Thrombolytic Therapy in Canine Pulmonary Embolism Comparative Effects of Urokinase and Recombinant Tissue Plasminogen Activator 1 - 3

RICHARD M. PREWITT,4 CONRAD HOY, ANN KONG, SHIAN GU, DAVID GREENBERG, ROBERT COOK, S. MING CHAN, and JOHN DUCAS

Introduction Several basic (1-4) and clinical (5, 6)

studies have demonstrated the efficacy of recombinant tissue plasminogen activator (rtPA) in venous thromboembolism. A recent canine study of embolic pulmonary hypertension compared thrombolytic efficacy of 1 mg/kg of rtPA infused over 15 (rtPA 1 S) and 90 min (1). During infusion, rate of pulmonary thrombolysis and rate of decrease of pulmonary artery pressure (Ppa) were greatest with rtPA 1S ' However, the differences in thrombolytic rate did not correspond to the difference in drug concentration. Similarly, another recent study (2) of canine pulmonary embolism demonstrated that, during drug infusion, rates of pulmonary thrombolysis were similar when 1 mg/kg of rtPA was infused over 5 or 15min. These data suggest an upper limit to the dosethrombolytic rate relationship with rtPA. One randomized, controlled clinical trial compared an investigational dose of rtPA (100 mg over 2 h) and an FDA-approved dose of urokinase (initial bolus of 4,400U/kg followed by 4,400U/kg/h) in treatment of pulmonary embolism (7). At 2 h, thrombolysis and hemodynamic improvement were greatest with rtPA. The investigators concluded that at the dose regimes employed, rtPA acts more rapidly and is safer than urokinase (UK). Conceivably, the more rapid initial rate of thrombolysis with rtPA was due to the relatively low dose of UK employed. The current study was designed to compare rate of pulmonary thrombolysis and corresponding pulmonary hemodynamic effects of rtPA and UK. Two doses of each drug were employed. UK was given over 15 min in doses of 30,000 and 60,000 U/kg. This design allowed us to directly compare, for the first time, two "adequate" doses of UK. Two dosing regimes of rtPA were also employed, 1 and 2 mg/kg over 15 min. Our previous studies demonstrated that the form290

SUMMARY We compared thrombolytic and pulmonary hemodynamic effects of recombinant tissue plasminogen activator (rtPA) and urokinase (UK) In canine mlcropulmonary thromboembolism. Dogs were embolized with radioactive autologous blood clot to Increase mean pulmonary artery pressure (from 13 to 34 mm Hg, p < 0.005) and decrease cardiac output (2.5 to 1.6 L min, p < 0.005). Four groups of six dogs were treated. We employed two doses of UK, 30,000 U/kg (UK 3o) and 60,000 U/kg (UKeo) , and two doses of rtPA, 1 mg/kg (rtPA 1) and 2 mg/kg (rtPA 2) . Drugs were infused over 15 min. Rate and extent of pulmonary thrombolysis were assessed by continuously counting over both lung fields with a gamma camera. Compared with treatment with UK, both rtPA regimes significantly increased thrombolysis. Mean total pulmonary thrombolysis was 14 and 23% with UK3 0 and UKeo, respectively, and 35 and 43% with rtPA 1 and rtPA 2 • Corresponding to the increased thrombolysis, pulmonary hemodynamics Improved most with rtPA. From 90 min to 3 h, pulmonary artery pressure was significantly lower with both rtPA regimes than with either UK regime. These results indicate, at least In the model employed, that compared with treatment with UK, pulmonary thrombolysis and corresponding hemodynamic improvement are greatest with rtPA. AM REV RESPIR DIS 1990; 141:290-295

