REVIEW Inherited epidermolysis bullosa: Updated recommendations on diagnosis and classification Jo-David Fine, MD, MPH, FRCP, a,b Leena Bruckner-Tuderman, MD, PhD, c Robin A. J. Eady, DSc, FRCP, FMedSci, d Eugene A. Bauer, MD, e Johann W. Bauer, MD, f Cristina Has, MD, c Adrian Heagerty, MD, FRCP, g Helmut Hintner, MD, f Alain Hovnanian, MD, PhD, h Marcel F. Jonkman, MD, PhD, i Irene Leigh, CBE, DSc, FRCP, FRSE, FMedSci, j M. Peter Marinkovich, MD, e,k Anna E. Martinez, FRCPCH, l John A. McGrath, MD, FRCP, FMedSci, d Jemima E. Mellerio, MD, FRCP, d,l Celia Moss, DM, FRCP, MRCPCH, m Dedee F. Murrell, MD, FACD, n Hiroshi Shimizu, MD, PhD, o Jouni Uitto, MD, PhD, p David Woodley, MD, q and Giovanna Zambruno, MD r Nashville, Tennessee; Freiburg, Germany; London and Birmingham, England; Stanford, Palo Alto, and Los Angeles, California; Salzburg, Austria; Toulouse, France; Groningen, The Netherlands; Dundee, Scotland; Sydney, Australia; Sapporo, Japan; Philadelphia, Pennsylvania; and Rome, Italy Background: Several new targeted genes and clinical subtypes have been identified since publication in 2008 of the report of the last international consensus meeting on diagnosis and classification of epidermolysis bullosa (EB). As a correlate, new clinical manifestations have been seen in several subtypes previously described. Objective: We sought to arrive at an updated consensus on the classification of EB subtypes, based on newer data, both clinical and molecular. Results: In this latest consensus report, we introduce a new approach to classification (‘‘onion skinning’’) that takes into account sequentially the major EB type present (based on identification of the level of skin cleavage), phenotypic characteristics (distribution and severity of disease activity; specific extracutaneous features; other), mode of inheritance, targeted protein and its relative expression in skin, gene involved and type(s) of mutation present, andewhen possibleespecific mutation(s) and their location(s). Limitations: This classification scheme critically takes into account all published data through June 2013. Further modifications are likely in the future, as more is learned about this group of diseases. Conclusion: The proposed classification scheme should be of value both to clinicians and researchers, emphasizing both clinical and molecular features of each EB subtype, and has sufficient flexibility incorporated in its structure to permit further modifications in the future. ( J Am Acad Dermatol 2014;70:1103-26.) Key words: classification; diagnosis; electron microscopy; epidermolysis bullosa; gene; genetics; monoclonal antibodies. From the Vanderbilt University School of Medicine, Nashville a ; National Epidermolysis Bullosa Registry, Nashville b ; University Medical Center Freiburg c ; St John’s Institute of Dermatology, King’s College London and Guy’s and St Thomas’ Hospital National Health Service Foundation Trust d ; Stanford University School of Medicine e ; Paracelsus Private Medical University, Salzburg f ; Heart of England Foundation Trust, Birmingham g ; INSERM and Department of Genetics, Toulouse h ; University Medical Center Groningen, University of Groningen i ; University of Dundee j ; Dermatology Service, Palo Alto Veterans Affairs Medical Center k ; Great Ormond Street Hospital for Children National Health Service Foundation Trust, London l ; Birmingham Children’s Hospital and University of Birmingham m ; St George Hospital and University of New South Wales, Sydney n ; Hok- kaido University School of Medicine, Sapporo o ; Thomas Jeffer- son University, Philadelphia p ; University of Southern California, Los Angeles q ; and Istituto Dermopatico dell’ Immacolata, IDI-IRCCS, Rome. r Funding sources: None. (Although the authors have acknowl- edged in other unrelated publications their extramural support for their own epidermolysis bullosaerelated research pro- grams, none of these has provided funding for the Consensus Conference or the generation of this report.) Conflicts of interest: None declared. Accepted for publication January 25, 2014. Reprints not available from the authors. Correspondence to: Jo-David Fine, MD, MPH, FRCP, Division of Dermatology, c/o Vanderbilt Health One Hundred Oaks, Dermatology Suite 26300, 719 Thompson Ln, Nashville, TN 37204. E-mail: [email protected]. Published online March 31, 2014. 0190-9622/$36.00 Ó 2014 by the American Academy of Dermatology, Inc. http://dx.doi.org/10.1016/j.jaad.2014.01.903 1103
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
REVIEW
Inherited epidermolysis bullosa: Updatedrecommendations on diagnosis and classification
Jo-David Fine, MD, MPH, FRCP,a,b Leena Bruckner-Tuderman, MD, PhD,c
Robin A. J. Eady, DSc, FRCP, FMedSci,d Eugene A. Bauer, MD,e Johann W. Bauer, MD,f Cristina Has, MD,c
Adrian Heagerty, MD, FRCP,g Helmut Hintner, MD,f Alain Hovnanian, MD, PhD,h
Marcel F. Jonkman, MD, PhD,i Irene Leigh, CBE, DSc, FRCP, FRSE, FMedSci,j M. Peter Marinkovich, MD,e,k
Anna E. Martinez, FRCPCH,l John A. McGrath, MD, FRCP, FMedSci,d Jemima E. Mellerio, MD, FRCP,d,l
Jouni Uitto, MD, PhD,p David Woodley, MD,q and Giovanna Zambruno, MDr
Nashville, Tennessee; Freiburg, Germany; London and Birmingham, England; Stanford, Palo Alto, and
Los Angeles, California; Salzburg, Austria; Toulouse, France; Groningen, The Netherlands; Dundee,
Scotland; Sydney, Australia; Sapporo, Japan; Philadelphia, Pennsylvania; and Rome, Italy
From
N
M
Ki
N
Sc
Sa
IN
M
of
M
N
C
H
ka
so
Lo
ID
Background: Several new targeted genes and clinical subtypes have been identified since publication in 2008of the report of the last international consensusmeeting ondiagnosis and classification of epidermolysis bullosa(EB). As a correlate, new clinical manifestations have been seen in several subtypes previously described.
