Abstract

Reactivation of human β-herpesviruses (cytomegalovirus [CMV], human herpesvirus [HHV]–6, and HHV-7) in nonimmunocompromised hosts is rare. Because these viruses are susceptible to reactivation by cytokines and stress-related mechanisms, the incidence of their reactivation was investigated among 120 patients during stress related to critical illness and compared with findings among 50 healthy volunteers. Human β-herpesvirus DNA was found in 65% of critically ill patients (60% men; mean age, 63 years) who required admission to an intensive care unit for medical (40%) or surgical (53%) indications or trauma (7%). HHV-6 reactivation was higher in critically ill patients than in healthy volunteers (54/101 vs. 0/50; P=.001). All patients except 1 were confirmed as HHV-6 variant A (mean virus load, 5066 copies/106 peripheral blood leukocytes). The reactivation of HHV-6A did not affect disease severity and outcome. No significant reactivation of HHV-7 or CMV was demonstrated among the critically ill patients. These findings contribute to the less-defined epidemiology of HHV-6A infection

Human β-herpesviruses (cytomegalovirus [CMV], human herpesvirus [HHV]–6 variants A and B, and HHV-7) are ubiquitous viruses that infect the majority of the human population (⩽70% for CMV and ⩽95% for HHV-6 and HHV-7) [1–3]. Primary infections with these viruses in the immunocompetent host generally cause benign illnesses. CMV causes up to 20% of infectious mononucleosis, and HHV-6 causes roseola infantum and febrile seizures in young children [3]. The clinical spectrum of HHV-7 illness is less defined but is assumed to be similar to that of HHV-6 [3]. Following primary infection, the β-herpesviruses remain latent or maintain low-level viral replication that is efficiently controlled by the immune system [4]

During periods of immunosuppression, reactivation from latency or increased viral replication may occur and cause disease [2, 4]. Organ invasion by CMV is manifested as sight-threatening retinitis in patients with AIDS and as hepatitis, pneumonitis, or enterocolitis in organ transplant recipients [4]. HHV-6 and HHV-7 are emerging pathogens that cause CMV-like illnesses in transplant recipients [2]

In contrast, β-herpesvirus reactivation in immunocompetent hosts is rarely documented. Recent observations, however, suggest that CMV reactivation may occur in normal hosts after exposure to stress characterized by excessive release of cytokines and cathecolamines [5–8]. Because of the similarity of the genomes of HHV-6 and HHV-7 to CMV, we hypothesized that reactivation of these viruses may also occur following stress-induced cytokine release. Furthermore, the clinical effects of β-herpesvirus reactivation in immunocompetent hosts are not defined. Thus, the goal of our study was to determine the incidence of β-herpesvirus reactivation following illness-related stress and to assess its effects on clinical outcome

Patients and Methods

Study populationWe enrolled 120 patients admitted to the medical and surgical intensive care units (ICUs) of St. Mary’s Hospital (Rochester, MN) between 1 January and 28 February 2000. We excluded all patients who were immunocompromised from cancer, hematologic disorders, or infection with human immunodeficiency virus and those who had received immunosuppressive agents or procedures. The medical records of all patients were reviewed for patient demographics, underlying illnesses, clinical symptoms, and patient outcome. Blood cell counts and the serum levels for creatinine, blood urea nitrogen, bilirubin, aspartate aminotransferase, and alanine aminotransferase were assessed on the days of ICU admission and blood collection for polymerase chain reaction (PCR) testing. Fifty healthy volunteers (mean age, 31 years [range, 20–43 years]; 55% men) were evaluated as control subjects

Clinical samplesIn all, 5 mL of blood was collected from all critically ill patients on the fourth day following the onset of illness or a procedure that required ICU admission. The blood samples were processed into peripheral blood leukocytes (PBLs) by the histopaque 1119 procedure (Sigma), aliquoted at 106 cells per vial, and stored at −70°C before testing

Detection of CMV and HHV-6 and HHV-7 DNAWe defined β-herpesvirus reactivation as the significant detection of CMV, HHV-6 DNA, or HHV-7 DNA, compared with findings in healthy volunteers. All critically ill patients were tested for CMV DNA, but only 101 patients had samples for HHV-6 and HHV-7 DNA detection. We analyzed the PBLs of the 50 healthy volunteers for all β-herpesviruses

CMV DNA was detected with 106 PBLs by using the COBAS Amplicor CMV Monitor (Roche Diagnostics) PCR assay. This automated test is sensitive, specific, and reproducible and quantifies CMV DNA by amplifying 265 bp of the CMV DNA polymerase gene UL54 [9]

