Bidirectional With Multiple Cavopulmonary Risk Factors Shunt in Patients Florentino J. Vargas, MD, Gustav0 Mengo, MD, Juan P. Gallo, MD, Carlos B. Skerl, MD, Maria E. Ranzzini, MD, and Juan C. Vazquez, MD Unit r)t Pediatric CarLiio\,Isculal. ~~~rg~~r\. tlclspital Italianlr, Ru~,nil\ ,\IRY, ;\I-g:r,ntin,l Rnck~~ro~r~rl. Twenty-five patients in whom a modified Fontan operation was deferred because of multiple risk factors underwent a bidirectional cavopulmonary shunt. Two or more of the following risk factors were present in all: age less than 1 year, severe pulmonary artery distortion, impaired left ventricular function, subaortic obstruction, anomalous systemic-pulmonary venous connection, atrioventricular valve incompetence, and increased mean pulmonary artery pressure. Met/rods. Adhitional procedures included take-down of systemic-pulmonary artery shunt, atria1 septectomy, pulmonary artery reconstruct’ion, bulboventricular foramen enlargement, and atrioventricular valve repair. lic~lts. There were three hospital deaths (12’4 ). Mean follow-up is 21 months. There was no late mortality. Mean oxygen saturation increased from 71% to 83’: Results obtained from pulmonary artery reconstruction, enlargement of bulboventricular foramen, and atrioventricular valve repair were satisfactory for the patients in whom these risk factors were present preoperatively. Ventricular function also improved in the survivors in whom it was previously deteriorated, this being related to the suppression of the sources of ventricular volume overload. (‘or~lrrsior~s. In this risk group of patients for a modified Fontan operation, a bidirectional cavopulmonary shunt provided adequate palliation at reasonable low risk. Early bidirectional cavopulmonary shunt would minimize complications originating from systemicpulmonary shunts such as pulmonary artery distortion and the potential harm of chronic ventricular volume overload. S Material ince the original report ot the USC’ of the Fontan principle in tricuspid atrc\ia, it ha5 been w,idel\, applied to a variety of complex form5 of congenital heart disease\. ‘The original crittsria tor st,lt>ction of patient5 a\ proposed bv Choussat and .l\sociates hn\,e been reviewed and the indication5 t~*panded [I I. I low;ever, ct’rtain risk factors prevail II\ a cao~ ot morbidit\ and mortalitv. These include 5~1 tnrc di\tortion of thtx puImonary arteries [2, .3], impairc,d v<lntricular function 141, subnortic obstruction 151, anomalous s\steniic-palmooat-\’ vc‘nous connection lh], atrio\,cntricul‘~r (,AL ) valve incompetence 171, increasc,d mean pulmonary, artet-v pressure (PAP) 171, and I opt ag:c 1I, 71. A poor \ut-glc‘? result after a modified tontan ijperation ~wuld hc c+pectcd if some of thcs\c ri\k tactor+ arc prt>sent simuIt;~neouhlv in the preoprrativ~~ ;Is~c~ssnient. For tht+c> patients, creation of a hidir~ctlc~nal ~-;1vol?Lllfiion,Ir\ and\tomosis (BCP) Ma5 ad\,ocattzd 15, Xl ‘15 an alternati\zc> either as a part of a \taged mc)diiit,J f ontan proccdurt~ OI as a dcfinitivt> palliation. lo thi\ rtlport LZ’L’ anal\/<% the, immediate and niidtcrm outconic in a \erics ot patit,nt\ in whom multiplc~ riik tactor\ IIC’I-I’ ttncountered. and Methods Since 1991 a total of 25 patients underwent BCP for palliation of the following forms of cyanotic congenital heart diseases: univentricular heart (16), hypoplastic right \ entricle (6), and mitral stenosislatresia with small left ventricle (3). I&trictivc pulmonary blood flow was present in all (either pulmonary stenosis/atresia or pulmonary artery banding). ‘Twenty-two of the 25 patients (SS”/“) had undergone one or more prior systemic-pulmonary artery shunt\ (B-T shunt), and a previous pulmonary artery band \\‘a\ present in 2. Mean preoperative oxygen saturation \\CIS ‘ 71”,~. Significant degrees of cyanosis or incapacity detc~rmined the surgical indication. All patients were ~~~~~~udcd from a modified Fontan type of procedure bc~au~~ ot the presence of two or more of the following i-i\h factors (Table 1): I. :\ge. Eight (32”,,) patients were aged less than 12 months. Ages for the series ranged from 4 months to 9 \‘t’ars (mean, 2.5 years). 7. Severe pulmonary artery distortion. Pulmonary artery anatomv was considered inadequate in 7 patients r2X”,,) (sig 1). Anatomic findings as displayed from the preoperative angiograms were severe kinking and diffusc~ narrowing of the main right pulmonary artery l-1), of both main right and left pulmonary arteries (2), and anatomic discontinuity between right and left 0003-4975/95/$9.50 \SDI 0003-4975(95)00815-2 cardiographic examination. Eight additional patients had moderate degrees of AV valve insufficiency. 