Objective: We sought to arrive at an updated consensus on the classification of EB subtypes, based onnewer data, both clinical and molecular.
Results: In this latest consensus report, we introduce a new approach to classification (‘‘onion skinning’’)that takes into account sequentially the major EB type present (based on identification of the level of skincleavage), phenotypic characteristics (distribution and severity of disease activity; specific extracutaneousfeatures; other), mode of inheritance, targeted protein and its relative expression in skin, gene involved andtype(s) of mutation present, andewhen possibleespecific mutation(s) and their location(s).
Limitations: This classification scheme critically takes into account all published data through June 2013.Further modifications are likely in the future, as more is learned about this group of diseases.
Conclusion: The proposed classification scheme should be of value both to clinicians and researchers,emphasizing both clinical and molecular features of each EB subtype, and has sufficient flexibilityincorporated in its structure to permit further modifications in the future. ( J Am Acad Dermatol2014;70:1103-26.)
Each of the many subtypes of inherited epider-molysis bullosa (EB) is currently defined by its modeof transmission and a combination of phenotypic,ultrastructural, immunohistochemical, and molecu-lar findings.1 Much has been learned about thenatural history and etiopathogenesis of EB sincethis disease was first formally classified, based on
CAPSULE SUMMARY
d Many new genes and clinical phenotypesof inherited epidermolysis bullosa havebeen characterized since the lastinternational consensus meeting ondiagnosis and classification waspublished in 2008.
d A new classification system (‘‘onion skin’’approach) has been created thatsequentially takes into account theepidermolysis bullosa type, mode ofinheritance, phenotype,immunofluorescence antigen mappingfindings, and mutation(s) present in eachpatient.
d Detailed summaries on the typicalfindings in each recognizedepidermolysis bullosa subtype areprovided for use by clinicians andresearchers.
electron microscopic fea-tures described in 1962,2 asa result of the application ofepidemiology, cell biology,immunology, and molecularbiology to the study of largenumbers of clinically well-characterized patients.
Four international con-sensus meetings on its diag-nosis and classification havebeen held since 1988, the lastin Vienna, Austria, in 2007.1
Since then, several new phe-notypes and causative geneshave been identified. In June2013 a number of leadingEB experts met in London,United Kingdom, to reviewthe collective data on thisdisease and to reformulatethe means whereby patientswith EB can be subclassified,with increasing focus on themolecular origin of each sub-
type whenever possible.
MAJOR EB TYPESWe recommend that the currently used names for
the 4major EB typeseEB simplex (EBS), junctional EB(JEB), dystrophic EB (DEB), andKindlerebe retained,so as to maintain continuity with decades of clinicaland basic scientific literature and to prevent anyconfusion or ambiguity arising among patients; EBsupport organizations; medical, nursing, and otherclinical colleagues; governmental agencies; andthird-party insurers. Changing those names wouldadd little or no value and might prove counterpro-ductive to the diagnosis and care of these patients.
EBS encompasses all subtypes of EB havingmechanical fragility and blistering confined to theepidermis. When the classification system was lastrevised, EBS was further separated into suprabasaland basal subgroups, based on the histopathologicsite of cleavage within the epidermis.1 In the pastEBS was referred to by some as ‘‘epidermolytic’’although this term is inaccurate, because cellularlysis is not a primary feature of any type or subtype of
EB. JEB includes all subtypes of EB in which blistersdevelop within the mid portion or junction, the so-called lamina lucida, of the skin basementmembranezone (BMZ). DEB (in the past referred to occasion-ally as ‘‘dermolytic’’ EB) includes all EB subtypes inwhich blistering occurs within the uppermost dermis(ie, just beneath the lamina densa of the skin BMZ).
Finally, Kindler syndrome,which was added to the EBclassification in 2008, de-scribes a specific entity thatis characterized by the pres-ence of clinical phenotypicfeatures unique among EB(most notably photosensi-tivity) and blistering thatarises in multiple levelswithin and/or beneath theBMZ, rather than within adiscrete plane, as occurs inall other EB types.
DIAGNOSTIC TESTINGIN EB
Diagnostic testing andclassification in EB beginwith the identification of thelevel of skin cleavage viaimmunofluorescence anti-gen mapping (IFM) and/ortransmission electron micro-scopy on preferably newly
induced blisters. The use of monoclonal antibodiesdirected against components of the skin BMZ andepidermal antigens can further facilitate subclassifi-cation, because skin samples from most of the EBsubtypes vary in the intensity of antigen staining (ifeven present) of specific structural proteins, corre-sponding to the presence of a mutation within itsassociated gene. Details of the immunohistochem-ical and ultrastructural findings in each of the majorEB subtypes may be found in our last consensusreport, published in 2008.1
Once the level of skin cleavage and the antigenstaining profile have been determined, pursuit ofmutational analysis is recommended, if available andaffordable, because this will permit the most precisesubclassification. The latter information will becomecritical for genetic counseling in the future, if andwhen molecular treatments become a reality. Fornow, molecular fingerprinting provides the mostprecise way of ascertaining mode of transmission,enabling the clinician to accurately perform geneticcounseling. Reliance solely on mutational findingsfor accurate clinical prognostication, however, must
be done with care, because considerable variationmay exist in disease severity and the natural historyof patients within even a single EB subtype or
Fig 1. A schematic representation of the epidermlocation of specific proteins pertinent to the paththe level in which blisters develop in different EBthe epidermis, the basal keratinocytes, and abo(spinous and granular layers), which are coveredattached to the dermis by the bilayered basemenlamina densa (red bar). On the left, the level of b(EBS ) suprabasal, the blisters form within the miwhich protein is mutated. In EBS basal, the cleavajunctional EB (JEB), the separation takes place wi(DEB), within the sublamina densa region within(KS ), cleavage can occur within the basal keratinbelow the lamina densa. On the right, the localizindicated. Transglutaminase 5 is present in thePlakoglobin and desmoplakin are desmosomal proplakophilin 1, which is expressed mainly in theplectin, BP230, exophilin 5 and kindlin-1 are foIntegrin a6b4, integrin a3, and collagen XVII aredomains emanating from the plasma membranelucida. Laminin 332 is a lamina lucida protein ananchoring fibrils, is found in the sublamina densa
kindred, because of the influence of environmentaland/or modifying genetic factors.
