- Email: [email protected]
Polymerase chain reaction diagnosis of primary human herpesvirus-6 infection in the acute care setting
POLYMERASE CHAIN REACTION DIAGNOSIS OF PRIMARY HUMAN HERPESVIRUS-6 INFECTION IN THE ACUTE CARE SETTING DANIELLE M. ZERR, MD, MPH, LISA M. FRENKEL, MD, MEEI-LI HUANG, PHD, MARGARET RHOADS, BA, LONG NGUY, BS, MARK A. DEL BECCARO, MD, AND LAWRENCE COREY, MD
Objective To evaluate the potential utility of identifying primary human herpesvirus (HHV)-6 infection in an emergency department setting by determining the frequency of HHV-6 viremia, diagnostic testing, and empiric treatment of serious bacterial infection (SBI) in HHV-6 viremic children, and concurrent SBI and HHV-6 viremia. Study design Children under age 2 years and who had a blood specimen taken for evaluation of fever were tested for HHV-6 by polymerase chain reaction (PCR). HHV-6 viremia was defined as detection of HHV-6 DNA in acute plasma. Results A total of 32 of the 181 subjects (18%) had HHV-6 viremia. Children with HHV-6 viremia frequently underwent procedures for diagnosis and empiric treatment of SBI: 60% had bladder catheterizations, 6% had lumbar punctures, 47% had radiographs, 32% received empiric antibiotics, and 34% were hospitalized. Four of the 32 children with HHV-6 viremia (12.5%) were diagnosed with SBI, although none had a positive culture of blood or cerebrospinal fluid. Conclusions Rapid diagnosis of HHV-6 viremia may not serve to adequately differentiate infants with and without SBI in acute care settings. Although no children with HHV-6 viremia had bacteremia or meningitis, it appears that additional criteria are needed to increase the specificity of HHV-6 PCR testing before withholding evaluation for SBI. (J Pediatr 2006;149:480-5)
oung children evaluated for fever in emergency departments (EDs) frequently undergo laboratory testing and receive empiric antibiotics, and some are hospitalized for observation out of concern for serious bacterial infection (SBI). Only 3% to 11% of young febrile children are typically found to have a SBI,1 and as such, alternative etiologies for the fever are common. HHV-6 infects ⬎ 90% of children by 2 years of age.2 Primary infection with HHV-6 frequently results in physician visits3 and has accounted for up to 21% of ED visits for fever in young children.4 Children generally recover uneventfully from primary HHV-6 infection and have not been observed to develop concurrent bacterial infection. For example, a study of 160 children with primary HHV-6 infection evaluated in an ED found that none had SBI.4 These data From the Departments of Pediatrics and suggest that a specific and sensitive test for primary HHV-6 infection might offer a Laboratory Medicine, University of Washington, Children’s Hospital and Regional diagnostic tool in the ED to reduce the need for other testing, antibiotic treatment, and Medical Center, and Fred Hutchinson Canhospitalization. cer Research Center, Seattle, WA. A natural history study of HHV-6 infection demonstrated that HHV-6 DNA levels This clinical research was conducted in accordance with guidelines for human experin saliva are low or undetectable early in the course of primary infection and subsequently imentation as specified by the Children’s 3 increase, reaching a plateau that is maintained for months. As such, a single determiHospital Institutional Review Board. nation of salivary HHV-6 alone is not a useful marker of primary infection. Conversely, None of the authors has any potential conflict of interest, real or perceived. detection of HHV-6 DNA in plasma, serum, or whole blood has been associated with Supported by in part by National Institutes 5-9 primary infection in immunocompetent children. of Health grant K23 AI001679 and by a We sought to determine the frequency of HHV-6 plasma viremia detected by grant from Children’s Hospital Fund for Excellence. polymerase chain reaction (PCR) in young children presenting to an ED with fever, to Submitted for publication Nov 7, 2005; last evaluate to what degree invasive medical testing and management takes place in these revision received Mar 15, 2006; accepted children, and to determine whether SBI occurs in children with HHV-6 plasma viremia. May 17, 2006.
Y
ED HHV IQR
480
Emergency department Human herpesvirus Interquartile range
PCR SBI
Polymerase chain reaction Serious bacterial infection
Reprint requests: Dr. Danielle M. Zerr, Children’s Hospital, W-8851, 4800 Sand Point Way NE, Seattle, WA 98105. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2006 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2006.05.027
METHODS Subjects All patients presenting to the ED of the Children’s Hospital and Regional Medical Center, Seattle between 8:00 AM and midnight from January 2001 through June 2002 who had a temperature of 38°C or higher, were under age 2 years, and were having blood drawn to evaluate for bacterial infection were potentially eligible for enrollment. Children were excluded from the study if they had been diagnosed with an immune deficiency or cancer or were receiving immunosuppressive medications, or if 1 parent was not able to communicate in English. Potentially eligible children were approached by the ED staff. If the parents were interested in the study, the ED staff summoned study personnel to the ED to further screen and, when appropriate, enroll the infant into the study. A total of 12,344 patients under age 2 years were brought to the ED between 8 AM and midnight during the study period; of these, 1639 had a diagnosis that included fever. Of these, 683 had a blood culture obtained, a marker for having blood drawn to evaluate for fever. Study personnel evaluated 252 patients who were screened by ED staff and expressed interest in the study. Of these, 182 children were enrolled. Of the 70 children not enrolled, 12 (17%) were ineligible, 54 (77%) had parents who declined participation for various reasons, and 4 (6%) lived out of the area or were scheduled to move before follow-up could be completed. After obtaining written informed consent from the subject’s parents, demographic and clinical information was collected, and blood and saliva were obtained for HHV-6 testing. Saliva also was collected approximately weekly by the subject’s parents during the 8 weeks after enrollment into the study. Saliva was collected using Sno strips as described previously.3 Definitions ● ●
HHV-6 viremia: Detection of HHV-6 DNA in the acute plasma specimen by real-time PCR. Viral shedding in saliva consistent with primary HHV-6: Absent or low levels (⬍ 12,000 copies of HHV-6 DNA/ mL) of HHV-6 DNA at the time of acute illness, followed by increasing levels in subsequent specimens, including at least 1 sample with ⬎ 1000 copies/mL. No more than 2 negative samples could be present between the first sample and the subsequent positive sample; however, temporal gaps in saliva samples were allowed. The HHV-6 saliva pattern was determined only for subjects who had a saliva specimen during acute illness and at least 1 follow-up specimen unless the acute saliva sample had ⬎ 12,000 copies/mL, indicating established infection. This definition was derived from a natural history study of primary HHV-6 infection that documented initial low-level HHV-6 DNA levels in saliva that increased and established a plateau over the weeks and months after primary infection.3 In the
●
natural history study, 82% of first-positive saliva samples in subjects with primary HHV-6 infection had ⬍ 12,000 HHV-6 DNA copies/mL, whereas 80% of samples from subjects with established HHV-6 infection had ⬎ 12,000 copies DNA/mL and 87% of samples from subjects with established HHV-6 infection had ⬎ 1000 DNA copies/ mL. SBI: Physician-diagnosed and treated bacteremia/sepsis, meningitis, urinary tract infection, pneumonia, bone and joint infection, or bacterial gastroenteritis, as documented in the medical record.
