TASRJN

ltmwo

hwbgs

gGvev

MhtamSmtcadOtiHbipA

A

PS

Atn

©P

he Retrograde Aortic Arch in the Hybridpproach to Hypoplastic Left Heart Syndrome

erban C. Stoica, MD, Alistair B. Philips, MD, Matthew Egan, MD,oberta Rodeman, RN, Joanne Chisolm, RN, Sharon Hill, ACNP,

ohn P. Cheatham, MD, and Mark E. Galantowicz, MD

ationwide Children’s Hospital, Columbus, Ohio

stistiAso

2tsimt

PED

IAT

RIC

CA

RD

IAC

Background. Before palliative stage 2 for hypoplasticeft heart syndrome, the coronary and cerebral circula-ions are often dependent on retrograde perfusion by

eans of the aortic arch. Results of hybrid palliationith a focus on patients exhibiting retrograde aortic arch

bstruction (RAAO) were analyzed.Methods. From July 2002 to March 2008 66 consecutive

ybrid procedures for hypoplastic left heart syndromeere performed. Patients requiring RAAO interventionased on cardiology–surgery consensus were defined asroup 1 (n � 16), whereas all other hypoplastic left heartyndrome patients formed group 2 (n � 50).

Results. At birth there were no differences betweenroups in terms of demographics or cardiac function.roup 1 had more patients with aortic atresia (94%

ersus 58%; p � 0.01), and 69% of patients had initialchocardiographic comments regarding incipient RAAO

ersus 26% in group 2 (p � 0.007). The type of ductal

aNturm

hmtpotgishecmmtFete

al, 700 Children’s Dr, Columbus, OH 43205; e-mail: mark.galantowicz@ationwidechildrens.org.

2009 by The Society of Thoracic Surgeonsublished by Elsevier Inc

tent, balloon versus self-expandable, did not influencehe subsequent development of RAAO. Before RAAOntervention (mean age, 74 days), group 1 patients hadignificantly more tricuspid regurgitation. The mainreatment for RAAO in group 1 was coronary stentnsertion, with 3 patients having a reverse central shunt.t a mean follow-up of 611 days, group 1 had reduced

urvival interstage (56.3% versus 88%; p � 0.005) andverall (43.7% versus 70%; p � 0.03).Conclusions. Clinically important RAAO occurred in

4% of the hypoplastic left heart syndrome patients inhis series. If RAAO is detected at birth or early inter-tage, a Norwood operation is now favored. Palliativenterventional catheterization remains very important

id and late interstage for continuing the hybrid strategyoward comprehensive stage 2.

(Ann Thorac Surg 2009;88:1939–47)

© 2009 by The Society of Thoracic Surgeons

ore than 25 years after the first successful stagedreconstruction, the management of patients with

ypoplastic left heart syndrome (HLHS) remains a mul-idisciplinary challenge. Several strategies in recent yearsre having a noticeable impact on the disease burden,ost notably the right ventricle to pulmonary artery

ano modification, a hybrid approach, and the closeonitoring of patients between stages 1 and 2 of pallia-

ion [1–4]. This diversification is exciting, but owing to aombination of approaches, heterogeneity of patients,nd relatively small numbers in each center, the merits ofifferent surgical strategies require further follow-up.ur center embraced the hybrid collaboration, and after

he adoption period we are now able to use a reproduc-ble technique with good results in the average-risk

LHS patient [5–7]. Hybrid stage 1 palliation consists ofilateral pulmonary artery banding, placement of a stent

n the persistent ductus arteriosus (PDA) without cardio-ulmonary bypass, and then balloon atrial septostomy.fter the child matures a comprehensive stage 2 is done

ccepted for publication June 25, 2009.

resented at the Forty-fifth Annual Meeting of The Society of Thoracicurgeons, San Francisco, CA, Jan 26–28, 2009.

ddress correspondence to Dr Galantowicz, Nationwide Children’s Hospi-

t approximately 6 months of age in the form of aorwood-type reconstruction combined with a bidirec-

ional cavopulmonary anastomosis. It is postulated butnproved that the whole sequence of management mayeduce the overall morbidity and mortality by avoidingajor open heart surgery in the neonatal period.A legitimate concern is that until stage 2 reconstruction

ybrid patients have coronary and cerebral perfusion byeans of retrograde flow through the aortic arch, with

he possibility of subsequent obstruction at this site. Inatients with aortic atresia this flow is entirely dependentn retrograde perfusion. Retrograde aortic arch obstruc-ion (RAAO) can occur in utero, postnatally before sur-ical treatment begins, or afterward in acute or more

nsidious fashion at the time of PDA stenting or duringubsequent follow-up [8, 9]. We report results with theybrid approach in a series of HLHS patients with anmphasis on the development of this particular compli-ation. In defining RAAO here we have taken a prag-atic approach. As discussed below, a comparison isade between patients who required a specific interven-

ion for RAAO and all other HLHS patients in the series.or the benefit of capturing the multiple facets of thisntity we have included here all HLHS patients treated onhe hybrid pathway at our center, including the initial

xperience previously described as “the learning curve” [5].

