Abstract
Resiquimod, a Toll-like receptor 7 and 8 agonist, stimulates production of cytokines that promote an antigen-specific T helper type 1 acquired immune response. Animal and phase II human trials showed posttreatment efficacy in reducing recurrent herpes lesion days and/or time to first recurrence. Three phase III randomized, double-blind, vehicle-controlled trials of topical resiquimod to reduce anogenital herpes recurrences were conducted in healthy adults with ≥4 recurrences within the prior year. Participants applied resiquimod 0.01% gel or vehicle gel 2 times per week for 3 weeks to each recurrence for 12 months. Trials 1 and 2 had 2:1 resiquimod-vehicle randomization. Trial 3 had 1:1:1 randomization for resiquimod and 500 mg valacyclovir orally twice daily for 5 days (RESI-VAL), resiquimod and oral placebo (RESI-PLA), and vehicle and oral placebo (VEH-PLA). The median time to first recurrence was similar for resiquimod and vehicle (trial 1, 60 and 56 days, P = 0.7; trial 2, 54 and 48 days, P = 0.47; trial 3, 51 [RESI-VAL], 55 [RESI-PLA], and 44 [VEH-PLA] days, P = not significant [NS]). The median time to healing of initial treated recurrence was longer for resiquimod (trial 1, 18 compared to 10 days, P < 0.001; trial 2, 19 compared to 13 days, P = 0.16; trial 3, 14 [RESI-VAL], 16 [RESI-PLA], and 8 [VEH-PLA] days, P < 0.001). In trials 1 and 2, moderate to severe erythema and erosion/ulceration at the application site were more common in resiquimod recipients. In conclusion, no posttreatment efficacy of resiquimod 0.01% gel was observed. Increased application site reactions and initial recurrence healing time are consistent with resiquimod-induced cytokine effects.
INTRODUCTION
Approximately 3% of the U.S. adult population has been diagnosed with anogenital herpes (1). Although anogenital herpes recurrences can be treated with episodic antiviral use and reduced with daily suppressive antiviral use, no currently available treatment impacts the natural history of the disease by reducing subsequent recurrences once nucleoside analogue therapy is stopped.
Resiquimod (R-848), an imidazoquinolinamine, is an investigational Toll-like receptor 7 (TLR-7) and TLR-8 agonist which stimulates an innate immune response with production of cytokines that lead to a subsequent antigen-specific T helper type 1 acquired immune response (2–5). Application of topical resiquimod to anogenital herpes lesions is hypothesized to act as an endogenous therapeutic herpes simplex virus (HSV) vaccine, with resiquimod acting like an adjuvant in the presence of endogenous antigen (HSV). The induced immunity could potentially delay or reduce future HSV recurrences. Resiquimod has been shown in an animal model (6) and phase II human trials (7, 8) to have some efficacy in reducing anogenital herpes recurrences. We report the results of three phase III randomized controlled trials of topical resiquimod to reduce anogenital herpes recurrences.
MATERIALS AND METHODS
Participants applied resiquimod 0.01% gel or vehicle gel 2 times per week for 3 weeks to each recurrence of anogenital herpes for 12 months. In trial 1, conducted in 38 European centers, and trial 2, conducted in 19 U.S. centers, both from 2000 to 2002, participants were randomized 2:1 to resiquimod versus vehicle. In trial 3, conducted in 18 U.S. and 7 Canadian centers from 2001 to 2002, participants were randomized 1:1:1 to (i) resiquimod 2 times per week for 3 weeks and 500 mg valacyclovir by caplet orally twice a day (BID) for 5 days, (ii) resiquimod 2 times per week for 3 weeks and oral placebo BID for 5 days, or (iii) vehicle 2 times per week for 3 weeks and oral placebo BID for 5 days. Each trial was approved by each trial center's independent ethics committee, and all participants gave written informed consent.
Participants in all three trials were healthy, 18 to 65 years of age, and HIV negative and had a history of frequently recurrent anogenital herpes. For inclusion in any of the three trials, participants had to have had ≥4 genital herpes recurrences per year in the past year or, if on suppressive oral nucleoside therapy, prior to starting suppressive therapy and at least one recurrence in the 3 months prior to screening and were willing to refrain from using topical or oral antivirals during the eligibility period and to use them only as directed by the investigator during the treatment period. Exclusion criteria included ever having used resiquimod (R-848) or imiquimod, pregnancy or breastfeeding, receipt of an HSV vaccine within the past 2 years, treatment of anogenital warts within the 4 weeks prior to enrollment or expected during the trial period, allergy to study gel excipient, hemoglobin of ≤9.4 g/dl, granulocyte count of ≤1.5 × 109 cells/liter, platelet count of ≤100,000 platelets per microliter, any serum chemistry value of grade 2 (moderate) or higher as defined in the protocol, or, in the 4 weeks prior to the screening visit, any use of an investigational or cytotoxic drug, interferon therapy or inducer, immunomodulator, systemic antiviral drug other than acyclovir, valacyclovir, or famciclovir, or systemic or high-dose inhaled corticosteroid.