PCR Analysis Plasma and saliva specimens were batched to test for HHV-6. DNA was extracted from the samples using Qiagen columns.10,11 Saliva, collected with Sno strips, was processed as described previously.3 DNA was extracted from 100 L of plasma and eluted into 50 L of buffer, of which 20 L was submitted to PCR analysis. Thus, for each subject, DNA from 40 L of plasma was tested for HHV-6. Detection of HHV-6 DNA was accomplished with a real-time fluorescent probe quantitative PCR assay, as described previously.10,11 HHV-6 DNA-positive samples with sufficient DNA were then further typed as HHV-6 A or B.11 The investigator performing the PCR analysis (LN) was blinded to the subjects’ identity and clinical findings. PCR results were not shared with the subjects’ clinical providers. Statistical Analysis Statistical analyses were performed using SPSS version 10.1 (SPSS Inc, Chicago, IL). Proportions were compared between groups using Pearson’s 2 test or, if the assumptions for the 2 test were not met, Fisher’s exact test. Continuous variables were compared between groups using the MannWhitney U test.
RESULTS Plasma HHV-6 PCR Of the 182 subjects who were eligible and enrolled in the study, the plasma specimens from 181 were adequate and included in the study. Thirty-two (18%) of these 181 subjects had HHV-6 DNA detected in plasma, or HHV-6 viremia. Their median plasma viral load was 7530 copies/mL (range, 90 to 699,750 copies/mL; interquartile range [IQR], 1298 to 21,393 copies/mL). Saliva HHV-6 PCR Saliva specimens during acute illness were obtained from 178 of the 181 subjects, and at least 1 additional saliva specimen was collected from 139 patients. Twenty of the 31 viremic subjects with an acute saliva sample (65%) had detectable HHV-6 DNA in the acute saliva, versus 41 of the 147 nonviremic patients with acute saliva specimens (28%). There were 150 children (29 with viremia and 121 without
Polymerase Chain Reaction Diagnosis Of Primary Human Herpesvirus-6 Infection In The Acute Care Setting
481
Figure 1. Salivary HHV-6 DNA levels over time in (A) 31 subjects with HHV-6 plasma viremia versus (B) 41 subjects without HHV-6 plasma viremia but with detectable HHV-6 in their acute saliva specimen and thus likely established HHV-6 infection. Each box plot represents the median (middle line) and IQR (outer lines), and each “I” bar represents the highest and lowest values, excluding the outliers. The number of children providing samples for PCR each week is shown below the x-axis.
viremia) who had adequate saliva specimens to enable detection of a salivary pattern characteristic of primary HHV-6 infection, that is, low salivary levels of HHV-6 DNA in their initial specimen, followed by increasing levels (see Methods).3 Twenty-one of the 29 viremic subjects (72%), versus 10 of the 121 nonviremic subjects (8%) (P ⬍ .001), had a salivary pattern consistent with primary infection. To explore whether quantifying acute salivary levels might facilitate the diagnosis of primary HHV-6, salivary HHV-6 patterns were compared between the 31 subjects who had viremia and saliva specimens available and the 41 subjects who likely had established infection (ie, those without HHV-6 viremia but with HHV-6 DNA detected in their acute saliva specimen) (Figure 1). Subjects with likely established infection had relatively high levels of HHV-6 DNA in their initial saliva specimens (median HHV-6 DNA level, 42,800 copies/mL; IQR, 5160 to 206,890 copies/mL) that remained relatively stable over time, whereas subjects with HHV-6 viremia had lower initial levels (median, 940 copies/ mL; IQR, 0 to 11,690 copies/mL) that increased over time (P ⬍ .001 for comparison of initial levels of salivary HHV-6 DNA). In these 72 subjects, HHV-6 DNA level in plasma was inversely correlated with the HHV-6 DNA level in the acute saliva sample (Figure 2; Pearson correlation P ⬍ .001).
HHV-6 Type HHV-6 type was determined in the 63 subjects (28 viremic and 35 nonviremic) with sufficient sample DNA available for testing. Only type B HHV-6 was identified in these subjects; no type A virus was found. Clinical Presentation of Children With HHV-6 Viremia The children with HHV-6 viremia were significantly older than those without viremia and more likely to have rash 482
Zerr et al
Figure 2. Scatterplot of HHV-6 DNA plasma level by corresponding HHV-6 DNA acute saliva level in 31 subjects with HHV-6 plasma viremia and 41 subjects with likely established HHV-6 infection (without HHV-6 plasma viremia but with detectable HHV-6 in their acute saliva specimen). Levels ⬍ 10 copies/mL actually represent 0. A small random number was added to the 0 so that they could be easily visualized on the graph rather than falling on the axes. A linear regression line demonstrates the inverse correlation between the 2 variables. , Viremia; ⫹, Established.
characterizing their illness at the time of presentation. There were no differences in the frequency of rhinorrhea, cough, vomiting, diarrhea, seizure, and otitis media between the 2 groups (Table I). The children with HHV-6 viremia were more likely to have lower platelet and white blood cell counts but higher hematocrit levels than those without HHV-6 viremia.