‘‘ONION SKIN’’ APPROACH TOCLASSIFICATION
Initial attempts at the classification of patients withEB at the bedside are based on personal and familyhistory and the presence or absence of specific clinicalfeatures, both cutaneous and extracutaneous. Onlylater, once more sophisticated laboratory testing hasbeen performed, is it possible to subclassify thesepatients more accurately. As with any other disease, aphysician sees, listens to, and examines a patientsuspected of having EB; generates a differentialdiagnosis; and thenpursues increasingly sophisticatedlaboratory tests as the needs demand. This approach
is, the skin basement membrane zone, theogenesis of epidermolysis bullosa (EB), andtypes. The scheme depicts the cell layers ofve them the suprabasal keratinocyte layersby the horny layer ( pink). The epidermis ist membrane consisting of lamina lucida andlister formation is indicated. In EB simplexddle/upper epidermal layers, depending onge plain is within the basal keratinocytes. Inthin the lamina lucida, and in dystrophic EBthe uppermost dermis. In Kindler syndromeocytes, at the level of the lamina lucida orations of the relevant mutated proteins areuppermost cell layers of the epidermis.teins that are panepidermal, compared withsuprabasal epidermis. Keratins 5 and 14,
und mainly within the basal keratinocytes.transmembrane proteins with extracellularof the basal keratinocytes into the laminad collagen VII, the major component of theregion.
Table II. Morphologic and molecular features of the major epidermolysis bullosa simplex subtypes6-19
Major EBS subtype Usual level of blister formation Protein affected and pattern of IF staining Mutated gene(s)
*Suprabasal EBS encompasses all EBS subtypes in which blister formation arises above the level of the basal keratinocyte.yBiallelic homozygous nonsense mutations in the desmocollin-3 gene have been associated with hereditary hypotrichosis and recurrent
scalp vesicles.zKRT6C mutations cause focal palmoplantar keratoderma with minimal nail changes and skin blistering.9
xAcral peeling skin syndrome, which arises subcorneally, may also be caused by mutations in the genes encoding cystatin A (CSTA),6,7
N-acetylgalactosamine-4-O-sulfotransferase (CHST8),8 or comeodesmosin (CDSN ) (although the latter 2 are associated with generalized
peeling skin).
Table I. The major epidermolysis bullosa types and subtypes
Level of skin cleavage Major EB type Major EB subtypes Targeted protein(s)
suggests a simple yet elegant means of standardizingnomenclature and subclassification of EB, givenrecent advances in our understanding of this groupof diseases at the molecular level and taken in contextwith clinical findings that have been observed inrobust, well-characterized patient cohorts.
Analogous to peeling an onion, we propose thatthe classification and subclassification of patientswith EB begin with their separation into 1 of the 4major EB groups, based on the level (intraepidermal[EBS]; within [JEB] or beneath [DEB] the skin BMZ;or mixed pattern [Kindler syndrome]) within whichblisters develop (Table I and Fig 1). The next level ofsubclassification takes into account the clinicalphenotypic features present in a given patient,most notably the distribution (localized vs general-ized) and relative severity of cutaneous and extrac-utaneous disease involvement. As a correlate, thepresence of diagnostically useful skin findings (eg,exuberant granulation tissue; mottled pigmentation;
pseudosyndactyly; other) may permit further sub-classification at this level. Each patient then can befurther subclassified on the basis of the mode oftransmission and, if identifiable, by the specific geneinvolved, the latter initially determined by means ofimmunohistochemical techniques (IFM, usingEB-pertinent monoclonal antibodies) and later bymutation analysis (Tables II to VII). In some clinicalsettings, some EB investigators prefer to pursuemolecular screening without first obtaining IFMresults. In the evaluation of a patient with clinicallyobvious severe generalized recessive DEB (RDEB),this is reasonable and cost-effective. This is not thecourse taken by most investigators, however, for EBsubtypes in which multiple genes may causeidentical phenotypes (ie, severe generalized JEB),given the current cost and time involved in screeningmultiple genes. In the future, marked reduction inthe cost of performing massive parallel sequencingof EB targeted genes may argue for more immediate
Table III. Epidermolysis bullosa simplex clinical subtypes
*Rare variants.yNo mutations detected in genes encoding for transglutaminase 5 and collagen type VII in the original patient who was characterized
(JA McGrath, unpublished data, 2013).
J AM ACAD DERMATOL
VOLUME 70, NUMBER 6Fine et al 1107
molecular testing. In the absence of molecular data,in those rare patients on whom the level of skincleavage is ascertained solely by electronmicroscopy, pertinent ultrastructural features(eg, alterations in structure and/or numbers ofanchoring fibrils or hemidesmosomes, or thepresence of abnormal keratin filament aggregates)would be used in classification of the patient.
