As outlined in the preceding manuscript (1), the use of low stringency polymerase chain reaction (PCR) had provided evidence for persistent viral infection in several patients diagnosed as having the chronic fatigue syndrome (CFS). The viral sequences were detected in patients' blood using a set of PCR primers reactive with regions of the gene coding the gp 64 late antigen (LA) of human cytomegalovirus (CMV) (2). Actual infection with CMV was considered unlikely, however, since PCR probing for other CMV related genes gave negative results. Moreover, the CMV LA reactive primers were not specific for CMV but could amplify various regions within other human herpesviruses, including human herpesvirus-6, Herpes simplex virus, Epstein-Barr virus and varicella zoster virus.

Support for a viral origin of CFS, and for a spectrum of CFS related disorders ranging from mild illness to profound neurological disturbance was provided by studies on a stereotactic brain biopsy on a woman with an unexplained progressive dysphasia. The brain biopsy tested positive for the CMV cross-reactive sequences by PCR. Enveloped viral particles were seen within the cytoplasm of neural cells. The nature of this virus could not be determined although the appearance was considered consistent with an atypical herpesvirus or retrovirus.

These studies have been extended since the time of the Cambridge Symposium. Progress has been made in both the use of low stringency PCR for viral detection and in the cultivation of novel cytopathic viruses from a large number of patients with CFS and with related neurological and neuromuscular diseases.

Polymerase Chain Reaction

The finding of viral particles in a PCR positive brain biopsy specimen led to efforts to identify other patients with unexplained neurological illnesses. In addition to the generic herpesviral PCR probes, the laboratory also used primers designed to amplify the tax gene of human T lymphotropic virus (HTLV) types I and II. Again, the PCR assays using the HTLV primers were run under conditions of reduced stringency so as to detect the possible presence of HTLV related sequences. The actual primers and detecting probe used differed only slightly from those described by Ehrlich and colleagues (3).

The PCR assays were established such that extracts of blood from normal healthy donors would consistently yield negative results by dot blot hybridization on PCR products. Multiple PCR products of varying sizes were, however, visualized on ethidium stained agarose gels when DNA from normal individuals was used as template. A strong positive hybridization by dot blot and a well defined band of the expected size were readily identified when HTLV I infected cells were used as the source of template DNA in the PCR. Blood samples from several CFS patients gave weak, but positive, dot blot hybridization signals. On Southern blot analysis, the reactive PCR products were shown to be of varying sizes and not to correspond to those seen with the positive HTLV control. The diagnostic utility of this assay was critically dependent on standardizing the exact assay conditions each time new primers and detecting probes were synthesized. Only in this way could there be the assurance of conditions that would not yield positive hybridization results with blood from normal individuals. Both positive and negative controls were run in each assay.

The results using the HTLV primers generally paralleled those using the CMV LA primers. Thus, while there were no positive responses seen when testing normal individuals, several patients gave clearly positive results. Of particular interest was the striking findings of CMV-LA and HTLV positive PCR results in a patient with severe residual neurological impairment three years following an atypical encephalitis-like illness. A clinical diagnosis of Herpes simplex encephalitis had been made but only after several days in the hospital. The clinical diagnosis had been delayed because of essentially normal findings on examination of cerebrospinal fluid (CSF). The individual responded poorly to Acyclovir therapy. Similar PCR findings were encountered in a 6 months old child in whom an encephalopathic illness slowly developed. By PCR, a 4+ response was seen with the HTLV primer set and a 2+ response with the CMV LA primer set. Yet the CSF chemistry was unremarkable. Detailed clinical descriptions of these and other cases will be presented elsewhere. The findings did, however, prompt an intense effort to attempt to culture virus from PCR positive CFS patients.

Viral Culture

A CFS patient repeatedly tested positive using both sets of PCR primers. The patient's illness was initially characterized by intense headaches, generalized myalgia and fever. She was hospitalized with an admitting diagnosis of encephalitis/meningitis. Her CSF examination, however, was unremarkable. The patient was treated with antibiotics which were discontinued after she developed watery diarrhea which subsequently subsided. The patient was discharged after 7 days with a clinical diagnosis of unexplained viral infection. Since that time she has experienced severe fatigue necessitating elimination of weekend social activities and a marked reduction in her capacity for normal work. Her cognitive functions have become impaired, especially her capacity to name items (dysnomia). Her short term memory is also defective. Moreover, she has near complete amnesia for the period of hospitalization and for events immediately preceding her initial illness. Her blood was cultured on a variety of cell lines including human fibroblasts. A cytopathic effect (CPE), characterized by the formation of rounded cell syncytia with prominent lipid inclusions, was noted. The CPE is not typical of any of the commonly isolated viruses in clinical laboratories. Nor do the affected cells stain using commercially available typing antisera. The viral infected cells were examined by electron microscopy. Numerous enveloped viral particles were seen within intracytoplasmic vacuoles. Non-enveloped, developing viral particles were seen within the nucleus. The CPE and electron micrographic appearances were suggestive of foamy or spumaviruses (4).

We have repeatedly cultured the same virus from this individual's blood as well as from her CSF. Virus producing a similar type of CPE in culture has been cultured from a number of additional patients referred for testing because of CFS-like illness. The rate of viral positivity has varied somewhat depending on the referring clinician. It has ranged from 40% to over 70%. Although, the culture results have yet to be correlated with clinical features, we know of no viral culture positive patients in whom the symptoms were considered to be vague and questionable. Nor have we seen positive cultures from normal, asymptomatic individuals.

Discussion

The culture findings add considerable support to the evidence from PCR that many CFS patients are persistently infected with a virus that shares some features with both herpesviruses and retroviruses. This could be explained by a recombinant virus containing both retroviral and herpesviral sequences. Possibly, the herpesviral related sequences could impart spumaviral and neurovirulence characteristics to an otherwise non-neurovirulent retrovirus. This suggestion would also be consistent with the findings reported by DeFreitas and her colleagues of HTLV-II gag region sequences in blood of CFS patients (5). The exact classification of foamy-cell inducing viruses that have been cultured from CFS patients will best be decided on the basis of sequencing data and efforts towards this goal are underway.

Acknowledgment

I wish to thank Khalid Ahmed, Li Cheng Zeng, Manju Roy, Anton Mayr and Peymon Javaherbin for helping to establish the viral culture and PCR assays. The work was supported by funds from the CFIDS Association, Inc., North Carolina.

References

1. Martin WJ. Detection of viral related sequences in CFS patients using polymerase chain reaction. Proceedings of the Cambridge Symposium (this issue).

2. Shibata D, Martin WJ, Appleman MD, et al. Detection of cytomegaloviral DNA in peripheral blood of patients infected with human immunodeficiency virus. J Infect Dis 1988; 158:1185-1192.

3. Ehrlich GD, Greenberg S, Abbott MA. Detection of human T-cell lymphoma / leukemia viruses in "PCR Protocols: A Guide to Methods and Applications." Ed Innis MI, Gelfand DH, Sninsky JJ, and White TJ. Academic Press 325-336; 1990.

4. Hooks JJ, Gibbs CJ. The foamy virus. Bacteriol Rev 1975; 39:169-185.

5. Defreitas E, Hilliard B, Cheney PR, et al. Retroviral sequences related to human T lymphocytotropic virus type II in patients with chronic fatigue syndrome. Proc Natl Acad Sci USA. 1991; 88:2922-2926.