Chronic herpes simplex virus (HSV) ulcers were one of the first described manifestations of AIDS,1 and HIV-infected persons have higher clinical anogenital HSV recurrence rates,2 higher anogenital HSV-2 shedding rates,2,3 and higher oral HSV shedding rates3 than HIV-uninfected persons. Higher rates of both HSV shedding and HSV clinical recurrences among HIV-infected persons compared with HIV-uninfected persons are presumed to be caused by decreased host ability to control HSV replication. However, there is considerable variability in HSV-2 reactivation rates in HIV-infected persons as in HIV-uninfected persons, even among those with similar CD4 T-cell counts. In general, severe persistent HSV-2 genital ulcer disease has been associated with CD4 T-cell counts <50 cells per cubic millimeter.4 Recent data in immunocompetent adults have shown that both oral and anogenital HSV reactivations occur more frequently and are cleared more rapidly than previously appreciated, with 21% of oral and 24% of anogenital reactivations lasting ≤6 hours and 39% of oral and 49% of anogenital reactivations lasting ≤12 hours.5 Because the frequency and duration of short HSV reactivations in HIV-infected persons might differ from immunocompetent adults and because their presence might affect investigational HSV treatment strategies in HIV-1-coinfected persons designed to decrease HIV-1 transmission,6 we performed an observational pilot study to examine the frequency of short episodes of oral and anogenital HSV reactivation among HIV-infected persons.
Study Participants and Procedures
HSV-2-seropositive HIV-1-seropositive participants aged ≥18 years were recruited in 2006-2007 from a pool of prior research study participants known to be able to comply with an intensive study protocol and asked to collect oral and anogenital swab specimens for HSV DNA polymerase chain reaction (PCR) at home 4 times per day for 60 days. Swab specimens were collected at ∼6-hour intervals: upon awakening, in the midmorning, in the afternoon, and at bedtime. Participants recorded in a diary exact swabbing times and any symptoms present and were instructed not to take any herpes antiviral medication during the study. As previously described,3,5,7,8 anogenital swabs were obtained by rubbing a polyester fiber-tipped swab across first the penile and then the perianal skin for men or first across the posterior cervical/vaginal, then vulvar, then perianal areas for women. Oral swabs were performed by rubbing 1 swab across the buccal mucosa and tongue. Separate swabs were collected from any oral or anogenital lesions noted by the participant. Participants were seen in the clinic every 3 weeks for collection of samples and diary review. Blood was drawn for both CD4 T-cell count and serum HIV-1 RNA level within 1 month of starting the swabbing study, and antiretroviral use was obtained by history at the time of study initiation. Highly active antiretroviral therapy (HAART) use was defined as receipt of at least 3 drugs in 2 of the following categories: protease inhibitor, nucleoside analogue, or nonnucleoside analogue, at the time of study initiation. This study was approved by the University of Washington institutional review board, and all participants gave written informed consent.
HIV-1 seropositivity was confirmed by standard enzyme-linked immunoassay and Western blot. HSV serologic testing was performed by Western blot.9 Absolute CD4 T-cell counts were determined by flow cytometry. HIV-1 RNA levels were measured using real-time reverse transcriptase-polymerase chain reaction technology (detection level, 30 HIV-1 RNA copies per milliliter). Oral, anogenital, and lesion swabs were placed into separate vials containing 1 mL of PCR transport medium and stored at −20°C until laboratory processing. HSV DNA was detected using a quantitative real-time PCR assay and expressed as copies per milliliter of transport medium.10,11 As previously described, the initial PCR assay uses type-common primers to the HSV glycoprotein B gene, with positive samples subsequently analyzed using type-specific primers to determine whether DNA detected was that of HSV-1, HSV-2, or both.10,12 An internal control was included in the PCR reaction to ensure that HSV-negative findings were not due to inhibition. Samples were considered positive for HSV if we detected ≥3 copies of HSV DNA per 20 μL of specimen (ie, ≥150 copies of HSV DNA per milliliter of transport media).11 Laboratory personnel were blinded to clinical data.