Schematically, this progressive approach may besummarized as: major EB type / phenotype(severity and distribution) / mode of transmission/ ultrastructural site of cleavage and associatedfindings / protein involved (with or withoutspecific IFM findings listed) / gene involved andmutational type / specific mutation present.
A contracted version of the onion skin classifica-tion scheme (eg, omitting the listing in the patient’sclassification of the mode of transmission and thetargeted protein, once the type of mutation[s] presenthas or have been determined) might be used inroutine clinical settings, based on the particularneeds of the clinician. It is also understood that theremay be considerable overlap among the severitygrades of mild, intermediate, and severe, and thatsome patients with ‘‘mild’’ subtypes of EB may beextremely disabled despite relatively limited areas ofskin involvement. These terms are used to define theusual degree of severity that has been seen within
each of themajor andminor EB subtypes, when largecohorts have been studied.
Table VIII compares this proposed nomenclaturefor several major EB subtypes with that adopted in2008.
The following clinical scenarios illustrate how thismight be done in individual patients.
1. A patient with intraepidermal blister formation,blistering confined to the palms and soles, anda family history consistent with autosomaldominant transmission would initially be classi-fied as having localized EBS. Once mutationalconfirmation is completed, the final diagnosis,using the onion skin method, would be: EBS,localized, KRT5 mutation (missense mutation).
2. A child with generalized blistering and skin cleav-age within the lamina lucida of the BMZwould beclassified as having JEB, generalized. In the pres-ence of specific phenotypic findings (eg, markedgrowth retardation, severe anemia, exuberantgranulation tissue, upper airway involvement,and severe intraoral disease) and IFM demonstra-tion of absent staining of laminin-332, thediagnosis would be: JEB, severe generalized,laminin-332 negative. Once mutational confirma-tion is available, final subclassification wouldbecome: JEB, generalized severe, laminin-332
Table IV. Morphologic and molecular features of the major junctional epidermolysis bullosa subtypes
Major JEB subtype
Usual level of
blister formation Targeted protein(s) Protein staining pattern Mutated gene(s)
*Rare variants.yA few patients with JEB and ITGB4 mutations have been reported who had generalized skin involvement and a fatal course but lacked
pyloric atresia. In one of these kindreds, an affected sibling with the same mutation and blistering also had pyloric atresia.zNot all patients with JEB, localized, having laminin-332 mutations have an inversa phenotype (Kiritsi and Has, unpublished data, 2013).xCompound heterozygous mutations in LAMA3 and its isoform a3a (LAMA3A) may occur in some patients who display clinical features of
both JEB, generalized intermediate, and JEB-LOC syndrome.
Table V. Morphologic and molecular features of dystrophic epidermolysis bullosa subtypes
DEB subtype Level of blister formation Protein affect and pattern of IF staining Mutated gene(s)
DDEB (all subtypes except bullousdermolysis of the newborn)
Dermal (sublamina densa) Collagen VII: normal or reduced COL7A1
DDEB and RDEB, bullousdermolysis of the newborn
Dermal (sublamina densa) Collagen VII: granular stainingwithin basal keratinocytes andreduced/absent staining alongthe DEJ during active disease;normal DEJ staining when olderand if becomes clinically inactive
COL7A1
RDEB, generalized severe Dermal (sublamina densa) Collagen VII: absent or markedlyreduced
COL7A1
RDEB, generalized intermediate Dermal (sublamina densa) Collagen VII: reduced COL7A1RDEB, localized Dermal (sublamina densa) Collagen VII: reduced or normal COL7A1RDEB, all other subtypes Dermal (sublamina densa) Collagen VII: reduced COL7A1
3. An adult with generalized blistering and subla-mina densa blistering would be initially classifiedas having generalized DEB. A history of identicaldisease in 2 preceding generations suggests
dominant DEB. Final subclassification might be:dominant DEB, generalized, COL7A1 mutation,mutation type specified (ie, missense mutation).
4. A child with severe generalized blistering andscarring, sublamina densa blistering, growthretardation, severe anemia, marked digital web
Table VI. Dystrophic epidermolysis bullosa clinicalsubtypes
DEB, major
subtypes All subtypes
Targeted
protein(s)
DDEB DDEB, generalized (DDEB-gen)DDEB, acral (DDEB-ac)*DDEB, pretibial (DDEB-pt)*DDEB, pruriginosa (DDEB-pr)*DDEB, nails only (DDEB-na)*DDEB, bullous dermolysis ofthe newborn (DDEB-BDN)*
formation, esophageal strictures, and lack of apositive family history would initially be classi-fied as having RDEB, generalized severe. Aftercompletion of IFM and mutational analysis, finalsubclassification would be: RDEB generalizedsevere, collagen VII absent, COL7A1 nonsense/nonsense mutations.
5. A patient clinically meeting the diagnosis ofinverse RDEB would eventually be subclassifiedas: RDEB inversa, collagen VII reduced, COL7A1mutations (ie, nonsense/missense mutation).
It should be emphasized that lack of knowledgeas to the specific type(s) of mutation present doesnot preclude classification, because the onionskin approach reveals only as many layers as canbe elucidated in a given patient. For example, anewborn with JEB having decreased laminin-332staining and yet lacking informative phenotypicfeatures may be best classified as JEB generalized,decreased laminin-332 staining, until additional databecome available. Such patients in the past wereclassified by some experts as having generalized JEB,indeterminate subtype, until further clinical orlaboratory findings permitted more accurate sub-classification. Similarly, before mutational analysis, aseverely affected autosomal recessive case withgeneralized skin and extracutaneous disease activityand absent staining of collagen VII in the skin would
be classified as: RDEB generalized severe, collagenVII absent. It should be noted that although inclusionof the specific mutational type and site of mutationwould be truly definitive, issues related to patientprivacy may prevent use of such data in any format,most notably publication, that can be accessed byanyone other than the patient, his or her physician,and the diagnostic laboratory.
