- Email: [email protected]
Repeat sternotomy after reconstruction of the pericardial sac with glutaraldehyde-preserved equine pericardium
J
THORAC CARDIOVASC SURG
1987;93:616-9
Repeat sternotomy after reconstruction of the pericardial sac with glutaraldehyde-preserved equine pericardium The risk of repeat sternotomy is higher than that of the initial sternotomy, especially if the pericardial sac was left open at the first intervention. In 200 consecutive patients with a pericardial defect after open heart operations, the pericardium was closed with a glutaraldehyde-preserved equine pericardial patch. Precardiac adhesions at reoperation were assessed in four groups of patients on a scale of 6, ranging from o (no adhesions) to 5 (calcified or ossified adhesions). Group I comprised 13 patients in whom the pericardium was left open at the first operation and an equine pericardial patch was implanted at reoperation. Group Ia included the first five Group I patients who underwent reoperation less than 1 year (early reoperation) after the initial procedure. Group Ib included the other eight patients of Group I, who underwent reoperation more than 1 year (late reoperation) after the first procedure. Group II comprised nine patients who were reoperated on after reconstruction of the pericardial sac with a glutaraldehydepreserved equine pericardial patch. After a mean follow-up of 20.2 months, the incidence of patch-related complications was 1 %. Statistical analysis shows less severe adhesions on reoperation in Group II patients (pericardial defect patched) than in Group I patients (pericardial defect left open): mean grade of adhesions 1.6 ± 0.9 (Group II) versus 3.2 ± 0.6 (Group I), p < 0.001. Precardial adhesions with the pericardium left open were similar in patients having early and late reoperations: mean grade of adhesions 3.0 ± 0.7 (Group Ia) versus 3.4 ± 0.5 (Group Ib), no significant difference. Therefore, the glutaraldehydepreserved equine pericardial patch can be considered a suitable material for primary closure of the pericardial sac in patients with inadequate autologous pericardium.
Ludwig v. Segesser, M.D., Naima Jornod, and Bernard Faidutti, M.D., Geneva, Switzerland
Wth the expanding volume of cardiac surgery throughout the world, reoperations have become more and more common. The various reasons for repeat sternotomies include primary palliation in congenital heart disease, failure of valve substitutes, infection on cardiac prostheses, evolution of arteriosclerotic lesions, and occurrence of new lesions in coronary arteries or aorta-coronary bypass grafts. Postoperative epicardial adhesions impose a major risk during repeat cardiac operations because of the danger of damage to the heart, great vessels, or extracardiac grafts during attempts of resternotorny.': 2 ComFrom the Department of Surgery, Geneva University Hospital, Geneva, Switzerland. Received for publication Nov. 26, 1985. Accepted for publication April 16, 1986. Address for reprints: Ludwig v. Segesser, M.D., Department of Surgery, Geneva University Hospital, 1211 Geneva 4, Switzerland.
616
plete closure of the pericardial sac after open heart operations has been reported to lower the incidence of complications at reoperation.':" However, primary closure of the pericardial sac is not always feasible, particularly after resection of autologous pericardium for correction of congenital cardiac disease and suture reinforcement or after implantation of extracardiac conduits. Various pericardiaI substitutes, either biologicalor synthetic, have been recommended by several authors." However, the literature contains few clinical studies on reoperations. Therefore, we report our experience with repeat sternotomies after reconstruction of the pericardial sac with glutaraldehyde-preserved equine pericardium in comparison with operations in which the pericardium was left open. Methods Between March 1982 and September 1985, 200 patients underwent cardiac operations resulting in a pericardial defect that was closed with a glutaraldehyde-
Volume 93 Number 4
Repeat sternotomy after reconstruction of pericardial sac
April 1987
6 17
Table I Mean age (yr)
Groups of patients I.
