gastrostomy tube feeding in children with epidermolysis bullosa: consideration of key issues

Gastrostomy Tube Feeding in Children withEpidermolysis Bullosa: Consideration of Key

Issues

Lesley Haynes, Dip. Diet,* Jemima E. Mellerio, B.Sc., M.D., FRCP,�and Anna E. Martinez, MRCP, MRCPCH�

Departments of *Dietetics and �Epidermolysis Bullosa, Hospital for Children, NHS Trust, London, UK

Abstract: Complications of severe forms of epidermolysis bullosa (EB)almost invariably lead to chronic malnutrition, jeopardizing immune status,growth, iron status, bone health, wound healing, and quality of life. Althoughgastrostomy tube (G-tube) feeding has successfully addressed the difficul-ties of providing nutrition and medications in some children attending ourcenter, others have developed problems such as abdominal distension, poorfeed tolerance, and leakage of gastric contents with persistent localized skinulceration, posing enormous challenges to skin management and nutritionalmaintenance. Suspicions that G-tube placement and feeding cause or ex-acerbate these problems has led to a decline in placements at our center overthe last 10 years. We therefore recognized that it should not be rejectedwithout due consideration of why some patients seem more prone to com-plications than others. Thus, information on selected issues and outcomesof G-tube placement was obtained from records of 66 patients undergoingsurgery between 1989 and 2008. The complex interrelationships of the se-quelae of severe EB, changes in practice over 20 years and lack of data forpatients treated early in the series make it impossible to draw firm conclu-sions at this stage, however, our scrutiny provides valuable information onwhich to base debate and future studies. It also offers well as useful insightsfor fellow professionals involved in nutrition support in children with severeEB.

Chronic malnutrition is extremely common inchildren with severe epidermolysis bullosa (EB) suchas generalized recessive dystrophic EB (RDEB) andjunctional EB (JEB) (1). Strategies to supplement oralnutritional intake are seldom maintained satisfactorilyin the long term (2), and over the past 20 years,

placement of gastrostomy feeding tubes (G-tubes) hasbecome increasingly routine in EB centers worldwideusing differing devices, insertion techniques, andfeeding regimens. The few publications documentingthe outcomes of G-tube placement in children with EBhave described generally reported positive outcomes,

Address correspondence to Lesley Haynes, Dip. Diet, DieteticDepartment, Hospital for Children, NHS Trust, Great OrmondStreet, London, WC1N 3JH, UK, or e-mail: [email protected].

DOI: 10.1111/j.1525-1470.2011.01612.x

� 2012 Wiley Periodicals, Inc. 277

Pediatric Dermatology Vol. 29 No. 3 277–284, 2012

including improved growth (3,4). If complications suchas leakage or poor healing around the insertion sitearose, these were reported as being successfully ad-dressed and not regarded as contraindications to theprocedure. Data published in 1996 by our own center(5) reflected this view, with the greatest numbers ofplacements in our patients being undertaken between1994 and 2001, but since then, concerns regardingcomplications thought to be associated with G-tubeplacement have led to a marked fall in placementnumbers in our patients, with a concomitant increasein the use of esophageal dilatations (EDs). Because G-tube feeding is often the only feasible way that carerscan consistently nourish and medicate their childrenand because children and carers viewed G-tubeplacement as a positive intervention despite the prob-lems, we recognized that it would be unjustifiable todowngrade its role as part of holistic EB managementwithout due investigation of the associated problems.For this reason, and as part of service development, aretrospective appraisal of selected problems with andoutcomes of G-tube placement was undertaken.

METHODS

Aspects of G-tube placement in our patients were iden-tified and collated using information extracted frompatients’ hospital notes and recorded during routineconsultations. These included the following details ofchildren (n = 66)whohadG-tubes placed between 1989and 2008:

• number of placements in each year• EB type, sex, and age at placement• reason(s) for placement and intended usage of G-tube (for feeding, medications, or both)

• type of device placed and method of placement• status at evaluation (whether currently attendingour hospital or other: transferred to other pediatricEB service or adult EB service, lost to follow-up,deceased)

Of those currently attending our center at the time ofevaluation (n = 21), the following additional informa-tion was noted:

• number of years since original surgery for G-tubeplacement

• incidence of pain, distension, leakage at G-tube en-try site in the past week

• number of episodes of insertion site infection in thepast 12 months

• changes in growth centiles since G-tube was placed• children’s and parents’ overall satisfaction with G-tube

Our intention was to highlight important problemsand explore associations, after which more-in-depth re-search would be undertaken.

