CLINICAL CONFERENCE IN PULMONARY DISEASE
Recurring Pulmonary Emboli* Clinical Conference in Pulmonary Disease from Northwestern University Medical School, Chicago Richard L. Hughes, M.D., F.C.C.P. Associate Proiessor of Clinical Medicine Stewart Mann, M.D. Fellow, Pulmonary Section Lewis J. Smith, M.D., F.C.C.P. Assistant Professor of Medicine
David Green, M.D. Professor of Medicine and Director of Atherosclerosis Program John Moran, M.D., F.C.C.P. Associate Professor of Surgery James Yao, M.D. Professor of Surgery
Dr. Hughes: Most internists are familiar with the management of acute pulmonary emboli. Our present ability to detect clots and the routine use of continuous heparin have erased many of the uncertainties which used to plague physicians in their attempts to manage this disease. However, relatively few of us have had adequate experience with those rare patients who continue to develop emboli despite adequate anticoagulant therapy. We recently encountered three patients who demonstrated some of the complications associated with
was begun on therapy with the antiplatelet drug, sul6npyrazone (Anturane) , resulting in marked reduction in warfarin requirements. He was admitted to the MICU for a pulmonary hypertension protocol using oxygen, ibuprofen, phentolamine, and hydralazine. Pulmonary pressure was 105/ 40 mm Hg. He responded only to administration of ibuprofen, with a 10 percent fall in pressure. He expired shortly thereafter. Autopsy revealed multiple old and new emboli. The left main pulmonary artery was 80 percent occluded by an organized thrombus. The second patient is a 54-year-old laborer in good health until 1975, when he noted pleuritic chest pain and shortness of breath shortly after a faD on his coccyx. A lung scan demonstrated multiple pulmonary emboli and he was placed on therapy with warfarin. Over the next four years, he had recurrent episodes of pleuritic pain, with perfusion scans compatible with recurrent emboli, despite therapeutic doses of warfarin. In 1980, he noted the gradual onset of bilateral leg edema. He was transferred to this hospital where venous flow studies were consistent with bilateral deep vein thromboses. Pulmonary artery pressure was 47/14 mm Hg, but angiography did not reveal any emboli. Echocardiography of the heart revealed normal chamber size without evidence of thrombi. A venogram revealed complete occlusion of the lYe with large collaterals (Fig 1). His pulmonary hypertension was presumed to be secondary to recurrent pulmonary emboli. He was discharged and continued on warfarin. The third patient is a 73-year-old retired policeman in perfect health until fracturing his left clavicle six months prior to admission. He was placed in an immobilization brace for six weeks. One month later, he had the sudden onset of dyspnea and pleuritic pain. V/ Q scans were compatible with emboli and he was begun on therapy with heparin, to which he showed some resistance, requiring up to 3,000 units/hr. A repeat lung scan showed new defects. He was discharged on warfarin therapy, with a prothrombin time 2l' times control. Eight days later, his prothrombin time was normal, Administration of warfarin was increased to 17.5 mg per day, and subcutaneous heparin 5,000 units bid was added. Three weeks later he was readmitted with
recurring pulmonarv emboli. We will use them to
develop guidelines for anticipating which patients may be refractory to anticoagulant therapy, and how to treat those complications which arise. Dr. Mann: The first patient is a 29-year-old salesman, in good health until 1971. when he fractured his right leg and developed superficial phlebitis. Three months later.. while on therapy with warfarin, he noted the onset of mild dyspnea. Evaluation revealed thrombophlebitis, elevated pulmonary artery pressure and multiple pulmonary emboli. Administration of warfarin was continued and an inferior vena cava (IVC) umbrella inserted. Antithrombin In levels were "decreased." In 1976, he had a pulmonary artery pressure of 95/35 mm Hg, marked bicuspid regurgitation, ascites, and V/0 scans compatible with continuing pulmonary emboli. In 1980, a trial of therapy with tolazoline (Priscoline), an alpha-adrenergic blocking agent, was unsuccessful. Antithrombin III level was 69 percent (normal range: 80-120 percent). Abdominal pain, acute psychotic withdrawal and increasing ascites required frequent hospitalizations. His thromboxane: prostacyclin ratio was 11 ( normal 1-2) . He ·From the Department of Medicine, Divisions of CardioThoracic and Vascular Surgery, and the Atherosclerosis Program, Rehabilitation Institute of Chicago, Northwestern University Medical School, Chicago
230 HUGHES ET AL
CHEST, 81: 2, FEBRUARY, 1982
FIGURE 2. Patient 3. PA radiograph of chest taken on admission. Arrows mark the most medial aspect of the hemothorax, which did not change on decubitus film. The heart is shifted to the left and is probably normal in size. The pleural disease on the left is old, as are the diaphragmatic plaques.