7.5 8.8 6.6 5.0
Pericardium left open la. Pericardium left open, early reoperation « I yr) lb. Pericardium left open, left reoperation (> I yr) II. Pericardium closed with equine xenograft
± 4.8 ± 3.3
± 5.6 ± 4.2
Mean interval 5.4 4.8 6.4 3.7
± 2.9 yr ± 1.3 mo ± 3.2 yr ± 2.3 mo
Mean grade of adhesions * 3.2 3.0 3.4 1.6
± 0.6 ± 0.7 ± 0.5 ± 0.9
"Minimum = 0; maximum = 5.
Table II. Differences between groups Groups of patients I.
Mean age
Mean interval
Mean grade of adhesions
Pericardium left open II. Pericardium closed with equine xenograft
13 9
NS
>1 yr
P <0.001
la. Pericardium left open, early reoperation II. Pericardium closed with equine xenograft
5 9
NS
NS
P <0.01
la. lb.
5 8
NS
1 yr
Pericardium left open, early reoperation Pericardium left open, late reoperation
NS
Legend: NS. No significant difference.
preserved equine pericardial patch (Xenomedica, Lucerne, Switzerland). In this series, 182 patients or 91% were in the pediatric age group (less than 16 years old) and 18 patients or 9% were in the adult age group. In the pediatric age group were 53% boys and 47%
girls. An equine pericardial patch was used during the first operation in 169 pediatric patients and during the second operation in 13. The indication for implantation of a glutaraldehyde-preserved equine pericardial patch was always inadequate autologous pericardium for closure of the pericardial sac without important tension because of its previous use for repair or implantation of extracardiac conduits. In the pediatric age group this situation occurred mainly after complete repair of Fallot's tetralogy (53%), the Senning procedure in transpositions (7%), repair of double-outlet right ventricle (6%), repair of atrioventricular canal (5%), and the Rastelli procedure (3%). The glutaraldehyde-preserved equine pericardial patch was rinsed three times for 5 minutes in 1 L of saline solution, tailored to the shape and size required, and sutured to the edges of the patient's own pericardium. Intrapericardial and retrosternal sump drainage was always maintained for about 36 hours. To date, nine of the children and one of the adults has required reoperation (5% of the patients with equine pericardial patch closure) for various reasons. Anterior epicardial adhesions at reoperation were graded on a scale of 6: Grade 0 = no adhesions, Grade 1 = blunt adhesions, Grade 2 = moderate adhesions, Grade 3 = severe adhesions, Grade 4 = very severe
adhesions, and Group 5 = calcified or ossified adhesions. Patients were divided into four groups: Group I comprised the pediatric patients who received an equine pericardial patch at the second operation (n = 13). Epicardial adhesions were graded at the first reoperation and the equine pericardial patch was implanted afterward. Group fa included the Group I patients who required operation early «1 year) after the first operation (n = 5). Epicardial adhesions were graded at the first reoperation. Group Ib comprised the Group I patients who were reoperated on late (> 1 year) after the first operation (n = 8). Epicardial adhesions were graded at the first reoperation. Group II included the pediatric patients who were reoperated on after reconstruction of the pericardial sac with a glutaraldehydepreserved equine pericardial patch (n = 9). Adhesions between the epicardium and equine pericardial patch were graded at reoperation, and samples of the implant were taken for histologic examination. After cardiac repair, the pericardial sac was closed with a new glutaraldehyde-preserved equine pericardial patch. Results After a mean follow-up of 20.2 months (range 1 to 42 months) there were three infections of the equine pericardial patch. In one case a Dacron patch used for reconstruction of the right ventricular outflow tract may have been infected, but the infection was probably not due to the equine pericardial patch. An abscess was limited to the space between the two patches. The major
The Journal of
618
Segesser, Jornod, Faidutti
Fig. 1. Histologic section from equine pericardial patch' 2 months after implantation. Wavy collagen bundels and few fibrocyte cells can be seen. There is no indwelling organization or neovascularization. (Hematoxylin and eosin stain; original magnification X 150.)
