Drug Effects on Function in the Ferret lschemic Hindlimb

EDWARD 0.

WESELCOUCHAND CHRISTINE D. DEMUSZ

In this article a new model of peripheral occlusive arterial disease is described. The lower hindlimb of an anesthetized ferret was fixed to a holder, the distal end of the Achilles tendon attached to an isometric force transducer, and a passive preload of 100 g was applied to the muscle preparation. The hindlimb was stimulated to contract isometrically via supramaximal electrical stimulation of the sciatic nerve. During the initial period, when femoral blood pressure equaled aortic blood pressure, net contractile force peaked within 1 or 2 min (372 f 24g, n = 20) and gradually declined to about 85% of peak over 20 min. Following 60 min of ischemia (induced by partial occlusion of the abdominal aorta), when blood pressure in the contralateral femoral artery was about 45 mm Hg, the 15-min area under the force-time curve (AUC) was 33.2 + 2.5% (n = 4) of the initial value. To validate the utility of this model in the search for treatment of peripheral vascular diseases, two drugs were tested. Pentoxifylline, which is clinically effective, attenuated the loss of function observed during ischemia in a dose-related manner, but nifedipine, which is without clinical benefit, had no effect at a dose that was extremely hypotensive. Because femoral perfusion pressure was controlled, systemic hemodynamic effects of test compounds are minimized as potential mechanisms of action, simplifying the evaluation of novel pharmacotherapy for treatment of ischemic diseases.

Key Words:

Ferret hindlimb;

Peripheral vascular disease; Ischemia;

Pentoxi-

fylline

INTRODUCTION Intermittent claudication is an increasing problem in the aging population, where it can severely limit mobility and quality of life. The symptoms, usually characterized by leg pain upon walking, are caused by occlusive peripheral arterial disease and occur in about 10% of those over 65 (Kannel and McGee, 1985). With the exception of pentoxifylline (Ward and Clissold, 19871, pharmacologic therapy of the symptoms has been largely disappointing (Coffman and Mannick, 1972; Verstraete, 1982), and the search for better chemotherapy has been hampered by the lack of relevant animal models. For example, pentoxifylline was reported to be effective in only one preclinical model of peripheral vascular disease (Angersbach and Ochlich, 1984). In this study, pentoxifylline improved contractile force in a feline ischemic hindlimb preparation. However, important hemodynamic parameters, like systemic blood

From the Cardiovascular Pharmacology Department, Bristol-Myers Squibb Co., Pharmaceutical Research and Development Division, Wallingford, Connecticut. Address reprint requests to: Edward 0. Weselcouch, Department 307, Bristol-Myers Squibb Co., Wallingford, CT 06492. 255 Journal of Pharmacological Methods 0 1990 Elsevier Science Publishing

23, 255-264 (1990) Co., Inc., 655 Avenue of the Americas, New York, NY lWl0

256

E. 0. Weselcouch

and C. D. Demusz

pressure, were not controlled or reported. Hemodynamic changes can affect function in this model and, therefore, need to be controlled so that treatment can be properly evaluated. In this report, we describe a novel model of occlusive peripheral vascular disease and show that the effects of two reference agents, pentoxifylline and nifedipine, correlate well with the clinical experience of these drugs. Furthermore, we demonstrate that the protective effects of pentoxifylline in this model are independent of effects on systemic blood pressure, a clear advantage over previous models of the disease. MATERIALS

AND

METHODS

Drugs

Pentoxifylline was a gift of Hoechst-Roussel Pharmaceuticals Inc. (Somerville, NJ). Nifedipine was purchased from Sigma (St. Louis, MO). Both compounds were dissolved in a vehicle of DMF:ethanol:saline (1:3:6) at a concentration sufficient to deliver the total dose in a volume of 0.1 mL/kg. Animals

Castrated male ferrets were purchased from Marshall Farms (North Rose, NY) and housed three to a cage. Ferret chow and tap water were provided ad libitum. At the time of the experiments, the ferrets weighed between 1 and 2.5 kg. Surgical

