Objective. Sin Nombre hantavirus (SNV) is the leading causative agent of hantavirus cardiopulmonary syndrome (HCPS) in the United States and Canada. Relatively few cases of HCPS have involved children. This report describes the clinical characteristics of a series of pediatric cases of SNV infection in the United States and Canada from 1993 through March 2000.
Methods. We analyzed clinical and laboratory data on 13 patients who were ≤16 years old with SNV infection from 1993 through March 2000 identified from a database at the University of New Mexico.
Results. The patients ranged from 10 to 16 years of age, with a median of 14. Fifty-four percent were female. Fifty-four percent were Native American. The most common prodromal symptoms were fever, headache, and cough or dyspnea (100%); nausea or vomiting (90%); and myalgia (80%). The most common physical findings at admission were tachypnea (67%) and fever (56%); hypotension was seen in 33% of patients. On admission, all patients manifested thrombocytopenia (median platelet count: 67 000/mm3) and elevated lactate dehydrogenase (median level: 1243 IU/L), and >85% of patients had elevated levels of serum aspartate aminotransferase, alanine aminotransferase, and hypoalbuminemia. Leukocytosis and hemoconcentration were seen in less than one third of patients at admission. HCPS developed in 12 of the 13 patients (92%), and 4 of those 12 died (33% case-fatality ratio). The majority of HCPS patients (8 of 12 [67%]) were critically ill and required mechanical ventilation. Extracorporeal membrane oxygenation was used in 2 patients, 1 of whom survived. An elevated prothrombin time (≥14 seconds) at admission was predictive of mortality.
Conclusions. Infection with SNV in children and adolescents causes HCPS with a clinical course and mortality rate similar to that described in adults. We believe that early recognition of HCPS in children and adolescents and appropriate referral to tertiary care centers that are experienced with HCPS are important in reducing mortality.
Hantavirus cardiopulmonary syndrome (HCPS), also known as hantavirus pulmonary syndrome, is a viral zoonotic disease. It is characterized by a febrile prodrome progressing to respiratory distress from noncardiogenic pulmonary edema and in severe cases to cardiogenic shock. HCPS is seen primarily in adults and has a case-fatality ratio of 38%. The majority of deaths result from hypoxemia and cardiac dysfunction with marked hypotension and ventricular arrhythmias. For this reason, some experts1 prefer the term “hantavirus cardiopulmonary syndrome” to the term “hantavirus pulmonary syndrome.” The Sin Nombre hantavirus (SNV) is the primary causative agent of HCPS in the United States and Canada.
As described by the Centers for Disease Control and Prevention (CDC),2 the HCPS prodrome typically consists of fever, chills, myalgia, headache, and gastrointestinal symptoms. The CDC's clinical case definition of HCPS is a febrile illness (ie, temperature >38.3°C) characterized by bilateral diffuse interstitial edema that may radiographically resemble adult respiratory distress syndrome, with respiratory compromise requiring supplemental oxygen, developing within 72 hours of hospitalization, and occurring in a previously healthy person. Typical laboratory findings include hemoconcentration, thrombocytopenia, left shift in the white blood cell count (WBC), neutrophilic leukocytosis, and circulating immunoblasts. Laboratory criteria for diagnosis include detection of hantavirus-specific immunoglobulin M or rising titers of hantavirus-specific immunoglobulin G, detection of hantavirus-specific ribonucleic acid sequence by polymerase chain reaction in clinical specimens, or detection of hantavirus antigen by immunohistochemistry.2 A few cases of SNV infection leading to febrile illness without respiratory compromise have been reported3–5; however, the majority of cases progress to HCPS as described above.
Typically, the disease affects healthy adults in rural settings, where there is peridomestic or occupational exposure to aerosols of rodent excreta. The deer mouse (Peromyscus maniculatus) is the main rodent reservoir for SNV. Most cases of HCPS have occurred in the southwestern United States, although confirmed cases have been reported in 30 states.6
Because relatively few cases of SNV infection have involved children, the disease has been described primarily as it occurs in adults. Since the emergence of HCPS in 1993, 15 (5.5%) of 274 cases reported to the CDC have involved children ≤16 years of age (J. Young, Special Pathogens Branch, CDC, personal communication, November 2000). Several case reports of children with hantavirus infection in the United States have appeared in the literature.3,,57–10 Because no case fatalities have been reported in children younger than 14 years, it has been hypothesized that younger children and adolescents who are infected with SNV are less likely to develop serious illness than are older adolescents and adults. This report describes the clinical characteristics and outcomes of a relatively large series of pediatric cases of SNV infection in the United States and Canada from 1993 through March 2000.