0003-4975/09/$36.00doi:10.1016/j.athoracsur.2009.06.115

Ib

P

PWcidmJse(aepprbhtdphdtepdc

itapsc

Fgpd3(8Gft

TH

C

MAW

A

B

A

AW

1940 STOICA ET AL Ann Thorac SurgRETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS 2009;88:1939–47

PEDIA

TR

ICC

AR

DIA

C

t was through this experience that important insights haveeen made influencing our current treatment algorithm.

atients and Methods

atient Populatione carried out a retrospective review of prospectively

ollected data. The study was approved by the localnstitutional review board. In view of the retrospectiveesign with absence of patient identifiers, the require-ent for individual patient consent was waived. Between

uly 2002 and April 2008, 76 consecutive cases of hybridtage 1 palliation were performed in our hospital. Afterxcluding 1 HLHS case with extremely high-risk featuressignificant comorbidities, including large congenital di-phragmatic hernia), another HLHS patient arriving inxtremis at 3.5 months of age from overseas, and 8atients with non-HLHS complex single ventricles, 66atients with HLHS were retained for analysis. Theeason for excluding non-HLHS patients was to avoidias from comparing them with HLHS patients withypoplastic arches and therefore, in theory, more prone

o RAAO. The comprehensive stage 2 operation wasone at an average age of 6 months. Interstage allatients had cardiology monitoring every 1 to 2 weeks. Aigh index of suspicion was maintained to detect evi-ence of cardiac dysfunction. Tricuspid regurgitation,

ransatrial septal flow, ventricular function, and gradi-nts in the PDA, retrograde aortic arch, and across theulmonary bands were evaluated regularly with echocar-iography. When abnormal echocardiographic variables

ig 1. (A) Angiographic appearance of retro-rade aortic arch obstruction in an interstageatient found to have tricuspid regurgitation,ecreased right ventricular function, and a5-mm Hg gradient across the aortic arch.B, C, and D) Same patient treated with an-mm-long coronary stent expanded to 5 mm.radient disappeared and right ventricular

unction and tricuspid regurgitation returnedo baseline.

orrelated with clinical findings of cardiac failure orAt

schemia or with ST changes on the electrocardiogram,he clinical diagnosis of RAAO was always entertained,nd patients proceeded to cardiac catheterization. Allatients were discussed in our weekly cardiology–urgery conference to obtain consensus regarding thelinical management strategy. For the purposes of this

able 1. Characteristics of Patients With Hypoplastic Lefteart Syndrome

haracteristicGroup 1(n � 16)

Group 2(n � 50)

pValue

ale sex (%) 9 (56) 32 (64) 0.76ge at H1 (days) 8.1 � 6.2 10.3 � 9.1 0.28eight at H1 (kg) 3.1 � 0.6 2.9 � 0.6 0.53Weight �2.5 kg at

H1 (%)4 (16) 15 (30) 1.00

Weight �2.0 kg atH1 (%)

0 (0) 3 (6) 1.00

natomyAA (%) 15 (94) 29 (58) 0.01AA/MS (%) 6 (38) 12 (24) 0.34Intact atrial septum 2 (13) 2 (4) 0.24

alloon/self-expandablePDA stents

3/13 16/34 0.36

ge at retrogradeintervention (days)

73.9 � 52.2 Not applicable

ge at C2 (days) 165.6 � 60.1 171.7 � 39.7 0.78eight at C2 (kg) 5.4 � 1.3 5.7 � 0.8 0.32

A � aortic atresia; C2 � comprehensive stage 2; PDA � persis-ent ductus arteriosus; H1 � hybrid stage 1; MS � mitral stenosis.

aiigcoc[

tsbg

TOihntuFotisoesPGJISmrthfrtap

F

TF

V

I

P

a

ms4iew

LR

1941Ann Thorac Surg STOICA ET AL2009;88:1939–47 RETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS

PED

IAT

RIC

CA

RD

IAC

nalysis we defined two groups on the basis of a targetedntervention to treat RAAO (group 1) or no interventionn the remainder of the hybrid population (group 2). Inroup 1 the treatment for RAAO was individualized andould consist of interventional catheterization, surgery,r both. It therefore included patients who had a reverseentral shunt, as described by Caldarone and associates8], and those who had an early comprehensive stage 2 to

able 2. Echocardiographic Data During Steady-Stateollow-up of Groups 1 and 2

ariableGroup 1(n � 16)

Group 2(n � 50) p

nitial assessmentRV functiona 1.2 � 0.6 1.3 � 0.7 0.49Patients with comment regarding

retrograde arch (%)11 (69) 13 (26) 0.007

reintervention assessmentRV functiona 1.5 � 0.7 1.3 � 0.7 0.15Tricuspid regurgitationb 2.2 � 0.8 1.6 � 0.9 0.03Retrograde velocity (m/s) 2.9 � 0.6 2.3 � 0.6 0.003PDA stent gradient (mm Hg) 2.5 � 0.7 2.3 � 0.6 0.21LPA band gradient (mm Hg) 4.0 � 0.7 4.3 � 0.5 0.17RPA band gradient (mm Hg) 3.6 � 0.4 4.0 � 0.4 0.07Atrial septal gradient 2.9 � 2.8 3.7 � 3.5 0.47

RV function measured on a scale of 1 to 4: 1 � normal, 2 � mild, 3 �oderate, and 4 � severe dysfunction. b Tricuspid regurgitation mea-

ured on a scale of 0 to 4: 0 � none, 1 � trivial, 2 � mild, 3 � moderate,� severe. In group 1 the last echocardiogram before retrograde arch

ntervention was reported at a mean age of 68 days. For group 2 anchocardiogram done at the same age was chosen and cumulatively thisas reported at an average of 72 days of life.