Analyses were done using SAS (SAS Institute Inc, Cary, NC). We used identical definitions of HSV shedding, shedding rate, and shedding episode as in our previous work examining shedding in immunocompetent adults.5 Briefly, HSV shedding was considered to have occurred if an oral or anogenital sample or the corresponding lesion sample was positive for HSV at a given time. If both the oral or anogenital sample and the corresponding lesion sample were positive for HSV, the sample with the higher HSV DNA copy number was used in further analyses. Shedding rates were defined as number of swab specimens with HSV DNA detected divided by number of swab specimens collected. Number of days with samples collected was defined as total days from first to last day of specimen collection for both oral and anogenital samples, minus any days during which no samples were collected; expected number of days per participant was at least 60. On each day with any genital (or oral) sampling done, 4 genital (or oral) swabs were expected. A shedding episode of known duration was defined as 1 or a series of HSV-positive swab specimens that were collected immediately before and after at least 2 HSV-negative swab specimens. Episode start time was estimated as the chronological midpoint between the last HSV-negative and first HSV-positive swab specimen and episode stop time as the midpoint between the last HSV-positive and first HSV-negative swab specimen.
Because not all time points had swab specimens available, some subjects shed HSV for an unknown period. These are referred to as shedding episodes of uncertain duration because the episodes may have been longer than observed. For these episodes, we assumed missing swab specimens 2 time points before and 2 time points after the positive swab specimen(s) were HSV negative and estimated start and stop times as above. Any shedding episode (of known or uncertain duration) could include 1 missing or 1 HSV-negative swab specimen within the episode.
Generalized estimating equations were used to determine whether there was a difference in average episode duration between episodes of known and uncertain duration, whether there was an association between average per-episode maximum HSV DNA copy number and episode duration, and whether concurrent oral and anogenital shedding occurs more frequently than would be expected by chance alone. Wilcoxon rank sum tests were used to test for differences in median CD4 count and HIV viral load among persons taking and not taking HAART and to examine person-level differences in shedding rates and number of reactivations by anatomic site. Fisher's exact test was used to test for associations between episode length and anogenital symptoms and lesions. Generalized estimating equations were also used to examine associations between measures of HIV disease severity (CD4 count, HIV viral load, and HAART use) and duration of or maximum HSV copy number in shedding episodes. CD4 count was dichotomized at 200 cells per cubic millimeter and HIV viral load at 10,000 copies per milliliter based on prior studies indicating this degree of immunosuppression and HIV viral replication were associated with longer HSV shedding duration.2
Twenty participants collected oral and anogenital samples for a median of 62 (range 32-79) days, with 20 (100%) participants collecting samples for at least 30 days, 18 (90%) for at least 50 days, and 15 (75%) for at least 60 days. Participants, who included 19 men and 1 woman, had a median CD4 count of 426 (range 29-1066) cells per cubic millimeter and a median HIV-1 RNA level of 2280 (range <30 to 150,000) copies per milliliter (Table 1). Five (25%) had undetectable HIV-1 RNA levels, 12 (60%) were taking antiretroviral therapy [for a median of 4.4 years (range 1.1-9.9 years) before study entry], and 12 (60%) met the 1993 Centers for Disease Control and Prevention AIDS surveillance case definition.13 Of the 12 receiving antiretroviral therapy, 7 (58%) were taking a nonnucleoside reverse transcriptase inhibitor-based regimen (4 nevirapine and 3 efavirenz), 4 (33%) a protease inhibitor-based regimen (1 ritonovir-boosted lopinavir, 1 ritonovir-boosted nelfinavir, 1 ritonavir-boosted atazanavir, and 1 unboosted fosamprenavir), and 1 (8%) a triple nucleoside reverse transcriptor inhibitor-based regimen. Three participants had a CD4 count <200 cells per cubic millimeter at study entry. Median CD4 counts among persons taking and not taking HAART were similar (473 versus 423 copies/mL, P = 0.51), but median HIV viral load among persons taking HAART was lower than among persons not taking HAART (239 versus 32,700 copies/mL, P = 0.01). Seven of 8 (88%) participants not on HAART and 1 of 11 (9%) participants on HAART with viral load data had a HIV viral load above 10,000 copies per milliliter (P = 0.001), whereas 1 of 8 participants (13%) not on HAART and 2 of 11 participants (18%) on HAART with CD4 data had CD4 <200 cells per cubic millimeter (P = 0.74).
