A short, 8-week course of imiquimod 5% cream versus podophyllotoxin in the treatment of anogenital warts: A retrospective comparative cohort study

Introduction

Anogenital warts rank among the most frequent sexually transmitted infections.¹ Treatment modalities for the management of anogenital warts are divided into provider-administered (such as cryotherapy) and patient-applied (imiquimod, podophyllotoxin, and sinecatechins).^2,3

Imiquimod and podophyllotoxin are the most commonly used patient-applied treatments for anogenital warts. To our knowledge, there are only two studies comparing imiquimod and podophyllotoxin in the treatment of anogenital warts which concluded that both treatments were equally effective.^4,5 Reports on the cost-effectiveness of anogenital wart treatments favor podophyllotoxin over imiquimod, because imiquimod is more expensive and the duration of treatment with imiquimod is longer.^6,7 On the other hand, because imiquimod enhances the immune responses against human papillomavirus, recurrence rates following treatment with imiquimod are lower, compared with podophyllotoxin.⁷ Thus, the choice between the two agents is not clear cut.

Imiquimod can be used for up to 16 weeks.^2,3 Shorter courses (4-week, 8-week, and 12-week) have been evaluated and produced results comparable to the 16-week course.^8,9

In this study, we compared response rates between a short, 8-week course of imiquimod 5% cream and podophyllotoxin in patients with anogenital warts. Apart from general response rates, we also assessed factors that may affect response to treatment. We decided to evaluate an 8-week course of imiquimod instead of the licensed 16-week course, because we need more data on this more practical and less expensive imiquimod regime.

Methods

In this retrospective cohort study, we reviewed the database of all patients with anogenital warts who presented to the sexually transmitted infections unit of “Andreas Sygros” hospital for skin and venereal diseases, between January 2012 and December 2016. The study was approved by the institutional review board.

The inclusion criteria were the following: (a) otherwise healthy males and females age 18 years and older with a clinically diagnosed first episode of anogenital warts lasting no more than 3 months and (b) monotherapy with either an 8-week course of imiquimod 5% cream or a 4-week course of podophyllotoxin. Exclusion criteria included (a) receiving any treatment for anogenital warts before presentation to our department, (b) presence of either intra-anal/intravaginal warts or warts of the urethral meatus, (c) presence of giant condylomata (Buschke–Lowenstein tumors), (d) presence of only one wart or wart area larger than 6 cm², and (e) pregnancy.

Anatomical sites of anogenital warts were divided into two groups. The first group included sites covered by dry, keratinized skin, such as pubic area, penile shaft, scrotum, groin, and the outer surface of labia majora of the vulva (“keratinized” sites). The second group included sites covered by moist, partially keratinized skin, such as perianal area, perineum, preputial cavity, inner surface of labia majora, and labia minora of the vulva (“partially keratinized” sites). Disease extent was defined as small, if the number of lesions was between 2 and 4 or affected area was <1 cm²; medium, if lesions were between 5 and 10 or area between 1 and 4 cm², and large, if lesions were more than 10 or area affected was 4–6 cm².

Imiquimod was applied three times weekly. Podophyllotoxin was applied each week twice daily for 3 consecutive days, followed by 4 days of rest. Podophyllotoxin 0.5% solution was prescribed for lesions on “keratinized” sites and podophyllotoxin 0.15% cream was prescribed for lesions on “partially keratinized” sites, in agreement with the recent European guidelines.²

For the comparison of the response rates between the different treatment groups, Chi-square test was used. We also performed a stratified analysis using Chi-square test for the comparison of response rates among the categories of location (“keratinized” sites vs. “partially keratinized” sites). For the comparison of medians between independent groups of patients, we used Mann–Whitney U-test.

Estimated odds ratios (OR) and 95% confidence intervals (CIs) were determined using a logistic regression model. To account for the observational nature of the study and possible confounding, inverse probability of treatment weighted logistic regression was also applied, with weights stabilized by the marginal probabilities of receiving each treatment. Analyses were conducted in R, version 3.5.1, and IPT weights were estimated using package “ipw.”

Results

In total, 347 patients were included in the study. Demographic and baseline characteristics of the treatment groups are shown in Table 1.

Complete clearance rates and more than 50% reduction in wart area for the groups as a whole and separately for lesions on “keratinized” and “partially keratinized” sites are shown in Tables 2 and 3. For lesions on “keratinized” sites, the difference between imiquimod and podophyllotoxin both in clearance rates (7.6% and 27.9%, respectively) and in >50% reduction in wart area (54.3% and 76.0%, respectively) was statistically significant.

The results of the univariate logistic regression for patients with >50% reduction in wart area versus <50% reduction in wart area are presented in Table 4.

