Intraoperative Stents to Rehabilitate Severely Stenotic Pulmonary Vessels

Transcription

1 Intraoperative Stents to Rehabilitate Severely Stenotic Pulmonary Vessels Ross M. Ungerleider, MD, Troy A. Johnston, MD, Martin P. O Laughlin, MD, James J. Jaggers, MD, and Peter R. Gaskin, MD Division of Pediatric Cardiology, Children s Hospital, Seattle, Washington, Division of Pediatric Cardiology, Department of Pediatrics, and Division of Thoracic Surgery, Department of Surgery, Duke University Medical Center, Durham, North Carolina Background. Patch enlargement of severe branch pulmonary artery stenosis (PAS) or pulmonary vein ostial stenosis (PVS) can be technically challenging. Recurrences are common and exposure may require long periods of cardiopulmonary bypass (CPB). Methods. Since 1993, we performed 31 procedures on 27 patients with endovascular stents placed intraoperatively under direct surgical vision: 22 patients with tight PAS and 5 patients with PVS. Selection for intraoperative (vs catheterization laboratory) stent placement was prompted by: (1) the need for a concomitant cardiac surgical procedure (16 cases); (2) limited vascular access for catheterization laboratory stent placement (11 cases); or (3) rescue of patients with complications after attempted placement of stents (4 cases). Results. In this group of very complex and challenging patients there were 5 hospital deaths (hospital survival, 81%). Follow-up of survivors has ranged from 1 month to 7 years (mean years). There have been 3 late deaths (late series survival, 70%). No complication or death was related to stent placement. Surviving patients have had significant clinical improvement. Mean pulmonary gradient (postoperative vs preoperative echo) has fallen in all survivors and has decreased from a mean of 66 mm Hg preoperatively to 28 mm Hg postoperatively (p 0.01). All pulmonary arteries are appreciably enlarged and will be easier to deal with at a later date if necessary. One patient (DORV, HLHS ) with pulmonary vein stents has gone on to a successful Glenn procedure. The other two surviving patients with PV stents have occlusion of their proximal PVs on follow-up catheterization; thus only 1 of 5 patients with stents for PVS has had a successful outcome. Four patients have had repeat surgery. Stents have produced no impediment to subsequent surgical procedures, and the pulmonary arteries were easy to work with. Conclusions. Intraoperative stenting provides an attractive option for rehabilitation of pulmonary vessels. Direct vision insertion on CPB is extremely quick and immediately effective, limiting the CPB exposure required to treat this problem. Once stented, vessels remain open and are amenable to future surgical intervention as necessary. Outcome is better for patients with PAS versus those with PVS. (Ann Thorac Surg 2001;71:476 81) 2001 by The Society of Thoracic Surgeons Patch enlargement of severe branch pulmonary artery stenosis (PAS) or pulmonary vein ostial stenosis (PVS) can be technically challenging. Recurrent stenosis is common, often caused by compression of the patch by surrounding structures such as the aorta (in the case of the right pulmonary artery) or lung (in the case of the left pulmonary artery). Exposure adequate for suturing to extremely small vessels (1 to 2 mm) can be hampered by excessive pulmonary collateral flow. In these cases, periods of hypothermic low flow cardiopulmonary bypass (CPB) or even circulatory arrest may be necessary, thus prolonging and increasing the risks of CPB. Endovascular stents, which can be placed under fluoroscopic control in the catheterization laboratory, have become an accepted adjunctive therapy for the treatment of these problems in selected patients [1, 2]. Limited Presented at the Poster Session of the Thirty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 31 Feb 2, Address reprint requests to Dr Ungerleider, Division of Cardiothoracic Surgery, Oregon Health Sciences University, L-353, 3181 SW Sam Jackson Park Rd, Portland OR vascular access and small patient size can prevent the application of this technique to all suitable candidates. This report reviews our institutional experience with the placement of endovascular stents into the pulmonary vessels under direct vision by the surgeon. Material and Methods In 1993, we began treating selected patients with PAS using endovascular stents placed intraoperatively under direct surgical vision. More recently, we have experimented with the placement of stents directly into the pulmonary veins for patients with severe PVS. Our experience encompasses 27 patients who underwent 31 operative procedures that incorporated the use of an endovascular stent. All procedures were performed at Duke University Medical Center. The patients ranged in age from 7 days to 14 years (mean years) and ranged in weight from 2.2 to 41.7 kg (mean kg). The anatomic diagnoses of the patients included truncus arteriosus (4 patients), pulmonary atresia/ ventricular septal defect (7), tetralogy of Fallot (8), pul by The Society of Thoracic Surgeons /01/$20.00 Published by Elsevier Science Inc PII S (00)

