Efficacy of hydrocolloid dressing for hand-foot skin reaction: J-SUPPORT 1701 APRON trial

Abstract

Purpose

Since hand-foot skin reaction (HFSR) is a dose-limiting toxicity associated with certain multikinase inhibitors, prophylactic measures to manage this adverse effect require improvement. This study aimed to evaluate the efficacy of depressurization using hydrocolloid dressing by comparing the effects of adding hydrocolloid dressing to standard prophylactic moisturizing versus moisturizing alone.

Methods

We conducted a phase 3 randomized self-controlled study for patients with unresectable colorectal cancer, gastrointestinal stromal tumors, or hepatocellular carcinoma who planned to receive either regorafenib or sorafenib. The primary endpoint was an incidence of grade 2 or higher HFSR (soles of the feet only) assessed using NCI CTCAE v.4.0. HFSR grading was performed according to pre-determined criteria, comprising a blinded central review of photographs of the foot, pain intensity (numerical rating scale), and patient-assessed interference score.

Results

Between January 2019 and January 2021, 50 patients were enrolled. 48/50 patients received regorafenib, and two received sorafenib. The incidence of grade 2 or higher HFSR was significantly lower in the hydrocolloid dressing group than in the control group (20% vs. 50%, p < 0.0001, 95% CI of risk difference: [-42.7%, -17.3%]). Time to grade 2 or higher HFSR was significantly longer in the hydrocolloid dressing group (hazard ratio 0.32, p = 0.0017). The incidence of moderate, severe, or very severe in patient-reported HFSR, assessed using PRO-CTCAE, was also lower in the hydrocolloid dressing group (10% vs. 32%, p = 0.0002).

Conclusions

Our study showed that the prophylactic use of hydrocolloid dressing was effective for HFSR.

Trial Registration

J-SUPPORT1701, registered in December 2018 with the University Hospital Medical Information Network (UMIN000034853).

Supplementary Information

The online version contains supplementary material available at 10.1007/s00520-026-10902-9.

Introduction

Multikinase inhibitors perform targeted inhibition of multiple kinases, including vascular endothelial growth factor receptor (VEGFR) and platelet-derived growth factor receptor (PDGFR). While they cause fewer systemic side effects compared to cytotoxic anticancer drugs, they frequently induce localized adverse events, particularly hand-foot skin reaction (HFSR), which is a dose-limiting toxicity [1–5]. Although the detailed mechanism is unclear, it is hypothesized that vascular repair functions are impaired when VEGFR and PDGFR are inhibited. This leads to damage in the blood vessels and fibroblasts of the skin's epidermis on the hands and feet, where pressure and friction are common in daily life, resulting in the manifestation of HFSR [6]. HFSR begins with mild redness and discomfort, followed by abnormalities in the stratum corneum, edema, blisters, bleeding, and pain. Additionally, the incidence and severity of HFSR are higher in Asians, including Japanese, than in Western populations [2]. Unlike HFSR caused by fluoropyrimidine-based drugs, HFSR caused by multikinase inhibitors tends to worsen more rapidly and severely, often leading to treatment discontinuation [7]. Therefore, preventing and mitigating the worsening of HFSR, which significantly impairs quality of life, is clinically important [8].

Topical moisturizers are commonly recommended as standard supportive care for the prevention of HFSR; however, the evidence regarding the optimal type of moisturizer and its efficacy remains inconsistent. In a multicenter phase III trial involving patients with advanced hepatocellular carcinoma receiving sorafenib, prophylactic application of a 10% urea cream significantly reduced the incidence of both any-grade and grade ≥ 2 HFSR [9]. In contrast, a more recent double-blind randomized trial did not demonstrate a reduction in HFSR incidence with twice-daily application of a 10% urea cream [10]. Similarly, a randomized, double-blind, placebo-controlled trial evaluating twice-daily 10% urea cream for the secondary prophylaxis of HFSR in renal cell cancer patients on sunitinib therapy also failed to show a significant benefit [11].Although a meta-analysis of clinical trials published up to 2021 suggested a potential benefit of urea-based prophylaxis [12], these more recent data, together with a retrospective comparative study from Japan showing a significant reduction in the incidence of any-grade HFSR with prophylactic 0.3% heparinoid ointment[13], indicate that the superiority of urea cream over other moisturizers remains uncertain. Additionally, an exploratory study involving hepatocellular carcinoma patients treated with sorafenib showed a favorable trend in the prophylactic use of topical steroids compared to the non-application group; however, no statistically significant difference was observed [14], and its effectiveness has not been demonstrated.