Diagnostic Evaluation and Management for SBI The children with HHV-6 viremia underwent procedures for diagnosis of SBI: 60% underwent bladder catheterThe Journal of Pediatrics • October 2006
Table I. Demographic and clinical factors among children with and those without HHV-6 viremia HHV-6 viremia
Demographic information Median age in months (IQR) Age ⬎ 2 months Male Nonwhite Enrolled fall/winter Symptoms before enrollment Median maximum temperature (IQR) Maximum temperature ⬎39.3°C* Rhinorrhea Cough Vomiting Diarrhea Rash Seizure Otitis media Laboratory values at enrollment Median WBC count (IQR) WBC count ⬍5,000/mm3† Median platelet count (IQR) Platelet count ⬍300,000/ mm3‡ Median hematocrit (IQR) Hematocrit ⬎31.5%§
Yes (n ⴝ 32)
No (n ⴝ 149)
8.3 (4.1, 14.7) 29 (91%) 18 (56%) 10 (31%) 16 (50%)
4.5 (1.5, 10.8) 99 (66%) 88 (59%) 55 (37%) 88 (59%)
39.8 (39.4, 40.2) 26 (81%) 17 (55%) 17 (53%) 14 (44%) 13 (41%) 14 (44%) 1 (3%) 3 (9%)
39.1 (38.6, 39.7) 55 (43%) 74 (50%) 78 (52%) 60 (40%) 49 (33%) 34 (23%) 6 (4%) 17 (11%)
9.8 (5.6, 12.4) 6 (21%) 341 (222, 422) 13 (46%) 35 (33, 36) 26 (93%)
11.9 (8.1, 16.7) 9 (7%) 431 (317, 526) 28 (21%) 33 (31, 35) 97 (71%)
P value .009 .006 .77 .53 .35 ⬍.001 ⬍.001 .72 .94 .72 .40 .02 .81 .74 .03 .02 .002 .004 .02 .02
*Maximum temperature of 39.3°C represents the 50th percentile derived from the 159 subjects with data. †WBC count of 5,000/mm3 represents the 10th percentile derived from the 165 subjects with data. ‡Platelet count of 300,000/mm3 represents the 25th percentile derived from the 164 subjects with data. §HCT of 31.5% represents the 25th percentile derived from the 165 subjects with data.
ization, and 47% underwent radiographic examination. In addition, 32% received empiric antibiotics, and 34% were hospitalized (Table II). The children hospitalized with HHV-6 viremia ranged in age from 1 to 22 months, and the duration of hospitalization ranged between 1 and 4 days. Diagnoses given as rationales for hospitalization included fever, fever and hematochezia, presumed viral thrombocytopenia, suspected group A streptococcal infection, viral infection, viral pneumonia, viral gastroenteritis and secondary dehydration (n ⫽ 2), bacterial pneumonia, and pyelonephritis (n ⫽ 2).
SBI in Children With HHV-6 Four of the 32 children with HHV-6 viremia (12.5%; 95% confidence interval, 3% to 29%) were diagnosed with an SBI, although none of these children had a positive blood culture (Table II). The 4 cases of concurrent SBI and HHV-6 viremia are as follows: ●
A 15-month-old toddler was diagnosed with gastroenteritis. The illness was characterized by bloody diarrhea, and stool culture was positive for Campylobacter jejuni. The acute plasma had 100 copies/mL of HHV-6 DNA, whereas the acute saliva had 42,000 copies/mL. Saliva results over time were not consistent with primary infection.
●
●
An 11-month-old infant was diagnosed with urinary tract infection. The child was diagnosed with roseola by the primary care provider 2 weeks before enrollment. Antibiotics were initiated for otitis media 3 days before enrollment. The child was brought to the ED with fever, vomiting, and decreased oral intake and urine output. Evaluation at enrollment revealed abnormal urinalysis (11 to 20 white blood cells/highpower field and many bacteria), peripheral white blood cell count was 26,000/mm3, and C-reactive protein level was 31.4 mg. Urine culture was obtained after administration of antibiotics and was negative. The acute plasma specimen had 177 copies/mL of HHV-6 DNA; the acute saliva specimen had 0 copies/mL. Saliva results over time were consistent with primary infection. An 8-month– old infant was diagnosed with urinary tract infection. Seven days before enrollment, the child had a fever, followed 5 days later by resolution of fever and onset of rash. The fever subsequently returned, and the child was brought to the ED. Urinalysis revealed 0 to 5 white blood cells/high-power field and moderate bacteria, and a catheterized urine specimen yielded ⬎ 103 colony-forming units/mL of Klebsiella pneumoniae. The acute plasma specimen had 2678 copies/mL of HHV-6 DNA, whereas the acute saliva specimen had 0 copies/mL. Saliva results over time were consistent with primary infection.
Polymerase Chain Reaction Diagnosis Of Primary Human Herpesvirus-6 Infection In The Acute Care Setting
483
Table II. Management and outcomes of children with and without HHV-6 viremia HHV-6 viremia
Diagnostics/management Bladder catheterization Lumbar puncture Blood culture X-ray Empiric antibiotics* Hospitalized Follow-up visit with primary physician Diagnosed with serious bacterial infections† Bacteremia/sepsis Meningitis Urinary tract infection Pneumonia Gastroenteritis Septic arthritis Mastoiditis
Yes (n ⴝ 32)
No (n ⴝ 149)
P value
19 (60%) 2 (6%) 29 (91%) 15 (47%) 10 (32%) 11 (34%) 19 (66%) 4 (13%) 0 0 2 (6.3%) 1 (3.1%) 1 (3.1%) 0 0
95 (65%) 49 (33%) 130 (87%) 78 (52%) 51 (34%) 81 (54%) 105 (86%) 34 (23%) 4 (2.7%) 1 (0.7%) 20 (13.4%) 5 (3.4%) 2 (1.3%) 1 (0.7%) 1 (0.7%)
.58 .002 .60 .57 .83 .04 .009 .19
*“Empiric antibiotics” was defined as antibiotics given empirically for the possibility of serious bacterial illness. This did not include antibiotics given for diagnosed bacterial infections such as otitis media, pneumonia, etc. †See text for clinical details of the four patients with SBI.
●
A 5-month-old child was diagnosed with pneumonia. Seven days before enrollment, the child developed fever, followed 2 days later by resolution of fever and onset of rash. Two days before enrollment, fever returned with worsening cough and mild respiratory distress (grunting, increased work of breathing). Chest radiography revealed no focal infiltrate. The acute plasma specimen had 1520 copies/mL of HHV-6 DNA; the acute saliva specimen had 5080 copies/mL. Saliva results over time were consistent with primary infection.
Three of the 4 children described above experienced illness consistent with roseola (ie, fever followed by rash with defervescence) in the 1 to 2 weeks before their enrollment into the study, suggesting recent primary HHV-6 infection, and they also had salivary HHV-6 patterns that were consistent with recent infection.