Frequency of Mucosal HSV Shedding
Oral samples were collected for 1201 days (95% of the expected 1269 days) and 4559 time points (95% of the expected 4804 time points) and anogenital samples for 1199 days (94% of the expected 1269 days) and 4544 time points (95% of the expected 4796 time points). Both oral and anogenital samples were collected at all 4 time points on 87% of days and at 3 of 4 time points on 7% of days, thus 94% of days had at least 3 daily oral and anogenital samples collected. HSV DNA was detected from oral samples on 58 (5%) days and 120 (3%) time points and from anogenital samples on 199 (16%) days and 535 (12%) time points (P = 0.002, indicating more frequent anogenital than oral shedding; Table 2). Nine participants (45%) had at least 1 oral sample in which HSV was detected and 15 (75%) at least 1 anogenital sample in which HSV was detected.
HSV typing was available for 109 (91%) of 120 oral samples and 533 (99.6%) of 535 anogenital samples; the copy number in the remaining samples was too low for typing. HSV-1 alone was found in 46 oral samples (42% of all oral samples and 46% of oral samples from HSV-1/HSV-2-seropositive participants) and 4 anogenital samples (1% of all anogenital samples and 5% of anogenital samples from HSV-1/HSV-2 seropositive participants), both HSV-1 and HSV-2 in 11 oral samples (10% of all oral samples and 11% of oral samples from HSV-1/HSV-2-seropositive participants) and 33 anogenital samples (6% of all anogenital samples and 20% of anogenital samples from HSV-1/HSV-2-seropositive participants), and HSV-2 alone in 52 oral samples (48% of all oral samples and 43% of oral samples from HSV-1/HSV-2-seropositive participants) and 496 anogenital samples (93% of all anogenital samples and 75% of anogenital samples from HSV-1/HSV-2-seropositive participants) (Table 2).
Number and Duration of Shedding Episodes
We identified 36 separate episodes of oral HSV shedding in 9 participants and 82 episodes of anogenital HSV shedding in 15 participants. Complete 4 times daily sampling allowed calculation of oral shedding duration for 29 episodes (81%) and anogenital shedding duration for 66 episodes (80%). The median duration of an oral HSV reactivation with complete sampling was 8 hours (range 4 hours - 11 days) and an anogenital reactivation with complete sampling, 11 hours (range: 4 hours to 14 days) (Table 2). Of the 29 oral episodes of known duration, 10 (34%) lasted ≤6 hours, 17 (58%) lasted ≤12 hours and 19 (66%) lasted ≤18 hours (Fig. 1A). No oral shedding episodes of known duration consisted of sole HSV-1 shedding; 17 (59%) were HSV-2 oral shedding episodes [median duration: 12 (range 5-36) hours], 6 (21%) included both HSV-1 and HSV-2 shedding [median duration 34 (range 6-253) hours], and 6 (21%) could not be typed [median duration: 6 (range 4-8) hours]. Of the 66 anogenital shedding episodes of known duration, 19 (29%) lasted ≤ 6 hours, 35 (53%) lasted ≤ 12 hours and 43 (65%) lasted ≤ 18 hours. Sixty-three (95%) contained HSV-2 only, 2 both HSV-1 and HSV-2, and 1 could not be typed (Fig. 1B). The median maximum copy number of HSV DNA detected during both oral and anogenital episodes increased with episode duration (for episodes lasting ≤24 hours and >24 hours 102.9 versus 104.7 copies orally, P = 0.009, and 103.3 versus 104.9 copies anogenitally, P < 0.001, Fig. 1C, D). Seven participants (35%) had at least 1 oral shedding episode which lasted ≤ 6 hours, 10 (50%) at least 1 anogenital episode ≤ 6 hours, 9 (45%) at least 1 oral episode ≤ 12 hours, and 11 (55%) at least 1 anogenital episode ≤ 12 hours. Of 3 people with CD4 <200 cells per cubic millimeter, 1 had 1 oral episode and 4 anogenital episodes lasting ≤6 hours, and 2 had at least 1 oral and at least 1 anogenital episode lasting ≤12 hours.
