Open versus endoscopic saphenous vein harvesting: wound complications and vein quality.

BACKGROUND: The saphenous vein is an important conduit for coronary artery bypass grafting. Wound complications from traditional open vein harvesting occur often. Minimally invasive endoscopic saphenous vein harvesting may decrease wound complications. Vein quality may be an issue with endoscopic harvesting. METHODS: We reviewed 568 patients who had bypass grafting and saphenous vein harvesting either endoscopic (group A, n = 180) versus open (group B, n = 388). Both groups were demographically similar and management identical. Wound complication was defined by the need for intervention and included lymphocele, hematoma, cellulitis, edema, eschar, and infection. Multiple vein segments were obtained from 8 patients, 4 from each group, and examined histologically. RESULTS: Wound complications were significantly less in group A (9/180, 5%) versus group B (55/388, 14.2%), p value equal to or less than 0.001. Open harvesting (p< or =0.001), diabetes (p< or =0.001), and obesity (p< or =0.02) were risk factors for wound complication by univariate analysis. By multiple logistic analysis, open harvesting (p< or = 0.0007) and diabetes (p< or =0.0001) were independent risk factors for wound infection. Histologic evaluation of vein samples showed that there was no difference between the groups and vascular structural integrity was maintained. CONCLUSIONS: Endoscopic saphenous vein harvesting was associated with fewer wound complications and infections. Vein quality was not adversely effected because of endoscopic harvesting.

Improved early results after aortic valve replacement: analysis by surgical time frame.

Primary aortic valve replacement was performed in 430 patients. It was an isolated procedure in 339 and was combined with coronary artery bypass grafting in 91. Of these patients, 282 underwent operation from 1970 through 1976 (time frame 1) and 148 from 1980 through 1985 (time frame 2). They were divided into subgroups by age, New York Heart Association functional class, combined coronary artery bypass graft, and valvular lesion. Overall hospital mortality was 7.7% (time frame 1 = 10.6% versus time frame 2 = 2.0%; p less than 0.01). Overall, functional class III or IV was the strongest predictor of hospital mortality (p less than 0.001). Association of coronary artery bypass graft was the next strongest predictor of hospital mortality (p less than 0.01), and it retained its predictive value in time frame 2. Overall, hospital mortality was higher in patients older than 55 years (10.5% versus 3.5%; p less than 0.05). There were no hospital deaths in patients younger than 55 years in time frame 2. Type of valvular lesion was not a predictor of hospital mortality. Hospital mortality in patients receiving cardioplegia was 2%. Cardioplegia use has lessened the effect of age and functional class as predictors of hospital mortality after primary aortic valve replacement. Earlier operation in time frame 2 played a substantial role in the overall improvement of early results.

Tricuspid valve replacement for Ebstein's anomaly in childhood with a Starr-Edwards caged-ball prosthesis. 21-year follow-up.

A 21-year follow-up after tricuspid valve replacement (TVR) with a Starr-Edwards caged-ball prosthesis in a ten-year old boy is described. TVR is performed for Ebstein's anomaly, with strict indications in childhood. Despite the current preference for bioprostheses, good performance can be expected from the Starr-Edwards caged-ball valve.

Results of intraaortic balloon pumping after cardiac surgery: experience with the Percor balloon catheter.

One hundred thirty-eight patients undergoing an open-heart procedure required an intraaortic balloon pump (IAPB) postoperatively. In Group I (N = 45), the AVCO femoral conduit surgical technique was used; in Group II (N = 93), the Percor balloon was inserted either in the operating room after groin cutdown (open insertion) or percutaneously in the intensive care unit (percutaneous insertion). IABP usage increased in Group II (3% versus 1.6%; p less than 0.001). Immediate mortality was 40% (55/138). Use of the Percor balloon in Group II resulted in lower immediate mortality (32/93 or 34% versus 23/45 or 51%; p less than 0.06). Delayed mortality from multiorgan failure was 11.6% (16/138). Immediate percutaneous insertion at the bedside rather than a return to the operating room for open insertion yielded lower mortality (2/8 or 25% versus 6/7 or 86%; p less than 0.05). Open insertion of the Percor balloon decreases the failure rate of insertion compared with both the AVCO femoral conduit technique (7/85 or 8.2% versus 5/45 or 11%) and percutaneous insertion. It has more complications than the AVCO femoral conduit technique (7/85 or 8.2% versus 2/45 or 4.4%) and less than percutaneous insertion.