FIGURE 1. Patient 2. Retrograde venogram of IVC demonstrating occlusion (presumably thrombotic) at level of renal veins . Left renal vein was not visualized, but the left kidney appears to be functioning normally . The small vessel (arrow) beneath the IVC is a collateral from the right iliofemoral vein . Ascending venograms demonstrated complete occlusion of the IVC at its biturcation. acute shortness of breath and weakness. His prothrombin time was 2~ times control, PTT 80 seconds, and hemoglobin 9 grams. His chest x-ray film demonstrated a large right pleural effusion (Fig 2), and his warfarin was stopped. He was transferred to this hospital, where a thoracocentesis revealed frank blood. A chest tube was inserted, and warfarin therapy restarted. Venous How studies were normal in both arms and legs . Contrast venography was normal in all extremities except the right calf, where there was evidence of old thrombotic disease. Echocardiography suggested a thrombus in his right ventricle. Following removal of his chest tube, he developed a low grade fever, and failed to expand his right lung. A portion of his 7th rib was removed and open drainage instituted, with successful evacuation of his pleural space, but no expansion of his entrapped rlght lung. Decortication was required six weeks later ( Fig 3 ) .
Dr. Smith: All of today's patients had secondary pulCHEST, 81: 2, FEBRUARY, 1982
monary hypertension. There are four noncardiac mechanisms by which pulmonary hypertension may develop (Table 1). For want of a better heading, primary pulmonary hypertension is listed under obliteration. Currently, it is only in this entity that drug therapy has shown some promise. In normal individuals, the pulmonary vasculature is relatively insensitive to blood flow: increased flow does not produce increased pressure. It is only when 65 percent or more of the pulmonary vascular bed is destroyed that pulmonary artery pressures increase at rest. The factors known to regulate the flow of blood through the lungs are listed in Table 2. The autonomic nervous system, predominantly the sympathetic system, appears to have a relatively small role in modulating flow. Hypoxia produces large increases in pulmonary artery pressure by direct arterial constriction. This is almost certainly the body's major mechanism of matching ventilation and perfusion in the diseased lung. Hypercarbia, in contrast, produces only small increases in Table I-Noncardiac Cauaea 01 Pulmonary Hyper'enaion Classification
Recurrent «mbol i
Emphysema Interstitial disensrSurgery
10 Pulmonary hypertension
RECURRING PULMONARY EMBOLI 231
FIGURE 3. Patient 3. Close-up views of right chest demonstrating resolution of hemothorax. A. Ten days after admission following removal of first chest tube, there is air in the pleural space. Volume of effusion has decreased by about 30 percent. B. One month later, follow ing removal of rib segment and open drainage. Two drainage tubes are in place (arroui). Both pleural surfaces are shaggy, and the right lung has not expanded further. C. Six weeks later, follow ing decortication. The lung is 80 percent expanded, with only a small residual of air outside the thickened surface of the lung (arrow).