The median number of HSV reactivations of known duration among those who shed during the 60-day sampling period was 3 (range: 1-8) oral and 4 (range: 1-18) anogenital reactivations per person. Excluding the 11 participants who were not observed to shed orally, the median oral HSV reactivation rate was 1.4 (range: 0.4-9.5) reactivations per 30 days or 16 reactivations annually. Excluding the 5 participants who did not shed anogenitally, the median anogenital HSV reactivation rate was 2.1 (range: 0.5-8.6) reactivations per 30 days or 26 reactivations annually (P = 0.17 for comparison between oral and anogenital reactivation rates).
Associations Between Shedding and Symptoms
No oral HSV reactivation was accompanied by symptoms or lesions. Three of 63 anogenital episodes (5%) of known duration with symptom and lesion information were associated with lesions and 5 (8%) with symptoms. Shorter anogenital shedding episodes were less likely to be symptomatic than longer ones. None of 45 anogenital shedding episodes lasting ≤24 hours had lesions, compared with 3 of 21 anogenital episodes (14%) >24 hours (P = 0.029). Similarly, only 1 of 45 anogenital episodes (2%) lasting ≤24 hours was associated with symptoms compared with 4 of 21 anogenital episodes (19%) >24 hours (P = 0.032).
Concurrent Oral and Anogenital Shedding
Swab samples were collected concurrently from both oral and anogenital sites at 4499 time points, with HSV detected on both oral and anogenital swabs concurrently at 89 time points. Concurrent shedding occurred in 4 participants (shedding data from 2 of these 4 are shown in Fig. 2) and occurred more frequently than would be expected by chance alone: oral HSV shedding was detected on 17% (89 of 532) of time points when HSV was detected anogenitally but on only 1% (30 of 3967) of time points when anogenital shedding was not occurring (P < 0.001). Of the 89 time points with concurrent shedding, most (n = 37, 42%) involved HSV-2 at both sites, 24 (27%) involved oral HSV-1 with both HSV-1 and HSV-2 detected anogenitally, 14 (16%) involved oral HSV-1 and anogenital HSV-2, 6 (7%) involved both HSV-1 and HSV-2 orally and anogenital HSV-2, 6 (7%) involved oral untypeable virus and anogenital HSV-2, 1 (1%) involved both HSV-1 and HSV-2 at both sites, and 1 (1%) involved oral HSV-2 and both HSV-1 and HSV-2 anogenitally.
Effect of Immune Status on HSV Shedding
In our population, CD4 count, HIV viral load, and antiretroviral use did not affect the duration of HSV shedding episodes (the proportion of episodes ≤6 hours versus >6 hours, Table 3). However, univariate analyses showed both antiretroviral use and a plasma HIV RNA <10,000 copies per milliliter to be associated with a lower average per-episode maximum HSV copy number. In multivariate regression including both HIV viral load and antiretroviral use, HIV viral load ≥10,000 copies per milliliter was found to be associated with a 45% increase in mean per-episode maximum log HSV copy number (103.1 HSV copies per milliliter among those with HIV viral load <10,000 copies per milliliter compared with 104.6 HSV copies per milliliter among those with HIV viral load ≥ 10,000 copies per milliliter, 95% confidence interval: 33% to 59% increase, P < 0.001), whereas the association between maximum HSV copy number and antiretroviral use was no longer significant.
