Pericardial Decompression

Transcription

1 Pericardial Decompression T. Z. Lajos, M.D., H. E. Black, M.D., R. G. Cooper, M.D., and J. Wanka, M.D. ABSTRACT Pericardial tamponade and chronic pericardial effusion were treated in 32 patients by creating a subxiphoid pericardial window under local anesthesia in preference to pericardiocentesis or pericardiectomy. Chest roentgenograms, fluoroscopy, and cardiac catheterization as standard clinical methods of diagnosis have been largely replaced by echocardiography as the most sensitive method for detecting pericardial effusion. Eleven patients had pericarditis of viral, uremic, or purulent origin. Two had intrapericardial hemorrhage following catheter perforation of the heart. Four had pericardial effusion associated with cardiomyopathy and rheumatoid arthritis. In the remaining 15 patients malignancy was the cause of tamponade. In 10 patients we attempted to control the effusion initially with pericardiocentesis. Five of these (50%) required an additional procedure. In 22 patients subxiphoid pericardial decompression was the primary method of treatment; there were no fatalities during or immediately following operation. None of these patients has had any recurrence of tamponade or effusion for up to three years. ericardiocentesis has been the accepted method for diagnosis and treatment of chronic pericardial effusion or tamponade for years [l]. Primal Certain limitations of this method have been borne out recently. -ily because of its lack of relative safety, diagnostic accuracy, and therapeutic benefits [2], it has been replaced by direct pericardial decompression in our institute. Our past experience has stimulated us to use the subxiphoid pericardial approach routinely for decompressing pericardial tamponade of differing etiology. Diagnosis Accumulation of fluid within the pericardium in our patients was usually of gradual development. It often appeared as a medical emergency in a previously stable patient. Pericardial pain may be inconspicuous, and increasing dyspnea may not be evident since many of these patients are chronically ill and on constant bed rest. Physical examination generally demonstrates significant abnormalities in the internal jugular venous pulse when it is inspected in a good light at 45 degrees of inclination. Prominent a and v waves have both been noted. We routinely place emphasis on the inspiratory loss of pressure as recorded by a sphygmomanometer cuff with auscultation over the brachial artery. Excluding patients with chronic obstructive lung disease and those with arrhythmias such as atrial fibrillation, this physical sign has been reliably diagnostic. A fall of more than 15 mm Hg From the Departments of Surgery and Medicine, State University of New York at Buffalo and the Buffalo General Hospital, Buffalo, N.Y. Accepted for publication June 24, Address reprint requests to Dr. Lajos, 100 Higll St., Buffalo, N.Y VOL. 19, NO. 1, JANUARY,

2 LAJOS ET AL. in systolic pressure with inspiration generally indicated a very noticeable collection of pericardial fluid that was producing restriction or tamponade. Distant heart sounds have been present in patients with a fluid collection of approximately 700 ml. Although the chest roentgenogram may frequently show a fairly typical cardiac silhouette, it does not distinguish adequately between patients having isolated pericardial disease and those with cardiac decompensation and cardiomyopathy. The value of echocardiography in the diagnosis of pericardial effusion has been well documented [4, 51, and during the past two years it has become the definitive investigative tool in the group of patients under discussion. This technique has the advantage of being a bedside procedure and does not require the transportation of sick patients from an intensive care setting. The procedure takes only a few minutes and becomes technically easier with increasing size of the effusion. Echocardiography was performed on 16 of these patients using a commercially available machine (Unirad), a 2.25 mhz transducer with a repetition rate of 1,000 per second. Patients were examined in whatever position was convenient, but preferably pictures were recorded in both the supine and left lateral positions. The mitral valve is identified first, and then the transducer is aimed superiorly and medially to outline the aortic root, left atrium, and left atrial posterior wall. The transducer is then directed inferiorly and laterally until the mitral valve echoes become faint; at this time clear delineation of the posterior wall indicates the endocardium, myocardium, and epicardium of the left ventricle. In all our patients it was relatively easy to demonstrate an echo-free fluid-filled space posterior to the epicardial surface with outlining of the parietal pericardium (Figure). Methods Since 1968 we have used basically two methods for decompressing the pericardium. All patients who had mediastinal tamponade following openheart operations were excluded from this series. With the first method, subxiphoid pericardiocentesis was performed for pericardial tamponade with a large-bore e%-inch-long needle or a 5-inchlong plastic needle subsequently connected to a Pleurevac suction for 24- to 48-hour emptying of the pericardium. Two of 10 patients in whom this method was used required subsequent pericardiocentesis, and 3 needed subxiphoid decompression. The method has definite disadvantages, since the needle may stimulate ventricular extrasystoles by touching the epicardial surface, puncture the thin wall of the right ventricle, or injure the coronary vessels. We have therefore used this method only in acute emergencies. The second method consisted of primary subxiphoid pericardial decompression performed by operative intervention under local anesthesia. It was done in 22 patients, in 3 as a secondary procedure. The skin, subcutaneous tissue, and rectus sheath were infiltrated with 60 to 80 ml of 0.5% Xylocaine (lidocaine), and an incision was made over the xiphoid. 48 THE ANNALS OF THORACIC SURGERY

