Rapid diagnosis of varicella zoster virus infection in acute facial palsy

Acute facial palsy caused by varicella zoster virus (VZV) reactivation is usually accompanied by auricular vesicles and vestibulocochlear dysfunction (Ramsay Hunt syndrome).¹ However, the pathognomonic vesicles of this disease are sometimes absent (zoster sine herpete) or may appear several days after the onset of facial palsy. Such patients are often initially misdiagnosed with Bell's palsy (idiopathic facial palsy) and treated without antiviral agents until the vesicles appear.^2,3 Previous studies demonstrated that early administration of acyclovir and prednisone within 3 days of the onset of facial palsy is critical for a good prognosis in terms of facial palsy and hearing loss.² This underlies the importance of early diagnosis of VZV infection in patients without pathognomonic vesicles so as to not miss the timing of antiviral treatment. In the current study, the viral genomes of VZV were analyzed using PCR on exudate specimens from auricular skin, peripheral blood mononuclear cells (PBMCs), and tear fluid secreted from lacrimal glands that were innervated by the greater superficial petrosal nerve of the facial nerve.

Methods. Patients and specimens. Fifty-one patients with acute unilateral facial palsy and 14 healthy volunteers were examined. all patients visited the outpatient clinic of Ehime University within 5 days after the onset of facial palsy. Twelve patients presented herpetic vesicles on the pinna at their initial visit. The remaining 39 patients had no herpetic rash on presentation, although 2 reported hearing impairment or vertigo. Serum antibody titers against VZV were examined twice, in the acute (within 5 days of the onset of facial palsy) and convalescent (more than 2 weeks after the onset) phases, by the complement fixation test. All patients and healthy volunteers gave informed consent.

Auricular vesicles were collected from 12 patients with typical Ramsay Hunt syndrome. In 39 patients without herpetic eruption, exudate of auricular skin was collected by absorbing it with a sterilized Schirmer's strip after scratching the interior surface of the geniculate zone on the auricle with a needle (figure 1). Blood and tear from the affected side on the face were obtained from all 51 patients (Pharmacia LKB, Sweden). The PBMC fraction was isolated in a Ficoll-Paque gradient. All specimens were collected at the patient's initial visit.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Collection of exudate of auricular skin. After scratching the interior surface of the geniculate zone on the auricle with a needle, exudate of auricular skin was absorbed with a sterilized Schirmer's strip.

DNA extraction and PCR. The auricular skin exudate and tear fluid were washed out from the Schirmer's strips with 100 µL Tris-EDTA (10 mM Tris-HC1, pH 7.5, 1 mM EDTA). These samples and PBMCs were treated and amplified by PCR as described previously.⁴ Synthetic primers corresponding in sequence to the parts of VZV gene 29 (sense, 5′-TACGGGTCTTGCCGGAGCTGGTAT-3′; anti-sense, 5′-AATGCCGTGACCACCAAGTATAAT-3′) were used for one round of PCR amplification (35 cycles; 95 °C, 80 seconds; 52 °C, 90 seconds; 73 °C, 1 second).⁵ The PCR products were separated by agarose gel and detected by Southern blot analysis. Rigid precautions against contamination in the sample processing included the use of water controls replacing DNA samples in all amplifications. The limit of detection for the PCR amplification was approximately 10 femtograms.

Results. Twelve patients presented herpetic vesicles on the pinna at their initial visit. Seven patients developed herpetic vesicles 2 to 7 days (mean 4.6 days) after their initial visit and were eventually diagnosed with Ramsay Hunt syndrome. Herpetic vesicles did not appear in 32 patients during the follow-up periods; however, serum antibody titers against VZV were elevated more than fourfold in six patients who were retrospectively diagnosed as having zoster sine herpete. According to the clinical signs and serologic tests, final diagnoses were Ramsay Hunt syndrome in 19 patients, zoster sine herpete in 6 patients, and Bell's palsy in 26 patients.

Specific VZV DNA was detected by both ethidium bromide staining and Southern blot analysis in all 12 vesicle specimens (100%), in 6 of 12 PBMC specimens (50%), and in 9 of 12 tear fluid specimens (75%) obtained from typical Ramsay Hunt patients. VZV DNA was also detected in five of the seven specimens of the auricular skin exudate (71%), in two of the seven PBMC specimens (29%), and in two of the seven tear fluid specimens (29%) obtained from Ramsay Hunt patients without vesicles at the initial visit. Furthermore, VZV DNA was detected in four of the six specimens of the auricular skin exudate (67%), in two of the six PBMC specimens (33%), and in two of the six tear fluid specimens (33%) obtained from patients with zoster sine herpete. In contrast, no VZV DNA was detected in exudate or in PBMC or tear fluid specimens obtained from 26 patients with Bell's palsy or the 14 healthy volunteers (figure 2). The table summarizes clinical data and results of PCR in six patients with zoster sine herpete and seven patients with Ramsay Hunt syndrome without herpetic eruption at the initial visit.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Detection of varicella zoster virus (VZV) DNA in auricular skin exudate, peripheral blood mononuclear cells (PBMCs), and tear fluid obtained from patients with Ramsay Hunt syndrome without vesicles and zoster sine herpete. Ethidium bromide staining (top) and Southern blot analysis (bottom) of PCR products are shown. VZV DNA fragments were amplified from the exudate of auricular skin, PBMCs, and tear fluid. Plasmid DNA used as a carrier for DNA extraction is seen in every lane except lanes M and P. Lanes 1 to 10 correspond to Patients 1 to 10 in the table. M = molecular size marker of 100-bp ladder; P = positive control.

Discussion. In our clinical observation of 325 patients with Ramsay Hunt syndrome, one-third developed herpetic eruption 2 to 14 days after the onset of facial palsy.⁶ These data suggest that, for many patients with zoster sine herpete and Ramsay Hunt syndrome without vesicles, the possibility exists to be misdiagnosed with Bell's palsy at the initial visit and treated without antiviral agents.

Serologic tests are conventionally used to diagnose VZV infection; however, seroconversion of VZV immunoglobulin (Ig)G antibody takes more than 2 weeks to be confirmed. IgM or IgA antibody titers to VZV occasionally elevate in the acute phase of Ramsay Hunt syndrome, but their diagnostic value for zoster sine herpete remains obscure.⁷ VZV antigen skin test and cell counts in CSF were proposed as early diagnostic tools for VZV infection.⁸ However, the specificity of these latter tests was questionable given that a negative VZV skin test and CSF cell count were also found in a small number of patients with Bell's palsy.⁸ Thus, although these aforementioned tests may help to assist an early diagnosis, they are neither specific nor direct methods for identification of VZV infection.

In the current study, we detected VZV DNA in auricular skin exudate in 67% of the patients with zoster sine herpete and in 71% of patients with Ramsay Hunt syndrome before the appearance of vesicles. Once latent VZV reactivates in the sensory ganglia, it spreads intra-axonally to the epithelial cells of the auricular skin and replicates there, forming pathognomonic vesicles. Thus, VZV might be present in the nerve endings or epithelial cells of the skin before appearance of vesicles. Gilden et al.⁹ also detected VZV DNA in PBMCs and CSF by PCR in two patients with prolonged radicular pain without zoster rash. Taken together, our results suggest that subclinical involvement of the nerve, skin, blood, and tear fluid are present in VZV reactivation and that PCR can reveal these subclinical stages of the disease. Because this method takes only 5 hours to provide a result, PCR analysis of VZV DNA in exudate of the auricular skin can be a rapid and useful diagnostic tool for identification of VZV infection in acute peripheral facial palsy without herpetic eruption.