Inaccuracy of Certain Commercial Enzyme Immunoassays in Diagnosing Genital Infections With Herpes Simplex Virus Types 1 or 2

Transcription

1 Microbiology and Infectious Disease / INACCURACY OF CERTAIN HSV ELISA TESTS Inaccuracy of Certain Commercial Enzyme Immunoassays in Diagnosing Genital Infections With Herpes Simplex Virus Types 1 or 2 Rhoda Ashley Morrow, PhD, 1,2 and David Friedrich 2 Key Words: Genital herpes; Herpes simplex virus serology DOI: /AKWHF62YYJ054HA6 Abstract Type-specific serologic results may be inaccurate if not based on glycoprotein G (gg). Commercial tests based on crude antigen (Zeus Scientific, Raritan, NJ; Wampole Laboratories, Cranbury, NJ; DiaSorin, Stillwater, MN) and one using gg-1 and gg-2 (Focus Technologies, Cypress, CA) were compared with Western blot on serum samples from patients with culture-documented first symptomatic episodes of genital herpes simplex virus (HSV) type 1 (n = 17) or HSV-2 (n = 49) infection or recurrent genital episodes (HSV-1, 30; HSV-2, 49). Concordance with Western blot results was 56% for Zeus, 63% for Wampole, 52% to 54% for DiaSorin, and 83% for Focus. Sensitivity and specificity. respectively, for HSV-1 were 77% and 53% (Zeus), 91% and 35% (Wampole), 98% and 8% (DiaSorin), 94% and 70% (DiaSorin predominant antibody), and 83% and 90% (Focus); for HSV-2 they were 88% and 81% (Zeus), 92% and 83% (Wampole), 96% and 54% (DiaSorin), 38% and 98% (DiaSorin predominant antibody), and 98% and 96% (Focus). Type-specific serologic testing for HSV should be performed with gg-based tests for accurate diagnosis of symptomatic genital herpes. Most patients with genital herpes are unaware of their infection status, partly because the disease can be mild or mistaken for other syndromes by clinicians and patients. 1,2 The recent treatment guidelines from the Centers for Disease Control and Prevention (CDC) for sexually transmitted diseases state that type-specific serologic tests are useful in confirming a diagnosis of genital herpes and to diagnose persons with unrecognized infection and to manage sex partners of persons with genital herpes. 3 A number of laboratories have developed and described tests that can discriminate accurately between antibodies to herpes simplex virus (HSV) type 1 and HSV-2. 4 Accurate type-specific HSV serologic testing depends on the detection of antibodies to 2 related but antigenically distinct viral envelope glycoproteins: gg-1 (from HSV-1) and gg-2 (from HSV-2). Three commercial tests based on gg have been approved by the US Food and Drug Administration for use in the diagnosis of genital herpes. A gg-based enzyme-linked immunosorbent assay (ELISA; HerpeSelect, Focus Technologies, Cypress, CA) is appropriate for large-scale reference and diagnostic laboratory testing 5,6 ; a strip immunoblot (HerpeSelect HSV-1 and HSV-2 immunoblot; Focus) is useful in low-volume settings; and the POCkit-HSV-2 (Diagnology, Belfast, Northern Ireland) is a rapid point-of-care test for clinics. 7,8 Despite the availability of gg-based HSV serologic tests and the knowledge that assays must be based on gg to accurately discriminate between antibodies to HSV-1 and HSV- 2, 9,10 several companies still manufacture and market ELISA kit pairs that use crude antigen preparations of HSV- 1 infected cell proteins in one plate of the pair and HSV-2 crude protein mixtures in the other plate. These kits still are purchased by laboratories to perform type-specific tests, Am J Clin Pathol 2003;120: DOI: /AKWHF62YYJ054HA6 839

2 Morrow and Friedrich / INACCURACY OF CERTAIN HSV ELISA TESTS including proficiency test panels from the College of American Pathologists. We ordered kits from 4 of 5 companies indicated as the source of test results in a recent College of American Pathologists proficiency survey. These kits were used to test convalescent serum samples from groups of well-characterized patients with lesions due to new genital herpes infections or from patients with culture-documented lesions from recurrent genital HSV-1 or HSV-2 symptomatic episodes. Serum samples from patients without a history of genital herpes but with HSV-2 antibodies as documented by Western blot type-specific serologic testing also were tested to determine the sensitivity of the tests for detecting