part of the equine pericardial patch was free of infection. Therefore, the incidence of the equine pericardial patch infection was assessed as 1% (2/200). One adult with a Dacron tube graft of the ascending aorta had a sternal infection in the early postoperative period and the sternum had to be redivided for proper healing. Although a retrosternal infection had developed, the implanted xenologous pericardial patch resisted the infection and fatal infection of the synthetic ascending aorta was avoided. This patient is well after 2 years of follow-up. There was no evidence of any other patch-related complications in this series of 200 consecutive patients. Table I summarizes the grade of adhesions that developed in the four groups of patients undergoing reoperation. In the 13 patients in Group I (mean age 7.5 ± 4.8 years), the pericardium was left open at the first operation. At reoperation after a mean interval of
Thoracic and Cardiovascular Surgery
5.4 ± 2.9 years, the epicardial adhesions were graded 3.2 ± 0.6 (severe). Group la includes five patients from Group I (mean age 8.8 ± 3.3 years) who underwent reoperation less than I year after the first intervention. In these patients, whose pericardium was left open after the primary operation, the epicardial adhesions were graded 3.0 ± 0.7 (severe) after a mean interval of 4.8 ± 1.3 months. Group Ib comprises eight patients from Group I (mean age of 6.6 ± 5.6 years) who underwent reoperation more than 1 year after the first intervention. In these patients in whom the pericardium was left open, epicardial adhesions were graded 3.4 ± 0.5 (severe) after a mean interval of 6.4 ± 3.2 years. Group II includes the nine patients of the pediatric age group who underwent reoperation after pericardial sac closure with a glutaraldehyde-preserved equine pericardial patch. Mean age of the patients in this group was 5.0 ± 4.2 years. Mean interval between implantation of the equine pericardial patch and reoperation was 3.7 ± 2.3 months. There were few adhesions between the pericardial patch and the sternum, and there were no problems during repeat sternotomy. The mean grade of the adhesions between the equine pericardial patch and the epicardium was 1.6 ± 0.9 (blunt to moderate). In some cases without adhesions white epicardial thickening occurred (the coronary arteries were not obscured in these cases). Histologic sections of explanted equine pericardial patches showed wavy collagen bundles without signs of inflammatory reactions. There was no indwelling organization or neovascularization (Fig. 1). The results of the statistical analysis of the different groups are summarized in Table II. The mean grade of adhesions was lower in Group II patients, whose pericardial sac was closed with an equine pericardial patch, than in Group I patients, whose pericardium was left open (II versus I, p < 0.001). There is no significant difference in mean age of the two groups but there is a difference in the mean interval (5.4 ± 2.9 years in Group I and 3.7 ± 2.3 months in Group II). The mean grade of adhesions in Group II patients was lower than in Group la patients, who had the pericardium left open and early reoperation (II versus Ia, p < 0.01), but no significant difference was noted between mean age and mean interval in these two groups (Ia and II). No significant difference in the mean grade of adhesions was noted between Group I patients who had early reoperation (Group la) and those who had late reoperation (Group Ib). Discussion Anterior epicardial adhesions at reoperation are less severe after reconstruction of the pericardial sac with glutaraldehyde-preserved pericardium (Group II) than
Volume 93 Number 4 April 1987
Repeat sternotomy after reconstruction of pericardial sac 6 1 9