5. Subaortic obstruction. Seven patients (28%) had a restrictive ventricular septal defectlbulboventricular foramen causing subaortic obstruction, which was defined from hemodynamic criteria in 2 (gradient %50 mm Hg either basal or isoprenaline induced) or by morphologic criteria in 5 (angiographic or echocardiographic demonstration of a 50% or greater reduction in end-systolic caliber of the narrowest subaortic segment relative to aortic size) (Fig 2A). Five had a I)-loop and 2 an L-loop type of ventricular arrangement. Signs of severe ventricular hypertrophy were present in all. h. Anomalous systemic/pulmonary venous connection. Four patients had a bilateral superior vena cava. In 1, an interrupted inferior vena cava drained to the right ~upt~rior vena cava via hemiazigos continuation. Four patients had the pulmonar! vtxins draining to the right \idtb ~)f a common atrium, and the pulmonary venous IC,~UIn was directed to the right superior vena cava in 111,I ~a& mean puln~on.~r\~ artery pressure. Mean I’.\I’ was greater than 18 nlm Hg in 12 patients (48%). 1’11or \,t,ntricular function (7), A\/ \,alve incompetence I IO), rebtrictive foramen ovate (coexisting with left AV \,al\,e stenosis or atresia) (3), and previous BT shunts 112) bvt’re identified as concurrent causes for the incremented mean PAP. Pulmonarv vascular resistances iz t’rc Itlb!, than 4 Wood unitbln?. Bidirec-tional cavopulmonar): anastomosis was pertormed through a median sternotomy and cardiopulmonarv b! pabs in all patients. The azygos vein was ligated, c%xcept in a patient with interrupted inferior vena cava. Bilateral BCP was performed in patients with bilateral superior cava. The aorta was cross-clamped to perform an atrial heptectomy in 9 patients, for AV valvoplasty in 2, and for enlargement of the ventricular septal defect/ bulbo\~entricul,r foramen in 7. In the latter, the ventricular septal defect was enlarged toward the apex in 5 (D-v~~~tricular loop) and to the inferior part of the defect in 2 (L-ventricular loop) through a transaortic approach. Pulmonary artery distortion was treated by augmentation with .tutologous pericardium (7). Pulsatile pulmonar\’ blood flow was obtained from a previous contralaterai B7 shunt, which M’as left open in 10, and from the ventricltb (antc,grade flow) in 3. In all patients with the combinaltion of diastolic overload, depressed ventricular tunctitrn, and increased mean pulmonary artery prespure, no zourcc’ of pulsatile pulmonary blood flow was Itaft (B’I ligation). Pulmonary artery was ligated or closed bv o\.tsrsewing the pulmonarv valve leaflets from inside in -I patients. Kesults I hree ot the 25 patients died (12”,ll). Three or more risk tactor\ were present in each, including depressed ventl-ocular tunction, increased PAP, pulmonary artery dis- Ann Thorac Surg 1995;60:5558-62 tortion, and restricted bulboventricular foramen in 1, depressed ventricular function, AV valve insufficiencv, and increased PAP in another, and restrictive bulbovcntricular foramen, severe pulmonary arterv- distortion, anomalous systemic/pulmonarv venous connection, and increased mean PAP in the remaining 1 (see Table I). One patient was reoperated on within the first 23 hour\ postoperatively to ligate a previous BT shunt, which had been left open (pulsatile flokv) due to increased superior vena cava pressure and excessive pulmonarv blood flokv. In another patient with an L-loop ventricular arrangement in whom enlargement of bulbo\:cntricular foramen was done, complete AV block developed and a permanent pacemaker was implanted before discharge. Chvlothorax developed in 1 patient but responded to medical therapy. Persistent pleural effusions were not present in the remaining. Midterm follow-up ranged from 8 months to 3.8 years (mean, 2.1 months). There has been no late mortalit!. Mean ox)gt’n saturation of the survivors is 83”~~. The\ are routinelv receiljing diuretics, digoxin, and aspirin. Two-dimensional color Doppler studies showed remarkable improvement of the ventricular function in the survivors of the group with poor preoperative ventricular function (mean postoperative fractional shortening, 0.31; mean velocity ot circumferential fractional shortening, 1.5 c/s). Satisfactory size of the augmented pulmonar\ arteries was displayed by echocardiographic examination, M.ith unrestricted flow into the repaired pulmonary artery through the BCP. This correlated