DISCUSSION We found HHV-6 DNA plasma viremia, suggesting HHV-6 primary infection, in 18% of young febrile children evaluated in the ED for bacterial infection. Almost 1/3 of children with HHV-6 viremia received empiric antibiotics, 34% were hospitalized, and 13% were diagnosed with SBI. The frequency of HHV-6 viremia observed is consistent with the observation by Hall et al4 that HHV-6 accounts for 10% to 21% of ED visits among young febrile children. That subjects with HHV-6 viremia in our study frequently underwent invasive medical testing, received empiric antibiotics, and were hospitalized suggests a potential role for rapid HHV-6 diagnostic tests in reducing these costly interventions. Developing useful rapid diagnostic tests may be com484
Zerr et al
plicated, however, because, in contrast to the findings of Hall et al,4 SBI was diagnosed in 4 (13%) of our subjects with HHV-6 viremia. Although these children did not have positive cultures of blood or cerebrospinal fluid, they did have clinical and/or microbiological evidence of bacterial infection. Clinical history and HHV-6 results in these children suggest that primary HHV-6 infection may have occurred in the weeks just before enrollment in our study. First, in 3 of the 4 subjects, the clinical history provided by the parents suggested that roseola, and thus possibly primary HHV-6 infection, occurred 1 to 2 weeks before presentation to the ED. Moreover, the salivary HHV-6 patterns in these 3 children were consistent with primary HHV-6 infection. Finally, the level of HHV-6 DNA in the plasma of 2 of the 4 children was relatively low (⬍ 200 copies/mL). Although it is possible that the low levels were falsely positive, it is also possible that a very sensitive PCR of plasma may have detected remaining viral nucleic acids after recent primary infection. A study of febrile infants age 90 days and younger, applying the same PCR system as used in the present study, also demonstrated HHV-6 viremia coincident with an SBI.12 In that study, 5 of 47 infants (10.6%) had HHV-6 DNA detected in serum by PCR, and 1 of those 5 HHV-6 –positive infants had Escherichia coli bacteremia and urinary tract infection. Similar to children with HHV-6 viremia and SBI in our study, the HHV-6 DNA level in the serum of the infant with E. coli bacteremia was relatively low compared with the levels in his study-mates with HHV-6 viremia and no SBI. To allow clinicians to comfortably recommend withholding laboratory testing and empiric antibiotics from young febrile children, a rapid diagnostic test with high specificity for primary HHV-6 infection without bacterial infection is The Journal of Pediatrics • October 2006
needed. It appears that real-time PCR of plasma is a sensitive means of identifying primary HHV-6 infection but that it also carries a lower specificity for early infection compared with an approach involving viral culture and seroconversion.4 It is possible that quantifying a threshold level of HHV-6 DNA in acute plasma, or a combination of acute plasma and acute saliva DNA threshold levels, could deliver this specificity; however, sensitivity for detecting primary HHV-6, and thus the “payoff” of testing, would likely diminish. Constraining the target population could also improve the performance of any diagnostic test for primary HHV-6. For instance, in our study, had we tested only children with fever without obvious source (eg, children without respiratory distress or bloody diarrhea), then 2 of the 4 children with HHV-6 viremia and concomitant SBI would have been excluded. Although too many such constraints risk the requirement of cumbersome decision analyses, it may be practical to consider such an approach in young children who have fever without a source, a population that frequently receives diagnostic evaluation, empiric antibiotics, and hospitalization out of concern for SBI. In summary, our preliminary study indicates that PCR of plasma can identify primary HHV-6 infection, a common illness among children in the ED, and that a high proportion of children with HHV-6 viremia may undergo diagnostic testing, receive empiric antibiotics, and be hospitalized out of concern for SBI. Our results also suggest that HHV-6 viremia as detected by PCR may extend for 1 to 2 weeks beyond the onset of HHV-6 infection. Given this prolonged detection of HHV-6 DNA and the high attack rate of HHV-6, it should be expected that events of relatively high frequency, such as SBI in young children, will at times be coincident or in rapid succession with primary HHV-6 and HHV-6 viremia. Clinical and/or virologic criteria are needed to increase the specificity of DNA PCR testing for early primary HHV-6 infection before it can be used to withhold evaluation and therapy for SBI. We are grateful to the families who participated in this study. We also thank the nurses and physicians in the Children’s Hospital
ED for identifying study subjects; to Rena Bornemann, Jaime Nell Manansala, Amanda Allpress, and Laurie Newman for their help in enrolling those subjects; and to Dimitri Christakis for his helpful review of the manuscript.
REFERENCES 1. Baraff LJ, Bass JW, Fleisher GR, Klein JO, McCracken GH, Powell KR, et al. Practice guideline for the management of infants and children 0 –36 months of age with fever without source. Ann Emerg Med 1993;22:1198-210. 2. Okuno T, Takahashi K, Balachandra K, Shiraki K, Yamanishi K, Takahashi M, et al. Seroepidemiology of human herpesvirus 6 infection in normal children and adults. J Clin Microbiol 1989;27:651-3. 3. Zerr DM, Meier AS, Selke SS, Frenkel LM, Huang M, Wald A, et al. Virologic and clinical characteristics of primary human herpesvirus 6: a population-based study. N Engl J Med 2005;352:768-76. 4. Hall CB, Long CE, Schnabel KC, Caserta MT, McIntyre KM, Costanzo MA, et al. Human herpesvirus-6 infection in children: a prospective study of complications and reactivation. N Engl J Med 1994;331:432-8. 5. Secchiero P, Carrigan DR, Asano Y, Benedetti L, Crowley RW, Komaroff AL, et al. Detection of human herpesvirus 6 in plasma of children with primary infection and immunosuppressed patients by polymerase chain reaction. J Infect Dis 1995;171:273-80. 6. Clark DA, Kidd IM, Collingham KE, Tarlow M, Ayeni T, Riordan A, et al. Diagnosis of primary human herpesvirus 6 and 7 infections in febrile infants by polymerase chain reaction. Arch Dis Child 1997;77:42-5. 7. Bland RM, Mackie PL, Shorts T, Pate S, Paton JY. The rapid diagnosis and clinical features of human herpesvirus 6. J Infect 1998;32:161-5. 8. Chiu SS, Cheung CY, Tse CY, Peiris M. Early diagnosis of primary human herpesvirus 6 infection in childhood: serology, polymerase chain reaction, and virus load. J Infect Dis 1998;178:1250-6. 9. Suga S, Yoshikawa T, Kajita Y, Ozaki T, Asano Y. Prospective study of persistence and excretion of human herpesvirus-6 in patients with exanthem subitum and their parents. Pediatrics 1998;102:900-4. 10. Ryncarz AJ, Goddard J, Wald A, Huang ML, Roizman B, Corey L. Development of a high-throughput quantitative assay for detecting herpes simplex virus DNA in clinical samples. J Clin Microbiol 1999;37:1941-7. 11. Zerr DM, Gupta D, Huang ML, Carter R, Corey L. Effect of antivirals on human herpesvirus 6 replication in hematopoietic stem cell transplant recipients. Clin Infect Dis 2002;34:309-17. 12. Byington CL, Zerr DM, Taggart EW, Nguy L, Hillyard DR, Carroll KC, et al. Human herpesvirus-6 infection in febrile infants 90 days and younger. Pediatr Infect Dis 2002;21:996-9.