after reconstruction with the pericardium left open (Group I). There is no significant difference in tightness of adhesions in patients with the pericardium left open and early reoperation (Group Ia) in comparison to patients with the pericardium left open and late reoperation (Group Ib). The increased morbidity and mortality of cardiac reoperation as a result of catastrophic hemorrhage is well documented, 1,2,10 and primary closure of the pericardial sac has been advocated by several authors.v'" When there is inadequate autologous pericardium for closure of the pericardial sac without important tension, pericardial substitutes have been recommended." Pericardial substitutes allow safer resternotomy because of separation of the posterior sternal wall from the cardiac cavities and the great vessels. The present study confirms that there are few epicardial adhesions with glutaraldehyde-preserved pericardium." For best results, the glutaraldehyde-preserved pericardium should be rinsed and handled according to specific directions. I I In coronary artery bypass operations pericardial meshing as proposed by Milgalter and associates" might be preferable to xenograft pericardial patches, as the former seems to produce less epicardial thickening masking the underlying coronary anatomy. Still, clinical reports on pericardial meshing are not yet available. Furthermore, in the case of substantial pericardial defects, as after repair of congenital cardiac disease, pericardial meshing might be insufficient to cover the necessary surface area. The reported 1% infection rate for the equine pericardial patch in the present series is relatively important, although the patients survived without sequela after removal of the implanted patch. However, one also must consider the case in which the equine pericardial patch was resistant to anterior mediastinitis and shielded the Dacron tube replacing the ascending aorta. At reoperation after a comparable interval, adhesions were significantly less severe in patients whose pericardial sac was closed with an equine pericardial patch (Group II) than in patients whose pericardium was left open (Group Ia). The fact that there was no decrease in the tightness of the adhesions with time (similar mean grade of adhesions in patients with pericardial sac left open after 5 months [Group Ia] and after 6 years [Group Ib]) further supports implantation of a pericardial substitute. The tightest and most dangerous adhesions seem to occur between the posterior sternal wall and synthetic implants such as Dacron fabric used for patches and tube grafts. The relative stiffness of the adhesions is a
further risk for transection with an oscillating or Sarns type saw at resternotomy. Therefore, the posterior sternal wall should be separated from synthetic implants, great vessels, and cardiac cavities whenever possible. In conclusion, glutaraldehyde-preserved equine pericardium can be considered a suitable material for patch closure of the pericardial sac when primary closure is not feasible. REFERENCES I. English TAH, Milstein BB. Repeat open intracardiac operation. J THORAC CARDIOVASC SURG 1978;76:56-60. 2. Merin G, McGoon DC. Reoperation after insertion of aortic homograft as a right ventricular outflow tract. Ann Thorac Surg 1973;16:122-6. 3. Cunningham IN, Spencer FC, Zeff R, Williams CD, Cukingnan R, Mullin M. Influence of primary closure of the pericardium after open-heart surgery on the frequency of tamponade, postcardiotomy syndrome, and pulmonary complications. J THORAC CARDIOVASC SURG 1975;70:11925. 4. Culliford AT, Cunningham IN, Zeff RH, Isom OW, Teiko P, Spencer Fe. Sternal and costochondral infec~ tions following open-heart surgery: a review of 2,594 cases. J THORAC CARDIOVASC SURG 1976;72:714-20. 5. Kohanna FH, Adams PX, Cunningham IN. Use of autologous fascia lata as pericardial substitute following open-heart surgery. J THORAC CARDIOVASC SURG 1977;74:14-9. 6. Gallo JI, Pomar JL, Artifiano E, Val F, Duran CMG. Heterologous pericardium for the closure of pericardial defects. Ann Thorac Surg 1978;26:149-54. 7. Bonnabeau RC, Armanious AW, Tarnay TS. Partial replacement of pericardium with dura substitute. J THORAC CARDIOVASC SURG 1973;66:196-201. 8. Inglis AE, Richter RC, Alonso DR, Gay W A. Glutaraldehyde-preserved pericardial allografts for pericardiaI closure. J Surg Res 1984;36:50-4. 9. Yakirevich VS, Abdulali SA, Abbott CR, lonescu MI. Reconstruction of the pericardiaI sac with glutaraldehydepreserved bovine pericardium. Texas Heart Inst J 1984;11:238-42. 10. Dobell ARC, Jain AK. Catastrophic hemorrhage during redo sternotomy. Ann Thorac Surg 1984;37:273-8. II. Skinner JR, Hyunchul K, Toon TS, Kongtahworn C, Phillips SJ, Zeff RH. Inflammatory epicardial reaction to processed bovine pericardium: case report. J THORAC CARDIOVASC SURG 1984;88:789-91. 12. Milgaiter E, Uretzky G, Siberman S, et al. Pericardial meshing: an effective method for prevention of pericardiaI adhesions and epicardial reaction after cardiac operation. J THORAC CARDIOVASC SURG 1985;90:281-6.
