Feasibility of Using Human Herpesvirus Six (HHV-6) IgG Levels
as a Screen for HHV-6 Associated Chronic Fatigue Syndrome

Konstance K. Knox, Ph.D. and Donald R. Carrigan, Ph.D.
Wisconsin Viral Research Group
Milwaukee, Wisconsin

Final Progress Report
October 23, 2006

Background:
Data generated by our laboratory has demonstrated the existence of a subpopulation of patients with CFS that have abnormal, active infections with HHV-6. Currently, our laboratory uses three separate diagnostic technologies to evaluate patients with CFS for HHV-6 infections:

1. nested HHV-6 PCR using plasma samples
2. rapid HHV-6 culture of blood samples
3. HHV-6 immediate early protein antigenemia using purified peripheral blood leukocytes

Examples of the data that we have obtained are summarized below. A cross-sectional study of CFS patients seen in a single clinical practice (Daniel Peterson, M.D.) and a set of samples from healthy control individuals using these three methods gave the following results.

 

Diagnostic Method	CFS Patients	Healthy Controls	Statistical Significance
Plasma PCR	9/441 (20%)	0/68 (0%)	p = 0.00012
Rapid Culture	13/58 (22%)	1/55 (2%)	p = 0.001
Antigenemia	25/58 (43%)	4/55 (7%)	p = 0.0001

1 Number positive in assay/Total tested
2 Comparison of patients and controls by two-sided Fisher's Exact Test

While these results have proven useful in helping to define an important and interesting subpopulation of patients with CFS, these technologies are labor intensive, expensive to perform and difficult to expand for high volume testing of patient samples. An inexpensive screening technology for identifying those CFS patients who may have an abnormal HHV-6 Infection is needed. Further, since HHV-6 and Epstein-Barr virus (EBV) are known to interact at the subcellular, and perhaps systemic, levels we studied diagnostic methods for it in parallel to those for HHV-6.

Results
Are elevated levels of IgG antibodies specific for HHV-6 or EBV of diagnostic value in CFS patients?

The first approach used in the development of a new and efficient diagnostic technology for HHV-6 and EBV was to determine whether CFS patients have higher serum IgG antibody levels than healthy control subjects. To this end, the median antibody titers for both HHV-6 and EBV were determined for a group of 33 healthy control subjects. These sera were obtained from a commercial source (Analytical Biological Services Inc; Wilmington, Delaware). HHV-6 antibody levels were determined by an indirect immunofluoresence assay (IFA) using HSB-2 T leukemia cells infected with two different strains of HHV-6 variant A (HHV-6A). These strains are
HHV-6A_GS and HHV-6A_WA. Antibodies specific for EBV capsid antigens were detected by IFA using a commercially obtained FDA approved kit (Epstein-Barr Virus VCA IFA IgG Test; Focus Diagnostics; Cypress, California).

It was determined that the median titers of HHV-6 and EBV serum IgG in healthy control subjects were 1:320 and 1:1920, respectively. Data illustrating these findings are shown in the table below.

 

Plasma samples from three different groups of patients with CFS were then analyzed for reactivity with HHV-6 and EBV infected cells by IFA. These patient groups were:

1. A cross section of CFS patients seen by Daniel Peterson, M.D. during a five month period in 1995. These patients were unselected since they were identified as sequential samples submitted for diagnostic testing for HHV-6 infection ("cross-section patients").
2. A group plasma samples from CFS patients seen by Dr. Peterson that were positive for HHV-6 by a nested polymerase chain reaction (PCR) assay in our laboratory ("PCR Positive Patients").
3. A set of plasma samples from CFS patients who have repeatedly tested negative for HHV-6 infections by the three methodologies discussed above ("Triple Negative Patients").

Results of these studies are summarized in the figures below.

 

For both EBV and HHV-6, no significant differences were observed between the "Cross-Section Patients" and the "Triple Negative Patients" and the healthy control subjects with respect to viral specific antibody titers above the median value. However, the percentage of the "PCR Patients" with HHV-6 antibody levels above the median (79%) was significantly higher (p = 0.02) than that of the controls. Specificity of this finding for HHV-6 is demonstrated by the fact that the "PCR Positive" patients did not significantly (p = 1.0) differ from the healthy controls with respect to their EBV specific antibody levels Therefore, selection of patients for positivity for HHV-6 infection by nested plasma PCR significantly coselected for increased levels of HHV-6 reactive antibodies.

In order to determine whether increased HHV-6 Specific antibody levels selects for positivity in any of the three standard HHV-6 diagnostic assays, the data obtained for the "Cross Section" patients were further analyzed. The three 2 x 2 contingency tables summarizing the analyses are shown below.

1 p= 1.00 by Two-sided Fisher's Exact Test; not significant
2 Number of samples

1 p = 1.00 by Two-sided Fisher's Exact Test; not significant
2 Number of samples

1 p = 0.37 by Two-sided Fisher's Exact Test; not significant
2 Number of samples

No significant relationship was seen between increased levels of HHV-6 reactive antibodies and any of the three diagnostic assays currently in use for HHV-6 infections. Therefore, while it appears that PCR positivity for HHV-6 DNA in plasma selects for increased levels of HHV-6 specific antibodies, the converse is not true. Increased antibody levels do not predict positivity for active HHV-6 infections.

Are other HHV-6 and EBV serological tests more useful that elevated IgG titers in the evaluation of CFS patients?

Given the failure of increased IgG levels to provide a useful screening procedure for HHV-6 and EBV infections in patients with CFS, a number of other antibody assays were assessed. The first of these was a test for EBV early antigen specific IgG, which is sometimes considered a marker for reactivation of EBV infections. Plasma samples from a cross-section of CFS patients from Dr. Daniel Peterson's practice were compared to a set of commercially obtained serum samples from healthy control subjects (Analytical Biological Services Inc; Wilmington, Delaware). The IFA assay used was an FDA approved diagnostic kit (MerifluorR EBV-EA IgG IFA/IFT; Catalog Number EA101/4870181; Meridian Bioscience Inc; Cincinnati, Ohio) which was used exactly as instructed by the manufacturer. All samples were run at a 1:40 dilution. Results are summarized in the figure below.

 

IgG reactive with the EBV early antigen was present in 70% to 80% of both the healthy controls and CFS patients, and the small difference observed between the two groups was not statistically significant Further, when the immunofluorescence staining intensity was quantified on a 0 (negative) to 4+ (strong positive) scale, no significant difference was seen between the patients and controls (p = 0.23 by Mann-Whitney Test).

The results obtained with the same plasma samples from the cross-sectional CFS patients in several different assays are summarized in the table below. Known positive control samples were included for each assay.

1 Manufacturer
2 Wisconsin Viral Research Group
3 Number Positive/Number Tested

None of the three assays described above produced a significant number of positive samples with the CFS plasma samples.