Meningitis and encephalitis caused by HSV infection may be difficult to recognize because they are only rarely associated with clinical evidence of extraneural infection [4]. Even when meningitis results from reactivation of latent HSV type 2 infection in patients with known genital herpes, coincident herpetic skin lesions are seldom documented [5, 6]. This absence of extraneural manifestations increases the importance of detecting HSV in cerebrospinal fluid or brain tissue or both. Viral cultures rarely show HSV type 1 in cerebrospinal fluid during viral encephalitis [4]. Herpes simplex virus type 2 is more commonly cultured from cerebrospinal fluid during meningitis associated with the first episode of genital HSV type 2 infection but is rarely cultured during meningitis associated with recurrent genital herpes [5, 7-9].

The failure of culture techniques has led to the use of new diagnostic techniques. These include the detection of antibodies to HSV in cerebrospinal fluid by immunoblot analysis [10] or enzyme-linked immunosorbent assay [5, 11] and the detection of HSV DNA using the polymerase chain reaction (PCR) [12-17]. The PCR assay is extremely sensitive and highly specific for the diagnosis of HSV infections of the central nervous system [12-17]. With a single exception [18], it has not been possible to isolate HSV from cerebrospinal fluid of patients with recurrent meningitis [5-9]. However, HSV DNA has been found in samples of cerebrospinal fluid from isolated patients with benign recurrent lymphocytic meningitis [16, 19, 20]. Therefore, we used PCR DNA amplification and type-specific antibody detection in cerebrospinal fluid samples to investigate the role of HSV in a large number of patients with the syndrome of benign, recurrent, self-limited aseptic meningitis [21].

Methods

Top Methods Results Discussion Author & Article Info References

Specimen Preparation and Polymerase Chain Reaction

Specimens submitted for PCR analysis for HSV were subjected to precautionary procedures to avoid contaminating the samples and the DNA amplification reactions with exogenous DNA [22]. The PCR assay for HSV was done as described previously [16] with minor modifications. Specimens were analyzed shortly after receipt by the laboratory. They were treated with the chelating InstaGene DNA purification matrix (BioRad Laboratories, Inc.; Hercules, California) before amplification. Briefly, the aqueous volume of the InstaGene slurry was first decreased by one half; to 100 µL of the decreased slurry, 20 µL of cerebrospinal fluid was added, and incubation steps were carried out according to the manufacturer's guidelines. Polymerase chain reaction mixtures contained 20 µL of InstaGene-treated cerebrospinal fluid in a total volume of 50 µL. Each reaction contained 1X PfuI buffer, 200 µmol/L each of four deoxynucleoside triphosphates, 1.25 units of PfuI (Stratagene; La Jolla, California), and 0.8 µmol/L of each primer, HSVPolA1 and HSVPolA2 [16]. The samples were amplified for 45 cycles.

All DNA amplification experiments included a negative water control and a positive sensitivity control, which consisted of 20 molecules of the plasmid pGX146. This plasmid contains the Kpn I x fragment of the HSV type 1 strain 17 DNA polymerase gene [23]. Appropriate clinical controls to compare with our test specimens cannot be obtained. In order to verify the specificity of the DNA amplification procedure, we used negative control specimens from patients with neurologic diseases who did not have herpes as specified (Table 1, patients 1 to 6) and used positive control specimens from patients with herpes simplex infections who were diagnosed by an alternate method (Table 1, patients 7 to 13). Experiments for DNA amplification were prepared in parallel with blindly coded cerebrospinal fluid specimens. Cerebrospinal fluid specimens from patients with benign recurrent lymphocytic meningitis were tested in parallel with one positive control specimen of cerebrospinal fluid Table 1, in addition to the pGX146 and water controls. To test for the presence of PCR inhibitors in specimens that yielded negative results in DNA amplification experiments, reconstruction experiments were done in which InstaGene DNA preparations were spiked with 20 molecules of pGX146 and subjected to amplification. Restriction digests of DNA amplification products were typed, as previously described [16].

Detection of Anti-Herpes Simplex Virus Antibodies

Anti-HSV antibodies were detected as previously described [16, 24, 25]. Results were scored as positive for HSV type 1 if a predominance of antibodies to HSV type 1 (at least four bands) was apparent and if no antibodies to HSV type 2 gG were detected. Results were scored as positive for HSV type 2 if a predominance of antibodies to HSV type 2 (at least four bands) was detected including a band corresponding to HSV type 2 gG. Herpes simplex virus type 2 "atypical" profiles consisted of four or more HSV type 2 bands and a lack of detectable antibody to HSV type 2 gG. Such profiles are apparent in 8% to 10% of patients with culture-documented HSV type 2 genital infections [25, 26]. Weak profiles refer to those that are scored using a 1:25 dilution of cerebrospinal fluid rather than the usual 1:50 dilution.