Procedure

Ferrets were anesthetized with pentobarbital (45 mg/kg, IP). The trachea was exposed and cannulated, and the animal was mechanically respired. Respiration was started at a rate of 25 breaths per minute and at a volume of about 26 mL per breath, and was adjusted if necessary following measurement of blood gases and pH (Corning 168 pH/Blood Gas Analyzer) (pH = 7.36 ? 0.02, pCO;? = 32.0 4 2.5 mm Hg, pOZ = 87.1 + 4.2, n = 20). A carotid artery and the left femoral artery were cannulated for measuring arterial blood pressure (pressure transducers: carotid, Statham p23Gb; femoral, Statham p23Gb). A jugular vein was cannulated for administration of experimental compounds and supplemental anesthesia. The abdomen was opened with a midline incision, and the abdominal aorta caudal to the renal arteries was dissected free of surrounding tissue. An hydraulic flow occluder (In Vivo Metric Systems, CA) was placed around the aorta and the abdomen closed. The sciatic nerve was isolated on the dorsal side of the right hindlimb, ligated, and severed proximal to the ligation. The right lower hindlimb was fixed to a holder, and the distal end of the right Achilles tendon was severed. One end of a silk ligature was tied around the tendon, and the other was attached to an isometric force transducer (Grass FT03C). A passive preload of about 100 g was applied to the hindlimb preparation. Contractions of the hindlimb were induced by electrical stimulation of the sciatic nerve. The nerve was stimulated at supramaximal voltage for a duration of 0.5 msec at a frequency of 3.5 Hz (Grass Stimulator SD9 connected through a Grass Stimulus isolation Unit SU15, Grass Instruments, Quincy, MA).

Drug Effects in the Ferret lschemic Hindlimb

Experimental

Protocol

Following completion of the surgical preparation, the hindlimb was stimulated to contract for about 30 set to check that the Achilles tendon was securely attached to the force transducer. Fifteen minutes later, the experiment was started. The protocol is illustrated in Figure 1. During conditions of normal flow to the hindlimb (femoral blood pressure = systemic blood pressure), the preparation was stimulated to contract for a period of 20 min. At the end of this period, hindlimb ischemia was produced by occlusion of the abdominal aorta. The occlusion was adjusted so that femoral blood pressure was reduced to between 40 and 50 mm Hg. The results from preliminary experiments indicated that ischemia in this pressure range would cause a predictable and consistent loss of function; at pressures not much below 40 mm Hg, the hindlimb rapidly lost all ability to contract, and at pressures not much above 50 mm Hg, little function was lost. Thirty minutes into the ischemia, the ferret was treated with the experimental compound or vehicle. The occlusion was adjusted to maintain femoral blood pressure around 45 mm Hg. Thirty minutes later, a second 20-min period of stimulation was begun. This 30-min interval between dosing the ferret and performing the experiment allowed the hemodynamic effects of treatment to diminish. At the end of the second period of stimulation, the ferret was sacrificed with a bolus injection of saturated KCI. Measurement

of Blood Flow

Blood flow in the gastrocnemius muscle was measured in separate experiments by the method of local tissue clearance of ‘33Xe (Gosselin and Audino, 1971). Ferrets 100

Ischemia 0

20

40

80

80

100

Minutes

FIGURE 1. The effect of pentoxifylline (10 mg/kg, IV) on contractile force in the ferret ischemic hindlimb. The open symbols represent forces observed during the initial normal flow period, and the closed symbols are forces observed during ischemia. The circles are the vehicle-treated animals, and the triangles are animals treated with pentoxifylline.

