Use of Intravascular Stents in Systemic Venous and Systemic Venous Baffle Obstructions
Short-term Follow-up Results
Background Balloon-expandable intravascular stents are well accepted in the management of arterial obstructions. This study was undertaken to detail the immediate and short-term results of intravascular stent implantation in systemic venous and systemic venous baffle obstructions in children.
Methods and Results Between September 1991 and June 1994, 12 patients had 21 stents implanted in 13 systemic venous obstructions, 1 patient having stents placed in 2 separate obstructions. In the baffle group, 4 of 13 obstructions were at the superior vena cava/right atrial junction after atrial baffling for transposition of the great arteries. One of 4 patients had complete obstruction requiring transseptal needle perforation before stent implantation. There was an immediate gradient reduction from 12±8.4 mm Hg (range, 4 to 20 mm Hg) to 1.3±1.9 mm Hg (range, 0 to 4 mm Hg, P=.05). The obstructed segment diameter increased from 3.5±3.9 mm (range, 0 to 8.5 mm) to 16±2.7 mm (range, 14 to 20 mm, P=.002). In the central vein group, 9 of 13 obstructions were in large central veins. Three of 9 patients had complete obstruction requiring transseptal needle perforation before stent implantation. There was an immediate gradient reduction from 10.3±8.5 mm Hg (range, 0 to 20 mm Hg) to 0.8±1.1 mm Hg (range, 0 to 3 mm Hg, P=.005). The obstructed segment diameter increased from 1.3±1.1 mm (range, 0 to 2.8 mm) to 9.4±1.7 mm (range, 7.6 to 12 mm, P<.001). There were no acute complications in either group. In the follow-up group, patients were scheduled for clinical follow-up at 3, 6, and 12 months with echocardiography or magnetic resonance imaging (MRI) at 3 or 6 months and for repeat cardiac catheterization at 12 months. All stents were patent by echocardiography or MRI when studied at follow-up. Cardiac catheterization in 6 of 12 patients, 2 to 13 months after stent, demonstrated that all stents remained patent without compression or fracture. Follow-up and immediate poststent gradients were not significantly different (1±1.6 versus 0.7±1.2 mm Hg, P=NS). Neointimal hyperplasia (5 of 6 patients) reduced the stent lumen only from 12.5±4.7 mm (range, 8 to 20 mm) to 10.6±4.7 mm (range, 4.5 to 17.7 mm, P=NS). No stents required redilation. One of 18 stents placed in series had “unlocked” and rotated in the venous lumen but remained fully patent to flow.
Conclusions Balloon-expandable intravascular stents can be safely and effectively used to relieve systemic venous and systemic venous baffle obstructions, even when obstruction is complete. Short-term follow-up suggests excellent continued patency, but further follow-up is required to observe for progression of neointimal hyperplasia. We postulate that balloon-expandable intravascular stents will become the treatment of choice for the relief of selected systemic venous and venous baffle obstructions in the pediatric population.
Balloon-expandable intravascular stents have gained increasing acceptance in the treatment of vascular obstructions, particularly in pulmonary1 2 and coronary arterial circulations.3 Recently there has also been considerable literature regarding the use of balloon-expandable intravascular stents as a component of the “transjugular intrahepatic portosystemic stent-shunt” procedure4 and various systemic venous obstructions in adults.5 This is the only FDA-approved study for stents in congenital heart disease, and relatively little has been written on use of balloon-expandable intravascular stents for the relief of systemic venous obstructions in the pediatric age group.1 2 There is only one case report of their use in a baffle obstruction after Mustard procedure in an adult.6
Surgical relief of systemic venous obstructions7 8 and post–Mustard procedure baffle obstructions6 requires general anesthesia, generally requires cardiopulmonary bypass, is technically difficult, and may be unrewarding. Likewise, isolated balloon dilation of systemic venous obstructions7 and systemic venous baffle obstructions6 may result in only transient relief of obstruction. All patients in this group were either symptomatic and/or large and multiple venous access was considered critical for the immediate or future intracardiac procedures (dilations, further stents, stent redilation, or pacemaker access). With these facts or considerations in mind, since September 1991 we have been implanting stents in baffle and systemic venous obstructions. Stents were used as the primary approach in these lesions because of (1) the authors’ favorable animal experience with venous stents,9 (2) the dismal long-term results from dilation alone in venous stenosis in the literature7 and in the authors’ own experience, and (3) predictable 100% restenosis in totally occluded or newly created venous channels opened only by dilation. This study was undertaken to assess the safety and efficacy of this procedure.