Experts recommend topical steroid therapy to prevent the worsening of HFSR; however, no prospective clinical studies have demonstrated its usefulness [15]. Nevertheless, since no effective treatment other than dose reduction or discontinuation of the primary therapy has been established, topical steroids are used in clinical practice. Additionally, a small RCT involving patients with renal cell carcinoma who developed Grade 1 HFSR following sorafenib treatment reported that the group receiving hydrocolloid dressings, a wound covering material, demonstrated significant prevention of symptom progression to Grade 2/3 compared to the group using a cream containing 10% urea [16].

Hydrocolloid dressings are a type of wound covering material commonly used for wounds, including pressure ulcers. They primarily aid in the healing of wounds that reach the dermis, providing protection and promoting wound healing. By adhering to the skin, these dressings can reduce external friction and shearing, thereby potentially alleviating pain.

Factors like pressure and friction on the hands and feet worsen HFSR, and we hypothesized that wound dressings could not only manage symptoms after the onset of HFSR but may also have a preventive effect. The current standard supportive therapy for preventing HFSR is moisturizing. This study investigated the preventive effect of prophylactic hydrocolloid dressing application in patients undergoing multikinase inhibitor therapy for colorectal cancer, gastrointestinal stromal tumors (GIST), and hepatocellular carcinoma.

Materials and methods

This was an intra-individual RCT, and the protocol was approved by the Japan Supportive, Palliative and Psychosocial Oncology Group (J-SUPPORT) under the designation J-SUPPORT1701, registered in December 2018 with the University Hospital Medical Information Network (UMIN000034853), and published as a protocol paper [17]. Approval was also obtained from the Clinical Research Review Board of the National Cancer Center Hospital (2018–160). This study was conducted in accordance with the Declaration of Helsinki and its later amendments.

Participants

This study involved patients with colorectal cancer, GIST, or hepatocellular carcinoma who began outpatient treatment with either regorafenib or sorafenib from January 2019 onwards. The standard treatment regimen for regorafenib was 160 mg once daily for three consecutive weeks, followed by one week off, constituting one cycle. For sorafenib, the standard dose was 800 mg daily for four consecutive weeks, constituting one cycle. The dosage was adjusted based on the patient’s condition at the discretion of the attending physician. The eligibility criteria for participants included (1) diagnosis of colorectal cancer, GIST, or hepatocellular carcinoma, (2) unresectable advanced and recurrent disease, (3) treatment with either regorafenib or sorafenib, (4) aged 20 years or older, (5) performance status of 0 to 2 according to the Eastern Cooperative Oncology Group Performance Status Scale criteria, (6) ability to communicate verbally, (7) functional capability to walk, (8) absence of any skin lesions on the soles of the feet that would hinder evaluation, and (9) ability to provide written informed consent. Exclusion criteria included mental illness or psychiatric symptoms that would interfere with daily life and make participation in the study difficult.

Patient enrollment, randomization, and blinding

When the decision to initiate treatment was made during an outpatient visit, the attending physician informed the researchers about the candidate. The researchers then provided the participants with an informed consent form and obtained their written consent. After consent was obtained and before the initiation of the first treatment, a nurse entered the eligibility criteria into Viedoc 4, an electronic data collection (EDC) system (Viedoc Technologies AB, Sweden), and completed the enrollment. At the data center, the soles of the left and right feet were randomly assigned to either the trial or the standard supportive therapy group using the minimization method. The cancer treatment drug was set as a stratification factor for the randomization.

This study involved the intervention of applying dressing agents, and blinding was not implemented for the clinicians, nurses, or patients. However, HFSR severity was assessed by trained dermatologists who evaluated the condition using photographs in a blinded manner.