Results are summarized in Tables 1 and 2.

NUMBERS OF G-TUBE PLACEMENTS, EB

TYPE, SEX, AND AGE

Sixty-four children (30 male, 34 female) underwent G-tube placement at our center between 1989 and 2008, thelargest number of placements being in 1994 and 1997(Fig. 1). Fifty-seven children (89%) had been diagnosedwith severe generalized RDEB, the remaining seven(11%) comprised one child with Herlitz JEB (HJEB),two with non-Herlitz JEB (NHJEB), three with variantsof EB simplex, and one with plakophilin deficiency.Another two boys with severe generalized RDEBunderwent placement elsewhere, with all subsequent careat our center, and they are therefore included in theevaluation. Ages at placement ranged from 5 months to15 years and 10 months (Fig. 2).

Forty-five patients (68%) had transferred to otherUK EB services (pediatric and adult), were lost tofollow-up (overseas patients), or died. Of the remain-ing 21 children (32%) with a G-tube, 19 continued touse it for delivery of food and medications. Theremaining two patients were currently relying on oralintake, but neither they nor their parents were suffi-ciently confident in the adequacy of their long termoral intake to request G-tube removal.

REASONS FOR G-TUBE PLACEMENT AND

INTENDED USAGE

Continual parental stress surrounding oral intake andgrowth failure were major reasons for placement in 54children (82%), of whom 36 (67%) were experiencingintractable painful defecation (with or without consti-pation). Of the total 66 children, 23 (35%) were refusingoral medications and supplements. Two children (3%)had devices placed proactively in an attempt to forestallthese situations, and in an additional two, placementwassolely for administration of medications.

INITIAL DEVICE PLACED, TYPE, LENGTH OF

TIME SINCE PLACEMENT, NUMBER OF

TIMES DEVICE CHANGED, AND REASON FOR

CHANGE

Fifty-one children (77%) received Corpak G-tubes(Corpak MedSystems, Wheeling, IL), 11 (17%) re-ceived Mic-Key G-tubes (Kimberly-Clark, Roswell,

278 Pediatric Dermatology Vol. 29 No. 3 May ⁄ June 2012

GA), and in four children (6%), the type was notrecorded. Early in the series, three devices wereplaced endoscopically, two initially using a Malecotcatheter. Thereafter, primary placement of low-profiledevices using an open operative technique was

employed. Of the children with a G-tube at evalua-tion (n = 19), all had undergone open placement.Five children had had the G-tube for <8 years, ninefor 8 to 10 years, and five for longer than 10 years(Fig. 3).

