Clinicopathological correlates of obstructed right-sided porcine-valved extracardiac conduits

J THORAC

CARDIOVASC SVRG

81:591-601,1981

Clinicopathological correlates of obstructed right-sided porcine-valved extracardiac conduits Thirteen of 308 patients (4.2%), who had received right-sided valved extracardiac conduits at the Mayo Clinic from November. 1972, to April, 1977, have required conduit replacement because of obstruction. Patients were 5 to 16 years old at initial operation and 8 to 20 years old at reoperation; the duration of conduit implantation was 27 to 79 months (mean 50). Four patients (31%) were asymptomatic; exertional dyspnea was present in eight (62%) and dizziness was reported in one (8%). Signs of conduit obstruction included increasing intensity of murmurs in 11 (85%), cyanosis in two (15%), and heart failure in one (8%). Peak pressure gradients from the right ventricle to the pulmonary arteries ranged from 50 to 140 mm Hg (mean 87) and correlated well with the degree of conduit obstruction. Catheter pullback allowed accurate localization of stenosis within the conduit, whereas angiography alone did not. The site of major obstruction was in the proximal conduit in five (38%), at the valve in nine (69%), and in the distal conduit or side branches in six (46%); stated differently, major stenosis affected the valve alone in five (38%), the graft alone in four (31%), and both the valve and the graft in four (31%). Valvular changes leading to stenosis included thrombosis, commissural fusion, and calcification, and changes consistent with insufficiency included tears, fusion of cusps to the conduit wall, and, in one case, infective endocarditis. Within the conduit, nonvalvular obstruction was due to formation of a thick fibrous peel (or neointima). Progressive thickening of the peel appeared to be due to organization of thrombus between the peel and conduit and not due to luminal mural thrombus. In one case, the peel formed a flap-valve, causing even further obstruction. Since many patients are asymptomatic, and since late conduit stenosis may develop unpredictably by several mechanisms, long-term follow-up is necessary.

Kishan C. Agarwal, M.D.,* William D. Edwards, M.D.,** Robert H. Feldt, M.D.,*** Gordon K. Danielson, M.D.,**** Francisco J. Puga, M.D.,**** and Dwight C. McGoon, M.D.,**** Rochester, Minn.

T

he surgical use of valved extracardiac conduits has allowed total repair of complex congenital cardiac anomalies (e.g., complete transposition, truncus arteriosus, and pulmonary atresia with ventricular septal defecr'r") which otherwise would have received only palliative procedures. Our early experience with frozen From the Departments of Anatomic Pathology, Pediatrics, and Surgery, and from the Mayo Graduate School of Medicine, Mayo Clinic and Foundation, Rochester, Minn. Received for publication Aug. 28, 1980. Accepted for publication Sept. 16, 1980. Address for reprints: William D. Edwards, M.D., Department of Anatomic Pathology, Mayo Clinic, Rochester, Minn. 55905. *Fellow in Pediatric Cardiology. **Department of Anatomic Pathology. ***Division of Pediatric Cardiology. ****Division of Cardiothoracic Surgery.

irradiated aortic homografts was disappointing because of progressive calcification and subsequent valvular stenosis. 6. 7 Even for "fresh" aortic homografts, late postoperative calcification and problems in availability still pose major obstacles. 8 Synthetic (corrugated woven Dacron) tubular grafts containing glutaraldehyde-fixed porcine aortic valves" were developed to avoid these shortcomings of homograft conduits. However, long-term postoperative follow-up indicates that conduit stenosis may well be a significant late complication'v " and that the porcine valve is not necessarily the major site of obstruction. Potential sites and mechanisms of conduit stenosis include the following:

0022-5223/81/040591+11$01.10/0 © 1981 The C. V. Mosby Co.

