Pulmonary embolism after cardiac surgery.

OBJECTIVES: We examined the incidence of pulmonary embolism after cardiac surgery. BACKGROUND: Because venous thromboembolism is considered to be an uncommon complication after cardiac surgery, its incidence was documented in a consecutive series of 1,033 patients who underwent cardiac surgery over a 5-year period. METHODS: Parallel cohorts of patients in a tertiary referral center were evaluated and the incidence of pulmonary embolism was compared in subgroups of patients undergoing coronary bypass surgery, valve surgery and combined procedures. RESULTS: Pulmonary embolism developed in 33 (3.2%) of the 1,033 cardiac surgical patients, within 2 weeks of a coronary bypass operation in most; it did not develop in any patient who had isolated valve replacement surgery (p < 0.05). The diagnosis of pulmonary embolism was established by pulmonary angiography in 24 patients, ventilation/perfusion lung scan in 3, postmortem examination in 5 and clinical examination in 1 patient. Important risk factors for pulmonary embolism included prolonged postoperative recovery, obesity and hyperlipidemia. The mortality rate was 18.7% in patients with in contrast to 3.3% in those without pulmonary embolism (p < 0.01). CONCLUSIONS: Although pulmonary embolism is rare after isolated valve replacement, it is not an uncommon complication after coronary bypass surgery.

Significance of changes in resting left ventricular ejection fraction after coronary artery bypass grafting.

The prognostic significance of changes in resting left ventricular ejection fraction was examined in 102 patients who underwent successful coronary artery bypass grafting. Between preoperative and early postoperative radionuclide ventriculography, mean resting left ventricular ejection fraction improved from 47.2% to 53.9% (p less than 0.01). Left ventricular ejection fraction increased by 5% or greater in 64 patients (63%), remained unchanged (within 4%) in 31 (30%), and decreased by at least 5% in 7 (7%). During 14 to 39 months (mean 27 months) of clinical follow-up, patients with normal preoperative left ventricular ejection fraction had a lower prevalence of recurrent angina, congestive heart failure, and mortality resulting from cardiovascular disease. Cardiovascular morbidity and mortality occurred with equal frequency for patients who did and did not show early postoperative improvement in left ventricular ejection fraction (36% versus 39%). Among 69 patients who had a third radionuclide ventriculography at late follow-up, left ventricular ejection fraction was less than the early postoperative value in 69% and less than the preoperative result in 36%. Patients with early postoperative improvement in left ventricular ejection fraction were more likely to retain resting left ventricular contractile function, at least at the preoperative level (71% versus 46%).

The superiority of continuous cold blood cardioplegia in the metabolic protection of the hypertrophied human heart.

The effects of sanguineous and asanguineous cardioplegia on the generation of myocardial acid in the hypertrophied human heart during aortic clamping and reflow were elucidated by continuous intraoperative monitoring of myocardial pH in 42 patients undergoing valve replacement, with or without coronary bypass. The patients were divided into three groups: Group I (n = 14) received intermittent crystalloid cardioplegia; group II (n = 14) received intermittent blood cardioplegia; and group III (n = 14) received continuous blood cardioplegia. The groups were matched according to six previously elucidated determinants of myocardial acidosis. Measurements were made of myocardial pH, hydrogen ion concentration ([H+]), and the difference in pH units between myocardial pH and the pH of neutrality of water at the corresponding temperature (delta pHn). Throughout aortic clamping, myocardial pH in groups I and II fell significantly by 0.46 +/- 0.08 and 0.15 +/- 0.07 units, respectively (p less than 0.001) between the groups). In contrast, myocardial pH remained statistically unchanged throughout aortic clamping in group III (p less than 0.001 compared to groups I and II). Similar relationships were observed in [H+] and delta pHn during aortic clamping. During the early reflow, myocardial acidosis was observed in all three groups and delta pHn in group III increased from -0.26 +/- 0.10 at the end of aortic clamping to -0.57 +/- 0.07 during reperfusion (p less than 0.03). Patients in groups II and III required significantly less inotropic and mechanical cardiac support than patients in group I (p = 0.017). Hence, although continuous blood cardioplegia does not completely prevent acid accumulation during reflow, it provides better metabolic protection of the hypertrophied human heart than either intermittent crystalloid or intermittent blood cardioplegia.

Regional changes in myocardial acid production during ischemic arrest: a comparison of sanguineous and asanguineous cardioplegia.