er regime results in relatively rapid pulmonary thrombolysis (1, 2). Because results from that study suggested that a further increase in rtPA concentration would not significantly affect the rate of thrombolysis, in the current study, the high dose was employed to test this hypothesis, i.e., rate and extent of thrombolysis will be similar when 1 and 2 mg/kg of rtPA is given over 15 min. Methods Twenty-four dogs (weighing 13to 22 kg) were anesthetized intravenously with pentobarbital (30 mg/kg) and supplemented as required to maintain apnea. Each dog was mechanically ventilated in the supine position via endotracheal tube with 100% O 2 at a tidal volume of 20 ml/kg. Rate was adjusted to maintain Pacoz between 25 and 45 mm Hg. Metabolic acidosis was treated as required with sodium bicarbonate to maintain arterial pH greater than 7.28. A catheter was inserted into the right femoral artery. Measurements of mean blood pressure (BP) were obtained from this catheter, and 200 ml of blood were drawn for formation of autologous clot. An intravenous line was inserted into the right femoral vein for infusion of hetastarch to maintain blood volume. A pigtail catheter (Cordis Corp., Miami, FL) was inserted via the left femoral artery into the left ventricle for measurement of left ventricular end-

diastolic pressure (LVEDP). A second intravenous line was inserted into the left femoral vein for injection of autologous clot. A thermistor-tipped, flow directed Swan-Ganz" catheter (Electro Catheter Corp., Rahway, NJ) was inserted via the external jugular vein and positioned in that proximal artery. Measurements of cardiac output (CO) and mean pulmonary arterial pressure (Ppa) wereobtained from this catheter. A second Swan-Ganz catheter was inserted by the same route into the right atrium to obtain measurements of right atrial pressure (PRA), and to inject saline boluses for determination of CO by computer (Columbus Instruments, Columbus, OH). The experimental preparation is illustrated in figure 1. Each catheter was connected to a Statham P23 transducer (Gould Inc., Oxnard,

(Received in originalform April 18, 1989 and in revised form June 27, 1989) i 1 From the Departments of Medicine and Pharmacology, University of Manitoba, and the Health Sciences Centre, Winnipeg, Manitoba, Canada. 2 Supported by the Medical Research Council of Canada. 3 Correspondence and requests for reprints should be addressed to Dr. Richard M. Prewitt, Department of Medicine/Section of Cardiology, Health Sciences Centre, 700 William Ave., GF223, Winnipeg, Manitoba R3E OZ3 Canada. 4 Manitoba Heart Foundation Research Professor.

THROMBOLYTIC THERAPY IN PULMONARY EMBOLISM

Endotracheal Tube

Blood Pre ••ure Cath.

y~~ ~~ L." V••'.'''.C.'h. (

(

lnfu.ion Line

Normal Saline

Radioactive Clot

Fig. 1. The experimental preparation is illustrated.

CAl leveled to midchest. Transducer output was displayed on a 12-channel Electronics for Medicine oscillograph (PPG Biomedical Systems, Pleasantville, NY). In an attempt to minimize bleeding, thrombostat (Parke-Davis Canada, Scarborough, ON) at a concentration of 1,000U/ml isotonic saline was applied in all animals to surgical wounds.

Radioactive Autologous Blood Clot Formation A low specific activity 99mTc sulfur colloid preparation was created by boiling 3.0 ml of 1 N HCI, 3.0mlofNa2S203-5 H 20, and 1.0to 1.5 GBq (IGBq = 27 mCi) of 99 mTc pertechnetate in 9.0 ml of saline for 3.5 min. 99mTc sulfur colloid (TSC) was chosen to label clot because of its known affinity for fibrin strands and because the small particles (0.1 urn) when released are rapidly cleared by the reticuloendothelial system (serum half time approximately 2 min) making background correction unnecessary (8,9). After ice-bath cooling for 5 min, 0.3 ml of human serum albumin and 8.0 ml of phosphate buffer were added. High quality preparations were confirmed using instant thin-layer chromatography in methyl ethyl ketone (98.4 ± 0.3%). Autologous clot was formed by slowly dripping 100 ml of freshly drawn unheparinized dog blood with 7.0 ml (350 MBq) of TSC and 10,000 U (10 ml) of thrombin into a shielded 500-ml pyrex beaker. The mixture was allowed to stand for 90 min to 2 h until the clot had a "jellolike" consistency. The serum was decanted and discarded. The low specific activity TSC ensured a large number of particles with adequate distribution in the clot matrix. The use of a small volume of TSC and the glass-walled container were necessary to develop a solid thrombus. The clot was then cut into approximately 1-ml aliquots and loaded into 6O-ml syringes before injection.

