Epidermolysis Bullosa Oropharyngeal Severity (EBOS) score: A multicenter development and reliability assessment Giulio Fortuna, DMD, PhD,a,b,d Nita Chainani-Wu, DDS, PhD, MPH,e Francina Lozada-Nur, DDS, MS, MPH,e Massimo Aria, PhD,c Rodrigo Cepeda-Valdes, MD,d Annamaria Pollio, DMD,b M. Peter Marinkovich, MD,a,f Adriana E. Martinez-Salazar, MD,d Michele D. Mignogna, MD, DMD,b Anna L. Bruckner, MD,g and Julio Cesar Salas-Alanıs, MDd,h,i Stanford, San Francisco, and Palo Alto, California; Naples, Italy; Monterrey, Mexico; and Aurora, Colorado Background: Epidermolysis bullosa (EB) is a genetic mucocutaneous disorder characterized by blister formation upon mild trauma. All 4 EB types may show oropharyngeal lesions involving either hard or soft tissues. Currently, there are very few data on EB scoring that include the oropharyngeal cavity. Objectives: We sought to develop an oropharyngeal severity score that was objective, valid, reliable, reproducible, easy to perform, and appropriate for all EB types. Methods: In this study, oral medicine specialists developed a new score, the EB Oropharyngeal Severity (EBOS) score. This measured oropharyngeal disease activity (erythema, atrophy, blisters, erosion/ ulceration) and structural damage (microstomia, ankyloglossia, scarring phenotype beyond microstomia and ankyloglossia, enamel hypoplasia). It was tested on 92 patients with different types/subtypes of EB, and interobserver and intraobserver reliability were assessed. Results: The EBOS mean total score was 12.9 6 10.9 (range: 0-34). Both interobserver and intraobserver reliability for total score on all patients with EB were considered excellent (intraclass correlation coefficient 0.94; 95% confidence interval 0.90-0.96 and intraclass correlation coefficient 0.90; 95% confidence interval 0.84-0.94, respectively). Even analyzing each single parameter of the disease activity and structural damage, a substantial to excellent correlation was found in the interobserver (except for 4 sites) and intraobserver reliability. A significant correlation was found between EB types/subtypes and the EBOS median score (P \ .001), but not between age and the EBOS mean total score in each group. Limitations: The sample size was small and the number of EB subtypes was limited. Conclusions: The EBOS score seems to represent an instrument capable of truly quantifying the oropharyngeal severity in different types/subtypes of EB, demonstrating excellent interobserver and intraobserver reliability. ( J Am Acad Dermatol 2013;68:83-92.) Key words: dental; epidermolysis bullosa; oral; prognosis; reliability; score. From the Department of Dermatology, Stanford University School of Medicine, Center for Clinical Sciences Researcha; Oral Medicine Unit, Department of Odontostomatological and Maxillofacial Sciences,b and Department of Mathematics and Statistics,c Federico II University of Naples; DebRA (Dystrophic Epidermolysis Bullosa Research Association) Mexico Foundationd; Department of Orofacial Sciences, School of Dentistry, University of California, San Franciscoe; Division of Dermatology, Veterans Affairs Palo Alto Healthcare Systemf; Departments of Dermatology and Pediatrics, University of Colorado, School of Medicineg; Dermatology Service, Instituto Tecnol ogico y de Estudios Superiores de Monterreyh; and Department of Dermatology, Facultad de Medicina y Hospital Universitario Dr Jose E. Gonzalez, Universidad Autonoma de Nuevo Leon, Monterrey.i Drs Bruckner and Salas-Alanıs equally contributed to the work and need to be considered as last authors.
Funding for this work was provided by the Department of Dermatology, Stanford University School of Medicine, Center for Clinical Sciences Research; Office of Research and Development, Palo Alto Veterans Affairs Medical Center (Dr Marinkovich); DebRA Mexico Foundation; and Oral Medicine Unit, Department of Odontostomatological and Maxillofacial Sciences, Federico II University of Naples. Conflicts of interest: None declared. Accepted for publication April 8, 2012. Reprint requests: Giulio Fortuna, DMD, PhD, Department of Dermatology, Stanford University School of Medicine, Center for Clinical Sciences Research, Marinkovich Laboratory, 269 Campus Dr, Stanford, CA 94305. E-mail:
[email protected]. Published online May 10, 2012. 0190-9622/$36.00 Ó 2012 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2012.04.009
83
J AM ACAD DERMATOL
84 Fortuna et al
JANUARY 2013
Epidermolysis bullosa (EB) encompasses a group In light of an important recent study, which has of inherited mucocutaneous disorders characterized validated a system on EB quality of life,18 and conby the occurrence of blisters onto the skin and sidering the importance and impact of the orophamucous membranes after mild mechanical trauma.1 ryngeal component on the global health of patients Recently, EB has been classified into 4 major types, with EB, we have developed a new separate score, based on the split at the ultrastructural level: called EB Oropharyngeal Severity (EBOS) score. Our intraepidermal or epidermolytic (EB simplex), intrapurpose was to quantify and monitor the oropharynlamina lucida or lucidolytic geal involvement with an in(junctional EB [JEB]), sublastrument that was as much as CAPSULE SUMMARY mina densa or dermolytic possible: (1) objective, to (dystrophic EB [DEB]), and guarantee an effective and Children with any type of epidermolysis mixed (Kindler syndrome), practical report of clinical bullosa are at higher risk of developing and about 30 different signs far from subjective paoropharyngeal lesions, involving either subtypes.2 tient’s perception; (2) valid, hard or soft tissues. Many patients with EB with content