Topical photodynamic therapy for genital warts: Systematic review and meta‐analysis

High recurrence rate and adverse effects were common shortcomings in treatments for genital warts, but a number of clinical studies have found that photodynamic therapy (PDT) had some unique superiority compared to other treatments. This study aims to evaluate the efficacy and safety of PDT for genital warts in comparison to alternative treatments. Seven databases and ClinicalTrials.gov were searched and seven randomized controlled trials were finally included. There was no significant difference in clearance rate of urethral lesions between the PDT group and the CO2 laser therapy (CO2‐LT group) (Odds ratio [OR] 0.676, 95% confidence interval [CI] 0.071–6.426, P = 0.734), but the overall clearance rate of all genital warts was slightly lower for PDT (OR 0.574, CI 0.335–0.985, P = 0.044). The recurrence rate after PDT was significantly lower than after CO2‐LT (OR 0.318, CI 0.220–0.460, P < 0.001), and adverse effects were also slighter after PDT (OR 0.015, CI 0.003–0.074, P < 0.001). In addition, a modification of parameters of light irradiation provides the potential to alleviate pain. This study shows that PDT is an effective treatment for genital warts with good tolerability and safety, especially for cavitary lesions. Some modified protocols probably have the potential to reduce adverse effects, but still need further investigation.

Genital warts, which is also referred to as condylomata acuminata, is one of the most common warty proliferating growth with high contagiosity caused by the infection of human papillomaviruses (HPV), and some of the lesions induced by oncogenic HPVs can develop into intraepithelial neoplasia and cancer [1]. Most of genital warts lesions develop in the anogenital region, and the overall annual incidence of anogenital warts was around 1.945‰ [2], leading to a heavy burden to health service systems and negative psychological effects in patients [3]. Current treatments for genital warts include topical applied medicine and ablative treatments, such as podophyllotoxin, imiquimod, CO 2 laser therapy (CO 2 -LT) and cryotherapy [4]. However, treatment for meatal lesions is particularly difficult because of the low accessibility. Besides, due to the intractable latent HPV infection, a high recurrence rate, which is generally up to 20% to 30%, becomes non-negligible limitation of these treatments [4]. Furthermore, the adverse effects of ablative treatments can be severe, such as infections, scars, and stenoses of anatomical luminal structures.
Photodynamic therapy (PDT) is a minimally invasive therapeutic procedure that shows of potential applications in skin cancers and several other cutaneous diseases [5,6]. The PDT principle can be described as follows: The photosensitizers are preferentially taken up by the highly proliferating epithelia cells, then the photosensitizers and their intracellular derivates generate a large amount of reactive oxygen species upon stimulation with suitable light sources. Subsequently, the abnormal cells are eliminated [7]. PDT was first applied for the clinical treatment of skin cancer in 1970s [7,8]. Noticing the similar proliferation characteristics of genital warts and skin tumors, Frank et al reported the efficacy of PDT conducted on seven anogenital condylomata acuminata patients in 1995 [9]. Later in 2000, Wang et al reported the novel application of 5-aminolevulinic acid (ALA)-PDT for 18 cases of urethral condylomata acuminata and found that PDT was well-tolerated and could remove the urethral condylomata acuminata lesions efficiently with low recurrence [10]. In 2004, Wang et al reported the first pilot study on the use of ALA-PDT for urethral condylomata acuminata. In addition, this study investigated the histological and cellular changes induced by ALA-PDT and partially revealed the mechanism of the treatment [11]. ALA-PDT was approved for the treatment of condylomata acuminata in China by the National Medical Products Administration in 2007 with a level of recommendation A and grade of evidence I [12]. Hereafter, PDT was applied in millions of patients with genital warts at more than 3000 hospitals in China. Besides, ALA-PDT has already become the first-line treatment for urethral condylomata acuminata in China.
During the past two decades, a lot of studies have been conducted to evaluate the efficacy and safety of PDT for genital warts by our teams and other experienced scholars and colleagues, including some phase III-IV clinical trials. Based on these evidences, PDT has been routinely used for the treatment of urethral condylomata acuminata and other genital warts in China [13]. This systematic review and meta-analysis aims to summarize these studies and determine the efficacy and safety of PDT for genital warts.

| MATERIALS AND METHODS
This systematic review was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, and registered with the PROSPERO international prospective register of systematic review entitled "Topical photodynamic therapy for condylomata acuminata: systematic review and meta-analysis" (CRD42022322200).