Strut failure of the Starr-Edwards cloth-covered metallic ball prosthesis. Report of two asymptomatic cases and review of the overall experience.

Two cases of asymptomatic strut failure of an aortic Starr-Edwards cloth-covered metallic ball prosthesis are reported. Strut failure can be asymptomatic for years. Available information on the overall worldwide experience of six cases is reviewed and suggestions made for early diagnosis and treatment of this rare complication.

Survival and "event-free" analysis of 785 patients with Björk-Shiley spherical-disc valves at 10 to 16 years.

Seven hundred eighty-five patients underwent Björk-Shiley spherical-disc valve replacement from 1970 to 1976. There were 268 mitral valve replacements (MVR), 227 aortic valve replacements (AVR), 65 double-valve replacements, and 225 "combined" procedures. A 97.2% follow-up (mean, 12 years) was achieved. With an operative mortality of 4.1% for MVR, 8.4% for AVR, 15.4% for double-valve replacement, and 12.4% for combined procedures, the 12-year survival was most closely related to age at valve replacement: age less than 50 years, 70%; age 50 through 59 years, 52%; and age 60 years or more, 38%. Twenty-four patients (3.1%) (6 who had MVR, 5 who had AVR, 1 who had double-valve replacement, and 12 who had combined procedures) had a thrombosed valve 1 to 134 months postoperatively; this is equal to 0.36 thrombosed valve per 100 patient-years. One hundred eighteen embolic episodes occurred in 94 (13%) of the operative survivors or 1.8 emboli per 100 patient-years. There were major bleeding complications in 0.5% of patients and minor bleeding complications, in 4.0%. Endocarditis appeared in 30 patients (4.2%) or 0.4 episode per 100 patient-years and paravalvular leaks, in 20 patients (2.8%). The event-free survival by age group and valve site at 5, 10, and 12 years is presented. Events included death, thrombosed valves, strokes, bleeding, emboli, paravalvular leaks, and endocarditis. There were 5.3 events per 100 patient-years excluding operative deaths.

Life-threatening chylothorax after left internal mammary artery dissection: therapeutic considerations.

Persistent chylothorax developed in a 53-year-old man after left internal mammary artery (LIMA) takedown and required surgical intervention. After an unsuccessful supraclavicular approach, left-sided standard thoracotomy showed thick adhesions around the LIMA takeoff with a diffuse oozing rather than an identifiable discrete leak. A possible leaking point was stitched, the area was sealed with fibrin adhesive, and complete remission ensued. Operation for chylothorax after LIMA takedown is challenging. A left-sided standard thoracotomy with minimal dissection and use of fibrin adhesive rather than blind stitching are recommended.

The implanted defibrillator: relation of defibrillating lead configuration and clinical variables to defibrillation threshold.

Forty-two defibrillating lead systems for the automatic implantable defibrillator were implanted and tested in 41 patients. Two basic lead configurations were used: 1) spring-patch, consisting of a transvenous superior vena cava spring electrode as the anode and an apical or left lateral ventricular patch electrode (either small [13.9 cm2] or large [27.9 cm2]) as the cathode; and 2) patch-patch, consisting of an anterior right ventricular patch as the anode and a posterior left ventricular patch as the cathode. Of the 42 lead systems, 10 were spring-patch and 32 were patch-patch combinations. The defibrillation threshold for the patch-patch combinations (9.8 +/- 6.5 J, mean +/- standard deviation) was significantly (p less than 0.01) lower than that for the spring-patch combinations (19.1 +/- 10.3 J). Subgroup analysis revealed the lowest defibrillation thresholds for patch-patch combinations with at least one large patch. Total surface area of defibrillating leads was strongly negatively correlated with the defibrillation threshold (p less than 0.005). Analysis of the relation of clinical variables to defibrillation threshold revealed that only amiodarone therapy was independently associated with a significantly (p less than 0.05) higher defibrillation threshold. Thus, surface area of the defibrillating leads is a critical determinant of the defibrillation threshold for the implanted defibrillator. Patch-patch lead systems with at least one large patch may provide an increased safety margin for defibrillation. Conversely, amiodarone therapy is associated with higher defibrillation thresholds and may decrease the margin of safety.