After the screening visit, participants entered a 12-week eligibility period during which they had to have a qualifying recurrence to continue. Within 36 h after development of the qualifying recurrence, participants presented to the trial center, were randomized, and received a study drug in a blind manner. Participants treated each recurrence with the same regimen during the 12-month trial.
Randomization was in blocks of 6 and stratified by sex and trial center. To ensure proper randomization by trial center, subjects presenting with a qualifying recurrence were given a unique subject identification number in consecutive numerical sequence for that trial center, and study drug kits were issued in sequence. Study drug (resiquimod or vehicle gel), supplied in single-use sachets containing 225 mg of gel, was applied topically to lesions just before bedtime and washed off after 8 to 10 h. During the initial treatment cycle, participants returned to the trial center on days 3, 8, 15, and 22. On day 22, they received study drug for use during the next recurrence. For subsequent recurrences, participants self-initiated treatment within 24 h of lesion recurrence (prodrome-only and erythema-only events were not treated) and returned to the trial center within 72 h after onset for clinical confirmation. For the first recurrence after the initial treated recurrence, participants were seen on days 1, 8, 15, and 22. For all other subsequent recurrences, participants were seen on day 1 and 1 to 3 days after their last application of study drug. Additional scheduled visits were at weeks 12, 24, 36, and 52, unless these visits overlapped with a treatment cycle. Recurrences were assessed independently by participant diary and by the investigator during each clinic visit. Compliance with study drug was measured by participant diary and return of used and unused study drug sachets. Safety was measured by evaluation of adverse events, anogenital site assessments (local skin signs and local symptoms), and clinical laboratory tests. Anogenital site assessments occurred at each clinic visit during the initial and first subsequent treatment cycles, at the day 1 visit and the visits 1 to 3 days after the last dose for other subsequent recurrences, and at the last trial visit. During anogenital site assessments, investigators recorded the most severe signs for each category: erythema, edema, vesicles, erosion/ulceration, and scabbing (investigator-assessed signs). Participants were asked, without prompting for specific terms, if they had anogenital site signs since their last clinic visit (subject-reported signs). Information about size and number of lesions was collected as part of lesion assessments.
Laboratory methods.
HSV serology was done at the University of Washington by Western blotting (9). HSV PCR was done at Pasteur, Cerba, France (trial 1), at Children's Hospital Medical Center, Cincinnati, OH, USA (trial 2), and at Viridae Clinical Sciences, Vancouver, British Columbia, Canada (trial 3). HSV PCR swabs were collected at the day 1 visit for all recurrences.
Statistical analysis.
A clinical recurrence was defined as a period of days with lesions preceded and followed by at least 1 day without lesions. For trials 1 and 2, the primary endpoint was time to first subject-reported recurrence; for trial 3, it was annual recurrence rate. Secondary endpoints, depending on the trial, included time to first recurrence, annual recurrence rate, percentage of subjects who had no subsequent recurrences after the qualifying recurrence, total number of days the subject had a lesion present (lesion days), and time interval between the first and second recurrence. Recurrences were measured three ways: (i) subject-reported recurrences based on dates of lesion onset and resolution, (ii) investigator-confirmed recurrences, and (iii) PCR-confirmed recurrences. Kaplan-Meier analysis was used to calculate median time to first recurrence, defined as time from qualifying recurrence healing date to start date of the next recurrence, with the log rank test used to assess differences between groups. If the randomization recurrence healing date was missing, we used the randomization recurrence start date. Persons were censored on their last visit date. A Cox proportional hazards model stratified by site was used to test the treatment effect. Covariates considered included sex, age (≤40 versus >40 years), race (white versus non-white), baseline HSV-1 serostatus, and number of recurrences in the past year (≤8 versus >8).