I. Early postoperative and intraoperative

A. Conduit too small for patient B. Extrinsic compression by sternum C. Obstructed proximal or distal anastomosis

591

The Journal of Thoracic and Cardiovascular Surgery

592 Agarwal et al.

Table I. Clinical and pathological data from 13 patients with obstructed right-sided porcine-valved extracardiac conduits Case No. I 2

3 4 5

6 7 8 9 10

II 12 13

Age. sex" 5 yr, 2 rno, M 14 yr, 2 mo, M 16 yr, 0 rno, F 5 yr, II rno, M

5 yr, 15 yr, II yr, 5 yr, 7 yr, II yr, 7 yr, 6 yr,

Timet (mo)

Peak systolic gradients. (mm Hg)

50

44

79 42

55 55

Percent obstruction (of cross-sectional area) Proximal

o

40

19

30 35

60

33

75

I mo, F

27

79

81 87

I mo, M 0 mo, F

58 36 55

90 100 100 105 110 114 140

4 mo, F

6 9 5 3 7 yr, 6

rno, M mo, F mo, F mo, M mo, M

60

58 50 73

* Age at initial placement of porcine-valved conduit.

I

Valve 87

I

Distal

o

55 60

28 19

53 44

40 23

I

Right

I

Left

v

V V V P P

50

#

#

94

60

20

90

83

44

30

36 36

83

97

87

63

56

40

81

36

91 16

67

95

75

85

o

Sites of majorobstruction§

V,D P, D D

P, V,D V

V,D V,D

P,

F. Female. M. Male.

t Time between conduit placement and reoperation. tFrom right (or pulmonary) ventricle to pulmonary arteries. §D, Distal conduit or right and left side branches. P, Proximal conduit. V, Porcine valve. lIRight side branch not available, but noted by surgeon to be totally occluded. HOnly obstructive proximal peel removed; valved conduit left in place.

II. Late postoperative A. Patient outgrows conduit B. Valvular stenosis C. Fibrous peel (neointima) or thrombus Reported herein are the clinical and pathological findings in 13 patients who have required replacement of obstructed right-sided valved Hancock conduits. Patients and methods Of 308 patients receiving right-sided porcine-valved extracardiac conduits at the Mayo Clinic from November, 1972, to April, 1977, 13 have required conduit replacement (from October, 1976, to April, 1980) because of conduit stenosis at the level of the graft, the valve, or both. The surgically excised specimens were available for review in 11 of these 13 patients. An additional two specimens were available in which the conduits had been initially inserted elsewhere but the obstructed conduits were subsequently replaced at our institution. Therefore, 13 specimens were available for detailed clinical and pathological evaluation. Excluded from this study were aortic homografts and porcinevalved conduits placed from the right atrium to the pulmonary trunk. Review of hospital charts provided clinical information concerning probable signs and symptoms of progressive conduit stenosis. The severity and sites of

conduit obstruction had been documented in all cases by cardiac catheterization and angiography. In addition, pressure gradients related to the conduit were measured again at reoperation in nine patients. The abnormalities of all 13 conduit specimens were reviewed in detail. Each porcine valve was evaluated grossly for distortion, calcification, tears, thrombosis, adhesions, commissural fusion, and degree of stenosis. The percent valvular obstruction was calculated by comparing the actual cross-sectional valve area to the anticipated normal valvular area (Table I); normal valvular diameters were considered to be 2 mm less than the conduit diameter. A roentgenogram of each valve was utilized to evaluate further the sites and extent of calcification. In selected cases, the valves were decalcified and sectioned for light microscopy (Fig. 1, right). Obstructive luminal fibrous peels were inspected grossly with respect to thickness, calcification, fenestrations or tears, thrombosis, detachment from the underlying conduit, and the degree of stenosis they produced. The percent stenosis of each conduit specimen, both proximal and distal to the porcine valve, was calculated by comparing the minimal luminal cross-sectional area to the anticipated normal cross-sectional area at that site (Table I); the stated conduit size was considered the normal diameter. In each case, multiple

Volume 81 Number 4 April. 1981

Valved extracardiac conduits

593

AORTA

/

/

/

CONDUIT

LUMEN

,

-.

