Topical treatments for Kaposi sarcoma: A systematic review

Abstract Background While treatment options exist for solitary or disseminated Kaposi sarcoma (KS) disease, there are currently no standardized clinical treatment guidelines for cutaneous KS. Objective This systematic review seeks to identify safe and effective topical treatments for cutaneous KS lesions. Methods We conducted a systematic review using peer‐reviewed articles from January 1970 to September 2021 published in the PubMed/MEDLINE and EMBASE databases. Results From the initial search that yielded 590 studies, 34 met the inclusion criteria and were selected. Of the 34 studies, seven were clinical trials, 26 were case reports/series and one was a multicentre study. A total of 634 patients were included in our review. The three most common topical treatments used for cutaneous KS were imiquimod, alitretinoin and timolol. Topical alitretinoin was used in three case reports and three clinical trials. Topical imiquimod was used in eight case reports, one prospective phase II cohort study and one comparative single‐blinded non‐controlled clinical study. Topical timolol was used in nine case reports/series. Our review also identified reports of less widely used topical treatments for cutaneous KS. These include topical diphencyprone (DPCP), all‐trans‐retinoic‐acid, rapamycin and bleomycin‐dimethylsulfoxide (BLM‐DMSO) which achieved variable response rates but have not been widely studied. Conclusion Topical alitretinoin, imiquimod and timolol demonstrated positive responses for cutaneous KS and the treatments were well tolerated. These three topical treatment modalities could be considered by clinicians when treating cutaneous KS.

for AIDS-related KS, to systemic chemotherapeutic agents for disseminated disease. 3 For patients with cutaneous KS, recent clinical guidelines have recommended radiotherapy, surgical excision, cryosurgery, laser therapy, and intralesional agents as treatment modalities. 2,4 Although these treatments are effective in most cases, they may not be feasible for all patients; their use is dependent on patient comorbidities, as well as the anatomical locations and extent of KS lesions. Topical treatment may be more appropriate for some cases of cutaneous KS, as they potentially allow higher drug levels at the tumour site, lower degrees of pain and scarring, and have better safety profiles. The objective of this systematic review is to characterize all the topical treatment modalities for cutaneous KS and evaluate the clinical efficacy and safety of each topical treatment.

| METHODS
This systematic review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. 5

| Eligibility criteria
We reviewed randomized controlled trials, retrospective cohort studies, prospective cohort studies, case series and case reports that utilized topical agents for treatment of cutaneous KS. The following inclusion criteria were used: studies that used topical treatments only, or studies that compared treatment groups between a topical agent and another local or systemic agent. Only articles written in English language were included. Studies were excluded if a non-topical treatment was used for cutaneous KS. Studies were also excluded if other systemic or local treatments were used in conjunction with topical treatments, or if they did not use clinical response as an outcome measure. Local treatments were defined as non-topical agents including radiotherapy, surgical excision, cryosurgery, laser, and intralesional agents. Additionally, review articles, study protocols, letters-commentaries-editorials, and guidelines-recommendations were excluded.

| Information sources and search strategy
A search of two databases, including PubMed/MED-LINE and EMBASE, from January 1970 to September 2021 was conducted. The following Medical Subject Headings and keywords were used appropriate to each database: 'Kaposi sarcoma' or 'Kaposi's sarcoma' and 'topical treatment' or 'topical administration' or 'topicals' or 'topical therapeutics'. Subsequent review of relevant article bibliographies was conducted to identify any additional studies. Full search strategy is available in the supplementary document.

| Selection process
All articles retrieved through database search were imported into Covidence software, where duplicates were automatically removed. Two reviewers (Kyaw Zin Htet and Michael A. Waul) independently assessed the eligibility of each article through an initial title and abstract review. Reviewers were blinded to each other's assessments. Irrelevant articles were excluded. Following title and abstract review, the full text of remaining articles was reviewed for eligibility. Full texts were assessed for eligibility independently and in duplicate. Any full text articles that did not meet eligibility criteria were excluded. Disagreements were resolved by a third author (Kieron S. Leslie) as a tie breaker.