Since 1993, the University of New Mexico (UNM) Department of Pathology has maintained a database of hantavirus infection cases documented with serologic testing performed at the UNM Health Sciences Center. The database includes patients who were seen at UNM Hospital as well as those hospitalized elsewhere. This database was reviewed for cases of hantavirus infection involving patients who were 16 years of age or younger. Supplemental information was obtained by medical chart review, communication with referring physicians, and in one instance from a published case report.8
Data were extracted by 2 authors (M.M.R. and R.B.R.) using a standardized data collection form and were analyzed using SAS software version 6.12 (SAS Institute, Inc, Cary, NC). Correlations between mortality and symptoms before admission, physical examination findings, or laboratory findings at admission were examined using Fisher's exact test (2-tail) or univariate logistic regression, where appropriate. Correlations between the development of respiratory failure and the aforementioned patient characteristics were examined similarly.
Serologic specimens were analyzed by Western blot and/or strip immunoblot assays. Our criterion for diagnosis was the detection of immunoglobulin M antibodies to SNV nucleocapsid (N) antigen. In all cases, immunoglobulin G antibodies to glycoprotein G1 antigens were present as well. Antibodies to glycoprotein G1 are specific for infection with SNV and are not seen with infections caused by other, closely related hantaviruses.11–13
As of March 31, 2000, the hantavirus database at UNM included a total of 175 patients with SNV infection confirmed by serologic testing at UNM Hospital. Of these, 13 patients (7%) were ≤16 years old. Ages ranged from 10 to 16 years (median: 14 years). Four of these 13 patients died (31%). The median age of those who died was 15 years (range: 14–15 years) versus 12.5 years (range: 10–16 years) for those who survived. Seven were female (54%) and 6 were male. Seven (54%) were identified as Native American; 2 (15%) were Hispanic; 1 (8%) was white; information on ethnicity was not available for the other 3 patients.
Medical records of 11 of these 13 patients were available for review; these patients are summarized in Table 1. Six patients were hospitalized at UNM Hospital, and 2 patients were hospitalized at Texas Tech University Health Sciences Center. For these patients, the database was most complete. The 2 patients who are not included in Table 1 died, and their information was limited; they were 14 and 15 years old.10
A previously healthy 11-year-old girl from rural Arizona (patient 4 from Table 1) presented to a regional hospital emergency department with a 2-day history of headache, myalgia, chest pain, sore throat, and fever. She was afebrile at the time of the visit. A throat culture was obtained, and she was discharged to home. She returned the following day with the same complaints and increasing shortness of breath. There were no ill contacts and no known rodent exposure. The patient's medical history was unremarkable.
At the emergency department on the day of admission, she was alert and in moderate distress with nasal flaring. She had a temperature of 39.0°C, pulse of 140 beats/min, respiratory rate of 64 breaths/min, and a blood pressure of 132/70 mmHg. Her percutaneous oxygen saturation was 61% in room air. She had markedly decreased breath sounds bilaterally, with mild retractions. Her abdominal examination was unremarkable. Her WBC was 15 300/mm3, her hematocrit was 46%, and her platelet count was 74 000/mm3. The WBC differential was 52% neutrophils, 27% lymphocytes, 13% bands, and 7% monocytes. Her chest radiograph revealed bilateral interstitial infiltrates. An initial arterial blood gas while on 3 L/min flow of oxygen by nasal canula showed the following: pH 7.43, pCO2 32 torr, pO2 61 torr, and HCO3 21 mmol/L.
She was intubated before transfer to UNM Hospital with a diagnosis of possible HCPS. Gram stain of a tracheal aspirate obtained at the time of intubation showed few white blood cells and many Gram-negative coccobacilli and diplococci suggestive of Haemophilus influenza infection. Before her transfer, she was treated empirically with nebulized albuterol and intravenous methylprednisolone, ceftriaxone, gentamicin, and erythromycin.
On admission to UNM Hospital, her temperature was 36.9°C, her heart rate was 96 beats/min, and her blood pressure was 120/55 mmHg. Her percutaneous oxygen saturation was 94% on a volume-controlled ventilator with a positive end expiratory pressure of 5, tidal volume of 500 mL (10 mL/kg), respiratory rate of 30, and an FIO2 of 0.60. Her physical examination was significant for tachycardia and diffuse rales bilaterally.