PERICARDIUM/

~···

•• PLANE OF TISSUE SECTION

./

"'PLANES OF/ TISSUE SECTION

LUMEN

~

'PERICARDIUM

\lI

COMMISSURE CUSP-··'"

PEEL----- •

CONDUIT

I J

_n_--~M'--~

••••••••• CONDUIT ---.-.•••• ~

CUSP

:::

Fig. 1. Orientation of tissue sections for microscopy. Left. Section of conduit graft with luminal fibrous peel. Pericardium refersto fibrous coaton pericardial surface of conduit. Right. Sections of opened porcine valve from

conduit, taken either longitudinally (radially) from midportion of cusp or horizontally (circurnferentially) at commissure.

sections of the fibrous peel were submitted for light microscopy (Fig. I, left), following decalcification when necessary. All tissues submitted for light microscopy were processed routinely and stained with hematoxylin and eosin, Mallory-Heidenhain trichrome, elastic-van Gieson, and the Alcian blue-periodic acid-Schiff reaction.

Results Clinical findings. Of the 13 patients (11 initially operated upon at the Mayo Clinic and two, elsewhere), seven were male and six were female. At the time of initial conduit placement, patients ranged in age from 5 to 16 years (mean 9 years), and at the time of conduit replacement, the age range was 8 to 20 years (mean 13 years). The 13 Hancock conduits were in place 27 to 79 months (mean 50 months) before progressive obstruction necessitated their excision (Table I). One patient (Case 6) previously had had an aortic homograft replaced after 4 years because of calcific obstruction of the valve. The diameter of the original Hancock conduits ranged from 16 to 25 rom (mean 20 mm). In one patient (Case 6), only the proximal peel was removed and the conduit was left in place. In the remaining 12 patients, the conduit was excised and was replaced with another conduit-an 18 rom and a 22 rom Hancock conduit in two patients and a 25 rom conduit in 10

patients (five Carpentier-Edwards, three Hancock, and two nonvalved Meadox grafts). The underlying congenital cardiac anomalies included four cases of complete transposition of the great arteries, four cases of pulmonary atresia with ventricular septal defect, two cases of truncus arteriosus, and one case each of univentricular heart, double-outlet left ventricle, and double-outlet right ventricle with atrioventricular discordance. The extracardiac conduit was placed from the right ventricle to the pulmonary trunk in eight cases, from the right ventricle to both the right and left pulmonary arteries in three (T-graft in two and right side branch in one), from the morphologically left ventricle to the pulmonary trunk in one case of atrioventricular discordance, and from a septated univentricular heart to the pulmonary trunk in one case. Four patients were asymptomatic, and conduit obstruction was discovered only at routine follow-up cardiac catheterization. Nine patients were symptomatic. Symptoms, which in eight patients were primarily manifestations of exercise intolerance, included tiredness, fatigue, exertional dyspnea, and leg pain. One patient had several episodes of dizziness. Signs of conduit stenosis in the 13 patients included congestive heart failure in one patient, cyanosis in two patients, and increasing intensity of cardiac murmurs in eleven; information concerning murmurs was not available in

594 Agarwal et al.

The Journal Thoracicand Cardiovascu Surg4

Fig. 2. Stenosisof porcine valvescausedby progressive calcification and commissural fusion. Valvesare viewed from above (a. d, and g) and from below, both grossly (b. e, and h) and roentgenographically (c,J. and i). a-c, Early calcification of porcine aorta and aortic valve commissures, with mild stenosis resulting from muscular band in lower valve cusp (Case 8). d-f, Moderate calcification resulting in rigid cusps and one fused commissure (Case 11). g-i, Severe calcification of porcine aorta and aortic valve cusps, with markedly deformed and rigid cusps and fusion of two commissures (Case 7). two patients. (Of the 13 patients, seven were thought to have residual ventricular septal defects at cardiac catheterization, but only five of these were verified at reoperation; in only two patients was interventricular shunting considered significant.) In the four asymptomatic patients, peak pressure gradients from the right ventricle to the pulmonary arteries ranged from 79 to 114 mm Hg (mean 96 mm Hg). In the nine symptomatic patients, peak pressure

gradients ranged from 50 to 140 mm Hg (mean 83 m Hg). At cardiac catheterization, continuous pressu recording during pullback of the catheter from the pu monary arteries into the right ventricle provided tI most accurate preoperative means of assessing tI site(s) of obstruction within the conduit (Table I). Rig ventricular angiography alone was not reliable localizing the sites of stenosis but did allow evaluatic of valvular regurgitation. In only six of 13 cases w