Transmural myocardial flow distribution during hypothermia. Effects of coronary inflow restriction.

Hypothermic coronary perfusion and blood cardioplegia have been used clinically to minimize intraoperative myocardial damage. However, pressure-flow characteristics in regions supplied by inflow-limiting collateral coronary arteries have not been investigated during hypothermic conditions. In this study tracer microspheres determined transmural myocardial blood flow distribution during cardiopulmonary bypass in normothermic empty, beating dog hearts (EBH), during hypothermic sanguineous perfusion at 15 degrees C (HP), and after hemodilution of cooled (15 degrees C) hearts to a hematocrit value of 20 vol% (HDL). Animals in Group I (N = 8) had normal hearts. Group II dogs (N = 9) had one region supplied predominantly by narrow collateral vessels (CR) and another nourished by normal coronary arteries (NR). Retrograde circumflex pressures were measured continuously for Group II as an additional index of CR perfusion. Flow characteristics in Group I hearts were always similar to the NR of Group II dogs. With HP, endocardial blood flow in the NR decreased from approximately 0.80 to 0.50 ml/min/gm. Subsequently, following HDL this flow increased to approximately 1.70 ml/min/gm, or over twice control levels. In comparison, flow to CR endocardium decreased even more during HP (0.12 ml/min/gm). Even though control flow levels were reestablished in CR endocardium by adding HDL, an unfavorable endocardial/epicardial ratio persisted. With both HP and HDL, retrograde circumflex pressure never changed from EBH values. These data suggest that a significant endocardial flow defect exists during periods of hypothermic sanguineous perfusion and may become more prevalent in regions subserved by inflow-limiting coronary vessels. Similar flow maldistributions may occur in patients if blood-containing cardioplegic solutions are used and during systemic hypothermia. Significant hemodilution helps minimize these imbalances and permits salutary effects of hypothermia to be delivered more evenly across the ventricular wall.

Myocardial blood flow and function during gradual coronary occlusion in awake dog.

The purpose of this study was to correlate dimensions and transmural blood flow in a segment of myocardium supplied by a coronary artery undergoing gradual closure. Nine adult dogs were instrumented with an electromagnetic flow probe, pneumatic occluder, and an ameroid constrictor on the circumflex coronary artery. Ultrasonic crystals were implanted 10 mm from the external surface of the left ventricle in a segment perfused by the circumflex artery. Regional blood flow was determined with 7- to 10-microns radiolabeled microspheres. Data were collected in the awake state at rest before closure began (control), during partial closure, and immediately after total closure. Seven of the nine animals were studied after occlusion during treadmill exercise. During both partial and total closure at rest the rate and extent of systolic shortening as well as the transmural blood flow were unchanged from control. During treadmill exercise mean flow increased. However, flow was redistributed away from the inner two layers causing deterioration in both the rate and extent of shortening of this segment. These data suggest that, although regional myocardial function and flow can be maintained at rest by the immature canine collateral circulation, these parameters are impaired markedly during augmented flow with exercise.

Measurement of global ventricular function in patients during cardiac operations using sonomicrometry.

In these studies, we evaluated the applicability of pulse-transit sonomicrometry for measuring changes in global cardiac activity in patients during cardiac operative procedures. In six patients two epicardial ultrasonic crystals (8 mm) were sutured across the left ventricular minor axis. Diastolic pressure-length data were recorded as left ventricles were filled passively to transmural pressures (P) from 0 to 20 mm Hg. Data were collected at the beginning of cardiopulmonary bypass and again 15 minutes following periods of induced global ischemia (29.8 +/- 0.8 minutes). Minor axis length data were normalized to Lagrangian strain (epsilon), and best-fit regression curves were obtained from P-epsilon by computer analysis. Nonlinear elastic constants, alpha and beta, were mathematically derived as additional curve descriptors. Decreases in ventricular compliance were demonstrated as leftward shifts in both computed and measured P-epsilon curves. Global ischemia appeared to effect a decrease in overall ventricular diastolic compliance in all patients studied (p =7E 0.01 at 5, 10, 15, 20 mm Hg). Simultaneously, no statistical change occurred in lo (62.93 +/- 2.80 mm), which represented end-diastolic length (EDL) at 0 mm Hg transmural pressure. Following coronary grafting several patients showed augmented systolic excursion when compared at similar EDL. For those analyses, shortening was compared at specific minor axis EDLs rather than filling pressures. These data indicate that experimentally developed sonomicrometry may safely provide accurate indices of systolic and diastolic ventricular properties during operations necessitating cardiopulmonary bypass. Thus various cardioplegic solutions, ischemic arrest periods, and inotropic agents may be evaluated more objectively.