257

258

E. 0. Weselcouch

and C. D. Demusz

were prepared as described above, and about 3 min prior to initiating exercise, 5 FL of a solution of ‘33Xe (New England Nuclear, Boston, MA) in normal saline was injected directly into the right gastrocnemius muscle through a 30-g needle. Clearance of ‘33Xe from the tissue depot was monitored with a 2-in. Nal scintillation detector connected to an amplifier system (EC & G Ortec, Oak Ridge, TN). The face of the detector was placed several millimeters from the injection site, and counts were accumulated for 20-set intervals over the next 10 min. Clearance rate constants were determined by linear regression of the counts for the 2-min period just before exercise and for the 3-min period following the initiation of exercise. Muscle blood flows were calculated from these rate constants, as described by Gosselin and Audino (1971). Two clearance trials were performed in each ferret: one with the femoral arterial pressure equal to systemic pressure, and a second when femoral pressure was reduced to 40-50 mm Hg. Data Acquisition

and Analysis

Systemic blood pressure, femoral blood pressure, and hindlimb contractile force were recorded on a Beckman R612 Dynagraph. In addition, a Tecmar A-D convertor (Scientific Solutions, Solon, Ohio) was used to record these measurements directly into an IBM PC-XT. Every 30 set during the stimulation of the hindlimb, the analog signals were sampled at IO-msec intervals for a period of 3 sec. The contractile force peaks were selected from these data. From each 3-set sample, the top five peak forces were averaged, and this value was recorded as the contractile force representative of that time period. The mean systemic and femoral blood pressures corresponding to those five time points were also averaged and used as the representative value. Baseline forces were similarly calculated from the five lowest contractile force values. Baseline forces (passive preload) were subtracted from the peak contractile forces for each time period, and these net contractile forces were normalized by expressing them as a percentage of the peak net contractile force observed during the normal flow period of stimulation. For the purpose of this report, the subsequent use of the terms force or contractile force will refer to these normalized net force values. In addition, the area under the force-time curve (AUC) was calculated for the first 15 min of stimulation for both the normal flow and the ischemic periods. The AUCs calculated for the ischemic periods were expressed as a percentage of their corresponding normal flow AUC. The effect of treatment on systemic and arterial blood pressures, peak contractile forces, and AUCs observed during ischemia were evaluated by one-way analysis of variance (ANOVA) using SAS (SAS Institute Inc., Cary, NC). When the ANOVA indicated significant differences (p < 0.05), the treatment group means were compared with the control group means with the Bonferroni test (Wallenstein et al., 1980). Other statistical tests are noted in the text. All values are expressed as mean ? SE.

Drug Effects in the Ferret lschemic Hindlimb

RESULTS The basic behavior of this model is shown in Figure 1. Following the initiation of stimulation during the normal flow period, contractile force reached a peak of 372 ? 24 g (n = 20) within 1 or 2 min and gradually declined to about 85% of the peak value during the 20-min period. During the subsequent period of stimulation during ischemia in the vehicle-treated ferrets, hindlimb contractile force reached only 77.6 + 1.9% of the initial peak force (n = 4, p < 0.001, paired t test versus initial peak force). Unlike in the normal flow conditions, contractile force decayed rapidly, reaching a plateau of about 20% of peak. Three experiments were performed to demonstrate that the 60-min interval between the first and second exercise period did not alter the performance of the hindlimb in the absence of ischemia. In these experiments, the peak force during the second period of stimulation was 95.2 + 3.8% of that seen during the initial period (n = 3, p > 0.05, paired t test versus initial peak force), indicating that this parameter is unaffected by the time interval between the exercise challenges, and the reduced peak force is due only to the ischemia. Although the plot of mean forces during ischemia (Figure I) suggests that this parameter was stable after about 5 min, in individual experiments force tended to vary considerably during this period. As shown in Figure 2, it was the small, spontaneous changes in femoral blood pressure that account for these changes in force. To minimize the influence of these small changes in femoral pressure on the eval-

202 30

35

Femoral

40

45

Blood Pressure

50

55

(mm Hg)

FIGURE 2. The effect of femoral blood pressure on contractile force in the ferret hindlimb during ischemia. The points represent spontaneous changes in force and femoral pressure observed during a single experiment and were measured from 8 to 20 min after the initiation of stimulation during ischemia. The line was determined by linear regression.