TABLE 1. Summary of Available Data on 47 Children at Time of Evaluation

Number* EB type SexYr G-tubeplaced

Age atG-tubeplacement, yrs

Reason forG-tubeplacement�

Type ofG-tube

Use ofG-tube Status

3 RDEB m 2000 8.7 1 Mic-Key F, M LFU4 RDEB m 2000 3.9 1, 3 ? F, M LFU5 RDEB m 1995 6.6 1, 3, 4 ? F, M GPE, died6 RDEB m 1989 10.5 1, 3, 4 Cor-Pak F, M Died8 RDEB m 1997 9.4 1, 3 Cor-Pak F, M TAS9 RDEB m 1996 10.8 1, 3, 4 Cor-Pak F, M TAS11 RDEB m 1994 8.9 1, 3 Cor-Pak F, M Died12 RDEB m 2001 7.4 1, 3 Cor-Pak F, M GNIU13 RDEB m 1995 14.3 1 Cor-Pak F, M TAS14 RDEB m 1991 4.7 1 Cor-Pak F, M Died15 RDEB m 1998 14.1 1 Cor-Pak F, M Died17 RDEB m 1996 9.8 1 Cor-Pak F, M TOPS18 RDEB m 1995 10.9 1, 3, 4 Cor-Pak F, M Died19 RDEB m 1992 12.1 1, 3 Cor-Pak F, M Died20 RDEB m 1995 11.3 1, 3 Cor-Pak F, M Died21 RDEB m 1996 0.4 1 Cor-Pak F, M Died22 RDEB m 1994 8.9 2, 3 Cor-Pak F, M Died23 RDEB m 1993 3.0 1 Cor-Pak F, M TAS24 RDEB m 1998 11.4 2 Cor-Pak F, M TAS25 RDEB m 1995 5.8 1 Cor-Pak F, M TAS26 RDEB m 1994 6.2 1, 3 Cor-Pak F, M TAS28 RDEB f 1997 13.5 1 Mic-Key F, M TAS30 RDEB f 1992 5.3 2, 3 Cor-Pak F, M Died31 RDEB f 1996 10.0 1, 3 Cor-Pak F, M TAS33 RDEB f 1999 1.7 1, 3, 4 Cor-Pak F, M Died34 RDEB f 2001 7.0 1, 3, 4 ? F, M Died35 RDEB f 1992 14.8 1, 3 Cor-Pak F, M Died36 RDEB f 1992 6.8 1, 3 Cor-Pak F, M Died37 RDEB f 1996 7.6 1, 3, 4 Cor-Pak F, M Died38 RDEB f 1997 1.9 1 Cor-Pak F, M TAS39 RDEB f 1994 10.8 1, 3, 4 Cor-Pak F, M Died40 RDEB f 1994 11.9 1, 3, 4 Cor-Pak F, M Died41 RDEB f 1994 9.5 1 Cor-Pak F, M TAS45 RDEB f 1993 12.9 2 Cor-Pak F, M TAS46 RDEB f 1991 6.6 1, 3, 4 Cor-Pak F, M TAS48 RDEB f 1995 3.5 1, 3, 4 Cor-Pak F, M Died49 RDEB f 1994 10.6 1, 3 Cor-Pak F, M TAS51 RDEB f 1996 5.7 1 Cor-Pak F, M TAS52 RDEB f 1994 7.9 2, 3 Cor-Pak F, M TAS53 RDEB f 1992 15.8 2 Cor-Pak F, M TAS54 RDEB f 1997 7.3 1, 3, 4 Cor-Pak F, M TAS55 RDEB f 1998 2.0 2 Cor-Pak F, M TOPS56 RDEB f 1990 5.1 1, 3 Cor-Pak F, M TAS60 RDEB f 1998 0.9 1, 3 Cor-Pak F, M Died62 NHJEB m 2003 13.0 1 Cor-Pak F, M TAS63 NHJEB f 1997 1.6 1, 3, 4 Cor-Pak F, M TOPS64 HJEB m 1997 4.2 1 Cor-Pak F, M GNIU

*Number applicable only to this review (not an EB registry number).�1, continual parental stress regarding feeding with growth failure; 2, continual parental stress regarding feeding with adequate growth;3, intractable constipation or painful defecation; 4, refusal of oral medications or supplements.EB, epidermolysis bullosa; G-tube, gastronomy tube; RDEB, severe generalized recessive dystrophic EB; NHJEB, non-Herlitz junctional EB;HJEB, Herlitz junctional EB; f, female; m, male; F, food; M, medications; LFU, lost to follow-up; GPE, G-tube placed elsewhere; TAS,transferred to adult services; GNIU, G-tube in place, but not in use; TOPS, transferred to other pediatric services.

Haynes et al: G-tube Feeding in Children with Epidermolysis Bullosa 279

TABLE2.Sum

maryof

Dataon

19ChildrenUsing

G-Tubes

atTim

eof

Evaluation

PreviousInform

ation

AtEvaluation

Number*

EB

type�

Sex

Yr

G-tube

placed

Ageat

G-tube

placement,

yrs

Reason

for

G-tube�

Type

of

G-tube

Use

of

G-tube

Yrs

since

G-tube

placed

EDs,

nMedications,

n

Problems

withG-tube#

Changein

growth

centiles

since

G-tube

placed**

Overall

satisfaction

withG-

tube�

�

Pain

in past

wks

Distension

in past

wks

Leakage

in past

wks

Infections

inpast

12mos

Weight

Height

Child

Parent

1REBS

m2005

10.0

1?

