Surgical Considerations A. Robert Cordell, M.D., Robert C. McKone, M.D., and M. Amjad Bhatti, M.D. ABSTRACT Thirty-five infants underwent pulmonary artery banding for cardiac defects producing excessive pulmonary blood flow with resultant refractory cardiac failure. Five died during or shortly after operation. Seven of the survivors subsequently underwent debanding and correction of defects, with 3 deaths. Problems with debanding because of extensive fibrosis at the bifurcation of the pulmonary artery and beyond caused 1 death. Of the other 2, 1 patient died from bronchopneumonia after atrial partitioning and ventricular septal defect closure, the other following correction of double-outlet right ventricle. Details of this experience are reviewed. Pulmonary artery debanding may require total band removal, pericardial patching, resection with anastomosis of the pulmonary artery, occasionally infundibulectomy, and even replacement of part of the pulmonary artery with a homograft or prosthesis. Residual gradients must be accepted at times. Transatrial closure of septal defects is desirable when technically feasible. Pulmonary debanding and defect correction may be formidable procedures requiring a variety of operative techniques, depending upon alterations in pathological anatomy. Banding therefore should be reserved for the infant under 10 pounds in weight when intensive medical management is unsuccessful. S ince the introduction of pulmonary artery constriction by Muller and Dammann [13] in 1951 as a means of reducing excessive pulmonary blood flow in desperately ill infants with ventricular septal defect, much has been learned regarding its application in congenital heart disease [3, 11, 15, 18, 25, 261. Accumulated experience to date certainly supports the concept that adequate constriction of the pulmonary artery may lead to significant improvement in patients with congestive heart failure, enhance their growth, and halt or decrease changes in pulmonary vascular resistance [2, 7, 12, 20-221. The procedure has been of value in isolated high-flow ventricular septal defect as well as in a variety of complex congenital anomalies resulting in excessive pulmonary blood flow. Banding of the pulmonary artery can be carried out with a mortality of approximately 15%, lower in isolated ventricular septal defect and higher in complex anomalies associated with ventricular septal defect [4, 6, 7, 201. Pulmonary artery banding in 35 infants was carried out at the Bowman From the Departments of Surgery and Pediatrics, Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, N.C. Supported by U.S. Public Health Service Grant No. HE 0-5257. Presented at the Eighteenth Annual Meeting of the Southern Thoracic Surgical Association, Tampa, Fla., Nov. 4-6, 1971. Address reprint requests to Dr. Cordell, Department of Surgery, Bowman Gray School of Medicine, Winston-Salem, N.C. 27103. 24 THE ANNALS OF THORACIC SURGERY
PATIENT EXPERIENCE WITH PULMONARY ARTERY DEBANDING Systolic Gradient Patient at Operation No. Age -(mm. Hg) Outcome VSD alone 1 7 yr. 4 mo. 77 65 Satisfactory 2 5 yr. 8 mo. 100-35 Satisfactory 3 7 yr. 9mo. 132-5 (3 Died 4 6 yr. 60-20 Satisfactory 5 6 yr. 100-50 Satisfactory - Other anomalies 6 1 yr. 7 mo. 64 0 VSD) - 7 5 yr. 4 mo. 74 50 outlet RV & VSD) (TGA & Died (double- Died VSD = ventricular septa1 defect; TGA = transposition of the great arteries; RV =right ventricle. Gray School of Medicine with a mortality of approximately 15% (5 of 35). As would be expected, most of the infants who died after banding had associated complex cardiac anomalies. Pulmonary artery banding as a definitive palliative procedure is employed in certain cardiac anomalies which are not otherwise reparable. Results in this group have been surprisingly good, with significant early improvement. Long-term evaluation is not yet possible [lo, 19, 20, 241. In patients with reparable lesions, however, consideration must be given to eventual removal of the band and definitive correction of the defect. In this group of patients, decisions must be made regarding the timing of operation, approach, details of debanding, and definitive correction of defects. Although experience reported to date has been meager, this information continues to interest the cardiothoracic surgeon to whom increasing numbers of such patients are