2 Ann Thorac Surg UNGERLEIDER ET AL 2001;71: INTRAOPERATIVE STENTS 477 Fig 1. Concomitant procedure. (A) Infant with pulmonary atresia and ventricular septal defect (VSD) and small central pulmonary arteries (PAs). The first procedure was to place a 6-mm pulmonary homograft conduit to these PAs, leaving the VSD open. (B) An angiogram of the infant 8 months later. The left PA has grown but the right PA remains small. The infant underwent VSD closure, a new right ventricular pulmonary artery homograft (10 mm), and intraoperative stenting of the right PA. At 4 years of age (C) the patient returned for conduit change. The stent in the right PA is visible and was dilated before conduit replacement. Right ventricular/left ventricular pressure ratio after successful conduit replacement was 0.35). Rescue was the term used for retrieval of errant stents that were delivered in the catheterization laboratory but ended up in undesirable locations. These stents were removed during surgery and then placed under direct vision into the appropriate location. Concomitant procedures included conduit change (5 patients), Rastelli (3), VSD or ASD closure (3), Glenn (1) pulmonary valve replacement (3), removal of errant stent (4), and aortic arch reconstruction (2 patients). In all patients, a preoperative angiogram demonstrated the area of narrowing in the pulmonary arteries and documented that the region in need of reconstruction was centrally located and was capable of identification under direct vision by the surgeon. The stent and balloon sizes were based on measurements made during this catheterization and were chosen by the cardiologist. Areas of distal (parenchymal branch) pulmonary artery stenosis were not addressed using intraoperative techniques. When placed intraoperatively, stents were placed into the pulmonary vessels under direct vision by the surgeon while the patient was on CPB. Management of CPB was determined according to the concomitant procedure or, if the procedure consisted of stent placement alone, was often done under normothermia on an electrically fibrillating heart. Most often, a small incision was made in the main pulmonary artery (or conduit) and, using cardiotomy suckers, the branch pulmonary artery monary atresia/intact ventricular septum (1), transposition of the great arteries (1), complex single ventricle (including heterotaxy syndromes, and unbalanced atrioventricular canal) (6 patients). A total of 42 stents were placed in the following locations: right pulmonary artery (5 patients), left pulmonary artery (10), bilateral pulmonary arteries (7), and pulmonary veins (5 patients). Although 70% to 80% of endovascular stents placed for congenital heart lesions at Duke are implanted in the catheterization laboratory using fluoroscopic guidance, our indications for placing stents in the operating room included (1) the need to perform a concomitant procedure (16 cases) (Fig 1); (2) limited vascular access (11 cases) (Fig 2); or (3) rescue (4 cases) (Fig 3). In many cases patients receiving concomitant procedures could have undergone preoperative catheterization and stent placement, but it was believed that it would be simpler and more efficient to place the stent in the operating room while the additional procedure was being performed. Patients with limited vascular access, either because of their small size or the absence of patent vessels required for the advancement of stent delivery systems, were referred for surgical implantation of stents. More recently, the safety of the transhepatic approach to cardiac chambers has diminished lack of vascular access as a reason to require surgical placement of stents. Fig 2. Limited access. (A) Angiogram of an infant who underwent neonatal tetralogy of Fallot repair. At that time he was noted to have a very tiny left pulmonary artery (PA) with substantial ductal tissue. This angiogram, obtained at 4 months, demonstrates virtual occlusion of the left PA. There is an extremely small communication with the distal pulmonary artery apparent on the actual angiogram. (B) Subsequent pulmonary venous wedge angiogram revealing a reasonable distal left PA. A stent was placed through the right ventricular outflow tract with the heart electrically fibrillating and using 10 minutes of CPB. (C) Follow-up angiogram of the rehabilitated left PA.