Finally, it is understood that it may be cumbersometo use complete naming in our routine interactionswith patients and their referring physicians. A severelyaffected patient with proven nonsense/nonsensemutations in COL7A1will still be referred to as havingRDEB generalized severe. The more precise onionskin designation of each patient, however, providesdocumentation to support suchadiagnosis andwill beinvaluable for clinical trials andbasic research, and thepublication of such results.
ELIMINATION OF EPONYMS AND OTHERPROPOSED NAME CHANGES
Since the first description of EB in 1886, as new EBphenotypes were recognized it became a commonpractice to attach the name(s) of the clinician(s) whofirst reported them. A similar tradition has existed inmany medical and surgical specialties in the namingof rare syndromes and surgical procedures.Unfortunately, although some eponyms immediatelybring to mind specific phenotypic features, and theiruse honors those who astutely first recognized theseconditions, eponyms have no immediate descriptivevalue, making their comparison with other relatedentities difficult for clinicians who are not experts ona given group of diseases. As a result, manyauthorities and journals now discourage thecontinued use of eponyms.
We recommended in previous consensus reportsthat several eponyms, to include Weber-Cockayne,Koebner, Hallopeau-Siemens, and Bart, be elimi-nated and substituted with descriptive terms (EBSlocalized, EBS generalized other, RDEB generalizedsevere, and EB with congenital absence of skin,respectively). The eponyms Cockayne-Touraine andPasini were previously dropped from the subclassi-fication of dominant DEB, because clinical andmolecular data increasingly suggested that such aseparation was artificial. In an effort to be consistent,we recommend that all remaining EB eponyms beeliminated. The only exceptions are Kindlersyndrome and EBS-Ogna, for lack of any bettersuggested names. JEB Herlitz will be renamed asJEB generalized severe and EBS Dowling-Meara willbecome EBS generalized severe. Table IX lists oldand new names for all EB subtypes previouslyassociated with eponyms. We have also listed within
Table VII. Types of mutations known in the major epidermolysis bullosa subtypes
EB type EB subtype Mutated genes Types of mutations known
EBS EBS, suprabasal TGM5 MS, Del, IndelsDSP NS, Del, MSPKP1 Spl, NS, Del, Indels, InsJUP NS, Spl
EBS, basal KRT5 MS, Del, Spl, NS, IndelsKRT14 MS, Del, NS, Spl, Indels, InsEXPH5 Del, NS, InsPLEC NS, Del, Ins, Spl, Indels, MSDST NS
JEB JEB, severe generalized LAMA3 NS, Del, SplLAMB3 NS, Del, Spl, InsLAMC2 NS, Del, Spl, Indels
JEB, generalized/localized LAMA3 MS, NS, Spl, InsLAMB3 MS, NS, Spl, Del, Ins, IndelsLAMC2 NS, Del, Indels, Ins, SplCOL17A1 NS, Del, Spl, Ins, MSITGB4 Del, Spl, MS
JEB, late onset COL17A1 MSJEB with pyloric atresia ITGB4 NS, MS, Del, Spl, Ins, Indels
ITGA6 Del, MS, NS, SplJEB with respiratory and renal involvement ITGA3 MS, Del, SplJEB-LOC syndrome LAMA3A Ins, NS
DEB RDEB, severe generalized COL7A1 NS, Del, Spl, Ins, Indels, MSRDEB, generalized and localized COL7A1 MS, NS, Del, Spl, Ins, IndelsDDEB (all subtypes) COL7A1 MS, Spl, Del
In many cases with recessive inheritance, 2 different mutations are present in 1 individual (compound heterozygosity).
This table depicts the common mutation constellations, but it is not exhaustive (source: Human Gene Mutation Database Professional
2013.2 [http://www.hgmd.cf.ac.uk/ac/index.php]). For the sake of simplicity, some very rare mutation constellations have been excluded.
Whenever possible, the order of the mutation types reflects their representation.
Of note, the new technologies of parallel sequencing using gene panels or exome sequencing allow screening for mutations in several
genes at the same time. As they become more widely available and at lower cost, they will likely replace at least some of the candidate gene
approaches for molecular genetic diagnosis of EB.
J AM ACAD DERMATOL
JUNE 20141110 Fine et al
our clinical summary tables themost commonly usedolder names for each subtype, to facilitate linkage tothe older literature. We realize that some cliniciansmay still choose to continue to use some of theseolder names. The changes that we are proposing inthis consensus report are therefore only recommen-dations from a large group of experts, composed ofboth clinical and basic investigators, who see, study,and treat large numbers of patients with EBworldwide.
When some EB eponymswere eliminated in 2008,we used the term ‘‘generalized other’’ to encompasspatients who had generalized disease activity butlacked the phenotypic and laboratory findings ofthose subclassified as having generalized severedisease. To clarify this and provide a less awkwardname than ‘‘other,’’ those subtypes referred to as‘‘generalized other’’ will be named ‘‘generalized
intermediate’’ to distinguish them from those withgeneralized severe disease activity. We alsorecommend elimination of the term ‘‘lethal’’ forpatients with acantholytic EBS; although JEB Herlitzis oftentimes also fatal in infancy or early childhood,the same term used synonymously with it added littlebut increased anxiety to the parents of a newborn.