Regional differences in myocardial acid production have not been characterized during administration of either asanguineous or sanguineous cardioplegia. To investigate this, miniature glass pH electrodes were placed in the right ventricular (RV) myocardium, the left ventricular subendocardial (LV endo) region, and the subepicardial (LV epi) region in a canine model. Multiple doses of either blood cardioplegia (Group 1; N = 11) or crystalloid cardioplegia (Group 2; N = 11) were administered during 4 hours of aortic cross-clamping. The accumulation of hydrogen ions during the cross-clamp period was greater in Group 2 than Group 1 in the LV endo region (629 +/- 79 nm/L versus 66 +/- 31 nm/L; p less than 0.001), the LV epi region (623 +/- 66 nm/L versus 72 +/- 32 nm/L; p less than 0.001), and the RV myocardium (814 +/- 296 nm/L versus 150 +/- 54 nm/L; p less than 0.05). Within each group, the time course of myocardial pH and the accumulation of hydrogen ions did not differ among the LV endo region, LV epi region, and the RV myocardium (p = not significant). These data indicate that transmural and interventricular differences in myocardial pH and hydrogen ion accumulation are not produced in the vented, arrested canine heart. In addition, when compared with asanguineous cardioplegia, blood cardioplegia globally and transmurally reduces acid accumulation during ischemic arrest.

Structural and metabolic correlates of cell injury in the hypertrophied myocardium during valve replacement.

To characterize the ultrastructural and metabolic changes occurring in the hypertrophied ventricle during cardiac operations in man, we studied 36 patients with valvular heart disease undergoing valve replacement, during which multiple doses of cold potassium cardioplegic solution were administered (Group I). Each patient had substantial ventricular hypertrophy according to measurements made of left ventricular mass, with a mean of 232.1 +/- 19.8 gm/m2 (normal: 92 +/- 16 gm/m2). Serial biopsy specimens were obtained from the left ventricular apex at the initiation of bypass, during the cross-clamp interval, and during reperfusion. Each specimen was scored from 0 to 4 according to ischemic changes in nuclear chromatin, mitochondrial swelling, myofibrillar edema, glycogen depletion, and overall cell morphology. Myocardial pH and temperature were measured continuously in the left ventricular free wall. During the cross-clamp period, ischemic injury was evidenced by changes in nuclear chromatin (0.38 +/- 0.10 to 1.25 +/- 0.21, p less than 0.0001), intracellular edema (0.43 +/- 0.06 to 0.97 +/- 0.14, p less than 0.002), overall cell morphology (0.37 +/- 0.06 to 0.97 +/- 0.14, p less than 0.001), and mitochondria (0.10 +/- 0.05 to 0.19 +/- 0.07, p less than 0.0001). During reperfusion, mitochondrial swelling increased further (0.19 +/- 0.07 to 0.35 +/- 0.08, p less than 0.0001) and glycogen stores were depleted (0.63 +/- 0.13 to 0.96 +/- 0.17, p less than 0.02), while the other structures remained unchanged. Myocardial pH declined during ischemic arrest from 6.89 +/- 0.04 to 6.40 +/- 0.04 (p less than 0.001). The changes in myocardial pH in Group I were compared to changes in myocardial pH in 10 patients (Group II) with no left ventricular hypertrophy undergoing isolated coronary bypass graft operations with the same protective techniques. In contrast to Group I, myocardial pH did not fall in Group II during ischemic arrest (6.98 +/- 0.06 to 6.94 +/- 0.05, p = not significant). Thus, with the use of current myocardial protective techniques, ultrastructural and metabolic changes indicative of ischemia are produced in the hypertrophied myocardium. The structural alterations consist of changes in nuclear chromatin and intracellular edema during the ischemic phase and by mitochondrial swelling during reperfusion.

Delayed sternal closure. An improved method of dealing with complications after cardiopulmonary bypass.

Intraoperative closure of the median sternotomy after cardiac operations in patients with complications, including severe postoperative bleeding, impaired cardiac function caused by myocardial edema, and cardiac dilatation, may lead to a critical and possibly fatal deterioration of hemodynamic function. In an effort to prevent this complication, we delayed mediastinal closure in 15 patients, covering the wound temporarily with a sheet of rubber latex (Esmarch bandage). An oval patch of this material was sized and sutured to the skin edges with a continuous suture. This technique provided easy and fast access to the mediastinal structures in four of the 15 patients who required multiple surgical interventions in the early postoperative period. Delayed closure was indicated for severe bleeding in 10 patients, heart compression in four patients, and severe postbypass arrhythmias in one patient. Definitive closure of the chest was delayed until satisfactory hemostasis was achieved or the heart size returned to normal. Thirteen of the 15 patients were long-term survivors, none of them had wound infections, and their wounds healed well. Delayed closure of the median sternotomy was an effective and safe approach in these groups of critically ill patients.

