Severe Acute Respiratory Syndrome in Children

Abstract

Background: 

Severe acute respiratory syndrome (SARS) is a febrile, respiratory tract illness caused by infection with the newly identified SARS-associated coronavirus. A notable feature of the 2003 global SARS outbreak was the relative paucity of cases reported among children. We reviewed the epidemiologic and clinical features of SARS in children and discuss implications of these findings for diagnosis, treatment and prevention of SARS.

Methods: 

We performed a literature search to identify reports of pediatric (younger than 18 years of age) patients meeting the World Health Organization case definitions for SARS and abstracted relevant clinical and epidemiologic information.

Results: 

We identified 6 case series reporting 135 pediatric SARS patients (80 laboratory-confirmed, 27 probable and 28 suspect) from Canada, Hong Kong, Taiwan and Singapore. Among laboratory-confirmed and probable SARS cases, the most common symptoms included fever (98%), cough (60%) and nausea or vomiting (41%); 97% had radiographic abnormalities. The clinical presentation of SARS in patients older than 12 years of age was similar to that in adults. However, patients 12 years of age or younger had milder disease and were less likely than older children to be admitted to an intensive care unit, receive supplemental oxygen or be treated with methylprednisolone. No deaths were reported among children or adolescents with SARS, and at 6 months after illness only mild residual changes were reported in exercise tolerance and pulmonary function. There is only 1 published report of transmission of SARS virus from a pediatric patient.

Conclusions: 

Children and adolescents are susceptible to SARS-associated coronavirus infection, although the clinical course and outcome are more favorable in children younger than 12 years of age compared with adolescents and adults. Transmission of SARS from pediatric patients appears to be uncommon but is possible.

Severe acute respiratory syndrome (SARS) is a febrile, respiratory tract illness that is caused by infection with a novel coronavirus (CoV), the SARS-associated CoV (SARS-CoV).^1–3 The disease first came to international attention in March 2003 following global dissemination of the outbreak by people infected from a single patient at a hotel and subsequently in a hospital in Hong Kong.⁴ The earliest SARS cases likely occurred in the southern part of China in November 2002.^5,6 The disease rapidly spread throughout the world before it was controlled with an intense global response. By July 2003, when the World Health Organization (WHO) declared the outbreak to be contained, 8098 cases and 774 SARS-related deaths had been reported to WHO from 29 countries.⁷

A notable feature of the global SARS outbreak was the relative paucity of cases reported among children. When pediatric cases of SARS did occur, the general perception was that children had a much milder and shorter course of illness than adults.^8–10 We review the epidemiologic and clinical features of SARS in children and adolescents from published case reports and discuss the implications of these findings for diagnosis, treatment and prevention of SARS.

METHODS

We performed a PubMed literature search using the terms “SARS and pediatric” and “SARS and children” to identify reports of pediatric and adolescent patients meeting the WHO case definition for suspect, probable or laboratory-confirmed SARS (Fig. 1). Six case series published in English were identified, describing SARS cohorts in Canada, Hong Kong, Taiwan and Singapore (Table 1). ^11–16 The case definition of SARS, age, gender, epidemiologic link, clinical features, radiographic features and outcome of patients younger than 18 years of age with SARS were abstracted from each paper. One case series included patients up to 20 years of age of which 4 were age 18–20 years and could not be extracted from the rest of the data reported in the paper.

FIGURE 1.: WHO SARS case definition.

TABLE 1: Epidemiologic Features of 135 Suspect, Probable and Confirmed Pediatric and Adolescent SARS Cases

EPIDEMIOLOGIC FEATURES

Of the 135 pediatric patients with SARS reported in the 6 publications, 80 had laboratory-confirmed SARS, 27 had probable SARS and 28 had suspect SARS (Table 1). Forty-six percent of the patients were male. Thirty-six percent of cases reported direct contact with an adult diagnosed with SARS. Sources of other exposures included living in a residential estate where transmission is believed to have occurred through household contact or environmental exposure (50%), nosocomial exposure (10%) or travel to a WHO-designated SARS-affected region (3%). Less than 2% of cases had no exposure to a known source of SARS.^{11–13,15,16} Although SARS-CoV infection in the probable or suspect cases was not verified by a laboratory test, all patients in these categories had documented epidemiologic links to sources of SARS that are known to be associated with risk for infection.