Results

Top Methods Results Discussion Author & Article Info References

Between November 1990 and 15 June 1993, 156 cerebrospinal fluid specimens were received for PCR analysis for HSV. Clinical information accompanying these specimens was often limited. Approximately 80% of specimens were submitted because of presumed encephalitis or encephalopathic syndrome, 10% for presumed neonatal herpes simplex infection, 5% for myelitis or neuropathy or both, and 5% for miscellaneous diagnoses. Twenty of the cerebrospinal fluid specimens were from patients with a presumptive diagnosis of recurrent viral meningitis. The specimens were submitted directly or shipped on dry ice by overnight mail to the Diagnostic Virology Laboratory of the University of Colorado Health Sciences Center between November 1990 and 15 June 1993. Thirteen of these 20 [65%] specimens analyzed for HSV DNA were obtained during the acute phase of the illness, were not previously exposed to prolonged thermal trauma, and met the following preestablished diagnostic criteria: 1) at least three episodes of meningitis occurring in the absence of other clinically significant neurologic signs and symptoms that resolved spontaneously without residual sequelae; 2) predominant cerebrospinal fluid lymphocytic pleocytosis; 3) negative or normal test results after culturing cerebrospinal fluid specimens for viruses, bacteria, Mycobacterium tuberculosis, and fungi; normal cytologic appearance; and the absence of treponemal or cryptococcal antigens; 4) no signs or symptoms of a uveomeningitic syndrome on routine ophthalmologic examination; 5) no evidence by history, examination, or laboratory tests to suggest the presence of collagen vascular disease, vasculitis, or drug-related hypersensitivity meningitis; 6) no evidence by computed tomography or magnetic resonance imaging [or both] of intracranial mass lesions; and 7) no evidence of herpetic skin lesions.

Seven of the 20 submitted samples from patients were excluded for the following reasons: polymorphonuclear pleocytosis (3 patients), fewer than three attacks of meningitis (1 patient), systemic lupus erythematosus (1 patient), convalescent specimen (1 patient), and inadequate specimen handling (1 patient). Clinical information was obtained from chart review and from the referring physicians. Genital HSV cultures and complete pelvic examinations were not done on all patients; therefore, we cannot exclude the possibility that some patients without evident herpetic skin lesions were asymptomatically shedding HSV from the genital tract. Patients who previously had herpetic skin lesions were not excluded but were considered as a separate group from those meeting all seven criteria.

Of the 13 patients Table 2 evaluated in this study, 9 were women. Patients had 3 to 9 attacks of meningitis (mean, 4.6 attacks) during a time interval of 2 to 21 years (mean, 8.4 years). Attacks of meningitis lasted 3 to 14 days (mean, 6.3 days). Cerebrospinal fluid specimens contained 48 to 1600 cells/mm³ (mean, 443 cells/mm³) with 58% to 98% lymphocytes. In the samples of cerebrospinal fluid, glucose levels were normal and protein levels ranged from 0.41 to 2.4 g/L (41 to 240 mg/dL), with a mean level of 1.22 g/L (122 mg/dL).

In 11 patients (84.6%; CI, 55% to 98%), HSV DNA was detected by PCR; the amplified sequence of DNA in 10 of these specimens was HSV type 2. In all of the cerebrospinal fluid specimens with HSV type 2 DNA, HSV type 2 antibody was also present. The cerebrospinal fluid from one patient had HSV type 1 DNA and HSV type 1 antibodies. Herpes simplex virus DNA could not be detected in cerebrospinal fluid specimens from two patients, although both contained antibody to HSV type 2. Thus, only 2 of 13 patients had benign recurrent lymphocytic meningitis that was not associated with an HSV infection in the central nervous system, as defined by the presence of HSV DNA.

To ensure that failure to detect HSV DNA in the cerebrospinal fluid from patients 9 and 10 Table 2 was not caused by the presence of PCR inhibitors, known amounts of pGX146 template DNA were added to specimens and amplification was done. These experiments consistently showed detection of DNA amplification products from an input of as low as 20 molecules of pGX146 by DNA hybridization. Therefore, our failure to detect HSV DNA in these two specimens was not the result of inhibition of the amplification process.

Only 3 of the 11 patients with HSV DNA and antibody in their cerebrospinal fluid had a history of recurrent genital HSV infection. The 3 patients studied for this series who had recurrent genital HSV did not have mucocutaneous manifestations at the time of the meningitis attack, and fewer than 50% of their previous attacks were associated with herpes genitalis. A single patient with HSV type 1 antibody had a history of oral HSV infection (Table 2, patient 7).