The significance of the late fall in myocardial PCO2 and its relationship to myocardial pH after regional coronary occlusion in the dog.

After acute regional coronary occlusion, myocardial tissue PCO2, as measured by mass spectrometry, rises, reaches a peak, and then gradually falls. This late fall in myocardial tissue PCO2 could be due to (1) a gradual increase in tissue blood flow (and hence improved carbon dioxide washout), (2) a gradual consumption of tissue bicarbonate, (3) a gradual reduction in the production of carbon dioxide due to progressive cellular damage, or (4) an artifact caused by the continued presence of the mass spectrometer probe in the ischemic tissue. To determine which of these four mechanisms is responsible for the late fall in myocardial tissue PCO2, we subjected 27 anesthetized open-chest dogs to 3-hour occlusion of the left anterior descending coronary artery. Both myocardial tissue PCO2 and intramyocardial hydrogen ion concentration were measured in the myocardial segment supplied by the left anterior descending coronary artery. Ten dogs (group 1) were killed after the occlusion (occlusion I), and 11 dogs (group 2) underwent reocclusion (occlusion II) at the same site after a 45-minute period of reflow. Regional myocardial blood flow was measured periodically by the intramural injection of 127Xe. Changes in myocardial tissue PCO2 and hydrogen ion concentration were related to ultrastructural changes in the tissues adjacent to the myocardial tissue PCO2 probe. Regional myocardial blood flow remained unchanged throughout the 3-hour occlusion, ruling out increased carbon dioxide washout as a cause for its late fall. Tissue hydrogen ion concentration, as measured by a new lead glass electrode, correlated well with myocardial tissue PCO2, with the reduction in regional myocardial blood flow, and with ischemic damage assessed histologically. Myocardial hydrogen ion concentration also exhibited a late fall after the occlusion, from a peak of 199.8 +/- 27.8 nmol/liter to 91.9 +/- 12.1 nmol/liter (mean +/- SEM). This ruled out consumption of tissue bicarbonate as the cause for the late fall in myocardial tissue PCO2. Peak rise in myocardial tissue PCO2 after occlusion II (71.2 +/- 7.9 mm Hg) was significantly lower than peak myocardial tissue PCO2 after occlusion I (116.7 +/- 13.9 mm Hg, P less than 0.001). The difference between these latter two values, as well as the magnitude of fall in myocardial tissue PCO2 during occlusion I, related directly to the degree of histological damage observed.(ABSTRACT TRUNCATED AT 400 WORDS)

Observations on 100 patients with continuous intraoperative monitoring of intramyocardial pH. The adverse effects of ventricular fibrillation and reperfusion.

Intramyocardial pH and temperature data recorded in 100 patients undergoing cardiac operations were analyzed to elucidate the effects of ventricular fibrillation and reflow. All patients underwent a single period of aortic clamping. Systemic hypothermia (25 degrees C) and intermittent cold crystalloid K+ cardioplegia were employed for myocardial protection. Baseline myocardial pH was 6.88 +/- 0.03 at a temperature of 36.5 degrees +/- 0.2 degree C. During the period of hypothermic ventricular fibrillation prior to aortic clamping, ventricular fibrillation did not affect myocardial pH in 45 patients (Group 1). In 21 patients (Group 2), it caused a significant drop in intramyocardial pH despite cooling. Group 2 patients had a higher incidence of valvular heart disease and left ventricular hypertrophy. They also exhibited low intramyocardial pH values during the subsequent periods of aortic clamping and reflow, indicating inadequate myocardial protection. During the period of reflow, reperfusion acidosis (pH less than 6.8 at 32 degrees C) was encountered in 39 patients (Group B) as opposed to 37 patients (Group A) whose pH remained well above 6.8 during that period. Group B patients had a higher incidence of valvular heart disease and left ventricular hypertrophy, tended to have more ischemic anterior walls prior to cardiopulmonary bypass, sustained longer periods of aortic clamping, had intramyocardial pH evidence of suboptimal protection during aortic clamping, were affected more adversely by ventricular fibrillation during reflow, and tended to have a higher operative mortality. Thus: Depending on the underlying myocardial disease, the adequacy of protection during aortic clamping, and the conditions of reflow, intramyocardial pH in man can fall significantly during ventricular fibrillation and reflow. The metabolic correlate of injury with reflow is a reperfusion acidosis that can reach as low as pH 5.98. When encountered, reperfusion acidosis can be minimized by prompt defibrillation.

