A systematic literature review of the disease burden in patients with recessive dystrophic epidermolysis bullosa

Background/objective Recessive dystrophic epidermolysis bullosa (RDEB) is a genetic collagen disorder characterized by skin fragility leading to blistering, wounds, and scarring. There are currently no approved curative therapies. The objective of this manuscript is to provide a comprehensive literature review of the disease burden caused by RDEB. Methods A systematic literature review was conducted in MEDLINE and Embase in accordance with PRISMA guidelines. Observational and interventional studies on the economic, clinical, or humanistic burden of RDEB were included. Results Sixty-five studies were included in the review. Patients had considerable wound burden, with 60% reporting wounds covering more than 30% of their body. Increases in pain and itch were seen with larger wound size. Chronic wounds were larger and more painful than recurrent wounds. Commonly reported symptoms and complications included lesions and blistering, anemia, nail dystrophy and loss, milia, infections, musculoskeletal contractures, strictures or stenoses, constipation, malnutrition/nutritional problems, pseudosyndactyly, ocular manifestations, and dental caries. Many patients underwent esophageal dilation (29–74%; median dilations, 2–6) and gastrostomy tube placement (8–58%). In the severely affected population, risk of squamous cell carcinoma (SCC) was 76% and mortality from SCC reached 84% by age 40. Patients with RDEB experienced worsened quality of life (QOL), decreased functioning and social activities, and increased pain and itch when compared to other EB subtypes, other skin diseases, and the general population. Families of patients reported experiencing high rates of burden including financial burden (50–54%) and negative impact on private life (79%). Direct medical costs were high, though reported in few studies; annual payer-borne total medical costs in Ireland were $84,534 and annual patient-borne medical costs in Korea were $7392. Estimated annual US costs for wound dressings ranged from $4000 to $245,000. Patients spent considerable time changing dressings: often daily (13–54% of patients) with up to three hours per change (15–40%). Conclusion Patients with RDEB and their families/caregivers experience significant economic, humanistic, and clinical burden. Further research is needed to better understand the costs of disease, how the burden of disease changes over the patient lifetime and to better characterize QOL impact, and how RDEB compares with other chronic, debilitating disorders. Supplementary Information The online version contains supplementary material available at 10.1186/s13023-021-01811-7.


