US Food and Drug Administration-approved enzyme-linked immunoassays (EIA) for determining type-specific herpes simplex virus (HSV) serostatus are widely used in clinical practice. We compared the performance of such assays with the University of Washington Western blot (UW WB) in patients who sought confirmation of their HSV serology result.
We reviewed charts of all persons evaluated at the Westover Heights Clinic in Portland, Oregon, from July 2010 through September 2015, who had a HSV EIA, followed by UW WB.
Of 864 persons, 47% were women. The median age was 36 years (range, 18–73 years). Using UW WB to define infection status, 286 (33%) persons were HSV-1 seropositive only, 104 (12%) were HSV-2 seropositive only, 134 (16%) were both HSV-1 and HSV-2 seropositive, 235 (27%) were HSV seronegative, and 105 (12%) had indeterminate results. Compared with the UW WB as the criterion standard, EIA was 70.2% sensitive and 91.6% specific for HSV-1, and 91.9% sensitive and 57.4% specific for HSV-2.
Among 278 persons who were HSV-1 seropositive by EIA, 255 were confirmed by the UW WB (positive predictive value [PPV], 91.7%). Of the 360 persons that were HSV-1 seronegative by the EIA, 252 were seronegative by UW WB (negative predictive value [NPV], 70.0%). Among 381 persons with HSV-2 EIA seropositivity, 193 tested HSV-2 seropositive by the UW WB (PPV, 50.7%). Of the 270 persons HSV-2 seronegative by EIA, 17 were seropositive with the UW WB (NPV, 93.7%). Among 261 persons with an EIA HSV-2 index value = 1.1–2.9, 39.8% of results were confirmed by UW WB, compared with 78.6% of the 70 persons with an EIA index value of 3 or greater (P < 0.001). The risk of false-positive HSV-2 EIA results was higher in those with HSV-1 antibody (47.1% vs 37.1%, P = 0.036).
US Food and Drug Administration-approved EIAs have poor PPV for HSV-2 and poor NPV for HSV-1 in clinical practice. More accurate rapid type-specific HSV antibody tests are needed.
A performance evaluation of herpes simplex virus (HSV) enzyme-linked immunoassay among adults in a private sexual health clinic in an urban setting, seeking to confirm their HSV serology result showed that enzyme-linked immunoassays have poor positive predictive value for HSV-2 and poor negative predictive value for HSV-1.
From the *Department of Medicine, †Department of Laboratory Medicine, University of Washington, Seattle, WA; ‡Westover Heights Clinic, Portland, OR; §Department of Biostatistics, University of Washington; ¶Vaccine and Infectious Diseases Division, Fred Hutchinson Cancer Research Center; and ∥Department of Epidemiology, University of Washington, Seattle, WA
Source of Funding: NIH/NIAID P01 AI-030731 and K24 AI-071113.
Conflict of Interest: A.W. has received research support from Vical Incorporated and Genocea Biosciences. A.W. is a consultant to AiCuris, Amgen Inc., Merck & Co. Inc, and GlaxoSmithKline plc. CJ has received research support from Sanofi-Pasteur, Genocea Biosciences, and Vical. C.J. is a consultant to Novavax. A.S.M. is a consultant to AiCuris and for Immune Design Corporation. All other authors report no conflict of interests.
Acknowledgements: The authors would like to acknowledge Dr. Rhoda A. Morrow who reviewed manuscript draft.
Presentations: This work was presented in part at the 2015 STI and HIV Congress in Brisbane, Australia and at the 2016 STD Prevention Conference in Atlanta, Georgia.
Correspondence: Elfriede Agyemang, MD, MPH, University of Washington, Harborview Medical Center, Mailstop 359928, 325 Ninth Ave, Seattle, WA 98104. E-mail: firstname.lastname@example.org; email@example.com.
Received for publication April 16, 2017, and accepted July 2, 2017.