subclinical infection. Performance results varied widely across tests and among the different clinical groups. In general, tests not based on gg were inadequate to detect new infections accurately and lacked specificity for diagnosing recurrent HSV-2 infections. Materials and Methods Patient Groups Serum samples were obtained with informed consent from patients seeking care at the Virology Research Clinic or the University of Washington Family Medicine Center, Seattle. Convalescent serum samples from patients with primary genital herpes due to HSV-1 (n = 17) or HSV-2 (n = 29) were selected to include samples drawn 28 to 90 days (median, 49 days) after the onset of symptoms. Earlier or later specimens were not used in the study. Western blot 11 was used to characterize patients as having primary infections, meaning that patients test result showed seronegativity within the first 2 weeks of symptom onset and later developed full profiles of antibody to the infecting virus type. Convalescent serum samples from 20 HSV-1 patients with nonprimary genital HSV-2 were obtained 28 to 70 days (median, 44 days) after the onset of symptoms. These 20 patients all had antibody to HSV-1 in acute serum samples and then showed seroconversion to HSV-2 by Western blot analysis. Subjects were enrolled in protocols approved by the University of Washington Institutional Review Board and gave informed consent for participation in studies of the natural history of genital herpes. Serum samples obtained within 6 months after a culturedocumented recurrence were from patients with genital HSV- 1 (n = 30) or genital HSV-2 (n = 49). Western blot revealed that none of the patients with HSV-1 had HSV-2 antibody. All patients with HSV-2 recurrent herpes had HSV-2 antibody revealed by Western blot, and, in addition, 31 had HSV-1 antibody, possibly from previous oral herpes infections. Lesion samples obtained during primary, nonprimary, or recurrent episodes were cultured on human diploid fibroblast cells or mink lung cells, and isolates were typed using typespecific monoclonal antibodies in an indirect fluorescent antibody assay as previously described. 12,13 The final group of serum samples were obtained as part of a study to determine viral shedding patterns among asymptomatic patients who were diagnosed as having HSV-2 by type-specific serologic testing (Western blot). 14 The 23 asymptomatic patients were followed up with daily selfadministered sampling for anogenital shedding of HSV-2, and all had at least 1 sample that was positive for HSV-2 by polymerase chain reaction, culture, or both. Serum samples were frozen at 20 C. All serum samples were thawed and refrozen once before use in the study. Serologic Testing We ordered 96-well plate ELISA kits from DiaSorin (Stillwater, MN), Zeus Scientific (Raritan, NJ), Wampole Laboratories (Cranbury, NJ), and Focus Technologies (Cypress, CA) based on the reported use of these kits by at least 1 laboratory in the spring 2002 College of American Pathologists proficiency survey. Each company offers 2 kits: one for HSV-1 and one for HSV-2; both were purchased. Tests were performed according to the manufacturers instructions. DiaSorin offers 2 methods to determine the serotype of antibodies. DiaSorin-1 in this report refers to calculation of the predominant antibody according to the manufacturer s formula. In brief, the ELISA value is determined by comparing the sample optical density against the kit calibrator optical density in a best-fit linear regression model. If the ELISA value for HSV-2 is greater than HSV-1, the predominant antibody is HSV-2. If the ELISA value for HSV-1 is greater than HSV-2, the predominant antibody is HSV-1. While DiaSorin claims that the predominant antibody to HSV-2 is highly predictive (98%) for HSV-2 infection, the kit insert also stipulates that the validity of predominant HSV-1 antibody has not been determined. DiaSorin-2 reflects the use of cutoff ELISA values for HSV-1 and HSV- 2 for each kit (HSV-1 and HSV-2) without regard for the test result of the other type. Data were analyzed for test accuracy using the patient s herpes infection status as the gold standard for the 145 patients with symptomatic infections and Western blot status for the 23 asymptomatic patients. Sensitivity and specificity were determined based on Western blot results. Sensitivity was the number of positive test results that were positive by Western blot (true positive) divided by the sum of the truepositive plus false-negative test results expressed as a percentage. Specificity was the number of true-negative test results divided by the sum of the true-negative and 840 Am J Clin Pathol 2003;120: DOI: /AKWHF62YYJ054HA6

