Methods

A randomised, double-blind, placebo controlled study comparing cryotherapy and podophyllotoxin with cryotherapy and placebo (cryotherapy alone) was conducted at five sexual health clinics (see list below). Podophyllotoxin was administered as a 0.15% cream (Wartec/Warticon, Stiefel) or matching placebo.

All patients received cryotherapy, applied using a liquid nitrogen spray (Cry-Ac; Brymill Cryogenic Systems, UK) to obtain a minimum of a 45-s freeze (from start of application), repeated after a thaw in a standardised manner. Cryotherapy was repeated weekly for up to 12 weeks. Starting the day after the cryotherapy, podophyllotoxin cream or placebo was applied twice daily for three consecutive days, and repeated weekly for up to 4 weeks, or until all warts had cleared if earlier. If further treatment was required, only cryotherapy was used between week 4 and week 12, but after week 12 treatment was at the discretion of the clinician.

Patients were evaluated at baseline when the number, size (area judged against a calibration template), location and type (hyperkeratotic, non-keratotic, mixed; reference images were provided) of warts were documented. Follow-up assessments were performed at clinic visits at weeks 1, 2, 3, 4, 12 and 24, when wart clearance, relapse and new warts were recorded separately, as well as treatment adherence. If patients did not attend for follow-up assessments, they were contacted by letter and telephone and encouraged to re-attend. If this failed, self-assessed clearance was recorded, together with information on any treatment received outside of the trial. Self-assessment was documented on a proforma returned by the patient to the clinic, or by the investigator after a telephone interview.

Principal entry and exclusion criteria

Entry requirements were age 18–70 years, male and female patients with at least two (maximum 30) external anogenital warts with a combined area of at least 10 mm², which in the opinion of the investigator could be appropriately treated with cryotherapy with or without podophyllotoxin cream. Warts were either previously untreated, or not treated for at least 4 months.

Patients were excluded if known to be HIV positive (HIV testing was not a requirement), had concurrent internal anogenital warts, or had individual warts greater than 4 cm² in area.

Sample size calculation

The sample size for the study was based on an estimate of a 45% response rate in the cryotherapy-alone arm and a 70% response in the podophyllotoxin combination arm. Sixty-eight patients in each group would have provided 80% power to detect such a difference, or 85 patients after allowing for a 20% loss to follow-up (170 patients in total).

Randomisation and analysis

Patients were randomly assigned in equal numbers to each group, stratified by sex and history of anogenital warts, in blocks of four. Treatment packs were supplied prelabelled by the clinical trials unit at the manufacturer (Stiefel). Study numbers and corresponding treatment packs were then allocated in sequence by site investigators, against a local register. Blocks were reallocated between centres if required by the rate of recruitment. The primary endpoints were clearance of all anogenital warts at 4 and 12 weeks, including warts that had been present at baseline and any warts that may have appeared, or recurred, during follow-up. Secondary endpoints were clearance at 24 weeks, recurrence at 12 weeks in those cleared at 4 weeks and recurrence at 24 weeks in those clear at 12 weeks.

The statistical analysis plan was defined before data processing. Primary and secondary endpoint analyses were to be performed on an intention-to-treat (ITT) population comprising those patients who were randomly assigned, received at least one cryotherapy treatment and attended at least one follow-up assessment. In the event of an observation being missing (due to non-attendance) the analysis would use the last observation carried forward (LOCF) process to impute the status at subsequent time points. This assumes that patients who had cleared warts had not relapsed or developed new warts, and those that had not cleared did not do so over the period in question. To see the effect of using LOCF, clearance rates are also reported using only observed data at each time point.

The primary efficacy endpoints are summarised by treatment group and visit, with differences between the groups investigated by linear logistic regression analysis, adjusted for randomisation group, centre and treatment group. An exploratory analysis of the primary outcome measure was planned, using a linear logistic regression model adjusting for randomisation group, centre, baseline wart area and history of warts, and any covariates significant at a 10% level in the univariate analysis. Further subgroup analyses were permitted but, given the small numbers, these would be exploratory and interaction effects would not be investigated.

Consent and approval

All patients gave written informed consent and the study was reviewed by the London Multicentre Research Ethics Committee. The study was monitored by an independent data and safety monitoring committee (DSMC), including a planned review of interim results with respect to efficacy and statistical power.

Results

Study recruitment was stopped at 149 patients, of whom 140 (70 in each group) started treatment and attended for at least one follow-up visit (ITT population). In the patients randomly assigned to podophyllotoxin combination or cryotherapy alone, respectively, 67.1% and 62.9% were men, and 30.0% and 25.7% had a previous history of anogenital warts (table 1). Patients in the combination arm had a higher proportion of keratotic warts and the estimated area of warts was larger.

