Journal of Orthopaedic Research 8297-303 Raven Press, Ltd., New York 0 1990 Orthopaedic Research Society
Effects of Exercise on Blood Flow to Canine Articular Tissues Peter A. Simkin, *Alice Huang, and Richard S. Benedict Departments of Medicine and *Physiology and Biophysics, University of Washington, Seattle, Washington, U.S.A.
Summary: We measured articular blood flow by standard microsphere techniques in normal adult dogs at rest and during treadmill exercise. All animals but one underwent P-adrenergic blockade as part of another experimental protocol. Expressed in p,Vmin/g & SEM, baseline flow values to articular tissues were: knee synovium 26 ? 4, femoral condyle 130 ? 21, tibia1 plateau 182 & 29, articular fat pad 9 & 2, knee ligaments 17 3, menisci 34 ? 6, wrist synovium 19 ? 4, distal radius 65 ? 13, and h a t e bone 59 & 13. Blood flow increased with exercise in all soft tissues of both the knee (stifle joint) and the wrist (radiocarpal joint). Geometric mean exercisehest flow ratios ranged from a low of 1.44 (p < 0.05) in the menisci of the knee to a high of 7.25 (p < 0.001) in the synovium of the wrist. In contrast, blood flow did not rise in juxtaarticular bones and fell significantly in femoral condyles (mean flow ratio 0.71, p < 0.005). These findings indicate that articular soft tissues derive increased perfusion from the redistribution of blood flow that accompanies short-term exercise. In contrast, flow to juxtaarticular bone does not increase under these conditions. Key Words: Blood flow-Microsphere-Synovium-Bone-Dog.
*
Synovial joints permit motion to occur in the mammalian musculoskeletal system. Thus, the central role of these structures involves activity, yet most articular physiology has focused on knee joints at rest (12,23). This inappropriate emphasis on the resting state has been likened to physiologic studies of the stopped heart (16). An understanding of the inactive joint is valuable, but its utility lies largely in the baseline it provides for interpretation of the articular response to physical activity. How, for instance, does exercise affect the flow of blood to articular tissues? This question is of particular interest in view of the long-standing concept that cardiac output is redistributed away from most organs-except the heart, brain, and skeletal muscle-during normal exercise (20).
The opportunity to study articular blood flow at rest and during treadmill exercise arose in the course of unrelated studies of myocardial blood flow in a larger series of normal adult dogs. The work employed microsphere methodology that is also well accepted as the most appropriate means for evaluating perfusion of bone and connective tissues in vivo (17,25). Comparable values in repeated control experiments provide reassurance that the capillary embolization inherent in this method does not alter physiologic vascular responses within bone (6). In the present primary experiments, myocardial perfusion was assessed at rest and during treadmill exercise. P-Adrenergic receptor blockade was also employed in all but one of the animals available for the present study (8). The normal cardiovascular response to exercise includes increases in heart rate, systolic blood pressure, and cardiac output with a concurrent decrease in peripheral vascular resistance. Each of these changes also occurs after @-blockade,although their magnitude is mod-
Received October 26, 1988; accepted October 12, 1989. Address correspondence and reprint requests to Dr. P. A. Simkin at Department of Medicine, RG-28, University of Washington, Seattle, WA 98195, U.S.A.
297
298
P . A . SIMKIN ET AL.