validity evaluVery few reports have been published may have systemic complicaated by experts in the field of aimed at developing an epidermolysis tions and mucosal lesions, oral medicine; (3) reproducbullosa scoring system that give any such as genital, ocular, ible and reliable worldwide weight to the oropharyngeal and oropharyngeal ones.3-5 among the same and differcomponent. These may affect both hard ent oral health care providers and soft tissues, showing dif(oral medicine specialists, This study provides a new, objective, ferent features and degrees dermatologists, otorhinolareasy-to-perform, and reproducible of severity.6 In all EB types yngologists, pediatricians); scoring system for the oropharyngeal oral soft tissues are fragile, (4) easy to use, allowing clicomponent in children with resulting in frequent blister nicians to calculate the total epidermolysis bullosa, which has and/or erosion formation, score very quickly in order to demonstrated excellent interobserver accompanied, in some EB avoid prolonged stress to the and intraobserver reliability. subtypes, by a scarring pheEB patients; and (5) appronotype and, although rarely, priate for all EB types/ oral milia.7-10 Similarly, oral subtypes. hard tissues may show either a marked developmentally compromised enamel or minor structural deMETHODS fects with areas of surface pitting and furrowing.11-15 Study design and patients Although until now no consensus statement on This was a multicenter study collecting data from EB severity score has been established, two reports 92 patients with EB between September 2010 and have been published aimed at developing an EB September 2011 coming from the EB clinic at Lucile scoring system.16,17 These attempted to develop a Packard Children’s Hospital, Palo Alto, CA; the adult method of scoring EB severity but evaluated too Bullous Disease Clinic, Stanford, CA; and dermatolmany variables (eg, skin, height, weight, mucous ogy clinic at Instituto Tecnol ogico y de Estudios membranes, nutritional status, cancer) and gave Superiores and DebRA Mexico Foundation, little, if any, weight to the oropharyngeal compoMonterrey, Mexico. The Department of Orofacial nent. Hence, there is a need to develop independent Sciences, School of Dentistry, University of mucosal scoring systems, because a single, allCalifornia, San Francisco, and the Oral Medicine inclusive score appears to be inadequate, being Unit, Department of Odontostomatological and unable to truly reflect the severity of clinical condiMaxillofacial Science, Federico II University of tions and to correlate with disease prognosis. Our Naples, Italy, cooperated with them. All patients concern regarding the oropharyngeal involvement is provided their written informed consent. This study that previously16,17 it has been evaluated solely on was approved by the ethical committees of Stanford the basis of subjective clinicians’ observations and University and Instituto Tecnol ogico in Monterrey. patients’ reports (presence/absence of blisters/eroAll patients were enrolled based on the following sions over an undefined period of time, eg, inclusion criteria: always, several per month,7 occasionally, frequently, persistently8). Such drawbacks should render these 1. Patients of both genders, all ages and races, with scoring systems less reliable and reproducible, causthe presence of typical mucocutaneous lesions of ing the notion of ‘‘score’’ to collapse. any EB type/subtype, as previously reported.2 d
d
d
J AM ACAD DERMATOL VOLUME 68, NUMBER 1
Abbreviations used: CI: DEB: EB: EBOS: ICC: IQR: JEB: RDEB:
confidence interval dystrophic epidermolysis bullosa epidermolysis bullosa Epidermolysis Bullosa Oropharyngeal Severity intraclass correlation coefficient interquartile ranges junctional epidermolysis bullosa recessive dystrophic epidermolysis bullosa
2. Diagnosis of EB based on skin biopsy specimen with routine histology and immunofluorescence antigen mapping, and/or, when available, electron microscopy, and/or DNA analysis. 3. Patients able to give consent if older than 18 years. For minor patients consent was given by their parents or guardians. At the time of admission exclusion criteria encompassed: 1. Patients who had used topical corticosteroids and/or topical and/or systemic antifungal therapy during the 3 weeks before the study, as such medications might have a significant effect on the clinical appearance of the oropharyngeal mucosa. 2. Patients with history of or current oropharyngeal malignancy and/or potentially malignant disorders. Generation and refinement of scoring items Originally, in designing the EBOS, 3 oral medicine specialists (G. F., F. L-N., and M. D. M.) discussed the possible choices and, based on their clinical experience in the field of oropharyngeal blistering diseases, decided to assign a score only to the number of sites involved and to microstomia, ankyloglossia, and enamel hypoplasia. After an accurate review of the literature on clinical features of oropharyngeal lesions in patients with EB and a discussion with two other oral medicine specialists (N. C-W. and A. P.), this score was abandoned as not really indicative of disease severity. Content validity was accurately reviewed and the EBOS was refined by introducing the nature of oropharyngeal lesions and the presence of a scarring phenotype in other parts of the oral cavity, beyond microstomia and ankyloglossia. This new instrument to measure the oropharyngeal severity score of EB was approved by 4 dermatologists (A. L. B., M. P. M., R. C-V., and J. C. S-A.). Eventually, a more appropriate EBOS was redesigned, including two different scores: disease activity and structural damage (Fig 1).