With the addition of self-reports, information was available on 119 (85.0%) at week 12 and 121 (86.4%) at week 24. Use of study cream declined over the 4 weeks of the blinded phase of the treatment protocol, to 31.1% in the podophyllotoxin combination and 54.8% in the cryotherapy alone (placebo cream) groups; application of cryotherapy or cream was recorded in 42.9% and 61.4%, respectively in the fourth week of follow-up. Reasons for withdrawal and the number followed up are detailed in fig 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Study flow diagram. *The safety population included all patients randomly assigned who received at least one cryotherapy treatment and one application of blinded medication. The intention-to-treat (ITT) population included all patients who were randomly assigned, had at least one treatment, and had at least one follow-up assessment. The per-protocol population comprised all those in the ITT population who did not have a major protocol violation, principally non-attendance for follow-up. ITT efficacy analyses used last observation carried forward as described in the Methods section. **Numbers with observed data increased when patients missed an earlier assessment but re-attended for later time points. ***Patients may have had more than one reason for withdrawal.

Clearance rates at weeks 4 and 12, the primary endpoints, were 60.0% at both times in the podophyllotoxin combination group, and 45.7% at both times in the cryotherapy-alone group, but these differences were not statistically significant (relative risk (RR) 1.31, 95% CI 0.95 to 1.81). The clearance rate at week 24 was very similar in both groups (68.6% and 64.3%, respectively; RR 1.07, 95% CI 0.84 to 1.35). When adjusted for baseline number and area of warts, history and treatment centre differences between the groups remained non-significant (table 2).

Loss to follow-up was similar in both groups, but comparison of the LOCF and observed data indicates that the clearance rate was lower in those subsequently lost to follow-up, particularly in the cryotherapy-alone arm (table 2).

In a multivariate analysis of factors associated with wart clearance, there were associations with history of warts and sex (table 3). At week 4, patients with a previous history of warts were more likely to clear (p = 0.009) than patients for whom this was their first diagnosed episode, although this effect was not seen at week 12. Consequently, an exploratory analysis of clearance rates was performed, split by wart history. At week 4, patients with a previous history of warts were more likely to have cleared with the podophyllotoxin combination (unadjusted odds ratio (OR) 4.80, 95% CI 1.03 to 22.3; p = 0.045); however, at week 12, the effect was no longer significant (2.03; CI 0.56, 7.31; p = 0.28). In the group with a first episode of warts, there was no difference in clearance at week 4 or week 12 by treatment allocation (OR 1.31, 95% CI 0.60 to 2.87; p = 0.50 at week 4, and OR 1.69, 95% CI 0.77 to 3.72; p = 0.19 at week 12, with podophyllotoxin combination vs cryotherapy alone, respectively).

At week 4, men were more likely to show clearance than women (p = 0.001). However, in a further analysis of data split by sex, the unadjusted OR for podophyllotoxin combination therapy versus cryotherapy alone were similar in men and women and were not significant (OR 1.75, 95% CI 0.52 to 5.82; p = 0.37 for women and OR 1.79, 95% CI 0.76 to 4.25; p = 0.19 for men). At week 12, women were more likely to have cleared with podophyllotoxin combination, but this was of borderline statistical significance (OR 3.12, 95% CI 0.96 to 10.2; p = 0.059). There was no effect in men (OR 1.36, 95% CI 0.59 to 3.09; p = 0.47).

The effect of podophyllotoxin combination treatment was also investigated in an exploratory analysis according to wart type. Ablative therapies have been recommended in preference to topical agents for more highly keratinised warts,10 but combination therapy might be still more effective. When analysed separately, there was no difference between the effect of podophyllotoxin combination therapy in patients classified as having either hyperkeratotic or non-keratotic warts at week 4 (OR 1.5, 95% CI 0.38 to 6.00 and OR 1.38, 95% CI 0.51 to 3.76, respectively), or at week 12 (OR 1.29, 95% CI 0.32 to 5.19 and OR 1.24, 95% CI 0.45 to 3.38).

More patients receiving the podophyllotoxin combination had at least a 50% reduction in wart area by weeks 4 and 12, but this difference was not statistically significant (data not shown). There was some evidence for patients in the cryotherapy-alone arm experiencing a greater number and larger area of new and/or relapsed warts. Between week 4 and week 12, more patients who had cleared warts at week 4 in the cryotherapy-alone arm had new or recurrent warts noted at week 12, compared with the podophyllotoxin combination arm, although this was not statistically significant (11/32, 34.4% vs 12/42, 26.6%; p = 0.67). After week 12, further treatment in both groups was at the discretion of the physician (who remained blinded to the earlier treatment allocation). Recurrence rates between week 12 and week 24 were similar (6/32, 18.8% and 7/42, 16.7%, respectively).