Multivariate analysis revealed that only type of treatment and location of lesions were independent predictors of response to treatment [Table 5]. More specifically, treatment with podophyllotoxin [OR (95% CI): 1.9 (1.11–3.23), P = 0.018] and location on “partially keratinized” sites [OR (95% CI): 3.75 (1.88–7.49), P = 0.0001] were statistically significantly associated with > 50% reduction in wart area.

Due to the observational nature of the study and nonrandom assignment of treatments, an inverse probability of treatment weighted logistic regression model was estimated. Stabilized inverse probability of treatment weights were calculated with marginal treatment probabilities as stabilizing factors and age, gender, extent of lesions, and location of lesions as probability of treatment factors. inverse probability of treatment weighted diagnostics demonstrated that weighting resulted in a balanced sample between treatment groups in important predictors for treatment and response [Table 6]. The inverse probability of treatment weighted model estimated a slightly smaller treatment effect of podophyllotoxin [OR (95% CI): 1.79 (1.08–2.97), P = 0.024], while the rest of the variables were found to have similar effects compared with the unweighted analysis [Table 5].

Discussion

In everyday clinical practice, imiquimod and podophyllotoxin are often considered first-line treatments for anogenital warts.

Imiquimod is shown to have a clearance rate ranging widely between 35% and 75% for treatment periods up to 16 weeks.^5,8-14 In our study, the imiquimod clearance rate after 8 weeks of treatment was 25.2%. However, the clearance rate was 50%, for lesions on “partially keratinized” sites, and 7.6%, for lesions on “keratinized” sites. So, if imiquimod is prescribed for patients with warts on “partially keratinized” sites, our study showed that even an 8-week course can lead to a clearance rate comparable to the clearance rates mentioned in the literature for treatment up to 16 weeks.

Podophyllotoxin 0.5% solution has been shown to have a clearance rate between 54% and 85% after use for 4 weeks.^5,15-18 For podophyllotoxin 0.15% cream, clearance rates of 63%¹⁹ and 75%¹⁸ have been reported. In this study, the clearance rate was 27.9% for podophyllotoxin solution (for lesions on “keratinized” sites) and 52.2% for podophyllotoxin cream (for lesions on “partially keratinized” sites). The low clearance rate that we observed for podophyllotoxin 0.5% solution can be attributed to the fact that we used it only for lesions on “keratinized” sites, which, according to the results of our study, are more resistant to treatment.

Previous studies comparing imiquimod and podophyllotoxin in the treatment of anogenital wartsinclude a meta-analysis⁴ (that evaluated three placebo-controlled trials of imiquimod and nine placebo-controlled trials of podophyllotoxin) and a head-to-head comparison in a randomized, open-label trial that included 45 patients.⁵ Both studies concluded that imiquimod and podophyllotoxin were equally effective. In this study, we compared podophyllotoxin with a shorter, 8-week imiquimod course. For lesions on “keratinized” skin, the difference between imiquimod and podophyllotoxin both in clearance rates and in >50% reduction in wart area was statistically significant. For lesions on “partially keratinized” skin, no difference was detected between the two treatments.

In univariate analysis, lesions on “partially keratinized” sites and female patients were statistically significantly associated with a >50% reduction in wart area. In the multivariate analysis, however, gender of patients was no longer associated with response to treatment and only location of lesions on “partially keratinized” sites and treatment with podophyllotoxin were independent predicting factors of >50% reduction in wart area. These findings reflect the fact that patients with lesions on “keratinized” sites, that is, with “difficult-to-treat” lesions, were mostly men and had been treated more frequently with podophyllotoxin, compared with imiquimod. Men more often have lesions on “keratinized” sites, such as the penile shaft and the pubic area, while women usually have lesions on “partially keratinized” sites, such as the perineum, inner surface of labia majora, and labia minora of the vulva. In the multivariate model that took into account the location of lesions, gender was no longer a predictor of response to treatment and, on the contrary, type of treatment was. So, our study showed that location of lesions and not patients’ gender predicts response to treatment.

The main limitation of our study is its retrospective design. We mitigated this shortcoming by basing the main results of our study on an inverse probability of treatment weighted analysis, creating a sample in which treatment assignment was independent of (measured) baseline covariates.

Conclusion

Our study showed that lesions’ location plays a central role on response to patient-applied treatments for anogenital warts. For lesions on “partially keratinized” sites, an 8-week course of imiquimod was equally effective to the standard 4-week course of podophyllotoxin. For lesions on “keratinized” sites, the 8-week course of imiquimod was less effective, compared with podophyllotoxin, so either podophyllotoxin or the standard “up to 16 weeks” imiquimod regime should be preferred. Lesions on “partially keratinized” sites responded more favorably to both treatments.