Intervention method

Nurses evaluated the assigned supportive therapy on the soles of the feet once a week on the evaluation days, starting from the day before or the day of initiation of the primary treatment until the end of the evaluation period. Prior to this, nurses provided instructions to the patients regarding self-care practices.

Trial supportive therapy group (Intervention group: Hydrocolloid dressing application group)

On the day before or the day of the primary treatment initiation, a nurse applied commercially available rectangular hydrocolloid dressings (6 cm × 7 cm) to the predicted sites of HFSR (the first toe, the base of the first to the fifth toes, and the heel). The dressing applied to the first toe was trimmed to match the anatomical shape, whereas two intact dressings were placed side by side across the base of the first to the fifth toes. For the heel, the dressing was applied in its original shape, with small peripheral slits added to allow better conformity to the curvature of the foot (Supplementary Fig. S1 in Online Resource 1). The continuous application period for each dressing was generally set for one week (with an allowance of an additional two days). The nurse replaced the dressings on evaluation days and continued the application until the end of the evaluation period. During the dressing changes, the entire sole was cleansed with foam to ensure cleanliness. If the dressing came off before the next evaluation day, the patient or caregiver was instructed to reapply it to the same area after receiving prior instruction from the nurse, and one replacement set of dressings was provided for this purpose. The dressing is self-adhesive and waterproof, allowing continuous wear during daily activities until the scheduled weekly visit. For areas not covered by the dressing, the patient applied a heparinoid moisturizing agent twice daily. The hydrocolloid dressing application continued even when HFSR (Grade 1 or higher) developed, and steroid ointment was not used.

Standard supportive therapy (Control group: moisturizing only)

From the day of initiating the primary treatment, patients applied a protocol-specified cream-type heparinoid moisturizing agent containing 0.3% heparinoid twice daily, approximately 0.5 g per foot per application. The heparinoid moisturizing agent was selected because of its low-irritation profile, allowing continued use even when epidermal detachment associated with HFSR occurred. When HFSR (Grade 1 or higher) developed, patients applied a very strong or the strongest topical steroid ointment twice daily to the affected areas.

Evaluation method

The evaluation sites for the study were both soles of the feet, with baseline assessments conducted on the day before or the day of initiating the primary treatment. Follow-up evaluations were performed weekly until the fourth week, coinciding with outpatient visits. HFSR severity was assessed using the criteria for “palmar-plantar erythrodysesthesia syndrome” from the National Cancer Institute’s (NCI) Common Terminology Criteria for Adverse Events, version 4.0 (CTCAE Ver. 4.0). The criteria involved a comprehensive Grade determination based on (1) image assessment, (2) pain intensity, and (3) the degree of impact on daily living activities. The final CTCAE grade at each assessment time point was determined using a predefined grading matrix that integrated these components according to the study protocol. The image-based assessment was centrally evaluated by three blinded dermatologists using standardized photographs of each sole (erythema, edema, hyperkeratosis, blisters, desquamation, and bleeding). Inter-rater agreement exceeding 80% was confirmed before study initiation. Two dermatologists independently conducted the primary evaluations, and discrepancies were resolved by a third dermatologist based on majority agreement. For subjective pain intensity assessment, patients rated their symptoms using an 11-level Numerical Rating Scale (NRS). Similarly, the patient-reported degree of impact on daily living was assessed using a three-level Verbal Rating Scale (VRS), a Likert-type scale, with options including “1. No problems walking,” “2. Pain present but able to walk,” and “3. Difficulty walking due to pain.” Adverse events related to the intervention were monitored and evaluated with a focus on expected local reactions associated with the hydrocolloid dressing, including skin erosion, erythema, pruritus, and contact dermatitis. Patient-reported outcomes related to HFSR were assessed using the single “Hand-foot syndrome” symptom term from the National Cancer Institute’s Patient-Reported Outcome version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). This item provides an integrated evaluation of hand–foot symptoms, including cracking, peeling, redness, and pain. Patients rated the worst severity experienced during the past 7 days using a five-level verbal rating scale (“none,” “mild,” “moderate,” “severe,” and “very severe”), which was collected via ePRO. PROs were electronically collected using a tablet provided by the facility during each treatment cycle visit. Clinical data entry, data management, and central monitoring were conducted using Viedoc 4, while patient-reported outcomes were gathered through ePRO and Viedoc Me. The J-SUPPORT data center was responsible for data collection and management.