Annualized recurrence rate was calculated by counting the total number of clinical recurrences occurring after the randomization recurrence and dividing by observation time, defined as time in years from the randomization recurrence clear date (or randomization recurrence start date if a clear date was unavailable) to the last visit date. Recurrence length was defined as the number of days during which lesions were present during a clinical recurrence. Recurrences of uncertain duration were those which did not have a clear date. To calculate a minimum recurrence length for these recurrences, we substituted the last date during that recurrence when the participant was seen for the clear date. Per-person lesion rates were calculated as the number of days with lesions divided by the number of days of observation, defined as the number of days from the start of the initial randomization recurrence to the last day of observation. The Wilcoxon rank sum test (trials 1 and 2) and the Kruskal-Wallis test (trial 3) were used to compare, by treatment arm, the median annualized subject-reported recurrence rate, median recurrence length, and median lesion rate. Fisher's exact test was used to compare, by treatment arm, the percentage of participants without subsequent recurrences. The Prentice-Williams-Peterson (PWP) gap time model (10), adjusted for treatment, sex, age, race, number of recurrences in the past year (≤8 versus >8 recurrences/year), and baseline HSV-1 serostatus, was used to model the time interval between the first and second subsequent subject-reported recurrences. In this model, the analysis of the second subsequent recurrence was necessarily restricted to subjects who had a first subsequent recurrence.
Power calculations were obtained using a simulation of 5,000 trials. For each subject, the gamma distribution was chosen to represent the number of days to each subsequent recurrence. The vehicle group was assumed to have a gamma distribution with a shape parameter of 2 and a scale parameter of 42, corresponding to a median time to first recurrence of 70 days and approximately 4 recurrences after the initial treated recurrence during the 1-year trial period. The resiquimod treatment group was assumed to have a shape parameter of 2 and a scale parameter of 63, corresponding to a median time to first recurrence of 105 days and approximately 2.6 recurrences after the initial treated recurrence. It was assumed that 40% of subjects would discontinue prior to completing the treatment period, and these subjects were assigned a value of 18 recurrences. For trials 1 and 2, 20% of subjects were assumed to have dropped out prior to experiencing a recurrence. For trials 1 and 2, 160 resiquimod and 80 vehicle subjects provided 96% power to detect a 50% difference in the median number of days to the first subsequent recurrence and 92% power to detect a 50% difference in the total number of recurrences. For trial 3, 105 subjects per treatment arm provided 81% power to detect a 50% difference in the total number of recurrences. A two-sided alpha level of 0.05 was used for each power calculation. Unless otherwise stated, all analyses were “intention to treat,” including all randomized subjects who were dispensed study drug.
Safety was measured by evaluations of adverse events, local skin signs and symptoms, time to recurrence healing, and size of lesions. Local skin signs (erythema, edema, vesicles, scabbing, and erosion/ulceration, as assessed by the investigator and participant) and local symptoms (pain, numbness/tingling, burning, and pruritus/itching, as assessed only by the participant) were summarized by maximum severity during the initial treatment cycle. The Kruskal-Wallis test was used to compare maximum severity scores across treatment groups. The Wilcoxon rank sum test was used to compare the median total investigator-confirmed lesion size during the initial treatment cycle between treatment groups. Time to healing of the qualifying recurrence was summarized using Kaplan-Meier survival methods. The Kruskal-Wallis test was also used to compare across all three trials by treatment group both trial and treatment completion rates and the proportion of participants who discontinued treatment due to adverse events or local skin symptoms or signs. Statistical analyses were conducted using Intercooled Stata 9.1 (College Station, TX).
RESULTS
The numbers of people who were assessed for eligibility, randomized, completed treatment, and completed trial are shown by trial and randomization arm in Fig. 1. Among those randomized, trial completion rates were 70%, 72%, and 70% in trials 1, 2, and 3, respectively, including 67%, 68%, and 69% who completed both trial and treatment and 3%, 3%, and 2% who completed trial but discontinued treatment during the trial, respectively (Fig. 1). Across trials, 307/451 (68%) resiquimod, 198/264 (75%) vehicle, and 73/103 (71%) RESI-VAL recipients completed the trial (P = 0.14). In all three trials, the median percentage of expected doses applied was 100% in all treatment groups. Across trials, the percentage of persons that applied all 6 doses during the initial treatment cycle was 89 to 94% in the resiquimod arm, 89 to 93% in the vehicle arm, and 89% in the RESI-VAL arm.
FIG 1.
Participant flow diagram for three randomized controlled trials.
Within each of the three trials, resiquimod, vehicle, and RESI-VAL participants were similar in baseline characteristics (Table 1). Consistent with previously published data (11, 12), male circumcision rates in the European trial (trial 1) were lower than those in the North American trials (trials 2 and 3). Observation time (from the start date of randomization recurrence to the healing date of the first recurrence or the last visit, whichever was later) was similar in all arms of all trials, with an overall median of 367 days (range of 1 to 473 days, interquartile range [IQR] of 262 to 378 days).
