Long-term safety and efficacy of Oleogel-S10 (birch bark extract) in epidermolysis bullosa: 24-month results from the phase III EASE study

Abstract

Background

Epidermolysis bullosa (EB) is a group of rare and severe genetic disorders characterized by persistent skin fragility and open wounds. EB manifests as cutaneous and mucosal blistering, erosions and impaired wound healing.

Objectives

To determine the long-term efficacy, tolerability and safety of Oleogel-S10 (birch bark extract) in dystrophic EB (DEB) and junctional EB (JEB) in the 24-month open-label phase (OLP) of the EASE study.

Methods

EASE was a double-blind randomized controlled phase III study consisting of two phases: a 90-day double-blind phase (DBP) and a 24-month OLP. Patients from both former treatment groups in the DBP entered the single-arm OLP (n = 205). Patients received Oleogel-S10 on all partial-thickness EB wounds. OLP endpoints included the incidence and severity/relatedness of adverse events (AEs), maximum wound infection severity, changes in body surface area percentage (BSAP) of wounds, EB Disease Activity and Scarring Index (EBDASI), pain, itch, disease severity and quality-of-life outcomes.

Results

The OLP data demonstrated that Oleogel-S10 target wound treatment adherence was > 99% and mean (SD) treatment duration was 584.7 (246.1 days). Seventy-two per cent of patients in the OLP were aged < 18 years and 86.8% had DEB; recessive DEB predominated (78.0%). AEs were reported in 77.1% of patients and were typically mild-to-moderate in severity. Severe and serious AEs were seen in 18.0% and 24.4% of patients, respectively. AEs resulted in the withdrawal of 7.8% of patients (n = 16), including three with treatment-related AEs. Nine deaths were reported; none were attributable to the treatment. The incidence of target wound infections was low (n = 7); five were mild-to-moderate in severity and two were severe. In patients treated with Oleogel-S10 throughout, mean (SD) BSAP changes from DBP baseline at 3, 12 and 24 months were −4.3% (8.1) (P < 0.001), −5.9% (8.6) (P < 0.001) and −3.7% (9.0) (P = 0.003), respectively. Similarly, significant changes in EBDASI skin activity score from DBP baseline were observed: −3.9 (8.3) (P < 0.001), −5.1 (8.2) (P < 0.001) and −3.0 (8.3) (P = 0.007) at 3, 12 and 24 months, respectively.

Conclusions

These data support an encouraging long-term safety profile of Oleogel-S10 and a sustained reduction in wound burden over at least 24 months of Oleogel-S10 treatment.

Graphical Abstract

Lay Summary

Epidermolysis bullosa (or ‘EB’ for short) is a group of rare genetic diseases. They are characterized by fragile skin and blistering in response to minor injuries. People with EB often develop wounds that may never fully heal, leading to significant pain and permanent scarring. Currently, there is no cure for EB. Treatment is limited to wound management. For example, ‘Oleogel-S10’ (or birch bark extract) is a gel that contains dry extract from birch bark. The ‘EASE study’ looked at the effect of Oleogel-S10 on EB wounds. The first phase of the EASE study showed that Oleogel-S10 was safe and healed wounds faster than another gel. Based on this information, Oleogel-S10 has been approved in Europe and the USA for the treatment of EB wounds.

In this paper, we report data from the follow-up phase of the EASE study on how safe and effective Oleogel-S10 gel is. A total of 205 people with different types of EB entered this phase of the study and received Oleogel-S10 for up to 2 years. Oleogel-S10 was well tolerated with few side effects. The most common side effects related to Oleogel-S10 included wound complications (for example, an increase in size, wounds re-opening and pain), itchy skin and wound infection. The faster wound healing seen with Oleogel-S10 led to important reductions in wound burden with long-term use over 2 years.

The results show that Oleogel-S10 is safe to use over a long period of time and helps wounds heal better.

What is already known about this topic?

• The rare genetic disease epidermolysis bullosa (EB), characterized by fragile skin and impaired wound healing, results in severe morbidity and mortality.

• The European Commission and the U.S. Food and Drug Administration have approved Oleogel-S10 for the treatment of EB wounds based on data from the pivotal EASE study.

• Oleogel-S10 has a reassuring safety profile and demonstrated accelerated wound healing, with significantly more patients having complete wound closure by day 45 compared with control gel.

What does this study add?

• This study reports on the second phase of EASE, the 24-month open-label phase, where 205 patients were treated with Oleogel-S10.

• The data support Oleogel-S10 as a well-tolerated topical treatment with a reassuring long-term safety profile.

• Sustained improvement in total wound burden was seen with Oleogel-S10 treatment over the additional 24 months.