Polymerase Chain Reaction Diagnosis Of Primary Human Herpesvirus-6 Infection In The Acute Care Setting
485
Objective To evaluate the potential utility of identifying primary human herpesvirus (HHV)-6 infection in an emergency department setting by determining the frequency of HHV-6 viremia, diagnostic testing, and empiric treatment of serious bacterial infection (SBI) in HHV-6 viremic children, and concurrent SBI and HHV-6 viremia. Study design Children under age 2 years and who had a blood specimen taken for evaluation of fever were tested for HHV-6 by polymerase chain reaction (PCR). HHV-6 viremia was defined as detection of HHV-6 DNA in acute plasma. Results A total of 32 of the 181 subjects (18%) had HHV-6 viremia. Children with HHV-6 viremia frequently underwent procedures for diagnosis and empiric treatment of SBI: 60% had bladder catheterizations, 6% had lumbar punctures, 47% had radiographs, 32% received empiric antibiotics, and 34% were hospitalized. Four of the 32 children with HHV-6 viremia (12.5%) were diagnosed with SBI, although none had a positive culture of blood or cerebrospinal fluid. Conclusions Rapid diagnosis of HHV-6 viremia may not serve to adequately differentiate infants with and without SBI in acute care settings. Although no children with HHV-6 viremia had bacteremia or meningitis, it appears that additional criteria are needed to increase the specificity of HHV-6 PCR testing before withholding evaluation for SBI. (J Pediatr 2006;149:480-5)
oung children evaluated for fever in emergency departments (EDs) frequently undergo laboratory testing and receive empiric antibiotics, and some are hospitalized for observation out of concern for serious bacterial infection (SBI). Only 3% to 11% of young febrile children are typically found to have a SBI,1 and as such, alternative etiologies for the fever are common. HHV-6 infects ⬎ 90% of children by 2 years of age.2 Primary infection with HHV-6 frequently results in physician visits3 and has accounted for up to 21% of ED visits for fever in young children.4 Children generally recover uneventfully from primary HHV-6 infection and have not been observed to develop concurrent bacterial infection. For example, a study of 160 children with primary HHV-6 infection evaluated in an ED found that none had SBI.4 These data From the Departments of Pediatrics and suggest that a specific and sensitive test for primary HHV-6 infection might offer a Laboratory Medicine, University of Washington, Children’s Hospital and Regional diagnostic tool in the ED to reduce the need for other testing, antibiotic treatment, and Medical Center, and Fred Hutchinson Canhospitalization. cer Research Center, Seattle, WA. A natural history study of HHV-6 infection demonstrated that HHV-6 DNA levels This clinical research was conducted in accordance with guidelines for human experin saliva are low or undetectable early in the course of primary infection and subsequently imentation as specified by the Children’s 3 increase, reaching a plateau that is maintained for months. As such, a single determiHospital Institutional Review Board. nation of salivary HHV-6 alone is not a useful marker of primary infection. Conversely, None of the authors has any potential conflict of interest, real or perceived. detection of HHV-6 DNA in plasma, serum, or whole blood has been associated with Supported by in part by National Institutes 5-9 primary infection in immunocompetent children. of Health grant K23 AI001679 and by a We sought to determine the frequency of HHV-6 plasma viremia detected by grant from Children’s Hospital Fund for Excellence. polymerase chain reaction (PCR) in young children presenting to an ED with fever, to Submitted for publication Nov 7, 2005; last evaluate to what degree invasive medical testing and management takes place in these revision received Mar 15, 2006; accepted children, and to determine whether SBI occurs in children with HHV-6 plasma viremia. May 17, 2006.
Y
ED HHV IQR
480
Emergency department Human herpesvirus Interquartile range
PCR SBI
Polymerase chain reaction Serious bacterial infection
Reprint requests: Dr. Danielle M. Zerr, Children’s Hospital, W-8851, 4800 Sand Point Way NE, Seattle, WA 98105. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2006 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2006.05.027
METHODS Subjects All patients presenting to the ED of the Children’s Hospital and Regional Medical Center, Seattle between 8:00 AM and midnight from January 2001 through June 2002 who had a temperature of 38°C or higher, were under age 2 years, and were having blood drawn to evaluate for bacterial infection were potentially eligible for enrollment. Children were excluded from the study if they had been diagnosed with an immune deficiency or cancer or were receiving immunosuppressive medications, or if 1 parent was not able to communicate in English. Potentially eligible children were approached by the ED staff. If the parents were interested in the study, the ED staff summoned study personnel to the ED to further screen and, when appropriate, enroll the infant into the study. A total of 12,344 patients under age 2 years were brought to the ED between 8 AM and midnight during the study period; of these, 1639 had a diagnosis that included fever. Of these, 683 had a blood culture obtained, a marker for having blood drawn to evaluate for fever. Study personnel evaluated 252 patients who were screened by ED staff and expressed interest in the study. Of these, 182 children were enrolled. Of the 70 children not enrolled, 12 (17%) were ineligible, 54 (77%) had parents who declined participation for various reasons, and 4 (6%) lived out of the area or were scheduled to move before follow-up could be completed. After obtaining written informed consent from the subject’s parents, demographic and clinical information was collected, and blood and saliva were obtained for HHV-6 testing. Saliva also was collected approximately weekly by the subject’s parents during the 8 weeks after enrollment into the study. Saliva was collected using Sno strips as described previously.3 Definitions ● ●
HHV-6 viremia: Detection of HHV-6 DNA in the acute plasma specimen by real-time PCR. Viral shedding in saliva consistent with primary HHV-6: Absent or low levels (⬍ 12,000 copies of HHV-6 DNA/ mL) of HHV-6 DNA at the time of acute illness, followed by increasing levels in subsequent specimens, including at least 1 sample with ⬎ 1000 copies/mL. No more than 2 negative samples could be present between the first sample and the subsequent positive sample; however, temporal gaps in saliva samples were allowed. The HHV-6 saliva pattern was determined only for subjects who had a saliva specimen during acute illness and at least 1 follow-up specimen unless the acute saliva sample had ⬎ 12,000 copies/mL, indicating established infection. This definition was derived from a natural history study of primary HHV-6 infection that documented initial low-level HHV-6 DNA levels in saliva that increased and established a plateau over the weeks and months after primary infection.3 In the
●
natural history study, 82% of first-positive saliva samples in subjects with primary HHV-6 infection had ⬍ 12,000 HHV-6 DNA copies/mL, whereas 80% of samples from subjects with established HHV-6 infection had ⬎ 12,000 copies DNA/mL and 87% of samples from subjects with established HHV-6 infection had ⬎ 1000 DNA copies/ mL. SBI: Physician-diagnosed and treated bacteremia/sepsis, meningitis, urinary tract infection, pneumonia, bone and joint infection, or bacterial gastroenteritis, as documented in the medical record.
