An Open-Label Single-Arm Clinical Trial of Propranolol Cream in Infantile Hemangioma

Infantile hemangioma (IH) is a common tumor in infants that gradually resolves and is often followed up for observation. However, for cosmetic reasons, parents often opt for treatment. Oral propranolol, the rst-line therapy for IH, shows several side effects, including hypotension, bradycardia, and hypoglycemia. No clinical studies on topical propranolol have been conducted using standardized procedures. We evaluated the ecacy and safety of topical propranolol in patients with IH. This multicenter, open-label phase II study was conducted from June 2019 to December 2020 and involved 8 Japanese infants aged 35–150 days with proliferating IH. Patients were treated with 5% propranolol cream twice daily. We examined the ecacy rate based on central evaluation (complete or near-complete healing of the target hemangioma) at weeks 24 and 12, respectively, compared to baseline values. The ecacy rate at week 24 was 68.8% (95% condence interval: 44.1–85.9%). The surface area, maximum diameter, and color intensity of the target IH decreased over time. Adverse event and drug-related adverse event rates were 87.5% and 0%, respectively. Propranolol cream (5%) is effective and safe in Japanese patients with IH and may be considered a rst-choice treatment for small and supercial IHs in cosmetically problematic areas.


Introduction
Infantile hemangioma (IH) is the most common soft-tissue tumor in infants, occurring in 4-5% of infants.
Tumors of IH occurring within the rst week of life grow rapidly over the rst 3-6 months of life. By the age of 4 years, 90% of IHs regress spontaneously [1,2]. Therefore, IH tumors are often followed up for observation [1,2].
Most IHs occur in the periphery of the head and neck areas [3], which can seriously affect a child's appearance and psychology [4]. Even when associated vital, functional, and cosmetic problems are mild, parents are often concerned about the cosmetic aspects because most IHs form on the body surface, and thus, the parents request treatment for IH [4]. In 2008, a French patient with hypertrophic obstructive cardiomyopathy complicated by IH was treated with propranolol, a nonselective beta-blocker, which ameliorated the hemangioma [5]. This led to the recognition of oral propranolol therapy against IH [5]. Today, oral propranolol is approved worldwide and is the rst-line treatment for IH in patients with a potential risk of dysfunction and serious condition [1,2]. However, several side effects have been reported, including hypotension, bradycardia, sleep disturbance, hypoglycemia, hyperkalemia, and respiratory symptoms [1,2]. These side effects occur because of the systemic effects of propranolol, as it is an oral formulation, although the currently approved oral propranolol therapy is highly effective against IH [1].
Several reports are available on topical application of propranolol for IH [6]. However, to date, no clinical trials on topical propranolol have been conducted using standardized procedures. Previous studies only included patients with regressed IH, involved longer treatment periods to target the regressed period, were clinical trials with complex or subjective evaluation methods without standardized procedures, or were observational studies [6]. Therefore, this study was conducted to evaluate the e cacy and safety of 5% propranolol topical cream in Japanese patients with IH.

Patients
A total of 8 patients participated in the study between July 2019 and July 2020. All patients received and completed 24 weeks of study treatment and were analyzed ( Figure 1). The patient characteristics are summarized in Table 1. There were 2 males and 6 females. Among them, 2 patients had a gestational age of less than 37 weeks. The mean (±SD) age at the start of propranolol therapy was 91.8 (±39.7) days. The IH was located on the scalp, back, upper and lower extremities, and genitalia. The clinical type of all patients was super cial. One patient with IH had ulcer.

E cacy
The e cacy rate of propranolol at weeks 24 and 12 is summarized in Tables 2 and S1. According to the centralized assessments, the e cacy rate at week 24 from baseline was 68.8% (95%CI: 44.1-85.9%), and the lower limit of the 95%CI exceeded the e cacy criteria of 12%. In comparison, the e cacy rate at week 12 from baseline was 31.3% (95% CI: 14.1-55.9%). Two assessors showed a Cohen's Kappa coe cient of 0.714, which was virtually identical. As shown in Table 2, the mean surface area (3.6, 3.5, and 2.7 cm 2 at weeks 0, 12, and 24, respectively), mean maximum diameter (2.8, 2.5, and 2.2 cm at weeks 0, 12, and 24, respectively), and mean color intensity (21.3, 14.7, and 9.2 at weeks 0, 12, and 24, respectively) decreased over time. All cases of IH in patients treated with propranolol before treatment and at weeks 12 and 24 are shown in Figure 2