erately diminished (3,15,19). The presence of P-adrenergic receptor blockade in the present series afforded us the opportunity to evaluate the effects of a-adrenergic mechanisms on blood flow to exercising canine joints. MATERIALS AND METHODS Ten adult, mixed-breed dogs weighing 17.5-26.8 kg (mean = 21.9 kg) were used. Each dog was trained to trot on a treadmill, and had been splenectomized at least 3 weeks earlier to ensure a stable hematocrit during exercise. Splenectomy also heightens the sympathetically mediated vascular response during exercise and thus provides a model that more closely resembles the physiology of normal humans (28). On the day of the flow experiment, each animal was briefly anesthetized with fentanyl citrate (i.v., 0.25 mg/kg for induction, plus 1.0-mg supplements every 30 min, or as needed) and nitrous oxide (80% in oxygen, ventilation with Harvard respirator pump (Model 607; Harvard Apparatus, Inc., South Natick, MA, U.S.A.) to permit placement of a catheter in the left atrium for microsphere injection (transseptal approach, after catheterization of the right side of the heart). An additional catheter for withdrawal of reference blood samples was positioned in the left carotid artery. After catheterization, anesthesia was reversed with the narcotic antagonists naloxone hydrochloride (0.2 mg/kg, i.v.) and naltrexone hydrochloride (0.1 mg/kg, i.m.), and a 2-h period was allowed for recovery. According to the design of the primary experiments, propranolol hydrochloride (2 mg/kg, i.v.) was given to 9 of the 10 dogs. Heart rate responses to isoproterenol hydrochloride (bolus i.v. doses 0.003-10.0 pg/kg) were recorded to confirm Pblockade of the myocardium. Because tissue blood flows from the unblocked animal were in the ranges of data from the P-blocked dogs, all observations were analyzed together. The first of five left atrial infusions of radiolabeled microspheres (New England Nuclear, Boston, MA, U.S.A.) was administered with the animals at rest, usually lying down. The microspheres, usually 2.24 X lo6 in number, were 15 pm in diameter and were labeled with 46Sc,95Nb, lo3Ru, I13Sn, 51Cr, or 14'Ce. Subsequent injections of microspheres labeled with a different radionuclide were made near the end of each of four 5-min periods of treadmill exercise. Beads were diluted to 2.5 ml with normal saline and were dispersed with periodic
J Orthop Res, Vol. 8, No. 2 , 1990
ultrasonication and vortex mixing. The adequacy of these methods was confirmed by close agreement in blood flow measurements to the left and right renal cortex (8). Treadmill speed and ramp inclination were adjusted between microsphere injections to elicit a range of heart-rate responses. These conditions were chosen on the basis of prior training sessions to provide heart rates characteristic of the following four exercise levels: (a) moderate warm-up, (b) near-maximal, (c) intermediate between warmup and near-maximal, and (d) intermediate between resting and warm-up. As expected, the effects of exercise included increases in heart rate (to as high as 291 beatdmin), systolic blood pressure, and fractionated plasma catecholamine concentrations (8). The animals were sacrificed with intravenous pentobarbital sodium at the end of each experiment and the fresh cadavers were dissected for appropriate tissue samples. Both stifle joints of each dog were exposed by 10-cm midline incisions. Articular fat samples from the infrapatellar fat pad, ligament samples including both cruciates as well as medial collaterals, both menisci, and synovial samples were excised and placed in tared 10-ml scintillation vials. Synovial tissue was taken primarily from the lateral aspect of the suprapatellar pouch. Tissue harvested from the femur included articular cartilage and consisted primarily of the medial and lateral condyles to a depth of approximately 1 cm. Similar amounts of bone tissue were removed from the tibia1 plateaus. Bone tissue was sectioned in a grid-like fashion to facilitate packing in scintillation vials. Both radiocarpal joints were similarly exposed for harvest of synovial tissue, the distal radius, and the intact lunate (intermediate carpal) bone. Free synovium was difficult to obtain from this joint and specimens so labeled contained significant amounts of underlying loose and dense connective tissue. The radius was sectioned as above to a depth of approximately 1 cm. All 18 samples from each dog were individually weighed and several drops of formaldehyde were then added as a preservative. Each sample was counted in a Packard Auto-Gamma Scintillation Spectrometer (Model 5035, Packard Instrument Co., Downers Grove, IL, U.S.A.) for either 1,000 or 2,000 s, depending on the radioactivity of the microspheres. Data obtained from gamma counting were processed on the DEC-10 computer available through the John Locke Computer Center at the University
ARTICULAR BLOOD FLOW IN EXERCISE
of Washington. The counts from the different nuclide labels were separated by simple “stripping” on the basis of the emission spectra of pure samples of each nuclide, yielding the counts and number of beads for each isotope in each tissue sample. These values were then calibrated according to the number of beads in the reference sample of arterial blood, and local blood flow was calculated in microliters per minute per gram of tissue. Because of the small number of beads in many samples, values from articular tissues of the left side were averaged with those from the right to provide tissue flow values for each dog. Although myocardial oxygen consumption and blood flow varied in accord with exercise levels, we could identify no similar relationships with the small number of beads available to quantify flow to articular tissues. The data from all four levels of exercise were therefore combined and compared with data from the same animal at rest. The number of beads in each tissue varied among isotopes and reliability increases at higher bead numbers. Therefore, the mean level of blood flow was calculated after values at each exercise level were weighted according to the number of beads counted. For example, two different isotopes at different exercise levels might yield flow rates of 100 and 50 p,l/min/g in the same specimen. If 400 beads were counted in the first instance and only 100 in the second, the weighted average of these observations would be 90 p,l/min/g [(400 x 100 + 100 x 50)/500]. The effects of exercise were judged by analysis of the geometric mean of exercisehest blood flow ratios. Mean values significantly greater or less than 1 .O, by two-tailed t test, indicated respective increases or decreases in tissue blood flow during exercise.
