Intralesional immunotherapy for non-genital warts: A systematic review and meta-analysis

Protocol and registration

We conducted a single-arm meta-analysis of prospective studies on the treatment response of intralesional immunotherapy for treating non-genital warts. We reported this study following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines⁶ and this study was registered with PROSPERO, an international database of prospectively registered systematic reviews (https://www.crd.york.ac.uk/PROSPERO/, CRD42020163379).

Databases

We performed a comprehensive search using predefined search terms in MEDLINE, EMBASE, Web of Science, and Cochrane Library databases from the respective dates of database inception to January 4, 2020. The main keywords used were “intralesional,” “immunotherapy,” “purified protein derivative,” “Trichophyton,” “BCG,” “bacillus Calmette-Guerin,” “MMR,” “measles-mumps-rubella,” “candida,” “Mycobacterium,” “vitamin D,” “Corynebacterium,” “INF,” “interferon,” “Propionium,” “Propionibacterium,” “tuberculin,” “vaccine,” “vaccination,” “wart” and “verrucae.” All prospective and experimental studies were included and the reference lists of identified relevant review articles were scanned manually as well. All articles identified using this search process were screened independently by two reviewers (H. J. J. and H. R. P.). In cases of discrepancy between the two main reviewers, a final decision was reached by consensus with two other reviewers (J.M.B. and J.H.L).

Study selection

All relevant clinical studies that reported the treatment response of intralesional immunotherapy by a single injection of the immunotherapeutic agent into the largest lesion in each treatment session were examined. The inclusion criteria were¹ prospective study design (randomized or non-randomized controlled trials and open trials);² participants of all age groups with a diagnosis of a non-genital wart;³ at least one intralesional immunotherapy or control (saline or sterile water injection) group;⁴ at least 10 participants in each treatment arm, regardless of dropout rate; and,⁵ outcomes measured based on complete response or no response (0 or <25%). Conversely, the exclusion criteria were¹ retrospective or observational study design;² different outcome measures;³ other interventions or combined; and⁴ unavailability of the corresponding authors. We excluded studies involving genital warts, as the causative virus and conventional therapeutic modalities are different from those of non-genital warts.

Data extraction, quality assessment and outcome measures

For the meta-analysis, two reviewers extracted the following predefined variables: authors, country, year of publication, study type, the immunotherapeutic agent used, numbers of treated patients, treatment protocols and outcome. Quality assessment of the analytic studies was performed.

The primary outcome was the treatment response rate of intralesional immunotherapy in patients with non-genital warts. Complete response was defined as the complete disappearance of all lesions including both the injected and nearby satellite lesions. The treatment response rate was calculated as the number of participants who achieved complete response divided by the total number of participants who completed the individual study. The no response rate was assessed by dividing the total number of patients who had no or minimal treatment response (<25%) by the total number of participants who had completed the individual study. Because the immune reaction to the intralesional injection can be effective not only for adjacent warts but also for anatomically distant warts, a secondary outcome, defined as distant complete response, was the clearance of distant warts located in an anatomically different body part, away from the injection site. In addition, the recurrence rate was analysed by compiling the studies that reported recurrence. Studies with fewer than 10 patients showing complete response after treatment were excluded.

Meta-regression for age and sex

Meta-regression analyses were conducted to determine whether the estimated treatment response rates varied according to age or sex of patients and were performed by setting the mean age and female to male ratio of participants in each study as moderator variables, respectively.

Safety profiles

We noted all reported adverse events. We recorded common events and their frequencies for each study and searched for serious adverse events.

Statistical analyses

The rates of corresponding treatment responses of the included studies were pooled by generic inverse variance weighting and were combined using a random-effects model.⁷ Heterogeneity was assessed using the I² value and subset analyses.⁸ We used a funnel plot of sample size against log odds to determine publication bias because conventional funnel plots can be asymmetric in the absence of publication bias, especially in studies for extreme proportional metrics.⁹ Meta-analyses were conducted using R (version 3.6.1; R Foundation for Statistical Computing, Vienna, Austria) software with “metagen” and “metafor” packages.

Search results

We initially identified a total of 1,348 records through database searching and six additional records from the reference lists of related articles; 727 duplicates were removed and 494 were deleted after reviewing the titles and abstracts [Figure 1]. A total of 133 full-text articles was assessed in terms of eligibility; of these, 79 were excluded after full-text evaluation. The remaining 54 studies fulfilled the inclusion criteria and were included in the final analysis [Table 1].