| DATA COLLECTION PROCESS
All included studies were extracted using a prespecified extraction spread sheet by two independent reviewers (Kyaw Zin Htet and Michael A. Waul). The data from the search were extracted and the following information when available was ascertained: patient age and sex, number of cases, HIV status, type of KS, known presence of extracutaneous involvement, drug response, drug regimen, timeline of sustained response and drug-related adverse side effects. The primary

| Study risk of bias assessment
The risk of bias was assessed depending on study design, utilizing the Newcastle-Ottawa Quality Assessment Scale (NOS) for case control studies, cohort studies and randomized controlled trials. 6 Case reports and case series were assessed using a modified NOS. 7 This was performed independently by two independent reviewers (Kyaw Zin Htet and Michael A. Waul) and any discrepancies were reviewed by a third reviewer (Kieron S. Leslie). The NOS tool includes eight points that evaluate the selection, comparability and exposure described in the studies. Points are awarded based on each evaluation criterion and tallied to form a final quality score. The minimum score is 0 and maximum score is 9.
For purposes of this review, high quality was defined by scores ranging from 7 to 9, moderate quality was defined by scores ranging from 4 to 6, and poor quality was defined by scores less than 4. The breakdown of NOS scores for each included article is illustrated in Tables S5 and S6 of the supplementary document.

| Synthesis methods
Each included individual article was rated using the modified Oxford Centre for evidence-based medicine rating scale. 8 Extracted outcome data, NOS score and ratings for each article are summarized in tabulated formats.

| RESULTS
The search results yielded a total of 589 studies that were published from January 1970 to September 2021. After duplicate studies were removed, remaining studies were screened based on title, abstract and full text to assess for eligibility. An additional two studies were identified through bibliography review and included, to produce a total of 34 studies eligible for inclusion. Of these studies, 26 were case reports or series and 8 were randomized clinical trials (RCTs). The selection process is shown in the format of a PRISMA diagram in Figure 1.

| Demographics
In the 34 studies included in this systematic review, the majority of patients were male 609/633 (96.2%), with only 24/633 female (3.8%). Two studies did not have gender specifications. 9,10 Within our included studies, there were 544 cases of AIDS-related KS, 120 cases of classic KS, 6 cases of iatrogenic KS, and 1 case of endemic KS. One study did not report any KS classifications, 9 and another grouped 17 cases of endemic and AIDS-related KS together. 11

| Topical alitretinoin for KS treatment
Three case reports and three clinical trials reported topical alitretinoin (9-cis-retinoic acid) for the topical treatment of cutaneous KS. [12][13][14][15][16][17] Five studies reported using 0.1% topical alitretinoin gel as the dosage 12,14-17 and in one clinical trial, the concentration of topical alitretinoin ranged from 0.05% to 0.1%. 13 The duration of the treatment regimen ranged from 8 to 78 weeks with a median of 12 weeks. There were varying drug frequencies, with twice daily being the most common regimen (Table 1). Of three patients in the case reports, two achieved complete clinical response and one had no response. From a total 194 patients across two clinical trials using topical alitretinoin for cutaneous KS, 2 reported complete response (1.0%), 68 reported partial response (35.1%), 94 reported stable disease (48.5%), and 32 reported progressive disease (16.5%). 12,17 Both trials reported a statistically significant difference (p = 0.00003 12 and p = 0.002 17 ) in outcomes between topical alitretinoin and the control group. In the third clinical trial of 115 patients, there was a statistically significant difference in treatment response between topical alitretinoin and the untreated control group (p < 0.001). 13 However, the level of response achieved in the treatment group was not clearly delineated. 13 One study reported no recurrence of cutaneous KS after 1.5 years follow up 15 and another study reported remission of 90%-100% of the lesions after 3-5 months 16 Four studies did not follow up for the examination of cutaneous KS recurrence (Table 1). [12][13][14]17 In all six studies, of 314 patients treated with topical alitretinoin, 212 patients (67.5%) experienced rash, 50 patients (15.9%) experienced skin disorders and 46 patients (14.6%) experienced pruritus as the most common side effects. The nature of the rash and skin disorder was not clearly defined.