3 Microbiology and Infectious Disease / ORIGINAL ARTICLE false-positive results (positive by the test measure but negative by Western blot) expressed as a percentage. Serum samples with equivocal results were excluded from further analyses. Results Diagnosis of New HSV-1 Infections Samples from 66 patients with newly acquired HSV-1 (n = 17) or HSV-2 (n = 49) were tested. Equivocal test results occurred with all 4 of the tests in a few patients. Although such a result might indicate early seroconversion, the incidence of equivocal results was not higher in patients with new infections than in those with established infections. Thus, equivocal results were censored from analysis of accuracy for detecting antibodies in newly infected persons. Table 1 gives the numbers of serum samples with correct results for each kit by category of infection. HSV-1 antibodies were present in most patients with primary HSV-1 according to the DiaSorin test (15/17 [88%]) and in more than half according to the Wampole test (10/16; 63%). The Zeus test was positive for HSV-1 antibody in only 3 (19%) of 16 patients with HSV-1. The gg-based Focus test was positive for HSV-1 in 13 (81%) of 16. Incorrect results of the Zeus, Wampole, and Focus tests were all false-negative, while the DiaSorin test showed antibodies to HSV-2 and HSV-1 in 1 of 17 and gave false-negative results in 1 of 17 by DiaSorin-1. When DiaSorin-2 results were used, 1 of 16 samples had false-negative results, and 2 of 16 were positive for HSV-1, but also for HSV-2. Diagnosis of New HSV-2 Infections HSV-2 infection was reflected by positive HSV-2 test results in most patients with primary HSV-2 by the Focus test (26/28 [93%]), while in 1 patient (4%), only HSV-1 antibody was detected. Predominant HSV-2 antibodies were detected by the DiaSorin-1 method in 22 (76%) of 29 patients; however, 6 patients (21%) were mistakenly given a diagnosis of HSV-1 with this method. DiaSorin-2 results were positive for HSV-2 in 27 (96%) of 28; however, 23 (85%) of the 27 serum samples were, in addition, falsely positive for HSV-1. The Wampole and Zeus tests detected far fewer HSV-2 seroconversions: 19 (70%) of 27 and 12 (50%) of 24, respectively (Table 1). Nonprimary infections are first episodes of HSV-2 genital herpes that occur in patients with preexisting HSV-1 antibodies. Of 20 such patients, the Zeus and DiaSorin-2 tests showed high proportions of positive results for HSV-2: 19 (95%) of 20. Wampole test results were positive in all 18 serum samples that had definitive HSV-2 results. In contrast, the DiaSorin-1 predominant antibody method detected none of the 20 HSV-2 seroconverters (Table 1). The Focus ELISA showed HSV-2 seroconversion in 13 (76%) of 17. Overall, correct serologic results (detecting antibody of the infecting type) were obtained in 57% (34/60) to 92% (59/64) of serum samples from first episodes of HSV-1 or HSV-2 when crude antigen based tests were used. Correct serologic results were obtained in 85% (52/61) of first episodes when the gg-based Focus ELISAs were used (Table 1). Diagnosis of Recurrent HSV-1 Episodes All serum samples from culture-documented HSV-1 recurrences had HSV-1 predominant antibody to HSV-1 shown by DiaSorin-1. However, 15 patients had detectable antibody to HSV-2 so that when the predominant antibody calculation was not applied (DiaSorin-2), almost half of these HSV-1 subjects seemed to have dual antibody Table 2. The Zeus test was positive for only HSV-1 in 11 (42%) of 26 of those with recurrent HSV-1, while 5 (19%) were positive for both HSV-1 and HSV-2 and 10 (38%) were negative Table 1 Correct Diagnoses in 66 Patients With Newly Acquired Genital Herpes * Primary HSV-1 Primary HSV-2 Nonprimary HSV-2 Total First Episodes Company Episodes (n = 17) Episodes (n = 29) Episodes (n = 20) (n = 66) Zeus 3/16 (19) 12/24 (50) 19/20 (95) 34/60 (57) Wampole 10/16 (63) 19/27 (70) 18/18 (100) 47/61 (77) DiaSorin method 1 15/17 (88) 22/29 (76) 0/20 (0) 37/66 (56) DiaSorin method 2 13/16 (81) 27/28 (96) 19/20 (95) 59/64 (92) Focus 13/16 (81) 26/28 (93) 13/17 (76) 52/61 (85) HSV, herpes simplex virus. * Data are given as number with correct diagnosis/number tested (percentage). For proprietary information, see the text. Test results were equivocal for HSV-1 in 1 sample each (Zeus, Wampole, Focus). A correct diagnosis was considered to be HSV-1 positive, only. Serum samples with positive HSV-1 but equivocal test results for HSV-2 (1 by DiaSorin-2; 1 by Wampole) were counted as correct diagnoses. Test results were equivocal for HSV-2 in 5 samples by Zeus, 2 by Wampole, and 1 each by DiaSorin method 2 and Focus. Correct test results for all tests were those that were positive for HSV-2. Test results were equivocal for HSV-2 in 3 samples by Focus and in 2 by Wampole. Correct test results for all tests were those that were positive for HSV-2. DiaSorin method 1, predominant antibody calculation used; DiaSorin method 2, independent testing results of HSV-1 and HSV-2 enzyme-linked immunosorbent assay. See also the Materials and Methods section. Am J Clin Pathol 2003;120: DOI: /AKWHF62YYJ054HA6 841

4 Morrow and Friedrich / INACCURACY OF CERTAIN HSV ELISA TESTS Table 2 Correct Diagnoses in Patients With Culture-Documented Genital Herpes Recurrences * Company HSV-1 (n = 30) HSV-2 (n = 49) Zeus 11/26 (42) 49/49 (100) Wampole 22/28 (79) 49/49 (100) DiaSorin method 1 30/30 (100) 16/49 (33) DiaSorin method 2 14/29 (48) 49/49 (100) Focus 27/30 (90) 49/49 (100) HSV, herpes simplex virus. * Data are given as number with correct diagnosis/number tested (percentage). Correct diagnosis refers to detecting the antibody to the virus type causing the recurrent genital lesion. See the Materials and Methods section and the Table 1 footnotes for a description of the DiaSorin methods. For proprietary information, see the text. All 30 patients had only HSV-1 by Western blot analysis. Samples with equivocal values by enzyme-linked immunosorbent assay were not included. All patients had HSV-2 antibody by Western blot analysis, while 31 had antibody to HSV-1 as well as to HSV-2. for both HSV-1 and HSV-2. The Wampole test detected only HSV-1 antibody in 22 (79%) of 28 patients with only HSV-1 infection but was positive for both types of antibody in 4 (14%) and negative for both HSV-1 and HSV-2 antibodies in 2 (7%). The Focus test was positive for HSV-1 only in 27 (90%) of 30 subjects with recurrent HSV-1 but was negative for both HSV-1 and HSV-2 antibody in 2 (7%) and positive for both in 1 subject (3%). Diagnosis of Recurrent HSV-2 Episodes All HSV-2 kits detected antibodies to HSV-2 (Table 2) after recurrent episodes. However, if DiaSorin-1was used, only 33% (16/49) had predominant HSV-2 (Table 2). Asymptomatic Patients With HSV-2 Antibody The CDC has suggested that type-specific serologic testing be used to identify patients who lack a definitive clinical diagnosis of genital herpes. 3 Partners of persons with genital herpes also may benefit from testing to ascertain their infection status if they have not been diagnosed. 15 We tested 23 patients without a history of HSV-2 but with HSV-2 antibody shown by Western blot analysis who were shown to have true-positive results for HSV-2 infection by later polymerase chain reaction or culture of genital secretions. 14 All 23 had positive results for the HSV-2 antibody with the Focus and Wampole tests, while 22 (96%) of 23 had positive results with DiaSorin-2, and 22 (96%) of 23 had positive results with the Zeus test for the HSV-2 antibody. However, when DiaSorin-1 was applied, HSV-2 antibodies were identified correctly in only 5 (22%) of 23 serum samples. Sensitivity and Specificity of ELISAs To compare test performance of these kits against a single serologic standard, we calculated the sensitivity and specificity for each test against the results of Western blot testing without regard for the clinical status of the patient Table 3. Sensitivity for HSV-1 antibody was high for the Wampole (91%) and both DiaSorin test methods (94%- 98%). However, the specificity was low for each test (8%- 70%), indicating a high probability of falsely positive HSV-1 test