pulmonary artery pressure when hydrogen ion concentration remains constant. Acidemia, like hypoxia, produces large increases in pressure and is the major cause of the elevated pulmonary artery pressure that occurs with hypercapnea. Catecholamines have a major role in regulating pulmonary blood How. Pulmonary vessels are liberally supplied with both alpha (vasoconstrictor) and beta (vasodilator) adrenergic receptors. Several lines of evidence strongly suggest that the alpha adrenergic system predominates. 1 Histamine, another potent vasoconstrictor, appears to mediate the hypoxic response, and operates through both the adrenergic and the HI receptors within the lung. The exact role of serotonin remains uncertain. Angiotensin II, unlike oxygen and histamine, constricts the systemic as well as the pulmonary vasculature. Prostaglandins Table 2--Regulation 01 Pulmonary Blood FIOtfl
have profound and opposing effects on most vascular beds in the body. There is some evidence that the ratio of the metabolites of thromboxane and prostacyclin determines the responsiveness of the pulmonary bed to prostaglandin inhibitors (ie, the higher the ratio the better the response). In our first patient, Dr. Watkins at Massachusetts General Hospital found this ratio to be 11 (thromboxane 600 pgm/ml: prostacyclin 55 pgm/rnl ). This was the principal reason ibuprofen was administered during this patient's protocol. The major forms of therapy that have been tried in patients with pulmonary hypertension are listed in Table 3. Each of these drugs has been effective in isolated cases. Unfortunately, there are insufficient data to be able to recommend one over anTable 3--Drug Therapy 01 Pulmonary Hyperteruion· Oxygen
Autonomic nerv es
Sympath et ic
Alpha adrenergic blockers'
Hypoxia Hypercarbia Acidemia
Catecholamines Histamine Serotonin Angiotensin II Pro staglandin
232 HUGHES ET AL
Pro stacyclin (PGI .)" Prostaglandin synthetase inhibitors "None of the drugs listed have proven consistently effective.
CHEST, 81: 2, FEBRUARY, 1982
Table 4--Labora,ory Criteria for In Vivo Tlarombode Aedl1i',.
Circulating endothelial cells Elevated ratio of BTG to PF 4 Increased ratio of VIII :Ag to VIII:C Decreased platelet survival Spontaneous platelet aggregation Increased platelet-collagen affinity
Circulating platelet aggregates Elevated fibrin first derivatives Elevated fibrinopeptide A, FDP Reduced AT-III Increased inhibitors of fibrinolysis
Increased platelet coagulant activity VIJI;Ag==Factor VIII-related antigen; VIII:C==Factor VIII coagulant activity; BTG==Beta Thromboglobulin; PF4== Platelet Factor 4; FDP=fibrin degradation products; ATIII == antithrombin III: see text for details.
other. Most of the beneficial effects have been demonstrated in patients with primary pulmonary hypertension especially those with the more labile, and presumably earlier, stage of the disease. Treatment of the secondary forms of pulmonary hypertension have been largely unsuccessful. Although one patient with recurrent emboli has responded acutely to ibuprofen (D.W. Watkins, personal communication ), there are no longterm studies available, and prostaglandin inhibitors have not been consistently effective in other patients. Dr. Green: For this conference, I want to confine my remarks to the recognition of the hypercoagulable state. Some laboratory tests that are indicative of in vivo thrombotic activity are shown in Table 4. Deep vein thrombosis may arise in vessels whose lining endothelium has been damaged, either by mechanical trauma, immune complex deposition, hypoxia, or some other mechanism. Such endothelial damage may be recognized by observing an increase in circulating endothelial cells, or by finding elevated levels of factor VIII-related antigen. This antigen is synthesized by endothelium and circulates in a 1:1 ratio with factor VIII-coagulant activity. Increases in this ratio are seen in patients with extensive endothelial cell damage and have been reported in patients with deep vein thrombosis. 7 Exposure of subendothelial connective tissue is a potent stimulus for platelet adhesion and aggregatten, The accelerated loss of platelets from the circulation creates a decreased platelet survival time, and reflects the deposition of platelets on the subendothelium. Decreased platelet survival has been documented in subjects with venous thrombosis," Activation of the coagulation cascade may generate trace amounts of thrombin, producing