Epidermolysis bullosa (EB) is a group of rare, severe genetic disorders characterized by ongoing persistent skin fragility and open wounds.^1,2 Patients living with EB have an altered wound healing profile characterized by uncontrolled inflammation and an inability to restore the epidermal barrier, resulting in a vicious cycle of delayed wound healing in the most severe forms: dystrophic EB (DEB) and junctional EB (JEB).^2–4

Impaired wound healing and uncontrolled inflammation in EB cause severe morbidity and increased mortality in people living with EB.^3,5–7 Wounds in EB may cover large areas, typically up to a third of a person’s body in severe forms.^8,9 Worsening EB severity (patient-reported) significantly correlated with larger wounds, increased use of opioid and other analgesics, infections, squamous cell carcinoma (SCC) and a poor quality of life (QoL).¹⁰ There is no cure for EB, and the management of wounds is a fundamental priority, regardless of underlying genetic abnormalities.¹¹ Patients with EB and their carers report that wound care is a major burden, and reducing the number and severity of wounds are highlighted as top priorities.⁹

Oleogel-S10 (birch bark extract/birch triterpenes) is a topical gel containing 10% triterpene birch bark extract (including betulin, lupeol, erythrodiol, betulinic acid and oleanolic acid) combined with sunflower oil.¹² Triterpenes in the Oleogel-S10 formulation have demonstrated antibacterial, antimycotic, antiviral, anti-inflammatory and antitumoral properties.^12–16 Preclinical studies have demonstrated that triterpene extract modulates the three key phases of wound healing, from the inflammatory phase to new epidermal barrier formation, ultimately speeding up the wound repair process.¹⁴ Importantly, wound healing in triterpene extract-treated keratinocytes derived from patients with recessive DEB (RDEB) and JEB has also been shown to be associated with increased inflammatory factors that promote keratinocyte migration.¹⁷ The efficacy of Oleogel-S10 in EB has been demonstrated in the largest EB trial conducted to date, the EASE study.^18,19 The EASE study consisted of a 90-day double-blind phase (DBP), followed by a 24-month open-label phase (OLP). In the DBP, EASE met its primary endpoint whereby Oleogel-S10 demonstrated a statistically significant increase in the proportion of target wounds healed within 45 days of treatment (41%) compared with vehicle control gel [29% (P = 0.01); hereafter referred to as control gel], with an even greater effect observed in patients with RDEB and those aged < 18 years.^18,20

The objective of the EASE OLP was to analyse the long-term safety and efficacy of Oleogel-S10 for the treatment of EB over 24 months.

Patients and methods

Study design

The EASE (BEB-13) study was a phase III study with a double-blind randomized controlled parallel-group design to compare the efficacy, safety and tolerability of Oleogel-S10 with a control gel in patients with DEB, JEB or Kindler EB (KEB). The DBP of EASE was completed in 2020. Details of the study design and inclusion/exclusion criteria have been published previously.^18,19 At the end of the DBP (day 90), patients in both treatment arms entered the single-arm OLP to receive Oleogel-S10 for an additional 24 months on all areas affected by partial-thickness EB wounds (Figure 1). This design was used to limit how long patients in the control gel group did not receive active treatment. Therefore, the OLP included two groups of patients: those formerly treated with Oleogel-S10 in the DBP and those formerly treated with the control gel (referred to as Oleogel-S10 and former control gel groups, respectively). As patients in the two groups received Oleogel-S10 treatment for different durations, the two groups were not compared directly in the OLP. The target wound was defined as the single investigator-selected partial thickness wound for evaluation, with up to four additional wounds selected by the investigator (additional wounds).^18,19 Further endpoints were measured across all partial-thickness EB wounds overall and these are referred to as other wounds.^18,19 The OLP was completed on 27 May 2022.

Figure 1

The EASE study design [double-blind phase (DBP) and open-label phase (OLP)].^18,19 Patients with epidermolysis bullosa simplex (EBS) were excluded following the implementation of protocol V4. Prior to this, two patients with EBS were enrolled. D, day; DEB, dystrophic epidermolysis bullosa; EB, epidermolysis bullosa; JEB, junctional epidermolysis bullosa; M, month; OLP, open-label phase. ^aNo patients with Kindler EB were enrolled.

Endpoints and patient-reported outcomes

OLP safety endpoints included the incidence, severity and relatedness of adverse events (AEs). Efficacy endpoints included body surface area percentage (BSAP) and EB Disease Activity and Scarring Index (EBDASI) skin activity score (Section 1: skin)²¹ at months 3, 12 and 24 (skin damage was not assessed in this analysis). The total EBDASI score was amended to 100 as anogenital regions and buttocks were excluded from examination. The incidence and severity of wound infections (identified based on wound infection AEs and/or use of topical and/or systemic antibiotics) were measured at OLP baseline and at month 24. Patient-reported outcomes evaluated at months 12 and 24 included QoL [EuroQoL-5D (EQ-5D)] and disease severity from clinician and patient/family perspectives (iscorEB).^22,23 Endpoints measured at month 3 only included target and additional wound closure (by clinical assessment) of wounds that were not closed at OLP baseline; procedural and background pain [Wong-Baker FACES^® pain rating scale or the Face, Legs, Activity, Cry, Consolability (FLACC) scale];²⁴ itch evaluated before wound dressing changes [Itch Man Scale (for patients aged ≥ 4 to 13 years) or Leuven Itch Scale (for patients aged ≥ 14 years)] and analysed separately by domain (frequency, duration, severity, consequences, distress, surface area); impact of wounds on sleep; number of days missed from school or work in the previous 14 days; and treatment satisfaction.