results with these kits. The Zeus test had low sensitivity (77%) and low specificity (53%) for HSV-1. The gg-based Focus test had 83% sensitivity and 90% specificity for HSV-1 (Table 3). The false-negative HSV-1 results occurred in 3 patients with primary HSV-1 infection, in 1 patient with nonprimary infection, in 2 patients with recurrent type 1, and in 12 of 31 culture-documented HSV-2 recurrent episodes in patients with dual antibody shown by Western blot. The false-positive HSV-1 Focus results occurred in 3 patients with primary HSV-2 infection and in 2 patients with recurrent HSV-2, with HSV-2 shown only by Western blot. Serum samples from these subjects were barely above the cutoff for positive in the Focus HSV-1 test (median, 2.5; cutoff, 1.10). Table 3 Comparison of ELISA and Western Blot Test Results * HSV-1 HSV-2 Sensitivity Specificity Sensitivity Specificity Concordance Zeus 81/105 (77) 26/49 (53) 94/107 (88) 38/47 (81) 86/154 (56) Wampole 96/105 (91) 17/48 (35) 98/106 (92) 39/47 (83) 97/153 (63) DiaSorin method 1 103/110 (94) 38/54 (70) 43/112 (38) 51/52 (98) 88/164 (54) DiaSorin method 2 106/108 (98) 4/53 (8) 107/111 (96) 27/50 (54) 84/161 (52) Focus 86/104 (83) 47/52 (90) 104/106 (98) 48/50 (96) 130/156 (83) ELISA, enzyme-linked immunosorbent assay; HSV, herpes simplex virus. * Data are given as follows (as determined by comparison with Western blot results for only HSV-1 [first 2 columns] or HSV-2 only [third and fourth columns]): sensitivity, number of true-positive results/number of true-positive plus false-negative results; specificity, number of true-negative results/number of true-negative plus false-positive results; concordance, number with same results/total number tested. Numbers in parentheses are percentages. Sensitivity and specificity calculations for HSV-1 and HSV-2 were made without regard for the test results for HSV-2 and HSV-1, respectively. Concordance for HSV-1 and HSV-2 results for ELISA and Western blot are in the final column. Equivocal results for either test were not counted. See the Materials and Methods section and Table 1 footnotes for a description of the DiaSorin methods. For proprietary information, see the text. 842 Am J Clin Pathol 2003;120: DOI: /AKWHF62YYJ054HA6

5 Microbiology and Infectious Disease / ORIGINAL ARTICLE The HSV-2 test sensitivities were 88%, 92%, and 96% for the Zeus, Wampole, and DiaSorin kits, respectively (Table 3), but the associated specificities were low (81%, 83%, and 54%, respectively). The predominant antibody method for DiaSorin detected only 43 (38%) of 112 serum samples with HSV-2 antibody shown by Western blot, but specificity was high (98%). The gg-based Focus test was 98% sensitive and 96% specific against the Western blot test. One of the 2 falsely positive Focus HSV-2 test results was in the nonprimary HSV-2 group; Western blot had not shown seroconversion in this serum sample; thus, by clinical criteria, these results were not falsely positive for HSV-2 but, rather, falsely negative by Western blot. For this group of patients, the Focus HSV-2 ELISA had 98% specificity when clinical data were considered. Overall concordance with Western blot results is shown for each test in Table 3. This is a highly rigorous test because both the HSV-1 and HSV-2 test results had to be in accord with the respective Western blot result to be counted as concordant. About half (52%-63%) of the non-gg-based ELISA results were concordant with Western blot. The Focus test was concordant with Western blot in 83% of the serum samples. Discussion The recent CDC treatment guidelines for sexually transmitted diseases recommend that type-specific testing based on gg be used for diagnosing genital herpes. 