THORAC CARDIOVASC SURG
1987;93:616-9
Repeat sternotomy after reconstruction of the pericardial sac with glutaraldehyde-preserved equine pericardium The risk of repeat sternotomy is higher than that of the initial sternotomy, especially if the pericardial sac was left open at the first intervention. In 200 consecutive patients with a pericardial defect after open heart operations, the pericardium was closed with a glutaraldehyde-preserved equine pericardial patch. Precardiac adhesions at reoperation were assessed in four groups of patients on a scale of 6, ranging from o (no adhesions) to 5 (calcified or ossified adhesions). Group I comprised 13 patients in whom the pericardium was left open at the first operation and an equine pericardial patch was implanted at reoperation. Group Ia included the first five Group I patients who underwent reoperation less than 1 year (early reoperation) after the initial procedure. Group Ib included the other eight patients of Group I, who underwent reoperation more than 1 year (late reoperation) after the first procedure. Group II comprised nine patients who were reoperated on after reconstruction of the pericardial sac with a glutaraldehydepreserved equine pericardial patch. After a mean follow-up of 20.2 months, the incidence of patch-related complications was 1 %. Statistical analysis shows less severe adhesions on reoperation in Group II patients (pericardial defect patched) than in Group I patients (pericardial defect left open): mean grade of adhesions 1.6 ± 0.9 (Group II) versus 3.2 ± 0.6 (Group I), p < 0.001. Precardial adhesions with the pericardium left open were similar in patients having early and late reoperations: mean grade of adhesions 3.0 ± 0.7 (Group Ia) versus 3.4 ± 0.5 (Group Ib), no significant difference. Therefore, the glutaraldehydepreserved equine pericardial patch can be considered a suitable material for primary closure of the pericardial sac in patients with inadequate autologous pericardium.
Ludwig v. Segesser, M.D., Naima Jornod, and Bernard Faidutti, M.D., Geneva, Switzerland
Wth the expanding volume of cardiac surgery throughout the world, reoperations have become more and more common. The various reasons for repeat sternotomies include primary palliation in congenital heart disease, failure of valve substitutes, infection on cardiac prostheses, evolution of arteriosclerotic lesions, and occurrence of new lesions in coronary arteries or aorta-coronary bypass grafts. Postoperative epicardial adhesions impose a major risk during repeat cardiac operations because of the danger of damage to the heart, great vessels, or extracardiac grafts during attempts of resternotorny.': 2 ComFrom the Department of Surgery, Geneva University Hospital, Geneva, Switzerland. Received for publication Nov. 26, 1985. Accepted for publication April 16, 1986. Address for reprints: Ludwig v. Segesser, M.D., Department of Surgery, Geneva University Hospital, 1211 Geneva 4, Switzerland.
616
plete closure of the pericardial sac after open heart operations has been reported to lower the incidence of complications at reoperation.':" However, primary closure of the pericardial sac is not always feasible, particularly after resection of autologous pericardium for correction of congenital cardiac disease and suture reinforcement or after implantation of extracardiac conduits. Various pericardiaI substitutes, either biologicalor synthetic, have been recommended by several authors." However, the literature contains few clinical studies on reoperations. Therefore, we report our experience with repeat sternotomies after reconstruction of the pericardial sac with glutaraldehyde-preserved equine pericardium in comparison with operations in which the pericardium was left open. Methods Between March 1982 and September 1985, 200 patients underwent cardiac operations resulting in a pericardial defect that was closed with a glutaraldehyde-
Volume 93 Number 4
Repeat sternotomy after reconstruction of pericardial sac
April 1987
6 17
Table I Mean age (yr)
Groups of patients I.