F,M

4.6

05

00

0None

gg

68

2RDEB

m1993

1.7

1,3

Cor-Pak

F,M

16.1

311

10

65

Constant

––

610

7RDEB

m1999

3.6

1,4

Mic-K

eyF,M

8.9

014

40

0<

2i

i7

10

10

RDEB

m1998

5.1

1,3

Cor-Pak

F,M

10.6

11

70

10

0<

2i

i10

10

16

RDEB

m2001

3.8

2,4

Cor-Pak

F,M

8.3

28

08

8None

ii

10

727

RDEB��

m2002

1.5

1,4

Mic-K

eyF,M

6.7

913

50

7<

2g

i9

929

RDEB

f2001

1.9

1,3,4

Mic-K

eyF,M

8.2

513

33

1None

ii

810

32

RDEB

f2000

0.8

2,4

Cor-Pak

F,M

8.4

316

00

0<

2i

i10

10

42

RDEB

f1997

3.5

2Cor-Pak

F,M

12.4

44

20

0None

ii

10

10

43

RDEB

f1997

3.4

1,3

Cor-Pak

F,M

11.8

713

20

6None

ii

89

44

RDEB

f2000

2.7

1Mic-K

eyF,M

8.4

812

67

0None

gi

610

47

RDEB

f2001

4.7

1,3,4

Mic-K

eyM

9.0

35

010

0None

ii

10

10

50

RDEB

f2003

3.1

2,4

Mic-K

eyM

5.5

810

45

3<

2–

i9

10

57

RDEB

f1999

1.7

1,3,4

Cor-Pak

F,M

9.0

14

15

10

810

2–4

ii

58

58

RDEB

f2001

2.2

1,3

Cor-Pak

F,M

8.1

811

28

10

None

–g

57

59

RDEB

f1998

1.2

2,3

Mic-K

eyF,M

10.7

10

24

75

02–4

ii

59

61

Plakophilin

deficiency

m1999

1.5

1,3,4

Cor-Pak

F,M

9.8

07

05

0<

2g

–10

5

65

EBSDM

m2006

12.8

1,3,4

Mic-K

eyF,M

3.5

02

00

0None

–i

86

66

EBSDM

m2006

2.1

1,3

Mic-K

eyF,M

3.4

06

00

0None

–g

–9

*Number

applicable

only

tothisreview

(notanEBregistrynumber).

�Gastrointestinaltube(G

-tube)

placedelsewhere.

�1,continualparentalstress

regardingfeedingwithgrowth

failure;2,continualparentalstress

regardingfeedingwithadequate

growth;3,intractableconstipationorpainfuldefecation;4,

refusaloforalmedications

⁄supplements.

–Allesophagealdilatations(EDs)

after

G-tubeplacementexceptpatient50,whohadoneED

before

G-tubeplacement.

#0,noproblems;10,extrem

eproblems.

**g,

increase;i,decrease;–,nochange.

��0,extrem

elydissatisfied;10,extrem

elysatisfied.��

GPE,G-tubeplacedelsewhere.

EB,epidermolysisbullosa;REBS,recessiveEBsimplex;RDEB,severegeneralizedrecessivedystrophicEB;NHJE

B,non-H

erlitz

junctionalEB;HJE

B,Herlitz

junctionalEB;EBSDM,

EBsimplexDowling-M

eara;f,female;m,male;F,food;M,medications;LFU,lost

tofollow-up;TAS,transferredto

adultservices;GNIU

,G-tubein

place,butnotin

use;TOPS,

transferredto

other

pediatric

services.


ESOPHAGEAL DILATATIONS

At evaluation,EDhadbeenundertaken in 14of 15 of thechildren with RDEB after G-tube placement, one ofwhom had undergone one ED before. The number ofEDs ranged from two to 14 (mean 7).