TABLE 1.
Summary of trials and demographic characteristics of participantsa
| Characteristic | Value for: |
||||||
|---|---|---|---|---|---|---|---|
| Trial 1 | Trial 2 | Trial 3 | |||||
| Location | Europe | United States | United States and Canada | ||||
| Randomization | 2:1 RESI:VEH | 2:1 RESI:VEH | 1:1:1 RESI-VAL:RESI-PLA:VEH-PLA | ||||
| Total no. of participants | 255 | 246 | 317 | ||||
| Randomization group (no. of participants) | RESI (n = 170) | VEH (n = 85) | RESI (n = 171) | VEH (n = 65) | RESI-VAL (n = 110) | RESI-PLA (n = 104) | VEH-PLA (n = 103) |
| Median age (range) in yrs | 40 (19–65) | 41 (21–64) | 38 (18–64) | 37 (19–64) | 37 (19–67) | 36 (19–62) | 38 (20–62) |
| No. (%) of men | 82 (48) | 38 (45) | 66 (39) | 28 (37) | 36 (33) | 36 (35) | 34 (33) |
| No. (%) of participants by race | |||||||
| White | 160 (94) | 76 (89) | 142 (83) | 66 (88) | 99 (90) | 94 (90) | 92 (89) |
| Black | 5 (3) | 7 (8) | 27 (16) | 9 (12) | 10 (9) | 9 (9) | 9 (9) |
| Native American | 1 (1) | 1 (1) | 0 | 0 | 0 | 0 | 1 (1) |
| Asian/Pacific Islander | 4 (2) | 1 (1) | 2 (1) | 0 | 1 (1) | 1 (1) | 1 (1) |
| No. (%) of men circumcised | 13 (16) | 7 (18) | 54 (82) | 26 (93) | 28 (78) | 29 (81) | 24 (71) |
| Median no. (range) of annual anogenital herpes recurrences | 7 (3–24) | 8 (4–30) | 6 (3–36) | 7 (4–36) | 8 (4–40) | 6.5 (4–28) | 8 (4–24) |
| No. (%) of HSV-1-seropositive participants | 104 (61) | 49 (58) | 78 (46) | 49 (65) | 52 (50) | 72 (65) | 53 (51) |
| No. (%) of HSV-2-seropositive participants | 156 (92) | 78 (92) | 157 (92) | 72 (96) | 98 (95) | 105 (95) | 101 (97) |
| No. (%) of participants with positive HSV PCR swab, qualifying recurrence | 126 (74) | 62 (73) | 78 (46) | 41 (55) | 71 (69) | 85 (77) | 75 (72) |
| No. (%) of participants with positive HSV PCR swab, any time during trial | 150 (88) | 76 (89) | 127 (74) | 58 (77) | 94 (90) | 98 (89) | 88 (85) |
RESI, resiquimod; VEH, vehicle; VAL, valacyclovir; PLA, oral placebo; HSV, herpes simplex virus.
Across all three trials, 818 participants had 3,779 clinical recurrences of anogenital herpes. Participants had a median of three recurrences (range of 1 to 22) per person. Recurrence duration was known for all but 174 (4.6%) recurrences and was a median of 11 (range of 1 to 75) days. The median minimum duration for the 174 recurrences of uncertain duration was 17 days (range of 1 to 31 days), and the median duration for all 3,779 recurrences (using the minimum duration if the total duration was unknown) was 11 (range of 1 to 75) days. Of all 3,779 recurrences, 1,952 (52%) were HSV positive, 867 (23%) were HSV negative, and 960 (25%) did not have a swab collected for virologic evaluation. PCR-positive recurrences had a longer median duration than both PCR-negative recurrences and recurrences from which no PCR swab was collected (13, 11, and 8 days, respectively, P < 0.001).
Time to first recurrence and between the first and second subsequent recurrences.
Time to first subject-reported recurrence after the initial randomization recurrence was similar in all trial arms in all three trials (Table 2 and Fig. 2a, b, and c). The Cox proportional hazards model for time to first subject-reported recurrence, both the crude unadjusted model and after adjusting for sex, age, race, number of recurrences in the past year, and baseline HSV-1 serostatus, also showed no statistically significant treatment difference with respect to time to first subject-reported recurrence in any of the three trials (data not shown). Results were similar when investigator-confirmed or virologically confirmed recurrences were used as the endpoint. No treatment differences were detected for the time interval between the first and second subsequent recurrences in any of the three trials.
TABLE 2.