Twelve patients with 13 clinically symptomatic obstructions or >50% obstruction of large systemic veins or atrial baffles at angiography were considered for stent implantation. Clinical assessment, chest radiography, and echocardiography (occasionally at a referring institution) were performed in all patients and magnetic resonance imaging in 1 patient before cardiac catheterization. Parental (and where applicable, patient) informed consent was obtained according to institutional review board protocol.
All patients were premedicated initially with meperidine and promethazine. When additional sedation was required, various combinations of midazolam, ketamine, and meperidine were administered. Routine hemodynamic and saturation data were collected during right-heart catheterization from the femoral vein and the internal jugular or brachial veins as required. Prograde left-heart catheterization via a patent foramen ovale or transseptal puncture, with or without retrograde left-heart catheterization, was performed where indicated.
Intravascular Stent Implantation
The diameters of the obstructed segment and the vessel proximal and distal to the obstruction, including those in patients with complete obstruction, were measured after angiography with a cardiomarker catheter (USCI). The obstruction was then crossed with an end-hole catheter.
Before stent implantation, balloon dilation (predilation) was performed in patients with extremely tight obstructions to permit subsequent passage of the long sheath. In these patients, predilation was performed with a 4- or 6-mm balloon. The stent implantation procedure was as described previously by O’Laughlin et al.1 2 Briefly, activated clotting time was measured after the obstruction was crossed and predilated where required. Heparin (50 to 100 U/kg) was administered, and additional heparin was administered to maintain activated clotting time of 250 to 300 seconds. A 0.038- or 0.035-in Superstiff exchange wire (Meditech Inc) was passed through the end-hole catheter across the obstruction. A long 11F sheath was then advanced distal to the obstruction. For stent delivery and implant, a balloon catheter was used with an 8F shaft and the balloon diameter equal to the vessel either proximal or distal to the obstruction (whichever was smaller). After the balloon on the catheter shaft had been flushed and rewrapped, the Palmaz P308M stent (Johnson and Johnson Interventional Systems) was secured firmly to the middle of the balloon by hand crimping. The hand crimping and use of balloon catheters with an 8F (2.3-mm) shaft prevented the extensive distortion and “overcompression” of the stent on the balloon that occur when smaller catheter shafts are used. The balloon/stent combination was advanced through the long sheath and centered on the stenosis. This position was verified by angiography through a second venous catheter. The long sheath was withdrawn so that it was completely clear of the balloon, and further angiography was performed to confirm precise stent placement before balloon inflation. The balloon was inflated to its recommended inflation pressure, expanding the stent. If there was a residual “waist” in the stent, the stent was redilated to 17 atm with a high-pressure Blue Max balloon (Meditech Inc) with a diameter similar to that of the initial balloon. Where obstructions were >2 cm in length, stents were placed in series, each overlapping by at least several millimeters until the obstruction was traversed completely by stents. An end-hole catheter was passed over the wire distal to the stent and the wire, and then the catheter was removed.
The first of four intravenous doses of Cefazolin (12.5 mg/kg) was administered, and postimplantation angiography and hemodynamic assessment were performed. Poststent chest radiographs, echocardiograms, and magnetic resonance imaging (where indicated) were performed the day after catheterization. Although no thrombosis was encountered in the animal model, in which no anticoagulation was used, because of the extensive experience and adamant recommendations for anticoagulation for stents used in the adult, we elected to use the antiplatelet regimen popular when the study started (aspirin and dipyridamole; 1991). Since there have still been no thromboses or complications on this regimen, we have continued with the protocol. Aspirin (80 mg/d) and dipyridamole (1 to 2 mg · kg−1 · d−1 in two or three divided doses) was begun before discharge and continued for 6 months.