There were no treatment-related serious adverse events. The proportion of patients experiencing any adverse event was similar in both groups (72.9% podophyllotoxin combination, 70.0% cryotherapy alone; p = 0.85). Application site events occurred more frequently in the combination group (64.3% vs 44.3%; p = 0.027), with pain being the most frequent event in both groups (34.3% vs 18.6%; p = 0.055). No patient withdrew from the study because of treatment-related local events, but more patients in the podophyllotoxin combination arm (n  =  13, 18.6%) stopped blinded study medication than in the cryotherapy-alone arm (n  =  4, 5.7%; p = 0.036).

Discussion

Cryotherapy and podophyllotoxin are used in combination in current practice, but this has not been evaluated before in a randomised controlled trial. Overall, there was no statistically significant difference between the treatments with respect to the primary endpoints of clearance of warts at 4 and 12 weeks. There was some evidence for an improved outcome with podophyllotoxin combination treatment, with approximately a third increase in the chance of clearing warts by 4 and 12 weeks (60.0% vs 45.7%; RR 1.31). Clearance at the end of 24 weeks, during which further treatment could be given at the discretion of the clinician, was similar in both groups (68.6% podophyllotoxin combination; 64.3% cryotherapy alone). Clearance was defined as resolution of both warts present at baseline and any new or relapsed warts observed during follow-up. These results are consistent with published reports on podophyllotoxin or cryotherapy alone.11 12 The incidence of new or relapsed warts appearing after the end of the blinded treatment phase favoured the podophyllotoxin combination arm but the difference was not statistically significant.

One study has investigated combination therapy with cryotherapy and podophyllin, rather than podophyllotoxin, in an open comparative trial in which treatment allocation for patients starting treatment was changed at random at weekly intervals.9 The study included 409 patients; 76 received cryotherapy and podophyllin and 85 received trichloracetic acid and podophyllin (the remainder received monotherapy). Either combination therapy (podophyllin 77.6%; trichloracetic acid 74.1%) was more effective than podophyllin alone (58.2%), but was similar to cryotherapy alone (75.3%). This study did not include the use of podophyllotoxin, which is more effective that podophyllin.7

The population recruited in this study was mixed, with men and women and first episode as well as repeat cases and few entry restrictions, to reflect as nearly as possible routine clinical practice. Randomisation was stratified by sex and history of warts, both factors shown to be related to treatment response. Men appeared to be responding better at 4 weeks, but there was no difference at 12 or 24 weeks. Similarly, those with a history of warts responded better. This may reflect a difference in immunological status in those with previous HPV-related disease. Although the groups were well matched on the stratification criteria, there were more patients with keratotic warts in the group given podophyllotoxin combination therapy. Current treatment guidelines recommend the use of ablative therapies for keratinised warts;10 in this study, there was no evidence of a difference in the treatment response according to the type of wart, although the number available for this subgroup analysis was small.

Podophyllotoxin combination therapy was well tolerated. There were more reports of application site events in the combination treatment group, but this is to be expected and is consistent with an additive effect of using podophyllotoxin. More patients discontinued use of the cream in the active treatment group (18.6% vs 5.7%).

The study achieved high follow-up rates, including the use of patient-reported assessments for those study subjects who missed follow-up evaluations in clinic. We used an analysis based on LOCF to account for missing assessment data. This may overestimate the effect of treatment if clearance is followed by early relapse and this is not captured due to missed assessments. More patients failed to attend for follow-up in the podophyllotoxin combination arm, but this was largely compensated for by additional self-reports. Follow-up rates are critical to the reliability of evaluations of wart treatments.7 Self-assessment may underestimate the presence of persistent or recurrent warts.13 Nonetheless, the simple endpoint of complete clearance may be less likely to lead to misclassification. Future studies based on a similar design but with fewer follow-up assessments may allow even higher follow-up rates, thus allowing wart treatments to be evaluated more easily and reliably in larger populations, providing greater statistical power.

Although patients receiving podophyllotoxin combination therapy had almost one third greater chance of clearance this difference did not reach statistical significance. At a planned interim review by the DSMC it was noted that the response rate in the comparator arm at 4 weeks was lower than predicted in the study design. This led to a recommendation that the number of patients recruited should be increased, but neither time nor resources were available to allow this. A substantially larger study would have been required to confirm whether combination therapy was indeed superior, or to exclude an important treatment effect. There remains the possibility that podophyllotoxin combination therapy may benefit at least some patients; further work is required to refine optimum treatment regimens.