Question
What is the performance compared to shell vial culture of an automated sample extraction procedure combined with an internally controlled real-time PCR assay for detection of herpes simplex virus (HSV) and varicella zoster virus (VZV)?

Design
This study describes the comparison of highly automated, internally controlled real-time PCR assays for detection of HSV-1, HSV-2, and VZV with conventional shell vial culture on specimens from patients suspected of HSV or VZV infections. 

Participants
One hundred eighty-two consecutive specimens from patients with clinical signs and symptoms suggestive of HSV or VZV infection were tested.  Specimens included swabs from skin (n=21), oro-facial (n=31), or ano-genital (n=56) vesicles or lesions, eyes (n=9), or throat (n=17), and bronchoalveolar lavages (BAL) (n=16).  

Description of Tests and Diagnostic Standard
All specimens except BAL were placed into 2 mL of virus transport medium and centrifuged.  One mL of supernatant was used for virus culture on human diploid embryonic lung fibroblasts in shell vials.  Virus antigen detection and typing were performed on day 2 for HSV and days 3 and 7 for VZV using monoclonal antibodies specific for HSV-1, HSV-2, or VZV (Argone Biosoft, Varilhes, France).  Total nucleic acids were extracted from 0.2 mL of each specimen using the MagNA Pure LC automated extraction reagents and instrument (Roche Diagnostics, Penzberg, Germany).  A fixed, low concentration of phocine herpes virus (PhHV) was added into each sample before extraction as an internal control.  Specific primers and TaqMan fluorescent probes for detection of HSV-1 (glycoprotein G gene), HSV-2 (glycoprotein D gene), VZV (gene 38), and PhHV (glycoprotein B gene) were used as previously described.  Four different PCRs were performed separately using the ABI Prism 7700 instrument (Applied Biosystems, Nieuweerkerk a/d Ussel, The Netherlands).  Each sample was tested in duplicate in each assay.  Assay thresholds were set at 10 times the standard deviations (SD) of the mean baselines calculated for cycles 3 to 15.  If the threshold cycle (Ct) value for the internal control was higher than 36.66 (2 SD above the mean Ct value), the amplification of the specimen was considered inhibited and the procedure repeated for that specimen.  Assay sensitivities were determined using serial dilutions of electron microscopy counted virus stocks of HSV-1 and HSV-2 and a quantified VZV genomic DNA solution.  HSV assay performance was also evaluated using a proficiency panel from Quality Control of Molecular Diagnostics (Glasgow, Scotland).  

Main Outcome Measures
The real-time PCR results were compared to those of the shell vial culture assays.

Main Results
All 3 assays were sensitive to at least 6 DNA copies per reaction (600 copies/mL of specimen prior to extraction).  For the HSV-1 and HSV-2 PCR assays, all proficiency panel samples containing at least 1-3 X 10³ genome/mL were detected.  A weak positive HSV-1 sample containing 3-9 X 10² genomes/mL was not detected.  No false positive results were obtained.  The PCR and culture results of the 182 clinical specimens are shown in the table.  The one HSV-1 culture positive specimen that was PCR negative was a genital swab that contained PCR inhibitors.  For the HSV PCR assays, the median Ct value for the culture negative specimens was 34.0 (range = 18.5-39.1), while the median Ct value for culture-positive specimens was 24.0 (range = 14.9-36.1) (P < 0.0001).  For HSV positive specimens, the detection rate compared to culture increased 33% and 37.5% for ano-genital and oro-facial specimen, respectively.  Throat specimens were detected equally by PCR and culture.  

Results of real-time PCR and culture for HSV-1, HSV-2, and VZV on 182 clinical specimens

Source of funding:  A grant from ZorgOnderzoek Nederland.