259

260

E. 0. Weselcouch

and C. D. Demusz

uation of the effect of treatment, the area under the force-time curve (AUC) was used as the index of function during ischemia. During ischemia, AUC was 33.2 + 2.5% of normal flow control values (n = 4, p < 0.001, paired t test versus AUC of control flow period). As was true for peak force, the 60-min interval between stimulation periods was not responsible for decrease in AUC. In the absence of ischemia, the AUC of the second period of stimulation was 96.0 + 3.7% of that seen in the normal flow period (n = 3, p > 0.05, paired t test). Effect of Treatment

Results are presented in Tables 1 and 2. Table 1 shows that pentoxifylline at 10 mg/kg significantly attenuated the loss of function during ischemia. Based on the plateau seen for pentoxifylline at 1 and 10 mg/kg, it appears that a maximum protective effect results in an AUC of about 55% of normal flow values. These protective

TABLE 1 Effect of Treatment on Peak Contractile Force and Area Under the Force-time Curve (AUC) During lschemia in the lschemic Hindlimb of Ferrets TREATMENT

MC/KC

N

PEAKFORCES

4

77.6

f

1.9

33.2

-c 2.5

0.1

4

73.4

f

a.8

32.8

f

1.0

4

85.0

f

6.6

50.7

2 5.4

10.0

4

92.7

f

7.1

54.7

*

0.3

4

86.2

2 5.8

41.3

+ 3.4

Vehicle Pentoxifylline

Nifedipine a Peak force force

observed

during

ischemia

during

the normal

b AUC expressed

expressed

as a percentage

AUCb

as a percentage

of the

a.0 4.2’

peak

flow period. of the AUC observed

during

the nor-

mal flow period. ’ Indicates

p < 0.05 for comparison

with vehicle-treated

animals.

TABLE 2 Mean Systemic and Femoral Blood Pressures Measured During Normal Flow and Periods of Hindlimb lschemia in Ferrets BLOOD PRESSURE (MM Ho NORMAL FLOW” TREATMENT

SYSTEMIC

MC/KG

a Normal period

when

FEMORAL

134*

7

1342

6

105 *

19

42.5

f

9

149*

7

110 t

ia

43.5

2 0.6

1.0

122 f 1252 135 f

0.3 flow refers to the period the abdominal

b p < 0.01 compared

SYSTEMIC

1422

10.0 Nifedipine

FEMORAL

0.1

Vehicle Pentoxifylline

ISCHEMIA~

when

10

118 *

10

94 f

16

44.4

+ 0.8

7

118 2

6

123 f

12

42.2

+ 1.2

45.7

f

126 2 10

IO

the abdominal

flow period

(paired

72?

aorta was not occluded.

aorta was occluded.

to normal

2.3

t test).

3b lschemia

1.5

refers to the

Drug Effects in the Ferret lschemic Hindlimb

effects are not due to difference in femoral or systemic blood pressure during ischemia (Table 2). The apparent increase in peak contractile force seen during ischemia following treatment was not significant (p = 0.231, ANOVA). Nifedipine was tested to determine if a potent vasodilator would be protective in this model. As shown in Tables 1 and 2, nifedipine, at a dose that was extremely hypotensive, was ineffective at preventing the loss of function in the ischemia hindlimb, suggesting that vasodilation alone is not effective in this model. Hindlimb

Blood Flow

Hindlimb blood flow was measured in three ferrets, and these results are presented in Figure 3. It is clear that when the abdominal aorta was not occluded, muscle blood flow was able to increase in response to exercise. However, when femoral blood pressure was controlled within the range of 40-50 mm Hg, not only was resting flow slightly reduced, but no increase in flow was observed during exercise. The lack of an exercise-induced hyperemia during ischemia strongly suggests that the vascular bed of the hindlimb muscles is maximally vasodilated and unable to respond further. DISCUSSION The goal of pharmacological therapy in the treatment of ischemic diseases is the improvement of tissue function, and most clinical trials of drugs have focused on such endpoints. For example, many trials of pentoxifylline, a drug used to treat peripheral vascular disease, have used increases in walking distance of patients suffering from intermittent claudication as the primary endpoint (Cameron et al., 1988; johnson et al., 1987; Ward and Clissold, 1987). Similar endpoints have been

U Normal Flow f&Xl Ischemia

Rest

Exercise

FIGURE 3. Muscle blood flow at rest and during exercise in the ferret hindlimb during normal perfusion (open bars) and during ischemia (striped bars). Only during periods of normal perfusion (femoral blood pressure = systemic blood pressure) was flow signifi~ntly increased in response to exercise (* indicates p < 0.05, ANOVA).