Fortuna et al 85
The first evaluated only clinical signs, as objective findings. Specifically, 4 were identified as key features of disease activity: erythema, atrophy, blister, and erosion/ulceration. These signs were not scored in terms of quality (mild, moderate, severe), because it was too subjective among physicians, or quantity (number of lesions), because it was too difficult and confusing to calculate, as usually oropharyngeal lesions tend to be confluent. Clinically, atrophy in EB appeared similar to that seen in progressive systemic sclerosis, ie, with vestibule obliteration, depapillated tongue, disappearance of palate rugae, and blanching of buccal mucosa and/ or soft palate. The second evaluated the presence or absence of 4 parameters: microstomia; ankyloglossia; intraoral scars beyond microstomia and ankyloglossia, such as vestibule obliteration; and enamel hypoplasia. These clinical features were considered more permanent, as a part of previous damage and, therefore, not necessarily reflecting current disease activity. Microstomia was evaluated with the maximal mouth opening by measuring the distance from the marginal edge of the central upper to lower incisors, along the interincisal line. In case of missing teeth in one or both jaws, measurement was done considering the distance between edentulous alveolar ridges, passing through a virtual line that connects the two craniometric points: nasion and gnathion. A patient with a maximal aperture less than 35 mm was considered as having microstomia, as previously reported.19 Ankyloglossia was evaluated by the ability of each patient to protrude the tongue over the lower incisors or edentulous alveolar ridge, move it over to the left and right side, and reach the premaxilla with the tip of the tongue. A patient unable to perform at least two of the above-mentioned movements was considered as having ankyloglossia. The EBOS score To increase the score sensitivity, the oropharyngeal cavity was divided into 13 different anatomic sites. The disease activity score evaluated each site affected by one or more clinical signs. We decided to assign 1 point to each clinical sign present in each anatomical site, leading to a total score ranging from 0 to 52. Conversely, the structural damage score evaluated the presence or absence of the 4 structural damages, assigning 2 points each to a total score ranging from 0 to 8 (Fig 1). The grading system was based on the sum of both scores, reaching a final total score ranging from 0 to 60, rather than on the virtually impossible task of determining accurately the percentage of each site involved by each type of lesion.
J AM ACAD DERMATOL
86 Fortuna et al
JANUARY 2013
Fig 1. Epidermolysis Bullosa Oropharyngeal Severity score.
Interobserver and intraobserver reliability Interobserver reliability was evaluated in all patients, who were scored independently by two different physicians on the same day: 42 patients were seen in Stanford, CA, by G. F. and A. L. B. and 50 patients in Monterrey, Mexico, by G. F. and J. C. S-A. The number of patients with EB varied from 3 to 5 per day. Conversely, intraobserver reliability was assessed on a randomly selected group of patients (34 of 92), scored by G. F. and J. C. S-A. Such patients were asked to come back after 3 hours and to not eat anything, use any kind of topical/mouthwash medication, drink alcohol, or smoke cigarettes. To minimize recall bias, during the 3 hours of interval, the scorers were asked to view 100 consecutive photographs of patients with oropharyngeal blistering diseases, such as pemphigus vulgaris, mucous membrane pemphigoid, and erosive lichen planus. Eventually, each patient was seen twice by the same physician, and, on the second round, in a different and random order compared with his/her first visit. This system was adapted from the method used to validate the Autoimmune Bullous Skin Disorder Intensity Score and Pemphigus Disease Area Index.20 Time for scoring was also recorded.
Statistical analysis Descriptive statistics of demographic characteristics and EB type/subtype distribution were calculated as mean 6 SD. The EBOS score was calculated as a mean value of the scores from two investigators for all 92 patients and the subgroup of 34 patients. Means, medians, and interquartile ranges (IQR) in each EB type/subtype and each oropharyngeal site were also calculated. Interobserver and intraobserver reliability for disease activity and structural damage (separately and grouped) were calculated by intraclass correlation coefficient (ICC) along with 95% confidence interval (CI). The ICC values were interpreted as follows: 0.00 to 0.20 = poor agreement, 0.21 to 0.40 = fair agreement, 0.41 to 0.60 = moderate agreement, 0.61 to 0.80 = substantial agreement, and 0.81 to 1.00 = excellent agreement.21 Spearman correlation coefficient was used to assess the relationship between age and the EBOS mean total score, and nonparametric Kruskal-Wallis analysis of variance was used to assess the relationship between EB types/subtypes and the EBOS median total score. P values of less than .05 were considered significant. Statistical analyses were performed using software (SPSS for Windows, Version 17.0, SPSS Inc, Chicago, IL).