3 Pericurdial Decompression Echocardiogram of 16-year-old boy with diagnosis of reticulum cell sarcoma clearly demonstrates a large anterior and posterior wall enusion. A pericardial window was created and approximately 1,000 ml of fluid was removed. (LA = left atrium; MV = mitral valve; LV = left ventricle.) This was exposed, the muscular and fascia1 attachments were severed, and the xiphoid process was removed by a rongeur. If further exposure was necessary, we also excised the lower part of the sternum with a rongeur. The pericardium was then opened under direct vision with the incision extended laterally to the left and right sides; normally, a 2 x 2-inch section of pericardium was excised and submitted for pathological study. Through this exposure, the pericardium was always well explored, and adherence and areas of possible infiltration could be well outlined. Two No. 28F red rubber catheters were placed in the pericardial space, one anteriorly along the right atrium and one posteriorly in the most dependent part of the pericardial sac behind the left ventricle. These were connected to a Pleurevac unit with 25 cm HzO suction. The incision was then closed in layers and a regular dry dressing applied. The tubes were removed between the second and fifth day, depending upon the amount of drainage. While we often obtained up to 800 ml of pericardial fluid immediately, continued drainage over the next 24 hours frequently yielded an additional 2,000 ml of fluid, the amount being related to the etiology of the disease (Table). Results In the past four years 32 patients have undergone pericardial decompression, diagnostic pericardiocentesis, or pericardial biopsy at our institution. Their ages ranged from 16 to 85 years with an average of 53 years. There were 21 women and 11 men. The patients were divided into four groups on the basis of etiology of the pericardial disease (see the Table). Group I included 11 patients, 5 with viral pericarditis, 5 with uremic pericarditis, and 1 with purulent pericarditis and pericardial tamponade. The patients who had viral and uremic pericarditis are long-term survivors, VOL. 19, NO. 1, JANUARY,

4 LAJOS ET AL. PERICARDIAL DECOMPRESSION IN 32 PATIENTS Group No. of Pericardial Effusion No. Etiology Patients (ml) I Pericardi tis Viral 5 Purulent 1 Uremic ,600 I1 Pericardial hemorrhage I11 Malignant pericardial effusion ,200 Carcinoma of breast 9 Carcinoma of lung 5 Reticulum cell sarcoma 1 Carcinoma of breast & ovaries 1 Carcinoma of breast & lungs 1 IV Rheumatoid arthritis Cardiomyopathy 3 I some living for more than five years without any recurrence of pericardial fluid. The purulent pericarditis in 1 patient developed following pneumococcal pneumonia. The initial procedure had been pericardiocentesis, which was followed by subxiphoid decompression because of her extremely poor condition. She gradually developed pericardial constriction while being prepared for operation with transfusions and albumin and diuretic therapy. She required a total pericardiectomy and had an uneventful recovery. Group I1 consisted of 2 patients who had tamponade with intrapericardial hemorrhage following cardiac catheterization and perforation by a pacemaker wire. Decompression was performed in 1 with a large-bore needle and in the other with a plastic needle left in the pericardial cavity. Both were discharged from the hospital following decompression. The 15 patients in group I11 underwent pericardial decompression for malignant pericardial effusions producing tamponade. Carcinoma of the breast and of the lung were the most common primary sites associated with tamponade. Subxiphoid pericardial decompression was the definitive procedure in these patients, and there were no recurrences of fluid in the follow-up period of up to three years. One patient with extensive reticulum cell sarcoma went on to develop diffuse tumor invasion of the heart and died four weeks after decompression. Group IV consisted of 4 patients, 1 with rheumatoid arthritis and 3 with cardiomyopath y. The 1 death (3.1%) associated with a pericardiocentesis in these 32 patients stimulated us to change our initial approach, and we now perform the subxiphoid pericardial decompression in the operating room under controlled circumstances once tamponade has been diagnosed. Another 4 patients (12.7%) died in the hospital due to progression of neoplastic disease or to pulmonary embolus. Twenty-seven patients (84y0) were discharged from the hospital with no evidence of recurrent pericardial effusion. We have never yet had a patient who had recurrent effusion. The one additional 50 THE ANNALS OF THORACIC SURGERY