| Eligibility criteria
Only randomized controlled trials (RCTs) evaluating the efficacy and safety of topical PDT in patients with genital warts providing information about one or more of the following outcomes were included: clearance of genital warts within 1 week, recurrence rates 3 months after treatment or later, pain and other adverse effects or treatment intolerability.

| Data sources and search strategy
A systematic search of PubMed, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), Web of Science, SinoMed, China National Knowledge Infrastructure (CNKI), Wanfang Data, and the ClinicalTrials.gov database was conducted from the earliest dates until February 28, 2022. The search string was ("photochemotherapy" or "photodynamic therapy") and ("condylomata acuminata" or "veneral warts" or "genital warts") and ("randomized controlled trial" or "controlled clinical trial"), no language limitation was set. Then duplicated articles were removed. Titles and abstracts of the remaining studies were subsequently screened, and the full-text articles were further reviewed to assess the eligibility for inclusion ( Figure 1). The risk for bias in the selected RCT was determined using the Cochrane Collaboration Risk of Bias tool [14].
The above processes were conducted by two independent investigators (Zhi Cao and Peiru Wang), the third investigator (Haiyan Zhang) would be consulted if there was disagreement between the two former investigators.

| Statistical analysis
Extracted outcomes were combined and a meta-analysis was performed using the Comprehensive Meta-Analysis (CMA) software version 3. Inconsistency and heterogeneity between studies was assessed using the I 2 test, where P < 0.05 was considered statistically significant. A random-effects model was used when I 2 > 50%, a fixedeffects model was used when I 2 < 50%. Odds ratio (OR) and 95% confidence interval (CI) were used as effect measures based on each of the different outcome parameters. Risk for bias of RCTs was assessed with Review Manager v5.4 (http://ims.cochrane.org/revman). To evaluate the publication bias, Egger's regression test and funnel plot was conducted using the CMA software.

| Study selection
A total of 2440 results were obtained from 7 databases and ClinicalTrials.gov. After removing the duplicated articles, 1415 articles were kept. Then 55 studies were selected out after screening of titles and abstracts. Seven RCTs were finally screened out for further analysis. All the RCTs had been conducted in China, most of the subjects included in the studies were middle-aged to elderly patients. Treatment protocols and the outcome evaluation are given in Table 1.  2.2 | Photodynamic therapy vs carbondioxide laser treatment for urethral condylomata acuminata Four RCTs had compared the efficacy and the safety and/or tolerability of PDT with those of CO 2 -LT [15][16][17][18]. These studies applied a similar treatment protocol, in which 20% wt/vol (in one case [16] 5% wt/vol) ALA solution (Fudan Zhangjiang Bio-Pharm Co. Ltd, Shanghai, China) was incubated for 3 to 4 h, covering the lesions and a 2 to 5 mm rim of surrounding urothelium that appeared to be normal, then the ALA was removed, and the lesions were irradiated with light at a wavelength between 590 and 690 nm. The treatments were conducted once a week until the lesions were eliminated or the maximum number of sessions [3,4] was reached (Table 1). A specific analysis for urethral condylomata acuminata patients was only possible in terms of clearance rate, because the recurrence rate and adverse effects were not reported separately for urethral condylomata acuminata and other genital warts in the four RCTs. Since the CO 2 -LT would remove all visible lesions, the clearance rate of warts in this group was 100% in all RCTs. For the PDT group, the clearance rates of urethral lesions were about 89.5% to 100.0%. There was no statistically significant difference in the clearance rate between the two groups (OR 0.676, CI 0.071-6.426, P = 0.734, Figure 2A).

| Photodynamic therapy vs carbondioxide laser treatment for all genital warts
Similar to urethral condylomata acuminata, no significant difference was found in the clearance rate between PDT and CO 2 -LT in other location of genital warts (OR 0.585, CI 0.341-1.004, P = 0.052, Figure 2B). But when combining data from all the locations, the overall clearance rate of CO 2 -LT was slightly superior compared to PDT (OR 0.574, CI 0.335-0.985, P = 0.044, Figure 2C), although the latter was also high, varying from 95.0% to 98.4%.
However, the overall recurrence rates were significantly different for the two treatments. The recurrence rate after PDT was 2.4% to 13.8%, and it was 17.4% to 33.3% after CO 2 -LT. The PDT group showed a significantly lower recurrence rate compared with the CO 2 -LT group (OR 0.318, CI 0.220-0.460, P < 0.001, Figure 2D).
To assess the safety and tolerability of the two treatments, the included four RCTs recorded the adverse effects, such as erythema, edema, infection, pain, bleeding, ulcer, scars, and hyperpigmentation. It was found that PDT was superior to CO 2 -LT in safety and tolerability (OR 0.015, CI 0.003-0.074, P < 0.001, Figure 2E).