CLINICAL FEATURES

Patients meeting the case definitions for laboratory-confirmed and probable SARS had similar clinical features, but patients with suspect SARS had a lower prevalence of constitutional symptoms, such as chills or myalgia, and lower rates of laboratory abnormalities (Table 2). The difference in clinical presentation of suspect SARS cases may reflect the nonspecificity of the clinical case definition and the inclusion of both SARS-CoV-infected and uninfected people in this group.

TABLE 2: Clinical Description of Pediatric SARS

Among laboratory-confirmed and probable pediatric SARS cases, the most common symptoms included fever (98%), cough (60%), nausea or vomiting (41%) and constitutional symptoms such as myalgia (29%), chills (28%) and headache (28%). In cohorts that included children identified from the Amoy Gardens residential estate outbreak in Hong Kong, diarrhea was not more common when compared with cohorts of children with other epidemiologic exposures (17% versus 29%). This observation is of interest given that diarrhea was found to be a prominent symptom among adults with SARS from the Amoy Gardens cohort, of whom 73% developed diarrhea.¹⁷

Radiographic abnormalities were noted in 97% of laboratory-confirmed and 96% of probable cases, although as in adults these abnormalities were often not detectable in the first few days of illness. Probable cases were more likely to have radiographic abnormalities at admission than were laboratory-confirmed cases (82% versus 48%; P < 0.05). Because almost all pediatric cases in both groups subsequently had radiographic evidence of pneumonia during illness, it is possible that this difference is a result of earlier hospital care sought for patients in the laboratory confirmed group although data on time from illness onset to hospitalization are not available. The most prominent radiographic findings included patchy infiltrates, opacities and areas of consolidation with multifocal lesions, predominantly in the lower lobes; the location of these lesions is consistent with the pattern seen in adults.¹⁶ The most common hematologic and biochemical abnormalities included lymphopenia, leukopenia, thrombocytopenia and elevated levels of lactate dehydrogenase and alanine aminotransferase.^13,14,16

Adolescents older than 12 years of age with laboratory-confirmed or probable SARS had a clinical presentation similar to that of adult patients with SARS, whereas children in this group 12 years of age or younger had a milder illness with a favorable outcome (Table 3). The difference in clinical presentation among children 12 years or younger and children older than 12 years of age was observed for both cases meeting the laboratory-confirmed and probable SARS case definitions. Compared with pediatric patients older than 12 years of age, constitutional symptoms such as myalgias, headache and chills were reported less frequently in children 12 years of age or younger. Younger children also were less likely to be admitted to an intensive care unit, receive supplemental oxygen or be treated with methylprednisolone.

TABLE 3: Comparison of Clinical Illness by Age in Probable and Laboratory-Confirmed Cases

DISTINGUISHING PEDIATRIC SARS FROM OTHER RESPIRATORY TRACT INFECTIONS

Although clinical and laboratory manifestations of pediatric SARS are nonspecific, certain features can help distinguish SARS from other respiratory tract illnesses. A comparison of 15 pediatric patients with laboratory-confirmed SARS and 15 age- and sex-matched patients with culture-confirmed influenza in Taiwan showed that the rates of fever, cough and constitutional symptoms such as chills and myalgia were similar between the 2 groups. However, patients with SARS had significantly less rhinorrhea [7% versus 93%; odds ratio (OR), 0.01; 95% confidence interval (CI), 0.00–0.09], less sputum production (7% versus 53%; OR 0.10; 95% CI 0.02–0.63), and less sore throat (20% versus 60%; OR 0.17; 95% CI 0.03–0.85) than patients with influenza. Conversely patients with SARS were more likely than those with influenza to demonstrate chest radiographic abnormalities (93% versus 38%; OR 22.4; 95% CI 2.2–227).¹³ In another comparison of 16 patients with serologically confirmed SARS and 32 age-matched patients with community-acquired pneumonia of other causes, an increased level of serum lactate dehydrogenase in the presence of a low neutrophil count and low serum creatine phosphokinase was suggestive of SARS-CoV infection.¹⁹ Additional studies are needed.