PA � left pulmonary artery; PDA � persistent ductus arteriosus;PA � right pulmonary artery; RV � right ventricular.

ig 2. Patient flow chart.

reat RAAO. For a functional comparison during inter-tage follow-up we chose the last echocardiogram doneefore RAAO treatment in group 1 and an echocardio-ram done at a similar mean age in group 2.

echniqueur hybrid strategy and results were recently described

n detail [6, 7]. Briefly, the neonate undergoes an initialybrid stage 1 palliation consisting of bilateral pulmo-ary artery bands and a PDA stent. The operation is done

hrough median sternotomy off cardiopulmonary bypasssing a 3.0- or a 3.5-mm Gore-Tex (W.L. Gore & Assoc,lagstaff, AZ) tube graft for creating the bands dependingn whether the patient’s weight was less than or greaterhan 2.5 kg, respectively. After bilateral pulmonary band-ng a sheath is inserted in the pulmonary artery and atent is deployed aiming to cover the PDA completely. Inur initial experience we used predominantly balloon-xpandable stents, but in more recent years newly de-igned self-expandable stents were favored. Types ofDA stent used are as follows: balloon expandable,enesis Blue, Genesis XD, premounted Genesis (Cordis/

ohnson & Johnson, Miami, FL), and Formula 418 (Cooknc, Bloomington, IN); and self-expandable, Precise,MART (Cordis/Johnson & Johnson), Protégé (ev3, Ply-outh, MN), and Zilver (Cook, Inc). In our initial expe-

ience we found that a balloon atrial septostomy at theime of the hybrid stage 1 is fraught with technical andemodynamic difficulties [5, 7]. Currently this is per-

ormed in more stable conditions when the neonate iseady for discharge, at about 2 weeks after undergoinghe hybrid procedure. The comprehensive stage 2 oper-tion consisted of: removal of the PDA stent and theulmonary bands, reconstruction of the aortic arch and

pdaarsp

psapRdmPdddvm

DF2d

SnFuct

R

PSadiohseb0

CTjd

T

PN

111

1

1

11

T

Bi

1942 STOICA ET AL Ann Thorac SurgRETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS 2009;88:1939–47

PEDIA

TR

ICC

AR

DIA

C

ulmonary arteries (if needed), reimplantation of theiminutive ascending aorta into the pulmonary artery,trial septectomy, and a bidirectional cavopulmonarynastomosis. The PDA stent was removed in all cases, theight pulmonary band site was incorporated in the Glennhunt, and a left pulmonary patch augmentation waserformed in approximately half of the cases.Interstage cardiac catheterization in this series of 66

atients was performed under general anesthesia inelected cases. The indication was diagnostic (eg, evalu-tion of physiology or associated anomalies such asulmonary vein stenosis) or therapeutic (eg, treatment ofAAO, PDA restenosis, or atrial septal restenosis). Theiagnosis of RAAO was confirmed by pressure gradienteasurement in the retrograde aortic arch using the Radi

ressureWire (Radi Medical Systems, Inc, Uppsala, Swe-en), and when appropriate, RAAO was treated byeploying a balloon-expandable coronary stent from theescending aorta (Fig 1). Angiographic improvement waserified after stenting, as well as by a drop in the directlyeasured RAAO gradient.

ata Analysisollow-up was complete at the time of censoring (June008). Continuous variables are reported as mean � stan-ard deviation and were compared using the unpaired

able 3. Individual Procedures and Outcomes in Patients Wh

atiento.

Age atIntervention (days)

Type of RAA Stent (dilated tomm)/Other Intervention(s)

1 122 Severe PDA in-stent and retrograarch stenosis, poor function aftstent dilatation and C2 broughforward

2 62 Retrograde stenosis shown oncatheterization day 51, C2performed early as a result

3 27 Cardiac arrest during H1, ECMO

4 168 Cordis Velocity (4.9)

5 63 Velocity RX heparin (5)6 163 Guidant Multi-Link Ultra (5.3)7 79 Guidant Multi-Link Vision (4.5)8 23 2 Guidant Multi-Link Vision (5)9 3 H1 and reverse BT shunt0 93 Guidant Multi-Link Ultra (4.8)1 22 Guidant Multi-Link Ultra (4.9)2 82 Guidant Multi-Link Ultra (4.7)

3 50 Guidant Multi-Link Ultra (4.8)

4 143 2 Guidant Multi-Link Ultra (5.2)

