Famciclovir is a nucleoside analogue recently licensed for the treatment of herpes zoster and recurrent genital HSV infection [17]. To determine the efficacy of famciclovir for the suppression of HSV reactivation in HIV-infected persons, we did a double-blind, placebo-controlled, crossover trial in 48 persons with HIV infection.

Methods

Top Methods Results Discussion Author & Article Info References

Study Participants and Design

Persons infected with both HIV and HSV were recruited into our study through advertisements in local newspapers and referrals from private physicians. The study was approved by the University of Washington institutional review board. We screened 123 persons and enrolled 48. Reasons for exclusion were lack of HSV antibodies (14%), inability to complete or lack of interest in completing 4 consecutive months of daily home culture (80%), and unwillingness to go without therapy during a recurrence of HSV (6%). At enrollment, each participant completed a standardized interview designed to record history of previous HSV reactivation, frequency of previous antiherpesvirus therapy, and extent of HIV disease. We excluded persons if they were younger than 18 years of age, were currently infected or had previously been infected with acyclovir-resistant HSV, had known gastrointestinal disorders affecting absorption, or had received suppressive antiviral therapy with acyclovir in the 6 months before enrollment. Patients were paid $300 for completion of the protocol; those who completed only the first arm of the protocol were paid $150.

At study entry, each participant was randomly assigned to receive either famciclovir tablets, 500 mg twice daily, or placebo tablets (identical in appearance to the famciclovir tablets) twice daily, for 8 weeks. This 8-week period was followed by a 7-day washout period, during which no pills were taken, and then by a second 8-week period during which participants received whichever regimen they had not received during the first phase of the trial. We chose this study design because it controlled for the variability between individual persons in CD4 cell count (which can affect the frequency of HSV reactivation [13]) and because a 7-day washout period had been shown not to influence subsequent HSV reactivation [18].

Famciclovir and placebo were dispensed in 28-day supplies (56 pills). During the entire 119-day study period, patients obtained once-daily cultures of the oropharynx, genitals (urethra and penile shaft in men; cervicovaginal area and labia in women), and rectum. Participants returned to the clinic at 28-day intervals so that we could review their daily symptom diaries, monitor compliance, distribute more study medication and culture supplies, and assess safety. All participants were also asked to return to the clinic during episodes of genital herpes. At these visits, we performed genital examinations to confirm the presence of lesions and obtained additional cultures of the lesions. Both participants and investigators were blinded to culture results until the conclusion of the study. Neither open-label oral acyclovir nor topical anti-HSV products for treatment of an HSV recurrence were allowed.

Collection of Daily Cultures

The methods used to collect swabs for HSV isolation have been described elsewhere [16, 18]. Each participant collected one specimen daily from each of four anatomic sites, using a separate Dacron swab for each culture. Oral-pharyngeal cultures were obtained by inserting a swab into the mouth and vigorously rubbing it along the gum line and over the palate. Men obtained urethral cultures by rubbing a swab over the urethral opening and obtained penile cultures by rubbing a separate swab along the entire ventral and dorsal shaft of the penis. Women obtained labial cultures by rubbing a swab over the entire labia majora and minora. Cervicovaginal cultures were obtained by rubbing a swab over the exocervix and posterior vaginal fornix. Rectal cultures were obtained by inserting the swab approximately 2 to 4 cm into the anus and gently rotating it. Each swab was placed in a separate vial that contained viral transport media, was labeled with site and date, and was stored in the refrigerator. Participants were instructed to obtain the cultures at the same time each day, preferably upon awakening. Cultures were picked up by a courier within 36 hours of collection and were transported to the virology laboratory of the University of Washington, where they were immediately inoculated into tissue culture. Participants also filled out daily diary cards that recorded the site-specific presence or absence of symptoms (pain, tingling, numbness or itching, presence of lesions) and the number of pills taken each day. The cards were collected monthly and reviewed by the study clinician. Unused cards of medication were collected, and pill counts were done to confirm compliance records.

Laboratory Studies

Serologic and T-Cell Analysis

Seropositivity for HIV was confirmed by using standard enzyme-linked immunoassays and Western blot assays. CD4 cell subset analysis was done by using flow cytometry. Tests for antibodies to HSV were performed with Western blot analysis on serum specimens obtained at study entry [11, 19]. Patients were classified into three groups: those who were seropositive for HSV-1 only, those who were seropositive for HSV-2 only, and those who were seropositive for both HSV-1 and HSV-2.

