Microwave therapy for the treatment of plantar warts

Single-centre non-randomised experimental retrospective study.

A retrospective cohort study was undertaken using medical records from patients who received microwave therapy between July 2018 and March 2022 at a single podiatry clinic in Milford, Auckland (New Zealand). At present, only a handful of microwave therapy units are being used throughout New Zealand, with the Milford clinic being the first to offer it to patients in July 2018. All patients who received microwave therapy from this date were invited to share their clinical treatment records for the purposes of this study. An a-priori sample size calculation was not undertaken, and sample size was informed by the number of patients who had been offered microwave therapy in this timeframe and consented to the study. Patients were included if they had one or more plantar wart(s) diagnosed by an experienced podiatrist and underwent microwave therapy treatment. Ethical approval was obtained from the Auckland University of Technology Ethics Committee on 27^th July 2021 (AUTEC 21/250). Data were only included from patients who provided written informed consent/assent for their clinical treatment records to be used for this study.

Patients were excluded from receiving microwave therapy at the podiatry clinic if they were pregnant or breast-feeding, had metal implants in their feet or ankles, or were < 5 years of age. During the initial appointment a podiatrist recorded information about the location(s) and duration(s) of the plantar wart(s), and details about previous treatment(s).

Microwave therapy was delivered using the Swift® (Emblation Ltd., UK) device, which is a hand-held tool that produces microwave heat energy within the 8 GHz range. Podiatrists at the clinic completed the manufactures standard training in using microwave therapy. The microwave energy was delivered directly to the infected tissue through a single use applicator tip to reduce the risk of cross-infection (Fig. 1). A standardised treatment procedure was followed for all patients. The first microwave therapy application at the first treatment session delivered 10 J using one application of 5W for 2 s. The wattage, frequency, and number of subsequent applications is determined by the level of pain tolerated by each patient. A 10-point scale was used at each appointment to collect patient-reported pain on arrival, during treatment, and on departure. Patients received a standard of three treatment sessions four weeks apart unless the lesion(s) resolved prior. A final review was undertaken at 12 weeks after the final treatment session where response to treatment was assessed by a podiatrist and recorded as either “resolved” or “unresolved”. Resolved lesions were defined as those no longer visible and where natural dermatoglyphics were restored. Before and after photographic evidence was used to support assessment. If unresolved at the 12-week review, patients were given the option continue further treatment sessions at 4 weeks apart until resolved, or to discontinue treatment.

Data from clinical treatment records of all included patients were extracted into a standardised extraction form, including patient demographics, relevant medical history, and microwave therapy and treatment parameters.

The distribution of all continuous outcomes (lesion duration, total treatment sessions, average applications per session, average watts per application, average frequency setting per application, average pain, and average joules per treatment session) for both resolved and unresolved lesions were reviewed for normality using both visual inspection of histograms and formal tests of normality for skewness and kurtosis. Following confirmation of approximate normal distributions, clinical and treatment characteristics for participants were described as mean (SD) for continuous data and n (%) for categorical data. To determine clinical characteristics (age, gender, immunosuppression, impaired healing, multiple vs single wart, location of lesion, lesion diameter) that were associated with plantar wart resolution, a generalised linear model with binomial distribution and a logit link function was used. This approach was chosen as it also allowed for the inclusion of a participant-specific random effect which accounted for repeated measures taken from participants who had multiple lesions [20]. Person-specific factors (including age, sex, immune status, etc.,) result in a high-level of within-person dependence between lesions, meaning that data from different lesions within the same person will be highly correlated. The Nakagawa conditional R² was computed which incorporates both fixed and random effects [21]. Adjusted estimates and p-values were reported. No multicollinearity was present between the clinical characteristics. All statistical analyses were performed in R (v4.1) using the lme4 package.

A total of 135 eligible patients who received microwave therapy between July 2018 and March 2022 were invited to share their clinical treatment records for this research. A total of 47 (35%) patients responded to the invitation, including 2 who declined to participate and 45 who consented to participate. The characteristics for the 45 included patients are shown in Table 1. The majority of participants were female (n = 25, 56%) with a mean age of 46.6 years (range 6 to 74 years). Most participants had multiple plantar warts and had tried previous treatments which were unsuccessful including salicylic acid, cryotherapy, and cryotherapy.

A total of 150 plantar warts were treated with microwave therapy across the 45 included patients (mean (SD) 3.3 (2.6) plantar warts per patient). Lesion-level characteristics and microwave therapy treatment details are summarised in Table 2. The majority of lesions were located on the plantar heel (n = 38, 25%) followed by the first metatarsal head, fifth metatarsal head, and hallux. The average lesion duration prior to treatment with microwave therapy was 35.2 months. Unresolved lesions received more treatment sessions (P = 0.043), a greater average number of microwave therapy applications per treatment session (P = 0.023), and higher average joules per treatment session (P = 0.034) compared to those with resolved lesions. However, average pain during treatment was significantly higher (P = 0.009).

In 32 patients full lesion resolution was achieved and in 10 patients lesion resolution was not achieved. Three patients with multiple plantar warts had both resolved and unresolved lesions. Of the total 150 total plantar warts treated with microwave therapy, 125 (83.3%) warts resolved and 25 (17%) warts did not resolve. Figure 2 presents examples of plantar lesions pre- and post-treatment. In the generalised linear model only decreasing age was a significant predictor of plantar wart resolution (odds ratio (95% confidence interval), 0.84 (0.71, 1.00), P = 0.046). The model explained 99.7% of the variance in plantar wart resolution (conditional R²).