5 73 Cook Formula 418 (5.2)6 14 Guidant Multi-Link Vision (4.1)

he 4th column (survival) indicates where the patient is still alive to date

T � Blalock-Taussig; C2 � comprehensive stage 2; ECMO � extranterstage; PDA � persistent ductus arteriosus; RAA � retrograde aort

tudent’s t test. Categorical variables, shown as absoluteumber and percentage, were compared with a two-tailedisher’s exact test. The Kaplan-Meier actuarial method wassed for follow up analysis, and the survival curves wereompared with the log rank test. Probability values smallerhan 0.05 were considered statistically significant.

esults

atient Characteristicsixteen patients (24%) exhibited RAAO as defined abovend received a targeted intervention at a mean age of 74ays. The characteristics of the two groups are presented

n Table 1. There were no differences in terms of sex, age,r weight at either stage 1 or stage 2 operations. As forigh-risk anatomic features [10, 11], the intact atrialeptum and aortic atresia–mitral stenosis variants werequally distributed between the groups, but in group 1 allut 1 patient had aortic atresia (94% versus 58%; p �.01).

ardiac Functionable 2 shows echocardiographic data at baseline and

ust before RAAO intervention for group 1. There was noifference in right ventricular function initially, but a

veloped Significant Retrograde Arch Obstruction (Group 1)

Survival (days)/Stage of Death

Cause of Death orSubsequent Clinical Outcome

128/C2 Early C2, off bypass on ECMO, no recovery

78/C2 Early C2, off bypass on ECMO, no recovery

27/H1 Unknown congenital, severe retrogradearch stenosis detected at autopsy

Alive C2 at 237 days, ultimately transplanted forfailing Fontan

Alive C2 at 221 daysAlive C2 at 217 daysAlive C2 at 139 daysAlive Reverse BT shunt day 102, C2 at 205 days70/IS Sudden death at home, unmonitoredAlive C2 at 177 days

141/IS Sudden death at home, unmonitored146/IS Reverse BT shunt day 137, cardiac arrest 9

days later134/IS Stent angioplasty day 95, ECMO day 109,

cerebral hemorrhage159/IS Listed for Tx day 157, sudden cardiac arrest

2 days laterAlive C2 at 114 days28/IS Cardiac arrest

for those who died it shows the age and stage of death.

o De

,

deert

and

corporeal membrane oxygenation; H1 � hybrid stage 1; IS �ic arch; Tx � transplantation.

hqpRawfpbca

CgFRgaptsbtmoeoitpapAps1pafathocudhfwrtcmtiis

st

sfealdopn

2(gim

C

Itpochicor

Fv

1943Ann Thorac Surg STOICA ET AL2009;88:1939–47 RETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS

PED

IAT

RIC

CA

RD

IAC

igher proportion of patients in group 1 had specificualitative comments about the retrograde arch com-ared with group 2 (69% versus 26%; p � 0.01). BeforeAAO treatment, group 1 patients had a higher gradientt this level along with more tricuspid regurgitation butithout statistically significant differences in ventricular

unction, atrial septal gradient, and PDA gradient. Twoatients in group 2 were catheterized interstage on theasis of suspected RAAO, but the diagnosis was notonfirmed and no specific treatment for the retrograderch was required.

linical Outcomes and Survivalroup 1. The clinical course of the patients is described inigure 2. Table 3 has further details about the type ofAAO intervention and the clinical course of patients inroup 1. Patient numbers in this section refer to group 1nd Table 3. Patient 3 was unstable during the hybridrocedure; despite resuscitative measures including ex-

racorporeal membrane oxygenation, the patient did noturvive. At autopsy severe congenital stenosis was notedetween the transverse and the descending aorta. Al-

hough the diagnosis of RAAO was established postortem, this patient belongs in group 1 for the purposes

f this study. Patients 1 and 2 are representative of ourarly experience and examples of unsuccessful attemptsf an early comprehensive stage 2. Both patients exhib-

ted RAAO physiology with cardiac dysfunction, leadingo stage 2 palliation being brought forward. Neitheratient could be weaned from cardiopulmonary bypass,nd despite extracorporeal membrane oxygenation sup-ort both patients died of multisystem organ failure.fter this initial unfavorable experience our strategy wasalliative stenting for RAAO as a bridge to correctiveurgery beyond 4 months of age. Seven patients in group

went on to comprehensive stage 2 by means of thisathway. Patient 9 was diagnosed with RAAO prenatallynd after discussing all possibilities with the patient’samily they opted for a hybrid procedure, which included

reverse central shunt. After an uneventful postopera-ive course the patient experienced sudden death atome interstage. Patient 11 was another unmonitoredut-of-hospital sudden death. Patient 12 had a reverseentral shunt 2 months after RAAO stenting and had annheralded cardiac arrest on the floor on postoperativeay 9. Patient 8 was successfully bridged to a compre-ensive stage 2 after a reverse central shunt was placed

or recurrent RAAO after developing severe stenosisithin the retrograde stent. Finally, of the 12 patients

eceiving stents for RAAO, 1 was listed for transplanta-ion and died on the waiting list, and 3 had suddenardiorespiratory arrests, with temporary extracorporealembrane oxygenation support for patient 13. Among

hese 12 RAAO stented patients, survivors had stentmplantation at an average age of 95 days versus 62 daysn nonsurvivors, but the difference was not statisticallyignificant.