299
RESULTS Mean weights of the harvested tissues (in grams) were: knee synovium 0.79, femur 9.76, tibia 9.08, fat 1.84, ligaments 0.96, menisci 1.66, wrist synovium 0.34, radius 3.46, and h a t e 3.36. Thus, the osseous samples were significantly heavier than the soft-tissue specimens. The resting blood flow levels i n juxtaarticular bone were also much greater than those found in adjacent soft tissues. The lowest soft-tissue flow was that of the infrapatellar fat pad (9 2 p,l/min/g) whereas the highest was that in the menisci (34 2 6 p,l/min/g). In contrast, bone flow rates ranged from 59 2 13 p,l/min/g in the lunate to 182 2 29 p,l/min/g in the tibial plateau. The combination of small samples and low blood flows means that relatively few radioactive beads were counted in the soft tissue specimens. All values are listed in Table 1 . Blood flow in the canine knee synovium increased over resting flow by a factor of nearly three during exercise (2.66, p < 0.005). Articular fat and ligament blood flow showed similar increases (2.77, p < 0.01, and 2.17, p < 0.005, respectively) over resting levels, and flow to the menisci also increased significantly (1.44, p < 0.05). In contrast, exercise-induced decreases in local blood flow were found in the femoral condyles and tibial plateau (0.71, p < 0.001, and 0.86, p < 0.10, respectively), although the latter change did not reach statistical significance. Wrist synovial blood flow increased more than seven times during exercise (7.25, p < 0.001). Differences between exercise and resting flows for the distal radius and lunate bones of the wrist (exercise/ rest ratios of 0.98 and 1.08, respectively) were not significant. These results are depicted in Fig. 1.
*
*
TABLE 1. Bloodflow to articular tissues at rest and during exercise; values are SEM, weights and bead numbers are per joint and per isotope, EIR is the geometric mean of individual exercisehest ratios, and p is the probability that the true EIR ratio is 1 .O
n
Weight (8)
Resting bead (no.)
Exercise bead (no.)
Resting flow (pVmin/g)
Exercise flow (pVmin/g)
E/R
P
Knee synovium Femoral condyle Tibia1 plateau Articular fat Knee ligament Menisci
10 10 10 7 9 10
0.79 2 0.29 9.76 2 0.36 9.08 2 0.58 1.84 2 0.67 0.96 2 0.08 1.66 2 0.14
8.1 2 1.5 545.9 ? 60.2 732.7 2 140.9 9.2 2 2.7 12.5 ? 3.5 25.4 2 5.5
13.1 2 3.5 218.5 2 43.9 377.6 ? 76.2 22.9 2 10.6 12.2 2 2.2 16.4 ? 4.2
26 2 4 130 2 21 182 ? 29 9 2 2 17 2 3 34 2 6
83 2 24 96 2 16 159 ? 27 30 2 6 41 2 8 46 2 9
2.66 0.71 0.86 2.77 2.19 1.44
Effects of Exercise on Blood Flow to Canine Articular Tissues Peter A. Simkin, *Alice Huang, and Richard S. Benedict Departments of Medicine and *Physiology and Biophysics, University of Washington, Seattle, Washington, U.S.A.
Summary: We measured articular blood flow by standard microsphere techniques in normal adult dogs at rest and during treadmill exercise. All animals but one underwent P-adrenergic blockade as part of another experimental protocol. Expressed in p,Vmin/g & SEM, baseline flow values to articular tissues were: knee synovium 26 ? 4, femoral condyle 130 ? 21, tibia1 plateau 182 & 29, articular fat pad 9 & 2, knee ligaments 17 3, menisci 34 ? 6, wrist synovium 19 ? 4, distal radius 65 ? 13, and h a t e bone 59 & 13. Blood flow increased with exercise in all soft tissues of both the knee (stifle joint) and the wrist (radiocarpal joint). Geometric mean exercisehest flow ratios ranged from a low of 1.44 (p < 0.05) in the menisci of the knee to a high of 7.25 (p < 0.001) in the synovium of the wrist. In contrast, blood flow did not rise in juxtaarticular bones and fell significantly in femoral condyles (mean flow ratio 0.71, p < 0.005). These findings indicate that articular soft tissues derive increased perfusion from the redistribution of blood flow that accompanies short-term exercise. In contrast, flow to juxtaarticular bone does not increase under these conditions. Key Words: Blood flow-Microsphere-Synovium-Bone-Dog.