Efficacy of resiquimod compared to vehicle in 3 randomized clinical trials (intent-to-treat analysis)a
| Characteristic | Value for: |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Trial 1 |
Trial 2 |
Trial 3 |
||||||||
| RESI | VEH | P value | RESI | VEH | P value | RESI-VAL | RESI-PLA | VEH-PLA | P value | |
| Median no. of days to first recurrence | 47 | 41 | 0.13 | 44 | 33 | 0.44 | 35 | 39 | 34 | 0.87 |
| Median annualized recurrence rate (no. of recurrences/yr) | 3.1 | 3.2 | 0.77 | 3.7 | 4.1 | 0.11 | 4.2 | 4.2 | 4.3 | 0.90 |
| Median lesion rate (%) | 20 | 14 | <0.001 | 20 | 19 | 0.24 | 22 | 21 | 16 | 0.007 |
| Median recurrence duration (no. of days) | 15 | 9 | <0.001 | 13 | 9 | <0.001 | 10 | 13 | 9 | <0.001 |
| Median no. of days to healing of initial recurrence | 18 | 10 | <0.001 | 19 | 13 | 0.16 | 14 | 16 | 8 | <0.001 |
RESI, resiquimod; VEH, vehicle; VAL, valacyclovir; PLA, oral placebo. Boldface indicates statistically significant results.
FIG 2.
Number of days to first herpes recurrence in trial 1 (a), trial 2 (b), and trial 3 (c) and number of days to resolution of randomization recurrence in trial 1 (d), trial 2 (e), and trial 3 (f).
Annualized recurrence rate, lesion rate, and recurrence duration.
The median annualized recurrence rates were also similar in all trial arms in all three trials (Table 2). The percentages of trial subjects with no recurrence after the initial randomization recurrence were similar between the randomized groups in all three trials (trial 1, 16% for resiquimod compared to 14% for vehicle, P = 0.40; trial 2, 16% for resiquimod compared to 11% for vehicle, P = 0.33; trial 3, 15% for RESI-PLA, 16% for VEH-PLA, 17% for RESI-VAL, P = 0.98). The median lesion rate (percentage of days with lesions) was higher in the resiquimod group than in the vehicle group in trial 1 (20% compared to 14%, P < 0.001), higher in the RESI-VAL and RESI-PLA groups than in the VEH-PLA group in trial 3 (22%, 21%, and 16%, respectively, P = 0.007), and similar in the resiquimod and vehicle groups in trial 2 (20% compared to 19%, P = 0.24). The median recurrence duration was longer in the resiquimod groups than in the vehicle groups in both trials 1 and 2 (Table 2) and longest in the RESI-PLA group, second longest in the RESI-VAL group, and shortest in the VEH-PLA group in trial 3 (13, 10, and 9 days, respectively, P < 0.001).
Pooled, subgroup, and sensitivity analyses.
Analysis of pooled data from the resiquimod and vehicle arms of trials 1, 2, and 3 also showed no difference in median days to first recurrence between the resiquimod and vehicle arms (44 and 34 days, respectively, P = 0.35). Using the pooled data, we conducted subgroup analyses to examine median days to first recurrence by treatment arm among all men, circumcised men, uncircumcised men, and all women. There was no statistically significant difference in time to first recurrence by treatment arm among any of these subpopulations, although there was a trend toward longer time to first recurrence among men receiving resiquimod than among those receiving vehicle (55 compared to 48 days, P = 0.08), which was limited to circumcised men (64 compared to 44 days, P = 0.10), with no difference seen among uncircumcised men (48 compared to 52 days, P = 0.45).
Across all three trials, 78 persons (9.5%) were observed for <60 days. Because theoretically these participants randomized to resiquimod may have had time to experience the cytokine-mediated inflammation and delayed healing of the randomization recurrence without having had enough time under observation to experience the potential benefits of resiquimod in delaying subsequent recurrences, we conducted a post hoc sensitivity analysis excluding these 78 persons and then reexamined lesion rate and median recurrence duration by trial and randomization group and number and percentage of trial participants with no subsequent recurrences after the initial randomization recurrence. The results for lesion rate for trials 1 and 3 were similar to the intent-to-treat analysis, but in trial 2 the difference in lesion rate between arms was no longer statistically significant (median lesion rate of 19% in the resiquimod group and 17% in the vehicle group, P = 0.17). The results for recurrence duration were identical to those of the intent-to-treat analysis. In all three trials, similar percentages of trial participants had no subsequent recurrences after the initial randomization recurrence (trial 1, 8.6% for resiquimod compared to 8.9% for placebo, P = 1.0; trial 2, 8.3% for resiquimod compared to 2.9% for placebo, P = 0.24; trial 3, 8.7% for RESI-VAL, 6.1% for RESI-PLA, and 9.6% for VEH-PLA, P = 0.68).
