ED IT O RIAL Clinical Heterogeneity in Epidermolysis Bullosa: Speculations on Causation and Consequence for Research A recent edit o rial. in thi s j 'o urn .al p.r escnted an extensive review on the role of ancho nn g fibnl changes 111 dystrophic epide rm olysis bullosa (EBD), s umm ari z- ing electron micros cop ic (EM), biochcmt ca l, and other eviden ce I 1] . In cluded vverc hi g hl y rele va nt pomt s on clinical diagnosi s, not the least of which was the difficulty in disc riminatin g the local fo rm s of the recessive t ype (R-EB D) from s oradic d om ina nt mutant (D- EUD) cases. T he focus her e w tll :e on the clinical variations in diff erent ca te go ries of EB, thetr poss ibl e ca usat ions, and conse quen ces . . . The clini ca l ap pearance of EB seems to have no end 111 tts variation . Without kn ow in g the cause for this, we may be unable to r eac h a sufficiently precise diagnosis to pr edict the fate of the ind ividua l patient. H ow mu ch mi sleading advice has been gtven pa re nts of the blistering in fa nt ' We are not fa r rem ove d from the days when pa re nt s of a child with severely blistenn g dom111ant mu ta nt case of the herp et if or m Dow lm g- M ea ra va net y (D-EBH- DM) we re to ld the child was go in g to die and the re curr ence nsk for additional childr en was 25% . Today, by aiJ ava il able c ntena a genera li zed R-EBD case is expected to d eve lop syncc ht ae a nd mi tten hands, yet some cases fa il to do so. Som ct tm cs tht s ts ascr ib ed to the effect of ph enyto in therap y, yet it occ ur s 111 th e abse nce of such therapy I Is this be ca use we do not yct recog ntze rhe inverse-d ys tr ophic ( R-EB D- 1) type in in fa ncy, or becau se of ex te nsiv e tr a um a protection in the first mo nth s of the chtld, or beca use of a peculiar genetic makeup of the mdtvtdu al as, for exa mple , different mutation s in the paternally and mat erna ll y de ri ve d EB genes? Most s urpri ses, howeve r, may be cxpc n enced in the grow ing numb ers of pat ients misdia gnosed 111 the past as havi ng EBD, yet n ow by EM or by Immun o flu orescence to h ave jun ctional blistenn g wtth (or wtth o ut ') hcmtdesm osome ab normalities. O ur goa l today sho uld be no less than to pinp oint the respon- sible mutat ed ge ne, pick it out of the D NA from the mdtvtdua l pat ient , cut it by restric ti on enzymes, and D NA-sequence tt. Th e demo nstrated ch ange on the molecular leve l ca n then be correlated to o ther bi ochemi ca l, mo rph olog ic, and clinical obse rvations. To do so we mu st kn ow precisely eith er where the gene is or what it does. At present we do not know either of t!1 esc facts for any EB ge n e. Nor do we kn ow h ow many sene11 c d1 sease entitles (how ma ny diff ere nt ge ne loc i) arc in vo lved, n or how many types of muta tions in and around each gene loc us (a ll e lt c mu t:ltlons) gtve diffe rent ph enotypic results (di sease 11ariant s). Twe nt y-ftve years ago it was w ritten in textb oo ks that EB was the bes t-d efined genetic disease gr oup in dermat ology . At that time it comprised 3 different diseases. A few years ago we co n- side red a minimum of 9 entities (gene loct) because EM had pinpo inted primary or early changes of 9 diff erent kind s, and 16 different clinicogenctic types when clim ca l stgns, predtlcctton stte s, and co ur se of disease were co nsidered [2]. H owever, the vanattons and clinical co mple xity do n ot cease. What arc the ca uses for tl115 , and ho w d oes this influence research in EB' *Winberg, Ge dd e- Da hl: Gcla tina sc ex press ion in ge nerali zed epider- mo ly sis bu ll osa simpl ex fibrobla sts. Submitted for publication . IN TRAE JJID ER MA L EP IDERMOLYS IS BULLOSA The t yp ica l EB s im plex (D-EBS) pat ien t pr esenting in warm weat her wit h