First report of intramyocardial pH in man. II. Assessment of adequacy of myocardial preservation.

Intramyocardial pH and temperature were continuously measured in the anteroseptal region in 40 patients undergoing aortic cross-clamping during cardiac operations. Myocardial protection was achieved with systemic cooling (25 degrees C) and multidose potassium cardioplegia (4 degrees C). A clinical myocardial preservation score was devised based on intraoperative and postoperative need for inotropic support, postoperative creatine kinase isoenzyme (CK-MB) and electrocardiographic changes, and radionuclide ventriculography. The patients were divided into three groups according to their preservation scores. Group I (n = 17) with good preservation (scores 0 to 2), Group II (n = 15) with fair preservation (scores 3 to 8), and Group III (n = 8) with poor preservation (scores 9 to 15). Baseline intramyocardial pH was similar in all groups (mean +/- SEM = 6.77 +/- 0.03). With the administration of cold potassium cardioplegia, intramyocardial pH rose above baseline in all three groups. The magnitude of this rise related directly to the adequacy of preservation and to the duration of the cross-clamp period. Patients with lowest preservation scores and shortest cross-clamp periods had the highest intramyocardial pH. In contrast, there was no relationship between myocardial temperature during cross-clamp and either intramyocardial pH or the preservation score. The integrated mean intramyocardial pH during cross-clamp was found to be the parameter that correlated most with the adequacy of preservation. The correlation between intramyocardial pH and myocardial temperature during the period of cross-clamping related to the length of this period; it was good (r = 0.76, p less than 0.01) in periods of 40 minutes or less and very poor in periods exceeding 60 minutes (r = 0.27, p greater than 0.10). It is concluded that (1) the magnitude of rise in intramyocardial pH during the period of aortic cross-clamping is a good indicator of the adequacy of myocardial preservation; (2) during periods of aortic cross-clamping exceeding 40 minutes, myocardial temperature is a poor indicator of adequacy of preservation, since progressive tissue acidosis may occur despite low myocardial temperatures; and (3) techniques and solutions that can effectively reduce the progression of tissue acidosis will, in most likelihood, enhance our ability to protect the ischemic myocardium during cardioplegic arrest.

Prognosis for aorta-coronary graft patency. A comparison of preoperative and intraoperative assessments.

For prognostication of aorta-coronary bypass graft patency, preoperative angiographic assessment of the recipient vessel and the regional left ventricle was compared with the intraoperative measurement of the internal diameter of the vessel and graft flow in 98 patients, in whom graft patency was evaluated at 1 year. Preoperative assessment of the recipient vessel did not correlate with the intraoperative measurement of the internal diameter or with the graft patency. Preoperative assessment of the regional left ventricular wall motion, on the other hand, was more valuable for prognosticating graft patency. Regions assessed as normal or hypokinetic had equally good graft patency, whereas akinetic or dyskinetic areas had only a 54% patency rate. Intraoperative measurement of the internal diameter of the vessel and graft flow were reliable predictors of graft patency: The larger the internal diameter and the greater the graft flow, the better the graft patency. Best surgical results can therefore be predicted when the internal diameter of the recipient vessel is 2.5 mm or more, graft flow is 100 ml/min or more, and the regional left ventricle moves well.

Long-term follow-up of limbs after use of intra-aortic balloon counterpulsation device.

There has been concern for the viability and function of the limb distal to the femoral artery access site after use of the intra-aortic balloon counterpulsation device (IABCD). We have studied the long-term (6 to 50 months) hemodynamic status in the lower limbs of 31 patients who received circulatory support with the IABCD. Since both limbs were classified according to the pressure changes induced by treadmill walking, comparisons with the unoperated leg could be affected. Nine of ten patients with insertion problems had satisfactory results at follow-up. One limb was lost as a result of the use of IABCD. Two patients had required femorofemoral bypass, three required use of the opposite leg because of gross femoral disease, one required thrombectomy, three lost distal pulses, and one developed groin phlebitis. All these patients (except the one requiring amputation) had satisfactory hemodynamics at follow-up. Of the 21 patients with no insertion problems, all had satisfactory hemodynamics at follow-up, and 16 of the 21 were normal. Three patients had better hemodynamic data in the balloon-access limb than in the contralateral limb at follow-up, suggesting that the surgery may have been salutary. We conclude that placement of the IABCD via the common femoral artery does not alter the long-term function of the ipsilateral limb if perioperative problems are avoided or handled appropriately.