Safety
AEs that occurred during the study treatment are summarized in Table 3. The incidence of any AEs and drug-related AEs were 87.5% (7 in 8 patients) and 0%, respectively. The AEs were mild elevation in liver enzymes, upper respiratory tract in ammation, abrasions, eczema, subcutaneous hematomas, urticarias, and infectious dermatitis. Mild elevations in liver function enzymes observed during the study period resolved spontaneously. Upper respiratory tract in ammation was mild in all cases and resolved with expectorant oral treatment. Eczema, which occurred in a different area from the target IH, resolved with moisturizer treatment. Urticaria that occurred in a different place from the subject IH was cured with oral antihistamines. Abrasions and subcutaneous hematomas occurred in areas other than the target IH and were cured without treatment. Infectious dermatitis occurred at the target IH site and quickly healed with oral antimicrobial treatment. None of the reported risks associated with propranolol, such as hypotension, bradycardia, hypoglycemia, or bronchial asthma, were observed in the patients during the study treatment (Table S2).
The plasma concentrations of propranolol in all patients are shown in Table 4. At weeks 4 and 24 of the study treatment, 37.5% (3 of 8 patients) and 87.5% (7 of 8 patients) of patients, respectively, showed propranolol in the plasma. The highest plasma propranolol concentration was 4.42 ng/mL at week 4 and 5.31 ng/mL at week 24.

Discussion
The present clinical trial demonstrated the e cacy and safety of 5% propranolol topical cream in Japanese patients with IH. The method for evaluating e cacy was similar to that used in previous clinical trials [7,8]. In the current clinical trial, the e cacy rate of 5% propranolol topical cream against IH after 24 weeks of treatment was 68.8%, which was higher than that of the placebo group in the phase II/III clinical trial [8]. Regarding safety, no severe AEs, which are often associated with propranolol, were observed during the study. E cacy was demonstrated by the success rate at week 24 and decrease in surface area, maximum diameter, and color intensity. We used a placebo value of 12% as our estimate, which was the placebo value for the oral drug in the reported clinical trial [8]. This value was used because there was no valid estimate that could be used for topical agents. The e cacy rate at week 12 was low (31.2%), suggesting some time is required for the drug to produce an effect. Moreover, the surface area, maximum diameter, and color intensity exhibited a decreasing trend over time. The CIE 2000 color difference formula used to calculate the color difference is characterized by its high correlation with visual inspection and is used for e cacy assessment of treatment of IH [7,8].
The cream was shown to be safe with no serious side effects such as hypotension, bradycardia, or hypoglycemia, unlike the oral formulation. In a clinical study of an oral liquid formulation in Japanese subjects, the median simulated trough concentrations in the plasma after repeated oral administration of propranolol in this study may have resulted in the low incidence of adverse drug reactions compared with oral drug administration. These results support the safety of topical propranolol cream.
The e cacy and safety of topical timolol or propranolol have been described in several uncontrolled case reports and case series; however, only one randomized, controlled trial comparing timolol with placebo has been reported [1,10]. Currently, no topical beta blockers are available for IH. A cream formulation was chosen in this study because a previous study comparing the effects of ointment, cream, and gel formulations of propranolol hydrochloride on porcine skin showed that creams exhibited higher retention of the drug effect on the skin surface, whereas peripheral vascular penetration was lower [11]. Topical beta blockers are expected to have a local effect because they are not subjected to the rst-pass effects of the liver, once safety and e cacy are con rmed [1,12].

Ethics
This study protocol was approved by the ethics committee of the Hamamatsu University School of Medicine in Japan. The study was conducted in accordance with the guidelines of the Declaration of Helsinki and other applicable regulations. Written informed consent was obtained from the legal guardians of all patients prior to the study. This trial was registered at the Japan Registry of Clinical Trials (jRCTs 041190041).

Participants
The study included 35-150-day-old patients who were diagnosed with infant hemangioma of the super cial type with a diameter of 1.5 cm or more. Patients were recruited from the Department of Pediatrics, Hamamatsu Medical Center, and Chutoen General Medical Center in Shizuoka, Japan. Patients were excluded if they had the following conditions: 1. IH affecting life or function, 2. Kasabach-Merritt syndrome, 3. receiving treatment for IH (e.g., laser therapy and propranolol oral therapy), 4. poorly controlled heart failure, 5. a history of bronchial asthma or bronchospasm, 6. hypoglycemia (blood glucose <40 mg/dL), 7. hypotension (systolic blood pressure <50 mmHg, diastolic blood pressure <30 mmHg), and 8. bradycardia (pulse <80 beats per min).

Study Design and Treatments
This was a multicenter, phase II, open-label, single-arm study conducted in Japan from June 2019 to December 2020. Patients with IH were treated with 5% propranolol cream twice daily for 24 weeks. The study medication was prepared as follows. Propranolol hydrochloride powder (Tokyo Chemical Industry, Tokyo, Japan), polyethylene glycol 400 (PEG 400, Fuji lm Wako Pure Chemical Corporation, Osaka, Japan), and puri ed water were mixed by inversion and further placed in a warm water bath at 50°C for 15 min. The hydrophilic cream was added to this aqueous solution and mixed using a self-rotating mixer at 100 ×g for 1 min. The parents were instructed to apply the cream using a ngertip onto the surface of the IH (su cient to just coat the tumor) twice per day and to gently rub it in. The condition of IH in participants was observed 8 times: at screening; baseline; and weeks 4, 8, 12, 16, 20, and 24.

E cacy and Safety Assessments
E cacy was assessed by two independent and trained assessors using standardized photographs with a de ned procedure. The assessors evaluated all photos; the date of visit and patient number were masked.
To measure e cacy parameters, we performed intrareader variability assessment in a previous clinical study [7]. Photographs of the target IH were taken according to a standardized procedure by an investigator assigned on day 1 and weeks 12 and 24 of treatment. A color chart for image correction, Casmatch (Bear Medic Corporation, Tokyo, Japan), was placed next to the tumor to adjust the brightness of all photographs. Casmatch, a diagnostic and evaluation method in the eld of dentistry, was used for image analysis of the e cacy evaluation [13,14]. Photographs were taken at least twice for each tumor with the tumor subject identi cation code, and color chart in focus. To determine the area, diameter, and color difference of the hemangioma from the photographs, image analysis was performed. First, color tone and size correction were performed using Casmatch in the images using Adobe Photoshop CC (Adobe Corporation, San Jose, CA, USA) and ImageJ (National Institutes of Health, Bethesda, MD, USA). Second, the tumor of interest was surrounded by a dot, and the area, length, diameter, and hues of red, green, and blue (RGB) were measured. The boundary between the tumor area and normal skin was determined by three investigators. In addition, four points of normal skin were selected to evenly surround the tumor followed by measurement of the RGB, and the average value of the four points was calculated. The color difference (dE* 2000 ) was calculated for the RGB of the tumor and normal skin.
Safety was assessed by analyzing adverse events (AEs, i.e., any adverse change in condition between the time of informed consent and the end of the trial), laboratory investigations (measurement of glucose levels from nger-prick blood samples and serum propranolol concentrations), physical examination, assessment of vital signs and neurodevelopment, and electrocardiography. The plasma concentrations of propranolol were determined using liquid chromatography coupled with tandem mass spectrometry.

Outcome Measures
The primary and secondary outcomes were the rate of e cacy of treatment based on central evaluation (complete or near-complete healing of the target hemangioma) at weeks 24 and 12, respectively, compared with baseline. The other secondary endpoints were changes in the area, maximum diameter, and color intensity of the target IH at weeks 12 and 24.

Power Calculation
The sample size was calculated using the minimax method for a phase II study [15]. The sample size is 6 when the minimum effective rate is 0.12, expected effective rate is 0.6, alpha error is 0.05, and power is 0.8. We considered 12% as our estimate for minimum effective rate, as the e cacy rate [95% con dence interval (CI)] of the placebo group in a phase II/III clinical trial was 3.6% [0.44-12.53] in a previous study [8]. Assuming an incidence of dropout and data rejection of 25%, 8 cases were required. For e cacy assessment, propranolol cream was considered as effective if the lower limit of the 95% CI for e cacy at week 24 was greater than 12%.

Statistical Analysis
The full analysis set or all patients administered the study treatment were the primary analysis set for all planned e cacy and safety analyses. All statistical results were descriptive [qualitative variables: number, percentage, and 95% CI; continuous data: number, mean, and standard deviation (SD)]. No statistical tests were performed.

Limitations
This study had some limitations. First, because of the small size, a control arm could be not set up, and patients with IH at all sites on the body could not be included. Second, the study was conducted only in Japanese patients. Third, the observation period was short (up to 24 weeks). Finally, unlike clinical trials [7,8], we used the easier and more portable Casmatch as a color scale. In this study, standardized procedures were used to evaluate color intensity and size with Casmatch.

Conclusions
In summary, we conducted an open-label, single-arm study of 5% propranolol topical cream in Japanese patients with IH. The e cacy of the cream was found to be similar to that of the oral formulation. Additionally, the safety of the cream was found to be superior to that of the oral formulation, with no observed AEs. Topical propranolol may be considered a rst-line treatment for small and super cial IHs in  Tables   Table 1 Baseline patient characteristics    Photographs of target IH in all patients at day 0 and weeks 12 and 24 Supplementary Files