METHODS

Study design

Data were collected as part of a cross sectional surveillance study to assess the prevalence of aciclovir resistance in genital herpes.⁹ Participants were enrolled at 22 STD and HIV clinics in seven urban areas across the United States from October 1996 to April 1998. Eligible participants were 18 years and older and were suspected of having active genital herpes. Participants’ genital lesions were cultured for HSV and positive cultures were typed.⁹ Participants also answered a short questionnaire.

Analysis

We analysed data from 940 participants who reported a previous occurrence of genital herpes and whose current lesions cultured positive for HSV. We calculated proportions of HSV-1 and HSV-2 according to various epidemiological characteristics and evaluated significance of univariate associations using Pearson’s χ² and Fisher’s exact tests. We used multivariate logistic regression models to identify independent risk factors for type specific infections. The final model included variables whose associated p value in the univariate analysis was less than 0.2, and an interaction term for sex and lesion site.

We used a stratified Cox proportional hazards model to detect differences in the duration of genital lesions (number of days the preceding genital herpes episode lasted) caused by HSV-1 compared with those caused by HSV-2. The model stratified by race and controlled for sex, age, ethnicity, HIV status, type of sexual partners in the preceding year (none, opposite sex, same sex, both sexes), and whether the preceding episode was known to be a recurrence. Because we modelled time until the lesion healed, longer lesion duration corresponded to a smaller hazard.

RESULTS

Among the HSV positive cultures, 4.2% were positive for HSV-1 and 95.8% for HSV-2. The proportion of HSV-1 was significantly higher in white people and individuals with 0–2 recurrences in the previous 12 months and, among men, in those with rectal/perirectal lesions rather than penile lesions (table 1). Although the proportion of HSV-1 was also higher among women and individuals reporting same sex partners only, these differences were not statistically significant. In multivariate analyses, the significant associations persisted; in addition, among those without rectal/perirectal lesions, the proportion of HSV-1 was significantly higher in women than in men (table 2).

There was little difference between the distributions of lesion duration for HSV-1 and HSV-2 (table 1). In a stratified Cox proportional hazards model, longer lesion duration was not significantly associated with virus type (HSV-1 compared with HSV-2: HR 0.95, 95% CI 0.65 to 1.38, p = 0.79), but was associated with male sex (HR 0.85, 95% CI 0.74 to 0.99, p = 0.04), and HIV seropositivity (HR 0.62, 95% CI 0.48 to 0.81, p<0.01).

DISCUSSION

In one of the largest studies to date of recurrent genital herpes, we examined the epidemiology of type specific HSV infection. Several aspects of the study were unique. We enrolled patients from many US geographical areas, included HIV infected patients, had sufficient patients to study recurrent genital HSV-1, and used statistical techniques that allowed for the identification of independent risk factors for type specific culture positivity and lesion duration. In this population, HSV-1 caused 4.2% of recurrent lesions. HSV-1 was more likely to be present in recurrent lesions of white people, individuals with few recurrences during the past year and, among men, in rectal/perirectal lesions rather than penile lesions. Lesion duration did not vary by HSV type, but did vary by sex and HIV status.

The percentage of recurrent lesions caused by HSV-1 (4.2%) was considerably less than the percentage of first episode lesions caused by HSV-1 (14.6%) in other participants at the same clinical sites (Jevitt et al, in preparation); this is consistent with reports that genital HSV-1 recurs less frequently than genital HSV-2.^6,7,10 This conclusion is also supported by the higher prevalence of HSV-1 among individuals with fewer recurrences during the past year (table 2). The proportion of recurrent lesions caused by HSV-1 (4.2%) was less than half that reported by Lafferty et al (9.9%).¹⁰ This difference does not appear to be explained by differences in sex practices, as our HSV-1 percentages were less within the subgroups of heterosexual men, heterosexual women, and men who have sex with men (MSM). The differences might be due to overall differences in the patient populations, because our study was from numerous clinical sites of wide geographical distribution within the United States whereas Lafferty et al¹⁰ enrolled patients at one STD clinic in Seattle.

The ratio of HSV-1 prevalence in black compared to white people was similar for recurrent (1.3% v 6.5%, odds ratio (OR) = 0.18) and first episode lesions (5% v 27%, OR = 0.1) (Jevitt et al, in preparation), suggesting that genital HSV-1 recurrence rates do not differ by race, but that white people are simply more likely to be genitally infected with HSV-1. This difference may be due to white people having a higher frequency of receptive oral sex¹¹ or being less likely to have acquired oral HSV-1 at an early age.¹²

Recurrent rectal/perirectal lesions in men were significantly more likely to be caused by HSV-1 than were recurrent penile lesions (OR = 8.4). This was also true for first episode lesions but differences by anatomical site were less marked (OR = 2.6) (Jevitt et al, in preparation), suggesting that HSV-1 is both more likely to infect the rectal/perirectal region and more likely to recur in individuals whose primary lesions were in this region. Among MSM, the higher prevalence of HSV-1 in rectal/perirectal lesions might be explained by oral/anal transmission of HSV-1, either by oral–anal contact or by use of saliva as a lubricant for anal sex.

Finally, recurrent lesion duration did not differ by viral type: a finding recently reported by Engelberg et al.⁷ Our study adds the observation that the lack of difference is not due to confounding by variables such as sex, race, sexual orientation, or lesion site. Our results are consistent with the previously established association between longer lesion duration in men¹³ and HIV infected individuals.²

In summary, we found that in the United States, recurrent genital HSV-1 is relatively rare in the general STD and HIV clinic setting, especially among black people. Among men, rectal/perirectal recurrent lesions are more likely to be caused by HSV-1 than are penile lesions. In addition, lesion duration depends on sex and HIV status but not virus type. These findings shed new light on the type specific epidemiology of recurrent genital HSV, and suggest that type specific testing can inform the prognosis and management of genital herpes.

DISCLAIMER

The use of trade names and commercial sources is for identification only and does not imply endorsement by the Centers for Disease Control and Prevention or the US Department of Health and Human Services.