MEDICATIONS

At evaluation, all children were taking prescribed sup-plements and medications (range 2–24). These fell intothree broad categories: micronutrient supplements,preparations affectinggastrointestinal (GI) function, andanalgesic and sedative agents.

INCIDENCE OF PAIN, ABDOMINAL

DISTENSION, LEAKAGE, AND INFECTION

AT G-TUBE ENTRY SITE

In the week before evaluation, rated on a scale of 0 to10 (0 = none, 10 = severe), the mean score for painwas 3, although two children rated it 10, and threechildren rated if five and higher. On the same scale,the mean score for distension was 4, with two childrenrating it 10, and eight children rating if five andhigher. On the same scale, the mean score for leakagewas 3, with two children rating it 10, and four childrenrating it five and higher. When leakage occurred, sixchildren reported that it caused a burning or stingingsensation in the surrounding skin, requiring theirdressings to be changed at least twice per day on aregular basis, with one child requiring at least fivechanges per day. In the year before evaluation, infec-tion had been experienced continually in one patient,two to four times per year in two, less than twice peryear in six, and never in 10.

WEIGHT AND HEIGHT CENTILES

Between G-tube placement and evaluation, the weightcentiles increased in four children (21%), remained thesame in five (26%), and decreased in 10 (53%). Heightcentiles increased in three children (16%), remained thesame in two (10%), and decreased in 14 (74%). Allchildren were difficult to measure with accuracy becauseof pain and fixed flexion contractures.

CHILDREN’S AND PARENTS’ OVERALL

SATISFACTION WITH G-TUBE

On a scale of 0 to 10 (0 = very dissatisfied through to10 = extremely satisfied), patients and parents ratedtheir overall satisfaction with G-tube feeding betweenFigure 3. Numbers of years gastronomy-tube in place at

time of evaluation.

Figure 1. Sixty-four gastronomy-tube placements, 1989–2008.

Figure 2. Numbers of gastronomy-tubes placed and ages ofchildren at placement.


5 and 10, with six children and nine parents rating it10 despite commenting that problems such as leakageaffected everyday life and family dynamics substan-tially.

DISCUSSION

Even in childrenwithout EB,G-tube placement has beenassociated with gastroesophageal reflux (GER), woundinfection, excess granulation tissue formation, and tubeleakage (6,7). Consequently, in children with EB, withpreexisting skin fragility, impaired wound healing,chronic inflammation, and potential GI tract involve-ment, such an intervention presents specific challenges(8). Despite this, improved parameters such as growth,bowel function, and overall quality of life (QOL) havebeen documented (3–5), so placement appears to offer apotentially important management choice in the absenceof any other feasible means to provide medium- to long-term nutritional support. The reasons for G-tube place-ment in the majority of our patients were sustainedparental stress regarding feeding, intractable constipa-tion or painful defecation, and refusal of oral medica-tions or supplements: factors that have a highly negativeeffect on QOL. G-tube placement successfully addressedthese problems, with children and parents emphasizingtheir relief at their cessation. The complications reportedby our patients, such as pain, abdominal distension,leakage, and infection around the G-tube insertion site,could reasonably be expected to detract markedly fromQOL, although even children who rated these as >5(more problematic than less) gave an overall satisfactionscore with theG-tube of 5 or higher (satisfied rather thandissatisfied), their parents being almost unanimously inmore-positive agreement. Because optimization of QOLis a major aim in the management of severe EB, animportant question arising fromour study is, ‘‘Could thecomplications experienced by our cohort of children bereduced, thereby improving their, and their parents’,QOL?’’ There is no straightforward answer to thisquestion, but the data from the children using theirG-tube at the time of evaluation provide useful startingpoints fromwhich to consider their relevance topotentialcomplications.