b li stering soles, toes, and so metime s palm s and fi ngers, ma y te ll abo ut having had occasional blisters elsew here in childh ood (mild ge nera lized Kobm:r type, EBS-K ), or deny this (loca li ze d Webe r- Cockay ne type, EBS - WC), and on ly ra rely describe persistent generali zed b li ster li ability (severe Kobner type, EBS-K). Low intr a- and hi gh inre rf amilial va ri ations indicate dif - fe rent mutations in diff ere nt fa milies I 3]. T here is as yet no proof that these occ ur in more than one gene locus. and linkage studies hav e only suggested this loc us (EBS2) is on chrom osome 1 [ 4]. H oweve r, the pooled EJ3S-K plus EBS-W C mutati on rate of 2 in one mi lli on gametes per ge ner at io n is 4 rim es hi gher than for any oth er dominant EB-t ype [21 and go nadal mo sa icism for new mut at ions is relatively frequent as disclosed by 2 or more patients in the ft rst affected gene ration (Amon -La mpre cht, Gedde- Dah l, Ledaa l, unpubli shed obse r vatio ns) [3). The EBS2 locus mu st fa ll into a very special ge ne ri c ca tegor y, or ge netic heteroge neity ex- ists. Low expressio n of fibroblast gelatinase activity has been re port ed in 3 of 9 D-EBS- WC cases [5 J, 3 of 9 D- EBS-K cases* [5]. but in none of 40 n on - EBS persons* [51. T he stgmfi ca n ce of th.is is unclear. Possible clinical hint s of heteroge nei ty ca n be that so me , but not all , families sh ow dram atic improveme nt with age; some , but not a ll , re port their serous blisters fr equ ently turn hemorrha gic; and rare cases displa y palm op lant ar ery th ema and initial blistering in palm bef ore foot blisterin g.* Multipl e erosions on the hands sho ul d alert the cl inician to as k about epiderma l fragili ty since t hi s is typical for a distinct entity, D-EBS - Og na 13]. caused by mu tation at the EBS ·f locus w hi ch is linked to the glutamic pyru vic transaminase loc us (C PT) on chr om oso me 8 or '16 (2]. The adolescent or adult pat ie nt who may clinica ll y be suspected of having EBS, but who f.1 il s to exacerbate in war m weat her and who had severe b li stering in in fa ncy and preschoo l yea rs whe n f eve r cleared the skin, wi ll in all prob ability by EM show basa l cell s ubnucl ea r tonofib rill ar aggrega ti on in pe ri lcsional skin bio p- sies. Epidermo l ys is bullosa herpc tif o rmi s Dow lin g-M cara (D- EB H- DM) is definitely a disease separate from the basal ce ll cy- to lys is EBS 12]. In inf ancy it is confused w ith j uncti ona l EB and in chil dhood with dys tr op hi c EB . Las tly, Ba rt's type of EB seems to f.1 ll in the intraepidcrma l group [2 ], but EM needs to be perf o rmed on the original fa mily for verifica ti on. JUN CTIONAL E PIDERMOLYSIS BULLOSA Fo ll ow ing Hcrlitz's de sc ription in 1935 of EB Ictal is as no nscar- ring and let hal within 3 mo nth s, and Pearson's dis cove ry in 1962 th at b li sters fo rmed in the int e rm e mbran e Uu nctional) space be- twee n the basa l ce ll plasma membrane and the e le ctro n-d ense de rm al membrane , jun ctional EB (EBJ) was for a long time co n- sidered to be synonymous wit h Herlitz's disease wi th occasional s ur vi vo r s. T he abno rmalit y of hemi desmosomcs , first rccogmzed by Has himot o ct al in 1976 (2 ], then all owed identification ofEBJ in 2 ways: either by a jun ctional split or, even in nonb listcrcd skin, by la ck of a s ubb asal dense plate and often other abno r- malities of the hemidesm oso m es. How ever, this was soo n fo und in patie nt s clinica ll y separate from the Herlit z type. A signifi cant po int was that several of the non-H crlitz EBJ pati ents were of con sang uineous parenta ge , indi ca tin g both parents ea rn ed the 0022-202X/86/$03.5U Copy ri ght © 1986 by T he Society for In ves ti gat iv e Dermatology. In c. 91