HHV-6 and HHV-7 DNA were detected by the LightCycler (LC) (Roche Molecular Chemicals) PCR assay. Nucleic acid for HHV-6 and HHV-7 detection was extracted from 106 PBLs by use of the Isoquick method (Orca Research). For HHV-6 detection, 5 μL of extracted nucleic acid was added to a cuvette with 15 μL of amplification master mix containing the primers and the probes, 0.03 U/μL platinum Taq, 0.2 nM dNTP, 0.2% uracil-DNA glycosylase, and 4 nM MgCl2. The primers and probes are located in the immediate early gene of the virus and are available from Roche Molecular Biochemicals as analyte-specific reagents (ASRs), referenced as LC HHV-6, primer/hybridization probes; LC HHV-6, template DNA (positive control); and LC HHV-6, recovery template (internal amplification control). Target nucleic acid sequences were amplified by use of the following protocol: sterilization/denaturation in 2 steps (37°C for 300 s and 95°C for 180 s), PCR for 45 cycles (95°for 0 s, 60°C for 12 s, and 72°C for 12 s), 1 cycle of melting process (95°C for 0 s, 54°C for 10 s, and 95°C for 0 s), and 1 cycle of cool-down process (40°C for 30 s)

The parameters for HHV-7 detection were similar to those described for HHV-6 detection except for the primers and probes. The primers and probes are located in the U10 and U11 genes of HHV-7 and are available from Roche Molecular Biochemicals as ASRs referenced as LC HHV-7, primer/hybridization probes; LC HHV-7, template DNA (positive control); and LC HHV-7, recovery template (internal amplification control)

The quantification of HHV-6 and HHV-7 DNA (using LC software version 3.1) compared the signals of the patient samples with serially diluted standard plasmid controls at concentrations of 103, 102, 101, and 100 genome copies per PCR input. A cuvette containing only the master mix solution served as the negative control. Results are reported in absolute quantities of viral genomic copies per PCR input. The LC-PCR detects as few as 5–10 copies of plasmid (pCR2.1) inserts of target DNA of HHV-6A, HHV-6B, and HHV-7. These assays are specific for HHV-6 and HHV-7 and do not amplify DNA from other herpesviruses (herpes simplex viruses 1 and 2, varicella-zoster virus, CMV, Epstein-Barr virus, and HHV-8)

HHV-6A and HHV-6B variant analysisThe LC hybridization probes detected both HHV-6 variants. The HHV-6 variants were differentiated into A or B, by the LC melting-curve analysis, when the temperature in the thermal chamber was raised from 54°C to 95°C, and the fluorescence was measured at every 0.2°C. The sequence difference between the PCR products resulted in shifts in the melting curve temperatures (72°C for HHV-6A and 66°C for HHV-6B; figure 1). The 3-bp mismatch between the probe and the HHV-6B variants resulted in melting-curves at 66°C. In a preliminary analysis, the assay could differentiate wild-type strain A (U1102) from strain B (Z29) (virus strains courtesy of J. Black, Centers for Disease Control and Prevention [CDC]). In addition, nucleic acid extracted from known reference strains of HHV-6A (08-729-000) and HHV-6B (08-703-000; Advanced Biotechnologies) correlated with LC melting curves and identifications obtained with the CDC reference strains

Figure 1

Differentiation between variants of human herpesvirus (HHV)–6 by the melting-curve feature of the LightCycler (Roche Molecular Chemicals) polymerase chain reaction assay. HHV-6 variant A is identified by melting curves detectable at 72°C and variant B by melting curves detectable at 66°C. “−d [F2/F1]/dT” denotes the first negative derivative of the fluorescence. QS, quantitative standard

Figure 1

Differentiation between variants of human herpesvirus (HHV)–6 by the melting-curve feature of the LightCycler (Roche Molecular Chemicals) polymerase chain reaction assay. HHV-6 variant A is identified by melting curves detectable at 72°C and variant B by melting curves detectable at 66°C. “−d [F2/F1]/dT” denotes the first negative derivative of the fluorescence. QS, quantitative standard

Statistical analysisThe frequency of β-herpesvirus reactivation and the clinical variables were compared among the various groups by the χ2 test. The levels of viral DNA were expressed as the geometric mean±SD. P⩽.05 was considered statistically significant