| Modified photodynamic therapy vs conventional photodynamic therapy for genital warts
To reduce pain and other adverse effects without decreasing efficacy, some studies have explored modified protocols of PDT in the treatment of all genital warts, and three RCTs of this topic were screened out for our study [19][20][21]. These RCTs decreased the power density of the light resource and prolonged the irradiating duration to maintain the energy density to applied by conventional PDT. It was shown that there was no significant difference in lesion clearance rate between modified PDT and conventional PDT (OR 1.392, CI 0.900-2.153, P = 0.137, Figure 2F).
Since one of the RCTs analyzed the severity of pain in a different method (visual analog scale [20]) from the other two (numeric rating scale [19,21]), and the inconsistency as well as heterogeneity of the remaining two RCTs were also significant (I 2 = 97.391, P < 0.001), we only present the outcome of severity of pain in Table 1. Outcomes from these RCTs supported that the modified PDT was superior to conventional PDT in decreasing pain.
Another well-conducted RCT had also evaluated the efficacy and severity of pain of modified PDT using a similar protocol [22], and it was shown that the severity of pain was the same as the control group who had taken tramadol before conventional PD. However, the original data of this study were unavailable, therefore it was not included in this study.

| Photodynamic therapy vs other treatments
Though there have been a considerable number of studies evaluating the efficacy and safety of PDT compared to other treatments, a few were conducted as RCT. We screened out one RCT, which compared the efficacy and safety between PDT and topical trichloracetic acid (TAA). The latter group accepted light irradiation after the application of TAA, and the parameters of irradiation were the same as in the PDT group. It was shown that there was no difference in complete response rate between the PDT and the TAA group, however, the recurrence rate of TAA was significantly higher than that of PDT (33% vs 0%).

| Heterogeneity and bias of included studies
Heterogeneity was negligible (I 2 = 0.000%) in all groups except for adverse effect assessment in PDT (I 2 = 72.476%) vs CO 2 -LT. Considering that the interventions and the population vary across different studies, a random effects model was used in the analyses of adverse effects of PDT vs CO 2 -LT, while the analyses of other outcomes were conducted with a fixed effects model. As shown in Figure 3, the risk for bias among the included RCTs was evaluated and summarized. Considering that the PDT treatment modality is quite different from CO 2 -LT, and the irradiating protocol was totally different between modified PDT and conventional PDT, it was unpractical to establish a blinding of participants and personnel. Therefore, though the selection bias was ranked to be "high risk of bias" in all seven RCTs, it would not disturb the analyses of this study. All of these RCTs did not mention a blinding processing during the evaluation of outcomes, so the detection bias was ranked to be "unclear risk of bias" in all seven RCTs. Besides, in the three RCTs in which the severity of pain was analyzed after PDT and modified PDT, the subjective evaluation of pain might bring bias into the study, so the item "Other bias" was ranked to be "high risk of bias." To evaluate the publication bias of this study, we conducted Egger's regression test (P = 0.094 > 0.05). Considering that the number of RCTs involved in this study was F I G U R E 3 Risk for bias assessment of randomized controlled trials (RCTs) included in this study. (A) Risk of bias graph of the seven RCTs included in this study. (B) Risk of bias summary of the seven RCTs included in this study. limited, the results might be inaccurate. A funnel plot including the seven RCTs was also made ( Figure 4) and presented an asymmetric distribution of the dots. The results suggest the possibility of publication bias, indicating that some negative studies might not have been published.

