfluenza viruses, adenoviruses, HCoV-NH, and human metapneumovirus. Our laboratory did not have access to materials from Hong Kong; therefore, the results cannot represent laboratory contamination from material obtained elsewhere. The percentage of positive specimens was similar to that described by Woo et al. (1 [0.25%] of 400) (14) and Sloots et al. (10 [3.1%] of 324) (15), which suggests that infection with HCoV-HKU1 may be uncommon or that the virus has properties that decreases the likelihood of detection, such as a brief period of viral shedding. Our study, the study by Sloots et al., and the original study by Woo et al. screened respiratory specimens submitted to a diagnostic laboratory. Therefore, HKU1 may be a common virus that causes symptomatic disease in only a relatively small percentage of infected persons. All HCoV-HKU1–positive specimens were collected from December 2001 to February 2002, which implies a winter distribution. The study by Sloots et al. also detected HCoV-HKU1 predominantly in the winter, although only respiratory samples submitted during winter months were screened. Whether the seasonal distribution of HCoV-HKU1 varies from year to year is not known.

Similar to the patients described by Woo et al., several HCoV-HKU1–positive patients had evidence of lower respiratory tract involvement (2 patients with pneumonia and 1 patient with bronchiolitis). Two of these patients had underlying illness. However, most patients identified in our study had only mild upper respiratory tract symptoms. Most HCoV-HKU1 infections in children, similar to other common HCoV infections, likely result in mild disease (4). The Australian study did not perform a detailed clinical review of HCoV-HKU1–positive patients, but the authors note that symptoms are consistent with those of acute respiratory tract illness (15).The severity of disease caused by SARS-CoV in children was also relatively mild for reasons that are not yet understood (17). Underlying illness and preexisting lung disease may predispose to a more severe clinical course.

Evidence of hepatitis in 1 child who tested positive for HCoV-HKU1 is an intriguing finding. HCoV-HKU1 is most closely related to the murine hepatitis virus, a virus that causes hepatitis as well as demyelinating disease in mice (18). Because of this patient's medical history (liver transplantation) and compromised immune status, many potential causes of hepatitis exist, though serologic assays and liver biopsy findings were unrevealing. Several reports have found coronavirus-like particles in stool of persons with gastrointestinal disease (19), which suggests that these viruses, like coronaviruses of animals, can cause disease of the gastrointestinal tract. Future studies will be needed to determine whether HCoV-HKU1, or other common human coronaviruses, play a role in liver disease.

Our study had several shortcomings. We limited our screening to respiratory specimens that were collected at the discretion of the medical team, we did not include a control group of asymptomatic children, and serum samples were not available for serologic assays. Nonetheless, our findings show that HCoV-HKU1 is circulating in New Haven, Connecticut, and is associated with both upper and lower respiratory tract disease and perhaps extrapulmonary disease.

The genetic variability of HCoV-HKU1 is unknown. The study by Sloots et al. suggests 2 genotypes when comparing the Australian isolates to the prototype Hong Kong strain (15). If multiple genotypes exist, they may not all be detected with the primer set used. This limitation would result in an underestimation of this virus in our study. However, the region of the replicase 1B gene targeted by the primers used (14) is highly conserved among other coronaviruses, and our screening was unlikely to have lacked sensitivity for that reason. Also, only rare polymorphisms were detected on the sequence analysis of the 9 individual isolates, which suggests that this region is highly conserved. However, to establish the true prevalence of HKU1, use of primers with known specificity and sensitivity for HCoV-HKU1 will be critical.

In conclusion, we show that HCoV-HKU1 circulates in the United States, and the strain identified in New Haven is similar to the original strain described from Hong Kong. Whether this newly recognized pathogen is responsible for a substantial proportion of respiratory tract disease in children remains to be determined. Future studies are required to determine the epidemiologic features and clinical spectrum of this newly recognized pathogen.