group 2. Seven deaths occurred in this group betweentages 1 and 2. There were 3 postoperative deaths after

he hybrid procedure. One patient with intact atrial l

eptum had balloon atrial septostomy on day 7 of lifeollowed by the hybrid on day 9 and arrested unexpect-dly on the night of the operation. Two patients died ofspiration before discharge and of necrotizing enteroco-itis, respectively. Between stages 1 and 2 one patientied of infection at another hospital, 2 patients hadut-of-hospital sudden death, and 1 other patient diedostoperatively after surgical repair of recurrent pulmo-ary vein stenosis performed at 6 months.As detailed above a total of 4 patients in groups 1 andhad postoperative deaths after the hybrid procedure

6%). Figure 3 shows actuarial survival for the tworoups, the figures to date being 56.3% versus 88%

nterstage (p � 0.005) and 43.7% versus 70% overall at aean follow-up of 611 days (p � 0.03).

omment

n this report we highlight a number of points related tohe development of RAAO in a series of consecutiveatients treated with a hybrid approach for HLHS. Firstf all, the majority of patients (76%) do not exhibit thisomplication. However, in patients with RAAO there isigher morbidity and mortality interstage and probably

nferior results overall. We should emphasize that in-luding our early experience there was a negative effectn overall survival compared with results reported moreecently [6]. As previously, it is a reflection of the lessons

ig 3. Kaplan-Meier survival curves for groups 1 and 2. (A) Sur-ival to stage 2; p � 0.005. (B) Survival overall; p � 0.03.

earned [5].

sdctceabnrs

cpfea

pdPehASaostapesbtmai

chRTatpt

ssemoept

Fo

FDaa

1944 STOICA ET AL Ann Thorac SurgRETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS 2009;88:1939–47

PEDIA

TR

ICC

AR

DIA

C

In the absence of a clear definition for what RAAOhould represent, we adopted a practical approach inefining it as a combination of echocardiographic orlinical indicators suggestive enough to warrant furtherargeted intervention. The treatment allocation reflectslinical consensus but is not immune from bias. How-ver, this classification allowed some clinical insight inton elusive and probably progressive entity, which cannote fully described by a snapshot characteristic. Group 1 isot homogeneous in terms of treatments applied andeflects our changes in management strategy based onerial accrual of patients.

An important lesson is that RAAO can exist as aongenital anatomic variant in a subset of HLHS patientsotentially manifesting itself in utero or postnatally be-

ore any surgical intervention. Patient 3 in group 1xperienced intermittent bradycardia in utero leading ton early delivery. After birth there were intermittent

ig 4. Autopsy specimen showing congenital retrograde aortic archbstruction (RAAO). (PDA � persistent ductus arteriosus.)

ig 5. Two-dimensional (left) and coloroppler echocardiograms (right) of retrograde

ortic arch obstruction (A) and normal archnatomy (B).

eriods of instability thought to be secondary to infectionelaying a hybrid stage 1. Immediately on deploying theDA stent, cardiac ischemia and collapse necessitatedxtracorporeal membrane oxygenation support. Withinours this support was stopped because of brain death.utopsy clearly demonstrated congenital RAAO (Fig 4).ubsequently our group learned how to evaluate thisrea by echocardiography, thereby ruling RAAO in orut before any surgical intervention (Fig 5). Using thiscreening we have seen that approximately 10% of pa-ients with HLHS have RAAO at birth. We believe this is

contraindication to a hybrid stage 1 and refer thoseatients for a traditional Norwood procedure. An inter-sting aside may be that this subgroup of patients reflectsome of the pretreatment variability reported in HLHSy neurologic imaging and neurocognitive studies. In

erms of patient size, it may be intuitive that RAAO isore prevalent in small patients, but we have not seen

ny relationship, which encourages us to continue treat-ng low-weight patients on the hybrid pathway.

Group 1 also includes 3 patients who had a reverseentral shunt [8]. The procedure was done at the time ofybrid in 1 patient and for the development of recurrentAAO after retrograde stenting for the other 2 patients.wo patients subsequently experienced sudden death,nd based on this limited experience, we have nowemporarily abandoned this solution. Perhaps the play ofressures in this unusual configuration is such that at

imes there is important and even fatal coronary steal.Another lesson learned was that an early comprehen-

ive stage 2, such as in patients 1 and 2, is not a goodolution for RAAO. Although Jaquiss and colleagues [12]legantly demonstrated how a cavopulmonary anasto-osis can work before 4 months of age after a Norwood

peration, we do not think that there is sufficient experi-nce to recommend it in the current context. The com-rehensive stage 2 is a bigger procedure than a stage 2 in

he traditional pathway with success dependent on opti-

mc

pRamsvaNbsilstoac

pcnagaaigfcaswt

gdbrwosojtfldaFtCimAab

w

atbsrdtditiimohwibwdsiofdrtpst

ahfadwrrcpripah

R

1945Ann Thorac Surg STOICA ET AL2009;88:1939–47 RETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS

PED

IAT

RIC

CA

RD

IAC

al cardiac and pulmonary function, which is not thease in a patient with RAAO.

