Detection of Herpes Simplex Virus and Varicella-Zoster Virus by Traditional and Multiplex Molecular Methods

Dejan Nikolic, MD, PhD; Debra Kohn; Belinda Yen-Lieberman, PhD; Gary W Procop, MD, MS


Am J Clin Pathol. 2018;151(1):122-126. 

In This Article

Materials and Methods

We collected and studied 188 routine clinical specimens submitted to our laboratory for traditional virologic studies for the detection of HSV or VZV. These were non–cerebrospinal fluid (CSF) specimens, consisting of specimens from lesions and/or mucous membranes. These were nonconsecutive specimens (ie, convenience sampling), collected over a period of approximately 2 months, which were enriched for specimens that tested positive by traditional methods. Only one residual specimen was tested per patient, regardless of the age or location of the patient (ie, inpatient or ambulatory).

We evaluated 95 specimens that were submitted for HSV culture-based detection using the ELVIS HSV Test System (Quidel, San Diego, CA) and 93 specimens that were submitted for VZV direct fluorescent antibody (DFA) detection (Millipore Sigma, Billerica, MA). The ELVIS HSV Test System is a culture-based system that uses a genetically engineered cell line for the rapid detection for the HSV virus. This system detects the growth of both HSV-1 and HSV-2 but does not differentiate these viruses. If differentiation is needed, then it would be achieved using HSV type-specific DFA reagents. DFA-based detection of VZV was the sole method of detection for VZV because of the poor sensitivity of culture that has been previously demonstrated.[15]Traditional studies were performed according to manufacturer's instructions and standard laboratory procedures.

Residual specimens were stored at –70°C and thawed just prior to molecular testing, which was performed in batches. Specimens were assessed with a laboratory-developed test designed to detect HSV-1, HSV-2, and VZV that used MultiCode analyte-specific reagents using the universal assay protocol from Luminex (Austin, TX) (ie, the ARIES HSV 1&2/VZV combination assay) and the US Food and Drug Administration (FDA)–cleared Solana HSV1 + 2/VZV Assay (Quidel), performed according to the manufacturer's instructions. In brief, the ARIES HSV 1&2/VZV combination assay (Luminex) is a cartridge-based polymerase chain reaction (PCR) assay that is a largely automated sample-to-answer application that is performed in the ARIES Molecular Diagnostics Platform (Luminex). The Solana HSV1 + 2/VZV Assay is an isothermal, helicase-dependent amplification assay that is simple to use and does not require a thermocycler.

Operators of the molecular assays were unaware of the results of the culture and DFA, and each specimen was tested only once in each assay. Our reference standard (ie, the standard used to determine whether a specimen contained either HSV or VZV) was to categorize any specimen that tested positive for a particular virus in two or more tests as containing the virus detected. This was regardless of whether the positive tests were from the traditional nonmolecular tests, the amplification assays studied, or the tests used for discrepancy testing. If two or more of the tests were positive for a particular virus (ie, HSV or VZV) for any specimen, then the specimen was considered positive. Positive tests results, regardless of the assay type, that could not be corroborated by another assay (ie, single positives) were considered false positives. The sensitivity of each test was determined by calculating the number of truly positive specimens detected by the test studied divided by that same number plus the number of truly positive specimens missed by the test (ie, false negatives); sensitivity = true positives/true positives + false negatives. The specificity of each test was determined by calculating the number of truly negative specimens characterized as such by the test studied divided by that same number plus the number of false positives for that test; specificity = true negatives/true negatives + false positives. The positive and negative predictive values were not calculated since these specimens were not sequentially collected; rather, we attempted to enrich the study set for VZV DFA and HSV ELVIS-positive specimens. The frequency of detecting true positives was compared between methods with χ 2 analysis using EpicCalc 2000 (

This methods comparison was done as part of test development and a quality improvement project, and therefore it was exempt from institutional review board review.