PROGNOSIS AND OUTCOME

To date, no fatal cases of SARS have been reported among children, and compared with adults the disease appears to be milder in children younger than 12 years of age. In the 135 cases we reviewed, the only patient younger than 12 years of age who required intensive care was a premature infant, born at 30 weeks of gestation and hospitalized at 8 weeks of age, who was managed successfully with continuous positive airway pressure.²⁰ Sore throat, high neutrophil count at presentation and peak neutrophil count of >10,000 cells per mL of blood have been found as independent predictive factors for severe illness in terms of requirements for oxygen and intensive care, regardless of the patient's age.¹⁶ Radiographic changes have been found to resolve more quickly in children than in adults,⁸ although opacities may require 1 month before returning to normal for the most severely affected children.²¹

The outcomes in children up to 6 months after disease onset, including exercise tolerance, pulmonary function and psychologic status, have been favorable.^16,22 In 1 study that examined radiographic and pulmonary function outcome in 47 children with SARS at 6 months after illness, mild abnormalities were detected by high resolution computerized tomography and pulmonary function testing in 34 and 11% of patients, respectively.²²

One report has described osteonecrosis in children who received steroids as part of a treatment regimen for SARS.²³ Although osteonecrosis has been a recognized risk of corticosteroid treatment of other pediatric illnesses including lymphoblastic leukemia,^24,25 the clinical significance and causality of this outcome with SARS are not well-established. For example, Chan et al conducted magnetic resonance imaging on 11 children with SARS and found evidence of osteonecrosis affecting multiple bones in 45% of patients, although all of these patients were free of symptoms of osteonecrosis. In addition, all of these children had been treated with prednisolone, and it is unclear whether the complication was associated with treatment or with the SARS illness. There are similar reports of osteonecrosis and avascular necrosis in follow-up studies of adult patients with SARS who were treated with corticosteroids,^26,27 and an association of osteonecrosis with cumulative steroid dose has been found in a study of 254 steroid-treated patients with confirmed SARS.²⁸

TRANSMISSION

In contrast to the high rate of secondary transmission from adult patients in the absence of infection control measures, transmission of SARS from pediatric SARS patients appears to be uncommon.²⁹ In 1 report, 8 children attended school while they were symptomatic but did not transmit infection to any of their classmates.¹² In addition, a serologic study of parents who had close contact with their SARS-positive children showed no evidence of virus transmission to the parents.³⁰ However, there is one published report of SARS transmission from an 11-year-old child to 3 adults and 1 other child.³¹ This 11-year-old index patient was not hospitalized or diagnosed with SARS until 14 days after onset of illness characterized by fever, cough and malaise. During this time, the child had substantial close contact with other household members who cared for the child, and no specific infection control measures were used. The large amount of time spent in the household while symptomatic, the high frequency of contact with household members and the lack of infection control practices while in the home likely contributed to the transmission of SARS from this patient.³¹ This cluster illustrates that, although it is uncommon, the possibility of SARS transmission from children should not be ignored, and infection control practices for pediatric patients should be similar to adult SARS patients (http://www.cdc.gov/ncidod/sars/guidance/I/index.htm).