presented for consideration for correction. Our experience thus far with pulmonary artery debanding and correction of defects is limited to 7 patients and is summarized in the table. Surprising variation in anatomy and problems imposed by prior banding prompt this report. Three of the patients died in the hospital, only 1 (Patient 3) as a direct result of debanding. Patients 6 and 7 had complex cardiac defects requiring longer perfusion. Admittedly, pulmonary artery debanding lengthened the perfusion time and probably was contributory in that respect to the deaths. Technical problems imposed by the band in Patient 3 will be discussed later. Comment Infants surviving the banding procedure generally do well for some years, with less cardiomegaly and congestive failure and increased growth and development. Eventually, however, increasing cardiac output in the VOL. 14, NO. I, JULY, 1972 25
CORDELL, MCKONE, AND BHATTI presence of a small, fixed pulmonary artery lumen produces more right-toleft shunting and increasing peripheral desaturation with its attendant problems. Debanding and defect closure ideally should be done when body weight is in the 15 kg. range but prior to the onset of marked desaturation and polycythemia. Such children represent relatively safe risks for perfusion and respiratory care without all the problems of intense cyanosis. Furthermore, the growth rate declines at this time, and further delay adds nothing to patient safety. Materials used for banding initially consisted of cotton umbilical tape, strips of Dacron graft material, or Teflon tape. Undue scarring and tissue incorporation from cotton and Dacron were described [l, 14, 171. Teflon is said to be relatively easily removed at the time of debanding [16]. Only cotton and Dacron were used in our patients. Marked local scarring with tissue incorporation was present, making removal of the band difficult in each instance. In addition, circumferential changes in the wall of the pulmonary artery produced undue thinning and friability, thus increasing technical difficulties. For these reasons, we are experimenting with Silastic bands and believe that the minimal tissue response and lack of incorporation in body tissue make such bands preferable. Removal at reoperation should be much easier. Results of long-term evaluation are not yet available. Migration of bands has been reported [23] and was demonstrated in Patient 3, the only patient in our series in whom death occurred as a direct result of technical problems associated with debanding. In this patient, the cotton band had migrated peripherally and produced severe stenosis in the area of the pulmonary artery bifurcation, with circumferential thinning and friability of the artery extending to each pulmonary trunk. Even though a large patch of anterior pericardium was used to enlarge the artery, adequate relief of stenosis was never attained and a large residual pressure gradient persisted. Bleeding from suture lines compounded the problem. Postmortem examination of the pulmonary arteries demonstrated that a bifurcated pulmonary artery homograft might have been inserted with success. Consequently, such grafts are being collected and stored for future use in such patients. Methods used for relieving the pulmonic stenosis produced by the band vary widely in published reports [5, 6, 91. Simple band removal and dilation of the pulmonary artery have been successful in some instances and might be done routinely when fairly nonreactive material has been used. Our experience thus far has dictated a variety of procedures, including resection with end-to-end anastomosis, insertion of a diamond-shaped pericardial patch, and anterior resection with transverse closure to restore an adequate lumen. We have not been able to simply remove the band and dilate the pulmonary artery, but we have measured the size of the artery with Hegar dilators following debanding and have insisted upon 12 mm. as the minimum satisfactory diameter. 26 THE ANNALS OF THORACIC SURGERY