Primary endpoint

The primary endpoint was defined as the incidence rate of HFSR graded 2 or higher observed at least once at the evaluation sites (both soles of the feet) during the evaluation period. This endpoint was established because the need for dose reduction, interruption, or discontinuation of the primary treatment due to HFSR is based on the occurrence of pain that significantly impacts activities of daily living classified as Grade 2.

Secondary endpoint

One of the secondary endpoints was the difference in the duration until the onset of HFSR graded 2 or higher on each sole after the initiation of the treatment. Other secondary endpoints included the intergroup differences in the worst pain intensity (NRS) caused by HFSR and the worst values of the NCI PRO-CTCAE HFSR, both evaluated on each sole. Furthermore, the incidence rate of HFSR graded 3 or higher observed at least once during the evaluation period was included. The study also assessed adverse events related to the intervention and compliance with the intervention as secondary endpoints. Compliance was evaluated based on adherence to predefined protocol procedures, including twice-daily application of the heparinoid moisturizing agent and maintenance of continuous dressing attachment. Adherence was recorded using a daily patient diary and confirmed by investigators. Compliance percentage was defined as the proportion of completed checklist items during the evaluation period.

Statistical analysis

The efficacy analysis included patients who underwent the primary treatment, received the intervention on the feet, and had at least one evaluation for HFSR. For the analysis of the primary endpoint, a generalized linear model was used, considering allocation adjustment factors as covariates and accounting for intra-individual correlation with robust estimation. The incidence rates of HFSR graded 2 or higher for each therapy, along with their point estimates and 95% confidence intervals, were calculated for intergroup comparison. No gatekeeping was performed for missing data, and the significance level was set at one-sided 2.5%.

For the secondary endpoints that were time-to-event type, estimation was performed using the Kaplan–Meier method, followed by the log-rank test. The intergroup differences in pain intensity and the grades of the PRO-CTCAE HFSR were assessed using the Wilcoxon signed-rank test and the Friedman test. The intergroup differences in the incidence rates of moderate or higher and severe or higher HFSR based on the PRO-CTCAE were analyzed using the same methods as the primary endpoint. Since all tests for secondary endpoints were exploratory, no adjustments for multiplicity were made, and the significance level was set at two-sided 5%.

Based on prior studies [16], it was anticipated that hydrocolloid dressings would reduce the incidence of HFSR graded 2 or higher by an absolute value of 25%. Assuming incidence rates of 5% and 30% in the trial and standard supportive therapy groups, respectively, with a one-sided significance level of 2.5% and a power of 80%, the sample size was calculated to be 45. Considering an expected rate of approximately 10% for non-evaluable cases, the required sample size was set at 50 (100 soles).

Results

Between January 2019 and June 2021, 50 participants were enrolled, and there were no disqualified cases. After random allocation, the intervention commenced. Although 11 patients discontinued the intervention, they were not excluded from the analysis; therefore, 50 participants (100 soles) were included in the analysis. The reasons for discontinuation were disease progression (n = 3), TKI-related adverse events unrelated to HFSR (n = 4), patient request for intervention to both feet (n = 1), and other reasons (n = 3) (Fig. 1). There were no adverse events related to the intervention, including skin erosion, erythema, pruritus, or contact dermatitis associated with the dressing, and over 70% of patients maintained a compliance of 80% or higher (Supplementary Table S1 in Online Resource 1).

Fig. 1.

Study flow

There was no significant bias in patient attributes between the two groups, with the median age of the 50 patients being 62.5 (range: 28–82) years, the proportion of female patients being 32% (n = 16 patients), and the proportion of male patients being 68% (34 patients). The distribution of cancer types was as follows: colorectal cancer in 39 cases (78%), GIST in 6 cases (12%), and hepatocellular carcinoma in 5 cases (10%). The performance status (PS) was 0 in 72% of cases, 1 in 26% of cases, and 2 in 2% of cases (Supplementary Table S2 in Online Resource 1). The hydrocolloid dressings were manageable throughout the study period using only one replacement set provided to each participant, with no reported practical issues related to dressing adherence or compatibility with usual footwear.