DIAGNOSIS

A comparison of clinical and epidemiologic features in laboratory-confirmed, probable and suspect SARS cases suggests that the clinical case definitions for SARS lacked both specificity and sensitivity among pediatric patients. In particular, the definition for suspect SARS appeared to have a low specificity. Patients with suspect SARS had a lower prevalence of constitutional symptoms such as chills or myalgia and lower rates of laboratory abnormalities than laboratory-confirmed and probable SARS cases. In addition, substantially fewer suspect cases reported histories of direct contact with a case of SARS compared with probable cases. Furthermore, many children classified as suspect SARS had an alternative diagnosis at discharge, such as viral fever, dengue fever or bronchitis.¹⁵ The clinical case definitions for SARS also had limited sensitivity, and Leung et al determined that inclusion of lower respiratory tract symptoms in a case definition for SARS would have excluded 34% of pediatric patients who were found to be positive for SARS-CoV by laboratory confirmation but did not have these symptoms.¹⁶

Given the suboptimal predictive value of the clinical case definition for SARS, laboratory diagnosis is crucial for successful identification of SARS-CoV infection. Classic methods for identifying respiratory tract viruses, including cell culture and antigen detection, are not applicable for routine diagnosis of SARS-CoV because of their lack of sensitivity and potential safety risk to laboratory personnel. Laboratory diagnosis is best accomplished with sensitive reverse transcriptase-polymerase chain reaction (RT-PCR) and serologic assays, but a clear understanding of the limitations of these tests is essential for their proper interpretation. In addition, experience with laboratory assays in pediatric patients is limited, and the information reported below primarily is based on experience in adult patients.

Validated and sensitive real time RT-PCR assays are available widely for SARS-CoV but may nevertheless be inadequate to detect the low levels of virus genome present during the first few days of illness. The sensitivity of RT-PCR can be improved by careful attention to specimen type and timing of collection. Blood/serum specimens collected within the first week after onset of illness during peak viremia may offer the best opportunity for early virus detection.^32,33 In infants, blood specimens collected by heel stick should provide adequate sample volume for testing. Lower respiratory tract specimens (eg, bronchoalveolar lavage, tracheal aspirates, sputum) collected in sufficient quantity and at different time points in the illness, but particularly between 1 and 2 weeks after onset of illness, have given the highest yield in adults.^17,34,35 In children, properly collected nasal aspirates or lavage may be more easily obtained and provide similar results; this should be evaluated. Studies also have shown that stool can be an excellent specimen for SARS diagnosis, particularly later in the illness, but data are lacking in children.^17,36,37

Detection of antibodies to the SARS-CoV is the most reliable indicator of infection but has limited diagnostic value during the acute illness.³⁸ Serologic data from adults suggest that most persons seroconvert within 2–3 weeks after onset of illness.¹⁷ However, some people do not develop detectable antibodies until >28 days after onset. Interpretation of serologic test results in infants may be complicated by the presence of maternal antibodies or lack of immune system maturity which could yield false positive or false negative results, respectively.

TREATMENT

Current treatment is based primarily on supportive care; there is not enough information on the efficacy and safety of treatments for SARS to recommend a specific regimen. In an attempt to control inflammation and reduce viral replication, corticosteroids and ribavirin were the most commonly used therapies during the 2003 outbreak. Interferon-α, a cytokine with antiviral activity, and combination protease inhibitors (eg, ritonavir and lopinavir), also were given. However, to date, no prospective, randomized, controlled trials have been conducted to verify the efficacy of any of these treatments, and many safety issues have been raised. Toxicity such as hemolytic anemia and hepatic dysfunction has been noted in adults after treatment of SARS with ribavirin.^39–42 Corticosteroid use for SARS has raised concerns of immunosuppressive effects and delayed viral clearance,⁴³ as well as potential longer term complications such as osteonecrosis or avascular necrosis of bone.^23,26,27 Convalescent plasma from patients with SARS and immunoglobulin have each been used in combination with ribavirin and corticosteroids, but the added benefit of this regimen is unclear. In the pediatric setting, a combination of corticosteroids and short term use of intravenous or oral ribavirin was well-tolerated without serious adverse events such as hemolytic anemia,^12,14 but evidence of the therapeutic efficacy of either agent is lacking.¹⁶