Close

Figure 1:

Flow diagram showing how eligible studies were identified in the present review

Export to PPT

Characteristics of included studies

A total of 54 studies of 3,446 enrolled patients was deemed finally eligible [Table 2]. Overall, 13 studies with 747 patients in the MMR group^10-22; 12 studies with 436 patients in the Candida group^23-34; nine studies with 311 patients in the Vitamin D group^25,35-42; six studies with 235 patients in the PPD group^38,43-47; six studies with 197 patients in the interferon-α group^48-53; four studies with 127 patients in the Mw group^54-57; two studies with 49 patients in the BCG group^13,58; and one study each of other therapeutic agents of smallpox vaccine,⁵⁹ mumps antigen,³⁴ mumps or Candida antigen,⁶⁰ mumps or Candida or Trichophyton antigen,⁵² HPV,⁶¹ Propionium,⁶² and zinc³⁷ were included. As a control group, 13 studies with 470 patients who received saline or sterile water injection were included.^{10,15,16,20,22,25,40,43,49,52,53,59,62} The mean treatment duration was 8.2 ± 6.0 weeks, and the mean follow-up duration was 4.6 ± 3.5 months (range, 1–25 months).

Treatment response of injected and nearby lesions

The overall complete response rate was 60.6% (95% confidence interval: 54.8–66.5%) among 53 studies with 2,548 patients [Figure 2a]. The complete response ratio of the Mw group was 73.1% (95% confidence interval: 55.6–90.6%) in four studies,^54-57,63 the Candida group was 62.6% (95% confidence interval: 53.3–71.9%) in 11 studies,^23,24,26-34 MMR group was 63.2% (95% confidence interval: 53.1–73.4%) in 13 studies,^10-22 PPD group was 62.7% (95% confidence interval: 42.4–83.0%) in six studies,^38,43-47 Vitamin D group was 54.2% (95% confidence interval: 32.7–76.0%) in eight studies,^35-42 interferon-α group was 51.9% (95% confidence interval: 27.0–76.9%) in six studies,^48-53 and BCG group was 50.9% (95% confidence interval: 5.0–96.8%) in two studies.^13,58 There was no significant difference depending on the immunotherapeutic agent used (p-value, 0.88). In the control group receiving saline/sterile water injection, the complete response rate was 17.3% (95% confidence interval: 10.0–24.5%) in 12 studies.^{10,15,16,20,22,40,43,49,52,53,59,62}

Close

Figure 2a:

Treatment response of intralesional immunotherapy. The complete response rate of intralesional immunotherapy on injected and adjacent warts

Export to PPT

Overall no response rate of intralesional immunotherapy was 16.6% (95% confidence interval: 12.4–20.8%) in 38 studies including a total of 1,719 patients. The no response rate was highest in the IFN-α group (48.1% [95% confidence interval: -2.2–98.4%] in two studies^49,52), followed by in the MMR group (20.9% [95% confidence interval: 7.5–34.2%] in six studies^{12,14,17,18,21,22}) and Candida group (14.6% [95% confidence interval: 8.8–20.3%] in eight studies^{24,25,28,29,31-34}), whereas the no response rate of the control group was 79.7% (95% confidence interval: 66.2–93.2%) in seven studies.^{22,25,40,43,49,52,62}

Treatment response of distant lesions

Among the studies included in this analysis, 15 described the treatment response of distant lesions located away from the mother wart. The overall distant complete response rate of the intralesional immunotherapy was 51.6% (95% confidence interval: 37.9–65.3%) [Figure 2b]. The distant complete response rate of the MMR group was 62.0% (95% confidence interval: 31.1–93.0%) and that of the Candida group was 42.0% (95% confidence interval: 7.3–76.7%). The distant complete response rate of Vitamin D (17.6%, [95% confidence interval: −0.5–35.8%]) and IFN-α (8.8%, [95% confidence interval: −0.7–18.4%]) showed no significant difference from that of the control group with saline injection (14.2% [95% confidence interval: −1.9–30.2%]).

Close

Figure 2b:

Treatment response of intralesional immunotherapy. The complete response of intralesional immunotherapy on distant warts located in other body parts

Export to PPT

Meta-regression for age and sex

There were no significant linear interactions between mean age and sex (female to male ratio) with changes in treatment response, and the coefficients for the variables were not statistically significant (p-value, 0.61 for age; 0.43 for sex).

Rate of recurrence among patients who have achieved complete response

A total of 47 studies reported recurrence after treatment, and the median follow-up period was six months (range, 0–12 months). The recurrence rates among studies were reported from 0 to 16.7%. The pooled recurrence rate was 2.0% (95% confidence interval: 1.1–2.9%).

Safety of intralesional immunotherapy

Of the 54 clinical studies included in this analysis, 51 reported occurrence of adverse events, and 41 presented the specific frequency of adverse events. The most common adverse event was pain, reported in 35 of 51 studies regardless of the immunotherapeutic agent used, with frequency varying from 2 to 100%. Flu-like symptoms were reported in 22 of the 51 studies, ranging in frequency from 2.5 to 100%. Other adverse events including erythema, oedema, and pruritus have been reported frequently, and lymphadenopathy, vasovagal syncope, dyspigmentation, eczematous reaction and ulceration have been noted as rare adverse events.