CHEST, 81: 2, FEBRUARY, 1982
spontaneous platelet aggregation, increased platelet responsiveness to other aggregating agents such as collagen, and exposure of platelet factor 3, which enhances clotting. These activated platelets release the contents of their alpha granules, which include B-thromboglobulin ( B-TG) and platelet factor 4 ( PF. ). The PF. has been a very brief plasma halflife, whereas B-TG circulates longer, resulting in elevated ratios of B-TG to PF•.9 Recently, Rossi et al 7 published the results of a prospective study of patients admitted to the hospital with acute spinal cord injury. Deep vein thrombosis developed in 72 percent of these patients, and coincident with the onset of thrombosis was the appearance of circulating platelet aggregates in the blood Thus, examination of the plasma for the presence of platelet aggregates is helpful in recognizing ongoing thrombosis. Other signs of intravascular coagulation include elevated titers of fibrin degradation products, fibrinopeptides'? and fibrinogen first derivative. 11 A major coagulation deficiency in the first patient was a reduced plasma concentration of antithrombin III. This factor is a major inhibitor of several activated clotting factors including IX, X, XI and thrombin.P His deficiency of antithrombin III could have been on a congenital basis, or secondary to disseminated intravascular coagulation or liver disease. In this case, cardiac cirrhosis secondary to the pulmonary hypertension and right heart failure was probably contributory, Longterm management of suoh patients is with warfarin. His ratios of B-TG to PF4 and VIII-related antigen to VIIIcoagulant were significantly elevated, as was the thromboxane: prostacyclin ratio. Therefore, sulfinpyrazone, which inhibits prostaglandin synthesis, was added to his drug regimen. Because this drug potentiates the action of warfarin.P the dose of the latter drug was reduced. A Physician: Which tests are most useful in establishing a hypercoagulable state, and do you consider a shortened PTT of any value?
Dr. Green: For routine testing, we would recommend antithrombin III levels, B-TG:PF4 ratio, and the 'platelet aggregate ratio. A shortened PTT usually reflects increased levels of coagulant factor VIII. Dr. Hughes: Should these tests be ordered on every patient who develops venous thromboses, and could you comment on subcutaneous heparin in patients who might be hypercoagulable?
RECURRING PULMONARY EMBOLI 233
Dr. Green: I do not think every patient should be tested for hypercoagulability. Obviously, any patient who develops recurrent pulmonary emboli while on anticoagulant therapy deserves investigation. Any patient without an obvious cause for his venous thrombosis, such as an injury, prolonged stasis, oral contraceptives, etc, should be studied with these techniques. Mini-dose heparin is very helpful in patients who are undergoing surgical procedures, the elderly with heart failure, or those on prolonged bed rest. It may help in preventing hypercoagulabiIity in acute illness. For longterm management, warfarin still seems to be the best, since there is some evidence that continuous heparin in conventional doses may actually predispose to recurrent emboli.v' Dr. Moran: I will confine my remarks to two rare complications of pulmonary emboli-hemothorax and a saddle pulmonary embolus producing cardiovascular collapse. Hemothorax is seen most often with trauma, and the enormous accumulation of blood seen in the third patient is really quite unusual. By the time we first saw this patient, about one-half of his thorax was obliterated, and his lung had already become entrapped by a fibrous peel, which we knew because his x-ray Blm findings did not change in the decubitus position. Blood is a hypertonic irritant, and causes increasing effusion once it is present in the pleural space. If left untreated, approximately 15-20 percent of hemothoraces will need decortication to free the lung. Hypoxemia, respiratory distress, and hypovolemic shock may complicate acute hemothorax. All patients with a signi6cant hemothorax should have the fluid evacuated as completely as possible, usually with a chest tube. One can treat a small hemothorax