J AM ACAD DERMATOL
Fortuna et al 87
VOLUME 68, NUMBER 1
RESULTS Patients’ characteristics During the study period 92 patients (48 [52.2%] female and 44 [47.8%] male) with a mean age of 15.4 years (range: 2 months-63 years) (SD 14.1; 95% CI 12.4-18.3) with different EB types/subtypes (Table I) were tested with the EBOS. Activity and damage score distribution The EBOS mean total score 6 SD was 12.9 6 10.9 (range: 0-33.5), whereas the mean total score 6 SD for disease activity and structural damage was 10.1 6 8.5 (range: 0-27.5) and 2.8 6 2.6 (range: 0-6), respectively. The range of EBOS mean score was 0 to 33.5 of a possible 60 maximum score, with a median of 13.00 and an IQR of 20.50 (Table II and Fig 2), whereas in the subsample (N = 34) the range distribution on time 1 (test) was 0 to 28, with a median of 13.00 and an IQR of 18.50 and on time 2 (retest) it was 0 to 26.5 with a median of 11.50 and an IQR of 16.50 (Table II and Fig 3). The highest EBOS score by site was measured for tongue, whereas the lowest was for upper and lower fornices, either on the total sample (Table III) or a subsample (Table IV). Interobserver and intraobserver reliability The interobserver and intraobserver reliabilities for total score on all patients with EB were excellent (ICC 0.94; 95% CI 0.90-0.96 and ICC 0.90; 95% CI 0.84-0.94, respectively). Even a comparison between the mean total score of patients with recessive DEB (RDEB) versus all other patients with EB showed excellent agreement, unlike the total score on all other patients with EB in the interrater assessment (ICC 0.58; 95% CI 0.30-0.77) (Table V). In the analysis of disease severity, either in the interrater or intrarater assessment, lower lip reached the highest agreement (ICC 0.89; 95% CI 0.84-0.92 and ICC 0.91; 95% CI 0.83-0.95, respectively), followed by tongue (ICC 0.87; 95% CI 0.81-0.91) in the interrater and by hard palate (ICC 0.87; 95% CI 0.76-0.93) in the intrarater (Tables III and IV). As far as structural damage is concerned, microstomia and ankyloglossia, either in the interrater (ICC 0.93; 95% CI 0.90-0.97 and ICC 0.89; 95% CI 0.84-0.93, respectively) or in the intrarater (ICC 0.94; 95% CI 0.89-0.97 and ICC 0.94; 95% CI 0.88-0.97, respectively) assessment, reached the highest agreement, (Tables III and IV). Correlation of EBOS score with EB types/ subtypes and age Lastly, a significant correlation was seen between EB types/subtypes and the EBOS median score
Table I. Characteristics of patients with epidermolysis bullosa Patients with EB
Age, y, mean 6 SD Male:female EB distribution
EBS-loc EBS-DM JEB-H JEB-nH-gen DDEB-gen DDEB-Pr RDEB-O RDEB-sev gen
Sample for interobserver analysis
Subsample for intraobserver analysis
15.36 6 14.12
19.76 6 15.78
44:48
11:23
No. (frequency %)
5 6 3 2 7 10 14 45
(5.4) (6.5) (3.3) (2) (7.6) (10.9) (15.2) (48.9)
No. (frequency %)
2 3 0 0 0 6 5 18
(5.9) (8.8) (0) (0) (0) (17.6) (14.7) (52.9)
DDEB, Dominant dystrophic epidermolysis bullosa; DM, DowlingMeara; EB, epidermolysis bullosa; EBS, epidermolysis bullosa simplex; gen, generalized; H, Herlitz variant; JEB, junctional epidermolysis bullosa; loc, localized; nH, non-Herlitz variant; O, generalized other; Pr, pruriginosa; RDEB, recessive dystrophic epidermolysis bullosa; sev, severe.
(Kruskal-Wallis analysis of variance = 71.626; P \ .001) (Table II) but not between age and the EBOS mean total score. Indeed, there was a decline of score in patients with EB simplex and JEB with age (EB simplex: r = 0.52, P = .107; JEB: r = 0.80, P = .13) and an increase in the rest of patients with EB (dominant DEB: r = 0.20, P = .42; RDEB generalized other: r = 0.20, P = .48; RDEB severe generalized: r = 0.03, P = .85), but none of them was statistically significant.
DISCUSSION Since it was first described in 1879,22 oropharyngeal involvement in patients with EB has been a key concern for clinicians everywhere and still represents one of the most important and challenging clinical manifestations in terms of evaluation, prognosis, and treatment. Although attempts have been made to improve the oral conditions of patients with EB through medical/surgical treatments,23-30 evaluation and prognosis of the oropharyngeal component in EB remains an enigma. All EB types/subtypes may experience all 4 clinical signs evaluated in this study, although with different frequencies and extensions: some had no or very few, mild lesions, whereas others had many severe, disfiguring lesions. It is therefore important that an instrument capable of distinguishing between active lesions and damage is available, to improve the global assessment of oropharyngeal severity in
J AM ACAD DERMATOL
88 Fortuna et al
JANUARY 2013
Table II. Epidermolysis Bullosa Oropharyngeal Severity total score distribution per epidermolysis bullosa type/ subtype EB types/subtypes
Mean score
Median score
Range distribution
IQR
Kruskal-Wallis ANOVA
0.70 1.25 7.30 4.00 0.00 1.25 7.00 21.90 12.90
0.00 0.00 7.50 4.00 0.00 0.25 10.50 23.00 13.00
0-3 0-6 4-11 2-6 0-0 0-4 2-22 7-34 0-34
1.00 1.50 6.50 4.00 0.00 2.50 8.50 11.00 20.50
P \ .001
EBS-loc EBS-DM JEB-H JEB-nH-gen DDEB-gen DDEB-Pr RDEB-O RDEB-sev gen Total score
Time 1*
EBS-loc EBS-DM DDEB-Pr RDEB-O RDEB-sev gen Total score
Time 2*
Mean score
Median score
Range distribution
0.00 2.50 2.10 8.70 20.30 12.60
0.00 1.50 2.50 10.50 21.25 13.00
0-0 0-6 0-3.5 3.5-13 8.5-28 0-28
IQR
Mean score
Median score
Range distribution
IQR
0.00 6.00 3.00 4.50 6.50 18.50
0.00 2.30 2.70 9.10 19.20 12.20
0.00 1.50 2.50 11.00 20.00 11.50
0-0 0-5.5 0-7 3.5-11 10-26.5 0-26.5
0.00 5.50 3.00 2.00 9.00 16.50
ANOVA, Analysis of variance; DDEB, dominant dystrophic epidermolysis bullosa; DM, Dowling-Meara; EB, epidermolysis bullosa; EBS, epidermolysis bullosa simplex; gen, generalized; H, Herlitz variant; IQR, interquartile range; JEB, junctional epidermolysis bullosa; loc, localized; nH, non-Herlitz variant; O, generalized other; Pr, pruriginosa; RDEB, recessive dystrophic epidermolysis bullosa; sev, severe. *Epidermolysis Bullosa Oropharyngeal Severity score for subsample of patients (N = 34) calculated on first (time 1) and second (time 2) rounds and used for intrarater reliability analysis.