VACCINES

Because of the severe consequences of SARS-CoV disease and the potential for SARS outbreaks to cause substantial social disruption and economic impact, vaccines are being developed against SARS-CoV. Vaccines based on whole SARS virus inactivated with formaldehyde, ultraviolet light and β-propiolactone are being evaluated in clinical trials in China, but safety concerns remain about exposure of production workers to live SARS-CoV in the manufacturing process and from the potential for disease from incompletely inactivated virus. The spike (S) protein of SARS-CoV is responsible for the virus binding to the cell, fusion and entry into cells and is a major inducer of neutralizing antibodies. Recombinant vaccines expressing the SARS-CoV S protein or a specific receptor binding domain of the S protein are being evaluated.⁴⁴ Other strategies to develop SARS vaccines also are being pursued.

LIMITATIONS

Some important caveats should be noted in interpretation of the findings of this review. First, we have summarized a variety of data collected by investigators in different countries without use of standardized data collection instruments, methods of assessment and management of patients, and diagnostic case definitions. Without consistent laboratory confirmation, these descriptions are likely to include cases that were not caused by SARS-CoV. Second, steroids and antiviral agents were used widely as empiric treatments for SARS in both adults and children. Although the effectiveness of these therapies is unknown, as a result of their use, many observed parameters such as biochemical levels and time required for chest radiograph resolution during the course of illness may not reflect the true natural history of SARS. We limited this review to English publications to allow us to review each case series consistently. However, experience with pediatric SARS appears to be similar across affected countries and is likely reflected by what has been published in the English literature.

CONCLUSIONS

Available data confirm that children are susceptible to infection with SARS-CoV, although the clinical course and outcome of illness are more favorable in children than in adults, particularly among children younger than 12 years of age. Transmission of SARS from children also appears to be less efficient than that from adults, a finding that might be related to the lower severity of illness among children. Pediatric SARS manifests with features of atypical pneumonia that are difficult to clinically distinguish from those of disease caused by other respiratory tract pathogens, but certain features can provide helpful clues in differential diagnosis. Although laboratory assays for SARS have improved, currently available tests do not allow exclusion of a diagnosis of SARS in the first few days of illness. The WHO clinical case definition for SARS may have limited sensitivity and specificity, especially in children. Thus, clinicians should consider clinical, epidemiologic and laboratory criteria in combination to ascertain the likelihood of SARS in a given patient. Although some therapeutic agents have shown promise in small clinical series or in vitro studies and although rapid progress has been made in development of vaccines against SARS-CoV, no specific therapeutic or preventive intervention is known to be effective, and prospective randomized controlled clinical trials are needed.

Whether another SARS outbreak will occur is impossible to predict, but reintroduction of SARS could happen periodically. Several animal species sold in markets of Guangdong, China have shown evidence of SARS-CoV infection and presumably were the source of a second cluster of SARS cases in that province in December 2003 and January 2004.^5,45,46 In 2005, 2 studies identified several types of coronavirus phylogenetically related to isolates of SARS-CoV from bats from southern provinces of China and the Hong Kong Special Administrative Region.^47,48 Although the risk of SARS-CoV-infected bats transmitting to humans is not known, these studies suggest bats as one likely reservoir of SARS-CoV. Laboratories that store specimens containing SARS-CoV or that use live virus for diagnostic or research purposes are also a potential source of a reintroduction of virus into humans. There have been 3 instances of laboratory-acquired infections in institutions conducting research with SARS-CoV,^49,50 1 of which occurred in Beijing and led to limited transmission in the community.⁵¹ In the event that a SARS reintroduction leads to a larger outbreak and involves children, the experience with SARS in children during the 2003 outbreak can be used as a guide to the diagnosis and management of infected children as well as to identify key unanswered questions about SARS in this population.