Protocol After obtaining baseline measurements, au-

291

tologous clot was injected over approximately 30 min into the left femoral vein. Each clot injection was followed by saline flush. Initially, clot was injected to increase Ppa to approximately 45 mm Hg. Over the next 20 min, each time the Ppa dropped below 45 mm Hg, clot was injected to return the pressure to approximately 50 mm Hg. The preparation was then allowed to stabilize for 1 h. Initially, over 15 to 20 min, Ppa decreased. Subsequently, the preparation became relatively stable. To confirm stability, hemodynamic measurements were taken every 15 min and in all instances, over the last two measurements, Ppa and CO varied less than 10070. The dogs were then randomly selected for treatment with tissue plasminogen activator (rtPA) (Genentech, Toronto, ON) at 1.0 mg/kg (rtPA 1) or 2.0 mg/kg (rtPA2) or with UK (Abbokinase; Abbott Laboratories, North Chicago, IL) at 30,000 IV/kg (UK30 ) or 60,000 IV/kg (VK6o) . At the beginning of the experiment, 24 slips of paper, each specifying one of the treatment groups, were placed in a container. At the point of randomization, one of the treatments was chosen by blindly selecting one of these slips. The two doses of each drug were given over 15 min. Each treatment was followed by a heparin infusion at 10 IV /kg/h for the remainder of the experiment. Infusion of hetastarch at 100 ml/h was started at the beginning of drug infusion and was continued until the conclusion of the experiment. Hemodynamic measurements (BP, Ppa, CO, PRA, and LVEDP) were taken at the following times: baseline (before clot injection), after embolization (immediately before treatment), and at 15,30,45,60, and 90 min and at 2, 2.5, and 3 h after the onset of drug infusion. Two 6-ml arterial blood samples were collected for fibrinogen measurement at the baseline condition, after embolization, and at 1 and 3 h after drug onset. The samples were added to 15ml Falcon 2057 polystyrene tubes (Falcon Plastics, Oxnard, CAl, each containing 650}l1 acid citrate {two parts 0.1 M citric acid, three parts 0.1 M Na citrate and 200 ul aprotinin (22 TID/ml; 900 kallikrein inhibitor units per TIV). The samples were immediately centrifuged at 2,500 rpm in either an IEC Centra-7R or an HN-5 centrifuge (International Equipment Co., Needham Heights, MA) and the plasma was separated and frozen at - 20° C. Fibrinogen estimation was done by the heat fractionation method of Goodwin using the modified Biuret method (10).

Assessment of Pulmonary Thrombolysis Monitoring of chest and abdominal activity was achieved with a Picker Dyra IV mobile gamma counter (Picker International Canada, Winnipeg, MB), using a parallel hole collimator coupled to a Medical Data Systems A2 mobile computer (Medtronic of Canada Ltd., Richmond, BC). Four-hour dynamic acquisition was in 64 x 64 matrix at a 120 s per frame rate. Thus, an image was obtained every 2 min over the course of the experiment.

Regions of interest were placed about the lung fields. To assess total pulmonary thrombolysis, counts in the lungs were summed over the 10 min just before administration of therapy and were compared with a decay-corrected image from the final 10 min. To assess rate of pulmonary thrombolysis, a marker was placed at the onset and end of drug infusion, and a linear best-fit curve between these coordinates was generated by the computer. Thus, the slope of the time-activity curve defined the rate of clot lysis during drug infusion. This relationship is expressed in terms of the percent decline of total lung counts per hour. In a previous 3-h study employing rtPA, clot lysis was shown to return to control levels during the second hour (1). Thus, to determine the point at which drug-induced clot lysis ceased in each dog, the slope of the timeactivity curve over the last hour was considered the control rate of lysis. The control slopes of all dogs were obtained and compared by one-way ANOVA. There were no significant differences between groups (table 3). Therefore, in each dog the "control" slope was back-projected along the time-activity curve. The point at which each curve reached its control rate of lysis was prospectively defined as the point at which clot lysis caused by drug treatment ceased.