hospital settings or in clinical trials of potential medications. To our knowledge, no data have been published on the grading of the individual clinical signs in each single oropharyngeal site in EB. Our results on 92 patients with EB show that the EBOS has an excellent interobserver and intraobserver reliability on both total score and partial score of disease activity and structural damage score, with a median score changing significantly upon different EB type/subtypes (P \ .001) (Table II), unlike age, which shows no correlation with the EBOS mean total score. Because we expected patients with RDEB to be more severe, to reduce the bias of sample dilution, we divided all 92 patients into two subgroups: RDEB versus all other EB type/subtypes. Even in this case the intrarater reliability was excellent in both groups, whereas the interrater reliability was excellent for patients with RDEB (ICC 0.87) and moderate for all other patients with EB (ICC 0.58) (Table V). We hypothesize that this difference of interrater reliability (excellent for RDEB vs moderate for all other EB) might be because oropharyngeal manifestations are very heterogeneous among all EB types/subtypes in terms of quality and quantity and, thus, it is likely that two scorers with different background/training are capable of reaching an agreement more easily on gross and more widespread lesions, as usually seen in
patients with RDEB, rather than on subtle and sporadic lesions, as tends to be observed in the remaining patients with EB. Interestingly, patients with RDEB reached the highest EBOS mean total score versus all other EB types/subtypes (Fig 2) as expected, in line with previous reports.6-10,19,23 Therefore, this result supports EBOS validity. Lastly, because we obtained such good results in terms of interrater reliability (either for all patients with EB combined or for the two subgroups [RDEB only and all EB except RDEB]) when the patients from Stanford, CA, and Monterrey, Mexico, were analyzed separately, we decided to group these patients together. The interrater reliability was shown to be poor to fair for upper and lower fornices (ICC 0.05 and 0.10, respectively), floor of the mouth (ICC 0.21), and oropharynx (ICC 0.34) (Table III), but was substantial to excellent either for disease activity or structural damage in the intrarater assessment. Unfortunately, two sites (soft palate and oropharynx) and one permanent damage (enamel hypoplasia) could not be evaluated, because dates were absent for both scorers or were present for only one. These disagreements between the two scorers may reflect differences in their ability to detect lesions in specific oropharyngeal sites such as upper and/or lower fornices, floor of the mouth, soft palate, and oropharynx and to classify the lesions based on their
J AM ACAD DERMATOL VOLUME 68, NUMBER 1
Fig 2. Box plot analysis of Epidermolysis Bullosa (EB) Oropharyngeal Severity total score distribution per EB type/subtype calculated on full sample of 92 patients. Lines in box denote medians, bar include at most 1.5 of interquartile distance, difference between first and third quartile of score. Circle indicates value of extreme outlier. DDEB, Dominant dystrophic EB; DM, Dowling-Meara; EBS, EB simplex; gen, generalized; H, Herlitz variant; JEB, junctional EB; loc, localized; nH, non-H variant; O, gen other; RDEB, recessive dystrophic EB; sev, severe.
clinical characteristics. Indeed, the different use of light may have led to underestimation or overestimation of clinical signs such as erythema, whose detection largely relies upon light direction, mostly in the most posterior oropharyngeal sites, such as soft palate and oropharynx. Lastly, the EBOS is a quick and easy-to-use tool, with the time taken for recording ranging from 1 to 5 minutes. Two parameters were not included in the EBOS. First, oropharyngeal malignancy was excluded because it was decided that this would need separate evaluation because prognosis could be severe. In addition, to our knowledge, only two reports have been made of EB and oropharyngeal cancer,31,32 and therefore we regarded this parameter to be currently unvalidated. Whether the lack of reports means that patients with EB rarely develop oropharyngeal cancer or simply that such lesions are undetected or underestimated remains unknown. In our study, none of the 92 patients presented with any oropharyngeal malignancy and/or potentially malignant disorders. Second, the presence of dental caries and/or periodontal disease (with subsequent dental loss) was not included, as these are not directly related to the disease. Indeed, although the percentage of dental caries appears to be higher in patients with EB versus a control group, it is not related to the disease.33 However, it would be necessary to
Fortuna et al 89
Fig 3. Box plot analysis of Epidermolysis Bullosa (EB) Oropharyngeal Severity total score distribution calculated on subsample of 34 patients at time 1 (test) and time 2 (retest) per EB type/subtype. Lines in box denote medians, bar include at most 1.5 of interquartile distance, difference between first and third quartile of score. Circle indicates value of extreme outlier. DDEB, Dominant dystrophic EB; DM, Dowling-Meara; EBS, EB simplex; gen, generalized; loc, localized; O, gen other; RDEB, recessive dystrophic EB; sev, severe.