Herpes Simplex Virus Culture and Sensitivity Testing

For isolation of HSV, each sample was inoculated in triplicate into 48-well microtiter plates that contained human diploid fibroblasts [20]. Each well was examined three times weekly for evidence of cytopathic changes of HSV infection. Cultures that showed cytopathic effects were confirmed and typed by using HSV-specific monoclonal antibodies.

The initial culture supernatant from each clinical isolate that was obtained while patients were receiving famciclovir was inoculated into human diploid fibroblast cells [21]. Cell-associated virus (1:200 dilution in MEM [minimal essential media]-10% fetal calf serum) was inoculated in duplicate onto freshly confluent diploid fibroblast cells in 24-well plates. Plates were rocked at 30 tilts per minute for 60 minutes; penciclovir (Smith-Kline Beecham Pharmaceuticals, Philadelphia, Pennsylvania) containing medium was then added at 11 serial log₂ dilutions from 40.96 to 0.04 µg/mL. Control wells with no penciclovir were also established. When controls without penciclovir demonstrated 4+ cytopathic effect, the Hybriwick kit (Diagnostic Hybrids, Athens, Ohio) was used to measure HSV DNA [22] according to the manufacturer's directions. The IC₅₀ value was estimated as the lowest concentration of drug that caused a 50% or greater reduction of mean cycles per minute hybridization of HSV-specific DNA probe to cell lysates. Values for IC₅₀ were calculated with a computer algorithm that used a Michaelis-Menten model and nonlinear least-squares curve fitting.

Definition of Terms

Total HSV shedding was defined as the total number of days on which HSV was isolated by culture (regardless of anatomic site) divided by the total number of days on which cultures were obtained. Asymptomatic HSV shedding was defined as the total number of days on which a participant reported no symptoms or lesions at an anatomic site from which HSV was isolated divided by the total number of days on which cultures were obtained. Symptomatic HSV shedding was defined as the total number of days on which a lesion or symptom was reported by the participant in the symptom diary at an anatomic site from which HSV was isolated divided by the total number of culture days. A recurrence of genital herpes was defined as lasting from the onset of lesions to the complete healing of lesions.

Statistical Analysis

Treatment effect was initially evaluated with intention-to-treat analyses that included all randomly assigned participants before crossover. Survival analysis was used to handle data censoring caused by early withdrawals. Time to first isolation of HSV (either HSV-1 or HSV-2) by culture, time to HSV-1 isolation, and time to HSV-2 isolation were examined. Differences in Kaplan-Meier survival curves between the famciclovir and placebo groups in the first treatment period were assessed by using the log-rank test. The relative risk for shedding was estimated by using the Cox proportional-hazards regression model with first-period treatment as the only predictor variable.

To determine the effect of famciclovir on reduction of HSV-1 and HSV-2 shedding rates and days with symptoms, we used crossover analysis methods to further analyze participants who successfully completed both arms of the study. Successful completion was defined as more than 28 days with cultures in each treatment arm. Because HSV shedding and symptom rates were not normally distributed, nonparametric tests were used. Wilcoxon rank-sum tests for carryover (residual) and period effects compared the sums and differences, respectively, of rates during placebo and famciclovir administration in participants who received famciclovir first compared with those who received placebo first [23]. In the absence of carryover and period effects, the Wilcoxon signed-rank test on the difference between placebo and famciclovir for each patient was used to examine treatment effect. This test is sensitive to a nonzero median difference or asymmetry between positive and negative differences.

Management of Protocol

After screening for eligibility, we assigned each patient to a treatment sequence (famciclovir followed by placebo or placebo followed by famciclovir) according to a computer-generated randomization code provided by the sponsor in blocks of four. Famciclovir and placebo tablets were identical in appearance and were presented to patients as four identical treatment packs containing weekly blister cards. The information on treatment assignments was held by the study sponsor and was not broken for any patient during the study. Neither the patients nor any personnel directly involved in conducting the study, monitoring the study, or reviewing the data knew the treatment assignments until the code was broken and the data were analyzed at the conclusion of the study.

Role of Study Sponsor

All data collection, analysis, and interpretation were done at the University of Washington independently of the study sponsor. In addition, preparation of the manuscript, decisions about what to include in the manuscript, and decisions about where to submit the manuscript for publication were all done independently of the industry sponsor. The final manuscript was submitted to the industry sponsor for internal review before submission; however, the authors were not obligated to use any suggested comments, revisions, or suggestions.