7.5 8.8 6.6 5.0
Pericardium left open la. Pericardium left open, early reoperation « I yr) lb. Pericardium left open, left reoperation (> I yr) II. Pericardium closed with equine xenograft
± 4.8 ± 3.3
± 5.6 ± 4.2
Mean interval 5.4 4.8 6.4 3.7
± 2.9 yr ± 1.3 mo ± 3.2 yr ± 2.3 mo
Mean grade of adhesions * 3.2 3.0 3.4 1.6
± 0.6 ± 0.7 ± 0.5 ± 0.9
"Minimum = 0; maximum = 5.
Table II. Differences between groups Groups of patients I.
Mean age
Mean interval
Mean grade of adhesions
Pericardium left open II. Pericardium closed with equine xenograft
13 9
NS
>1 yr
P <0.001
la. Pericardium left open, early reoperation II. Pericardium closed with equine xenograft
5 9
NS
NS
P <0.01
la. lb.
5 8
NS
1 yr
Pericardium left open, early reoperation Pericardium left open, late reoperation
NS
Legend: NS. No significant difference.
preserved equine pericardial patch (Xenomedica, Lucerne, Switzerland). In this series, 182 patients or 91% were in the pediatric age group (less than 16 years old) and 18 patients or 9% were in the adult age group. In the pediatric age group were 53% boys and 47%
girls. An equine pericardial patch was used during the first operation in 169 pediatric patients and during the second operation in 13. The indication for implantation of a glutaraldehyde-preserved equine pericardial patch was always inadequate autologous pericardium for closure of the pericardial sac without important tension because of its previous use for repair or implantation of extracardiac conduits. In the pediatric age group this situation occurred mainly after complete repair of Fallot's tetralogy (53%), the Senning procedure in transpositions (7%), repair of double-outlet right ventricle (6%), repair of atrioventricular canal (5%), and the Rastelli procedure (3%). The glutaraldehyde-preserved equine pericardial patch was rinsed three times for 5 minutes in 1 L of saline solution, tailored to the shape and size required, and sutured to the edges of the patient's own pericardium. Intrapericardial and retrosternal sump drainage was always maintained for about 36 hours. To date, nine of the children and one of the adults has required reoperation (5% of the patients with equine pericardial patch closure) for various reasons. Anterior epicardial adhesions at reoperation were graded on a scale of 6: Grade 0 = no adhesions, Grade 1 = blunt adhesions, Grade 2 = moderate adhesions, Grade 3 = severe adhesions, Grade 4 = very severe
adhesions, and Group 5 = calcified or ossified adhesions. Patients were divided into four groups: Group I comprised the pediatric patients who received an equine pericardial patch at the second operation (n = 13). Epicardial adhesions were graded at the first reoperation and the equine pericardial patch was implanted afterward. Group fa included the Group I patients who required operation early «1 year) after the first operation (n = 5). Epicardial adhesions were graded at the first reoperation. Group Ib comprised the Group I patients who were reoperated on late (> 1 year) after the first operation (n = 8). Epicardial adhesions were graded at the first reoperation. Group II included the pediatric patients who were reoperated on after reconstruction of the pericardial sac with a glutaraldehydepreserved equine pericardial patch (n = 9). Adhesions between the epicardium and equine pericardial patch were graded at reoperation, and samples of the implant were taken for histologic examination. After cardiac repair, the pericardial sac was closed with a new glutaraldehyde-preserved equine pericardial patch. Results After a mean follow-up of 20.2 months (range 1 to 42 months) there were three infections of the equine pericardial patch. In one case a Dacron patch used for reconstruction of the right ventricular outflow tract may have been infected, but the infection was probably not due to the equine pericardial patch. An abscess was limited to the space between the two patches. The major
The Journal of
618
Segesser, Jornod, Faidutti
Fig. 1. Histologic section from equine pericardial patch' 2 months after implantation. Wavy collagen bundels and few fibrocyte cells can be seen. There is no indwelling organization or neovascularization. (Hematoxylin and eosin stain; original magnification X 150.)