3 478 UNGERLEIDER ET AL Ann Thorac Surg INTRAOPERATIVE STENTS 2001;71: Fig 3. Rescue. (A) Stent that had been placed (in the catheterization laboratory) in the right pulmonary artery (PA) of a 4-year-old patient with RPA stenosis. The stent became dislodged during catheter withdrawal and ended up in the right ventricular cavity. (B) Roentgenogram taken after surgery, during which the stent was removed from the right ventricle and placed under direct vision into the RPA through a small incision in the main pulmonary artery. to be stented was identified. Small dilators were placed into the pulmonary artery to enlarge the lumen so that a stent delivery catheter could be inserted. Furthermore, use of dilators helped to provide the surgeon with a sense of where the stent should be situated. The balloon catheters were placed without the use of a wire. The balloons were inflated to the manufacturer s recommended pressure for approximately 30 seconds. The balloon diameter ranged from 4 to 15 mm with a mean diameter of 8 mm. Fluoroscopy was not used during the procedure. Placement under direct vision makes positioning of the stent (with experience) quite simple, and it was possible to place stents in a manner such that they did not obstruct access to the contralateral pulmonary artery. In our procedures the surgeon placed the stent across the stenotic area and the cardiologist selected the stent and inflated the balloon. Stents that are placed and not believed to be in the optimal location are easily removed. Flaring of the proximal stent helps to ensure that it will not embolize distally and makes reentry into the stented vessel during any subsequent catheterization easier. In the limited experience with stents placed in the pulmonary veins, we learned that it is prudent to place a few stitches securing the stent to the atrial wall. This helps to prevent embolization of the stent (along the direction of blood flow) into the atrium and subsequently into the systemic circulation (as happened in 1 of our patients). Results Of the 27 patients, 22 (81%) survived to leave the hospital. The 5 hospital deaths included 2 patients with truncus arteriosus who had successful cardiac repair but who died late of immunologic complications from severe DiGeorge syndrome: 1 patient after thymic transplant who developed graft versus host disease, and 1 with complications while awaiting thymus transplant. Two patients with pulmonary atresia had severe distal pulmonary artery stenoses that produced significant pulmonary hypertension despite adequate dilation of the main branch pulmonary arteries. Both of these patients died late after surgery from massive hemoptysis just before hospital discharge (1 patient) or from persistent low cardiac output with right heart failure (1 patient). The other hospital death occurred in an infant who underwent repair of truncus arteriosus with interrupted aortic arch. This infant had a large apical muscular ventricular septal defect, long-segment left pulmonary artery stenosis, and significant truncal valve insufficiency. She was brought to the operating room 10 days after repair of her truncus interruption because of persistent heart failure and inability to wean from the ventilator. During anesthetic induction she had a cardiac arrest, and resuscitation required emergent placement on CPB. She underwent closure of her apical VSD, truncal valve replacement, and stent placement in the LPA; however, she could not successfully be weaned from CPB. There have been 3 late deaths after hospital discharge. One was in a patient with tetralogy of Fallot and severe distal pulmonary artery stenosis. This patient underwent repair in the first weeks of life because of persistent cyanosis and subsequent stenting of both the right and left pulmonary arteries because of high right ventricular pressures and small branch pulmonary arteries. This allowed her to be weaned from support and discharged