Duration of initial treated recurrence.
The median number of days to healing of randomization recurrence was 6 to 8 days longer in the resiquimod-treated groups than in the vehicle-treated groups in all three trials, although this reached statistical significance only in trials 1 and 3 (Table 2 and Fig. 2d, e, and f). There was no statistically significant difference in days to healing of the initial recurrence between the RESI-VAL and RESI-PLA groups in trial 3 (14 compared to 16 days, P = 0.35), although both were longer than in the VEH-PLA group (8 days, P < 0.001 for comparison with both RESI-VAL and RESI-PLA).
Adverse effects.
Topical application of resiquimod to anogenital herpes lesions was systemically well tolerated. There were no deaths, and all serious adverse events (6 in trial 1, 9 in trial 2, and 9 in trial 3) were considered by the study site principal investigator to be probably not related to resiquimod. No clinically meaningful differences between treatment groups in the number of subjects reporting systemic adverse events and no increases in adverse events compared with the vehicle group were observed that might be associated with systemic exposure to resiquimod.
The percentage of participants with moderate or severe investigator-assessed local signs during the initial treatment cycle was higher in the resiquimod group than in the vehicle group for erythema (trial 1, 57% compared to 41%, P = 0.017; trial 2, 47% compared to 32%, P = 0.006) and erosion/ulceration (trial 1, 47% compared to 34%, P = 0.007; trial 2, 28% compared to 20%, P = 0.003) but not trial 3 (Table 3) and for scabbing in trial 2 only. No differences were seen for edema or vesicles, nor were differences seen in subject assessments of local skin sign severity or the percentage of subjects who graded local symptoms as moderate or severe during the initial treatment cycle. The median investigator-confirmed total lesion size during the initial treatment cycle was greater at days 3, 8, and 15 for resiquimod than for vehicle (trials 1 and 2 only). Combining results from all three trials, resiquimod recipients were most likely to discontinue drug, followed by RESI-VAL recipients and then vehicle recipients (161/451 [36%], 30/103 [29%], and 70/264 [27%], respectively; P = 0.03). Overall, 5 resiquimod, 1 vehicle, and 2 RESI-VAL participants discontinued drug due to adverse events (P = 0.35), and 11 resiquimod, 2 vehicle, and 1 RESI-VAL participants discontinued drug due to local skin reaction (P = 0.55).
TABLE 3.
Investigator-assessed local skin signs at site of drug application during initial treatment cycle
| Moderate to severe local skin sign | % of participants |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Trial 1 |
Trial 2 |
Trial 3 |
||||||||
| RESI | VEH | P value | RESI | VEH | P value | RESI-VAL | RESI-PLA | VEH-PLA | P value | |
| Erythema | 57 | 41 | 0.017 | 47 | 32 | 0.006 | 38 | 42 | 30 | NS |
| Scabbing | 14 | 8 | 0.044 | 16 | 21 | NS | 15 | 23 | 16 | NS |
| Erosion/ulceration | 47 | 34 | 0.007 | 28 | 20 | 0.003 | 31 | 32 | 25 | NS |
RESI, resiquimod; VEH, vehicle; VAL, valacyclovir; PLA, oral placebo; NS, not significant. Boldface indicates statistically significant results.
DISCUSSION
These three randomized controlled trials of resiquimod 0.01% gel applied topically to lesions of recurrent anogenital herpes two times per week for 3 weeks for each recurrence over 12 months did not show any posttreatment efficacy of resiquimod as measured by time to first recurrence, annualized recurrence rate, or lesion rate. Although systemically well tolerated, use of topical resiquimod resulted in increased application site reactions which were only partially ameliorated by valacyclovir and also increased time to healing. These findings are consistent with proinflammatory effects of resiquimod-induced cytokines and are helpful to review as “prime and pull” HSV-2 vaccination strategies are being contemplated (15).