Coronary stent insertion, on the other hand, hasroved to be a more durable and predictable solution toAAO. A few words of caution about this modality arelso warranted. If resorted to in early interstage, stentingay trigger an intimal hyperplasia response that is

elf-perpetuating. In patients 8, 12, and 13 further inter-entions for RAAO were indeed required in the form ofcentral shunt or stent angioplasty with only 1 survivor.onsurvivors had stent implantation at a younger ageut statistical significance and a cause and effect relation-hip are absent for now. In this experience, however,nterventional catheterization was invaluable in mid andate interstage to transition patients toward comprehen-ive stage 2. Regardless of the timing of RAAO interven-ion, the importance of interstage surveillance cannot beveremphasized [1, 5]. Unmonitored sudden death waslso present in group 2 and remains an occurrence in allontemporary series.

What causes clinically important RAAO in the firstlace and how is it best diagnosed? Our current seriesan offer some hints, but more targeted research iseeded to answer these questions. First of all groups 1nd 2 had few baseline differences in terms of demo-raphics, anatomy, and ventricular function. Group 1 hadsignificant clustering of cases with aortic atresia (94%),situation entirely reliant on retrograde arch flow. It is

nteresting to note that significantly more patients inroup 1 had some echocardiographic suggestions of riskor subsequent RAAO (Table 2). The hint was an echo-ardiographic gradient or a subjective comment suchs “narrowing of isthmus” or “juxtaductal shelf.” Thishows that early indicators of RAAO may be present, ande have embarked on studying these further so as to be able

o have a better triage of patients on the surgical pathways.In terms of PDA stents our choices parallel those of the

roup in Giessen, Germany [3]. We prefer the newesign of the self-expandable stent and will only use aalloon-expandable device if the PDA is stenosed andequires dilatation. The type of stent was not associatedith RAAO in this analysis, but to what extent the originf the retrograde aortic arch was covered initially by thetent is being examined in a separate study. In ourpinion it is not advisable to have too short a stent at the

unction of the PDA with the aorta because this can leado recoarctation and possible problems with retrogradeow. Crossing the retrograde orifice with an open-cell–esign stent has not been problematic in our experiencend allows further catheterization targeted at RAAO.rom a surgical point of view it was possible to removehe stent entirely in every case at comprehensive stage 2.ompared with another group with extensive experience

n the hybrid approach it is not clear why we are seeingore RAAO. It may be partly related to patient profile;kinturk and coworkers [3] reported fewer cases of aortictresia and more patients eventually amenable to aiventricular repair.This study has a number of limitations. First, RAAO

as not defined at the outset, and group 1 does not have

“before and after” comparison of treatment effects inhis retrospective design. Defining the groups on theasis of nonalgorithmic treatment allocation is anotherource of bias, but on the other hand it reflects the clinicaleality of treating these patients. The report is mainlyescriptive in presenting this problem and the steps we

ook to deal with it. There can be only speculativeiscussions at this stage about the anatomy of clinically

mportant RAAO and its true physiologic impact. Au-opsy data were very limited because of a lack of consentn all patients. With only 16 RAAO index cases, it is alsompossible to have more cause and effect insight because

ultivariate analysis remains out of bounds. Finally, it isf course too early to evaluate neurologic outcomes inybrid patients with and without RAAO. In the interime have taken the view that satisfactory cardiac function

s a surrogate of equally satisfactory cerebral functionetween stages 1 and 2. This assumption is in keepingith a detailed autopsy study showing that the aorticiameter distal to the origin of the innominate artery isignificantly bigger than the aortic dimension just prox-mal to this landmark [9]. Nonetheless, this study dem-nstrates satisfactory results with the hybrid approachor HLHS in a series of consecutive patients and is aetailed clinical report of RAAO. Our group is currentlyesearching early imaging indicators of RAAO as well ashe influence that a hybrid pathway has on cerebralerfusion and function. In a separate analysis we havehown that echocardiographic variables are predictive ofhe clinical course interstage [13].

In conclusion, a high index of suspicion is required atll stages for detecting and treating this complication. Itas become useful for us to think of three separate time

rames for the development of RAAO that may be man-ged differently. First, RAAO may be diagnosed echocar-iographically at birth or preoperatively. In this case weould view it as a contraindication to a hybrid approach,

ecommending instead a Norwood operation. Second, forapidly progressive early interstage RAAO one shouldonsider either a conversion to a Norwood procedure orlan the retrograde stenting such that the area can beestented if in-stent restenosis develops. Finally, if RAAOs detected in mid or late interstage, the transcatheterlacement of a retrograde stent can alleviate the stenosis,llowing the patient to transition to a successful compre-ensive stage 2 on the hybrid pathway.

eferences

1. Ghanayem NS, Cava JR, Jaquiss RD, Tweddell JS. Homemonitoring of infants after stage one palliation for hypoplas-tic left heart syndrome. Semin Thorac Cardiovasc SurgPediatr Card Surg Annu 2004;7:32–8.

2. Wernovsky G, Ghanayem N, Ohye RG, et al. Hypoplasticleft heart syndrome: consensus and controversies in 2007.Cardiol Young 2007;17(Suppl 2):75–86.