Statistical analyses

All categorical (binary and ordinal) data were summarized using frequency counts and percentages of patients. Continuous variables were summarized using number of observations, mean (SD) or median (range), unless otherwise specified. All estimations included a point estimate and the corresponding two-sided 95% confidence interval (CI). OLP efficacy and patient-reported outcome analyses were not powered for statistical significance.

Owing to a limited number of patients with BSAP and EBDASI scores obtained within the predefined visit windows during the OLP – due in part to COVID-19 ­restrictions – a post hoc analysis was conducted without visit windows to analyse the mean change from DBP baseline in BSAP and EBDASI at OLP timepoints; P-values from a two-sided paired t-test were reported.

Results

Baseline characteristics and demographics

A total of 205 patients entered the EASE OLP (Oleogel-S10 group, n = 100; former control gel group, n = 105) (Table 1); 6 patients from the control gel arm of the DBP entered the OLP early (Figure S1; see Supporting Information). Most patients (n = 141/205; 68.8%) completed the OLP, and target wound treatment adherence to Oleogel-S10 in the 24-month OLP was > 99% (Table 1), with a mean (SD) treatment duration of 584.7 (246.1) days (i.e. > 1.5 years), which was comparable regardless of former DBP treatment group. Discontinuations over the 2-year period were primarily due to withdrawal of consent, and withdrawals tended to decrease over time. Mean (SD) dose interruptions for Oleogel-S10 for any reason were low [2.7 (12.6) days] and the mean (SD) duration of dose interruption because of AEs was < 1 day [0.9 (6.1) days] (Table 1).

Table 1

Patient disposition, treatment duration and adherence in the open-label phase (OLP) of the EASE study

	Oleogel-S10 group	Former control gel group	All patients
Patient disposition
No. of patients who entered the OLP and received Oleogel-S10	100 (100)	105 (100)	205 (100)
Patients who completed the OLP	66 (66.0)	75 (71.4)	141 (68.8)
Patients who discontinued the OLP	34 (34.0)	30 (28.6)	64 (31.2)
Withdrawal of consent	18 (18.0)	15 (14.3)	33 (16.1)
AE	9 (9.0)	7 (6.7)	16 (7.8)
Other	5 (5.0)	6 (5.7)	11 (5.4)
Noncompliance	1 (1.0)	1 (1.0)	2 (1.0)
Progression of medical condition	1 (1.0)	1 (1.0)	2 (1.0)
Treatment duration and adherence
No. of patients evaluated for treatment duration	100 (100)	105 (100)	205 (100)
Treatment duration (days), mean (SD)	594.4 (235.3)	575.4 (256.8)	584.7 (246.1)
Median (range)	725.0 (14.0–841.0)	727.0 (7.0–840.0)	727.0 (7.0–841.0)
No. of patients evaluated for treatment adherencea	100 (100)	105 (100)	205 (100)
Treatment adherence (%), mean (SD)a	99.5 (2.6)	98.9 (5.0)	99.2 (4.0)
Median (range)	100.0 (82.9–100.0)	100.0 (59.9–100.0)	100.0 (59.9–100.0)
Total duration of interruptions overall (days), mean (SD)	1.4 (6.8)	3.9 (16.3)	2.7 (12.6)
Total duration of interruptions due to AEs (days), mean (SD)	0.4 (3.7)	1.4 (7.7)	0.9 (6.1)

Data are presented as n (%) unless otherwise stated. AE, adverse event. ^aRefers to adherence to the prescribed treatment regimen, specifically to the target wound.

The baseline demographics of patients who entered the OLP are shown in Table 2. The most common type of EB was DEB (n = 178/205; 86.8%), of which RDEB was the most common subtype (n = 160/205; 78.0%). Of the patients with RDEB, 55.1% (n = 113/205) had severe and 17.1% (n = 35/205) had intermediate forms, respectively. JEB was represented by 12.2% (n = 25/205) of patients, of whom 17 were classified as having intermediate disease. Only two patients (n = 2/205; 1.0%) with epidermolysis bullosa simplex (EBS) were included in the study; these patients were excluded from the subsequent analysis following a protocol amendment and no further patients with EBS were recruited. No patients with KEB were enrolled. Patients with DEB and JEB were balanced across groups. Most patients were aged < 18 years (n = 147/205; 71.7%), with the majority aged between ≥ 4 and < 12 years (n = 81/205; 39.5%) or between ≥ 12 and < 18 years (n = 50/205; 24.4%), with fewer children aged 0 to < 4 years included (n = 16/205; 7.8%) (Table 2).