Results

The demographics and clinical characteristics of our patients are presented in table 1: 60% were men; mean age was 63.2 years (range, 17–92 years); and 94% were white. Forty percent of the patients were admitted for medical indications, including an acute cardiac event (myocardial infarction and/or heart failure; 48%), infection (20%), seizures and cerebrovascular accidents (4%), and various illnesses (27%) such as pancreatitis, gastrointestinal bleeding, ketoacidosis, and noninfectious pulmonary illnesses. Fifty-three percent of the patients were admitted for surgical indications, including cardiac (e.g., coronary artery bypass grafting; 33%) and vascular (e.g., repair of aneurysm or bypass grafting for peripheral vascular disease; 28%) surgery, neurosurgery (e.g., tumor excision; 18%), and various urologic, orthopedic, and abdominal surgery (20%). Seven percent had trauma-related illness. The average duration of ICU admission was 7.7 days (range, 1–42 days); the overall mortality rate was 5.0%

Table 1

Characteristics of population studied.

Table 1

Characteristics of population studied.

Of the 101 patients tested for all 3 viruses, 65% had detectable β-herpesvirus DNA. HHV-6 DNA was detected in 54 of 101 critically ill patients (mean virus load, 5066 copies/106 PBLs), compared with none of 50 healthy volunteers (P=.001). All except 1 HHV-6 was determined to be variant A by the melting-curve feature of the LC assay. To further analyze the difference, the amplicon products were sequenced and found to correspond to the HHV-6A and HHV-6B reference strains, respectively (data not shown)

We observed similar rates of HHV-6 DNA detection among the 3 ICU groups (medical, 56%; surgical, 49%; trauma, 57%; P=.78). We detected higher mean HHV-6 DNA levels among patients in the medical group than in the other groups (medical, virus load of 6154 copies/106 PBLs; surgical, virus load of 4487 copies/106 PBLs; trauma, virus load of 2608 copies/106 PBLs), but these were not statistically significant. The rate of detection was also similar between cardiac or neurologic and noncardiac/nonneurologic illness (P=.38). Furthermore, the rate and level of detection of HHV-6 DNA were not associated with renal, respiratory, or hepatic dysfunction. The presence of HHV-6 reactivation did not influence mortality

HHV-7 DNA was detected more commonly (18/50 vs. 14/101; P=.002) but at a lower level (mean virus load, 66.8 copies/106 PBLs vs. 95.6 copies/106 PBLs) among healthy volunteers, compared with critically ill patients. CMV was not detected among the 50 healthy volunteers. Only 1 of the 120 patients had detectable CMV DNA

Discussion

Our study offers new insights into the interplay between stress and β-herpesvirus reactivation. We found significant and selective reactivation of HHV-6 but not of CMV and HHV-7 during illness-related stress. Furthermore, in our patients, HHV-6A was the predominant variant, a finding that contributes to the less-defined epidemiology of HHV-6A infection [3]

The absence of HHV-6 DNA among the healthy volunteers suggests that active HHV-6 replication does not occur in healthy persons. We hypothesize that the acute medical illnesses and surgical procedures among our patients mediated a cascade of reactions that resulted in HHV-6 reactivation. The precise mechanism of this reactivation has not been conclusively defined, but studies of a related β-herpesvirus (CMV) suggest the role of cytokines and chemical mediators [10, 11]

Although significant HHV-6A reactivation among critically ill patients was demonstrated, our analysis did not identify specific parameters associated with HHV-6 replication. Our study found no difference in the incidence and degree of HHV-6 reactivation regardless of age, thereby excluding this as the variable that would explain the higher HHV-6 detection rate among older critically ill patients, compared with younger healthy volunteers. In addition, renal, hepatic, neurologic, cardiac, and respiratory dysfunction were not associated with higher incidence or degree of HHV-6 reactivation. Our findings confirm a recent report [6] that, although 54% of patients with multiorgan-failure syndrome had HHV-6 reactivation, it did not adversely affect morbidity or patient outcome. We recognize that the uniform severity of our patients’ illnesses (i.e., severely ill patients at tertiary referral centers) could account for the inability to identify predictors of HHV-6 reactivation

The exact mechanism of the selective reactivation of HHV-6A but not -6B is unclear. However, our findings suggest that the reactivation characteristics of the 2 variants are different. Indeed, our observations reflect the epidemiologic differences between the 2 variants, and they support the hypothesis that the 2 variants may indeed represent 2 completely different viruses [3]. The majority of the illnesses known to be due to HHV-6 infection are attributed to variant B. Thus, our finding of the predominant HHV-6A reactivation during critical illness could contribute to a better understanding of the HHV-6 epidemiology. The availability of PCR assays that could differentiate the 2 variants will aid in defining the clinical spectrum of infection between the variants