Volume81 Number4 April. 1981

conduit stenosis correctly identified or localized by angiography alone. Pathological findings. Porcine valve. Valvular stenosis was either inherent or acquired. In no case was inherent stenosis the result of the initial placement of a conduit which was relatively too small for the patient. (In eight patients in whom intraoperative pressures in the right ventricle and in the pulmonary arteries had been measured following insertion of the initial Hancock conduit, the pressure gradient ranged from 6 to 25 mm Hg (mean 18 mm Hg). By the time of reoperation, peak pressure gradients in these eight patients ranged from 50 to 114 mm Hg (mean 80 mm Hg). Therefore, increases in gradients ranged from 30 to 108 mm Hg (mean 62 mm Hg). Inherent stenosis was present in one patient (Fig. 2, a to c) and was due to a muscular ridge within one aortic cusp; the degree of stenosis was considered mild. More commonly, valvular stenosis was acquired, being the site of major obstruction in nine of 13 (69%) and causing 75% or greater luminal obstruction in four of these nine (Table I). Calcification underlay acquired valvular stenosis in all nine conduits. It began as mineralization of the porcine aorta near the commissures (Fig. 2, a to c) and progressively involved the valvular commissures until they were fixed in a semiclosed position (Fig. 2, d to/). In severely stenotic valves, nodular calcific excrescences were present throughout the cuspid tissue and resulted in immobile and distorted valves (Fig. 2, g to 0. Osseous metaplasia was occasionally identified, not only in degenerating cuspid tissue but also in disrupted porcine aorta. Thrombosis also contributed to valvular stenosis in each instance by causing commissural fusion with fixation of cusps in a semiclosed position (Fig. 3). Such thrombi were almost invariably associated with calcification and focal disruption of cuspid tissue. It is unclear whether thrombosis preceded calcification, or vice versa, but both processes seemed to accompany degenerative changes in collagen and elastin. Valvular incompetence frequently coexisted with stenosis and was primarily due to cuspid tears or due to fusion of cusps to either the porcine aorta or a distal conduit peel. Tears, present in six of 13 cases (46%), resulted in flail valves and were most commonly encountered in cusps involved by platelike calcification with eggshell fragility (Fig. 4, a and b). In one case (Fig. 4, c to e), extensive tears and focal loss of valve tissue were considered to be due to healed infective endocarditis, and this was supported by the clinical history; bacterial stains of the valve were negative.

Valved extracardiac conduits

595

Fig. 3. Photomicrographs of stenotic porcine valves. Commissural fusion by partially organized thrombus (Thr) with focal calcification (Ca). Focal fragmentation and destruction (arrows) of cuspid tissue by calcific concretions. (a, Case 1, Mallory-Heidenhain trichrome, x 20; b, Case 11, elastic- van Gieson, x20.)

Thrombosis in some cases was present along the proximal (ventricularis) side of the valve (Fig. 5, a and b) but more commonly and more severely involved the distal (arterial) side. This resulted not only in wrinkling and retraction of cuspid tissue but also in fusion of cusps to the aortic or conduit wall, which focally fixed the valve in a semiopened position (Fig. 5, c to e). Conduit. In general, a fibrous peel (neointima) was present within each conduit, although the thickness varied considerably and caused little or no obstruction in some. However, the fibrous peel resulted in major obstruction in that portion proximal to the porcine valve (proximal conduit) in five of 13 cases (38%) and in that beyond the valve (distal conduit or side branches) in six of 13 (46%). Luminal obstruction of 75% or greater was calculated in four proximal conduits and in four distal conduits or side branches (Table I). Grossly, the peel was firm and gray white, with a relatively smooth luminal surface; however, fenestrations were characteristically present in the peel (Fig. 6). Although the peel was adherent to the surrounding conduit in some cases, in most it was separated from the

596

Agarwal et al.