Results

Top Methods Results Discussion Author & Article Info References

Study Group and Compliance

We enrolled 48 participants (45 men and 3 women) who were seropositive for both HSV and HIV (Table 1). The median age at study entry was 34 years (range, 22 to 65 years). Eighteen of 48 patients (37%) had asymptomatic HIV infection at study entry, and the rest had AIDS. The median CD4 cell count at study entry was 384 cells/mm³. Sixteen participants were receiving antiretroviral therapy; half of the 16 were initially assigned to receive famciclovir and half were initially assigned to receive placebo. Three participants (6%) had antibodies to HSV-1 only, 18 (38%) had antibodies to HSV-2 only, and 27 (56%) had antibodies to both HSV-1 and HSV-2. No notable differences were seen in HIV status, CD4 cell counts at study entry, demographic variables, HSV serologic profiles, or history of HSV recurrences for those initially assigned to receive placebo and those initially assigned to receive famciclovir. Half of the participants were initially assigned to receive placebo, and half were initially assigned to receive famciclovir; 37 of 48 participants (77%) completed the first arm of the study. The other 11 participants (6 receiving famciclovir and 5 receiving placebo) withdrew from the study within 28 days of enrollment. The incidence of adverse events was similar with famciclovir and placebo.

Intention-to-Treat Analysis of the First Treatment Period

All 48 enrolled participants were included in the intention-to-treat analysis, which considered all data collected up to the time of crossover (Table 2). The median number of days on which cultures were obtained before crossover was 55 for patients initially assigned to receive placebo and 54 for patients initially assigned to receive famciclovir. Participants took 98% of prescribed pills during this period, and symptom diaries were kept for a median of 56 days in both groups. Overall, 5 of 24 participants (21%) assigned to famciclovir and 17 of 24 (71%) assigned to placebo shed HSV at least once during the initial treatment phase. Famciclovir therapy delayed the time to detection of HSV-1 and HSV-2 (Figure 1). We isolated HSV on 122 of 1114 days (11%) for the placebo group and 9 of 1071 days (1%) for the famciclovir group. Among the 30 persons who were seropositive for HSV-1, 1 of 14 receiving famciclovir (7%) and 6 of 16 receiving placebo (38%) shed HSV-1. Among the 45 persons who were seropositive for HSV-2, 4 of 23 (17%) receiving famciclovir and 14 of 22 (64%) receiving placebo shed HSV-2. By using the Cox model, we estimated that the relative risks for any HSV shedding, HSV-1 shedding, and HSV-2 shedding for famciclovir recipients compared with placebo recipients were 0.15 (95% CI, 0.06 to 0.42; P < 0.001), 0.18 (CI, 0.02 to 1.45; P = 0.066), and 0.15 (CI, 0.05 to 0.45; P < 0.001), respectively. In other words, total HSV shedding and HSV-2 shedding were significantly lower while patients were receiving famciclovir. Famciclovir therapy also substantially reduced the total number of days with lesions (from 11% to 4%) and the total number of days with symptoms (from 18% to 5%).

View larger version (20K): [in this window] [in a new window]   Figure 1. Time to first isolation of herpes simplex virus (HSV). Top. Time to first isolation of HSV-1 or HSV-2 among all 48 participants in the first treatment period. Marks on the curves indicate censoring times due to study dropout or completion of the first period without a positive culture. Middle. Time to first isolation of HSV-1 among the 30 participants seropositive for HSV-1 in the first treatment period. Bottom. Time to first isolation of HSV-2 among the 45 participants seropositive for HSV-2 in the first treatment period.

Crossover Analysis

Participants Who Completed Both Arms of the Study

Of 37 participants who completed the initial phase of the trial, 33 crossed over to receive the alternate study regimen and 29 completed both regimens. Three of four late withdrawals were due to the time requirements of the study, and the fourth occurred because the participant moved from the area. For the 29 participants who completed both study periods, the median number of days on which cultures were obtained was 54 while patients were receiving famciclovir, 55 while patients were receiving placebo, and 7 during the washout period. Participants took 98% of the prescribed doses.

Only 2 (7%) of the 29 participants who completed both arms of the study had neither symptoms nor viral shedding at any time during the study. Days with symptoms and days with shedding were both reduced by famciclovir therapy. Of the 29 participants, 18 (62%) had fewer days with symptoms and 20 (69%) had less HSV shedding while receiving famciclovir. The median difference in days with symptoms was 13.8% of days, and the median difference in days with total HSV shedding was 5.4% of days (P < 0.001 for both). Neither carryover effects nor period effects were statistically significant (P > 0.2 for all), indicating that the observed differences were related solely to therapy.