PCR Analysis Plasma and saliva specimens were batched to test for HHV-6. DNA was extracted from the samples using Qiagen columns.10,11 Saliva, collected with Sno strips, was processed as described previously.3 DNA was extracted from 100 L of plasma and eluted into 50 L of buffer, of which 20 L was submitted to PCR analysis. Thus, for each subject, DNA from 40 L of plasma was tested for HHV-6. Detection of HHV-6 DNA was accomplished with a real-time fluorescent probe quantitative PCR assay, as described previously.10,11 HHV-6 DNA-positive samples with sufficient DNA were then further typed as HHV-6 A or B.11 The investigator performing the PCR analysis (LN) was blinded to the subjects’ identity and clinical findings. PCR results were not shared with the subjects’ clinical providers. Statistical Analysis Statistical analyses were performed using SPSS version 10.1 (SPSS Inc, Chicago, IL). Proportions were compared between groups using Pearson’s 2 test or, if the assumptions for the 2 test were not met, Fisher’s exact test. Continuous variables were compared between groups using the MannWhitney U test.
RESULTS Plasma HHV-6 PCR Of the 182 subjects who were eligible and enrolled in the study, the plasma specimens from 181 were adequate and included in the study. Thirty-two (18%) of these 181 subjects had HHV-6 DNA detected in plasma, or HHV-6 viremia. Their median plasma viral load was 7530 copies/mL (range, 90 to 699,750 copies/mL; interquartile range [IQR], 1298 to 21,393 copies/mL). Saliva HHV-6 PCR Saliva specimens during acute illness were obtained from 178 of the 181 subjects, and at least 1 additional saliva specimen was collected from 139 patients. Twenty of the 31 viremic subjects with an acute saliva sample (65%) had detectable HHV-6 DNA in the acute saliva, versus 41 of the 147 nonviremic patients with acute saliva specimens (28%). There were 150 children (29 with viremia and 121 without
Polymerase Chain Reaction Diagnosis Of Primary Human Herpesvirus-6 Infection In The Acute Care Setting
481
Figure 1. Salivary HHV-6 DNA levels over time in (A) 31 subjects with HHV-6 plasma viremia versus (B) 41 subjects without HHV-6 plasma viremia but with detectable HHV-6 in their acute saliva specimen and thus likely established HHV-6 infection. Each box plot represents the median (middle line) and IQR (outer lines), and each “I” bar represents the highest and lowest values, excluding the outliers. The number of children providing samples for PCR each week is shown below the x-axis.
viremia) who had adequate saliva specimens to enable detection of a salivary pattern characteristic of primary HHV-6 infection, that is, low salivary levels of HHV-6 DNA in their initial specimen, followed by increasing levels (see Methods).3 Twenty-one of the 29 viremic subjects (72%), versus 10 of the 121 nonviremic subjects (8%) (P ⬍ .001), had a salivary pattern consistent with primary infection. To explore whether quantifying acute salivary levels might facilitate the diagnosis of primary HHV-6, salivary HHV-6 patterns were compared between the 31 subjects who had viremia and saliva specimens available and the 41 subjects who likely had established infection (ie, those without HHV-6 viremia but with HHV-6 DNA detected in their acute saliva specimen) (Figure 1). Subjects with likely established infection had relatively high levels of HHV-6 DNA in their initial saliva specimens (median HHV-6 DNA level, 42,800 copies/mL; IQR, 5160 to 206,890 copies/mL) that remained relatively stable over time, whereas subjects with HHV-6 viremia had lower initial levels (median, 940 copies/ mL; IQR, 0 to 11,690 copies/mL) that increased over time (P ⬍ .001 for comparison of initial levels of salivary HHV-6 DNA). In these 72 subjects, HHV-6 DNA level in plasma was inversely correlated with the HHV-6 DNA level in the acute saliva sample (Figure 2; Pearson correlation P ⬍ .001).
HHV-6 Type HHV-6 type was determined in the 63 subjects (28 viremic and 35 nonviremic) with sufficient sample DNA available for testing. Only type B HHV-6 was identified in these subjects; no type A virus was found. Clinical Presentation of Children With HHV-6 Viremia The children with HHV-6 viremia were significantly older than those without viremia and more likely to have rash 482
Zerr et al
Figure 2. Scatterplot of HHV-6 DNA plasma level by corresponding HHV-6 DNA acute saliva level in 31 subjects with HHV-6 plasma viremia and 41 subjects with likely established HHV-6 infection (without HHV-6 plasma viremia but with detectable HHV-6 in their acute saliva specimen). Levels ⬍ 10 copies/mL actually represent 0. A small random number was added to the 0 so that they could be easily visualized on the graph rather than falling on the axes. A linear regression line demonstrates the inverse correlation between the 2 variables. , Viremia; ⫹, Established.
characterizing their illness at the time of presentation. There were no differences in the frequency of rhinorrhea, cough, vomiting, diarrhea, seizure, and otitis media between the 2 groups (Table I). The children with HHV-6 viremia were more likely to have lower platelet and white blood cell counts but higher hematocrit levels than those without HHV-6 viremia.
Diagnostic Evaluation and Management for SBI The children with HHV-6 viremia underwent procedures for diagnosis of SBI: 60% underwent bladder catheterThe Journal of Pediatrics • October 2006
Table I. Demographic and clinical factors among children with and those without HHV-6 viremia HHV-6 viremia
Demographic information Median age in months (IQR) Age ⬎ 2 months Male Nonwhite Enrolled fall/winter Symptoms before enrollment Median maximum temperature (IQR) Maximum temperature ⬎39.3°C* Rhinorrhea Cough Vomiting Diarrhea Rash Seizure Otitis media Laboratory values at enrollment Median WBC count (IQR) WBC count ⬍5,000/mm3† Median platelet count (IQR) Platelet count ⬍300,000/ mm3‡ Median hematocrit (IQR) Hematocrit ⬎31.5%§
Yes (n ⴝ 32)
No (n ⴝ 149)
8.3 (4.1, 14.7) 29 (91%) 18 (56%) 10 (31%) 16 (50%)
4.5 (1.5, 10.8) 99 (66%) 88 (59%) 55 (37%) 88 (59%)
39.8 (39.4, 40.2) 26 (81%) 17 (55%) 17 (53%) 14 (44%) 13 (41%) 14 (44%) 1 (3%) 3 (9%)
39.1 (38.6, 39.7) 55 (43%) 74 (50%) 78 (52%) 60 (40%) 49 (33%) 34 (23%) 6 (4%) 17 (11%)
9.8 (5.6, 12.4) 6 (21%) 341 (222, 422) 13 (46%) 35 (33, 36) 26 (93%)
11.9 (8.1, 16.7) 9 (7%) 431 (317, 526) 28 (21%) 33 (31, 35) 97 (71%)
P value .009 .006 .77 .53 .35 ⬍.001 ⬍.001 .72 .94 .72 .40 .02 .81 .74 .03 .02 .002 .004 .02 .02
*Maximum temperature of 39.3°C represents the 50th percentile derived from the 159 subjects with data. †WBC count of 5,000/mm3 represents the 10th percentile derived from the 165 subjects with data. ‡Platelet count of 300,000/mm3 represents the 25th percentile derived from the 164 subjects with data. §HCT of 31.5% represents the 25th percentile derived from the 165 subjects with data.