Discrepancy Analysis

We added a third level of analysis (ie, discrepancy testing) because the two nucleic acid amplification assays were expected to be more sensitive than the traditional methods (based on previous findings). We felt that this would provide an additional level of accuracy in characterizing the specimens, in the event that one of the specimens had a very low concentration of virus that was not detected by one of the molecular tests being studied and missed by the traditional method. Discrepant analyses were performed if the nucleic acid amplification test (NAAT) disagreed or when either HSV or VZV was detected by a traditional method, but the presence of the virus could not be confirmed by either of the NAATs.

Specimens with discordant results for HSV were assessed by bidirectional sequencing following PCR amplification using analytically validated primers directed to HSV sequences distinct from those of the ARIES HSV 1&2/VZV combination assay. The individuals performing the sequencing analysis were unaware of the results of any other assay. Briefly, following extraction using the NucliSENS easyMAG (bioMérieux, Durham, NC), specimens were subjected to PCR using the OneStep QIAGEN Reverse Transcriptase PCR Kit (Qiagen, Hilden, Germany). Amplicons were then treated with exonuclease I and shrimp alkaline phosphatase to remove unincorporated primers and deoxynucleoside triphosphates remaining from the PCR reactions. Bidirectional sequencing was performed using M13 forward and reverse primers with the Sanger dideoxy sequencing method. Dye-labeled terminator cycle sequencing reactions were performed using the BigDye Terminator v3.1 Cycle Sequencing Kit (ThermoFisher, Waltham, MA). Any unincorporated dye was removed using the BigDye Xterminator Purification Kit (ThermoFisher). Sample electrophoresis and sequencing analysis were performed on the 3730xl Analyzer (ThermoFisher) using the 3730xl Data Collection software (v 3.1.1) and Sequencing Analysis software (v 5.4). Sequences that (1) were at least 200 bases in length, (2) had a Phred score greater than or equal to 20 for at least 90% of the bases, and (3) contained fewer than 5% ambiguous base calls were considered for further analysis using the Basic Local Alignment Search Tool (BLAST; National Center for Biotechnology Information). The Phred quality score provides a measure of the likelihood of a base calling error and was assigned during base calling according to peak mobility and shape analysis. Phred scores are expressed on a logarithmic scale where a Phred score of 20, which was used here, denotes a 1/100 likelihood of error or 99% confidence. Acceptable matches to BLAST reference sequences were those with greater than 95% query coverage and identity and an Expected Value (E-Value) less than 10 to 30 compared with the reference sequence.

Specimens discordant for VZV were assessed by a previously described VZV real-time PCR assay.[15] In brief, the VZV-specific PCR assay for discrepancy analysis was performed on the Roche LightCycler 1.2 system (Roche Diagnostics, Indianapolis, IN). The primers and fluorescence resonance energy transfer probes used with this assay were as follows: forward primer, 5′-GAC AAT ATC ATATACATG GAATGT G-3′; reverse primer, 5′-GCG GTA GTA ACAGAG AAT TTC TT-3′; hybridization probe 1, 5′-CGA AAA TCCAGAATCGGAACT TCT T-FITC-3′; and hybridization probe 2, 5′-LC640-CCA TTA CAG TAA ACT TTA GGC GGT C-3′. The LightCycler PCR protocol consisted of 10 minutes at 95°C for DNA polymerase activation, 45 cycles of PCR amplification (95°C for 10 seconds, 60°C for 10 seconds, and 72°C for 20 seconds), a melting step (40°C to 95°C at 0.1°C/s), and a cooling step (40°C for 30 seconds). Specimens that demonstrated exponential amplification and had a melting temperature that corresponded with the melting temperature of the positive control were deemed to contain VZV.

Assessment of Unexpected Findings

The use of this multiplex assay afforded the opportunity to investigate unexpected findings (ie, when HSV was discovered in a specimen that was submitted for VZV testing and vice versa). We reviewed the medical records of patients wherein HSV was detected in a specimen submitted for VZV testing and vice versa. If the virus detected was in the differential diagnosis of the clinical team (eg, if VZV was detected in the specimen submitted for HSV testing, but a test was also ordered for VZV detection), then the result was deemed "expected." If the appropriate test was not ordered for the virus detected, then it was categorized as an unexpected finding.