with adeyuatc pulmonarv blood flow shown both from the chest roentgenograms and radionuclear perfusion studies in the prcx\iouslv compromised lung. Doppler echocardiogt-aph\, aIs;, showed good function of the AV valve in the \ur\,ivor in \vhom an AV valvoplasty was performt~d. For the remaining 8 patients with moderate Al \.al\,ct incompetence, this markedly improved or disappeared. Satisfactory enlargement of the ventricular septal dt~tt~ctibulboventricular foramen with relief of the previous subaortic obstruction cvas confirmed by two-dimensional color Doppler studies in all patients (Fig 28). Pulsatile flow in the pulmonary artery was confirmed in patients with a pulsatile type of BCP (either those in whom a previous BT shunt was left open or those with antegrade flow from the ventricle) (Fig .?I. Radionuclear studies showed preferential perfusion of the right lung in patients with either pulsatile or nonpulsatilc flop. No difference in perfusion between the lungs \\‘a~ observed in patients in whom bilateral BCP was performed. Two patients underwent completion of Fontan circulation successfully 2 and 2.7 years after BCP. Comment Expansion of the indication for the Fontan procedure to a \\idr spectrum of patients with complex cardiac malformations has yielded to a more precise knowledge of the risk factors involved and their significance in terms of survival and quality of life [l-7]. At this point, we assume that patients with concurrent risk factors are definitely at J higher risk, making a modified Fontan procedure disadvisable. Bidirectional cavopulmonary anastomosis appears as a good alternative for staging procedures or as a definitive palliation. This study concerns a group of patients excluded from a modified Fontan operation due to multiple risk factors in whom BCP was performed. Anv statistical inference from the data in this small series i\ hpen to criticism. Due to the presence of several of these factors in each of the patients who died (see Table I ), It is difficult to infer their individual relevance as a cause of mortality. Satisfactory augmentation with autologous pericardium was obtained in all patients with preoperative distortion of the pulmonary arteries. Interestingly, inadequate pulmonary anatomy was the result of previous palliative operation in all but 1 of the patients reported. This fact should lead us to emphasize an effort to initially avoid BT shunts when po>siblt~ 13, 31 01., more ft~asible, to shorten the time elapsed between a 87 shunt performed in the neonatal period and the BC‘P 191. Ventricular function improv~~i markcdlv in the sur\,ivors in whom it was depresst-d yrt,opt~rati\,cl\. ‘Two <It them had been refcrrcld to our hospital for tranbplantation. The modification ot the vtrlumc 01 tnrload after BC’I’ and BT ligation (with AL’ \al\(’ rcpnir when ncccs>ar\ i should contribute to impro\ CY~ \,vntricular prrformanct~ in these patients. Both elimination crf the* hourcc ot diastolic overload and impro\ c>ment ot thts m~ocardlal oxvgenatlon added to the bc*nt+it obtainc>J M*ith tht, procedure. In thih regard, it i\ again important to con-sider the advantages ot perfornling HCI’ and khunt ligation earlier in life to minimilt, the potentially harmtul effect of long-standing diastolic overload on thcssc \,rntrlcles 191. This is even more t!<scntial in patient:, bvith ‘1 single morphologic right vt~ntric‘lt~. Significant degree>\ (It AV valve incompetence M,ill r<‘quirts \ur$cal repair at tht, time of BCP proccdurt~. t-loL\c\;c~r, h paticant5 in this st‘rics’ preoperativelv sho\\~i moderate dts:f-~‘ch ot A\’ \*,iI\ t’ insufficicnci,, ‘which improv~~d atttbr f3C I’ and RI shunt ligation. Ai ~‘a5 pre\,iouslL, rt>portcd, tht’ mentit~nt~d postoperati\,tx elimination ot that diastolic ventricular overload co~rld explain this ~~il~~nc~tnt~nc~~l /l)l. Complex cardiac anomalies with buhaorti< clb~tructil>n reprv>ent a difficult therapeutic probicm C.ombination ot a modified Darnils-Ka\,~-St,llis~,l anastr)mosl\ \t,ith a BC‘I’ has becr~ adwcated ‘24 <I mtd~td of tr-<~,ltrw~>t ior thc5<* patients in the prcsencc of unrt’stric tcti pulmonar\ bloo~i flow 151. UnfortunateI\, in this scrit,s. thta prc~5t~ii~‘~’ of pulmonary stenosis or atr(+ia ~*~ludt~d thli .Iltt>rnati\ c, the r~nlargemtint of tht, vt,ntri<ular \c%ptal def’ctti bulbovcntricular foramc~n beiris th<a singlca option. Patients with I,-loop ventricular ;Irrangt~nl~*nt a~‘<~at hight:r ri\h ot injur\. to the conduction system. One patient in our series required permanent pacemaker implantation