Table 2

Baseline patient characteristics

	All patients at DBP baseline (n = 223)	All patients at OLP baseline (n = 205)
Age (years), mean (SD)	16.7 (14.2)	16.3 (14.1)
Median (range)	12 (0–81)	12 (0–81)
Age group (years)
0 to < 4	17 (7.6)	16 (7.8)
≥ 4 to < 12	85 (38.1)	81 (39.5)
≥ 12 to < 18	54 (24.2)	50 (24.4)
≥ 18	67 (30.0)	58 (28.3)
RDEB	175 (78.5)	160 (78.0)
Severe	124 (55.6)	113 (55.1)
Intermediate	39 (17.5)	35 (17.1)
Localized	7 (3.1)	7 (3.4)
Other	5 (2.2)	5 (2.4)
DDEB	20 (9.0)	18 (8.8)
JEB	26 (11.7)	25 (12.2)
EBS	2 (0.9)	2 (1.0)

Data are presented as n (%) unless otherwise stated. DBP, double-blind phase; DDEB, dominant dystrophic epidermolysis bullosa; EBS, epidermolysis bullosa simplex; JEB, junctional epidermolysis bullosa; OLP, open-label phase; RDEB, recessive dystrophic epidermolysis bullosa.

Safety and tolerability

AEs were experienced by 77.1% of patients (n = 158/205), and were typically mild or moderate (Table 3). Severe AEs were seen in 18.0% of patients (n = 37/205) and serious AEs in 24.4% (n = 50/205) of patients; two serious AEs were considered related to Oleogel-S10 (rash and wound infection) and one patient (rash) discontinued the study. Sixteen of 205 patients (7.8%) were withdrawn from the OLP due to AEs, including 3 who experienced treatment-related AEs (Table 3). Nine deaths were reported during the OLP (Oleogel-S10, n = 7; former ­control gel, n = 2); none were reported to be related to treatment (Table 3). Two deaths were due to sepsis, two due to pneumonia and one each due to progression of EB, cardiac failure, cardiorespiratory arrest, intestinal ischaemia and acute renal failure.

Table 3

Summary of adverse events (AEs) in the open-label phase (OLP) of EASE

	Oleogel-S10 group (n = 100)	Former control gel group (n = 105)	All patients (n = 205)
Any AEs	77 (77.0)	81 (77.1)	158 (77.1)
Any serious AEs	26 (26.0)	24 (22.9)	50 (24.4)
Any severe AEs	19 (19.0)	18 (17.1)	37 (18.0)
Any related AEs	8 (8.0)	17 (16.2)	25 (12.2)
Any serious related AEs	1 (1.0)	1 (1.0)	2 (1.0)
Any AEs leading to study withdrawal	9 (9.0)	7 (6.7)	16 (7.8)
Any related AEs leading to study withdrawal	0 (0)	3 (2.9)	3 (1.5)
Any serious AEs leading to study withdrawal	9 (9.0)	3 (2.9)	12 (5.9)
Any serious related AEs leading to study withdrawal	0 (0)	1 (1.0)	1 (0.5)
Any serious AEs leading to deatha	7 (7.0)	1 (1.0)	8 (3.9)
Any AEs due to wound complicationsb	38 (38.0)	46 (43.8)	84 (41.0)
Any AEs leading to drug withdrawal	9 (9.0)	7 (6.7)	16 (7.8)

Data are presented as n (%). ^aOne additional patient in the former control gel group died as a result of a non-treatment-emergent AE of sepsis with an onset date > 30 days after last date of study medication administration in the OLP. ^bWound complication encompassed the following events: increase in wound size vs. baseline (i.e. baseline of double-blind phase [DBP]); wound reopening; increase in wound size vs. the previous visit; other (including increase in wound burden, worsening of epidermolysis bullosa wound pain and wound odour); injury to the wound; wound worsening vs. baseline (i.e. baseline of the DBP).

AEs occurring in ≥ 2% of patients by system organ class are presented in Table 4. AEs most frequently reported in the OLP were wound complications (n = 84/205; 41.0%), anaemia (n = 37/205; 18.0%), wound infection (n = 21/205; 10.2%), staphylococcal wound infection (n = 21/205; 10.2%), pyrexia (n = 20/205; 9.8%), oesophageal stenosis (n = 19/205; 9.3%), bacterial wound infection (n = 16/205; 7.8%), pruritus (n = 14/205; 6.8%) and dysphagia (n = 13/205; 6.3%), all of which are consistent with the EB disease process (Table 4).

Table 4

Summary of adverse events (AEs) occurring in ≥ 2% of patients overall by system organ class