The Journal of Thoracic and Cardiovascular Surgery

Fig. 4. Incompetence of porcine valvesbecauseof tearsand fusionof cusps with wall of aortaor conduit. a. Tear in cusp (arrowhead) and fusionoftwo cuspsto aorticwall at commissure (arrow). Viewedfrom above(Case 10). b. Extensive tear of one cusp (arrowheads), resulting in flail prolapsing cusp. Viewed from below (Case 4). c-e, Suspected case of healed infective endocarditis, withdestruction, calcification, and retraction of cuspidtissue (Case 3) (c. from above; d. from below; e, opened as in Fig. I, with arrows at commissures). conduit by a layer of thrombus (Figs. 6 and 7, a). In one case (Fig. 7, b and c), an extensive thrombus separated the peel from the conduit and caused severe obstruction (analogous to an intramedial aortic dissection). In yet another case (Fig. 8), portions of both the proximal and distal peels had been lifted off the conduit and lay loose (as obstructing valvelike ridges) within the lumen. The side branches of T-grafts tended to be severely stenotic (Table I). Microscopically, the conduit peels were quite similar, regardless of thickness or apparent age. Each peel had two arbitrary regions, namely one nearer the lumen (luminal portion) and the other nearer the conduit mater-

ial (conduit portion). The luminal portion was densely collagenous and contained only a few fibroblasts and infrequent capillary channels (Fig. 9, a). Occasionally, shallow organized mural thrombi were identified, but these contributed very little to the thickness of the peel. The conduit portion usually consisted of active granulation tissue, with fibroblastic proliferation, neovascularization, focal clusters of leukocytes (lymphocytes, neutrophils, plasma cells), and early collagen deposition (Fig. 9, a and b). Between the peel and conduit, varying amounts of thrombotic material were present, ranging in age and content from fresh platelet-fibrin white thrombi to classic red thrombi to old

Volume 81

Valved extracardiac conduits

Number 4 April. 1981

597

Fig. 5. Photomicrographs of incompetent porcine valves. a and b, Small organized thrombus (Thr) along ventricularis (proximal) surface of cusp, resulting in wrinkling and retraction of valvular fibrosa (arrowheads) (b is close-up of boxed area in a). c-e. Organized thrombus (Thr] along ventricularis (proximal) and arterial (distal) surfaces of cusps, the latter resulting in fusion of cusps (arrowheads) to aortic wall (Ao). There is aortic disruption indo (a and b. Case 8, Mallory-Heidenhain trichrome; c-e, Cases II, 13, and 3, elastic-van Gieson; a, c-e, x7; b, x20.)

necrotic thrombotic debris. Neither capillaries nor fibroblasts with collagen anchored the fibrous peel to the synthetic conduit weave (Fig. 9, c). Fibrous tissue along the outer surface of the conduit, representing proliferative tissue of epicardial or parietal pericardial origin, also failed to anchor securely to the conduit weave (Fig. 9, c and d). As a result, in most cases both the peel and the overlying fibrous tissue could be easily detached from the conduit by blunt dissection. Whereas foreign body giant cells generally lined the conduit externally, they were distinctly uncommon at the interface between conduit and luminal peel. The presence, within the fibrous luminal portion of the peel, of an occasional giant cell with phagocytosed foreign graft material very strongly suggested that the peels had become thickened by organization of

thrombus between conduit and peel rather than by organization and progressive layering of luminal mural thrombi. Discussion