Regional myocardial dimensions following coronary artery bypass grafting in patients. Relationship of functional deterioration to graft occlusion.

The direct relationship between graft flow and regional midwall myocardial function has not been documented in patients. Therefore, the present study was designed to quantitate the effects of coronary artery bypass grafting on regional myocardial mechanics distal to a coronary artery obstruction. Twenty-one patients with subtotal or total occlusion of the left anterior descending (LAD) coronary artery underwent coronary artery bypass grafting. Following completion of the aortic and coronary anastomoses, two miniature ultrasonic dimension transducers (2.5 mm. diameter) were positioned within the minor axis of the anterior left ventricular free wall and were allowed complete freedom of movement. The transducers were placed at midwall depth, and areas of clinically apparent myocardial fibrosis were not utilized as sites of implantation. During control, 30 minutes following the termination of cardiopulmonary bypass, regional myocardial dimensions, pulmonary artery diastolic pressure, arterial pressure, and heart rate were recorded with all saphenous vein grafts open and after 30 seconds of single vein graft occlusion. These measurements were repeated during atrial pacing at a rate of 128 +/- 4 beats per minute. Data are mean +/- the standard error of the mean. During control, graft occlusion resulted in a regional decrease in systolic excursion from 1.3 +/- 0.1 to 1.0 +/- 0.2 mm. (p less than 0.01), as well as a decrease in the rate of shortening from 8.7 +/- 0.2 to 6.2 +/- 1.1 mm. per second (p less than 0.05); heart rate, mean arterial pressure, and diastolic pulmonary artery pressure remained unchanged. Graft occlusion with atrial pacing resulted in an exaggerated decrease in both regional systolic excursion, from 1.2 +/- 0.2 to 0.6 +/- 0.2 mm. (p less than 0.01), and rate of shortening, from 9.4 +/- 1.5 to 4.4 +/- 0.2 mm. per second (p less than 0.01). For the group of patients studied, end-diastolic lengths were unchanged with graft occlusion during control and atrial pacing. Moreover, with graft occlusion, isolated patients demonstrated regional dyskinesia as evidenced by holosystolic bulging. These studies in patients have documented for the first time that, despite a constant preload, afterload, and heart rate, regional myocardial function following coronary artery bypass grafting is dependent upon adequate graft flow, especially during stress.

The effects of intermittent ischemic arrest on the perfusion of myocardium supplied by collateral coronary arteries.

Six weeks after placement of an ameroid constrictor on the circumflex coronary artery, blood flow in a collateral region was compared with flow in myocardium supplied by normal arteries during cardiopulmonary bypass (80 mm Hg). Myocardial blood flow was determined using radionuclide-labeled microspheres (8 to 10 mu) before 10 minutes of ischemic arrest and after 1, 5, and 10 minutes of reperfusion. The retrograde circumflex pressure was monitored continuously and served as an additional index of perfusion of the collateral region. During reperfusion, endocardial flow in the collateral region remained unchanged despite a threefold increase in a similar layer having normal arteries (p less than 0.01). Following ischemic arrest, mean transmural and subendocardial hyperemic responses both persisted for longer than 10 minutes in normal regions. Simultaneously, peripheral circumflex pressures decreased at 1 and 5 minutes of reperfusion (p less than 0.001) but returned to control within 10 minutes. Persistently elevated endocardial flow in the normal arteries and the absence of a hyperemic response in the collateral region during an associated decrement in retrograde circumflex pressure may indicate incomplete flow repayment even after 10 minutes of reperfusion. Marked transmural flow imbalances despite adequate coronary perfusion pressures suggest that intermittent ischemic arrest may cause cumulative ischemia, and this occurrence may be detrimental especially in collateral regions of myocardium.