Introduction
Recessive dystrophic epidermolysis bullosa (RDEB) is a rare, severe form of dystrophic epidermolysis bullosa (DEB), a genetic collagen disorder characterized by skin fragility and scarring of the skin from birth onwards. Patients with this disorder inherit mutations in both alleles of COL7A1, the gene which produces type VII collagen (C7). This leads to absences or irregularities in C7 and alterations in the character and number of anchoring fibrils, which secure the skin's dermal layer to the epidermal layer [1]. Due to these mutations, patients with RDEB can experience blistering at the dermal layer with only minimal trauma.
An analysis of the United States (US) National Epidermolysis Bullosa (EB) Registry, funded and operated from 1986 to 2002, reported RDEB incidence of 3.05 cases per one million live births and prevalence of 1.35 cases per one million live births [2]. However, a more recent genotypic modeling of publicly available whole-exome and whole-genome sequencing estimated an incidence of 95 cases per one million births. This suggests that the National EB Registry estimates may be significantly understated, potentially due to underestimation of less severe cases of RDEB, likely mis-diagnosed as EB Simplex or de novo variants of Dominant Dystrophic EB (DDEB) [3]. This underestimation may also result in overestimation of systemic and severe manifestations.
RDEB is typically diagnosed clinically and often confirmed through assessment of immunofluorescence or electron microscopy on skin biopsy or by genetic testing [4]. RDEB is divided into several subtypes: severe (characterized by absent or markedly reduced C7), intermediate (characterized by reduced C7), and other, rarer, subtypes including inversa, localized, or pruriginosa [5].
Cutaneous signs of the disease include blistering and wounding in response to mechanical traumas, milia, atrophic scarring, dystrophic or absent nails, and alopecia. Pruritus is also a frequent complaint. Chronic wounding and fibrosis is generally believed to favor the frequent development of aggressive squamous cell carcinoma (SCC), which represents a leading cause of premature death in patients with severe and intermediate RDEB [4,6]. 1 Extracutaneous manifestations include anemia, growth retardation, dental caries, pseudosyndactyly, esophageal strictures, malnutrition, and ocular involvement [4].
Currently, there are no approved disease-modifying therapies for RDEB. Treatment of the disease is limited to management of symptoms and secondary complications, such as wound care, prevention of trauma, treatment of infections, pain and itch management, strategic wrapping of the hands and feet to prevent pseudosyndactyly, and early detection and treatment of SCC [1,7,8]. Gastrointestinal manifestations of the disease are managed through proactive nutritional support including gastrostomy feeding, esophageal dilation, and treatment of anemia [1,8]. Other areas of disease management include physical therapy and rehabilitation, psychosocial and group support, and extra schooling accommodations [1,7,9].
Due to the high unmet need for corrective treatments, the Food and Drug Administration released guidance for industry on the development of drugs for treatment of cutaneous manifestations of EB in June 2019, identifying drug development and trial design, population, and evaluation as issues specific to patients [10]. An increasing number of clinical trials are being conducted in cellbased therapies, gene and molecular therapies, protein replacement therapies, exon skipping molecular therapies, and drug-mediated premature termination codon read-throughs targeted to manage and treat RDEB [11]. Gene therapies, which involve the transfer of functional COL7A1 gene to the patient with RDEB, appear to be promising potential treatments, likely available in the near future [12][13][14].
Many narrative and expert reviews provide a commentary on RDEB; however, no reviews use a systematic method to evaluate the literature regarding burden of disease in this patient population. This paper aims to systematically review and synthesize the data regarding the clinical, humanistic, and economic burden of RDEB. Abeona Therapeutics, a company developing EB-101, investigational autologous COL7A1 gene-corrected keratinocytes sheets for the treatment of large and chronic RDEB wounds, initiated this review to fully characterize disease burden, and its employees (J.G and D.R.) co-authored the paper. Pharmerit conducted the literature review and generated the first draft. Tang et al. Orphanet J Rare Dis (2021) 16:175

Methods
The systematic literature review was conducted in accordance with Preferred Reporting Items for Systematic Review and Meta-Analyses guidelines [15]. The literature review was developed based on a predefined search and selection protocol. Search terms are provided in Additional file 1: Table S1. The search focused on relevant studies published as journal articles or conference abstracts through April 2, 2020 in MED-LINE and Embase (searched via ProQuest), written in English.
Studies comprised of ≥ 80% patients with RDEB from any country were included. Observational studies (retrospective or prospective) including cohort, case-control, cross sectional studies, and case series and trials were included. Case reports (sample size < five patients), notes, editorials, and commentaries were excluded. Systematic reviews were included for the purpose of identification of primary studies. Animal and preclinical studies were excluded.
An experienced reviewer (EL, ST) independently screened all titles and abstracts resulting from the search methodology to identify articles for full-text review.
Citations selected for full-text review were screened by the same reviewer for potential inclusion into the data extraction file and report. A second reviewer (EL, ST) verified the results of the title/abstract screening and fulltext review. A standardized table was used to extract and record relevant data from selected publications.

Summary of included studies
A total of 740 citations were identified, of which 229 full texts were screened and 65 were included within this review (56 manuscripts; nine conference abstracts or presentations) ( Fig. 1; Additional file 1: Table S2). The majority of studies included all ages (n = 36 studies), or children (n = 20 studies); one study was in an adult-only cohort, and eight did not report age. The average age ranged between three years to 30 years of age at the time of the study. Most patients had severe or intermediate subtypes; the proportion of patients with severe RDEB included in the identified studies ranged from 26 to 100%.

Clinical burden Symptoms
Forty-one studies reported on symptom prevalence (Table 1) and/or burden . Seven studies reported on the cumulative risk of symptoms at different ages including data from the US National EB Registry and Australasian EB Registry (Table 2) [6,[25][26][27][28][29][30].

Wound burden
A US single-center study of 40 RDEB patients reported the clinical differences between recurrent or chronic open wounds [56]. Recurrent wounds healed but blistered again easily while chronic wounds remained open for 12 weeks or longer. Chronic wounds were significantly larger than recurrent wounds (66.3 cm 2 vs. 44.7 cm 2 ; p < 0.01) and more painful (4.31 of 10 points vs. 3.59; p = 0.05). Larger wound size was correlated with increased pain and itch among both chronic and recurrent wounds.