Aortic homografts and porcine-valved extracardiac conduits have been utilized to reconstruct the right ventricular outflow tract in the repair of complex forms of congenital heart disease for which only palliative procedures were previously available. 1-5 Calcific stenosis of aortic homografts is a widely known late complication":" and has led to preferential usage in many institutions of Dacron conduits containing porcine aortic valves. 9 Only recently has long-term follow-up of this latter group revealed that obstruction of porcine-valved extracardiac conduits may well be a

The Journal of

598 Agarwal et al.

Thoracic and Cardiovascular

Surgery

Fig. 6. Conduit obstructionby proliferative luminal fibrous peel. T-graft (with proximal[Prox] portion and right [R] and left [L] distal side branches) has completeobstructionof normallumen of right branch, with development of false channel to right pulmonary artery via fenestration in peel (Case 7). a, Proximal peel separated from conduit by thrombus (Thr). Defect in proximal peel (arrow) leads to thrombus. b, Anterior half of conduit removed (except over porcine valve), revealingfibrous peel throughout. Defect in proximal peel (arrow) leads to thrombus (Thr), while defect in distal peel (black arrowheads) leads to false channel (white arrowheads). c, Obstructed right branch, with false channel (Fe). d, Marked concentric obstruction of left branch. significant late postoperative complication.tv P In almost all reported cases, obstruction has been attributed to calcific valvular stenosis. In one of the cases reported by Ciaravella and associates," nonvalvular conduit obstruction by a fibrous peel was described. (Although this case was reported from our institution, the surgically excised conduit was one of two that was sent to the manufacturer; therefore, it was not available for study and is not included in the present series.) In contrast to these published reports,10-14 in our experience major conduit stenosis involved the valve alone in only 38%. In 31%, obstruction was due primarily to a fibrous peel (or neointima), and in an additional 31%, major obstruction was caused both by valvular stenosis and by a thick fibrous peel. Although nonvalved conduits may be used in selected cases, clearly their use, when constructed of the same material as these conduits, may not eliminate late postoperative conduit obstruction in the majority of patients. Those patients with isolated valvular stenosis tended to have

the lowest pressure gradients, whereas those with both valvular stenosis and a thick fibrous peel usually had the highest pressure gradients (Table I). The distal side branches of T-grafts were characteristically severely stenotic. In general, the pressure gradient across the conduit correlated very well with the maximal degree of obstruction (Table I); in only two patients (Cases 1 and 8) did this generalization fail. Although the degree of conduit obstruction documented both clinically and pathologically was severe in all 13 cases, four patients were asymptomatic. Nine patients were symptomatic, but in eight the symptoms were rather nonspecific clinical manifestations of exercise intolerance. The most common sign of apparent conduit obstruction was increasing intensity of cardiac murmurs, heard in all cases in which such information was recorded. However, in many cases the murmur was attributed to residual ventricular septal defects. The true incidence of conduit stenosis may be much higher than this current study indicates, since (1) many

Volume 81 Number 4 April. 1981

patients with significant conduit obstruction may be asymptomatic or have rather nonspecific symptoms, (2) these patients may have murmurs resulting from conduit stenosis which are attributed to other hemodynamic alterations, and (3) some patients initially operated upon at the Mayo Clinic may have had replacement of obstructed conduits elsewhere. Since conduit stenosis may develop to a significant extent as early as 27 months after implantation, perhaps routine long-term follow-up by cardiac catheterization is indicated. Cardiac catheterization is currently the best method for clinically assessing the site and severity of conduit obstruction. Angiography does not appear reliable, and the history and physical examination alone often yield inadequate information upon which to make the diagnosis of conduit stenosis. Correlative studies utilizing two-dimensional echocardiography are not yet available. Porcine valvular stenosis was primarily due to calcification and thrombosis of commissures, resulting in rigid cusps held in a semiclosed position. Valvular changes suggestive of incompetence frequently accompanied the stenotic lesions and included cuspid tears, commissural thrombi holding the valves partially opened, and thrombosis of the valvular sinuses, resulting in thrombotic adherence of cuspid tissue to the underlying conduit or porcine aorta. Similar changes of degeneration, calcification, and thrombosis have been reported in isolated intracardiac porcine bioprostheses 14- 16 and are more common in children than adults.": 17. 18 Calcification of porcine valves in children has been attributed, in part, to increased calcium turnover; none of our 13 patients had known disturbances of calcium metabolism. In our series, the other element of significant obstruction was the fibrous peel (or neointima) which developed within the Dacron graft. Such a peel caused major obstruction in 62%, either with or without concomitant valvular stenosis. To our knowledge, the characteristics of fibrous peels have not been previously reported, although the potential for obstruction of tubular Dacron grafts is known. 19, 20* Although the pathogenesis of such peels is not entirely clear and will form the basis of a future communication, t one concept is quite clear: Progressive thickening of the peel occurs