Discussion

Top Methods Results Discussion Author & Article Info References

The availability of an HSV type-specific PCR assay in our diagnostic virology laboratory prompted the submission of cerebrospinal fluid specimens from patients with benign recurrent lymphocytic meningitis, the cause of which was unknown after extensive investigations. These included attempts at virus isolation, serologic examination of serum and cerebrospinal fluid, and neuroimaging.

Our patients with benign recurrent lymphocytic meningitis fit generally accepted diagnostic criteria for Mollaret meningitis [27-32]. However, the available cerebrospinal fluid specimens, typically obtained 1 to 2 days after the onset of illness, did not contain Mollaret cells, nor was there a documented transition from mainly polymorphonuclear cells to mainly lymphocytic cells. Because controversy continues about the importance of these features for the diagnosis of Mollaret meningitis [33], we refer to this condition as benign recurrent lymphocytic meningitis.

Although Mollaret suggested that disease in some patients might be caused by viruses [34], specific infectious agents have rarely been identified in patients with benign recurrent meningitis [18, 35, 36]. Herpes simplex virus type 1 has been cultured from the cerebrospinal fluid of one patient [18], and HSV type 2 has been identified as the cause of some episodes of benign recurrent lymphocytic meningitis associated with recurrent herpes genitalis [5, 6]. Herpes simplex virus DNA has been identified in cerebrospinal fluid from 5 patients with recurrent meningitis [16, 19, 20]: type 1 in 1 patient and type 2 in 4 patients.

Our study describes 13 patients with benign recurrent lymphocytic meningitis. Eleven patients had HSV DNA in their cerebrospinal fluid specimens (Table 2). This finding was absent in the cerebrospinal fluid specimens from the cohort of patients with other illnesses associated with meningeal inflammation Table 1 and was absent in 2 of our patients. In all patients who had HSV DNA in their cerebrospinal fluid, anti-HSV type 2 antibody was almost exclusively present in the cerebrospinal fluid. One patient had antibody to HSV type 1 and HSV type 2 in her cerebrospinal fluid. The seroprevalence for HSV type 1 in adults from North America exceeds 50% [37]. Therefore, the paucity of HSV type 1 antibody in cerebrospinal fluid suggests, but does not definitively establish, that the presence of HSV type 2 antibody in cerebrospinal fluid is the result of intrathecal synthesis and is not caused by passive transfer of serum antibody resulting from breakdown of the blood-brain barrier.

The absence of HSV DNA in the cerebrospinal fluid of two patients may indicate that additional pathogens are also capable of causing this syndrome, although delay in obtaining specimens or suboptimal specimen handling could have contributed to these specimens being falsely negative for HSV. In addition, during the period of this study, cerebrospinal fluid specimens were received from three patients who met our criteria for a diagnosis of meningitis but who had polymorphonuclear pleocytosis and had negative PCR test results for HSV. One of these specimens had antibody to HSV type 1, although two had no detectable antibody to HSV (data not shown). This suggests the possibility that benign recurrent polymorphonuclear meningitis may be caused by a different process than benign recurrent lymphocytic meningitis.

As has been previously described [5, 6], three patients had a history of recurrent genital herpes simplex. None had mucocutaneous lesions at the time of their meningitis. Genital HSV cultures and complete pelvic examinations were not done on all patients, and we cannot exclude the possibility that some patients without evident herpetic skin lesions were asymptomatically shedding HSV. On the basis of our results, it seems prudent to do complete genital examinations on patients, including viral culture and PCR analysis of swabs obtained from the male urethra or from the vagina or from genital lesions (or both).

The pathogenesis of recurrent HSV meningitis remains obscure. We believe that the syndrome occurs far more commonly as a result of HSV type 2 infection than HSV type 1 infection. This pattern is the opposite of that encountered with HSV encephalitis in adult patients, in which most patients (>90%) have disease caused by HSV type 1 [4]. Differences in the neurologic syndromes produced by HSV may reflect the anatomic sites of virus latency. Herpes simplex virus type 2 latency occurs primarily in sensory neurons of sacral dorsal root ganglia, whereas HSV type 1 latency occurs principally in the sensory neurons of the trigeminal ganglia [38-40]. In both cases, reactivation of virus typically results in mucocutaneous disease (herpes labialis or keratitis for HSV type 1 and genitalis for HSV type 2). Neurologic events appear to be exceedingly rare during reactivation. When they do occur, we postulate that herpes simplex virus type 1 in trigeminal ganglia may spread to brain parenchyma producing encephalitis, whereas HSV type 2 in sacral dorsal root ganglia may seed the cerebrospinal fluid subarachnoid space and produce meningitis.

The fact that HSV is the major causative agent in benign recurrent lymphocytic meningitis raises the possibility that attacks of meningitis may be aborted by therapy and that recurrent attacks might be prevented. However, treatment with acyclovir has not yet been shown to definitively alter the natural history of the disease [41].