3 Several manufacturers in the United States sell HSV tests that are not based on gg-1 for HSV-1 antibody detection and gg-2 for HSV-2 antibody detection. To our knowledge, the accuracy of such tests has not been determined directly in characterized patient groups. To study the ability of currently available HSV serologic test kits to accurately diagnose genital herpes due to HSV-1 or HSV-2, we selected serum samples from patients with culture-documented first episodes of genital HSV-1 or genital HSV-2 and from patients after episodes of culture-documented recurrent lesions. These serum samples were tested with kits based on crude antigen mixtures from Zeus, DiaSorin, and Wampole. For comparison, the same serum samples were tested using gg-based ELISAs (for HSV-1 and HSV-2) from Focus. The Focus tests (brand name, HerpeSelect) have received US Food and Drug Administration approval for type-specific serologic testing of adults and pregnant women. We used single serum samples obtained 28 to 90 days after symptomatic, lesional outbreaks of genital herpes. The non gg-based tests gave inaccurate results in a high proportion of patients seroconverting to HSV-1 or HSV-2. The Zeus and Wampole tests were insensitive for detecting early HSV infection (57% and 77%, respectively). The DiaSorin HSV-2 test seemed to have acceptable sensitivity for early episodes of HSV-2 (93%-95%) at the expense of specificity: 3 (19%) of 16 patients with primary HSV-1 infection had false-positive results for HSV-2. When the predominant antibody was calculated for DiaSorin, none of the patients with nonprimary HSV-2 were diagnosed correctly, in accord with previous results of other non gg-based tests. 9 One of the main reasons to use a type-specific serologic test is to confirm infection in newly symptomatic patients without a virologic diagnosis. In our opinion, the use of any of these tests to supplement or replace virologic testing of patients with genital herpes would produce confusion rather than clarity. The non gg-based tests generally were more sensitive for detecting antibody after a recurrent HSV-2 episode than for diagnosing first episodes. An exception was the DiaSorin kit when the predominant antibody method was used: only 33% (16/49) of HSV-2 positive serum samples were detected, albeit with high specificity (98%). The HSV-2 specificities for the Zeus, Wampole, and DiaSorin tests compared with Western blot were low (54%-83%). Thus, positive HSV-2 tests with these kits have a high probability of being falsely positive. In contrast with diagnosing HSV-2, the ability to diagnose antibody after recurrent HSV-1 episodes was low for all methods except DiaSorin-1. The Zeus test had a high falsenegative rate for recurrent HSV-1 (10/26 [38%]) and a high false-positive rate for HSV-2 results (5/26 [19%]) in this group. The Wampole test had a false-negative rate of only 7% for HSV-1 (2/28), but 4 (14%) of 28 patients with recurrent HSV-1 had false-positive results for HSV-2. All 15 of the incorrect HSV-1 diagnoses with the DiaSorin kit were false-positive results for HSV-2 antibody. The high falsenegative rate for HSV-1 by the non gg-based tests would result in continued diagnostic uncertainty, at best, for a large proportion of patients with HSV-1 genital herpes. On the other hand, the high number of patients with HSV-1 with false-positive HSV-2 results could lead to incorrect counseling and treatment for patients and their partners. The Focus HSV-2 ELISA was positive in 39 (87%) of 45 first episodes of HSV-2. The false-negative Focus test results occurred in serum samples obtained 28 to 76 days after infection (median, 48 days); false-negative findings of the other tests occurred in serum samples obtained 30 to 90 days after infection (median, 43 days for the DiaSorin test and 45 days for the Wampole and Zeus tests). Previous studies using various gg-based tests including the Western blot have shown that individual patients seroconvert over a range of times. 4,16 In a recent study, we found that 93% of patients with primary HSV-2 episodes had a positive result in the HSV-2 Focus ELISA by 3 months (90 days), as did 73% of patients with nonprimary HSV-2 episodes. The median Am J Clin Pathol 2003;120: DOI: /AKWHF62YYJ054HA6 843