3. Akinturk H, Michel-Behnke I, Valeske K, et al. Hybridtranscatheter-surgical palliation: basis for univentricular orbiventricular repair: the Giessen experience. Pediatr Cardiol2007;28:79–87.

4. Caldarone CA, Benson L, Holtby H, Li J, Redington AN, VanArsdell GS. Initial experience with hybrid palliation for

1

1

1

1

D

DTa

rhgiwSwpr6as

taitlflpts

tttdt

icioTaihbo

1946 STOICA ET AL Ann Thorac SurgRETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS 2009;88:1939–47

PEDIA

TR

ICC

AR

DIA

C

neonates with single-ventricle physiology. Ann Thorac Surg2007;84:1294–300.

5. Galantowicz M, Cheatham JP. Lessons learned from thedevelopment of a new hybrid strategy for the managementof hypoplastic left heart syndrome. Pediatr Cardiol 2005;26:190–9.

6. Galantowicz M, Cheatham JP, Phillips A, et al. Hybridapproach for hypoplastic left heart syndrome: intermediateresults after the learning curve. Ann Thorac Surg 2008;85:2063–70.

7. Galantowicz M, Cheatham JP. A hybrid strategy for theinitial management of hypoplastic left heart syndrome: tech-nical considerations. In: Sievert H, Qureshi SA, Wilson N,Hijazi ZM, eds. Percutaneous interventions for congenitalheart disease. Abingdon, UK: Informa Healthcare, 2007:531–8.

8. Caldarone CA, Benson LN, Holtby H, Van Arsdell GS. Mainpulmonary artery to innominate artery shunt during hybridpalliation of hypoplastic left heart syndrome. J Thorac Car-

diovasc Surg 2005;130:e1–2.

ccurred, catheter-based interventions do not effectively prevent

sdts

iei

a

DDmy

niwwtto

lseslc

vdoeaco

Bas

9. Ilbawi AM, Spicer DE, Bharati S, Cook A, Anderson RH.Morphologic study of the ascending aorta and aortic arch inhypoplastic left hearts: surgical implications. J Thorac Car-diovasc Surg 2007;134:99–105.

0. Lim DS, Peeler BB, Matherne GP, Kron IL, Gutgesell HP.Risk-stratified approach to hybrid transcatheter-surgicalpalliation of hypoplastic left heart syndrome. Pediatr Cardiol2006;27:91–5.

1. Glatz JA, Fedderly RT, Ghanayem NS, Tweddell JS. Impactof mitral stenosis and aortic atresia on survival in hypoplas-tic left heart syndrome. Ann Thorac Surg 2008;85:2057–62.

2. Jaquiss RD, Ghanayem NS, Hoffman GM, et al. Early cavo-pulmonary anastomosis in very young infants after theNorwood procedure: impact on oxygenation, resource utili-zation, and mortality. J Thorac Cardiovasc Surg 2004;127:982–9.

3. Fenstermaker B, Berger GE, Rowland DG, et al. Interstageechocardiographic changes in patients undergoing hybridstage I palliation for hypoplastic left heart syndrome. J Am

Soc Echocardiogr 2008;21:1222–8.

ISCUSSION

R CHRISTOPHER A. CALDARONE (Toronto, ON, Canada):hank you, President Chitwood and Secretary Wood, members,nd guests. I have no disclosures to make.Doctor Stoica and colleagues have made a comprehensive

eport of outcomes using a hybrid strategy for palliation ofypoplastic left heart syndrome, and, along with the Geissenroup, these authors are to be congratulated for pioneering an

nnovative strategy which challenges the more traditional Nor-ood strategy in this difficult group of patients. At the Hospital forick Children in Toronto, we have developed a hybrid program asell based on the experience of these groups and crafted ourrogram based on their lead. This is a tough group of patients, andegistry data typically is consistent with 2- to 3-year survival in the0 to 80% range. It is against this backdrop that the hybridpproach has been advocated as an alternative to the Norwoodtrategy and represents an important innovation in our field.

It is said that a great deal of medical innovation undergoeshree phases. First is a “boom” phase in which the innovationppears as a panacea which neatly solves a clinical problem ands often associated with rapid adoption by many centers. Afterhe boom phase comes the disillusionment phase when prob-ems surface, which suddenly render the innovation somewhatawed and perhaps not ideal for all patients, and, finally, a nichehase is established. This transition to a niche phase is aided by,

ypically, refinements in technique and improvements in patientelection.

I would submit that the Columbus group is striving to enterhe niche phase where the patient selection and refinements inechnique will establish the role of the hybrid as an alternativeo the Norwood strategy. Now is the time we must entertainebate and wrangle over selection criteria and refinements in

echnique. Within this context I have the following questions.The authors have identified retrograde flow through the

sthmus as the Achilles’ heel of the hybrid strategy. Theirurrent approach relies on rescue therapy using catheter-basednterventions after the problem is diagnosed. This commonlyccurs after detectible decreases in myocardial performance.welve patients were treated with this catheter-based strategynd 6 either died or had diminished myocardial function requir-ng transplantation. My basic question is whether the authorsave considered that their treatment of this problem may simplye ineffective. The data suggests that once the problem has

udden death or myocardial failure. Could the authors provideata to demonstrate the efficacy of these interventions at the

ime of intervention in terms of augmentation of isthmus dimen-ions and gradient reduction?