Data Analysis The total clot lysis, rate of clot lysis during infusion, time of clot lysis after infusion, and fibrinogen levels were compared between groups by a one-way ANOVA. Hemodynamic parameters were analyzed for a change with embolization by a paired t test, and for a change with time by a two-way ANOVA. These hemodynamic measurements, plus the percent change in Ppa from infusion onset, were compared between groups at each time interval by one-way ANOVA. If any oneway or two-way ANOVA was significant (p < 0.05), the Student Newman-Keuls test was used to determine which points were different. Results

Mean hemodynamic values before and after embolization in all 24 dogs are illustrated in table 1. Embolization caused a marked increase in Ppa (p < 0.0005) and a corresponding decrease in CO (p < 0.00(5). PRAincreased (P < 0.00(5) and BP and LVEDP remained constant. Individual and mean (± SE) values are plotted in table 2, which illustrates effects of the different treatment regimes on rate of pulmonary thrombolysis. The values depicted are those obtained during the IS-min infusion intervals. Note that compared with UK 3 0 , rate of pulmonary thrombolysis was increased with UK6 0 • Also, compared with UK30 , both rtPA regimes increased the rate of clot lysis. The mean value for thrombolysis was greater with rtPA 2 than with rtPA 1 and UK 6 0 •

292

PREWITT, HOY, KONG, GU, GREENBERG, COOK, CHAN, AND DUCAS

TABLE 1 HEMODYNAMIC EFFECTS OF EMBOLIZATION

Condition

BP (mmHg)

Ppa (mmHg)

CO (Llmin)

PRA (mmHg)

LVEDP (mmHg)

Pre-embolization Postembolization

142 ± 5 139 ± 4

13.4 ± 0.4 34.2 ± 0.9*

2.5 ± 0.2 1.6 ± 0.1 *

2.4 ± 0.3 5.0 ± 0.4*

6.4 ± 0.6 8.0 + 0.8

Definition of abbreviations: BP = blood pressure; Ppa = pulmonary artery pressure; CO pressure; LVEDP = left ventricular end-diastolic pressure. 'p < 0.0005 when compared with pre-emboli condition.

= cardiac

output; PRA

= right atrial

TABLE 2 EFFECTS OF RECOMBINANT TISSUE PLASMINOGEN ACTIVATOR (rtPA) AND UROKINASE (UK) ON RATE OF CLOT LYSIS DURING INFUSION rtPA 1

• p p

t

51 40 37 50 31 25 39 4t

UK 60,000

(r)

(% clot lysis/h)

(r)

(% clot lysis/h)

(r)

(% clot lysis/h)

(r)

0.990 0.996 0.989 0.841 0.980 0.935 0.955 0.025

48 56 37 66 43 46 49 4*

0.990 0.988 0.985 0.991 0.99 0.944 0.990 0.001

25 25 16 12 24 19 20 2

0.973 0.973 0.955 0.915 0.957 0.967 0.957 0.009

37 25 29 37 42 67 40 6t

0.991 0.987 0.979 0.986 0.986 0.996 0.988 0.002

(% clot lysis/h)

Mean ± SE

UK 30,000

rtPA2

< 0.001 when compared with UK 30,000. < 0.05 when compared with UK 30,000.

However, because of large intragroup appreciable amount of time after thromvariability, this difference was not signifi- bolytic therapy was discontinued. Note cant. In each instance, the count-time that the duration ofthis persistent thromcoordinates obtained during drug infu- bolysis was significantly greater with sion were well described by linear regres- both rtPA regimes than with UK 30 or sion analysis, and the values of the regres- U~o. As illustrated in figure 2, after drug sion coefficients (r) are also depicted in infusion, rate of thrombolysis gradually attenuated, and over the third hour, it was table 2. Typical time-activity curves obtained constant in all groups (table 3). Mean (± with each treatment regime are plotted SE) results are plotted in figure 3, which in figure 2. Note that during infusion, illustrates effects of treatment on extent rate of thrombolysis was relatively con- of total clot lysisover 3 h. Compared with stant. Further, as illustrated in table 3, U~o and UKJO , both rtPA regimessignifieffective thrombolysis persisted for an cantly increasedpulmonary thrombolysis.

g

o

7/8/88

100.0

U

;;i

o

UKJO

I-

Z W

rtPA,

. - 0 01 1

11.8% LYSIS

LL

34.9% LYSIS

0

88.2% REMAINING

§

65.1 % REMAINING

~ W -

-

-

-

- 1.631

ll.

OL

_

0.011

1:65 XlO'

X 10'

o

I-

o l~.~/r88~:::::->o