*Ben-Shachar O. Nicoloff OM, Edwards JE: Separation of neointima from Dacron graft causing obstruction. Case foIlowing Fontan procedure for tricuspid atresia. (Unpublished data.) tAgarwai KC, Edwards WO, Feldt RH, Danielson OK, Puga FJ, McGoon DC: Pathogenesis of obstructive fibrous peels in rightsided porcine-valved extracardiac conduits. (Unpublished data.)

Valved extracardiac conduits

599

Fig. 7. Thrombus between peel and conduit. a, Typical small thrombus (Case 4). b and c. Unusually extensive thrombus, with marked conduit obstruction, extending to level of porcine valve (Case 12) (b, cross-section; c, longitudinal section). from the surface in contact with the conduit and not from that in contact with the luminal flow of blood. Therefore, luminal mural thrombus probably plays little if any significant role in the development of severely obstructive fibrous peels. Rather, we hypothesize that, through fenestrations in the peel, luminal blood dissects portions of the peel away from the conduit and thereby forms many blind channels which may eventually coalesce. Apparently, stasis of blood is common in these channels, since

600 Agarwal et al.

The Journal of Thoracic and Cardiovascular Surgery

Fig. 8. Separation of peel from conduit, forming obstructive flap-valve, with blind channel between peel and conduit. Moderately obstructive flap (arrowheads) in distal conduit (a), and severely obstructive flap (arrowheads) in proximal conduit (b) (Case 10).

Fig. 9. Photomicrographs of conduit peels. a, Low-power view of entire thickness, showing densely fibrous (F) luminal portion with thin focal mural thrombus (MT). Active granulation tissue (GT) and recent thrombus (Thr] occupy the portion nearer the conduit (Case 8). b, Highly vascular granulation tissue in region of peel near the conduit (Case 6). c, Peel, conduit, and external fibrous coat (pericardium), showing lack of fibrous bridging between peel and pericardium (Case 3). d, Conduit and vascularized external ("pericardial") coat, with absence of either vascular or fibroblastic proliferation through the weave of Dacron conduit (C) (Case 2). (All, hematoxylin and eosin; a, x 10; band d, x50; c, x20.)

Volume 81 Number 4

Valved extracardiac conduits

60 1

April,1981

thrombus is characteristically found between the conduit and its peel at these sites. The thrombus becomes organized, and recurrent separation of the peel from the conduit allows progressive fibrous thickening. In this manner, the luminal portion of the peel is the older and more fibrotic layer, whereas the conduit portion is younger and contains granulation tissue and thrombus. Since vascular channels were not identified passing through the conduit from the external fibrous layer to the fibrous peel, it is highly unlikely that the thrombus between the peel and conduit results from rupture of transconduit capillaries. Perhaps with each heartbeat, the rhythmic, accordion-like rocking motion of the corrugated conduit contributes not only to the initial formation of fenestrations but also to the early loosening of the peel from the conduit. The ease of dissection of blood through cleavage planes between the peel and conduit is probably greatly facilitated by another factor: The Dacron weave of the conduit is made deliberately tight and nonporous for hemostatic reasons. However, this greatly hinders the formation of fibrous anchors between the conduit and peel, since plump active fibroblasts cannot easily penetrate the small interstices of the Dacron weave. As a result, the peel easily strips away from the underlying conduit.