Presently we believe that bands should be removed in their entirety. This is accomplished safely after cannulation for bypass is completed and everything is in readiness for perfusion. If the thinned and friable wall of the pulmonary artery is entered during removal of the band, bypass is begun and debanding continued. If the area of constriction is short and localized, it is probably wiser to resect that segment and do an end-to-end anastomosis. Use of an anterior roof of pericardium for patients in whom marked scarring and attachment to surrounding structures are present is simple and safe, particularly if performed over a Hegar dilator at least 12 mm. in diameter. We have been impressed with a steady progression in some patients toward typical tetralogy of Fallot to the point of marked infundibular muscular hypertrophy and beginning peripheral desaturation from right-to-left shunting. In such patients it was deemed necessary to perform infundibulectomy at the time of pulmonary debanding in order to reduce gradients. Experience with pulmonary valvular stenosis suggests that the associated infundibular hypertrophy may resolve with time, but long-term evidence of this is not presently available. Reduction of pulmonary outflow tract gradients has proved difficult and the results variable, as seen in the table. Despite all efforts, significant gradients have remained in some patients even though they are doing well clinically. Certainly, the abolition of significant gradients is desirable, but apparently this is not always immediately attainable. Repeat catheterization studies during prolonged follow-up observation are necessary to establish the sequence of these hemodynamic alterations. Reported changes in the pulmonary valve have been presumed due to trauma from the leaflets striking the constricted area [l, 5, 6, 91. The pulmonary valve cusps have been thickened and abnormal in every patient we have observed. We have not been able to demonstrate a gradient of more than 10 mm. across the valve, however, and in no instance was valvulotomy necessary. Septa1 defect closure usually can be accomplished by the transatrial approach, which possesses the obvious advantage of making ventriculotomy unnecessary [8]. Most ventricular septal defects are well visualized by means of an atriotomy parallel to the atrioventricular groove with retraction of the septa1 leaflet of the tricuspid valve. Seldom has it become necessary in our experience to divide the leaflet. The transatrial approach was used in Patients 4 and 5. Adequate infundibulectomy cannot be carried out through this approach, however, and at times it is technically difficult to assess the size of the reconstructed pulmonary artery without a ventriculotomy because of problems of exposure. Furthermore, in patients with a double-outlet right ventricle or very large ventricular septal defect (such as Patient 7), proper closure of the defect is difficult or impossible through a transatrial approach alone. Nevertheless, this approach adds little in time or operative trauma and should be used first to assess the septal defect. If exposure is unsatisfac- VOL. 14, NO. 1, JULY, 1972 27
CORDELL, MCKONE, AND BHATTI tory, ventriculotomy is then performed in a plane parallel to the muscle fibers, with care being taken to avoid coronary vessels. On occasion, both approaches are helpful for better suture placement in different areas of patch attachment. It would appear, therefore, that the approach for definitive correction of a septal defect will vary according to the anatomy, and the surgeon must be prepared to adapt his approach to the existing situation. There is no substitute for an adequate preoperative catheterization study with angiocardiography to allow surgical evaluation of the anatomy. Such a study is particularly important if a long period has elapsed since banding. In addition, the desperately ill situation of many of these infants prior to banding makes accurate and detailed study difficult. A complete study prior to correction therefore becomes even more important. Measurement of pressure gradients at operation both before and after definitive repair is desirable for its immediate prognostic value as well as for correlation with follow-up catheterization studies. As mentioned, it is difficult to know from clinical evaluation alone how much residual stenosis remains, and detailed catheterization studies are necessary. References 1. Berry, C. L. Changes in the wall of the pulmonary artery after banding. J. Pathol. 99:29, 1969. 2. Boruchow, I., Waldhausen, J. A., Miller, W. W., Rashkind, W. J., and Friedman, S. Pulmonary artery hypertension in infants with congenital heart disease. Arch. Szirg. 