	Oleogel-S10 group (n = 100)	Former control gel group (n = 105)	All patients (n = 205)
Any AEs	77 (77.0)	81 (77.1)	158 (77.1)
Injury, poisoning and procedural complications	45 (45.0)	51 (48.6)	96 (46.8)
Wound complicationa	38 (38.0)	46 (43.8)	84 (41.0)
Wound secretion	2 (2.0)	2 (1.9)	4 (2.0)
Infections and infestations	36 (36.0)	44 (41.9)	80 (39.0)
Wound infection	6 (6.0)	15 (14.3)	21 (10.2)
Staphylococcal wound infection	9 (9.0)	12 (11.4)	21 (10.2)
Bacterial wound infection	7 (7.0)	9 (8.6)	16 (7.8)
Skin infection	3 (3.0)	4 (3.8)	7 (3.4)
Influenza	1 (1.0)	4 (3.8)	5 (2.4)
URTI	2 (2.0)	3 (2.9)	5 (2.4)
Conjunctivitis	3 (3.0)	1 (1.0)	4 (2.0)
Nasopharyngitis	3 (3.0)	1 (1.0)	4 (2.0)
Otitis externa	2 (2.0)	2 (1.9)	4 (2.0)
Pneumonia	3 (3.0)	1 (1.0)	4 (2.0)
Pseudomonas wound infection	3 (3.0)	1 (1.0)	4 (2.0)
Gastrointestinal disorders	22 (22.0)	29 (27.6)	51 (24.9)
Oesophageal stenosis	8 (8.0)	11 (10.5)	19 (9.3)
Dysphagia	6 (6.0)	7 (6.7)	13 (6.3)
Diarrhoea	3 (3.0)	6 (5.7)	9 (4.4)
Toothache	4 (4.0)	1 (1.0)	5 (2.4)
Vomiting	3 (3.0)	2 (1.9)	5 (2.4)
Blood and lymphatic system disorders	16 (16.0)	24 (22.9)	40 (19.5)
Anaemia	16 (16.0)	21 (20.0)	37 (18.0)
General disorders and administration site conditions	17 (17.0)	16 (15.2)	33 (16.1)
Pyrexia	10 (10.0)	10 (9.5)	20 (9.8)
Asthenia	2 (2.0)	2 (1.9)	4 (2.0)
Skin and subcutaneous tissue disorders	11 (11.0)	20 (19.0)	31 (15.1)
Pruritus	5 (5.0)	9 (8.6)	14 (6.8)
Blister	0	4 (3.8)	4 (2.0)
Metabolism and nutrition disorders	11 (11.0)	14 (13.3)	25 (12.2)
Hypoalbuminaemia	5 (5.0)	5 (4.8)	10 (4.9)
Vitamin D deficiency	4 (4.0)	6 (5.7)	10 (4.9)
Malnutrition	2 (2.0)	3 (2.9)	5 (2.4)
Eye disorders	10 (10.0)	7 (6.7)	17 (8.3)
Ulcerative keratitis	3 (3.0)	2 (1.9)	5 (2.4)
Congenital, familial and genetic disorders	4 (4.0)	4 (3.8)	8 (3.9)
Syndactyly	2 (2.0)	2 (1.9)	4 (2.0)
Hepatobiliary disorders	4 (4.0)	2 (1.9)	6 (2.9)
Abnormal hepatic function	3 (3.0)	2 (1.9)	5 (2.4)

URTI, upper respiratory tract infection. ^aWound complication encompassed the following events: increase in wound size vs. baseline [i.e. baseline of the double-blind phase (DBP)]; wound reopening; increase in wound size vs. the previous visit; other (including increase in wound burden, worsening of epidermolysis bullosa wound pain, and wound odour); injury to the wound; wound worsening vs. baseline (i.e. baseline of the DBP).

Wound infections

Target wound infections, requiring topical or systemic antibiotics, occurred in few patients (Oleogel-S10 group, n = 4; former control gel group, n = 3) (Figure S2; see Supporting Information). The severity of target wound infections was mild (n = 2) and severe (n = 2) in patients who received Oleogel-S10, and all were moderate in patients in the former control gel group (n = 3). The incidence of additional and other wound infections was low through to OLP month 24 and the severity was predominantly mild-to-moderate in intensity.

Wound burden

Mean changes in BSAP from DBP baseline were significant at all OLP timepoints for patients who received Oleogel-S10 throughout the study. Mean (SD) changes from DBP baseline to DBP day 90/OLP baseline, and OLP months 3, 12 and 24 were −4.2% (6.9) (P < 0.001); −4.3% (8.1) (P < 0.001); −5.9% (8.6) (P < 0.001); and −3.7% (9.0) (P = 0.003), respectively, for Oleogel-S10 treated patients (Table 5); for former control patients these values were −2.6% (9.0) (P = 0.005); −1.7% (9.8) (P > 0.05); −3.7% (8.0) (P < 0.001) and −3.7% (7.2) (P < 0.001), respectively (Table 5). Mean BSAP values at each timepoint throughout the study for all treatment arms are shown in Figure 2.

Figure 2

Mean body surface area percentage (BSAP) from baseline of the double-blind phase (DBP) to month 24 in the open-label phase (OLP). Data shown are the absolute mean BSAP values at each timepoint. Assessments recorded within OLP visit windows were lower than expected largely due to the COVID-19 pandemic. Therefore, a post hoc analysis was produced without visit windows in the OLP. Day 90/OLP baseline: the last assessment before the first open-label treatment and at day 90 (eight patients did not have a day-90 visit). Refer to Table 4 for changes from DBP baseline data and statistical analysis. D, day; M, month.