UPDATED SUMMARIES OF THE CLINICALAND LABORATORY FINDINGS IN EBSUBTYPES
Tables X to XIII summarize the clinical findings oneach of the major EB subtypes. Similar tables on theminor or rarer subtypes may be found in Tables XIVto XXVI. Although we have made every effort toensure that the descriptions are correct for eachsubtype, we recognize that new clinical findings orassociations may be reported in the future and that
differences exist even among those having the sameEB subtype. Whenever possible, we have describedthose differing observations in footnotes in thetables. Sometimes striking differences in overalldisease severity and/or clinical findings may beseen within a single kindred in whom the samemutation is present, the result of environmental orepigenetic factors, or the influence of modifyinggenes. Although these tables represent our bestcomposite pictures of patients with EB within eachmajor and minor subtype, some differences may benoted in individual patients. They should be used asguidelines when assessing a patient but not asabsolute phenotypic criteria.
OTHER REPORTED BUT NOT YETADOPTED EB SUBTYPES
The recent literature contains several case reports,based on the findings in 1 or only a small number ofpatients, proposing the existence of other EBsubtypes. We critically reviewed each of thesereports and have incorporated as new entities thosethat presented sufficient documentation to meritinclusion at this time. Others must await morerigorous confirmation. As an example, 2 childrenreported as having pretibial EB and renal diseasewere shown to have mutations in the gene encodingfor CD151, a member of the tetraspanin familyknown to be a component of the hemidesmosome.3
Our review of the available information failed todocument sufficient electron microscopic andimmunohistochemical data to allow us to confirmthe precise level of blister formation within the skinBMZ that would permit us to distinguish betweenJEB and DEB, a major concern because pretibialEB is considered to be a subtype of DEB. Morerigorous characterization will hopefully resolve this.A pedigree was reported of patients with vesiclesresulting from a homozygous nonsense mutationin the DSC3 gene, encoding the desmosomalglycoprotein desmocollin-3, but no definitive clinicalor histopathologic evidence of blistering waspresented; the clinical images showed hypotrichosisand keratosis pilaris and histopathology revealedmild follicular plugging only.4,5 To date, there is
Table X. Clinical summary of epidermolysis bullosa simplex localized, generalized severe, and generalizedintermediate subtypes
EBS, localized EBS, generalized severe EBS, generalized intermediate
Former eponyms or names Weber-Cockayne EBS, Dowling-Meara; EBSherpetiformis
EBS, generalized other;non-Dowling-Meara,
EBS, KoebnerMode of transmission AD AD ADOnset (usual) Early childhood Birth BirthSkin distribution (predominant) Palms and soles Generalized (with relative
Other None Arciform (‘‘herpetiform’’)blistering; EB nevi (rare)
EB nevi (rare)
Relative inducibility of blisters Common Common CommonExtracutaneous involvement*Anemia Absent Variable AbsentGrowth retardation Absent Common AbsentOral cavitySoft-tissue abnormalities Erosions in about
25% in infancyCommon Variable
Enamel hypoplasia Absent Absent AbsentCaries Normal frequency Normal frequency Normal frequency
Risk* by age 30 y ofSquamous cell carcinoma None None NoneMalignant melanoma None None Very rareBasal cell carcinoma None None NoneDeath related to EB None Uncommon None
insufficient tissue or other laboratory studies toadequately support the diagnosis of EB.
For inclusion into further revised classificationschemes, any newly proposed EB subtype will needto have sufficient clinical, ultrastructural, immuno-histochemical, and ideally molecular characteriza-tion to demonstrate that it is sufficiently differentfrom previously described EB subtypes.
SUMMARYThis revised classification of EB incorporates
several new genetic subtypes. Where possible wehave replaced eponyms with descriptive terms. We
hope our proposed systematic onion skin approachusing successive layers of clinical, immunohisto-chemical, and molecular findings will prove usefulto both clinicians and researchers, and adaptable tofuture discoveries.
The authors wish to thank DEBRA International for itscontinued generous support of basic and clinicalresearch and international EB conferences such as this,its many member DEBRA organizations, and our thou-sands of patients worldwide whose patience, loyalty,and cooperation over many decades have permittedcollection of the data embraced by these consensusreports.
Table XI. Clinical summary of junctional epidermolysis bullosa generalized severe, generalized intermediate,and localized subtypes20-23
JEB, generalized severe JEB, generalized intermediate JEB, localized
*Relative frequencies: absent or none; rare; 11; 21; 31; 41.yCarriers with LAMA3 null mutations have enamel defects.22 Similarly, a mouse model for JEB has demonstrated that COL17A1 plays a key
role in enamel formation.23
zDeath occurs in about half of those with generalized severe JEB and generalized intermediate JEB within the first 2 y of life, with a further
increase in the cumulative risk of death in the former JEB subtype with increasing age. Although there are a variety of causes of death in
both JEB subtypes during infancy and early childhood, the most common ones are sepsis, upper airway occlusion, and failure-to-thrive, the
latter primarily arising in generalized severe JEB.xRare patients have had pseudosyndactyly, protein losing enteropathy, profound failure to thrive, low birth weight, and/or early death.kThe cumulative lifetime risk of squamous cell carcinoma has been estimated to be 18% in JEB generalized severe (per National EB Registry
data20), whereas cross-sectional analysis of the Groningen, The Netherlands, JEB cohort has revealed the presence of these tumors in
approximately 25% of those with generalized intermediate JEB.21
*Relative frequencies: absent; rare; 11; 21; 31; 41.yWhen present, the digital webbing is usually localized or partial, as contrasted with RDEB generalized severe, where the deformities are
usually bilateral and severe, with eventual fusion of all digits, marked flexion contractures, and bony resorption.zPer data from the US National EB Registry.20
xOf note, enamel abnormalities have been recently observed via scanning electron microscopy in the teeth in a murine model for RDEB
generalized severe. Although not as yet documented at the clinical level in human beings, as they have in junctional EB teeth, such
microscopic defects may explain the propensity of these patients for developing severe caries during early childhood.kAlthough rare cases have been reported to occur at an earlier age, data from the US National EB Registry have shown that the cumulative
risk of cutaneous squamous cell carcinomas arising in generalized severe and generalized intermediate RDEB subtypes usually begins during
the mid to late teenage years and then increases thereafter, strikingly so in the former RDEB subtype, with death from metastases from
squamous cell cancer occurring in the majority of those with generalized severe RDEB within the first 5 y after diagnosis and surgical
removal of the first primary tumor.{Documented rarely by age 12 y in this RDEB subtype, and with no further increased cumulative risk thereafter.#With the exception of rare cases succumbing to sepsis during infancy, the risk of premature death in patients with generalized severe and
intermediate subtypes of RDEB is most often a result of metastatic squamous cell carcinoma in young adulthood or later life. Death may also
result from CRF in a small minority of these patients.