ization, and 47% underwent radiographic examination. In addition, 32% received empiric antibiotics, and 34% were hospitalized (Table II). The children hospitalized with HHV-6 viremia ranged in age from 1 to 22 months, and the duration of hospitalization ranged between 1 and 4 days. Diagnoses given as rationales for hospitalization included fever, fever and hematochezia, presumed viral thrombocytopenia, suspected group A streptococcal infection, viral infection, viral pneumonia, viral gastroenteritis and secondary dehydration (n ⫽ 2), bacterial pneumonia, and pyelonephritis (n ⫽ 2).
SBI in Children With HHV-6 Four of the 32 children with HHV-6 viremia (12.5%; 95% confidence interval, 3% to 29%) were diagnosed with an SBI, although none of these children had a positive blood culture (Table II). The 4 cases of concurrent SBI and HHV-6 viremia are as follows: ●
A 15-month-old toddler was diagnosed with gastroenteritis. The illness was characterized by bloody diarrhea, and stool culture was positive for Campylobacter jejuni. The acute plasma had 100 copies/mL of HHV-6 DNA, whereas the acute saliva had 42,000 copies/mL. Saliva results over time were not consistent with primary infection.
●
●
An 11-month-old infant was diagnosed with urinary tract infection. The child was diagnosed with roseola by the primary care provider 2 weeks before enrollment. Antibiotics were initiated for otitis media 3 days before enrollment. The child was brought to the ED with fever, vomiting, and decreased oral intake and urine output. Evaluation at enrollment revealed abnormal urinalysis (11 to 20 white blood cells/highpower field and many bacteria), peripheral white blood cell count was 26,000/mm3, and C-reactive protein level was 31.4 mg. Urine culture was obtained after administration of antibiotics and was negative. The acute plasma specimen had 177 copies/mL of HHV-6 DNA; the acute saliva specimen had 0 copies/mL. Saliva results over time were consistent with primary infection. An 8-month– old infant was diagnosed with urinary tract infection. Seven days before enrollment, the child had a fever, followed 5 days later by resolution of fever and onset of rash. The fever subsequently returned, and the child was brought to the ED. Urinalysis revealed 0 to 5 white blood cells/high-power field and moderate bacteria, and a catheterized urine specimen yielded ⬎ 103 colony-forming units/mL of Klebsiella pneumoniae. The acute plasma specimen had 2678 copies/mL of HHV-6 DNA, whereas the acute saliva specimen had 0 copies/mL. Saliva results over time were consistent with primary infection.
Polymerase Chain Reaction Diagnosis Of Primary Human Herpesvirus-6 Infection In The Acute Care Setting
483
Table II. Management and outcomes of children with and without HHV-6 viremia HHV-6 viremia
Diagnostics/management Bladder catheterization Lumbar puncture Blood culture X-ray Empiric antibiotics* Hospitalized Follow-up visit with primary physician Diagnosed with serious bacterial infections† Bacteremia/sepsis Meningitis Urinary tract infection Pneumonia Gastroenteritis Septic arthritis Mastoiditis
Yes (n ⴝ 32)
No (n ⴝ 149)
P value
19 (60%) 2 (6%) 29 (91%) 15 (47%) 10 (32%) 11 (34%) 19 (66%) 4 (13%) 0 0 2 (6.3%) 1 (3.1%) 1 (3.1%) 0 0
95 (65%) 49 (33%) 130 (87%) 78 (52%) 51 (34%) 81 (54%) 105 (86%) 34 (23%) 4 (2.7%) 1 (0.7%) 20 (13.4%) 5 (3.4%) 2 (1.3%) 1 (0.7%) 1 (0.7%)
.58 .002 .60 .57 .83 .04 .009 .19
*“Empiric antibiotics” was defined as antibiotics given empirically for the possibility of serious bacterial illness. This did not include antibiotics given for diagnosed bacterial infections such as otitis media, pneumonia, etc. †See text for clinical details of the four patients with SBI.
●
A 5-month-old child was diagnosed with pneumonia. Seven days before enrollment, the child developed fever, followed 2 days later by resolution of fever and onset of rash. Two days before enrollment, fever returned with worsening cough and mild respiratory distress (grunting, increased work of breathing). Chest radiography revealed no focal infiltrate. The acute plasma specimen had 1520 copies/mL of HHV-6 DNA; the acute saliva specimen had 5080 copies/mL. Saliva results over time were consistent with primary infection.
Three of the 4 children described above experienced illness consistent with roseola (ie, fever followed by rash with defervescence) in the 1 to 2 weeks before their enrollment into the study, suggesting recent primary HHV-6 infection, and they also had salivary HHV-6 patterns that were consistent with recent infection.