Second, what are the characteristics of patients in whomn-stent restenosis has developed? Could failure to achieveffective restoration of isthmus patency have contributed to thencreased mortality in this group?

Again, I compliment the authors on their pioneering effortsnd look forward to their responses. Thank you.

R STOICA: I wonder if I could please have my last slide back.r Caldarone, first of all, thank you very much for your com-ent and questions. We do value very much the contributions

ou are making in this new field.Before answering your first question, I would like to put the

umbers in perspective, if I may. This complication is only seenn a quarter of the patients, and of those at risk, for example,ith aortic atresia, only half of the patients with aortic atresiaere in group 1 and twice as many were in group 2. We expect

hat these numbers at risk and the prevalence of this complica-ion will go down after the lessons we learned here and whenur diagnosis methods are continuously refined.I would argue that our treatment is also proactive and, to a

arge extent, effective. If we look at our current philosophy inubset 1 (early, preintervention diagnosis) before treatment ismbarked on, the Norwood strategy will be effective and willolve the problem of unnecessary attrition. Subset 3 (diagnosisate interstage) can also be treated effectively we think byatheter-based intervention.

I wouldn’t say that this is done late, because RV (rightentricular) function is still preserved. I will come back to theata. There is the group in the middle (subset 2, early interstager recurrent RAAO [retrograde aortic arch obstruction]), how-ver, for which we don’t have a good solution as yet. It is a grayrea, it is an evolving area, and we think that a Norwoodonversion is the best way forward until this problem is sortedut.In-stent stenosis does occur. We see it at about 6 to 8 weeks.

ut if the patient is old enough, for example 3 months onverage, 2 months later they can have a safe comprehensive

tage 2 with a good result. We rely heavily on interstage

mp

cea

wd

ep

TT

OTcitsta

tFs

oSttwr

ptydi

1947Ann Thorac Surg STOICA ET AL2009;88:1939–47 RETROGRADE ARCH IN HYBRID PROCEDURE FOR HLHS

©P

onitoring, and the diagnosis is made while RV function isreserved.In terms of using a reverse BT (Blalock-Taussig) shunt, this

omplication does not occur in everybody, so we don’t think thatverybody needs to be prophylactically treated for that. There is

urning Today’s Research Into Tomorr

ng opportunities for surgeons and partnered with other

fp

pdn(rrtn

soamihFit

ita

2009 by The Society of Thoracic Surgeonsublished by Elsevier Inc

e don’t understand very well, and in fact, 2 of our 3 patientseveloped this and died, but all of them had coronary steal.To conclude, I will say that this cohort reflects our early

xperience, the lessons learned from adopting a hybrid ap-roach and, more recently, for retrograde arch obstruction.

C

strong potential for coronary steal in this new physiology that Thank you very much.

horacic Surgery Foundation for Research and Education

ow’s Patient Care

PED

IAT

RIC

CA

RD

IA

ur patients don’t follow the details of our research.hey don’t discuss unexpected breakthroughs or techni-al setbacks. They are not always aware of how changesn health care policies impact research funding and labime. Nonetheless, the advances we make in thoracicurgery touch each and every one of them. New surgicalechniques and potent new drugs improve patient healthnd extend patient lives.That is an outcome everyone can understand, and it’s

he one that has continued to push the Thoracic Surgeryoundation for Research and Education (TSFRE) forwardince its inception in 1992.

TSFRE was founded by the four major thoracic surgeryrganizations: the American Association for Thoracicurgery (AATS), The Society of Thoracic Surgeons (STS),

he Southern Thoracic Surgical Association (STSA), andhe Western Thoracic Surgical Association (WTSA). As itas 17 years ago, the Foundation’s mission is to support

esearch and education in thoracic surgery.The Foundation, however, has not only maintained its

osition as a leading supporter of research and educa-ion, it has also expanded its reach. Over the past fewears, the Foundation has established a comprehensiveevelopment program, improved its public policy train-

oundations such as the LUNGevity Foundation to im-rove support for research training.Perhaps most importantly, the Foundation has chosen to

lay a leading role in changing the current training para-igm for thoracic surgeons by becoming a founding orga-ization of the Joint Council on Thoracic Surgery Education

JCTSE). Along with the AATS, the American Board of Tho-acic Surgery (ABTS), and the STS, TSFRE has committed itsesources to support and empower the JCTSE to overhaulhe current thoracic surgery training program and coordi-ate all thoracic surgery education in the United States.TSFRE is a pivotal force for the growth and vitality of our

pecialty and its role is increasing, particularly in the areasf research, academic career development, and postgradu-te education. The philanthropic participatory index forembers of the Foundation’s founding organizations is

mportant as these surgeons know that giving begins atome and TSFRE is their home for research and education.oundation supporters—through donations or network-

ng—can have a significant impact on the future of cardio-horacic surgery and the welfare of our patients.

If you would like to make a pledge or receive morenformation about giving to TSFRE, please visit www.sfre.org or call Donna Kohli, TSFRE Executive Director,

t 978-927-8330.

Ann Thorac Surg 2009;88:1947 • 0003-4975/09/$36.00