J AM ACAD DERMATOL
JUNE 20141114 Fine et al
Table XIII. Clinical summary of Kindler syndrome24
Kindler syndrome24
Mode of transmission AROnset BirthSkin distribution(predominant)
*Scale: absent or none; rare; 11; 21; 31; 41.yBased on the findings seen in the original well-characterized proband; subsequently, the other autosomal dominant kindred was proven to
have dominant dystrophic epidermolysis bullosa.zDeath usually occurs in children with acantholytic EBS in the neonatal period. As such, the possibility of the occurrence of extracutaneous
complications (eg, anemia) typically seen in older children with some severe epidermolysis bullosa subtypes cannot be fully excluded, were
a child to survive for several years, because only limited numbers of cases with this rare subtype have been reported.
J AM ACAD DERMATOL
JUNE 20141116 Fine et al
causes lethal acantholytic epidermolysis bullosa. Am J Hum
Genet 2005;77:653-60.
12. Hobbs RP, Han SY, van der Zwaag PA, Bolling MC, Jongbloed
JD, Jonkman MF, et al. Insights from a desmoplakin mutation
identified in lethal acantholytic epidermolysis bullosa. J Invest
Table XVI. Clinical summary of epidermolysis bullosa simplex skin fragility syndromes16,25,26
Skin fragility-woolly hair
syndrome (DSP mutations25)
Skin fragility-ectodermal
dysplasia syndrome
(PKP1 mutations)
Skin fragility, plakoglobin
deficiency (JUP mutations26)
Mode of transmission AR AR AROnset (usual) Birth or early infancy Birth BirthSkin distribution (predominant) Generalized Generalized GeneralizedSkin findings (frequency*)Blisters Superficial erosions and crusts,
not blistersSuperficial erosions and
crusts; blisters lesscommon
21 Erosions
Milia Absenty Absent AbsentAtrophic scarring Absenty Absent AbsentDystrophic or absent nails 21 41 41Granulation tissue Absent Absent AbsentScalp abnormalities Hypotrichosis, woolly hair Hypotrichosis, woolly hair Hypotrichosis, woolly hairKeratoderma Focal, punctate or striate, with
fissuringFocal, with fissuring Focal, with fissuring
Other Follicular hyperkeratosis overextensors
Perioral fissuring; circinatescaly erosions
Relative inducibility of blisters Variable Variable VariableExtracutaneous involvement*Anemia 11 Absent AbsentGrowth retardation 11 21 AbsentOral cavitySoft-tissue abnormalities Erosions Tongue fissuring AbsentEnamel hypoplasia None Absent AbsentCaries Normal frequency Normal frequency Normal frequency
Risk* by age 30 y ofSquamous cell carcinoma None None NoneMalignant melanoma None None NoneBasal cell carcinoma None None NoneDeath (all causes) Unknownx Unknown Unknownx
AR, Autosomal recessive; ?, there is some disagreement about whether those specific findings are correct as previously reported in the
literature.
*Scale: absent or none; rare; 11; 21; 31; 41.yMilia and atrophic scarring arose in an area of aplasia cutis on the ankle of 1 affected individual (Jemima Mellerio, MD, FRCP, personal
written communication, 2013).zBlepharitis; absent or sparse eyelashes.xDevelopment of cardiomyopathy might be progressive and lead to death but most reported cases are still children.kEsophageal stricture in 1 patient.
J AM ACAD DERMATOL
VOLUME 70, NUMBER 6Fine et al 1117
the dystonin gene coding for the coiled-coil domain of the
epithelial isoform of BPAG1 underlies a new subtype of
Risk* by age 30 y ofSquamous cell carcinoma None None NoneMalignant melanoma None None NoneBasal cell carcinoma None None NoneDeath related to EB None None None
AD, Autosomal dominant; AR, autosomal recessive; EB, epidermolysis bullosa; EBS, epidermolysis bullosa simplex; ?, there is some
disagreement about whether those specific findings are correct as previously reported in the literature.
*Relative frequencies: absent; rare; 11; 21; 31; 41.yAlthough historically patients with EBS-Ogna were reported to have a tendency for easy bruising, this has not been confirmed by several
other groups more recently, suggesting the possibility that this may not be a valid phenotypic marker of this disease but rather just an
Table XVIII. Clinical summary of epidermolysis bullosa simplex with mottled pigmentation, musculardystrophy, and pyloric atresia subtypes
EBS with mottled
pigmentation EBS with muscular dystrophy EBS with pyloric atresia
Mode of transmission AD AR AROnset (usual) Birth Blisters as early as birth; acral
predominance; musculardystrophy in infancy to
adulthood
Birth
Skin distribution (predominant) Generalized Generalized GeneralizedSkin findings (frequency*)Blisters 41 41 41Milia Rare 21 AbsentAtrophic scarring Absent 2-31 2-31Dystrophic or absent nails 11 41 AbsentGranulation tissue Absent Absent AbsentScalp abnormalities Absent Absent AbsentKeratoderma (palms and soles) 11 Focal Punctate or focal AbsentOther Mottled or reticulate
brown pigmentationNone Widespread congenital
absence of skinRelative inducibility of blisters Variable Variable CommonExtracutaneous involvement*Anemia Absent 11 31Growth retardation Absent 11 31Oral cavitySoft-tissue abnormalities Absent 1-21 31Enamel hypoplasia Absent 1-21 AbsentCaries Normal frequency Normal frequency Normal frequency
*Relative frequencies: absent or none; rare; 11; 21; 31; 41.
