Literature review > Issue_2 > Review Turner et al. 

An estimated 22% of Americans have antibody to HSV-2 [1]. Genital herpes is associated with significant clinical and psychosocial morbidity and epidemiologic studies suggest persons with HSV-2 are at elevated risk of HIV acquisition [2]. As a result, some authorities have advocated the development of public health genital herpes prevention programs including widespread serological testing for HSV-2 [3]. While Western blot has been the traditional serological "gold standard" for HSV-2 diagnosis, the test is costly and not widely available. Older, non-type specific serological EIAs do not reliably distinguish between HSV-1 and HSV-2. However, several type-specific tests for herpes glycoprotein g-2 have found the tests to be more specific, offering hope that these relatively inexpensive tests could be used in diverse populations [4]. Turner and colleagues used stored sera to assess the sensitivity and specificity of two type-specific tests with the goal of determining whether the test is sufficiently accurate to use in the "real world." Only one of these tests is now commercially available (HerpeSelect), so I will restrict specific comments to the performance of that test.

The study has three limitations that restrict one's ability to draw conclusions about the tests' performance. First, as the authors acknowledge, the Chiron RIBA has not been extensively studied. Although limited data support the test's accuracy compared to Western Blot, the use of the Chiron test introduces some uncertainty into the study results. Second, the sample size was not sufficient to narrowly define the sensitivity and specificity of the tests. Third, the study population had a very high prevalence of HSV-2 and was not drawn from the population in which the authors purport to show it can be used. Because the implications of a false positive test in an asymptomatic person are arguably more important than a false negative result, the main issue here is the test's specificity. Although the authors found that the test was 94.5% specific, the 95% confidence interval on this estimate was wide (83.9%-98.6%), and was not reported in the paper. The authors appropriately show the low positive predictive value of the test (66%) in a 10% prevalence population, while concluding that the test performs well in a high prevalence population like that they tested and, presumably, the population that might be tested in STD clinics. However, given the wide confidence interval on this estimate, using the lower bounds of the confidence estimate, the positive predictive value (72%) would not be acceptable to most clinicians even at a prevalence of 20%. Moreover, from the perspective of an unsuspecting HSV-2 negative patient getting a screening HSV-2 serological test, the authors' findings suggest a risk of receiving a false negative test result of 4.5%, with an upper bound potential risk of a false negative test as high as 13%. Given these uncertainties, can a clinician in good conscience recommend this test to asymptomatic patients without offering them any confirmatory testing?

This is a judgment call, and the issue of how to use gG-2 based tests remains uncertain. Further complicating the issue is the fact that the true HSV-2 status of persons who are Western blot negative and EIA positive is uncertain, since EIA may be more sensitive that Western Blot. A recent study documented that the HerpeSelect EIA becomes positive faster than the Western blot after HSV-2 acquisition [5]. Turner et al.'s conclusion, that confirmatory testing is not needed in relatively high prevalence populations, is shared fairly widely. For example, Whittington and colleagues studied the Meridian gG-2 test (now not available) among STD clinic patients and reported a positive predictive value of 95.8% (95% CI 91.6%-98.0%) [6]. However, the risk of HSV-2 varies considerably even within an STD clinic population, with higher prevalence among women and older persons. As a result, the risk of a false positive result varies in a way that is likely fairly predicable. My own opinion is that some confirmatory testing is probably required, at least in selected populations. It may be possible to use the index value of the EIA to triage the need for confirmatory testing. Prince observed that discordant EIA and Western blot results were more common among persons with low index values on the EIA and suggested that specimens with low values required confirmatory testing [7]. Although more data are needed on this issue, this seems like a prudent conclusion. A reasonable course is to report high index value EIAs as positive, and send low index value EIAs for confirmatory testing if such testing is available, and otherwise report these results as indeterminate, recommending that patients seek medical evaluation and culture testing if they have lesions, or repeat serological testing at some point in the future. The uncertainty on how to use existing EIAs suggests the need for new, confirmatory HSV-2 tests.

References:

1. Fleming DT, McQuillan GM, Johnson RE, et al. Herpes simplex virus type 2 in the United States, 1976 to 1994. N Engl J Med. 1997;337(16):1105-1111.

2. Wald A, Link K. Risk of human immunodeficiency virus infection in herpes simplex virus type 2-seropositive persons: a meta-analysis. JID. 2002;185:45-52.

3. Corey L. Raising the consciousness for identifying and controlling viral STDs: fears and frustrations--Thomas Parran Award Lecture. Sex Transm Dis. 1998;25(2):58-69.

4. Ashley RL. Performance and Use of HSV Type-specific Serology Test Kits. Herpes. 2002;9(2):38-45.

5. Ashley-Morrow R, Krantz E, Wald A. Time course of seroconversion by HerpeSelect ELISA after acquisition of genital herpes simplex virus type 1 (HSV-1) or HSV-2. Sex Transm Dis. 2003;30:310-314.

6. Whittington WL, Celum CL, Cent A, Ashley RL. Use of a glycoprotein G-based type-specific assay to detect antibodies to herpes simplex virus type 2 among persons attending sexually transmitted disease clinics. Sex Transm Dis. 2001;28(2):99-104.

7. Prince HE, Ernst CE, Hogrefe WR. Evaluation of an enzyme immunoassay system for measuring herpes simplex virus (HSV) type 1-specific and HSV type 2-specific IgG antibodies. J Clin Lab Anal. 2000;14(1):13-16.

about SDI | newsletters | grants | publications | literature reviews

WHO Home - WHO Search - TDR Home - SDI Home - SDI Contact us
(c) WHO/OMS 2001