In a guinea pig model of anogenital herpes, subcutaneous resiquimod reduced recurrent lesion days both during treatment and after treatment discontinuation (6). Three human, phase II, randomized, double-blind, vehicle-controlled trials of resiquimod have been published (7, 8, 13). In the first, resiquimod recipients applied resiquimod 0.05% gel once or twice weekly or 0.01% gel twice or thrice weekly topically to anogenital herpes lesions. Those assigned to resiquimod had a longer median time to first recurrence than those assigned to vehicle (169 compared to 57 days, P = 0.006), with the group receiving 0.01% gel thrice weekly having the longest median time to first recurrence (>195 days) (7). In the second trial, participants with anogenital herpes applied resiquimod 0.01% gel or vehicle gel topically to herpes lesions 2 times weekly for 3 weeks and then collected daily anogenital swabs for 60 days for HSV detection. Recurrences during the subsequent 7 months were treated with study gel. During the final treatment-free 60 days, participants again collected daily swabs to assess shedding. The median lesion and shedding rates were statistically significantly lower for resiquimod than for vehicle recipients during the initial and final sampling periods, and resiquimod recipients tended to have a longer time to first recurrence than vehicle recipients (median of 41 compared to 28 days, P = 0.25) (8). These data from smaller trials suggested that resiquimod modifies the natural history of established HSV infection. In the third trial, participants with anogenital herpes applied resiquimod 0.01% gel or vehicle gel topically to herpes lesions 2 times weekly for 3 weeks within 24 h of recurrence onset and had daily lesion assessments and sampling for HSV DNA PCR for 21 days or until investigator-determined healing. No difference was observed between arms in median time to healing, maximum severity scores for investigator- or participant-assessed local skin signs/symptoms, or time to cessation of viral shedding (13).
Why were the phase III results different from the published results of three phase II trials of topical resiquimod (7, 8, 13)? Differences between the trials in dosing methodology, timing, proportion of recurrences which were HSV positive by PCR, or severity of disease in participants could be partially or completely responsible for the different findings. In the phase II efficacy trials, participants reported to the trial center within 24 h of the qualifying recurrence and received investigator-measured amounts of study drug dispensed from a tube in the center during the day, whereas in the phase III trials, participants reported to the trial center within 36 h of the qualifying recurrence and applied study drug later that evening at home from a single-use sachet. The increased delay between epithelial viral replication and drug application could have adversely affected drug efficacy in the phase III trials, especially if it resulted in less time with both resiquimod and HSV present together on the anogenital mucosa. The phase III trials also had a lower day 1 virological yield than the phase II trials (74% in trial 1, 48% in trial 2, and 69% in the RESI-VAL and 77% in the RESI-PLA groups in trial 3, compared with 85% by culture in the Spruance phase II trial) despite a more sensitive method of detecting HSV, although the phase III trial PCR tests were performed in different labs for each trial, hampering direct comparison. However, the lack of efficacy seen in the phase III trials even when restricted to virologically confirmed recurrences suggests that lower virologic yield in the phase III than that in the phase II trials is not responsible for the different outcomes. Reported adherence, measured by participant diary and return of used and unused study drug sachets, was quite good in the phase III trials, with ≥89% of all trial participants in all three trials in all randomized groups applying all six topical doses during the initial treatment cycle. Thus, inadequate adherence is unlikely to account for lack of efficacy, although it is possible that self-application of the study drug in the phase III trials had an adverse effect on efficacy if participants did not effectively apply an adequate amount of drug directly to the herpes lesions. It is also possible that the relatively high dropout rate in the phase III trials masked a small effect or that the favorable results in the two much smaller phase II trials with efficacy assessments were due to chance alone.
The results are also consistent with resiquimod having a small effect on HSV-2 reactivation after being applied directly to anogenital mucosa with HSV-2 present. Because HSV shedding occurs more frequently than clinical reactivation of anogenital herpes (14), shedding may be a more sensitive (and objective) measure of the pharmacologic efficacy of resiquimod at reducing future recurrences. Additional studies would need to be conducted to optimize this antiviral effect and translate it into clinical benefit. Thus, further investigation of immunomodulation, perhaps with better TLR-7 and -8 agonists or other agents, as a treatment strategy for anogenital herpes is warranted.
ACKNOWLEDGMENTS
This work was supported by 3M Pharmaceuticals and by grants K23 AI071257, K24 AI071113, and P01 AI030731 from the National Institute of Allergy and Infectious Diseases at the National Institutes of Health.
H.B.S. and T.-C.M. were employees of 3M Pharmaceuticals at the time these trials were done. T.-C.M. performed consulting work for Medicis Pharmaceutical Corporation when they owned the rights to topical resiquimod.
Appreciation is expressed to Ping Liu, formerly of 3M Pharmaceuticals, for assistance with data management and analysis, as well as to all trial participants, investigators, and staff.