with observation, but it is absolutely necessary to drain a large one. One problem the physician faces is differentiating a bloody effusion from a hemothorax. I consider a significant hemothorax one that occupies three or more interspaces and has a hematocrit above 10. The cut-ofI point depends a little upon when you collect the Huid, since blood is diluted as soon as it enters the intrapleural space. Whether you call fluid with a hematocrit less than 5 a pleural effusion or a hemothorax is mostly a question of semantics, and requires your clinical assessment. Our third patient clearly had a hemothorax, regardless of the hematocrit of the fluid. In contrast to hemothorax, a bloodtinged effusion is quite benign in the intrapleural space. It is still necessary occasionally to remove massive pulmonary emboli surgically. The advent of streptokinase and urokinase has considerably reduced the indications for surgical embolectomy, but there are
234 HUGHES ET AL
still some patients who will require this intervention. Patients who develop a contraindication to anticoagulation or thrombolytic therapy, patients who are pregnant, and those who are dying from a falling cardiac output should be operated on. An embolectomy is one of the simplest operations that cardiac surgeons perform; it is quick and highly therapeutic. To prevent recurrence, the sternotomy incision should be extended to the umbilicus, the duodenum reflected and a clip placed on the Ne. This introduces Dr. Yao's role, which I consider very important in the management of recurrent pulmonary emboli. Dr. Y00: I would like to describe how we approach patients with recurrent pulmonary emboli and what our experience has been in their management. First, it is essential that the diagnosis of pulmonary emboli be confirmed before vena caval interruption. This is best done by pulmonary angiogram. We have found ventilation-perfusion scans to be unreliable. At this institution, 10 percent of patients (six of 62) who were considered for vena cava interruption because of recurrent emboli had abnormal lung scan findings, but normal pulmonary arteriograms. Surgical intervention was therefore abandoned in these patients and all remain in good health. The second patient discussed today was managed this way because of his normal pulmonary vessels. We do not interrupt the vena cava unless pulmonary embolism is confirmed by pulmonary angiography. This policy is supported by the experience of others. IS I should stress that the timing of the angiogram is critical. If it is performed within 24 to 72 hours of the acute episode, the angiogram will accurately detect a blood clot if it is present. If pulmonary arteriography is performed after 72 hours, the incidence of false-negatives increases. In most patients, an ascending venogram should also be performed to determine the source of the emboli. There are a number of ways to interrupt the vena cava. The most widely used is the so-called transvenous approach through the jugular vein. This approach avoids the need for general anesthesia. At present, we favor the use of the Greenfield filter. This device alIows maximal blood How through the filter with less tendency to develop total occlusion. Finally, I would like to emphasize that patients who have recurrent emboli after caval interruption must have venograms. If ascending venography fails to demonstrate the clot, a retrograde venogram via the axillary route should be done and this technique may detect clot above the clip (Fig 4). Surgical removal of the clot followed by ligation of the vena cava above the clip is necessary to eliminate the source of recurrent emboli.
CHEST, 81: 2, FEBRUARY, 1982
4. Venogr am of a clot occurring above an IVC clip . Diagram at right indicates extent of clot (arrows). This patient had a complete ligation and no further symptoms of pulmonary emboli . FIGURE
A Physician: How would you manage the patient who already has an Ive filter in place, who continues to have episodes suggestive of pulmonary emboli, but whose pulmonary angiogram is negative? Dr. Yao: If the pulmonary angiogram is negative, the patient probably is not having significant emboli and there is no need to completely interrupt the vena cava. However, if you cannot anticoagulate the patient, then complete interruption might be considered.