Table 5

Mean change in body surface area percentage from baseline in the double-blind phase (DBP) of the study

	DBP baseline	Day 90/OLP baseline	OLP month 3	OLP month 12	OLP month 24
All patients
n	222	196	171	139	128
Mean (SD) change from baseline	–	−3.4 (8.0)	−3.0 (9.1)	−4.8 (8.3)	−3.7 (8.1)
P-value	–	< 0.001	< 0.001	< 0.001	< 0.001
Oleogel-S10 group
n	109	100	85	67	59
Mean (SD) change from baseline	–	−4.2 (6.9)	−4.3 (8.1)	−5.9 (8.6)	−3.7 (9.0)
P-value	–	< 0.001	< 0.001	< 0.001	0.003
Former control gel group
n	113	96	86	72	69
Mean (SD) change from baseline	–	−2.6 (9.0)	−1.7 (9.8)	−3.7 (8.0)	−3.7 (7.2)
P-value	–	0.005	> 0.05	< 0.001	< 0.001

Parameter and model estimates based on Ancova for the change from baseline, with treatment group, epidermolysis bullosa subtype and target wound size class as fixed effects, and corresponding body surface area percentage score at baseline as a covariate. Assessments recorded within open-label phase (OLP) visit windows were lower than expected largely due to the COVID-19 pandemic. Therefore, a post hoc analysis was produced without visit windows in the OLP. Day 90/OLP baseline: the last assessment before the first open-label treatment and at day 90 (eight patients did not have a day-90 visit).

Mean (SD) EBDASI changes from DBP baseline are shown in Table 6. Statistically significant changes were seen at all OLP timepoints for patients who received Oleogel-S10 throughout the study. Mean (SD) changes for patients who received Oleogel-S10 from DBP baseline to DBP day 90/OLP baseline, and OLP months 3, 12 and 24 were −2.8 (7.5) (P < 0.001); −3.9 (8.3) (P < 0.001); −5.1 (8.2) (P < 0.001); and −3.0 (8.3) (P = 0.007), respectively. Mean (SD) changes for patients in the former control gel group from DBP baseline to DBP day 90/OLP baseline, and OLP months 3, 12 and 24 were −2.9 (7.5) (P < 0.001); −1.9 (9.4) (P > 0.05); −5.0. (8.3) (P < 0.001); and −2.5 (8.7) (P = 0.02), respectively. All available mean EBDASI scores at each timepoint are shown in Figure 3.

Figure 3

Mean Epidermolysis Bullosa Disease Activity and Scarring Index (EBDASI) skin activity scores from baseline of the double-blind phase (DBP) to month 24 of the open-label phase (OLP). Data shown are the absolute mean EBDASI skin activity score values at each timepoint. Assessments recorded within OLP visit windows were lower than expected largely due to the COVID-19 pandemic. Therefore, a post hoc analysis was produced without visit windows in the OLP. Day 90/OLP baseline: the last assessment before the first open-label treatment and at day 90 (eight patients did not have a day-90 visit). Refer to Table 5 for changes from DBP baseline data and statistical analysis. D, day; M, month.

Table 6

Mean change in Epidermolysis Bullosa Disease Activity and Scarring Index (EBDASI) skin activity score from baseline in the double-blind phase (DBP) of the study

	DBP baseline	Day 90/OLP baseline	OLP month 3	OLP month 12	OLP month 24
All patients
n	221	195	173	138	128
Mean (SD) change from baseline	–	−2.8 (7.5)	−2.8 (8.9)	−5.0 (8.2)	−2.7 (8.5)
P-value	–	< 0.001	< 0.001	< 0.001	< 0.001
Oleogel-S10 group
n	108	99	85	66	59
Mean (SD) change from baseline	–	−2.8 (7.5)	−3.9 (8.3)	−5.1 (8.2)	−3.0 (8.3)
P-value	–	< 0.001	< 0.001	< 0.001	0.007
Former control gel group
n	113	96	88	72	69
Mean (SD) change from baseline	–	−2.9 (7.5)	−1.9 (9.4)	−5.0 (8.3)	−2.5 (8.7)
P-value	–	< 0.001	> 0.05	< 0.001	0.02

Parameter and model estimates based on Ancova for the change from baseline, with treatment group, epidermolysis bullosa subtype and target wound size class as fixed effects and corresponding EBDASI skin activity score at baseline as covariate. Assessments recorded within open-label phase (OLP) visit windows were lower than expected largely due to the COVID-19 pandemic. Therefore, a post hoc analysis was produced without visit windows in the OLP. Day 90/OLP baseline: the last assessment before the first open-label treatment and at day 90 (eight patients did not have a day-90 visit).

No clear trends were seen in either treatment group with regard to disease severity (iscorEB) or QoL (EQ-5D) at DBP day 90/OLP baseline, month 12 or month 24 (Figures S3, S4; see Supporting Information).

Endpoints and patient-reported outcomes evaluated at open-label phase month 3

First complete wound closure was evaluated at month 3 in the OLP for patients with target wounds that did not achieve complete closure in the DBP. Six patients experienced first complete wound closure at this timepoint: three in the Oleogel-S10 group and three in the former control gel group. Images of wound closure in patients in the former control gel group treated with Oleogel-S10 are shown in Figure S5 (see Supporting Information).

Procedural pain (pain associated with dressing changes) and background pain data are reported in Figures S6 and S7 (see Supporting Information). Overall, the effects on procedural and background pain achieved in the DBP were generally maintained at month 3 of the OLP for the Oleogel-S10 group in patients aged ≥ 4 years. Former control patients aged ≥ 4 years demonstrated improvements in procedural pain at 3 months after switching to Oleogel-S10 [from 2.9 to 2.4 (FACES); Figure S6]. No clear trends were seen in younger patients (< 4 years, n = 17), but patient numbers were low (Figures S6, S7).