5 Perical-dial Decompression procedure, pericardiectomy, had to be performed for constriction, not for effusion. Pericardial decompression with pericardiocentesis was ineffective in 5 of 10 patients (lioyo) who required additional procedures. Comment In cardiac tamponade the heart is limited during diastole by increased pressure from the pericardium or from fluid or blood filling the pericardial space. In acute injuries a relatively small amount of blood-150 to 250 ml-may be sufficient to cause tamponade, as in our 2 patients with perforation of the heart. Immediate needle aspiration still has a place in these cases. In patients with chronic effusion, however, the pericardium becomes stretched over a period of time and may develop the capacity to contain several liters of fluid with minimal cardiodynamic effect. A large heart shadow on chest roentgenogram presents a problem in differential diagnosis between cardiomyopathy and pericardial effusion. Prior to this cardiac output is minimally reduced, but when this stage is reached the addition of a small volume may reduce cardiac output much further. Twenty-eight patients in our series reached this stage, but immediate operation prevented death due to tamponade. Treatment may be equally dramatic when removal of a small volume of blood or fluid allows a return of blood pressure and cardiac output to near-normal ranges. As pressure in the pericardial sac increases, compensatory mechanisms attempt to maintain circulatory dynamics. Increased sympathetic activity causes vasoconstriction, which tends to maintain systemic arterial pressure. The heart rate increases, and systolic ejection becomes more vigorous. Coronary blood flow may be affected because of the reduction in cardiac output and the pressure gradient between the aorta and the coronary circulation. Clinical shock is evident in tamponade. The patient s skin is cool and moist, heart sounds are distant, the pulse is rapid, and blood pressure may be normal or low. The striking clinical manifestation is venous distention at a time when other signs suggest peripheral circulatory failure similar to that in hemorrhagic shock. Cyanosis may be present as a result of venous stasis. Treatment of tamponade must not be delayed, as it is a life-threatening situation. In our series, patients who had tamponade belonged to groups I to 111, i.e., those with traumatic perforation of the heart, viral and uremic pericardial disease, and malignant pericardial effusion with acute exacerbation. Those patients with chronic pericardial effusion were in group IV and had cardiomyopathy and rheumatoid arthritis. There are three basic anatomical approaches to the pericdrdium when performing a limited pericardiotomy: (1) an anterior approach through the triangle of safety (Voinitsch) [IS]; (2) the transpleural or retropleural posterior approach [lo]; and (3) an inferior approach or subxiphoid pericardial window, which was described by Larrey [8] in 1829 and