4 Ann Thorac Surg UNGERLEIDER ET AL 2001;71: INTRAOPERATIVE STENTS 479 from the hospital. Unfortunately, her first year of life was marked by recurrent respiratory problems requiring frequent hospitalizations and she expired at home from an acute respiratory arrest at 11 months of age. The other two late deaths occurred in two patients with pulmonary vein ostial stenosis who underwent pulmonary vein stenting in the hope that they would be able to progress to Fontan physiology. Both of these patients died (at 5 and 7 months, respectively) after hospital discharge from unknown causes. None of the deaths (early or late) seemed to be related to the stents. Follow-up of the surviving patients has been from 2 months to 7 years (mean years). All surviving patients exhibit clinical improvement over the symptoms that led them to treatment. Four patients have undergone subsequent surgery; the stents did not interfere with subsequent surgery, even when performed at the stented area. Echocardiographic follow-up is available for 10 patients with a median follow-up of 2.3 years. The gradient across the stenosis decreased in all patients. The mean of the peak instantaneous gradients fell from 66 to 28 mm Hg (p 0.01). Ten follow-up catheterizations have been performed for the patients with pulmonary artery stents. Four of these patients had balloon dilation of the stents performed at follow-up. One had mild-to-moderate neointimal proliferation with a peak-to-peak gradient of 40 mm Hg across the stent in the proximal left pulmonary artery; this gradient was reduced to 2 mm Hg with balloon dilatation. One patient with pulmonary atresia with ventricular septal defect and hypoplastic pulmonary arteries still had a significant gradient (27 mm Hg) across the left pulmonary artery stent. There was no neointimal proliferation in the 5-mm stent, and dilation with a 6-mm balloon did not change the gradient. The third patient had 30 mm Hg peak-to-peak gradients across bilateral proximal pulmonary artery stents at the time of follow-up catheterization. These gradients did not change with balloon dilation and were believed to be related to abnormalities of the pulmonary arteries unrelated to the stents. The fourth patient, who had a Rastelli operation for pulmonary atresia with VSD (Fig 1C), underwent dilation of a stent in the RPA before elective conduit change. After conduit change, the pressures in the RV were 35% systemic. The remaining 6 patients had no need for further intervention at the time of their cardiac catheterization. Three of the patients with pulmonary vein stents have undergone catheterization. In 2 patients the pulmonary veins have become occluded despite the presence of stents. One patient has gone on to a Fontan procedure. The remaining 2 patients with stented pulmonary veins died late at home without follow-up regarding the status of their pulmonary veins. Although there were no deaths related to stents, there were eight complications among the 31 procedures: most of these occurred early in our experience and relate to learning how to deploy stents under direct vision. In 2 patients dilation of the stent resulted in laceration of the pulmonary artery; in both cases, the stent was removed Fig 4. Infant with tetralogy of Fallot and with pulmonary artery (PA) stenoses that produced systemic right ventricular pressures and abnormal convalescence after tetralogy of Fallot repair. (A) Stent inflated in the left PA does not fully expand along its distal portion because of inability to distend the distal left PA. The distal left PA is diffusely small. (B) Stent in the right PA, which has been placed too distal to enlarge the area of right PA ostial stenosis. (C) Patient was returned to surgery and a second stent was placed in the right PA (stent on a stent) to address the proximal right PA stenosis. The severe distal right PA stenosis is apparent at the branching of the upper and lower lobe branches. and the pulmonary artery was repaired with a patch. Based on these experiences we recommend limited dissection around the pulmonary artery to be stented and we do not recommend stenting across fresh suture lines, as this was related to laceration in 1 patient. One patient had inadequate inflation of the stent despite high sustained pressures on the dilation balloon (Fig 4A). This stent was placed in the left pulmonary artery of an infant after repair of tetralogy. This infant had severely narrowed distal pulmonary arteries and also underwent stenting of the right pulmonary artery. Figure 4B shows that this stent was placed too distally and the patient was subsequently returned to the operating room for repeat stenting (Fig 4C). This experience of stent malposition taught us to make sure that some of the proximal stent is visible in the orifice of the main PA at the time of stent inflation, because the stent shortens as it expands. This patient survived to leave the hospital but died in her first year of life from severe respiratory problems. Malposition of the stent requiring reintervention occurred in 1 other patient. Two stents that were placed intraoperatively dislodged and required intervention for retrieval and repositioning. One of these was from the pulmonary vein and led us to recommend placing a few sutures on stents in this location to prevent it from embolizing along