Participants with Antibodies to Herpes Simplex Virus Type 2 at Baseline Who Completed Both Arms of the Study

Of the 26 participants with HSV-2 who completed both arms of the study, 13 were initially assigned to receive placebo and 13 were initially assigned to receive famciclovir. Fifteen (58%) shed HSV-2 only while receiving placebo, 2 (8%) shed HSV-2 only while receiving famciclovir, and 4 (15%) shed HSV-2 during both arms of the study. All HSV-2 reactivations were anogenital. Five (19%) of 26 participants never had HSV-2 isolated from any anatomic site during the 17-week study period; 2 of the 5, however, had HSV-1 isolated from the oropharynx. The median rate of HSV-2 isolation was 5.6% of days (range, 0% to 50% of days) while participants were receiving placebo and 0% of days (range, 0% to 18% of days) while participants were receiving famciclovir (P = 0.001) (Table 3). Isolation of HSV-2 occurred on 9.7% of placebo days and 1.3% of famciclovir days, for an 87% reduction in viral shedding with famciclovir. Famciclovir reduced the frequency of total, symptomatic, and asymptomatic HSV-2 shedding. The effectiveness of famciclovir was similar in all of these analyses, for an 87% reduction in total shedding, a 76% reduction in asymptomatic shedding, and a 97% reduction in symptomatic shedding.

Twenty-five recurrences of genital herpes occurred among the 26 participants who were seropositive for HSV-2. Seven recurrences began while participants were receiving famciclovir, 16 began while participants were receiving placebo, and 2 began during the washout period. The median duration was 5 days for anogenital recurrences that began while participants were receiving famciclovir and 11 days for those that began while patients were receiving placebo. There were 85 days with anogenital symptoms during famciclovir therapy (5.8% of days during therapy) and 243 days with anogenital symptoms during placebo administration (16.5% of days during administration), for a 65% reduction in days with anogenital symptoms (P = 0.007). The percentages of days with herpetic lesions (4.9% compared with 13.8%) and of days with symptoms but no lesions (0.9% compared with 2.7%) were lower during famciclovir therapy than during placebo administration. For symptoms as well as shedding, no significant carryover or period effects were seen (P > 0.2 for both). In other words, findings were similar in participants who received famciclovir first and participants who received placebo first.

Participants with Antibodies to Herpes Simplex Virus Type 1 Who Completed Both Arms of the Study

Famciclovir also reduced HSV-1 reactivation. Among the 20 participants with antibodies to HSV-1, 7 (35%) shed HSV-1 only while receiving placebo, 1 (5%) shed HSV-1 only while receiving famciclovir, and 2 (10%) shed HSV-1 during both arms of the study. The remaining half of these participants never shed HSV-1. Shedding of HSV-1 occurred on 26 of 1084 placebo days (2.4%) and 6 of 1053 famciclovir days (0.6%) and was always oral-labial. All famciclovir days and 77% of placebo days with HSV-1 shedding were asymptomatic. Famciclovir reduced the percentage of days with oral symptoms from 7.3% with placebo to 1.3% with famciclovir. The reductions in shedding and symptoms were statistically significant (P = 0.02 for both), and no carryover or period effects were found (data not shown).

Shedding of Herpes Simplex Virus Type 2 during Famciclovir Therapy

Six participants shed HSV-2 while receiving famciclovir (Table 4). In two participants, isolated days of shedding were seen on days 22 and 46 of famciclovir therapy. A third participant shed HSV-2 on days 35 and 36 of therapy. Four days of shedding in two other patients began on days 1 and 2 of famciclovir therapy. Half of the episodes of breakthrough HSV-2 shedding were associated with missed or inadequate famciclovir doses. All HSV-2 isolates obtained during famciclovir treatment were susceptible to famciclovir in vitro (IC₅₀ 2.9 µg/mL).

Discussion

Top Methods Results Discussion Author & Article Info References

This placebo-controlled trial was designed to evaluate the effect of daily antiviral therapy on HSV reactivation in HIV-seropositive persons. Several new findings about the natural history and treatment of HSV in HIV-positive persons emerge from our observations. First, HSV reactivation is more common than previously appreciated [24]. Of the 29 patients who completed both arms of our study, 27 (93%) had either symptoms or HSV shedding at some time during the study. Among the 26 participants with antibodies to HSV-2, 21 (81%) shed HSV-2. Of the 20 with antibodies to HSV-1, half shed HSV-1. Moreover, we observed that most HSV shedding was asymptomatic; HSV-1 was isolated only from the oropharynx and HSV-2 only from perirectal mucosa or genital skin.