First, what are the criteria for, and aims of, G-tubeplacement in severe EB? Algorithms (9,10) have beendrawn up to aid in the general decision-making process,but they cannot address specific aspects of individualcases. Is early placement (before growth failure becomesestablished) likely to be associated with fewercomplications, better growth, and greater acceptance ofthedevice?Publisheddata (4,5) support this, althoughnolong-term studies have been published, so perhaps

insufficient time had elapsed for complications toemerge. In our experience, complications of severe EBintensify over time, so it would not be not surprising ifthis were true also of those associated with G-tubeplacement. In this context, if early placement jeopardizesthe ‘‘life’’ of the insertion site for reasons currently un-known, is this acceptable if there have been improve-ments in other parameters (e.g., reduction in painfuldefecation and easier administration of medications)?Conversely, if placement is delayed, how much under-nutrition and growth faltering are acceptable before G-tube placement becomes unavoidable, bearing in mindthe inevitable detrimental effects of this state on woundhealing, iron status, and immune status, among otherthings, and the need for adequateweight gain to promotepuberty, with its protective hormonal effect on bonehealth?

Second, how important is growth? In our study, theweight centiles of eight children (42%) were maintainedor increased and those of 11 (58%) decreased. Heightcentiles increased in three children (16%), remained thesame in three (16%), and decreased in 13 (68%). In theabsence of defined growth standards for children withEB, perhaps these results are not out of order in diseasewith this degree of severity. In some children, simplyproviding enhanced nutrition fails to promote longitu-dinal growth and causes central obesity and increased fatmass, which if excessive, can compromise mobility. Suchaltered body composition in severe EBmay be related toa defect in the pathway involving the growth hormoneand insulin-like growth factor 1 (GH ⁄ IGF-1) andhypothalamic–pituitary–gonadal axes, as has been doc-umented in children with other chronic inflammatoryconditions such as juvenile idiopathic arthritis andCrohn’s disease (11). Proinflammatory cytokines,notably interleukin 6, affect the GH ⁄ IGH-1 pathway,leading to growth hormone resistance in children withinflammatory bowel disease (12). Furthermore, serumconcentrations of IGF-1 and IGF binding protein 3(IGF-BP3) were found to be low in a series of childrenwith RDEB (13). Development of fatty liver is a recog-nized complication of chronic undernutrition and sub-sequent overfeeding. Although no post mortems wereperformed on our patients, six-monthly liver functiontests showed no evidence of abnormality.

Third, how relevant are factors such as insertiontechnique, G-tube design, method and frequency of tubechanges, and aftercare of the insertion site and its man-agement? In our patients, the same surgeon wasresponsible for placing virtually all of the G-tubes, andthere were no significant differences in outcomes ofsurgery undertaken by another surgeon. Insertion tech-niques used in EB centers internationally include open


surgical placement, percutaneous endoscopic placement,laparoscopically assisted endoscopic placement, andmore recently, nonendoscopic percutaneous image-guided placement (14) using different device designsmade fromvaryingmaterials andwith differingmeans ofsecuring and advice regarding rotation, timing ofreplacement, and method of replacement. No studieshave been published on long-term outcomes of differentapproaches.

Fourth, why does GI tract involvement develop,sometimes after several years, in some children withsevere types of EB? Freeman et al (8) reported a co-hort who demonstrated inflammatory changes in thelower GI tract and postulated that this may arisesecondarily to antigenic exposure in the gut lumen as aresult of mucosal fragility. Modified feeding formulas(e.g., protein hydrolysates), in conjunction withexclusion diets and drug therapy, were used withmixed success to address severe abdominal distensionand explosive diarrhea. If there is no overt GI tractinvolvement before or at the time of placement, couldfeed modification preempt the development of GIcomplications and problems around the insertion site?Is the timing of feeding relevant? Most G-tube-fedchildren with EB prefer nocturnal feeding, but this isunphysiological and can exacerbate GER, resulting innausea or pain, and reduce feed volume tolerance.Might by-passing the stomach and feeding lower downin the gut using a gastrojejunal tubes reduce GER? Isthere any relationship between the co-administrationof food, supplements, and medications and thedevelopment of GI problems and leakage at the G-tube insertion site? Supplements of zinc and iron arerecognized to cause nausea and vomiting and gastricirritation and constipation, respectively (1). Paradoxi-cally, several of the medications necessary to addressthe multiple complications of severe EB are implicatedin GI disturbances (15). Parents frequently co-admin-ister these for reasons of practicality (e.g., to save time,avoid mealtimes, avoid school hours, minimize thenumber of bolus flushes of water) even when profes-sional advice has counselled against this because ofpotential drug–nutrient interactions and altered bio-availablity (16).