Strictures and stenoses
In six studies, the proportion of patients with esophageal strictures ranged from 51 to 86% [30,31,[52][53][54][55]. A Mexican analysis of 14 patients reported a median of one stenosis per year, with 74% (14/19 stenoses) in the proximal region [32]. A single-center UK study of 57 patients with stenoses reported a median of two stricture sites at esophageal dilation, the majority of which were located in the cervical or thoracic region (percentage not given) [41]. A single-center study in Croatia reported that in six patients, each with a stenosis, 83% (5/6 stenoses) were located in the upper third of the esophagus, with the remaining stenosis in the lower third [40]. Strictures and stenoses of the anus, pulmonary artery, nostril, urethra, anterior commissure, pylorus, larynx, and rectum were reported in 0.2% to 15% of patients with RDEB in five studies (Table 1) [22,27,29,30,52]. In two analyses of the US National EB Registry (N = 422-425), the lifetime cumulative risk of esophageal strictures was much higher than risk of laryngeal stenoses and strictures (Table 2) [29,30]. Risk was higher in the severe subtype than the intermediate or inversa subtypes.

Malnutrition/failure to thrive
The proportion of patients with malnutrition/failure to thrive ranged from 25 to 72% while negative height and height velocity standard deviation scores were 94% and 89% (Table 1) [20,31,44,53,54]. In a UK analysis of 57 patients, etiologies of failure to thrive included reduction in dietary intake (due to dental involvement, pain from oral lesions), esophageal strictures, and heightened nutritional requirements secondary to extensive skin involvement [31]. A retrospective study of 157 German patients reported that approximately 50% of children with RDEB (exact numbers not reported) showed wasting (defined as weight below the third percentile) after the age of eight, and approximately 50% of children showed stunting (defined as height below the third percentile) after the age of ten; body mass index (BMI) in patients with RDEB fell in the underweight category (< 18.5 kg/m 2 ), with a median BMI of 13.8 kg/m 2 in men and 15.7 kg/ m 2 in women 20 years of age [55]. The nutritional characteristics of 12 patients with RDEB undergoing gastrostomy were described in a French, single-center study. Within the cohort, the mean estimated oral energy intake as a percentage of the recommended dietary allowance (adjusted for age and sex) was 56% (SD, 18) at time of gastrostomy feeding onset [20].

Pseudosyndactyly
The proportion of RDEB patients experiencing pseudosyndactyly ranged from 13 to 71% (Table 1), reported in three studies [16,43,52,54]. The US National EB Registry reported the lifetime cumulative risk of pseudosyndactyly of the hands, which was highest in patients with the severe subtype (Table 2) [28].

Congestive heart failure and cardiomyopathy
An international multicenter case review (N = 13) reported a mean age of cardiomyopathy diagnosis of 12.6 years [39]. Notably, six (46%) of these patients were deceased at the time of publication. Reported in 407 patients with RDEB included in the US National EB Registry, the cumulative risk of dilated cardiomyopathy and congestive heart failure tended to be low over the patient lifetime, though risk in patients with severe subtype increased with age (Table 2) [25].

Premature mortality
In a UK registry analysis of 79 patients with RDEB aged 16 and younger, nine children (11%) died. Causes of death included sepsis and organ failure (n = 5), failure to thrive (n = 2), bowel perforation (n = 1), and preconditioning for bone marrow transplantation (N = 1) [17]. The US National EB Registry analyzed the cumulative risk of childhood death from pneumonia, sepsis, respiratory failure, renal failure, and failure to thrive as in Table 2 [63].

Squamous cell carcinoma
Database analyses of the cumulative risk of developing SCC showed low risk during childhood and increases with age, with a high risk for development of and mortality from SCC by 40 years (Table 2) [6,17,26,53,63]. In data from the US National EB Registry, SCCs tended to develop in chronic skin wounds (77.8% of SCC sites in 59 RDEB patients with SCC) [26]. The median number of SCC sites per patient was 3-3.5 (range: 1 to 40 sites). Median age at diagnosis, provided in a singlecenter study of 14 RDEB patients with SCC in Spain, was 24 years [64].
In an Australian registry analysis of patients with RDEB diagnosed with SCC (N = 16), the median number of SCC sites per patient was 7 and ranged from 1 to 56 sites, with a median age of 29.5 years at first SCC [6]. The majority of SCCs (95%; sample size not reported) developed on the extremities; 70% of those developed on the hands or feet. The site of SCC tended to be in areas of chronic ulcers and non-healing wounds, though percentages were not provided. Almost 70% (11/16) of patients diagnosed with SCC experienced metastasis to regional lymph nodes (100% of patients with metastasis, 10/10 patients [data not available in one patient]), lungs (80%, 8/10), vertebrae (30%, 3/10), and liver, adrenal gland, and muscle (10%, 1/10) over the patient lifetime. Almost half of patients included in the study (7/16, 44%) underwent therapeutic amputation in their lifetime; median age at first amputation was 29 years.
In a single-center study in Spain, 35% of patients (8/23) underwent amputation due to SCC [64]. In a study of 59 patients with RDEB and SCC, surgical amputation of at least one limb was performed in 21% of patients with the intermediate subtype to 42% with the severe subtype of patients (sample size of subtypes not reported) [26]. Amputation of the leg was most common (intermediate subtype, 67%; severe subtype, 29%), followed by arm (

Procedures
Esophageal dilation and gastrostomy tube placement were the procedures most frequently reported upon in the literature. Ten studies reported on the use of esophageal dilations in patients with RDEB [17, 20, 23, 30, 40-42, 55, 65, 66], and ten studies reported on the use of gastrostomy tubes (Table 4) [17,23,30,31,39,49,54,55,65,66]. One single-center case review in the UK reported on treatment burden and satisfaction associated with gastrostomy tubes in RDEB (N = 57) [65]. Two-thirds of children with RDEB (10/15) and all of their parents (15/15) reported a satisfaction level with the gastrostomy tube of at least seven out of ten (indicating extreme satisfaction) over a median 8.9 years since placement. Almost half of patients (7/15, 47%) reported no gastrostomy site infections in the previous year; one-third reported fewer than two infections in the previous year (5/15, 33%). The remaining patients reported either 2-4 infections (n = 2) or constant infections in the previous year (n = 1).

Pseudosyndactyly release
Three studies reported on pseudosyndactyly release of the hands or feet in patients with RDEB [23,28,66]. Within a US and Canadian cohort of 238 RDEB patients, 62 (22%) underwent hand surgery. The median age for their first hand surgery was 8.1 years (IQR, 5.5-12.1 years; range, 3-25 years) [23]. Among 414 patients in the US National EB Registry, 151 (37%) underwent mitten repair of the hands and 11 (3%) underwent mitten repair of the feet [28]. The median number of hand surgeries performed was 3.0 (range: 1-22 surgeries). Data on the median number of foot surgeries was not available. A retrospective analysis of 25 children with RDEB reported that 27% of patients (sample size not provided) underwent pseudosyndactyly release with or without skin graft [66].
Patients with RDEB experienced significant impairment in overall quality of life (QOL) across multiple PROMs and domains (Table 5). RDEB patients had lower QOL than patients with other EB subtypes and patients with other skin diseases, especially compared with patients with more common diseases such as atopic dermatitis and psoriasis (Fig. 2).

Functioning and social activities
Patients with RDEB experienced limitations in functioning and social activities; many patients with RDEB required assistance or are unable to complete activities of daily living (  [53].