| DISCUSSION
This systematic review with meta-analysis suggested that the efficacy of PDT was slightly lower than that of CO 2 -LT, but the recurrence rate and adverse effects of PDT were significantly lower. Moreover, modified PDT protocols have the potential to reduce pain without decreasing efficacy, but more RCTs with high quality are required to verify this conclusion.
While high recurrence rate is one of the most common and intractable drawbacks of traditional treatments for genital warts, PDT has a significantly lower recurrence rate compared with CO 2 -LT, topical TAA, and other treatments. It has been reported that a photosensitizer could be taken up selectively by HPV-infected cells and induce apoptosis during PDT treatment [23]. Based on the fluorescence examination, it was recommended to apply the photosensitizer to not only the wart lesions but also to a surrounding 2 cm rim in clinical practice [12,24]. Both the photosensitizer and the irradiation of light could be applied in a large area, and consequently eliminate HPV infection in a field-clearing pattern. With these advantages, PDT can clear the latent HPV infection effectively and reduce the recurrence rate significantly.
Besides, immune reactions induced by PDT also play a crucial role in the clearance of genital warts, subclinical lesions, and latent HPV-infections. A large number of immune cells could be recruited and activated by PDT, including the dendritic cells, neutrophils, macrophages, T cells, and B cells [11,25]. The immune reaction would further promote the clearance of HPV-infected cells and decrease the recurrence [11].
Some anatomical meatal lesions pose another intractable problem in the treatment of genital warts due to the inaccessibility of the lesions and the high tendency to adverse effects, which may include scars, contracture, stenosis, and even malformation. It was reported that PDT has excellent accessibility in the urethra, anus, cervix, and other meatal position via the application of an optical fiber [11,26,27]. In addition, the adverse effects of PDT mainly included transient erythema, edema, slight erosion, and pain [12]. Fluorescence examination has confirmed that brick-red fluorescence was dominantly distributed in HPV-infected epidermis of the skin or mucosa after topical application of photosensitizer [28,29]. The clearly demarcated therapeutic area of PDT accounts for the low rate of long-term adverse effects, that is, no scars, contractures, or stenosis would be induced by PDT. These findings indicated that PDT is superior in accessibility and safety compared with other treatments in the treatment of meatal genital warts.
We have also noticed that the follow-up duration was relatively short in the analyzed RCTs, as the clearance rate of lesions was usually evaluated 1 week after the last treatment, and the recurrence rate was evaluated after 3 months. As the genital warts tend to relapse after conventional treatments [4,30], it is recommended that future RCTs evaluate the clearance rate and recurrence rate with longer follow-up durations of at least 6 months. Quantitative detection of HPV will also help verifying the clearance of HPV-infection and the long-term effects of PDT.
Pain is one of the most common adverse effects of PDT and it limits the application of PDT in clinical practice to some extent. Existing studies have tentatively confirmed that PDT-related pain could be reduced to a slight and tolerable extent by decreasing the power density and prolonging the light irradiation time. These studies provide evidence of modified PDT protocols with good quality and feasible strategies for the extensive application of PDT in clinical practice.
In addition, it was also found that many studies have evaluated the efficacy and safety of combined treatments in comparison with some conventional or first-line treatments, for instance, the efficacy of CO 2 -LT in combination with PDT compared with CO 2 -LT alone [31][32][33] or PDT alone [34], or the efficacy of cryotherapy plus PDT vs cryotherapy alone [35]. These studies provided large amounts of evidence and experience to improve the clinical outcome of the treatment of genital warts. However, this study was aimed evaluating the efficacy and safety of PDT alone for genital warts. A combination with other treatments might enhance the treatment efficacy but would have disturbed our evaluation. Therefore, these research studies were not included in our study.
One limitation of this study was that most of the included RCTs were conducted on Chinese patients. The reason might be that condylomata acuminata had been approved as an indication for PDT in China [12], but not in other countries conducting many RCTs. Besides, ALA is the only topical photosensitizer approved in China, therefore, the efficacy of PDT for genital warts in other countries or with other photosensitizers remains unclear. Though several studies from different countries have reported successful treatment of genital warts with methylaminolevulinate-PDT [36,37], further well-designed RCT studies are still limited to evaluate the efficacy and safety of PDT in these conditions. These factors contributed to the bias of this study. Other limitations of our conclusions originating from the high risk for bias of the seven RCTs retained, the small number of included RCTs, and the publication bias might not be negligible.
Our team was one of the earliest in the world who applied PDT in the treatment of urethral condylomata acuminata as well as other genital warts, and deeply recognized the efficacy, the adverse effects, and mechanism of this treatment. Nevertheless, in other countries, PDT in this field is still not extensively applied. More low-risk-for-bias RCTs involving more countries are still needed to help understand this treatment, in particular the efficacy of PDT for genital warts in different anogenital regions, the efficacy in immunodeficient patients, the possibility to optimize PDT protocols in order to minimize pain, and the costeffectiveness of this treatment.
Despite the above limitations, our results highlight that PDT is a favorable treatment for genital warts with high efficacy, low recurrence rate, and mild adverse effect, especially for urethral and other meatal lesions. In the future, more well-conducted RCTs are still needed to optimize the protocol and verify the outcome on a larger and wider population so as to provide further evidence so that PDT becomes more standardized and popular for the treatment of genital warts.