261

262

E. 0. Weselcouch

and C. D. Demusz

used for the clinical evaluation of other drugs targeted for these diseases, e.g., buflomedil (for review, see Clissold et al., 1987). Unfortunately, relevant animal models of peripheral vascular disease using functional endpoints are few, and the preclinical evaluation of these compounds has centered on indirect measurements of tissue function. For example, increases in tissue pOZ and pH in ischemic rat hindlimb (Angersbach and Ochlich, 1984) and increases in red cell flow velocity in rat mesentery microvessels (Ohshima and Sato, 1981) are a few of the reported effects of pentoxifylline in animal models. However, the link between these effects and an improvement in tissue function has not been well established. In a model similar to the ferret hindlimb, Angersbach and Ochlich (1984) reported that both pentoxifylline and BRL 30892 (denbufylline) improved contractile force in an acutely ischemic cat hindlimb. In these experiments, ischemia was induced by partial occlusion of the femoral artery such that contractile force was reduced by 30%. lntraduodenal dosing with pentoxifylline or BRL 30892 caused an increase in contractile force in a dose-related fashion. However, in these experiments, neither systemic blood pressure nor blood pressure distal to the femoral occlusion were controlled or reported. In the experiments reported here, the effective perfusion pressure (i.e., femoral arterial pressure) of the ferret ischemic hindlimb was controlled. This control is an important consideration, because it is clear from the data presented in Figure 3 that during this level of ischemia, the hindlimb is near maximally or maximally vasodilated and unable to dilate further to increase flow in response to exercise. Thus, flow is determined primarily by inlet pressure, in this case the pressure within the femoral artery, and any change in this pressure will result in a corresponding change in flow. Changes in flow under these conditions should result in a change in hindlimb function. Indeed, as demonstrated in Figure 2, contractile force of the ferret ischemic hindlimb is very sensitive to small changes in pressure, an observation entirely consistent with a fixed, maximally dilated vasculature. Assuming a similar situation in the feline model of Angersbach and Ochlich (1984), any change in systemic pressure would also result in a change in the perfusion pressure in the cat hindlimb. Pentoxifylline, which is mildly hypertensive in the ferret (systemic blood pressure increased from 8 to 23 mm Hg5 min after dosing), might be expected to raise systemic blood pressure in the cat, and even though such increases might be small, they could be sufficient to explain partially the results reported by Angersbach and Ochlich (1984). It is clear that without knowing the effect of treatment on systemic blood pressure, it is impossible to evaluate properly the results in their functional model of peripheral vascular disease. In the model reported here, the effective perfusion pressure of the ferret ischemic hindlimb was controlled, and, thus, changes in it were minimized as a factor in evaluating the changes in function caused by treatment. It seems unlikely that increases in collateral flow would be responsible for the protective effect of pentoxifylline. Collateral flow could influence muscle performance only if the vessels supplying the collateral flow entered the contracting muscles by a route that did not connect with the femoral system; any newly opened or dilated collateral vessels