part of the equine pericardial patch was free of infection. Therefore, the incidence of the equine pericardial patch infection was assessed as 1% (2/200). One adult with a Dacron tube graft of the ascending aorta had a sternal infection in the early postoperative period and the sternum had to be redivided for proper healing. Although a retrosternal infection had developed, the implanted xenologous pericardial patch resisted the infection and fatal infection of the synthetic ascending aorta was avoided. This patient is well after 2 years of follow-up. There was no evidence of any other patch-related complications in this series of 200 consecutive patients. Table I summarizes the grade of adhesions that developed in the four groups of patients undergoing reoperation. In the 13 patients in Group I (mean age 7.5 ± 4.8 years), the pericardium was left open at the first operation. At reoperation after a mean interval of
Thoracic and Cardiovascular Surgery
5.4 ± 2.9 years, the epicardial adhesions were graded 3.2 ± 0.6 (severe). Group la includes five patients from Group I (mean age 8.8 ± 3.3 years) who underwent reoperation less than I year after the first intervention. In these patients, whose pericardium was left open after the primary operation, the epicardial adhesions were graded 3.0 ± 0.7 (severe) after a mean interval of 4.8 ± 1.3 months. Group Ib comprises eight patients from Group I (mean age of 6.6 ± 5.6 years) who underwent reoperation more than 1 year after the first intervention. In these patients in whom the pericardium was left open, epicardial adhesions were graded 3.4 ± 0.5 (severe) after a mean interval of 6.4 ± 3.2 years. Group II includes the nine patients of the pediatric age group who underwent reoperation after pericardial sac closure with a glutaraldehyde-preserved equine pericardial patch. Mean age of the patients in this group was 5.0 ± 4.2 years. Mean interval between implantation of the equine pericardial patch and reoperation was 3.7 ± 2.3 months. There were few adhesions between the pericardial patch and the sternum, and there were no problems during repeat sternotomy. The mean grade of the adhesions between the equine pericardial patch and the epicardium was 1.6 ± 0.9 (blunt to moderate). In some cases without adhesions white epicardial thickening occurred (the coronary arteries were not obscured in these cases). Histologic sections of explanted equine pericardial patches showed wavy collagen bundles without signs of inflammatory reactions. There was no indwelling organization or neovascularization (Fig. 1). The results of the statistical analysis of the different groups are summarized in Table II. The mean grade of adhesions was lower in Group II patients, whose pericardial sac was closed with an equine pericardial patch, than in Group I patients, whose pericardium was left open (II versus I, p < 0.001). There is no significant difference in mean age of the two groups but there is a difference in the mean interval (5.4 ± 2.9 years in Group I and 3.7 ± 2.3 months in Group II). The mean grade of adhesions in Group II patients was lower than in Group la patients, who had the pericardium left open and early reoperation (II versus Ia, p < 0.01), but no significant difference was noted between mean age and mean interval in these two groups (Ia and II). No significant difference in the mean grade of adhesions was noted between Group I patients who had early reoperation (Group la) and those who had late reoperation (Group Ib). Discussion Anterior epicardial adhesions at reoperation are less severe after reconstruction of the pericardial sac with glutaraldehyde-preserved pericardium (Group II) than
Volume 93 Number 4 April 1987
Repeat sternotomy after reconstruction of pericardial sac 6 1 9