Using the Itch Man Scale, patients in both groups maintained reductions in itching below the DBP baseline value from OLP baseline to month 3 (Figure S8; see Supporting Information). Figure S9 (see Supporting Information) shows data from the Leuven Itch Scale for patients aged ≥ 14 years.

Slight improvements in the impact of wounds on sleep and the number of days missed from school or work were observed at OLP month 3 vs. DBP baseline for both the Oleogel-S10 and former control gel groups, with the majority of patients (n = 52/62; 83.9%) satisfied with their treatment at this timepoint (Figures S10–S12; see Supporting Information). Mean changes from baseline in all other endpoints are shown in Table S1 (see Supporting Information).

Discussion

The DBP of the EASE study demonstrated that Oleogel-S10 topical treatment accelerated the healing of partial-thickness wounds in EB.¹⁸ The clinical significance of accelerated wound healing was demonstrated by several of the secondary endpoints, including reductions in pain, frequency of dressing changes and total body wound burden, together with a low incidence of wound infection over the 90-day observation period.¹⁸ Here, we show that these improvements were sustained with long-term treatment in the OLP. Furthermore, 24–27 months of treatment with Oleogel-S10 reduced the extent of open wounds, as observed with BSAP and EBDASI skin activity score.

Long-term treatment with Oleogel-S10 was well tolerated, with no new safety signals identified in the OLP. Most of the AEs reported were associated with the patients’ underlying EB disease. Importantly, AEs related to treatment were lower than that observed in the DBP,¹⁸ indicating that longer exposure to Oleogel-S10 does not raise any safety concerns, and may reflect a long-term habituation to treatment in the improvement of wounds.

Due to the underlying skin fragility in EB, infections are common, particularly in intermediate and severe forms where wounding can be extensive and chronic.²⁵ Wound infections are associated with impaired wound healing and increase the risk of fibrosis and progression to SCC.²⁶ In the OLP, the incidence of target wound infections was low and typically mild, consistent with the DBP.¹⁸ This is encouraging as measures or treatments that promote healing are fundamental to the care of EB wounds.

Total wound burden is a clinically important parameter, and a reduction affects the morbidity experienced by individuals with EB.⁹ In this study it was encouraging to see that – given the nature of this chronic and escalating disorder in which patients’ fragile skin wounds cycle regularly – the reductions achieved in the DBP, in both BSAP and EBDASI skin activity score, were maintained for up to 24 months.

EBDASI skin activity assesses blistering, erosions and crusting. Minimal clinically important differences for improvement and deterioration in EB have previously been calculated for EBDASI activity scores.²¹ A 3-point decrease in the complete EBDASI activity score (total score = 276) was determined to be 75% sensitive and 73% specific for improvement. While direct comparisons cannot be made with this study because the full score was not employed, a reduction of 3 points at OLP month 24 with Oleogel-S10 is proportionately likely to reach the threshold for an important clinical difference.²¹ It will be equally valuable to follow the impact of Oleogel-S10 on the disease progression of patients who start treatment early in life and in the different subtypes. The encouraging observations of improving wound burden suggest that treatment with Oleogel-S10 may have long-term benefit. Real-world data are emerging and a recent study by Torres Pradilla et al. in 13 patients with DEB treated with Oleogel-S10 over 2 years reported similar reductions in BSAP (> 50%).²⁷

Pain experienced during dressing changes can be extremely distressing and time-consuming for patients with EB, with opiates often necessary for analgesia.^28–30 Overall, the effects on procedural pain achieved in the DBP were generally maintained at month 3 of the OLP for Oleogel-S10 patients aged ≥ 4 years. Complete resolution of pain, particularly when there is a chronic requirement for dressings, is unrealistic; however, maintaining a lower severity of pain as demonstrated here is encouraging.

Patients experiencing severe itch also present a major unmet need for effective treatment.³¹ In the DBP, Oleogel-S10 and the control gel improved itch and these improvements were sustained up to OLP month 3. However, the complexity of itch means that the data obtained are difficult to interpret and in younger children (those aged 0 to < 4 years) further complicated by small patient numbers.

Limitations of the EASE trial were largely related to disruptions caused by the COVID-19 pandemic, as well as the unique challenges of clinical trials in rare disease in which relatively small patient populations are inevitable. Patient-reported outcomes of iscorEB and EQ-5D were added into the protocol approximately 2.5 years after study initiation, resulting in slow incorporation and implementation. Few patients had baseline assessments, affecting the interpretation of results.