Delayed gastric emptying after vagotomy and drainage in the spinal cord injury patient.

In patients with pre-existing spinal cord injuries, truncal vagotomy and drainage procedures resulted in an unacceptable incidence of prolonged gastric stasis. After gastric resection without vagotomy or closure of perforated ulcers, spinal cord injury patients resumed normal gastrointestinal function within 10 days. After vagotomy, three patients were unable to eat for 25 to 57 days. A fourth patient with large gastric residual volumes died of aspiration on the 13th day after vagotomy.

Micturitional static urethral pressure profile: a method of recording urethral pressure profile during voiding and the implications.

Using a tri-lumen graduated 10F catheter we attempted to record static (lateral) pressures at successive points in the urethra, synchronous with intravesical pressure recorded during voiding. Based on our understanding of the essential physical principles described by many earlier investigators we attempted numerous studies in a predominantly male population, which included normal male and female subjects, and male subjects with bladder outlet obstruction of varied etiology. We also attempted to evaluate the effects of the Valsalva maneuver, augmenting the detrusor contraction and artificial distal obstruction (penile compression) on the micturitional static urethral pressure profiles. Studies also were performed to compare the static urethral pressure profiles obtained with the anterograde (catheter withdrawal) and with the retrograde (catheter insertion during voiding) techniques. All of these studies have helped in understanding the various factors, such as double obstructions, artifactual catheter obstructions and incompetent distal sphincter mechanisms, in the interpretation of the micturitional static urethral pressure profile.

Intramural PCO2: a reliable index of the severity of myocardial ischemic injury.

The present study was performed to evaluate the usefulness of changes in intramural oxygen (PmO2) and carbon dioxide (PmCO2) tensions shortly after coronary artery occlusion as indices of the severity of myocardial ischemic injury. In 14 open-chest, anesthetized dogs, a 60-min coronary artery occlusion was performed, during which PmO2 and PMCO2 were measured continuously with a mass spectrometer. Regional myocardial blood flow (RMBF) adjacent to the mass spectrometer probes was measured by the xenon-127 washout technique both before and 30 min after coronary artery occlusion. At the end of 60 min of occlusion, the dogs were killed, and biopsies for histological examination of 1-micron-thick sections were obtained from the tissue surrounding each mass spectometer probe. The decline in PmO2 during the 60-min occlusion bore no relationship either to the severity of ischemic injury as assessed by histological examination, or to the reduction of RMBF. In contrast, the magnitude of rise in PmCO2 during the 60 min of occlusion corresponded closely to both the severity of injury assessed histologically and the reduction of RMBF.

Assessment of the efficacy of interventions to limit ischemic injury by direct measurement of intramural carbon dioxide tension after coronary artery occlusion in the dog.

Although numerous interventions have been shown to exert a salutary effect on the ischemic myocardium, the severity of ischemia generally has been measured by indirect techniques. In the present investigation the effect of ischemia on intramural carbon dioxide tension (PmCO(2)) was measured directly in the open-chest, anesthetized dog with a mass spectrometer during repetitive 10-min coronary artery occlusions separated by 45-min periods of reflow; simultaneously, regional myocardial blood flow in the ischemic area was measured by (127)Xenon washout. In all dogs the increase in PmCO(2) from before to 10 min after the first occlusion (DeltaPmCO(2)) exceeded that during subsequent occlusions. In those dogs not receiving an intervention (controls), DeltaPmCO(2) during the third occlusion was similar to that during the second occlusion. When propranolol, hyaluronidase, and nitroglycerin were administered to different groups of dogs before the third occlusion, each caused significantly smaller elevations in DeltaPmCO(2) than those occurring during the control second occlusion, and the combination of all three interventions induced the smallest increase in DeltaPmCO(2). Regional myocardial blood flow rose with hyaluronidase and was unchanged with propranolol, nitroglycerin, and the three drugs in combination. In contrast to these beneficial interventions, isoproterenol infused with the third occlusion caused a higher DeltaPmCO(2) than during the control second occlusion. It is concluded, first, that interventions that modify the severity of ischemia can be evaluated by measuring intramural carbon dioxide tension; second, that propranolol, hyaluronidase, and nitroglycerin reduce ischemic injury, whereas isoproterenol increases it; and third, that the combination of propranolol, hyaluronidase, and nitroglycerin exerts an additive beneficial effect on ischemia.