We used a crossover trial design to allow measurement of the reduction of HSV reactivation by paired analysis as well as by the more standard intention-to-treat analysis. This design allowed us to evaluate patterns more accurately over time and independently of differences related to progression of HIV disease [13]. Both analytic methods revealed a treatment effect of famciclovir on HSV reactivation. In the intention-to-treat analysis of the first treatment period, the time to the first HSV reactivation was delayed in the famciclovir group. In the paired analysis, famciclovir resulted in a marked reduction in total, symptomatic, and asymptomatic HSV shedding. Daily famciclovir therapy reduced total HSV-2 shedding by 87%. In addition to the significant virologic benefit of therapy, the frequency of genital signs and symptoms was reduced by 65%. The percentage of days with genital lesions was reduced from 13.8% to 4.9%. Recurrences of genital herpes that began while participants were receiving placebo lasted a median of 11 days; recurrences that began while participants were receiving famciclovir lasted a median of 5 days. Thus, even if breakthroughs occurred, famciclovir shortened the course of the recurrences.

Famciclovir is the oral formulation of penciclovir and is 77% bioavailable [25]. Penciclovir is a selective substrate for the HSV-1, HSV-2, and varicella-zoster virus thymidine kinase. Viral thymidine kinase phosphorylates penciclovir to penciclovir monophosphate; this allows cellular enzymes to produce penciclovir triphosphate, which selectively inhibits the viral DNA polymerase [26]. Penciclovir triphosphate has an intracellular half life of 10 to 20 hours in HSV-infected cells [26]. The in vitro IC₅₀ of penciclovir for HSV-1 and HSV-2 is 0.4 to 0.6 µg/mL and 1.5 to 2.4 µg/mL, respectively. A single oral 500-mg dose of famciclovir resulted in maximum plasma concentrations of penciclovir ranging from 2.73 to 3.97 µg/mL [27].

Although famciclovir therapy was very effective, HSV-2 shedding still occurred during this therapy in 6 (23%) of 26 persons who were seropositive for HSV-2. Most of these episodes were single days of asymptomatic shedding. Of the five episodes of multiple days of shedding during famciclovir therapy, two occurred within the first 48 hours of therapy and one was associated with noncompliance. Isolates from these reactivations were sensitive to penciclovir in vitro [28]. Thus, breakthroughs during famciclovir therapy were due to shedding before the onset of the antiviral effects of the drug, noncompliance, or unknown causes. These findings are similar to those in previous studies of oral acyclovir in immunocompetent patients [16].

Famciclovir was well tolerated in our study. No patient discontinued therapy because of toxicity or side effects related to medication over 8 weeks of follow-up. Our dropout rate (40%) was greater than the 24% rate in our earlier study of similar design in HIV-seronegative women [16]. Participants who withdrew from the study reported complications of HIV as the main reason for dropping out.

Two (7%) of the 29 participants in the crossover analysis had no symptoms or shedding during the 17-week study period; this indicates that in a clinical setting, some patients may not benefit from daily suppressive antiviral chemotherapy. Their CD4 cell counts were 440 and 688 cells/mm³, respectively. On the other hand, half of the participants with HSV-2 who completed the crossover study had reductions in both anogenital symptoms and HSV-2 shedding while receiving daily famciclovir therapy. In these 13 participants, the median percentage of days with anogenital symptoms during placebo administration was 30% and the percentage of days with HSV-2 shedding was 13%; both of these measures were reduced to a median of 0 during famciclovir therapy. Reactivation of HSV itself in some persons has been associated with increased levels of HIV-1 RNA in plasma [29]. Thus, suppression of reactivation may also offer potential benefit in reducing chronic T-cell activation [30]. In summary, our data indicate that suppressive antiviral chemotherapy offers clinical benefit to HIV-infected persons with frequently recurring HSV. The cost–benefit ratios for intermittent and daily therapy have not been compared. Despite these positive findings and the widespread use of antiviral chemotherapy for HSV in HIV-infected persons, many questions still need to be answered before guidelines for the optimal management of HSV infection can be developed.

Presented in part at the Third Antiviral Conference on Human Retrovirus, January 1996, Washington, D.C.

Drs. Hu and Zeh: University of Washington, Department of Statistics, Box 354322, Seattle, WA 98195-4322.

Drs. Koelle and Corey: Fred Hutchinson Cancer Research Center, 1124 Columbia Street, M115, Seattle, WA 98104.

Dr. Saltzman and Mr. Boon: SmithKline Beecham, 1250 South Collegeville Road, PO Box 2089, Collegeville, PA 19426-0989.

Ms. Shaughnessy and Ms. Barnum: Virology Research Clinic, 1001 Broadway, Suite 320, Seattle, WA 98122.