DISCUSSION We found HHV-6 DNA plasma viremia, suggesting HHV-6 primary infection, in 18% of young febrile children evaluated in the ED for bacterial infection. Almost 1/3 of children with HHV-6 viremia received empiric antibiotics, 34% were hospitalized, and 13% were diagnosed with SBI. The frequency of HHV-6 viremia observed is consistent with the observation by Hall et al4 that HHV-6 accounts for 10% to 21% of ED visits among young febrile children. That subjects with HHV-6 viremia in our study frequently underwent invasive medical testing, received empiric antibiotics, and were hospitalized suggests a potential role for rapid HHV-6 diagnostic tests in reducing these costly interventions. Developing useful rapid diagnostic tests may be com484
Zerr et al
plicated, however, because, in contrast to the findings of Hall et al,4 SBI was diagnosed in 4 (13%) of our subjects with HHV-6 viremia. Although these children did not have positive cultures of blood or cerebrospinal fluid, they did have clinical and/or microbiological evidence of bacterial infection. Clinical history and HHV-6 results in these children suggest that primary HHV-6 infection may have occurred in the weeks just before enrollment in our study. First, in 3 of the 4 subjects, the clinical history provided by the parents suggested that roseola, and thus possibly primary HHV-6 infection, occurred 1 to 2 weeks before presentation to the ED. Moreover, the salivary HHV-6 patterns in these 3 children were consistent with primary HHV-6 infection. Finally, the level of HHV-6 DNA in the plasma of 2 of the 4 children was relatively low (⬍ 200 copies/mL). Although it is possible that the low levels were falsely positive, it is also possible that a very sensitive PCR of plasma may have detected remaining viral nucleic acids after recent primary infection. A study of febrile infants age 90 days and younger, applying the same PCR system as used in the present study, also demonstrated HHV-6 viremia coincident with an SBI.12 In that study, 5 of 47 infants (10.6%) had HHV-6 DNA detected in serum by PCR, and 1 of those 5 HHV-6 –positive infants had Escherichia coli bacteremia and urinary tract infection. Similar to children with HHV-6 viremia and SBI in our study, the HHV-6 DNA level in the serum of the infant with E. coli bacteremia was relatively low compared with the levels in his study-mates with HHV-6 viremia and no SBI. To allow clinicians to comfortably recommend withholding laboratory testing and empiric antibiotics from young febrile children, a rapid diagnostic test with high specificity for primary HHV-6 infection without bacterial infection is The Journal of Pediatrics • October 2006
needed. It appears that real-time PCR of plasma is a sensitive means of identifying primary HHV-6 infection but that it also carries a lower specificity for early infection compared with an approach involving viral culture and seroconversion.4 It is possible that quantifying a threshold level of HHV-6 DNA in acute plasma, or a combination of acute plasma and acute saliva DNA threshold levels, could deliver this specificity; however, sensitivity for detecting primary HHV-6, and thus the “payoff” of testing, would likely diminish. Constraining the target population could also improve the performance of any diagnostic test for primary HHV-6. For instance, in our study, had we tested only children with fever without obvious source (eg, children without respiratory distress or bloody diarrhea), then 2 of the 4 children with HHV-6 viremia and concomitant SBI would have been excluded. Although too many such constraints risk the requirement of cumbersome decision analyses, it may be practical to consider such an approach in young children who have fever without a source, a population that frequently receives diagnostic evaluation, empiric antibiotics, and hospitalization out of concern for SBI. In summary, our preliminary study indicates that PCR of plasma can identify primary HHV-6 infection, a common illness among children in the ED, and that a high proportion of children with HHV-6 viremia may undergo diagnostic testing, receive empiric antibiotics, and be hospitalized out of concern for SBI. Our results also suggest that HHV-6 viremia as detected by PCR may extend for 1 to 2 weeks beyond the onset of HHV-6 infection. Given this prolonged detection of HHV-6 DNA and the high attack rate of HHV-6, it should be expected that events of relatively high frequency, such as SBI in young children, will at times be coincident or in rapid succession with primary HHV-6 and HHV-6 viremia. Clinical and/or virologic criteria are needed to increase the specificity of DNA PCR testing for early primary HHV-6 infection before it can be used to withhold evaluation and therapy for SBI. We are grateful to the families who participated in this study. We also thank the nurses and physicians in the Children’s Hospital
ED for identifying study subjects; to Rena Bornemann, Jaime Nell Manansala, Amanda Allpress, and Laurie Newman for their help in enrolling those subjects; and to Dimitri Christakis for his helpful review of the manuscript.
REFERENCES 1. Baraff LJ, Bass JW, Fleisher GR, Klein JO, McCracken GH, Powell KR, et al. Practice guideline for the management of infants and children 0 –36 months of age with fever without source. Ann Emerg Med 1993;22:1198-210. 2. Okuno T, Takahashi K, Balachandra K, Shiraki K, Yamanishi K, Takahashi M, et al. Seroepidemiology of human herpesvirus 6 infection in normal children and adults. J Clin Microbiol 1989;27:651-3. 3. Zerr DM, Meier AS, Selke SS, Frenkel LM, Huang M, Wald A, et al. Virologic and clinical characteristics of primary human herpesvirus 6: a population-based study. N Engl J Med 2005;352:768-76. 4. Hall CB, Long CE, Schnabel KC, Caserta MT, McIntyre KM, Costanzo MA, et al. Human herpesvirus-6 infection in children: a prospective study of complications and reactivation. N Engl J Med 1994;331:432-8. 5. Secchiero P, Carrigan DR, Asano Y, Benedetti L, Crowley RW, Komaroff AL, et al. Detection of human herpesvirus 6 in plasma of children with primary infection and immunosuppressed patients by polymerase chain reaction. J Infect Dis 1995;171:273-80. 6. Clark DA, Kidd IM, Collingham KE, Tarlow M, Ayeni T, Riordan A, et al. Diagnosis of primary human herpesvirus 6 and 7 infections in febrile infants by polymerase chain reaction. Arch Dis Child 1997;77:42-5. 7. Bland RM, Mackie PL, Shorts T, Pate S, Paton JY. The rapid diagnosis and clinical features of human herpesvirus 6. J Infect 1998;32:161-5. 8. Chiu SS, Cheung CY, Tse CY, Peiris M. Early diagnosis of primary human herpesvirus 6 infection in childhood: serology, polymerase chain reaction, and virus load. J Infect Dis 1998;178:1250-6. 9. Suga S, Yoshikawa T, Kajita Y, Ozaki T, Asano Y. Prospective study of persistence and excretion of human herpesvirus-6 in patients with exanthem subitum and their parents. Pediatrics 1998;102:900-4. 10. Ryncarz AJ, Goddard J, Wald A, Huang ML, Roizman B, Corey L. Development of a high-throughput quantitative assay for detecting herpes simplex virus DNA in clinical samples. J Clin Microbiol 1999;37:1941-7. 11. Zerr DM, Gupta D, Huang ML, Carter R, Corey L. Effect of antivirals on human herpesvirus 6 replication in hematopoietic stem cell transplant recipients. Clin Infect Dis 2002;34:309-17. 12. Byington CL, Zerr DM, Taggart EW, Nguy L, Hillyard DR, Carroll KC, et al. Human herpesvirus-6 infection in febrile infants 90 days and younger. Pediatr Infect Dis 2002;21:996-9.
Polymerase Chain Reaction Diagnosis Of Primary Human Herpesvirus-6 Infection In The Acute Care Setting
485