A right pleural effusion and bilateral interstitial infiltrates were present on the admission chest radiograph. Laboratory studies on admission revealed a WBC of 11 400/mm3, hematocrit of 35%, platelet count of 74 000/mm3, an elevated serum aspartate aminotransferase of 109 IU/L (normal: 5–35), and alanine aminotransferase of 71 IU/l (normal: 5–35). Her serum lactate dehydrogenase was elevated at 1268 IU/L (normal: 300–600). Peripheral blood smear analysis revealed thrombocytopenia, >10% circulating immunoblasts among the lymphoid series, and a left shift in the granulocytic series without significant toxic changes. These features were consistent with hantavirus infection. A Western blot assay done on admission was positive for immunoglobulin M and immunoglobulin G antibodies against SNV. The positive serologic finding was confirmed by polymerase chain reaction analysis, which revealed circulating SNV ribonucleic acid.
The sputum culture from the referring facility grewMoraxella species. Blood and urine cultures from the referring facility were found to be negative. A repeat sputum culture from UNM hospital sent on the day of admission found normal oral flora. Nasopharyngeal swabs for respiratory syncytial virus, influenza A and B, parainfluenza, and adenovirus detected by fluorescent antibodies were negative.
Antibiotics were discontinued late on the day of admission when results from the Western blot were available. The patient made a rapid recovery. She was extubated to a face mask that delivered inspired oxygen of 40% on the second hospital day and was weaned to oxygen by nasal canula later that day. A chest radiograph from that day found minimal residual patchy air space disease. By 2 days after her admission, her thrombocytopenia also had resolved, to a platelet count of 164 000/mm3.
She was transferred back to the referring hospital on the fourth hospital day. At that time, she was receiving oxygen by nasal canula at 1 L/min flow with a percutaneous oxygen saturation of 91%.
Among the 10 patients for whom information was available, the mean duration of symptoms before hospitalization was 3.5 days (median: 3.5; range: 1–7). The most common symptoms, each present in at least 80% of patients at the time of admission, were, in descending order of frequency, fever, headache, nausea or vomiting, cough, shortness of breath, and myalgia (Table 2). All patients had respiratory complaints of either cough or shortness of breath. No patients had complaints of rhinorrhea or nasal congestion, although 4 (40%) complained of sore throat.
At the time of hospital admission, 6 (55%) of 11 patients were hypoxemic with percutaneous oxygen saturations below 90% in room air; 2 patients required oxygen by nasal canula, and 4 patients required mechanical ventilation either before admission or shortly thereafter. The most common physical examination findings on admission were tachypnea and fever (Table 3). Seven patients (78%) had respiratory findings of either tachypnea or rales at admission. Hypotension and tachycardia were relatively uncommon findings on admission, seen in 33% and 22% of patients, respectively. No patient had purpura or petechial rash, evidence of mucosal bleeding, or peripheral or periorbital edema.
Thrombocytopenia was observed at admission in all of 11 patients (100%) for whom this information was available (Table 4). Leukocytosis and hemoconcentration were less common, each present in 3 of 11 patients (27%). Of the 10 patients with differential WBC at the time of admission, 6 (60%) had at least 10% band forms, 3 (30%) had metamyelocytes, and 4 (40%) had atypical lymphocytes. Other laboratory abnormalities commonly seen at admission included elevated levels of lactate dehydrogenase, aspartate aminotransferase, and alanine aminotransferase and hypoalbuminemia (Table 4).
Three HCPS patients who were admitted to UNM hospital had peripheral blood smear analysis. All had thrombocytopenia, >10% circulating immunoblasts, and left shift in the granulocytic series without toxic changes.
Five patients had an initial urinalysis at the time of admission. The median urine specific gravity was 1.029 g/mL (range: 1.013 to 1.041). Three of 5 patients had proteinuria (≥2+) on admission. Urine dipsticks were positive for blood in 2 of 5 patients; microscopic examination revealed <3 red cells per high-power field for both.
Initial chest radiographs for 10 patients revealed interstitial or interstitial and alveolar infiltrates in 5 patients (50%), Kerley B lines or fluid in the fissures in 2 (20%), fluffy alveolar infiltrates in 1 (10%), and normal radiographs in 2 (20%). The 2 patients with initial normal chest radiographs developed interstitial edema within 48 hours.
Of the 13 patients, 5 became ill during the spring, 2 in the summer, 4 in the fall, and 2 patients (patients 9 and 10 from Table 1) in the winter. The majority of patients were previously healthy and without medical problems. One patient was taking erythromycin at the time of admission for acne vulgaris, and another patient had a history of asthma.
Three patients were examined by medical providers and discharged to home with mistaken provisional diagnoses before returning and being admitted. All survived. Two of the 3 developed respiratory failure and required mechanical ventilation.