after reconstruction with the pericardium left open (Group I). There is no significant difference in tightness of adhesions in patients with the pericardium left open and early reoperation (Group Ia) in comparison to patients with the pericardium left open and late reoperation (Group Ib). The increased morbidity and mortality of cardiac reoperation as a result of catastrophic hemorrhage is well documented, 1,2,10 and primary closure of the pericardial sac has been advocated by several authors.v'" When there is inadequate autologous pericardium for closure of the pericardial sac without important tension, pericardial substitutes have been recommended." Pericardial substitutes allow safer resternotomy because of separation of the posterior sternal wall from the cardiac cavities and the great vessels. The present study confirms that there are few epicardial adhesions with glutaraldehyde-preserved pericardium." For best results, the glutaraldehyde-preserved pericardium should be rinsed and handled according to specific directions. I I In coronary artery bypass operations pericardial meshing as proposed by Milgalter and associates" might be preferable to xenograft pericardial patches, as the former seems to produce less epicardial thickening masking the underlying coronary anatomy. Still, clinical reports on pericardial meshing are not yet available. Furthermore, in the case of substantial pericardial defects, as after repair of congenital cardiac disease, pericardial meshing might be insufficient to cover the necessary surface area. The reported 1% infection rate for the equine pericardial patch in the present series is relatively important, although the patients survived without sequela after removal of the implanted patch. However, one also must consider the case in which the equine pericardial patch was resistant to anterior mediastinitis and shielded the Dacron tube replacing the ascending aorta. At reoperation after a comparable interval, adhesions were significantly less severe in patients whose pericardial sac was closed with an equine pericardial patch (Group II) than in patients whose pericardium was left open (Group Ia). The fact that there was no decrease in the tightness of the adhesions with time (similar mean grade of adhesions in patients with pericardial sac left open after 5 months [Group Ia] and after 6 years [Group Ib]) further supports implantation of a pericardial substitute. The tightest and most dangerous adhesions seem to occur between the posterior sternal wall and synthetic implants such as Dacron fabric used for patches and tube grafts. The relative stiffness of the adhesions is a
further risk for transection with an oscillating or Sarns type saw at resternotomy. Therefore, the posterior sternal wall should be separated from synthetic implants, great vessels, and cardiac cavities whenever possible. In conclusion, glutaraldehyde-preserved equine pericardium can be considered a suitable material for patch closure of the pericardial sac when primary closure is not feasible. REFERENCES I. English TAH, Milstein BB. Repeat open intracardiac operation. J THORAC CARDIOVASC SURG 1978;76:56-60. 2. Merin G, McGoon DC. Reoperation after insertion of aortic homograft as a right ventricular outflow tract. Ann Thorac Surg 1973;16:122-6. 3. Cunningham IN, Spencer FC, Zeff R, Williams CD, Cukingnan R, Mullin M. Influence of primary closure of the pericardium after open-heart surgery on the frequency of tamponade, postcardiotomy syndrome, and pulmonary complications. J THORAC CARDIOVASC SURG 1975;70:11925. 4. Culliford AT, Cunningham IN, Zeff RH, Isom OW, Teiko P, Spencer Fe. Sternal and costochondral infec~ tions following open-heart surgery: a review of 2,594 cases. J THORAC CARDIOVASC SURG 1976;72:714-20. 5. Kohanna FH, Adams PX, Cunningham IN. Use of autologous fascia lata as pericardial substitute following open-heart surgery. J THORAC CARDIOVASC SURG 1977;74:14-9. 6. Gallo JI, Pomar JL, Artifiano E, Val F, Duran CMG. Heterologous pericardium for the closure of pericardial defects. Ann Thorac Surg 1978;26:149-54. 7. Bonnabeau RC, Armanious AW, Tarnay TS. Partial replacement of pericardium with dura substitute. J THORAC CARDIOVASC SURG 1973;66:196-201. 8. Inglis AE, Richter RC, Alonso DR, Gay W A. Glutaraldehyde-preserved pericardial allografts for pericardiaI closure. J Surg Res 1984;36:50-4. 9. Yakirevich VS, Abdulali SA, Abbott CR, lonescu MI. Reconstruction of the pericardiaI sac with glutaraldehydepreserved bovine pericardium. Texas Heart Inst J 1984;11:238-42. 10. Dobell ARC, Jain AK. Catastrophic hemorrhage during redo sternotomy. Ann Thorac Surg 1984;37:273-8. II. Skinner JR, Hyunchul K, Toon TS, Kongtahworn C, Phillips SJ, Zeff RH. Inflammatory epicardial reaction to processed bovine pericardium: case report. J THORAC CARDIOVASC SURG 1984;88:789-91. 12. Milgaiter E, Uretzky G, Siberman S, et al. Pericardial meshing: an effective method for prevention of pericardiaI adhesions and epicardial reaction after cardiac operation. J THORAC CARDIOVASC SURG 1985;90:281-6.