5 480 UNGERLEIDER ET AL Ann Thorac Surg INTRAOPERATIVE STENTS 2001;71: the direction of blood flow. One patient had reperfusion pulmonary edema requiring several days of ventilator support. Comment Percutaneous transcatheter insertion of endovascular stents for pulmonary artery stenosis has become an attractive alternative to conventional surgical techniques. The experience reported with stents placed in this manner has been favorable [1 5]. However, because of the relatively large sheaths required for delivery of the stent, the technique is limited to larger patients. Additionally, some patients present problems of limited vascular access because of thrombosis of the ileofemoral venous system. The increasing popularity of the transhepatic route for percutaneous access should circumvent the difficulties of small patient size and limited vascular access in these patients. However, our experience reported in this series suggests that intraoperative stent placement is also a viable option for selected patients. Intraoperative stent placement has been reported previously. Mendelsohn and colleagues, from the University of Michigan [6], reported experience with 10 patients and no mortality. A subsequent report by Coles and colleagues, from Toronto Sick Children s Hospital [7], provided a more cautious overview. They experienced 27% mortality (3 of 11 cases), and follow-up of the surviving patients disclosed a high rate of intimal proliferation and obstruction within the stent (5 of 7 surviving patients had stent related stenosis/thrombosis). This has not been our experience, nor is it the experience that has been obtained by following patients with percutaneously inserted stents [2, 4, 5, 8]. At the time that this experience was begun, transhepatic access to the venous system was not routinely employed and we found the simplicity of intraoperative placement of stents to be attractive. Because of the reliability and relative noninvasiveness of transcatheter stent insertion, we would now recommend catheterization laboratory stent placement in all cases in which this is feasible. However, our experience with this group of patients has encouraged us to use direct operative placement in selected cases when catheterization laboratory stent placement is not preferred or possible. The need to perform a concomitant operative procedure is certainly an indication for operative stent placement. With experience we learned how to place stents in the proximal pulmonary arteries in perfect position so that they would not obstruct access to the contralateral pulmonary artery. It is doubtful that stents could be positioned more accurately in the catheterization laboratory, and the time required to place them under direct vision is minimal. There is enthusiasm by many surgeons to treat pulmonary artery narrowing with patch angioplasty. Stents, however, can enlarge extremely narrow pulmonary artery segments (Fig 2) and quickly rehabilitate the pulmonary artery. It can be difficult and time consuming to suture a patch to enlarge a 1-mm pulmonary artery. The presence of significant pulmonary collateral flow might encourage the surgeon to use periods of low flow or even total circulatory arrest, thereby complicating the CPB strategy and exposing the patient to the potential for a more protracted convalescence. Furthermore, patch angioplasties can scar or get compressed by adjacent structures, resulting in recurrent stenosis. This is often the case when the aorta compresses the right pulmonary artery (often the aorta in patients such as those with pulmonary atresia/ventricular septal defect is quite large) and has led some investigators to suggest graft extension of the aorta when right pulmonary arterioplasty is planned [9]. Likewise, the left pulmonary artery is often compressed by surrounding structures. Endovascular stents