compare patients who have EB with a control group selected from different social backgrounds to accurately ascertain the incidence in both groups. Lastly, we decided not to evaluate associated symptoms such as pain and/or discomfort, and daily activity impairment such as ability to eat, drink, or speak in this score because we considered these to be subjective parameters more suited to evaluation of the quality of life and quality of oral health. This study has several limitations. The sample size was small and the number of EB subtypes was limited. It would be preferable to test this score in different hospital settings in a larger sample and a greater variety of different EB subtypes. Another possible limitation is the evaluation of interobserver and intraobserver reliability by only two physicians. However, we intentionally limited the physicians to two to avoid causing excessive pain in patients and overloading them with multiple visits by more than two physicians. In addition, the intraobserver reliability was tested on only a part of the total sample. Despite these limitations, the EBOS may offer many important future perspectives. First, a validated scoring system capable of objectively evaluating oropharyngeal disease severity may be useful even longitudinally in better understanding the disease progression, thereby representing a valid prognostic tool. Indeed, the EBOS may be used to follow up patients with EB from birth and assess how the oropharyngeal cavity will be affected over time to
J AM ACAD DERMATOL
90 Fortuna et al
JANUARY 2013
Table III. Interobserver reliability and score distribution of disease activity and structural damage Measure
Mean score
Median score
Range distribution
IQR
ICC
95% CI
1.00 1.22 0.12 0.10 0.70 0.73 1.01 1.12 1.14 0.45 2.13 0.10 0.24 10.08
0.50 1.00 0.00 0.00 0.50 0.50 0.75 1.00 0.75 0.00 3.00 0.00 0.00 9.75
0-4 0-4 0-1 0-1 0-2 0-2 0-3 0-3 0-4 0-2.5 0-4 0-1.5 0-2 0-27.5
2.00 2.50 0.00 0.00 1.50 1.50 2.00 2.00 2.12 0.62 3.50 0.00 0.50 16.75
0.79 0.89 0.05 0.10 0.76 0.77 0.78 0.81 0.85 0.66 0.87 0.21 0.34 0.91
0.69-0.85 0.84-0.92 0.00-0.26 0.00-0.30 0.65-0.84 0.67-0.84 0.67-0.85 0.72-0.87 0.74-0.91 0.53-0.76 0.81-0.91 0.01-0.40 0.15-0.51 0.87-0.94
0.97 0.97 0.77 0.08 2.77
0.50 1.00 0.00 0.00 3.00
0-2 0-2 0-2 0-2 0-6
2.00 2.00 2.00 0.00 6.00
0.93 0.89 0.82 0.85 0.95
0.90-0.97 0.84-0.93 0.74-0.87 0.79-0.90 0.92-0.97
Disease activity Upper lip Lower lip Upper fornix Lower fornix Upper gingiva Lower gingiva Right cheek Left cheek Hard palate Soft palate Tongue Floor of mouth Oropharynx Total Structural damage Microstomia Ankyloglossia SPh-O Enamel hypoplasia Total
CI, Confidence interval; ICC, intraclass correlation coefficient; IQR, interquartile range; SPh-O, scarring phenotype in other parts of oral cavity, beyond microstomia and ankyloglossia.
Table IV. Intraobserver reliability and score distribution of disease activity and structural damage on subsample of patients (N = 34) calculated on first and second times Time 1 Measure
Disease activity Upper lip Lower lip Upper fornix Lower fornix Upper gingiva Lower gingiva Right cheek Left cheek Hard palate Soft palate* Tongue Floor of mouth Oropharynx* Structural damage Microstomia Ankyloglossia SPh-O Enamel hypoplasia*
Time 2
Mean score
Median score
Range distribution
Mean score
Median score
IQR
Range distribution
IQR
ICC
95% CI
1.30 1.57 0.23 0.20 0.66 0.72 0.85 1.01 0.80 — 2.02 0.25 —
1 1.5 0 0 0.5 0.75 0.5 1 0.75 — 2.25 0 —
0-3 0-3.5 0-1 0-1 0-2 0-2 0-2.5 0-3 0-2.5 — 0-4 0-1.5 —
1.90 2.75 0.50 0.50 1.00 1.00 1.50 1.90 1.50 — 3.38 0.50 —
1.20 1.52 0.11 0.14 0.60 0.75 0.83 1.10 0.75 — 1.91 0.23 —
1.25 1.50 0.00 0.00 0.50 0.75 0.50 1.00 1.00 — 2.25 0.00 —
0-3 0-3.5 0-1 0-0.5 0-1.5 0-2 0-2.5 0-3 0-2.5 — 0-4 0-2 —
1.38 2.00 0.00 0.50 1.00 1.00 1.38 2.00 1.00 — 2.88 0.50 —
0.86 0.91 0.72 0.82 0.77 0.69 0.65 0.85 0.87 — 0.86 0.74 —
0.73-0.93 0.83-0.95 0.52-0.85 0.71-0.92 0.59-0.88 0.46-0.83 0.39-0.81 0.72-0.92 0.76-0.93 — 0.74-0.93 0.54-0.86 —
0.88 0.88 0.85 —
0 0.5 0.5 —
0-2 0-2 0-2 —
2 2 2 —
0.88 0.85 0.85 —
0.00 1.00 1.00 —
0-2 0-2 0-2 —
2.00 2.00 2.00 —
0.94 0.94 0.77 —
0.89-0.97 0.88-0.97 0.59-0.88 —
CI, Confidence interval; ICC, intraclass correlation coefficient; IQR, interquartile range; SPh-O, scarring phenotype in other parts of oral cavity, beyond microstomia and ankyloglossia. *Data were absent or present only for one scorer.