J AM ACAD DERMATOL
VOLUME 70, NUMBER 6Fine et al 1119
Table XIX. Clinical summary of autosomalrecessive epidermolysis bullosa simplex BP230deficiency subtype19
EBS, BP230 deficiency
Mode of transmission AROnset Birth or childhoodSkin distribution (predominant) AcralSkin findingsBlisters 21Milia AbsentAtrophic scarring AbsentDystrophic or absent nails 11Granulation tissue AbsentKeratoderma AbsentOther None
Relative inducibility of blisters MildExtracutaneous involvement*y
Anemia AbsentGrowth retardation AbsentOral cavitySoft-tissue abnormalities AbsentEnamel hypoplasia AbsentCaries Normal frequency
*Relative frequencies: absent; rare; 11; 21; 31; 41.yA reported case had neurologic symptoms that may be
attributable to co-existing CADASIL (cerebral autosomal
dominant arteriopathy with subcortical infarcts and
leukoencephalopathy) syndrome.
Table XX. Clinical summary of epidermolysisbullosa simplex exophilin 5 deficiency
EBS, exophilin
5 deficiency
Mode of transmission AROnset Birth or infancySkin distribution (predominant) GeneralizedSkin findings*Blisters Crusts, blistersMilia AbsentAtrophic scarring AbsentDystrophic or absent nails AbsentGranulation tissue AbsentKeratoderma AbsentOther Mild mottled
pigmentarychanges
Relative inducibility of blisters MildExtracutaneous involvement*Anemia AbsentGrowth retardation AbsentOral cavitySoft-tissue abnormalities AbsentEnamel hypoplasia AbsentCaries Normal frequency
Table XXI. Clinical summary of junctional epidermolysis bullosa with pyloric atresia and junctionalepidermolysis bullosa inversa subtypes29
JEB with pyloric atresia JEB, inversa
Mode of transmission AR AROnset Birth BirthSkin distribution (predominant) Generalized IntertriginousSkin findings (frequency*)Blisters 41 31Milia 11 11Atrophic scarring 31 31Dystrophic or absent nails 31 31Granulation tissue Absent AbsentKeratoderma Absent AbsentOther May be associated with large areas of aplasia cutis None
Relative inducibility of blisters Common CommonExtracutaneous involvement*Anemia Variable AbsentGrowth retardation Variable AbsentOral cavitySoft-tissue abnormalities Variable VariableEnamel hypoplasia Present PresentCaries Excessive Increased frequency
Table XXII. Clinical summary of junctional epidermolysis bullosa, late onset, and junctional epidermolysisbullosaelaryngo-onycho-cutaneous syndrome subtypes30
JEB, late onset30 JEB-LOC syndrome
Mode of transmission AR AROnset Young adulthood or later BirthSkin distribution (predominant) Variable Especially facial and neckSkin findings (frequency*)Blisters 41 21, with ErosionsMilia 41 11Atrophic scarring 21 21Dystrophic or absent nails 21 41Granulation tissue Absent 31Keratoderma Mild AbsentOther Hyperhidrosis; absent dermatoglyphs Increased incidence in Punjab
Relative inducibility of blisters* Variable 11Extracutaneous involvement*Anemia Absent 11Growth retardation Absent 11Oral cavitySoft-tissue abnormalities Variable 41 LarynxEnamel hypoplasia Present 31Caries Normal frequency 21
Table XXIII. Clinical summary of junctionalepidermolysis bullosa with respiratory and renalinvolvement subtype31,32
JEB with respiratory and
renal involvement
Mode of transmission AROnset Respiratory and renal
involvement at orshortly after birth;skin features withinthe first months
of lifeSkin distribution (predominant) Legs, buttocksSkin findings*Blisters 21Milia AbsentAtrophic scarring AbsentDystrophic or absent nails 21Granulation tissue AbsentKeratoderma AbsentOther Erosions 11
Relative inducibility of blisters VariableExtracutaneous involvement*Anemia 41Growth retardation 41Oral cavitySoft-tissue abnormalities AbsentEnamel hypoplasia Not applicableCaries Not applicable
Gastrointestinal tract AbsentRenal and genitourinary tract Congenital nephrotic
syndromeOcular AbsentPseudosyndactyly AbsentRespiratory tract Severe respiratory
distress, interstitialpneumopathy
Risk* by age 30 y ofSquamous cell carcinoma Not applicableMalignant melanoma Not applicableBasal cell carcinoma Not applicableDeath related to EB Within the first months
Relative inducibility of blisters Variable Variable VariableExtracutaneous involvement*Anemia Absent Absent RareGrowth retardation Absent Absent RareOral cavitySoft-tissue abnormalities 11 11 11Enamel hypoplasia Absent Absent AbsentCaries Normal frequency Normal frequency Normal frequency
*Relative frequencies: absent; rare; 11; 21; 31; 41.yThere may be an increased risk of squamous cell carcinoma at age[30 y.
J AM ACAD DERMATOL
JUNE 20141124 Fine et al
Table XXV. Clinical summary of dominantdystrophic epidermolysis bullosa localized, nailsonly, and dystrophic epidermolysis bullosaebullousdermolysis of the newborn subtypes
DDEB, localized,
nails only DEB-BDN
Mode of transmission AD AD or AROnset (usual) Childhoody Birth or