Research in EB is ongoing, and a number of therapeutic interventions are being explored.³² Recently, a topical herpes simplex virus 1 vector-based gene therapy to treat wounds in patients with DEB was approved by the U.S. Food and Drug Administration. Published results for this redosable, weekly application gene therapy are also encouraging.³³ Limitations to the maximum dose mean that for many patients the wounds treated only represent a portion of the total body surface area affected.^34,35 Data reported here from EASE, where all EB wounds were treated with Oleogel-S10 in patients with DEB and also in patients with JEB, showed that continued use for at least 24 months resulted in an overall reduction in total body wound burden. These data are clinically meaningful and important to consider, given the significant impact wound burden and wound care are reported to have on daily activities for patients, their caregivers and their families.^4,8

EASE – the largest study in EB – has demonstrated a favourable safety profile for Oleogel-S10 with long-term treatment. Furthermore, the long-term data from the OLP have demonstrated a sustained reduction in wound burden over 24 months. This is encouraging given the escalating vicious cycle of wound propagation, healing and reopening leading to fibrosis and SCC in the natural course of this chronic disease.

The potential for Oleogel-S10 to provide sustained wound healing from the age of 6 months may offer a change in wound trajectory for patients with DEB and JEB.

Acknowledgements

The authors thank the patients, their caregivers and the investigators who participated in the trial; Sandra Löwe, Janet Boylan, Miriam Barry and Simone Klusch from Chiesi for their assistance with the trial; and Susan Allen PhD, professional medical writer at Brandfish Ltd, for editorial and technical support in the preparation of the manuscript, funded by Chiesi. The authors also acknowledge the contributions of Maryanne Donovan from Chiesi for support with data analysis/interpretation and editorial assistance. Amryt Research Ltd funded EASE as a development clinical trial of Oleogel-S10, and was therefore responsible for study design, data collection, analysis and regulatory compliance. Amryt Pharmaceuticals DAC is a wholly owned subsidiary of CHIESI S.p.A. The Marketing Authorization holder for Filsuvez (birch bark extract) in the EU and UK is CHIESI S.p.A.

Funding sources

EASE was funded by Amryt Research Ltd, Dublin, Ireland (acquired by Chiesi Farmaceutici S.p.A. April 2023). Trial registration: NCT03068780, EudraCT 2016-002066-32; registered 3 March 2017; https://clinicaltrials.gov/study/NCT03068780 and https://www.clinicaltrialsregister.eu/ctr-search/trial/2016-002066-32/GB.

Data availability

Selected subgroup-level data, figures and listings can be made available on request from the corresponding author ([email protected]). These will be available from 1 year after completion of the trial and in perpetuity. The study protocol, statistical analysis plan and consent form are included in the supporting information of the original study data published in 2023 by Kern et al.^18,19

Ethics statement

EASE was conducted in accordance with the Declaration of Helsinki. The research protocol was approved by relevant institutional review boards and ethics committees for each study site.

Patient consent

Written informed consent was received for all patients.

Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher's website.

Author notes

The list of EASE investigators is provided in Appendix S1 (see Supporting Information).

Conflicts of interest: D.F.M. reports personal fees from Chiesi during the conduct of the study and other fees from AbbVie, Amicus, Amgen, Amryt Pharma, Arcutis, Arenan, Argenix, AstraZeneca, BMS, Castle Creek, Cutanea, CSL, Dermira, Eli Lilly, Evelo, Galderma, Genentech, Incyte, JCB Biopharma, Kiniksa, Novartis, Pfizer, Phoenix Tissue Repair, Pierre Fabre, Principia, ProQR/Wings, Regeneron, Rheacell, Sanofi, Scioderm, Shire, Takeda and UCB. D.F.M. is a co-patent holder with Stanford University for the use of topical sirolimus for epidermolysis bullosa simplex. J.S.K. reports grants from Chiesi during the conduct of the study and other grants from AbbVie, Amicus, Amgen, Amryt Pharma, Arcutis, Arena, Argenix, Astra Zeneca, BMS, Boehringer Ingelheim, Cutanea, CSL, DFG/Fresenius, Dompé, Eli Lilly, Evolo, Exopharm, Galderma, Genentech, Incyte, InflaRx, Kiniksa, Merck, Mitsubishi, Novartis, Pfizer, Principia, Regeneron, Sanofi, Scioderm, Shire, Takeda and UCB outside the submitted work. C.B. and M.F.F. report personal fees from Chiesi during the conduct of the study. A.L.B. reports grants and other support from Chiesi during the conduct of the study; grants and other support from Amicus/Scioderm, grants and other support from Castle Creek, and grants from EB Research Partnership and EM Medical Research Foundation, Fibrocell, Phoenix Tissue Repair, ProQR/Wings and Rheacell outside the submitted work. E.S. reports personal fees from Chiesi during the conduct of the study and personal fees from Bayer, BiomX, Kamari Pharma, Pierre Fabre and Sol-Gel Technologies outside the submitted work. D.K. reports receipt of grants/research support from Rheacell; personal fees from Amryt Pharma, Chiesi, Fibrx Derm, Rheacell and Xinnate; and other support from DEBRA International, the EB Research Partnership, the Fritz-Thyssen Foundation and the German Research Foundation. M.T.-P. reports grants/fees from AbbVie, Amryt Pharma, Avicanna, Chiesi, Eucerin, Galderma, Gencell Pharma, Novartis, Pfizer, Pierre Fabre, Roemmers, Sanofi and Uno Healthcare. T.C. and L.M. were employees of Amryt Research Ltd during the EASE study. C.D. was a paid consultant for Amryt Research Ltd.