Pressure-flow characteristics of the coronary collateral circulation during cardiopulmonary bypass. Effects of ventricualr fibrillation.

Even though ventricular fibrillation is used frequently during cardiopulmonary bypass (CPB), the effects of fibrillation on myocardial regions supplied by collateral vessels have not been determined. To study these effects, nine dogs with left ventricles (ameroid model) consisting of a region of myocardium supplied by collateral vessels (CR) and a region supplied by normal coronary arteries (NR) were subjected to normothermic CPB at two perfusion pressures. In both the empty beating heart (EBH) and empty fibrillating heart (EFH) regional myocardial flow was determined by tracer microspheres. Retrograde coronary pressure was measured via cannulation of the circumflex artery distal to the ameroid induced occlusion. When perfusion pressure was maintained at 80 mm Hg, retrograde coronary pressure was similar in the EBH (46 +/- 4 mm Hg) and in the EFH (48 +/- 3 mm Hg). During fibrillation subendocardial flow in the CR was unchanged, while flow in the NR increased (P less than 0.02). In addition, the endo/epi was greater in the NR than in the CR (P less than 0.01), a difference which did not exist in the EBH. The flow response to fibrillation in the CR could be produced in the NR by reducing the perfusion pressure to 50 mm Hg. These data suggest that during CPB, fibrillation exaggerates existing subendocardial perfusion deficits in collateral regions and the impaired flow response appears to be related to a low regional intravascular pressure.

Myocardial blood flow distribution in concentric left ventricular hypertrophy.

Regional myocardial blood flow during both control conditions and ischemia-induced vasodilatation was studied in eight chronically instrumented awake dogs. Seven of these animals had coarctation-banding of the ascending aorta performed at 6 wk of age, and the other dog had congenital subvalvular aortic stenosis. The mean left ventricular weight for the group was 157+/-7.6 g, and the left ventricular body weight ratio was 8.76+/-0.47 g/kg. None of the animals exhibited signs of congestive heart failure. During the control state, the mean left ventricular systolic pressure was 249+/-12 mm Hg and the left ventricular end-diastolic pressure was 11.5+/-0.5 mm Hg. The aortic diastolic pressure was 74+/-6 mm Hg. Mean left circumflex coronary artery blood flow was 71+/-6 cm(3)/min. In the animals with coarctation-banding, 52+/-6% of the flow occurred during systole. In the dog with congenital subvalvular aortic stenosis, 5% of the coronary flow was systolic. Mean transmural blood flow during resting conditions was 0.97+/-0.08 cm(3)/min per g, and the ratio of endocardial to epicardial flow (endo/epi) was 0.88+/-0.07. During reactive hyperemia, the mean transmural blood flow increased to 3.5+/-0.30 cm(3)/min per g; however, the endo/epi decreased to 0.52+/-0.06.THESE STUDIES DOCUMENT A DIFFERENCE IN TRANSMURAL BLOOD FLOW DISTRIBUTION BETWEEN THE NORMAL AND THE HYPERTROPHIED LEFT VENTRICLE: during resting conditions, in the normal ventricle, the highest flow occurs in the endocardial layer, whereas in the hypertrophied ventricle, the highest flow is in the middle layers with the endocardial flow less than the epicardial flow. During ischemia-induced vasodilatation, the abnormal endo/epi becomes accentuated markedly. These data demonstrate that, in situations requiring high flow, the endocardial layer of a heart with marked concentric left ventricular hypertrophy may not be perfused adequately.

A reproducible model of moderate to severe concentric left ventricular hypertrophy.

A reproducible model for the production of moderate to severe concentric left ventricular hypertrophy has been developed in this laboratory. Coarctation-banding of the ascending aorta was performed successfully in 10 puppies. There were no late deaths related to aortic rupture, and in the dogs surviving for 1 yr no evidence of congestive heart failure was present. A second operative procedure was performed in seven dogs for chronic instrumentation, and all survived. Severe supravalvular aortic stenosis with a marked peak systolic pressure gradient was noted in each dog. Postmortem examination revealed a substantial increase in left ventricular mass and in the ratio of left ventricular to body weight.