During the primary treatment, 48 participants were treated with regorafenib and two with sorafenib. Among these, the completion rate of the primary treatment without dose reduction or discontinuation during the study was 50% for regorafenib and 0% for sorafenib (Supplementary Table S3 in Online Resource 1).

The incidence rate of HFSR graded 2 or higher (on the soles) for the primary endpoint was 20% in the trial supportive therapy group, which was significantly lower (p < 0.0001, risk difference: −30%; 95% CI: −42.7 to −17.3). than that of the standard supportive therapy group (50%) (Table 1 and Supplementary Fig. S2 in Online Resource 1). The incidence of Grade 3 HFSR in the trial and standard supportive therapy groups were 0% and 8%, respectively.

Table 1.

Incidence of HFSR (CTCAE)

	Intervention	Control	p
Total	50 (100%)	50 (100%)
None	11 (22%)	17 (34%)
Grade 1	29 (58%)	8 (16%)
Grade 2	10 (20%)	21 (42%)
Grade 3	0	4 (8%)	0.3692 Friedman
		(Weighed)	0.18 Wilcoxon
Any	39 (78%)	33 (66%)	0.1718
Led to Tx modification	5 (10%)	7 (14%)
≥ Grade 2	10 (20%)	25 (50%)	< 0.0001
	Risk difference: −30% (95% CI: −42.7 – −17.3)		(one-sided)
Led to Tx modification	1 (2%)	6 (12%)

The time until the onset of HFSR graded 2 or higher was significantly longer in the trial supportive therapy group than in the standard supportive therapy group (median: not reached vs. 28 days; hazard ratio in the proportional hazards model: 0.32 [95% CI: 0.16–0.68]; p = 0.0017 by log-rank test) (Fig. 2).

Fig. 2.

Time to onset of HFSR ≥ Grade 2. HFSR, hand-foot skin reaction

The incidence rate of HFSR (sole only) classified as moderate or higher based on the PRO-CTCAE was 10% in the trial supportive therapy group, which was significantly lower (p < 0.0002) than that of the standard supportive therapy group (32%) (Table 2).

Table 2.

Patient-reported HFSR (PRO-CTCAE)

	Intervention	Control	p
Total	50 (100%)	50 (100%)
None	27 (54%)	21 (42%)
Mild	18 (36%)	13 (26%)
Moderate	5 (10%)	8 (16%)
Severe	0	6 (12%)
Very Severe	0	2 (4%)	0.0362 Friedman
		(Weighed)	0.002 Wilcoxon
Any	23 (46%)	29 (58%)	0.0997
Led to Tx modification	3 (6%)	7 (14%)
≥ Grade 2	5 (10%)	16 (32%)	0.0002
Led to Tx modification	1 (2%)	5 (10%)

Pain intensity was significantly lower in the trial supportive therapy group at all evaluation time points (Fig. 3). The NRS values for the trial supportive therapy group and the standard supportive therapy group were as follows: baseline (0.02 [mean] ± 0.14 [SD] vs. 0.02 ± 0.14), Week 1 (0.30 ± 1.11 vs. 0.62 ± 1.48, p = 0.03), Week 2 (0.36 ± 0.95 vs. 1.27 ± 2.42, p = 0.001), Week 3 (0.74 ± 1.64 vs. 1.72 ± 2.58, p < 0.001), and Week 4 (0.58 ± 1.40 vs. 1.17 ± 1.48, p = 0.01). At all-time points except baseline, the NRS in the trial supportive therapy group was significantly lower than in the standard supportive therapy group (Fig. 3).

Fig. 3.