| Topical timolol for KS treatment
Eight case reports/series documented the use of topical timolol as a treatment for cutaneous KS. 27 did not specify the regimen, 30 but six studies reported using 0.5% topical timolol 28,29,31-34 whereas one study used 0.1% topical timolol. 27 Treatment duration ranged from 4 to 24 weeks, with a median of 12 weeks. There were varying medication frequencies, with twice daily application the most common regimen (Table 3). Of 13 patients treated with topical timolol, eight patients (61.5%) had complete clinical response, four patients (30.8%) had partial response and one patient (7.7%) had no response. Seven studies reported no recurrence of KS after variable months of follow up with a median of 12 weeks and a range of 4-24 weeks (Table 3). There were no treatment-related side effects associated with the use of topical timolol. One patient reported asthenia and fever; however, the authors attributed these symptoms to HIV and active tuberculosis during treatment. 27

| Other topical treatments for KS treatment
Nicotine patch was used as a treatment modality in one clinical trial, and of 24 patients using nicotine patch, 2 achieved complete response (8.3%), 4 achieved partial response (16.7%), none achieved minor response (0%), 8 had no response (33.3%), and 7 had disease progression (29.2%). 35 In this study, there was no statistically significant response between nicotine patch and the untreated control lesions (p = 0.74). 35 Topical diphencyprone (DPCP) in 0.2% gel was used in a case report with two patients; both achieved complete response. 36 One case series used all-transretinoic acid as a topical treatment for KS in eight patients. 37 In this case series, one patient achieved complete response and seven achieved partial response. 37 Another case series used topical bleomycin-dimethylsulfoxide as topical KS treatment in eight patients, with all achieving partial response and the disappearance of pain. 38 Topical 0.5% rapamycin ointment provided complete response for one patient in a case study. 39 Topical BCG and cord factor was used for four patients in a case study and all achieved complete response. 40 In one multicenter study, 4,4 0dihydroxybenzophenone-2,4-dinitrophenylhydrazone (A007) was used a topical KS treatment in four patients; two achieved partial response and two had no response. 9 Topical calcipotriene was used in eight patients in a single case series; of eight patients, one achieved complete response, three achieved partial response and four had no response. 10 In one clinical trial of 17 patients using topical halofuginone as KS treatment, 6 patients achieved partial response with no statistical difference between the treatment and the control groups (p = 0.689). 41 For detailed clinical characteristics and outcomes of less widely used topical agents for cutaneous KS see Table 4.  The risk of bias for included studies assessed using Newcastle-Ottawa Quality Assessment Scale. The risk of bias for included studies assessed using Newcastle-Ottawa Quality Assessment Scale.