als a result of surgically induced complete AV block. Adcyuatc relief of subaortic obstruction was obtained in this seric5 as demonstrated by the postoperative twodimensional color Doppler studies. The presence of anomalous 5vstemic/pulmonarv venous return encountc*rcd in oL,r &ics did not q&sent an additional probIcLm in management. However, other forms of anomalous pulnionar\~ venous return imply a more complex surgical tcchniquc~ to be performed at the time of BCP. In the ~IWWIIW ot elevated mean PAP as a result of previous shunts, .A\’ valve insufficiency, or depressed left ventricular function, a good functioning BCP was obtained (with s.lti5factor\ superior vena cava pressures) after BCP p~~rformancc and either BT shunt ligation or AV valve plastv (\2 ith consequent impro\;ement in the ventricular tunition). s\nalogous benefit on the PAP was obtained by c,nlarging .I restrictive foramen ovale in cases with mitral \,al\ c’ \tcnosis/atresia and small left ventricle. NevertheIi’+, undr’sirablcx results arc to be expected if preoperati\c pulmonarv vascular resistances are greater than 4 111)cd unit5iml’. Low ,rgts did not show itself as a factor affecting mc~rtalitv. On the other hand, the advantages of performing RC I’ as carlv as possible have been discussed above IL)]. Iwclvc patients in our series were left with a pulsatile source of pulmonary blood flow. When deciding, it is diEi<-uft to predict preoperative/y if functioning of the id\.olIllllmonarv circulation will be compatible with a pul\atilt~ source of pulmonary blood flow. One of our patient\ rt’quirtxd ligation of a BT shunt during the first 24 hlrurs postc)ptrati\;el~. We have no evidence to support that this ht)modvnamic condition is preferable to a nonpuI5atilc~ situation. Wc can speculate that pulsatility 5562 c.iir,jr;\t.l).A Rt’ih t ,\ClC)R’, tt’15.i ttI<tt I IN C \\ \ \tit,\5 01’1 I klO\ \li\l I I \I 5ttL Ann l-horac 2 t be of benefit by avoiding the potential development of pulmonarv arteriovtanous fistula; it also could help provide normal growth of the contralateral pulmonary artery after BCP. These theoretic advantages need further clinical confirmation. Finally, although BCP appears a rc>awnable alternative in this group of patients, some doubts arise regarding their management in the long term [lOI. Some of them will never become ideal Fontan candidates, some will require coil embolization of vkmous collaterals, and perhaps some will be considered for transplantation. On the other hand, some patients \vill bt, suitable for modified Fontan operation, and in this c“aw it becomes difficult to define the proper timing for the scwmd stage. In our opinion, there is no clear adl,antagt) in offering a different kind of palliation to a patitant Lvho i\ doing reasonably well. could References 1. Mayer JE, Hclgason ti. Ionas I<:\, ~‘1 al t \tcnding the limits for modified Fontan procc~durt~~. I I horac Cardiovasr Surg lYXh;Y2:1021-8. 2. Mietus-Snyder \I, Lan :; r, x1‘l\cr II , 1011J~ IiA, C‘IstariLda AR. Childhood ~\.~tcmic-yllIn~c~n,~r\ -hunts: \ul>wqut*nt Surg 199j;6035558-62 3. 1. 5. 6. 7. 8. 9. IO. suitability for Fontan operation. Circulation 1987;76(Suppl 3):3Y -44. Alboliras ET, Chin AJ, Barber G, Helton JG, Pigott JD, Norwood WI. Pulmonary artery configuration after palliative operations for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 1989;97S78-85. Caspi J, Coles JC, R. , .ovitch M, et al. Morphological findings contributing to a failed Fontan procedure:twelve year experience. Circulation 1990;82(Suppl 4):177-82. Di Donato RM, Amodeo A, di Carlo DD, et al. Staged Fontan operation for complex cardiac anomalies with subaortic obstruction. J Thorac Cardiovasc Surg 1993;105:398-405. Vargas FJ, Mayer JE, Jonas RA, Castafieda AR. Anomalous systemic and pulmonary venous connections in conjuction with atriopulmonary anastomosis (Fontan-Kreutzer). J Thorat Cardiovasc Surg 1987;93:523-32. Trusler GA, Williams WG, Cohen AJ, et al. The cavopulmonary shunt: evolution of a concept. Circulation 1990;82(Suppl 4):131-8. Bridges ND, Jonas RA, Mayer JE Jr, Flanagan MF, Keane JF, Castafieda AR. Bidirectional cavopulmonary anastomosis as interim palliation for high-risk Fontan candidates: early results. Circulation 1990;82(Suppl 4):170-6. Chang A, Hanley F, Wernovsky G, et al. Early bidirectional cavopulmonary shunt in young infants. Circulation 1993;88: 149-50. Jonas R. Indications and timing for the bidirectional Glenn shunt versus the fenestrated Fontan circulation. J Thorac Cardiwasc Surg 1994;108:522-4.