In 12 (92%) of the 13 patients reviewed , HCPS developed. One patient did not have an oxygen requirement and so failed to meet the CDC clinical case description as described previously. He was a 12-year-old from a rural Arizona town and had a febrile illness characterized by prominent abdominal pain, nausea and vomiting, headache, myalgia, cough, and sore throat. Because of the patient's rural location, history of mice around the home, and presenting symptoms, the patient was tested for SNV infection.
Of the 12 patients with HCPS, 4 (33%) required oxygen by nasal canula and 8 (67%) required mechanical ventilation. For the patients who required mechanical ventilation, the average time from the onset of symptoms to endotracheal intubation was 3.8 days (median: 4.5; range: 1–6). Patients who were intubated but not treated with extracorporeal membrane oxygenation (ECMO) were intubated for an average of 4.6 days (median: 4; range: 2–8). Hypotension necessitating support with vasoactive infusions developed in 5 of the 10 HCPS patients for whom this information was available.
Two patients received ECMO support for hemodynamic deterioration despite resuscitation with fluids and vasoactive medications and mechanical ventilation. ECMO therapy was initiated after 1 day and 3 days of symptoms, respectively, for the 2 patients who received this treatment. The patient who received ECMO and survived required ECMO for 8 days and mechanical ventilation for 20 days. The other ECMO patient died after 7 days of ECMO support as a result of brain death caused by a prolonged cardiac arrest before ECMO initiation.
Only 1 patient developed significant bleeding during hospitalization; she developed a hemothorax as a complication of thoracentesis for pleural effusion. Although not common at the time of admission, leukocytosis eventually was seen in 7 of 11 patients (64%) during hospitalization, and hemoconcentration was seen in 6 of 11 patients (55%).
The clinical course of the 4 HCPS patients who died was characterized by pulmonary edema, hypotension, and ventricular arrhythmias. The ECMO case fatality was described above. One patient died en route to a hospital, and 2 patients died despite standard critical care. Of the group who survived, the average hospital stay was 9.9 days (median: 8; range: 3–28) to discharge to home or transfer (in 2 cases) to a regional hospital. Those who survived were without sequelae.
Predictors of Mortality
An elevated prothrombin time (≥14 seconds) at admission was associated with a fatal outcome (P = .04, Fisher's exact test). An elevated WBC (>13.5 × 103/mm3) on admission showed only a trend toward significance in association with mortality (P = .06, Fisher's exact test) as did age ≥14 years, the median age of our patients (P = .07, Fisher's exact test). Symptoms before hospitalization, duration of symptoms, specific physical examination findings on admission including hypoxemia, and other laboratory findings were not associated with mortality.
Predictors of Respiratory Failure
Hypotension at admission was associated with respiratory failure requiring mechanical ventilation (P = .02, Fisher's exact test) as was the absence of fever at admission (P= .05, Fisher's exact test). Symptoms before hospitalization, duration of symptoms, other physical examination findings on admission including presence of hypoxemia, and laboratory findings were not associated with respiratory failure.
In this case series of pediatric patients who were 10 to 16 years of age and infected with SNV, the clinical outcomes did not differ greatly from those described in adult cases. The case fatality ratio of 33% (4 of 12) for these pediatric patients with HCPS is comparable to the 38% case fatality rate (105 deaths in 274 cases) described for HCPS overall in the United States (J. Young, Special Pathogens Branch, CDC, personal communication, November 2000). As in adult cases, the majority of the 12 pediatric HCPS patients described herein (8 of 12 [67%]) were critically ill and progressed to respiratory failure. Consistent with the newer designation of hantavirus cardiopulmonary syndrome, at least half of the patients developed cardiogenic shock and required inotropic support.
The most frequent prodromal symptoms of our pediatric patients, particularly fever, headache, myalgia, and respiratory and gastrointestinal complaints, are comparable to those described in adults with HCPS. One exception is the common complaint of sore throat, seen in almost half of our patients, which has been described as an infrequent symptom among adults with HCPS.14,,15 Typical clinical laboratory findings early in the course include thrombocytopenia and elevated liver enzymes and lactate dehydrogenase. A left shift in the granulocytic series without toxic changes is often present, but leukocytosis and hemoconcentration are relatively late findings, observed in >50% of cases during the course of illness.