Fifth, how best can the advantages of ED andG-tubeplacement be jointly exploited? ED is increasinglyemployed to ease dysphagia and promote oral intake(14), but is not a substitute for G-tube placement. Ourpatients experienced mixed outcomes of serial EDs; theprocedure eased dysphagia and regurgitation of pooledsecretions in some, with parental reports of much hap-pier mealtimes. In others, GER increased, or aversion toswallowing was so entrenched that there was refusal to

widen the range of foods eaten. Any increase in oralintake did not translate into greater weight and heightgain, and in no cases did ED alleviate dysphagia suffi-ciently to warrant removal of the G-tube. Mortell andAzizkhan (14) reported that significant growth failurepersisted in a reported 25% to 35% of cases despitesatisfactory ED, but should ED be used proactively todelay or avoidG-tube surgery in selected patients (e.g., ifrobust contraindications for placement were estab-lished)?

Lastly, and perhaps most importantly, how doeschronic inflammation influence the course of severe EB?The role of proinflammatory cytokines in developmentofGI complications (11,12) andaltered growth andbodycomposition (13) has already been alluded to, andinflammation is very likely to play a critical role in thepubertal delay (13), abnormal bone mineralization (17),and refractory anemia (18) so frequently seen in thisgroup of patients.

CONCLUSIONS

Numerous variables potentially influence outcomes ofG-tube feeding inchildrenwith severeEB,and this reviewraises more questions than it can answer. Successfuloutcomes depend partly on whether the initial aims ofplacement have been realized (and these often varybetween patients and over time in the same patient) andpartly on comparison of factors such as growth with val-idated standards. No such standards currently exist, andfamilies’ (and professionals’) expectations of the out-comes often vary. Careful consideration of the potentialadvantagesanddisadvantages isneedednotonly toavoidthe possible replacement of one set of problems withanother, but also to ensure that professionals presentconsistent information andwork cohesively.

Our results show that not all children experiencedproblems and that the parents of those who did felt themto be acceptable if the alternative was removal of theG-tube with resultant chronic hunger and pain that theywere powerless to relieve. They unanimously acknowl-edged that the G-tube was a crucial means of nourishingandmedicating their children and were pleased that theyhad chosen this intervention. Parents whose childrendemonstrated little or no improvement in longitudinalgrowth (in some cases with associated central fat depo-sition and distorted body shape) reported that this,although disappointing, was acceptable. The childrenwere in general agreement, frequently expressing regretthat the G-tube was necessary but mainly accepting itsproblems, with most citing the major advantages to bethe removal of pressure to eat and swallow medications.The major disadvantage cited by children was pain


during G-tube changes, particularly when the insertionsite was eroded or infected.

The question remains ‘‘Could we do better?’’ Toattempt to answer this, further, and more detailed, datacollection needs to be undertaken onwhich professionalscan base best practice while appreciating each family’sneed for advice and support tailored to their specificsituation. A robustmeasure ofQOL is needed, but this isdifficult to design for a disease with complications thatincrease and intensify over time. There is currently noQOL measurement tool that is validated for use withchildren with EB, and we recognize the importance ofpursuing this. We also acknowledge that lack of datafrom a matched control group that did not undergoG-tube placement diminishes our ability to draw firmconclusions from our study. Establishment of collabo-rative studies between the major international pediatricEB centers would greatly facilitate future evidence-basedpractice, expedite the development of an appropriateQOLmeasurement tool, and aid decision-making in thisextremely challenging group of conditions.

ACKNOWLEDGMENTS

The authors would like to thank Derek Roebuck, Con-sultant Interventional Radiologist, for data on ED andMaya Abou-Khater, Dietitian, for preparing the initialinformation spreadsheet and entering data. J.E.M.acknowledges financial support from the Department ofHealth via the UK National Institute for Health Re-search Comprehensive Biomedical Research CentreAward toGuy’s andStThomas’NHSFoundationTrustin partnership with King’s College London and King’sCollege Hospital NHS Foundation Trust.