One of the investigators (C.E.M.) devised a technique for crossing total obstructions. The obstructed vessel was catheterized, and angiography was performed from both sides of the obstruction. A straightened transseptal sheath/dilator set was passed into the vein to the proximal end of the obstruction. With the tip of the dilator firmly against the obstruction, a straight “glide wire” (Terumo Wires, Meditech) was used to probe the occluded vessel, and the course of the wire was closely observed both on posterior-anterior and lateral fluoroscopy. In three cases in which the wire would not traverse the obstruction, a transseptal needle passed from the femoral vein through the long 6F or 7F transseptal sheath was used to perforate the obstruction, with the biplane fluoroscopy, angiogram “road maps,” and a catheter placed on the opposite end of the obstruction as guides. While the transseptal needle was advanced, small amounts (0.25 to 0.50 mL) of contrast were injected intermittently through the needle to define the anatomy. After the obstruction had been perforated with either the wire or needle, the long dilator and sheath were advanced over the wire or needle through the obstruction. The long sheath/dilator was exchanged for an end-hole catheter. The remainder of the procedure was as described above.
Restudy of patients was undertaken 12 months after stent implantation as a requisite of our stent protocol. Catheterization as described above was performed; in 1 patient, recatheterization was performed at the referring institution. Redilation was planned if the stent had a residual “waist” or if neointimal hyperplasia reduced the lumen to <80% of the vessel diameter proximal or distal to the stent (whichever was smaller).
Paired pre– and post–stent-implantation hemodynamic and angiographic data were analyzed with Student’s t test. A value of P<.05 was considered significant. The obstructed segment diameter was defined as the diameter of the narrowest portion of the obstructed segment. In each case, the percentage obstruction [1 minus the obstructed segment diameter divided by the diameter of the nearest normal proximal or distal vessel (whichever was the smaller) times 100] was calculated before and after stent implantation and at follow-up.
Twelve patients have had 21 stents placed in 13 obstructions within the systemic venous circulation. Results from patient 6 have been reported previously.2 Patient ages ranged from 3 months to 20 years (mean, 8.2±6.6 years) and weights from 5.5 to 60.6 kg (mean, 17.2±4.8 kg) (Table 1⇓). Patients 3a and 3b were the same patient, who underwent iliac stent implantation to allow access from the femoral vein before having a stent implanted at a superior vena cava/systemic venous atrial obstruction after Mustard procedure (Fig 1⇓). There was an immediate gradient reduction from 10.8±8.1 mm Hg (range, 0 to 20 mm Hg) to 0.9±1.3 mm Hg (range, 0 to 4 mm Hg, P<.001). The obstructed segment diameter increased from 2±2.4 mm (range, 0 to 8.5 mm) to 11.4±3.7 mm (range, 7.6 to 20 mm, P<.001), and the percentage obstruction decreased from 80±18% (range, 54% to 100%) to 2.3±8.3% (range, 0% to 30%, P<.001). Multiple stents were placed in 5 of 13 obstructions (Table 2⇓). Twelve of 13 obstructions were post prior surgical or catheterization procedures, of which 4 of 13 were systemic venous baffle obstructions after Mustard or Senning procedures (baffle group, patients 1 through 4 inclusive, Table 1⇓). The remaining 9 of 13 obstructions were in large systemic veins (central vein group, patients 5 through 11 inclusive, Table 1⇓).
In 1 of 4 patients, there was complete obstruction before stent implantation. Intravascular stent implantation resulted in an immediate gradient reduction from 12±8.4 mm Hg (range, 4 to 20 mm Hg) to 1.3±1.9 mm Hg (range, 0 to 4 mm Hg, P=.05) (Fig 2⇓). Similarly, there was a significant reduction in the angiographic percentage obstruction after stent implantation from a mean of 77±24% (range, 54% to 100%) to 0% in all patients (P<.001). The obstructed segment diameter increased from 3.5±3.9 mm (range, 0 to 8.5 mm) to 16±2.7 mm (range, 14 to 20 mm, P=.002) with stent implantation (Fig 3⇓). The range of balloon sizes used for stent dilation was 15 to 20 mm (mean, 18±2.9 mm). Residual waists at maximal inflation were seen in patients 1 and 2.
Symptomatology. Two of 4 patients presented with the superior vena cava syndrome, which was completely relieved by stent implantation. The remaining 2 patients in the baffle group were asymptomatic, with the obstruction discovered at attempted pacemaker lead implantation. Pacemaker lead implantation was subsequently achieved through the stent.
Central Vein Group
In 3 of 9 patients, there was complete obstruction before stent implantation. Intravascular stent implantation resulted in an immediate gradient reduction from 10.3±8.5 mm Hg (range, 0 to 20 mm Hg) to 0.8±1.1 mm Hg (range, 0 to 3 mm Hg, P=.005) (Fig 2⇑). There was also a significant reduction in the angiographic percentage obstruction after stent implantation from a mean of 81±18% (range, 57% to 100%) to 3.3±10% (range, 0% to 30%, P<.001). The obstructed segment diameter increased from 1.3±1.1 mm (range, 0 to 2.8 mm) to 9.4±1.7 mm (range, 7.6 to 12 mm, P<.001) with stent implantation (Fig 3⇑). The range of balloon sizes used for stent dilation was 8 to 12 mm (mean, 11±2 mm), with maximal stent expansion of 7.6 to 15 mm (mean, 9.4±1.7 mm). Residual “waists” in the stents at maximal expansion were seen in patients 5, 6, 9, 10, and 12.
Symptomatology. Six of 9 patients were symptomatic from their obstruction. Four of these 6 patients (patients 5 through 8) presented with symptomatology related to the superior vena cava syndrome. In patients 5 and 6, both with macrocephaly related to superior vena caval obstruction, there has been no progression of hydrocephalus, and the rate of head growth has been normal after stent implantation. Fig 4A⇓, showing patient 7, demonstrates complete superior vena caval obstruction and Fig 4B⇓ the same patient after stent implantation. Patient 8, with an obstructed superior vena cava/pulmonary artery anastomosis after a classic Glenn shunt, presented with progressive cyanosis and symptoms of the superior vena cava syndrome (Fig 5⇓). Symptoms related to the superior vena cava syndrome resolved completely after stent implantation, and the cyanosis, related to retrograde azygos vein flow, was markedly improved.
Patient 10 had heterotaxy and associated totally anomalous pulmonary venous return. The superior vena cava was congenitally narrowed in the area in which the four pulmonary veins entered. Hemodynamically, this created obstructed pulmonary venous return, manifesting as cyanosis and tachypnea. In this patient, there was considerable clinical improvement after stent implantation at the site of the superior vena caval narrowing.
Patient 12 presented with progressive cyanosis related to pulmonary arteriovenous fistulae of the right lung after a right classic Glenn shunt and baffling of the inferior vena cava to the left pulmonary artery (Fig 6⇓). In this patient, the right pulmonary artery pressure was 2 mm Hg higher than the right atrium. This gradient was abolished after stent implantation between the right atrium and the right pulmonary artery. After stent implantation, there was angiographic evidence of a portion of superior vena caval return being directed away from the right lung and the pulmonary arteriovenous fistulae, through the stent, to the left lung. This resulted in decreased right-to-left shunting through the pulmonary arteriovenous fistulae and a consequent increase in systemic saturation. In this patient, two stents also were implanted in a left pulmonary artery obstruction without complication.
All the infracardiac obstructions (patients 3b, 11, and 13) were asymptomatic, being recognized at diagnostic or interventional cardiac catheterization. In these patients, in whom more peripheral vascular access had been obtained, stents were implanted to maintain large venous access for anticipated future cardiac catheterization interventions (future redilations or stent implants).
No complications were incurred in either group at stent implant or follow-up. Specifically, there was no vascular rupture or hemorrhage. In the baffle group, there was no pulmonary venous obstruction secondary to stent implantation.
Six of 13 stented systemic venous obstructions have been restudied at cardiac catheterization 2 to 13 months after stent implantation (mean, 8.8±4.3 months). Because of this relatively small number of patients, the patients in the baffle group (3 of 4) and the patients in the central vein group (3 of 9) will be considered together (Table 2⇑).
All stents remained patent, with no evidence of stent fracture or compression, and none required redilation. Neointimal hyperplasia was noted in 5 of 6 stents (Fig 7⇓). Where present, the neointimal hyperplasia only minimally reduced the effective lumen size, from 13±4.7 mm (range, 8 to 20 mm) to 11±4.7 mm (range, 4.5 to 18 mm, P=NS). There was no significant difference between the immediate and follow-up poststent gradients (1±1.6 versus 0.7±1.2 mm Hg, P=NS). Patient 2 was electively restudied 2 months after stent implantation at the time of pacemaker lead implantation. Patient 10 was recatheterized 3 months after stent implantation after recurrence of cyanosis and tachypnea. Despite neointimal hyperplasia, there was no gradient across the stent. All four pulmonary veins, however, were narrowed at their entrance to the stent. Since the predominant problem was pulmonary vein stenosis, with no gradient across the stent, redilation of the stent was not undertaken. The patient died at subsequent attempted surgical relief of the pulmonary vein obstruction. Four of 6 patients have no gradient across the stent. Patient 4 had a 4 mm Hg residual gradient immediately after stent implantation and a 3 mm Hg gradient at follow-up. Follow-up in this patient revealed that the most superior of the four superior vena caval stents had “unlocked” from the adjacent stent and rotated 90° but remained unobstructed.
None of the 3 patients in the baffle group who have been restudied had evidence of secondary pulmonary venous obstruction. Patient 4 was found to have a 4 mm Hg mean gradient between the proximal and distal portions of the pulmonary venous atrium that was anatomically unrelated to the stent.
Intravascular stents can be used safely and effectively in the management of superior vena cava–to–systemic venous atrial baffle obstructions after either Senning or Mustard procedures, even when obstruction is complete. Surgical relief of such obstructions generally requires cardiopulmonary bypass and often fails to satisfactorily relieve the obstruction.6 Balloon dilatation may also provide only transient relief of such obstructions6 because of the elastic recoil of the obstructing tissue or restenosis. We incurred no complications in dilating the baffle obstructions to the size of the adjacent normal superior vena cava. Unfortunately, our series contains no inferior vena caval baffle obstructions, but we believe stent implantation in these lesions would be equally effective.
In children, intravascular stents can be used effectively in the acute relief of systemic venous obstruction, as suggested by O’Laughlin et al1 2 and supported by the literature on adults.5 10 Surgical relief of central systemic venous obstructions requires general anesthesia, is technically difficult, and, as in baffle obstructions, often fails to satisfactorily relieve the obstruction.7 8 One recent report7 described superior results with surgical relief of systemic venous obstructions compared with isolated balloon dilation and intravascular stent implantation. This report, however, failed to differentiate between those patients undergoing isolated balloon dilatation and those having implantation of intravascular stents, and there was a significant incidence of intermediate-term restenosis or occlusion (3 of 8 patients). Our study suggests results superior to these, although clearly the patient populations are different.
In our study, perforation of obstructions that were complete was uncomplicated, although this is clearly not without risk. Since all of these complete obstructions were acquired, we believed these obstructions would be surrounded by fibrous tissue to contain any extravasation of blood. The risks of this technique, however, relate not only to hemorrhage but also to inadvertent perforation of adjacent structures. Extreme care was taken to minimize these risks, with frequent small injections of contrast through the perforating needle to define the anatomy in both lateral and anteroposterior projections. This may prove to be an effective technique in selected acquired venous obstructions.
It has been postulated that pulmonary arteriovenous fistulae among patients with classic Glenn shunts may relate to the absence of normal hepatic flow to the anastomosed pulmonary artery.11 We postulated that the pulmonary arteriovenous fistulae might regress if the right lung received a portion of hepatic return, in the same way that pulmonary arteriovenous fistulae related to liver disease regress after liver transplantation.12 In our patient, since there were multiple microfistulous connections in the right lung, embolization of the fistulae was not feasible, and “reanastomosis” was considered the best option. This patient, despite the absence of a residual gradient across the stent, was also the only patient in our series with any angiographic obstruction after stent implantation. Since high-pressure balloons >12 mm in diameter are not yet available, this obstruction was dilated only to 3.5 atm. We anticipate that larger high-pressure balloons will be available by the time restudy is undertaken and plan redilation at that time. We await follow-up with regard to resolution of the arteriovenous fistulae.
Short-term follow-up of patients with superior vena cava–to–systemic venous atrial baffle obstructions and systemic venous obstructions has confirmed persistent stent patency. Neointimal proliferation recognized in 5 patients was not significant hemodynamically and has not required redilation. Further follow-up is required, however, to confirm that such neointimal proliferation remains nonobstructive. Recent reports suggest successful dilation of neointima within venous stents in animals13 and within pulmonary arteries in humans.14 On this basis, we anticipate successful dilation of neointimal proliferation if required.
We have found no evidence of stent compression or fracture, unlike similar procedures in adults15 and animals.16 The one stent that was found on follow-up to have rotated was most likely inadequately overlapped with its adjacent stent. Currently, we recommend at least 10 mm of overlap when stents placed in series are to be dilated >12 mm. This would ensure secure interlocking when the stents shorten with expansion. We would now even recommend placement of a third stent between two stents that have a very short (<3 mm) or no overlap after full expansion.
We conclude that intravascular stent implantation is safe and effective in the acute relief of obstructions of both systemic venous baffles and large systemic veins. Short-term follow-up suggests excellent maintenance of patency with only minimal neointimal hyperplasia. Continued observation will be required, however, to ensure that neointimal hyperplasia does not become obstructive. We postulate that balloon-expandable stents will become the treatment of choice for the relief of obstructions of large systemic veins and systemic venous baffles in the pediatric population.
Reprint requests to Charles E. Mullins, MD, Texas Children’s Hospital, Department of Pediatric Cardiology, 6621 Fannin, Houston, TX 77030.
Guest editor for this article was Joseph K. Perloff, MD, UCLA School of Medicine, Los Angeles, Calif.
- Received August 1, 1994.
- Revision received December 13, 1994.
- Accepted December 27, 1994.
- Copyright © 1995 by American Heart Association
O’Laughlin MP, Perry SB, Lock JE, Mullins CE. Use of endovascular stents in congenital heart disease. Circulation. 1991;83:1923-1939.
O’Laughlin MP, Slack MC, Grifka RG, Perry SB, Lock JE, Mullins CE. Implantation and intermediate-term follow-up of stents in congenital heart disease. Circulation. 1993;88:605-614.
Chatelain P, Meier B, Friedli B. Stenting of superior vena cava and inferior vena cava for symptomatic narrowing after repeated atrial surgery for D-transposition of the great arteries. Br Heart J. 1991;66:466-468.
Mullins CE, O’Laughlin MP, Vick GW, Mayer DC, Myers TJ, Kearney DL, Schatz RA, Palmaz JC. Implantation of balloon expandable intravascular grafts by catheterization in pulmonary arteries and systemic veins. Circulation. 1988;77:188-199.
Srivastava D, Preminger TJ, Spevak PJ, Lock JE. Development of pulmonary arteriovenous malformations following cavopulmonary anastomoses. Circulation. 1993;88(suppl I):I-149. Abstract.
Ing FF, Grifka RG, Nihill MR, Mullins CE. Redilation of intravascular stents in congenital heart defects. Circulation. 1993;88(suppl I):I-389. Abstract.