have the advantage of fighting for their territory. A concern about stent placement is that they fix the pulmonary artery at a small size and will not allow for growth. We have found that it is quite possible to redilate these stents as patients grow [10, 11]. Furthermore, the presence of a stent does not interfere with a suture line at the time of subsequent surgery [7]. We have operated on numerous patients with pulmonary artery stents and have found that it is quite easy to incise across them and to suture through their interstices. Our experience with pulmonary vein ostial stenosis is more dismal, and only 1 of our 5 patients received enough benefit to move on to a subsequent procedure. Two of these patients had progression of their pulmonary vein stenosis and, at subsequent catheterization, the veins proximal to the stent were occluded. The 2 remaining patients died at home without follow-up of their pulmonary vein stents. At the present time we do not recommend stenting of pulmonary vein ostial stenosis. The severity of disease in this series was mixed, and the majority of patients who had intraoperative stents placed because of small size or limited access were high-risk patients. They tended to be younger patients with less favorable anatomy for conventional surgical repair. Because stents can be placed quickly, reliably, and easily by the surgeon and can provide rapid pulmonary artery rehabilitation, they have a role in the options available to the surgeon for the treatment of these selected patients. In particular, we recommend them for small patients with extremely tight areas of pulmonary stenosis when vascular access is limited and the patient has significant clinical impairment from pulmonary artery narrowing. Furthermore, patients who require additional cardiac surgery can have stents placed during surgery with ease as an adjunct to their care. Long-term follow-up of stented patients will be necessary to know the implications of these recommendations regarding the eventual outcome for these patients. Because of some reports of stent intimal proliferation and thrombosis [7], we currently recommend treatment with antiplatelet therapy after stent insertion, regardless of whether the stent is inserted intraoperatively or in the catheterization laboratory. The technique of intraoperative stent placement is an example of the benefit of close cooperation between the pediatric cardiac surgeon and the pediatric

6 Ann Thorac Surg UNGERLEIDER ET AL 2001;71: INTRAOPERATIVE STENTS 481 interventional cardiologist that has helped lead to improved outcomes for some of our most complicated patients. References 1. Benson LN, Nykanen D, Freedom RM. Endovascular stents in congenital heart disease. Prog Cardiovasc Dis 1996;39: 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: Fogelman R, Nykanen D, Smallhorn JF, McCrindle BW, Freedom FM, Benson LN. Endovascular stents in the pulmonary circulation: clinical impact on management and medium-term follow-up. Circulation 1995;92: Hatai Y, Nykanen DG, Williams WG, Freedom RM, Benson LN. Endovascular stents in children under 1 year of age: acute impact and late results. Br Heart J 1995;74: Shaffer KM, Mullins CE, Grifka RG. Intravascular stents in congenital heart disease: short- and long-term results from a large single-center experience. J Am Coll Cardiol 1998;31: Mendelsohn AM, Bove EL, Lupinetti FM, Crowley DC, Lloyd TR, Fedderly RT, Beekman RH. Intraoperative, and percutaneous stenting of congenital pulmonary artery, and vein stenosis. Circulation 1993;88: Coles JG, Yemets I, Najm HK, et al. Experience with repair of congenital heart defects using adjunctive endovascular devices. J Thorac Cardiovasc Surg 1995;110: Ovaert C, Caldarone CA, McCrindle BW, et al. Endovascular stent implantation for the management of postoperative right ventricular outflow obstruction: clinical efficacy. J Thorac Cardiovasc Surg 1999;118: Curran RD, Mavroudis C, Backer CL. Ascending aortic extension for right pulmonary artery stenosis associated with ventricular to pulmonary conduit replacement. J Card Surg 1997;12: Ing FF, Grifka RG, Nihill MR, Mullins CE. Repeat dilation of intravascular stents in congenital heart defects. Circulation 1995;92: Morrow WR, Palmaz JC, Tio FO, Ehler WJ, vandellen AF, Mullins CE. Re-expansion of balloon-expandable stents after growth. J Am Coll Cardiol 1993;22: