Thrombolysis: Go with the Flow

editorials VOLUME 99 / NUMBER 3 / MARCH, 1991

Thrombolysis: Go with the Flow The experimental study by Prewitt and colleagues (see page 70S) is the first to highlight the relationship between systemic blood flow and thrombolysis. In a canine model of pulmonary embolism, either pretreatment with hydralazine or opening a systemic arteriovenous fistula led to >50 percent increase in cardiac output compared with controls, and resulted in a parallel (>50 percent) increase in the rate of thrombolysis. In earlier experiments these investigators had shown hydralazine to lessen pulmonary vascular resistance and hypothesized that this effect would accelerate pulmonary artery thrombolysis. IThe equivalent results in the present study using an arteriovenous fistula, without attendant decrease in pulmonary artery pressure, strongly points to flow per se as a key determinant of clot dissolution therapy. These findings have important implications for clinical management of cardiogenic shock and future investigation of adjunctive pharmacotherapy to thrombolysis. CARJ)JOCENIC SHOCK

In the setting of acute, massive myocardial injury or necrosis, the development of power failure and hypotension can be anticipated. Favorable interruption of the natural history of cardiogenic shock with a mortality in excess of SO percent in historical controls has been suggested by recent series of patients undergoing direct coronary angioplasty.2-4 However, in patients with cardiogenic shock treated with intravenous thrombolysis, the results thus far have been disappointing. In the Gruppo Italiano per 10 Studio della Streptochinasi (GISSI) trial, there was a 70 percent 30-day mortality for patients in Killip class 4 at time of presentation, irrespective of whether streptokinase or conventional therapy was given. ~ Similarly, with intracoronary streptokinase in the multicenter registry of the Society for Cardiac Angiography, a 67 percent in-hospital mortality was reported for patients with cardiogenic shock." In light of the current study by Prewitt et aI, a likely explanation for apparent failure of therapy is hypoperfusion. In the future, new approaches to thrombolysis in the setting of cardiogenic shock may do well to incorporate early intervention to maximize cardiac output. To this end, use of either intra-aortic balloon counterpulsation 7 or per-

cutaneous cardiopulmonary bypass has, in small series of patients, been accompanied by relatively favorable outcome. DRUG-DRUG INTERACTION

For any drug that changes systemic or regional blood flow, a careful reassessment will be necessary before coupling the agent with thrombolytic therapy. As cited by Prewitt et al, an example which we did not anticipate was the negative interaction between tissue plasminogen activator and the prostacyclin analog I\oprost. In a pilot clinical trial we found very low infarct vessel patency (44 percent) for these two agents given concomitantly9 and only later did experimental studies confirm that I\oprost has the potential to decrease the plasma tPA level 30-40 percent, likely mediated via markedly increased hepatic blood flow and clearance. 11'11 It is possible that this negative interaction is specific for the agents tested; a pilot study with streptokinase and prostaglandin E 1 suggested improved thrombolytic efficacy. 12 On the other hand, Rentrop et a)l3 have demonstrated that nitroglycerin given with intracoronary streptokinase improves left ventricular function beyond the level that each agent achieved as monotherapy. The mechanism for this interaction is not established, but may be due to nitroglycerin's amelioration of collateral blood supply, decreased afterload, or both. Likewise, the salutary effect for combined intravenous thrombolysis and captopril or thrombolysis with adenosine may, in part, be attributed to vasodilation and more extensive thrombolysis.I4·I~ Ongoing large scale clinical trials are testing the interaction of thrombolytic agents with angiotensin-converting enzyme inhibitors such as the International Studies ofInfarct Survival (ISIS-4) and GISSI-3 trials. The facilitation of thrombolysis mediated by hydralazine verified in the present study suggests this may be a key method by which a vasodilator offers incremental benefit. The study by Prewitt is fundamental to our understanding of how thrombolysis can be modulated. In order to accelerate infarct vessel recanalization in myocardial infarction, improvement in systemic blood flow and actual delivery of the plasminogen activator to the site of occlusion will be necessary. That thrombolysis in essence goes "with the flow" represents a significant, yet up until now neglected, future direcCHEST I 99 I 3 I MARCH. 1991

529

tion of investigational effort.

Eric J Topol, M.D.

Ann Arbor

Division of Cardiology, Department of Internal Medicine, University of Michigan Medical School. Reprint requests: Dr. 1bpo1, University Hospital, 1500 East Medical Center Drive, Ann Arbor 48109