99:716, 1969. 3. Ching, E., DuShane, J. W., McGoon, D. C., and Danielson, G. K. Total correction of ventricular septal defect in infancy using extracorporeal circulation. Ann. Thorac. Surg. 12:1, 1971. 4. Dobell, A. R. C., and Busse, E. F. G. Banding of the pulmonary artery. Dis. Chest 54:54, 1968. 5. Dobell, A. R. C., Murphy, D. R., and Gibbons, J. E. Pulmonary artery debanding. Ann. Thorac. Surg. 5:435, 1968. 6. Hallman, G. L., Cooley, D. A., and Bloodwell, R. D. Two-stage surgical treatment of ventricular septal defect: Results of pulmonary artery banding in infants and subsequent open-heart repair. J. Thorac. Cardiovasc. Surg. 52:476, 1966. 7. Horsley, B. L., Zuberbuhler, J. R., and Bahnson, H. T. Factors influencing survival after banding of the pulmonary artery. Arch. Surg. 101:776, 1970. 8. Hudspeth, A. S., Cordell, A. R., Meredith, J. H., and Johnston, F. R. An improved transatrial approach to the closure of ventricular septal defects. J. Thorac. Cardiovasc. Surg. 42: 157, 1962. 9. Hunt, C. E., Formanek, G., Castaneda, A., and Moller, J. H. Closure of ventricular septal defect and removal of pulmonary arterial band. Am. J. Cardiol. 26:345, 1970. 10. Idriss, F. S., Riker, W. L., and Paul, M. H. Banding of the pulmonary artery: A palliative surgical procedure. J. Pediatr. Surg. 3:465, 1968. 11. Kahn, D. R. Discussion of Ching and associates [3]. 12. Morrow, A. G., and Braunwald, N. S. The operative management of ventricular septal defect in infancy: Results of corrective operations performed after palliative pulmonary artery constriction (Abstract). Am. J. Cardiol. 21 : 112, 1968. 28 THE ANNALS OF THORACIC SURGERY
13. Muller, W. H., Jr., and Dammann, J. F., Jr. The treatment of certain congenital malformations of the heart by the creation of pulmonic stenosis to reduce pulmonary hypertension and excessive blood flow. Surg. Gynecol. Obstet. 95:213, 1952. 14. Osborn, J. R., Hall, R. J., Winn, D. F., Jr., Capper, R. S., and Blake, H. A. An unusual late complication of pulmonary artery banding. Circulation 34: 61, 1966. 15. Rashkind, W. J., Waldhausen, J. A., Miller, W. W., Dodd, P., and Friedman, S. Test band of the pulmonary artery with a balloon-tipped catheter. Circulation 39, 40 (Suppl. 1):I-201, 1969. 16. Reid, J. M., Barclay, R. S., Coleman, E. N., Stevenson, J. G., Welch, T. M., and McSwan, N. Pulmonary artery banding in congenital heart disease associated with pulmonary hypertension. Thorax 23:385, 1968. 17. Rohmer, J., Brom, A. G., and Nauta, J. Bands inside the pulmonary artery: A complication of the Dammann-Muller procedure. Ann. Thorac. Surg. 3: 449, 1967. 18. Sigmann, J. M., Stern, A. M., and Sloan, H. E. Early surgical correction of large ventricular septal defects. Pediatrics 39:4, 1967. 19. Somerville, J., Agnew, T., Stark, J., Waterston, D. J., Aberdeen, E., Bonham- Carter, R. E., and Waich, S. Banding of the pulmonary artery for common atrioventricular canal. Br. Heart J. 29:816, 1967. 20. Stark, J., Aberdeen, E., Waterston, D. J., Bonham-Carter, R. E., and Tynan, M. Pulmonary artery constriction (banding): A report of 146 cases. Surgery 65:808, 1969. 21. Takahashi, M., Lurie, P. R., Petry, E. L., and King, H. Clinical and hemodynamic effects of pulmonary artery banding. Am. J. Cardiol. 21:174, 1968. 22. 23. 24. 25. 26. Talner, N. S. Pulmonary arterial banding. Am. J. Cardiol. 26:441, 1970. Verel, D., Taylor, D. G., and Emery, J. L. Failure of pulmonary artery banding due to migration of the band. Thorax 25:126, 1970. Waldhausen, J. A., Boruchow, I., Miller, W. W., and Rashkind, W. J. Transposition of the great arteries with ventricular septal defect: Palliation by atrial septostomy and pulmonary artery banding. Circulation 39, 40 (Sup@. 1):1-215, 1969. Williams, G. D. Pulmonary artery banding: An improved technique with a new instrument. Surgery 69:478, 1971. Yao, J. K. Y., and Mustard, W. T. An extrapleural approach to the heart with particular reference to pulmonary artery banding and innominate artery. J. Thorac. Cardiovasc. Surg. 10:273, 1969. NOTICE FROM THE EDITOR Beginning with this issue, a new section, How to Do It, is being started. We hope to publish details of operative techniques that will be of help to our readers. The editors welcome any contributions to this section. VOL. 14, NO. 1, JULY, 1972 29