Severity of Pain (NRS). NRS, Numerical Rating Scale

Discussion

In this study, the prophylactic application of hydrocolloid dressings significantly reduced the incidence rate of HFSR graded 2 or higher compared to standard treatment, demonstrating its superiority (risk difference: −30%; 95% CI: −42.7 to −17.3) (Table 1 and Supplementary Fig. S2 in Online Resource 1). Additionally, an extension in the time until onset of Grade 2 HFSR or higher (Fig. 2) and the patient’s subjective evaluations (Table 2) also demonstrated similar superiority. A randomized phase 2 comparative trial showed that hydrocolloid dressing application was significantly superior in reducing the incidence rate of HFSR graded 2 or higher and extending the time until onset of Grade 2 HFSR, specifically among patients with renal cell carcinoma and Grade 1 HFSR [16]. The present study demonstrated the efficacy of a simpler approach that uses hydrocolloid dressings from the initiation of the multikinase inhibitor treatment in a larger patient population, including those with hepatocellular carcinoma. Furthermore, regarding pain intensity (NRS), the hydrocolloid dressing group showed significantly lower pain levels from the first week of treatment compared to the moisturizing-only group, and this effect was maintained throughout the treatment period (Fig. 3). This suggests that preventive application at the initiation of treatment can suppress the onset of HFSR at an earlier stage and lead to pain prevention. While a prophylactic application may be more favorable, further investigation is needed to determine whether the timing of application—preventive versus after the onset of Grade 1 HFSR—offers greater benefits, considering both preventive effect and cost. From a practical implementation perspective, a risk-adapted strategy may be reasonable, in which prophylactic use is prioritized for patients at higher risk of HFSR, such as those receiving agents associated with higher incidence rates or individuals with active daily lifestyles involving prolonged ambulation. HFSR is believed to be exacerbated by pressure and friction on the hands and feet in daily life [7]. Hydrocolloid dressings are wound-covering materials that provide moisturizing, protection, and irritation removal, likely contributing to their significant effectiveness when applied prophylactically to the common sites of HFSR. This indicates the general importance of moisturizing and protection in preventing the onset of HFSR. It is also crucial to establish a multidisciplinary system for continuous patient education and to enhance patient adherence to self-care practices related to moisturizing and protection.

The strength of this study lies in the use of an intra-individual randomized comparative trial, which eliminated the influence of external confounding factors related to daily life stimuli, thereby ensuring reliability and validity. However, while HFSR can also manifest on the palms and fingers, this trial focused on the soles, which are less affected by individual differences such as daily activities or the dominant hand. Consequently, the study could not address the effectiveness of preventive measures for HFSR occurring on the palms and fingers, highlighting a limitation of the research. In addition, because of the intra-individual trial design, global outcomes such as quality of life or patient satisfaction were not assessed, representing another limitation of the study.

Conclusions

We conducted an intra-individual randomized trial involving patients with colorectal cancer, GIST, and hepatocellular carcinoma undergoing multikinase inhibitor treatment. We show that the preventive application of hydrocolloid dressings to the common sites on the soles was more effective in preventing HFSR and alleviating symptoms compared to moisturization alone.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Asako Ryu, Sadamoto Zenda, Yoichi Shimizu, Atsuo Takashima, Arata Tsutsumida, Akira Takahashi, Michiko Arai, Chigusa Morizane, Takuhiro Yamaguchi, Yusuke Takagi, Tomoe Mashiko, Tempei Miyaji, Takashi Kawaguchi, and Naoya Yamazaki. The first draft of the manuscript was written by Asako Ryu, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the Kimura Foundation for Nursing Education [grant number FY2017]. The funder had no role in the study design; in the collection, analysis or interpretation of data; in the writing of the report; or in the decision to submit the article for publication.

Data availability

The data generated and analyzed in this study are available from the corresponding author upon reasonable request, subject to approval by the ethics committee.

Declarations

Ethics approval

This was an intra-individual RCT, and the protocol was approved by the Japan Supportive, Palliative and Psychosocial Oncology Group (J-SUPPORT) under the designation J-SUPPORT1701, registered in December 2018 with the University Hospital Medical Information Network (UMIN000034853), and published as a protocol paper [17]. Approval was also obtained from the Clinical Research Review Board of the National Cancer Center Hospital (2018–160).

Consent to participate

Written informed consent was obtained from all individual participants included in the study.

Consent to publish

The authors affirm that human research participants provided informed consent for publication of the images in Supplementary Fig. S1.

Previous presentation

A portion of this manuscript has already been presented at the American Society of Clinical Oncology (ASCO) Annual Meeting 2023, held from June 2 to June 6 in Chicago, Illinois.

Competing interests

The authors declare no competing interests.