Generic instruments
Instruments measuring physical functioning ABILIHAND Individual item scores range from 0 (impossible) to 2 (easy) Overall hand function score ranges from 0 to 42 Higher score indicates better hand function   When measuring intellectual development, all children with RDEB were labeled normal (18%) or above normal (82%) intellectual development. All children with EBS, JEB and DDEB were labeled as above normal; significance not reported When measuring affective development, a higher proportion of children with RDEB were labeled with immaturity (46%) than maturity (36%) or forced growth (18%). All patients with EBS and DEDB and 80% of patients with JEB were labeled as mature; significance not reported When measuring scholastic/working efficiency, a higher proportion of children with RDEB were labeled as adequate or high (73%) than inadequate (27%). All patients with EBS, JEB and DDEB were labeled as adequate or high; significance not reported When measuring drive display, a higher proportion of children with RDEB were labeled as adequate (64%) than coarctate (18%) or excessive (18%). All patients with DDEB, 80% of patients with JEB, and 33% of patients with EBS were labeled as adequate; significance not reported When measuring psychosocial development, a higher proportion of children with RDED were labeled with accommodating adjustment (46%) or assimilative adjustment (36%) than reported maladjustment (18%    Adapted from: Cestari [69], Eisman [70], Frew [71], Jeon [59]. DDEB, dominant dystrophic epidermolysis bullosa; EB, epidermolysis bullosa; EBS, epidermolysis bullosa simplex; JEB, junctional epidermolysis bullosa; QOL, quality of life; QOLEB, Quality of Life in Epidermolysis Bullosa; RDEB, recessive dystrophic epidermolysis bullosa. aQOLEB is an EB-specific patient-reported outcome measure with scores ranging from 0 (best possible function/highest possible QOL) to 51 (lowest possible function/worst possible QOL). bSkindex-29 is a dermatology-specific patient-reported outcome measure with scores ranging from 0 (best possible QOL) to 100 (worst possible QOL) RDEB, emotional distress, worsened physical wellbeing, and increased household expenditure [47].

Direct costs and healthcare resource use
Direct medical costs in patients with RDEB were high (Table 6) [59,62,[76][77][78]. Medical expenses varied considerably; a patient survey in Ireland (N = 5) reported median payer-borne total medical costs, consisting of costs for wound dressings, drugs, overnight hospital stays, and outpatient visits, to be $84,534 per year [78], and a patient survey in South Korea (N = 13) reported total patient-borne medical costs, comprising medical dressings and all other disease-related expenses, to be $7392 per year [59]. Costs of wound dressing materials ranged from $4000-$245,000 in a US cost exercise model incorporating patient age and material quality [77], while costs of dressing materials reported in patient surveys from Ireland, South Korea, and the UK ranged from $4296-$28,727 [59,62,76]. Variations in cost are likely due to small sample sizes, contrasting health systems, and differences in EB subtypes. Patients with the severe subtype or complex wounds tended to report higher expenses [62,76,78].
Hospital resource use was reported in two studies in South Korea [59] and the UK [62], both with small samples. Almost half of patients in a South Korean survey were hospitalized in the previous year due to RDEB (6/13, 46%) [59]. Five patients (39%) were hospitalized for more than seven days. A survey of UK patients with RDEB (N = 10) reported a median duration of hospital stay of four and a half days (range: 2-155 days) [62].
Duration of dressing changes The time required for wound care was considerable (Table 7) [53,59,79]. A single-center survey of patients (N = 11) in the UK reported a median time of 25.25 h per week (101 h per month) spent on wound dressings [76]. In a US and Canadian survey, the majority of patients (55/90, 61%) required the assistance of one person for dressing changes; 17% (15/90) required two assistants [79]. Only 22% (20/90) did not require assistance.

Overall financial burden
The overall financial burden of RDEB was reported in two studies [59,61]. In the Korean survey, over half of respondents (7/13, 54%) reported always experiencing economic burden due to dressing materials [59]. Similarly, in a US survey, half (16/32, 50%) of respondents reported a high or severe level of impact on finances due to their disease [61].
Data on the indirect costs of RDEB, including impact on employment and productivity loss in patients with RDEB and their families, were not identified.