J AM ACAD DERMATOL
Fortuna et al 91
VOLUME 68, NUMBER 1
Table V. Interobserver and intraobserver reliability of total score on all 92 patients with epidermolysis bullosa and on only patients with recessive dystrophic epidermolysis bullosa versus all others epidermolysis bullosa Agreement
Interrater
Intrarater
Measure
ICC
95% CI
Total score on all patients with EB Total score on only patients with RDEB Total score on all patients with EB except RDEB
0.94
0.90-0.96
0.87
0.73-0.93
0.58
0.30-0.77
0.90
0.84-0.94
0.78
0.64-0.89
0.84
0.54-0.95
Total score on all patients with EB Total score on only patients with RDEB Total score on all patients with EB except RDEB
CI, Confidence interval; EB, epidermolysis bullosa; ICC, intraclass correlation coefficient; RDEB, recessive dystrophic epidermolysis bullosa.
possibly prevent structural damage and ameliorate disease activity. Second, it appears that the severity of oropharyngeal lesions increases with the severity of cutaneous lesions,3 and it would therefore be interesting to correlate the EBOS with other cutaneous scores in the literature. Third, it would be interesting to correlate the EBOS with the type of mutation to determine whether there is an oropharyngeal phenotypic-genotypic correlation. Fourth, it may be useful to investigate any clinically meaningful change of the EBOS in response to medical and/or surgical oropharyngeal care interventions, particularly in clinical trials. In conclusion, in this preliminary study the EBOS was found to be a clinically valid and reliable tool for assessing oropharyngeal severity in patients with EB. However, to better confirm our results, further investigation and refinement by international groups are strongly recommended. We are very grateful to all children and adults with EB for their courage, strength, and patience. We would like to thank Dr Phuong Khuu, Dr Grainne O’Regan, all dermatology residents, Lisa Taylor, Lorraine Spaulding, Robin Newman, Gretchen Flanagan, Abigaille Augustin, and all staff of the EB clinic at Lucile Packard Children’s Hospital and the Bullous Disease Clinic at Stanford. We also would like to thank all DebRA Mexico staff, specifically Lic. Erika Salas, Myrna Lozano, Juan Salas, Dr Erika Garza, and Lic. Peggy Perez.
REFERENCES 1. Pai S, Marinkovich MP. Epidermolysis bullosa: new and emerging trends. Am J Clin Dermatol 2002;3:371-80. 2. Fine JD, Eady RA, Bauer EA, Bauer JW, Bruckner-Tuderman L, Heagerty A, et al. The classification of inherited epidermolysis bullosa (EB): report of the third international consensus meeting on diagnosis and classification of EB. J Am Acad Dermatol 2008;58:931-50. 3. Fine JD, Mellerio JE. Extracutaneous manifestations and complications of inherited epidermolysis bullosa, part II: other organs. J Am Acad Dermatol 2009;61:387-402; quiz 3-4. 4. Fine JD, Mellerio JE. Extracutaneous manifestations and complications of inherited epidermolysis bullosa, part I: epithelial associated tissues. J Am Acad Dermatol 2009;61:367-84; quiz 85-6. 5. Fu T, Lingala B, Kent K, Bachrach LK, Bruckner AL. Patterns of bone mineral acquisition in children with epidermolysis bullosa: a longitudinal study. Br J Dermatol 2011;165:1081-6. 6. Wright JT. Oral manifestations in the epidermolysis bullosa spectrum. Dermatol Clin 2010;28:159-64. 7. Wright JT, Fine JD, Johnson LB. Oral soft tissues in hereditary epidermolysis bullosa. Oral Surg Oral Med Oral Pathol 1991;71: 440-6. 8. Wright JT, Fine JD, Johnson LB, Steinmetz TT. Oral involvement of recessive dystrophic epidermolysis bullosa inversa. Am J Med Genet 1993;47:1184-8. 9. Serrano-Martinez MC, Bagan JV, Silvestre FJ, Viguer MT. Oral lesions in recessive dystrophic epidermolysis bullosa. Oral Dis 2003;9:264-8. 10. Azrak B, Kaevel K, Hofmann L, Gleissner C, Willershausen B. Dystrophic epidermolysis bullosa: oral findings and problems. Spec Care Dentist 2006;26:111-5. 11. Wright JT, Johnson LB, Fine JD. Development defects of enamel in humans with hereditary epidermolysis bullosa. Arch Oral Biol 1993;38:945-55. 12. Kirkham J, Robinson C, Strafford SM, Shore RC, Bonass WA, Brookes SJ, et al. The chemical composition of tooth enamel in recessive dystrophic epidermolysis bullosa: significance with respect to dental caries. J Dent Res 1996;75:1672-8. 