6 FRIEDMAN AND TURINA popularized in 1948 by Willius and Dry [14]. The disadvantages of direct pericardiocentesis have induced us to adopt the inferior pericardial approach whenever possible for diagnostic and therapeutic purposes in the operating room setting. When performed in the presence of a grossly bloody malignant pericardial effusion, pericardiocentesis is dangerous. It is difficult to determine whether the needle is in the cardiac cavity and is aspirating blood or pericardial effusion, even if hemoglobin studies, blood gas, or clotting studies are performed on the aspirated fluid. Since these patients are seriously ill, we do not believe a major operation with extensive pericardiectomy is indicated. In a comparable group of patients [3] pericardiectomy had 12.5% operative mortality, while ours was only 3.1%. It is our experience that with an inferior pericardial approach performed under local anesthesia there is minimal danger to the patient [6, 121. By draining the pericardial cavity completely with two No. 28 drainage tubes, one posteriorly and one along the right atrium, approximation of the visceral and parietal pericardium is achieved with ease, thus preventing the recurrence of bloody or serosanguineous fluid. The instillation of an appropriate chemotherapeutic agent will exert its effect through its cytotoxic properties. None of our patients have developed recurrence, and only 1 had a subsequent constrictive pericarditis due to purulent effusion. C hem o t h era!? e u t i c Considerations In the 15 patients with malignant pericardial effusion we utilized the chemotherapeutic agent that would have been anticipated to offer some degree of control of the basic malignant process [9]. Since there is no evidence that instilling chemotherapeutic agents into an involved area is more effective than systemically administering these agents for the particular malignancy in question, we utilized 5-fluorouracil to treat pericardial tamponade in 11 patients (see the Table) who had carcinoma of the breast in whom we expected the primary carcinoma to respond to the agent. Using this compound avoids the problem of adhesive pericarditis, which might be induced with Atabrine or occasionally with nitrogen mustard therapy. When locally instilled into the pericardial sac, 5-fluorouracil has been shown to be retained in an active form for 48 hours, a significantly longer period of time than when it is administered intravenously. Thus an improved effect might be anticipated such as one sees with prolonged infusion of this compound into specific organ systems. In 4 patients with primary squamous cell carcinoma of the lung or reticulum cell sarcoma, however, we continued to use nitrogen mustard, which at a dose of 25 mg is not only topically cytotoxic but will indeed produce adhesive pericarditis. A major disadvantage of nitrogen mustard is its topical irritation, which is almost always associated with reaccumulation of a significant amount of fluid in the 24 hours following instillation, thus 52 THE ANNALS OF THORACIC SURGERY

7 J. Pericardial Decompression necessitating further removal of fluid. In addition, this dose of nitrogen mustard is always associated with systemic hematologic toxicity, a side-effect that occurs to an insignificant degree when 5-fluorouracil is used in the pericardial sac. We have avoided any cytotoxic compounds in the treatment of nonmalignant pericardial disease, of course, instilling a solution of 50% dextrose in 15 patients instead, and most recently utilizing tetracycline in 2 patients without recurrence of fluid [ 111. We have not employed talcum in the treatment of effusions, as has been described by others [7]. With the advancement of technology and more elaborate monitoring equipment in the operating room, we strongly support the view that the pericardium should be approached through a subxiphoid incision for diagnostic and therapeutic purposes. The procedure has no associated mortality and minimum morbidity, and appropriate measures can be taken concomitantly to combat the etiology of the disease. With this minor procedure maximum benefits are achieved. Ref ercinces 1. Blalock, A., and Ravitch, M. M. Consideration of nonoperative treatment of cardiac tamponade resulting from wounds of heart. Surgery 14:157, Cassell, P., and Cullum, P. The management of cardiac tamponade: Drainage of pericardial effusion. Br J Surg 54:620, Dean, R. H., Killen, D. A., Daniel, R. A., and Collins, H. A. Experience with pericardiectomy. Ann Thorac Surg 15: 378, Feigenbaum, H. Clinical applications of echocardiography. Prog Cardiouasc Dis 14:531, Feigenbaum, H., Zaky, A., and Walclhausen, 1. A. Use of ultrasound in the diagnosis of pericardial effusion. Ann Intern Met1 65:443, Fontenelle, L. J., Cuello, L., and Dooley, B. N. Subxiphoid pericardial window. Am J Surg 120:679, Goldman, B. S., and Pearson, F. G. Malignant pericardial effusion. Can 1 Surg 8: 157, Larrey, D. J. Clin Chir 36:393, Lokich, J. I The management of malignant pericardial effusion. JAMA 224: 1401, Loucks, H. H. Suppurative pericarditis: Keport of two cases drained by the posterior route. Arch Surg 18:852, Rubinson, R., and Bolooki, H. Intrapleural tetracycline for control of malignant pleural effusion: A preliminary report. South Med J 65:817, Schlein, E. M., Bartley, T. D., Spooner, G. R., and Cade, R. A simplified surgical approach to therapy of uremic pericarditis with tamponade. Ann Thorac Surg 10:548, Shipley, A. M. Operative approach to heart and pericardium. Surg Gyn~co2 Obrtet 54:280, Willius, F. A., and Dry, T. J. A Hzytory of thc Heart and Circulation. Philadelphia: Saunders, 1948.