REFERENCES 1 Ducas J, Duval D, Dasliva H, Boileau ~ Prewitt RM. Treatment of canine pulmonary hypertension: effects of norepinephrine and isoproterenol on pulmonary vascular pressure-flow characteristics. Circulation 1987; 75:235-42 2 Lee L, Bates ER, Pitt B, Walton JA, Laufer N, O'Neill ww. Percutaneous transluminal coronary angioplasty improves survival in acute myocardial infarction complicated by cardiogenic shock. Circulation 1988; 78:1345-51 3 Heuser RR, Maddoux GL, Goss JE, Ramo BW: RaffGL, Shadoff N. Coronary angioplasty for acute mitral regurgitation due to myocardial infarction. Ann Intern Med 1987; 107:852-55 4 Brown TM Jr, Iannone LA, Gordon DF, Wickemeyer WJ, Rough RR. Percutaneous myocardial perfusion (PMR) reduces mortality in acute myocardial infarction (MI) complicated by cardiogenic shock [abstract]. Circulation 1985; 72:111-309 5 Gruppo Italiano per 10 Studio della Streptochinasi nell'lnfarto Miocardico (GISSI). Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Lancet 1986; 1:397402 6 Kennedy JW: Gensini GG, Timmis GC, Maynard C. Acute myocardial infarction treated with intracoronary streptokinase: a report of the society for cardiac angiography. Am J Cardiol 1985; 55:871-77 7 DeWood MA, Notske RN, Hensley GR, Shields J~ O'Grady W~ Spores J, et aI. Intraaortic balloon counterpulsation with and without reperfusion for myocardial infarction shock. Circulation 1980; 61:1105-12 8 Shawl FA, Domanski MJ, Punja S, Hernandez TJ. Emergency percutaneous cardiopulmonary (bypass) in cardiogenic shock. J Am Coli Cardioll989; 13:160A 9 Topol EJ, Ellis SG, Califf RM, George BS, Stump DC, Bates ER, et al. Combined tissue-type plasminogen activator and prostacyclin therapy for acute myocardial infarction. J Am Coli Cardioll989; 14:877-84 10 Kerins DM, Fitzgerald G, Fitzgerald D. The influence of platelet inhibitors on the efficacy and disposition of tissue plasminogen activators [abstract]. Clin Res 1988; 37:4A 11 Nicolini FA, Nichols Mehta JL. Prostacyclin analog, I1oprost, decreases thrombolytic potential of t-PA in canine coronary thrombosis [abstract]. Circulation 1989; 80:11-234 12 Sharma B, Wyeth R~ Gimenez HJ, Franciosa JA. Intracoronary prostaJdandin E. plus streptokinase in acute myocardial infarction. Am J Cardioll986; 58:1161-66 13 Rentrop K~ Feit F, Sherman w: Steey ~ Hosat S, Cohen M, et aI. Late thrombolytic therapy preserves left ventricular function in patients with collateralized total occlusion: primary end point findings of the second Mount Sinai-New York University Reperfusion Trial. J Am Coli Cardioll989; 14:58-64 14 Nabel EG, Topol EJ, Galeana A, Bates ER, Werns SW: Walton JA, et a1. A randomized, double-blind, placebo controlled, pilot trial of combined early intravenous captopril and tPA therapy in acute myocardial infarction. Circulation 1989; 80:11-112 15 Clarke B, Coupe ML. Adenosine: cellular mechanisms, pathophysiological roles and clinical applications. Int J Cardioll989; 23:1-10

ww:

530

Identifying Patients with High Risk

of High Cost

It services has been shown for many types of health care that a relatively small proportion of pa-

tients account for a disproportionately large share of resource use. This pattern appears again in the article by Oye and Bellamy in this issue (see page 685). For intensive care unit (ICU) patients, the high-cost group includes many patients who die. There is concern that large amounts of resources are being spent to treat chronically ill and/or elderly patients with acute illness or massively injured patients to salvage only a small proportion of these patients who are discharged alive from acute care hospitals. In some cases, families may demand this care in a futile attempt to save patients who are beyond help. For some ICU patients, the high cost far outweighs any benefit of these expenditures. Of course, it is one thing to see, ex post, that the medical care has been unsuccessful and quite another to predict outcome or cost in advance. Since it is the medical prognosis that is most helpful in guiding decision making, substantial efforts have been devoted to the development of models to predict the outcome of ICU care}·2 These efforts have met with considerable success in predicting hospital mortality for groups of patients. From a somewhat different direction, interest has developed in the characterization of severity of illness and its relation to resource use. Driven largely by concerns about reimbursement based on diagnosis-related groups, systems have been developed to measure severity of illness so that the existence of resource use variability can be recognized in reimbursement. Recent works, of which the article by Oye and Bellamy is an example, have begun to explore relationships among severity (as measured by predicted outcome), actual outcome, and resource use. Knaus et al2 found that, when severity was held constant, outcome improyed with resource use at low levels of resource use. 3 Other studies have emphasized the need to distinguish between actual outcome and predicted outcome. While it is true that many highcost patients die, some patients who die are not high cost and some high-cost patients do not die. A finding first reported by Detsky et al,4 confirmed by Rapoport et al,s and again reported in this issue by Oye and Bellamy, is that patients who have an outcome that is opposite to the predicted outcome tend to be highcost patients. Two implications of these findings might be noted. First, average resource use does not always rise with severity of illness (as measured by probability of mortality) for ICU patients. As severity of illness increases, average resource use rises and then falls, contrary to the (often implicit) assumption of most Editorials