13. Wright JT, Hall KI, Deaton TG, Fine JD. Structural and compositional alteration of tooth enamel in hereditary epidermolysis bullosa. Connect Tissue Res 1996;34:271-9. 14. Kirkham J, Robinson C, Strafford SM, Shore RC, Bonass WA, Brookes SJ, et al. The chemical composition of tooth enamel in junctional epidermolysis bullosa. Arch Oral Biol 2000;45:377-86. 15. Nakamura H, Sawamura D, Goto M, Kida M, Ariga T, Sakiyama Y, et al. Analysis of the COL17A1 in non-Herlitz junctional epidermolysis bullosa and amelogenesis imperfecta. Int J Mol Med 2006;18:333-7. 16. Tamai K, Hashimoto I, Hanada K, Ikeda S, Imamura S, Ogawa H, et al. Japanese guidelines for diagnosis and treatment of junctional and dystrophic epidermolysis bullosa. Arch Dermatol Res 2003;295(Suppl 1):S24-8. 17. Moss C, Wong A, Davies P. The Birmingham epidermolysis bullosa severity score: development and validation. Br J Dermatol 2009;160:1057-65. 18. Frew JW, Martin LK, Nijsten T, Murrell DF. Quality of life evaluation in epidermolysis bullosa (EB) through the development of the QOLEB questionnaire: an EB-specific quality of life instrument. Br J Dermatol 2009;161:1323-30. 19. Stellingsma C, Dijkstra PU, Dijkstra J, Duipmans JC, Jonkman MF, Dekker R. Restrictions in oral functions caused by oral manifestations of epidermolysis bullosa. Eur J Dermatol 2011;21:405-9. 20. Rosenbach M, Murrell DF, Bystryn JC, Dulay S, Dick S, Fakharzadeh S, et al. Reliability and convergent validity of
92 Fortuna et al
21.
22.
23. 24. 25.
26.
two outcome instruments for pemphigus. J Invest Dermatol 2009;129:2404-10. Chandran V, Gottlieb A, Cook RJ, Duffin KC, Garg A, Helliwell P, et al. International multicenter psoriasis and psoriatic arthritis reliability trial for the assessment of skin, joints, nails, and dactylitis. Arthritis Rheum 2009;61:1235-42. Arwill T, Olsson O, Bergenholtz A. Epidermolysis bullosa hereditaria, 3: a histologic study of changes in teeth in the polydysplastic dystrophic and lethal forms. Oral Surg Oral Med Oral Pathol 1965;19:723-44. Kramer SM. Oral care and dental management for patients with epidermolysis bullosa. Dermatol Clin 2010;28:303-9, x. Marini I, Vecchiet F. Sucralfate: a help during oral management in patients with epidermolysis bullosa. J Periodontol 2001;72:691-5. Brooks JK, Bare LC, Davidson J, Taylor LS, Wright JT. Junctional epidermolysis bullosa associated with hypoplastic enamel and pervasive failure of tooth eruption: oral rehabilitation with use of an overdenture. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;105:e24-8. Wright JT. Comprehensive dental care and general anesthetic management of hereditary epidermolysis bullosa: a review of fourteen cases. Oral Surg Oral Med Oral Pathol 1990;70:573-8.
J AM ACAD DERMATOL
JANUARY 2013
27. Wright JT. Epidermolysis bullosa: dental and anesthetic management of two cases. Oral Surg Oral Med Oral Pathol 1984;57: 155-7. 28. Penarrocha M, Larrazabal C, Balaguer J, Serrano C, Silvestre J, Bagan JV. Restoration with implants in patients with recessive dystrophic epidermolysis bullosa and patient satisfaction with the implant-supported superstructure. Int J Oral Maxillofac Implants 2007;22:651-5. 29. Penarrocha M, Rambla J, Balaguer J, Serrano C, Silvestre J, Bagan JV. Complete fixed prostheses over implants in patients with oral epidermolysis bullosa. J Oral Maxillofac Surg 2007;65:103-6. 30. Lindemeyer R, Wadenya R, Maxwell L. Dental and anesthetic management of children with dystrophic epidermolysis bullosa. Int J Paediatr Dent 2009;19:127-34. 31. Martinez L, Goodman P, Crow WN. Squamous cell carcinoma of the maxillary sinus and palate in epidermolysis bullosa: CT demonstration. J Comput Assist Tomogr 1992;16:317-9. 32. Baek JO, Lee HY, Oh SW, Lee JS, Kim SC, Lee JR, et al. A novel homozygous keratin 14 mutation in a patient with autosomal recessive epidermolysis bullosa simplex and squamous cell carcinoma of the tongue. Br J Dermatol 2010;162:880-2. 33. Wright JT, Fine JD, Johnson L. Dental caries risk in hereditary epidermolysis bullosa. Pediatr Dent 1994;16:427-32.