Drug Effects in the Ferret lschemic Hindlimb

that entered the femoral system would contribute to the femoral pressure, which was controlled at a constant level. Pentoxifylline has little vasodilator activity (Ward and Clissold, 1987), and, as described earlier, such activity should be of little utility in this model. Indeed, the failure of nifedipine to prevent loss of function in the ferret ischemic hindlimb clearly demonstrates that vasodilation is ineffective. This observation parallels the clinical experience where vasodilators are of little value in the treatment of peripheral vascular disease (Coffman, 1988; Coffman and Mannick, 1972). These results give little information regarding the mechanism of action of pentoxifylline in attenuating the loss of function in this model of occlusive arterial vascular disease. Pentoxifylline is reported to have hemorheologic effects that could increase flow in the absence of changes in vascular geometry (Schmalzer and Chien, 1984,1989; Ward and Clissold, 1987; Weselcouch and Baird, 19891, and such activity would most certainly be of benefit. Further experiments would be required to explain more fully the action of this drug in the ferret ischemic hindlimb. In conclusion, the ferret ischemic hindlimb represents a simple model of peripheral occlusive arterial disease. By the controlling of effective perfusion pressure in the ischemic muscle preparation, systemic hemodynamic effects of test compounds are eliminated as a possible mechanism of action, simplifying the evaluation of results. We have demonstrated that pentoxifylline is protective in the model and that a vasodilator, nif~dipine, is not-results that reflect the clinical experience with these drugs. Thus, this model may be used to screen test compounds for efficacy in the treatment of ischemic vascular diseases, like intermittent claudication.

REFERENCES Angersbach D, Ochlich P (1984) The effect of 7-(2’-oxopropyl)-I ,3-di-n-butyl-xanthine (BRL 38892) on ischaemic skeletal muscle ~02, pH and contractility in cats and rats. Arzneim Forsch 34:1274-1278. Cameron HA, Wailer PC, Ramsay LE (1988) Drug treatment of intermittent claudication: a critical analysis of the methods and finding of published clinical trials, 1965-1985 Br 1 C/in Pharmacol 26:569-576. Clissold SP, Lynch S, Sorkin EM (1987) Buflomedil. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in peripheral and cerebral vascular diseases. Drugs 33:430-460. Coffman ID (1988) New drug therapy in peripheral vascular disease. Med Clin N Am 72:2.59-265. Coffman JD, Mannick JA (1972) Failure of vasodilator drugs in arteriosclerosis obliterans. Ann Intern Med 76:35-39.

Gosselin RE, Audino LF (7971) Muscle blood flow and functional capillary density evaluated by isotope clearance. Pfluegers Arch Gesamte Physiol ~enschentiere 322:197-216. Johnson WC, Sentisse JM, Baldwin D, Hamilton j, Dion J (1987) Treatment of claudication with pentoxifylline: are benefits related to improvements in viscosity? ] Vast Surg 6:211-216. Kannel WB, McGee DL (1985) Update on some epidemiologicfeatures of intermittent claudication: The Framingham study. ] Am Geriatr Sot 33:1318. Ohshima N, Sato M (1987) Effect of pentoxifylline on microvascular blood flow velocity. Angiology 32:752-763. Schmalzer EA, Chien S (1984) Filterabjli~ of subpopulations of leukocytes: Effect of pentoxifylline. Wood 64542-546. Schmalzer EA, Chien S (1989) Effect of methylxanthines, cytochalasin B and FMLP on neutrophil

263

264

E. 0. Weselcouch filterability

and C. D. Demusz

in a constant flow system. C/in He-

morheol9:69-80.

Verstraete M (1982) Current therapy for intermittent claudication. Drugs 24:240-247. Wallenstein S, Zucker CL, Fleiss JL (1980) Some statistical methods useful in circulation research. Circ Res 47:1-9.

Ward A, Clissold SP (1987) Pentoxifylline, a review of its pharmacodynamic and pharmacokinetic properties, and its therapeutic efficacy. Drugs 34:50-97.

Weselcouch EO, Baird AJ (1989) The effect of BMY20014 and pentoxifylline on the filterability of human neutrophils and red blood cells. C/in Hemorheol9:973-982.

Drug effects on function in the ferret ischemic hindlimb.

In this article a new model of peripheral occlusive arterial disease is described. The lower hindlimb of an anesthetized ferret was fixed to a holder,...
701KB Sizes 0 Downloads 0 Views