The differential diagnosis for pediatric patients who present with fever, headache, myalgia, and respiratory and gastrointestinal symptoms is broad. Viral and bacterial pneumonia, sepsis syndrome with adult respiratory distress syndrome, and acute gastroenteritis are among the more likely clinical syndromes. The authors are aware of patients admitted to rule out HCPS and who subsequently had the diagnosis of viral respiratory illness (eg, respiratory syncytial virus), streptococcal pharyngitis, and sepsis attributable to S aureus. Depending on the region of North America where the patient may present and the exposure history, the differential diagnosis may be broad, including septicemic plague or tularemia, ehrlichiosis, leptospirosis, Colorado tick fever, relapsing fever, or (“spotless”) Rocky Mountain spotted fever.
Because the initial prodrome is nonspecific, clinically diagnosing pediatric HCPS with either mild disease or in the early prodrome phase presents a diagnostic challenge. SNV infection should be considered in pediatric patients from rural areas, especially in western North America, who present with fever, headache, myalgia, and respiratory and gastrointestinal symptoms, particularly if there is a history of possible rodent exposure. Infection is most common from spring through fall.
If HCPS is suspected, then a complete blood count with platelet count should be obtained. Thrombocytopenia is a key laboratory feature of HCPS. If thrombocytopenia or a rapidly decreasing platelet count is found, peripheral blood smear analysis and serology testing should be performed. At UNM hospital, a peripheral blood smear with 4 of the 5 criteria (thrombocytopenia, >10% circulating immunoblasts among the lymphoid series, left shift of granulocytic series, without toxic changes, and hemoconcentration) has been found to have a positive predictive value of 90% for HCPS. All cases with 5 of the 5 criteria have been confirmed serologically (K. Foucar, Department of Pathology, UNM School of Medicine, personal communication, September 2000). Pending results, patients need to be monitored closely for signs of cardiopulmonary compromise, which can develop rapidly with the onset of pulmonary edema. There is not yet a reliable early indicator as to which patients will develop more severe disease.
Because of the high proportion of HCPS patients who become critically ill and the rapid deterioration seen in many HCPS patients, we believe that early transfer to a tertiary care center that is capable of providing critical care and ECMO support should be strongly considered. In our experience, ECMO seems to be beneficial in the support of critically ill patients with severe HCPS,16 including pediatric patients. ECMO has been used in the treatment of 26 HCPS patients at UNM Hospital with a survival rate of 69% (M.R. Crowley, unpublished data). Criteria for the initiation of ECMO at UNM Hospital include parameters consistent with 100% mortality from our experience with HCPS.
HCPS is an uncommon serious viral zoonosis that causes respiratory failure and cardiovascular instability in children and carries a high case fatality rate of 33%. HCPS in pediatric patients has a similar presentation and outcome to that described in adults. HCPS occurs mainly in the rural western United States, but cases have been reported nationwide. Because of the high mortality and often fulminant deterioration of pediatric HCPS patients, medical providers who care for children should familiarize themselves with the clinical features of HCPS. Early recognition of this disease with prompt referral to tertiary care centers that have experience with HCPS likely will improve outcome and reduce mortality.
This study was supported by Public Health Service Grant RO1 AI 41692 and by the Defense Advanced Research Projects Agency.
We thank J. Rawlings, F. Koster, R. Servi, D. Goade, J. Young, and J. Hutchinson for their help in collecting data; C. Qualls for assistance with statistics; and TriCore Laboratories for technical assistance.
- Received June 28, 2000.
- Accepted April 4, 2001.
Reprint requests to (B.H.) Department of Pathology, University of New Mexico School of Medicine, Albuquerque, NM 87131. E-mail:
- HCPS =
- hantavirus cardiopulmonary syndrome •
- SNV =
- Sin Nombre hantavirus •
- CDC =
- Centers for Disease Control and Prevention •
- UNM =
- University of New Mexico •
- WBC =
- white blood cell count •
- ECMO =
- extracorporeal membrane oxygenation
- ↵American Academy of Pediatrics. Hantavirus cardiopulmonary syndrome. In: Pickering LK, ed. 2000 Red Book: Report of the Committee on Infectious Diseases. 25th ed. Elk Grove Village, IL: American Academy of Pediatrics; 2000:272–274
- Zavasky DM,
- Hjelle B,
- Peterson MC,
- et al.
- Khan AS,
- Ksiazek TG,
- Zaki SR,
- et al.
- Rawlings JA,
- Torrez-Martinez N,
- Neill SU,
- et al.
- Hjelle B,
- Jenison S,
- Torrez-Martinez N,
- et al.
- Jenison S,
- Yamada T,
- Morris C,
- et al.
- Bharadwaj M,
- Nofchissey R,
- Goade D,
- Koster F,
- Hjelle B
- Moolenaar RL,
- Dalton C,
- Lipman HB,
- et al.
- Copyright © 2001 American Academy of Pediatrics