Our study indicates that oral and anogenital HSV-1 and HSV-2 reactivation are even more common in HIV-infected persons than previously appreciated. In particular, subclinical oral shedding of HSV-1 and HSV-2 is quite common. Most mucosal HSV reactivations in HIV-infected persons are short and subclinical, with a median anogential HSV reactivation duration of 11 hours. Twenty-nine percent of anogenital HSV reactivations last ≤6 hours and more than half (53%) last ≤12 hours.
We also found that concurrent oral and anogenital HSV shedding, often of the same viral type but sometimes of different viral type, occurred more frequently than would be predicted by chance, supporting the findings of Kim et al.3 It is of interest that simultaneous reactivation from oral and anogenital mucosa, often with different subtypes, happens more frequently than would be predicted to occur, suggesting common systemic or mucosal host factors influencing shedding. Interestingly, oral HSV-2 shedding was as common as oral HSV-1 shedding; among HSV-1/HSV-2-seropositive participants, 46% of oral samples showed HSV-1 alone and 43% showed HSV-2 alone. We do not know whether oral HSV-2 shedding during episodes of oral-genital contact with a HSV-2-negative partner can lead to partner acquisition of genital HSV-2 infection, but this is certainly biologically plausible.
These data on HSV reactivation duration in HIV-infected persons are similar to what we found in immunocompetent hosts, who had a median anogenital HSV reactivation duration of 13 hours, with 24% of reactivations lasting ≤6 hours and 49% lasting ≤12 hours.5 We studied an HIV-infected population that was only moderately immunosuppressed (median CD4 count 426 cells/mm3) and our results may have been different if we had enrolled only HIV-infected persons with more marked immunosuppression (CD4 < 200 cells/mm3, for example), but our finding of some short anogenital HSV reactivations even in our few study participants with CD4 <200 cells per cubic millimeter suggests that even quite immunosuppressed HIV-infected persons can continue to have short rapidly cleared HSV reactivations. The proportion of HSV shedding episodes which were ≤6 hours did not differ by CD4 count, plasma HIV RNA, or HAART use, suggesting that level of immunosuppression does not markedly affect HSV episode length, although small numbers of participants limit our power to find subtle differences.
All but 1 of our participants were men, limiting our ability to draw definitive conclusions about short HSV reactivations in HIV-infected women. However, we have previously shown a greater number of short (<12 hour) episodes in immunocompetent men than immunocompetent women and a lower median HSV viral load at genital shedding episode onset in immunocompetent men than immunocompetent women (103.2 copies/mL versus 104.5 copies/mL, P < 0.0001).5 Genital HSV shedding rates among immunocompetent women have been shown in some studies to be ∼40% higher than among immunocompetent men.8,14 Together these data suggest that immunocompetent women may shed genital HSV more frequently and in longer episodes than immunocompetent men; whether the same is true of HIV-infected women is unknown. Our participants were also relatively old (median age 45 years), suggesting many likely acquired HSV-2 many years ago. Whether HSV shedding episode duration differs among HIV-infected individuals with more recently acquired HSV-2 requires further study.
Our findings have important implications regarding interactions between HIV-1 and HSV-2 and potential investigational HIV-1 prevention methods which would focus on HSV-2 treatment. Observational data show that HSV-2 coinfection may increase HIV-1 transmission,15 possibly by HSV stimulating transcription of latent HIV-1.16 In HIV-negative HSV-2-seropositive persons, HSV-2-specific CD4+ and CD8+ T cells and plasmacytoid and myeloid dendritic cells, including cells expressing the C-type lectin receptor DC-SIGN, persist at the dermal epidermal junction for months after lesion healing, even with daily antiviral therapy.17,18 The frequent short bursts of HSV reactivation and the high annual anogenital reactivation rate (median of 26 anogenital reactivations annually) that we demonstrated here in HIV-infected persons likely also lead to persistent anogenital mucosal immune activation, potentially contributing to HIV infectiousness. We know that acyclovir 400 mg orally twice daily, the dose used to assess whether HSV suppressive therapy in HIV-1/HSV-2-coinfected persons could help prevent HIV-1 transmission,19 does not completely suppress HSV reactivation.7,20 More potent therapies, or an effective HSV vaccine, are needed to assess whether complete suppression of HSV reactivation in HIV-1/HSV-2-coinfected persons, or prevention of HSV-2 infection, could help prevent HIV-1 transmission.