We did not find significant HHV-7 reactivation among critically ill nonimmunocompromised hosts. The detection rate among the young healthy volunteers was significantly higher than that of the critically ill patients. However, age did not seem to influence the rate of HHV-7 DNA detection. The low detectable levels (mean virus load, 66.8 copies/106 PBLs) among the healthy volunteers may reflect the detection of a latent virus or evidence of subclinical low level viral replication

CMV reactivation has been reported to occur 4 days after acute illness [8], a finding that was not confirmed by our study. The differences in patient population and the method of CMV detection may account for our contrasting findings. Our findings agree with a previous observation that CMV reactivation is not found among patients who receive mechanical ventilation [12]. We also expanded our study to 600 ICU patients but found only a 2.8% CMV DNA detection rate during the fourth day of critical illness (C.F. and C.V.P., unpublished data)

In summary, our study demonstrated significant reactivation of HHV-6A but not of HHV-6B, HHV-7, and CMV following illness-related stress. Our findings provide new insights into the epidemiology of HHV-6 infection and HHV-6A in particular. The exact mechanism for this selective reactivation is obscure and is subject to further investigation

References

1
Mendez
JC
Dockrell
DH
Espy
MJ
, et al.  . 
Human β-herpesvirus interactions in solid organ transplant recipients
J Infect Dis
 , 
2001
, vol. 
183
 (pg. 
179
-
84
)
2
Dockrell
DH
Paya
CV
Human herpesvirus-6 and -7 in transplantation
Rev Med Virol
 , 
2001
, vol. 
11
 (pg. 
23
-
36
)
3
Dockrell
DH
Smith
TF
Paya
CV
Human herpesvirus 6
Mayo Clin Proc
 , 
1999
, vol. 
74
 (pg. 
163
-
70
)
4
Paya
C
Prevention of cytomegalovirus disease in recipients of solid-organ transplants
Clin Infect Dis
 , 
2001
, vol. 
32
 (pg. 
596
-
603
)
5
Asadullah
K
Prosch
S
Audring
H
, et al.  . 
A high prevalence of cytomegalovirus antigenaemia in patients with moderate to severe chronic plaque psoriasis: an association with systemic tumour necrosis factor α overexpression
Br J Dermatol
 , 
1999
, vol. 
141
 (pg. 
94
-
102
)
6
Desachy
A
Ranger-Rogez
S
Francois
B
, et al.  . 
Reactivation of human herpesvirus type 6 in multiple organ failure syndrome
Clin Infect Dis
 , 
2001
, vol. 
32
 (pg. 
197
-
203
)
7
Humar
A
St Louis
P
Mazzulli
T
, et al.  . 
Elevated serum cytokines are associated with cytomegalovirus infection and disease in bone marrow transplant recipients
J Infect Dis
 , 
1999
, vol. 
179
 (pg. 
484
-
8
)
8
Kutza
AS
Muhl
E
Hackstein
H
Kirchner
H
Bein
G
High incidence of active cytomegalovirus infection among septic patients
Clin Infect Dis
 , 
1998
, vol. 
26
 (pg. 
1076
-
82
)
9
Sia
IG
Wilson
JA
Espy
MJ
Paya
CV
Smith
TF
Evaluation of the COBAS Amplicor CMV Monitor test for detection of viral DNA in specimens taken from patients after liver transplantation
J Clin Microbiol
 , 
2000
, vol. 
38
 (pg. 
600
-
6
)
10
Prosch
S
Wendt
CE
Reinke
P
, et al.  . 
A novel link between stress and human cytomegalovirus (HCMV) infection: sympathetic hyperactivity stimulates HCMV activation
Virology
 , 
2000
, vol. 
272
 (pg. 
357
-
65
)
11
Ritter
T
Brandt
C
Prosch
S
, et al.  . 
Stimulatory and inhibitory action of cytokines on the regulation of hCMV-IE promoter activity in human endothelial cells
Cytokine
 , 
2000
, vol. 
12
 (pg. 
1163
-
70
)
12
Stephan
F
Meharzi
D
Ricci
S
Fajac
A
Clergue
F
Evaluation of polymerase chain reaction of cytomegalovirus reactivation in intensive care patients under mechanical ventilation
Intensive Care Med
 , 
1996
, vol. 
22
 (pg. 
1244
-
9
)
Informed consent was obtained from patients or their guardians. The human experimentation guidelines of the US Department of Health and Human Services and of the Mayo Clinic institutional review board were followed in the conduct of this clinical research