6 Morrow and Friedrich / INACCURACY OF CERTAIN HSV ELISA TESTS time to seroconversion shown by the Focus HSV-2 test was 21 days for primary infection and 23 days for nonprimary infection. 16 Similar results were noted in the present study: 93% (26/28) of patients with primary HSV-2 infection were positive by 90 days (median, 45 days; range, days), as were 76% (13/17) of patients with nonprimary HSV-2 infection (median, 39 days; range, days). The longer median times to apparent seroconversion in the present study likely were due to our testing only serum samples obtained at least 4 weeks after infection, thereby missing earlier seroconversion. Of 23 patients with subclinical HSV-2 infections, only the Wampole and Focus tests accurately identified all 23 as having HSV-2 antibodies. However, the specificity of the Wampole test for HSV-2 in patients without HSV-2 infection was only 83%, whereas the specificity for the Focus HSV-2 ELISA was 96% against the Western blot and 98% against infection status. Thus, false-positive test results in patients without a history of genital herpes are almost certainly reduced by using a gg-based test. Our study was limited by its emphasis on symptomatic patients; however, this emphasis permitted the clinical and virologic status, rather than solely a serologic assay, to function as the gold standard for test performance. In isolated patient groups, individual non gg-based tests functioned adequately; however, overall, these tests had inadequate specificity for an accurate diagnosis of either HSV-1 or HSV-2 in almost all clinical scenarios. In the light of the consistently more accurate performance of the Focus ggbased ELISAs and the widespread availability of this test, we consider the use of non gg-based tests unnecessary and inadvisable for diagnosis of genital HSV infection. From the 1 Department of Laboratory Medicine, University of Washington; and 2 Children s Hospital and Regional Medical Center, Seattle. Sera and patient data were collected by Lawrence Corey, MD, and Anna Wald, MD, MPH. The study was supported, in part, by NIH Herpes Program AI-30731, National Institutes of Health, Bethesda, MD. Address reprint requests to Dr Morrow: Virology, Room G815, 8G-3, Children s Hospital and Regional Medical Center, 4800 Sand Point Way, NE, Seattle, WA References 1. Fleming DT, McQuillan GM, Johnson RE, et al. Herpes simplex virus type 2 in the United States, 1976 to N Engl J Med. 1997;337: Wald A, Ashley-Morrow R. Serological testing for herpes simplex virus (HSV)-1 and HSV-2 infection. Clin Infect Dis. 2002;35(suppl 2):S173-S Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines MMWR Morb Mortal Wkly Rep. 2002;51: Ashley RL. Type-specific antibodies to HSV-1 and -2: review of methodology. Herpes. 1998;5: Ribes JA, Hayes M, Smith A. Comparative performance of herpes simplex virus type 2 specific serologic assays from Meridian Diagnostics and MRL Diagnostics. J Clin Microbiol. 2001;39: Turner KR, Wong EH, Kent CK. Serologic herpes testing in the real world. Sex Transm Dis. 2002;29: Ashley RL, Eagleton M, Pfeiffer N. Ability of a rapid serology test to detect seroconversion to herpes simplex virus type 2 glycoprotein G soon after infection. J Clin Microbiol. 1999;37: Ashley RL, Wald A, Eagleton M. Premarket evaluation of the POCkit HSV-2 type specific serologic test in culturedocumented cases of genital herpes simplex virus type 2. Sex Transm Dis. 2000;27: Ashley R, Cent A, Maggs V, et al. Inability of enzyme immunoassays to discriminate between infections with herpes simplex virus types 1 and 2. Ann Intern Med. 1991;115: Martins TB, Woolstenhulme RD, Jaskowski TD, et al. Comparison of four enzyme immunoassays with a Western blot assay for the determination of type-specific antibodies to herpes simplex virus. Am J Clin Pathol. 2001;115: Ashley RL, Militoni J, Lee F, et al. Comparison of Western blot (immunoblot) and glycoprotein G specific immunodot enzyme assay for detecting antibodies to herpes simplex virus types 1 and 2 in human sera. J Clin Microbiol. 1988;26: Langenberg A, Zbanyszek R, Dragavon J, et al. Comparison of diploid fibroblast and rabbit kidney tissue culture and diploid fibroblast microtiter plate system for the isolation of herpes simplex virus. J Clin Microbiol. 1988;26: Lafferty WE, Krofft S, Remington M, et al. Diagnosis of herpes simplex virus by direct immunofluorescence and viral isolation from samples of external genital lesions in a high prevalence population. J Clin Microbiol. 1987;25: Wald A, Zeh J, Selke S, et al. Reactivation of genital herpes simplex virus type 2 infection in asymptomatic seropositive persons. N Engl J Med. 2000;342: Wald A. Testing for genital herpes: how, who, and why. Curr Clin Top Infect Dis. 2002;22: Ashley RL, 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: Am J Clin Pathol 2003;120: DOI: /AKWHF62YYJ054HA6