Footnotes
Published ahead of print 7 April 2014
REFERENCES
- 1.Xu F, Sternberg MR, Gottlieb SL, Berman SM, Markowitz LE, Forhan SE, Taylor LD. 2010. Seroprevalence of herpes simplex virus type 2 among persons aged 14-49 years—United States, 2005-2008. MMWR Morb. Mortal. Wkly. Rep. 59:456–459 [PubMed] [Google Scholar]
- 2.Ahonen CL, Gibson SJ, Smith RM, Pederson LK, Lindh JM, Tomai MA, Vasilakos JP. 1999. Dendritic cell maturation and subsequent enhanced T-cell stimulation induced with the novel synthetic immune response modifier R-848. Cell. Immunol. 197:62–72. 10.1006/cimm.1999.1555 [DOI] [PubMed] [Google Scholar]
- 3.Wagner TL, Ahonen CL, Couture AM, Gibson SJ, Miller RL, Smith RM, Reiter MJ, Vasilakos JP, Tomai MA. 1999. Modulation of TH1 and TH2 cytokine production with the immune response modifiers, R-848 and imiquimod. Cell. Immunol. 191:10–19 [DOI] [PubMed] [Google Scholar]
- 4.Hemmi H, Kaisho T, Takeuchi O, Sato S, Sanjo H, Hoshino K, Horiuchi T, Tomizawa H, Takeda K, Akira S. 2002. Small anti-viral compounds activate immune cells via the TLR7 MyD88-dependent signaling pathway. Nat. Immunol. 3:196–200. 10.1038/ni758 [DOI] [PubMed] [Google Scholar]
- 5.Jurk M, Heil F, Vollmer J, Schetter C, Krieg AM, Wagner H, Lipford G, Bauer S. 2002. Human TLR7 or TLR8 independently confer responsiveness to the antiviral compound R-848. Nat. Immunol. 3:499. 10.1038/ni0602-499 [DOI] [PubMed] [Google Scholar]
- 6.Bernstein DI, Harrison CJ, Tomai MA, Miller RL. 2001. Daily or weekly therapy with resiquimod (R-848) reduces genital recurrences in herpes simplex virus-infected guinea pigs during and after treatment. J. Infect. Dis. 183:844–849. 10.1086/319262 [DOI] [PubMed] [Google Scholar]
- 7.Spruance SL, Tyring SK, Smith MH, Meng TC. 2001. Application of a topical immune response modifier, resiquimod gel, to modify the recurrence rate of recurrent genital herpes: a pilot study. J. Infect. Dis. 184:196–200. 10.1086/322007 [DOI] [PubMed] [Google Scholar]
- 8.Mark KE, Corey L, Meng TC, Magaret AS, Huang ML, Selke S, Slade HB, Tyring SK, Warren T, Sacks SL, Leone P, Bergland VA, Wald A. 2007. Topical resiquimod 0.01% gel decreases herpes simplex virus type 2 genital shedding: a randomized controlled trial. J. Infect. Dis. 195:1324–1331. 10.1086/513276 [DOI] [PubMed] [Google Scholar]
- 9.Ashley RL, Militoni J, Lee F, Nahmias A, Corey L. 1988. 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. 26:662–667 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Prentice RL, Williams BJ, Peterson AV. 1981. On the regression analysis of multivariate failure time data. Biometrika 68:373–379 [Google Scholar]
- 11.Xu F, Markowitz LE, Sternberg MR, Aral SO. 2007. Prevalence of circumcision and herpes simplex virus type 2 infection in men in the United States: the National Health and Nutrition Examination Survey (NHANES), 1999–2004. Sex. Transm. Dis. 34:479–484 [DOI] [PubMed] [Google Scholar]
- 12.Alanis MC, Lucidi RS. 2004. Neonatal circumcision: a review of the world's oldest and most controversial operation. Obstet. Gynecol. Surv. 59:379–395. 10.1097/00006254-200405000-00026 [DOI] [PubMed] [Google Scholar]
- 13.Fife KH, Meng TC, Ferris DG, Liu P. 2008. Effect of resiquimod 0.01% gel on lesion healing and viral shedding when applied to genital herpes lesions. Antimicrob. Agents Chemother. 52:477–482. 10.1128/AAC.01173-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wald A, Zeh J, Selke S, Ashley RL, Corey L. 1995. Virologic characteristics of subclinical and symptomatic genital herpes infections. N. Engl. J. Med. 333:770–775. 10.1056/NEJM199509213331205 [DOI] [PubMed] [Google Scholar]
- 15.Shin H, Iwasaki A. 2012. A vaccine strategy that protects against genital herpes by establishing local memory T cells. Nature 491:463–467. 10.1038/nature11522 [DOI] [PMC free article] [PubMed] [Google Scholar]