Discussion
This study is, to our knowledge, the first published systematic literature review (SLR) to comprehensively describe the clinical, humanistic, and economic burden of disease in patients with RDEB. A total of 65 studies met inclusion criteria for this systematic literature review, and, together, the data indicated that the cost of disease care, including wound management, in patients with RDEB and their families is considerable. Significant time was spent dressing wounds and the patient-borne Patients with RDEB experienced a significant impact on QOL due to their disease compared to other EB subtypes, skin diseases, and healthy controls and experience severe limitations in function and social activities. Furthermore, the humanistic and economic burden of RDEB extended beyond the patient to affect families and their interpersonal relationships.
A considerable burden was associated with large wounds, associated pain and itch, and multiple other comorbidities including infection, anemia, strictures and stenoses, contractures, difficulty walking, and failure to thrive. Patients with RDEB also had an increased risk of premature mortality due to pneumonia, sepsis, organ failure, and failure to thrive. Many patients developed SCC and associated complications in adulthood which can frequently be lethal. Finally, patients with RDEB underwent serious and intensive surgeries to manage their disease including esophageal dilation, gastrostomy tube placement, pseudosyndactyly release, and amputation. Our review also found that approximately 1 out of every 4 patients with RDEB do not undergo any confirmatory diagnostic testing and 1 out of every 3 do not undergo genetic analysis, though both are recommended by clinical practice guidelines [23,80]. This is in contrast to the US and many western European countries, where the vast majority of RDEB patients are genotyped. Identification of the causative mutation via genetic testing provides patients with a definitive diagnosis, estimation of disease prognosis, and potential inclusion into clinical trials [81].
One systematic review of the natural history of RDEB reported preliminary results as the first stage of development of a longitudinal cohort study in the UK (PEBLES) [82]. The authors identified limitations such as small sample sizes, high numbers of single-center studies, limited longitudinal data, unclear or no identified RDEB subtypes, and mixing of results between RDEB and other EB subtypes. Furthermore, they identified limited-to-no data on subjective or psychosocial aspects of RDEB and the economic burden of the disease. An SLR conducted by Montaudié and colleagues on SCC and EB reported the highest incidence of SCC in EB to occur in patients with RDEB, development of SCC to arise primarily in upper and lower extremities, and in areas with chronic wounds [83]. Findings from both of these reviews are consistent with our own.

Limitations of the existing literature
Several limitations within the existing body of literature were identified. First, sample sizes of studies were generally small and many studies were single-center, making generalizations from the study population to the larger RDEB population difficult or impossible. Additionally, almost all studies were cross-sectional, without longitudinal assessment of clinical, economic, or humanistic burden over the patient's lifetime. The severe and lifelong nature of RDEB would best be assessed by following patients over months or years to evaluate changes over time.
Second, as there are likely RDEB patients with milder phenotypes who have been misdiagnosed [3] or were underreported, these milder RDEB presentations are likely underrepresented in the literature. Thus, it is likely that severe and systemic manifestations such as non-esophageal strictures and stenoses, pseudosyndactyly, and microstomia, are overreported. Furthermore, in some studies, there is no distinction made between RDEB subtype or severity of disease. Additionally, some studies also included other types of EB such as DDEB or EBS, making generalizations about RDEB more difficult.
Third, economic data was limited and cost comparisons across studies and populations were difficult due to differences in currencies, health systems, and cost definitions. Furthermore, indirect costs such as impact on employment and productivity loss were not available. Additionally, evidence on satisfaction and/or burden associated with treatment was minimal.
Finally, disease burden in the literature may have been applicable to RDEB patients but reported in a larger EB cohort, and thus, not met inclusion criteria into the review. For instance, Danial and colleagues highlighted itch as the most bothersome symptom among all EB patients, including those with RDEB, but these data were not included in our results as they were not specific to RDEB patients [57].
Despite these limitations, the available literature suggests that the clinical, humanistic, and economic burden of RDEB is substantial.

Future research directions
Further research on the long-term impact of RDEB is needed to better understand how the burden of disease changes over the patient lifetime and stratified by disease severity. Recent research has shown a moderate-to-major financial impact of disease on patients with RDEB and high out-of-pocket dressing costs [84], but quantification of the economic burden across populations, geographies and healthcare systems is needed to provide appropriate care for patients. Further evaluation of the presence of anxiety, depression, and other mental health disorders that impact the humanistic burden of RDEB is also needed. Additionally, future research should measure the change in burden of disease as disease modifying treatments for RDEB, such as gene therapies, enter the market and are utilized.

Conclusion
Collectively, the evidence identified in this review suggests a critical unmet need for RDEB treatment options that addresses the underlying disorder. RDEB is associated with significant humanistic and economic burden on patients and their families/caregivers in addition to the clinical burden. New therapies that target the underlying disorder and stand to reduce wound burden could help address the overall disease burden.