| DISCUSSION
This systematic review aimed to identify topical treatments for cutaneous KS and evaluate the clinical efficacy of each treatment modality. We also investigated the most widely used treatment regimens and dosages, along with the drug-related adverse effects. Based on our review, the three most common topical treatments were timolol, imiquimod and alitretinoin. These three drugs were easily administered topical medications with minimal drug-related adverse events.
Alitretinoin is the only FDA-approved topical medication for KS, with a labelled indication for AIDS-related cutaneous KS. As with other retinoids, alitretinoin binds to and activates intracellular retinoid receptors; this results in altered gene expression that affects cellular differentiation and proliferation. 13 Imiquimod is an immune response modifier that activates immune cells via Toll-like receptor seven agonist activity. It is FDA-approved for treatment for actinic keratoses, superficial basal cell carcinoma, and genital/perineal warts. However, it has been used for many off-label indications, including the treatment of cutaneous KS. Imiquimod's anti-angiogenic properties are thought to be mediated by cytokine induction (including IL-10 and IL-12), local upregulation of endogenous angiogenesis inhibitors (including TIMP, TSP-1), local downregulation of pro-angiogenic factors (including bFGF, MMP9), and promotion of endothelial cell apoptosis. 11 Timolol is a nonselective beta-blocker; as a topical ophthalmic agent it has a labelled indication for treatment of elevated intraocular pressure. The topical gelforming solution also has demonstrable efficacy for treatment of superficial infantile hemangiomas, and has been widely used by dermatologists for this purpose. 42 The mechanism by which beta blockade mitigates vascular proliferation is not completely understood, but is hypothesized to be mediated by vasoconstriction, inhibition of angiogenesis and induction of apoptosis. 28 Given that KS is a vascular tumour, it is not surprising that lesions would respond to topical timolol treatment. In addition, there are data to suggest that KS lesions may strongly express beta adrenergic receptors, 43 and therefore may be particularly susceptible to betablocker treatment. While the use of topical timolol yielded good response rates in our review, it should be noted that this evidence is based exclusively on case reports and case series. Stronger level of evidence, such as randomized controlled trials, are needed to fully evaluate the efficacy of this drug.
Our review also identified reports of less widely used topical treatments for cutaneous KS. These included topical diphencyprone (DPCP), all-trans-retinoic-acid, rapamycin and bleomycin-dimethylsulfoxide (BLM-DMSO) which achieved variable response rates but have not been widely studied.
The demographics of KS patients in our review are consistent with the published epidemiological data of this disease. 44 The majority of patients with CKS in our included studies was elderly males with a median age of 75 years, while AIDS-related KS was prevalent in HIV positive younger males with a median age of 38.5 years. Overall, our reviewed studies included a significantly higher number of male patients than female patients. Extracutaneous KS may involve the viscera and lymph nodes; it is more difficult to treat and has a poorer prognosis than exclusive cutaneous disease. 45 In most studies in our review, clinical outcome was not stratified based on the extracutaneous KS involvement.
The data on treatment response stratified by lesion size was inconclusive. One study reported a statistically significant positive correlation between small initial tumour area at the beginning of treatment and the area response at the end of treatment (p = 0.0246). 11 Another study did not find any correlation between initial size of the lesions and response rate. 26

| Limitations
Our systematic review was limited by the inclusion of very few RCTs. Secondly, the heterogeneity of studies in describing clinical outcomes made it difficult to provide summary statistics for each topical agent. We found that some studies used AIDS Clinical Trials Group Oncology Committee criteria while others used the modified KS staging system to define clinical response. 46,47 A few studies did not clearly define the criteria for determining clinical response, which could be a source of investigator bias. Additionally, some patients in our included studies were on concurrent medications and had undergone prior KS therapies such as cryotherapy, local radiation and excisional surgery. This could potentially confound clinical results and introduce bias in reporting clinical outcomes. In some studies, immunosuppressive medications were tapered at the time of KS diagnosis, which may also affect the validity of results. In addition, variable rates of follow-up to assess for cutaneous KS recurrence after topical therapy limited our ability to comment on sustained treatment efficacy. 48

| CONCLUSION
Alitretinoin, imiquimod and timolol were the three most common topical agents used to treat cutaneous KS. All three showed clinical efficacy with minimal drug-related adverse events. There were also reports of less widely used topical treatments which showed variable clinical efficacy for cutaneous KS. However, further randomized controlled trials are needed to more accurately determine clinical outcomes, regimens and adverse effects for these agents. In the absence of standardized clinical guidelines on cutaneous KS therapy, our study provided a comprehensive list of topical treatments, and a summary of the available data on their efficacy, to better inform clinicians regarding treatment decisions.