REFERENCES

1. Haynes L. Nutrition for children with epidermolysisbullosa. In: Murrell DF, Thiers BH, eds. Epidermolysisbullosa: Part II – diagnosis and management, DermatolClin 28, 2. Philadelphia: Saunders, 2010:289–301.

2. Allman S, Haynes L, MacKinnon P et al. Nutrition indystrophic epidermolysis bullosa. Pediatr Dermatol1992;9:231–238.

3. Ingen-Housz-Oro S, Blanchet-Bardon C, Vrillat M et al.Vitamin and trace metal levels in recessive dystrophicepidermolysis bullosa. J Eur Acad Dermatol Venerol2004;18:649–653.

4. Colomb V, Bodemer C. The benefits of gastrostomy andenteral feeding in children with recessive dystrophic

epidermolysis bullosa. Presentation at DebRA Interna-tional Conference, Avignon, 2007.

5. Haynes L, Atherton DJ, Ade-Ajayi N et al. Gastrostomyand growth in dystrophic epidermolysis bullosa. Br JDermatol 1996;134:872–879.

6. Kutiyanawala MA, Hussain A, Johnstone JMS et al.Gastrostomy complications in infants and children. AnnRColl Surg Engl 1998;80:240–243.

7. Jolley SG, TunellWP,HoelzerDJ et al. Lower esophagealpressure changes with tube gastrostomy: a causative factorof gastroesophageal reflux in children? J Ped Surg1986;7:624–627.

8. Freeman E, Koglmeier J, Martinez AE et al. Gastrointes-tinal complications of epidermolysis bullosa in children. BrJ Dermatol 2008;158:1308–1314.

9. Stehr W, Farrell MK, Lucky AW et al. Non-endoscopicpercutaneous gastrostomy in children with recessive dys-trophic epidermolysis bullosa. Pediatr Surg Int 2008;24:349–354.

10. Haynes L. Nutritional support for children with epiderm-olysis bullosa. J Hum Nutr Diet 1998;11:163–173.

11. Macrae VE, Wong SC, Farquharson C et al. Cytokineactions in growth disorders associated with pediatricchronic inflammatory diseases [review]. Int J Mol Med2006;18:1011–1018.

12. Wong SC,MacraeVE,McGroganP et al. The role of pro-inflammatorymatory cytokines in inflammatory boweldisease with growth retardation. J Pediatr GastroenterolNutr 2006;43:144–155.

13. Martinez AE, Allgrove J, Brain C. Growth and pubertaldelay in patients with epidermolysis bullosa. In: MurrellDF, Thiers BH, eds. Epidermolysis bullosa: Part II –diagnosis and management, Dermatol Clin 28, 2. Phila-delphia: Saunders, 2010:357–359.

14. Mortell AE, Azizkhan RG. Epidermolysis bullosa: man-agement of esophageal strictures and enteric access bygastrostomy. In: Murrell DF, Thiers BH, eds. Epidermol-ysis bullosa:Part II - diagnosis andmanagement,DermatolClin 28, 2. Philadelphia: Saunders, 2010:311–318.

15. PaediatricFormularyCommittee.BNFfor children [2010–2011]. London: BMJ Publishing Group, PharmaceuticalPress, and RCPCH Publications, 2010.

16. White R. Symposium 8: drugs and nutrition: how sideeffects can influence nutritional intake. Conference on‘Malnutrition matters’. Proc Nut Soc 2010;69:558–564.

17. Martinez AE,Mellerio JE. Osteopenia and osteoporosis inepidermolysis bullosa. In: Murrell DF, Thiers BH, eds.Epidermolysis bullosa: Part II - diagnosis and manage-ment, Dermatol Clin 28, 2. Philadelphia: Saunders,2010:353–355.

18. Hubbard L, Haynes L, Sklar M et al. The challenges ofmeeting nutritional requirements in children and adultswith epidermolysis bullosa: proceedings of a multi-disci-plinary team study day. Clin and Exp Derm 2011;36:579–584.


gastrostomy tube feeding in children with epidermolysis bullosa: consideration of key issues

Documents