2 3

4

5

6

7

REFERENCES Ross DN, Somerville J: Correction of pulmonary atresia with a homograft aortic valve. Lancet 2: 1446-1447, 1966 McGoon DC, Rastelli GC, Ongley PA: An operation for correction of truncus arteriosus. JAMA 205:69-73, 1968 Rastelli GC: A new approach to "anatomic" repair of transposition of the great arteries. Mayo Clin Proc 44: 112, 1969 Rastelli GC, McGoon DC, Wallace RB: Anatomic correction of transposition of the great arteries with ventricular septal defect and subpulmonary stenosis. J THORAC CARDIOVASC SURG 58:545-552, 1969 Kouchoukos NT, Barcia A, Bargeron LM, Kirklin JW: Surgical treatment of congenital pulmonary atresia with ventricular septal defect. J THORAC CARDIOVASC SURG 61:70-83, 1971 Merin G, McGoon DC: Reoperation after insertion of aortic homograft as a right ventricular outflow tract. Ann Thorac Surg 16:122-126, 1973 Moodie DS, Mair DD, Fulton RE, Wallace RB, Danielson GK, McGoon DC: Aortic homograft obstruction. J THORAC CARDIOVASC SURG 72:553-561, 1976

8 Saravalli OA, Somerville J, Jefferson KE: Calcification of aortic homografts used for reconstruction of the right ventricular outflow tract. J THORAC CARDIOVASC SURG 80:909-920, 1980 9 Bowman FO Jr, Hancock WD, Malm JR: A valvecontaining Dacron prosthesis. Its use in restoring pulmonary artery-right ventricular continuity. Arch Surg 107: 724-728, 1973 10 Rosenthal A, Rocchini AP, Keane JF, Castaneda AR, Nadas AS: Hemodynamics after surgical repair with right ventricle to pulmonary artery conduit (abstr). Circulation 52:Suppl 2:102, 1975 11 Norwood WI, Freed MD, Rocchini AP, Bernhard WF, Castaneda AR: Experience with valved conduits for repair of congenital cardiac lesions. Ann Thorac Surg 24:223232, 1977 12 Ciaravella JM Jr, McGoon DC, Danielson GK, Wallace RB, Mair DD, Ilstrup DM: Experience with the extracardiac conduit. J THORAC CARDIOVASC SURG 78:920-930, 1979 13 Bisset GS, Schwartz DC, Benzing G, Helmsworth J, Kaplan S: Late results of reconstruction of the right ventricular outflow tract with valved external conduits in children (abstr). Am J Cardiol 45:448, 1980 14 Geha AS, Laks H, Stansel HC Jr, Cornhill JF, Kilman JW, Buckley MJ, Roberts WC: Late failure of porcine valve heterografts in children. J THORAC CARDIOVASC SURG 78:351-364, 1979 15 Ferrans VJ, Spray TL, Billingham ME, Roberts WC: Structural changes in glutaraldehyde-treated porcine heterografts used as substitute cardiac valves. Transmission and scanning electron microscopic observations in 12 patients. Am J Cardiol 41: 1159-1184, 1978 16 Ashraf M, Bloor CM: Structural alterations of the porcine heterograft after various durations of implantation. Am J CardioI41:1185-1190, 1978 17 Silver MM, Pollock J, Silver MD, Williams WG, Trusler GA: Calcification in porcine xenograft valves in children. Am J Cardiol 45:685-689, 1980 18 Thandroyen FT, Whitton IN, Pirie D, Rogers MA, Mitha AS: Severe calcification of glutaraldehyde-preserved porcine xenografts in children. Am J CardioI45:690-696, 1980 19 Mair DD, Fulton RE, Danielson GK: Thrombotic occlusion of Hancock conduit due to severe dehydration after Fontan operation. Mayo Clin Proc 53:397-402, 1978 20 Gale AW, Danielson GK, McGoon DC, Wallace RB, Mair DD: Fontan procedure for tricuspid atresia. Circulation 62:91-96, 1980