Our findings also suggest that treatment of HIV-1 in HIV-1/HSV-2-coinfected persons might help prevent HSV-2 transmission. Plasma HIV RNA ≥10,000 copies per milliliter and lack of HAART use were associated with higher maximum HSV copy numbers during shedding episodes. These higher copy numbers likely lead to greater HSV-2 infectiousness.
In summary, we found that frequent short episodes of oral and anogenital HSV reactivation occur in HIV-infected persons, even those with relatively advanced immunosuppression, and that the median oral and anogenital HSV reactivation duration in HIV-infected hosts is surprisingly similar to that in HIV-uninfected hosts. Further study of the pathogenesis of HSV-induced chronic mucosal immune activation and its effect on HIV-1 transmission is warranted.
1. Siegal FP, Lopez C, Hammer GS, et al. Severe acquired immunodeficiency in male homosexuals, manifested by chronic perianal ulcerative herpes simplex lesions. N Engl J Med
2. Schacker T, Zeh J, Hu HL, et al. Frequency of symptomatic and asymptomatic herpes simplex virus type 2 reactivations among human immunodeficiency virus-infected men. J Infect Dis
3. Kim HN, Meier A, Huang ML, et al. Oral herpes simplex virus type 2 reactivation in HIV-positive and -negative men. J Infect Dis
. 2006;194:420-427. E-pub 2006 July 2012.
4. Bagdades EK, Pillay D, Squire SB, et al. Relationship between herpes simplex virus ulceration and CD4+ cell counts in patients with HIV infection. AIDS
5. Mark KE, Wald A, Magaret AS, et al. Rapidly cleared episodes of herpes simplex virus reactivation in immunocompetent adults. J Infect Dis
6. Lingappa JR, Kahle E, Mugo N, et al. Characteristics of HIV-1 discordant couples enrolled in a trial of HSV-2 suppression to reduce HIV-1 transmission: the partners study. PLoS One
7. Wald A, Corey L, Cone R, et al. Frequent genital herpes simplex virus 2 shedding in immunocompetent women. Effect of acyclovir treatment. J Clin Invest
8. Wald A, Huang ML, Carrell D, et al. Polymerase chain reaction for detection of herpes simplex virus (HSV) DNA on mucosal surfaces: comparison with HSV isolation in cell culture. J Infect Dis
9. 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
10. Jerome KR, Huang ML, Wald A, et al. Quantitative stability of DNA after extended storage of clinical specimens as determined by real-time PCR. J Clin Microbiol
11. Magaret A, Wald A, Huang M, et al. Optimizing PCR positivity criterion for detection of HSV DNA on skin and mucosa. J Clin Microbiol
12. Corey L, Huang ML, Selke S, et al. Differentiation of herpes simplex virus types 1 and 2 in clinical samples by a real-time taqman PCR assay. J Med Virol
13. CDC. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR
. 1992;41(no. RR-17).
14. 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
15. Gray RH, Wawer MJ, Brookmeyer R, et al. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. Lancet
16. Mosca JD, Bednarik DP, Raj NB, et al. Herpes simplex virus type-1 can reactivate transcription of latent human immunodeficiency virus. Nature
17. Zhu J, Koelle DM, Cao J, et al. Virus-specific CD8+ T cells accumulate near sensory nerve endings in genital skin during subclinical HSV-2 reactivation. J Exp Med
18. Zhu J, Hladik F, Woodward A, et al. Persistence of HIV-1 receptor-positive cells after HSV-2 reactivation is a potential mechanism for increased HIV-1 acquisition. Nat Med
19. Celum C, Wald A, Lingappa J, et al. Acyclovir and transmission of HIV-1 from persons infected with HIV-1 and HSV-2. N Engl J Med
20. Wald A, Zeh J, Barnum G, et al. Suppression of subclinical shedding of herpes simplex virus type 2 with acyclovir. Ann Intern Med