bolytic therapy may be preferable to anticoagulant because residual swelling and pain are common sequelae in axillary vein thromboses. This is especially true if swelling and pain occur acutely in the dominant hand. Dr. Green : I agree with Dr. Kwaan's last comment regarding thrombolytic therapy. The most common causes of axillary vein thromboses are trauma and malignancy elsewhere in the body. : We would like to thank Dr. W. Dav id Watkins, Massachusetts General Hospital. for the generous use of his laboratory and time. ACKNOWLEDG~1ENT
A Physician : How often do you see significant pulmonary emboli arising in the veins of the upper extremities? Dr. Yao: Approximately 10 percent of the time .!" Dr. Kwaan, would you like to make a comment? Dr. Kwaan: Yes, two comments. The first is to emphasize Dr. Moran's approach to the dying patient with a pulmonary embolus. Even if you can institute thrombolytic therapy, there is a lag period of two hours before thrombolysis begins, which may prove fatal to the patient. The second comment concerns upper extremity thromboses. Again, I agree that you should see the pulmonary embolus before instituting therapy. In this instance, throm-
CHEST, 81: 2, FEBRUARY, 1982
2 3 4 5 6
Bergofsky EH. Active control of the normal pulmonary circulation. In: Moser KM, ed . Pulmonary vascular disease. New York: Marcel Dekker, 1979 Ruskin JH, Hutter AM Jr. Primary pulmonary hypertension treated with oral phentolamine. Ann Intern Med 1979; 90:772-74 Daoud FS, Reeves JT, Kelly DB. Isoproterenol as a potential pulmonary vasodilator in primary pulmonary hypertension. Am J Cardiol 1978; 42 :817-22 Rubin LJ, Peter RH. Oral hydralazine therapy for primary pulmonary hypertension. N Engl J Med 1980; 3<69-73 klinke="" wp="" gilbert="" jal.="" diazoxide="" in="" primary="" pulmonary="" hypertension.="" n="" engl="" j="" med="" watkins="" wd="" peterson="" mb="" crone="" rk="" shannon="" dc=""/>
RECURRING PULMONARY EMBOLI 235
Levine L. Prostacyclin and prostaglandin E 1 for severe idiopathic pulmonary artery hypertension. Lancet 1980; 1:1083 Rossi EC, Green D, Rosen JS, Spies SM, Yao JST. Sequential changes in factor VIII and platelets preceding deep vein thrombosis in patients with spinal cord injury. Br J Haemat 1980; 45:143-51 Harker LA, Slichter SJ. Platelet and fibrinogen consumption in man. N Eng} J Med 1972; 287:999-1005 Kaplan KL, Owen J. Plasma levels of B-thromboglobulin and platelet factor 4 as indices of platelet activation in vivo. Blood 1981; 57: 199-202 Cronlund M, Hardin J, Burton J, Lee L, Haber E, Bloch KJ. Fibrinopeptide A in plasma of nonnaI subjects and patients with disseminated intravascular coagulation and systemic lupus erythematosus. J Coo Invest 1976; 58: 142-52
11 Alkjaersig N, Fletcher A, Joist H, Chaplin H Jr. Hemostatic alterations accompanying sickle cell pain crises. J Lab Clin Med 1976; 8:440-49 12 Rosenberg RD. Actions and interactions of anti-thrombin and heparin. N Eng! J Med 1975; 2~2:146-51 13 Weiss M. Potentiation of coumarin effect by sulfinpyrazone. Lancet 1979; 1:609 14 Wilson JE, Bynum LJ, Parken RW. Heparin therapy in venous thromboembolism. Am J Med 1981; 70:808-16 15 Novelline RA, Baltarowich OH, Athanasoulis CA, WaItman AC, Greenfield AJ, McKusick KA. The clinical course of patients with suspected pulmonary embolism and a negative pulmonary arteriogram. Radiology 1978; 126:561-67 16 Adams JT, DeWeese JA, Mahoney EB, Rob CG. Intermittent subclavian vein obstruction without thrombus. Surgery 1978; 63:147-65
Radiologic Approaches to the Chest and Abdomen The University of Wisconsin-Extension; Department of Continuing Medical Education; and University of Wisconsin School of Medicine, Department of Surgery, will present this course April 28-30 in Madison. For information, contact Ms. Sarah Z. Aslakson, 465B WARF Building, 610 Walnut Street, Madison 53706 (608:263-2856).
Cardiac Interventional Techniques and DiagnosticImaging This course will be held April 5-8 at the Barbizon Plaza Hotel, New York City, sponsored by the North American Society for Cardiac Radiology. For information, address: Dr. Thomas A. 80S, c/o Educational Resources Associates, Inc., 209 Harvard Street, Brookline, Massachusetts 02146 (617:738-8859).
NewApproaches in the Diagnosis and Management of Ischemic Heart Disease The Towsley Center for Continuing Medical Education, University of Michigan Medical School, will present this program on March 24 in Ann Arbor. For information, contact Ms. Wendy L. Frisch, Towsley Center for Continuing Medical Education, University of Michigan Medical School, Ann Arbor 48109 (313:764-2287).
23& HUIHES ET AL
CHEST, 81: 2, FEBRUARY, 1982