*
Synovial joints permit motion to occur in the mammalian musculoskeletal system. Thus, the central role of these structures involves activity, yet most articular physiology has focused on knee joints at rest (12,23). This inappropriate emphasis on the resting state has been likened to physiologic studies of the stopped heart (16). An understanding of the inactive joint is valuable, but its utility lies largely in the baseline it provides for interpretation of the articular response to physical activity. How, for instance, does exercise affect the flow of blood to articular tissues? This question is of particular interest in view of the long-standing concept that cardiac output is redistributed away from most organs-except the heart, brain, and skeletal muscle-during normal exercise (20).
The opportunity to study articular blood flow at rest and during treadmill exercise arose in the course of unrelated studies of myocardial blood flow in a larger series of normal adult dogs. The work employed microsphere methodology that is also well accepted as the most appropriate means for evaluating perfusion of bone and connective tissues in vivo (17,25). Comparable values in repeated control experiments provide reassurance that the capillary embolization inherent in this method does not alter physiologic vascular responses within bone (6). In the present primary experiments, myocardial perfusion was assessed at rest and during treadmill exercise. P-Adrenergic receptor blockade was also employed in all but one of the animals available for the present study (8). The normal cardiovascular response to exercise includes increases in heart rate, systolic blood pressure, and cardiac output with a concurrent decrease in peripheral vascular resistance. Each of these changes also occurs after @-blockade,although their magnitude is mod-
Received October 26, 1988; accepted October 12, 1989. Address correspondence and reprint requests to Dr. P. A. Simkin at Department of Medicine, RG-28, University of Washington, Seattle, WA 98195, U.S.A.
297
298
P . A . SIMKIN ET AL.
erately diminished (3,15,19). The presence of P-adrenergic receptor blockade in the present series afforded us the opportunity to evaluate the effects of a-adrenergic mechanisms on blood flow to exercising canine joints. MATERIALS AND METHODS Ten adult, mixed-breed dogs weighing 17.5-26.8 kg (mean = 21.9 kg) were used. Each dog was trained to trot on a treadmill, and had been splenectomized at least 3 weeks earlier to ensure a stable hematocrit during exercise. Splenectomy also heightens the sympathetically mediated vascular response during exercise and thus provides a model that more closely resembles the physiology of normal humans (28). On the day of the flow experiment, each animal was briefly anesthetized with fentanyl citrate (i.v., 0.25 mg/kg for induction, plus 1.0-mg supplements every 30 min, or as needed) and nitrous oxide (80% in oxygen, ventilation with Harvard respirator pump (Model 607; Harvard Apparatus, Inc., South Natick, MA, U.S.A.) to permit placement of a catheter in the left atrium for microsphere injection (transseptal approach, after catheterization of the right side of the heart). An additional catheter for withdrawal of reference blood samples was positioned in the left carotid artery. After catheterization, anesthesia was reversed with the narcotic antagonists naloxone hydrochloride (0.2 mg/kg, i.v.) and naltrexone hydrochloride (0.1 mg/kg, i.m.), and a 2-h period was allowed for recovery. According to the design of the primary experiments, propranolol hydrochloride (2 mg/kg, i.v.) was given to 9 of the 10 dogs. Heart rate responses to isoproterenol hydrochloride (bolus i.v. doses 0.003-10.0 pg/kg) were recorded to confirm Pblockade of the myocardium. Because tissue blood flows from the unblocked animal were in the ranges of data from the P-blocked dogs, all observations were analyzed together. The first of five left atrial infusions of radiolabeled microspheres (New England Nuclear, Boston, MA, U.S.A.) was administered with the animals at rest, usually lying down. The microspheres, usually 2.24 X lo6 in number, were 15 pm in diameter and were labeled with 46Sc,95Nb, lo3Ru, I13Sn, 51Cr, or 14'Ce. Subsequent injections of microspheres labeled with a different radionuclide were made near the end of each of four 5-min periods of treadmill exercise. Beads were diluted to 2.5 ml with normal saline and were dispersed with periodic
J Orthop Res, Vol. 8, No. 2 , 1990
ultrasonication and vortex mixing. The adequacy of these methods was confirmed by close agreement in blood flow measurements to the left and right renal cortex (8). Treadmill speed and ramp inclination were adjusted between microsphere injections to elicit a range of heart-rate responses. These conditions were chosen on the basis of prior training sessions to provide heart rates characteristic of the following four exercise levels: (a) moderate warm-up, (b) near-maximal, (c) intermediate between warmup and near-maximal, and (d) intermediate between resting and warm-up. As expected, the effects of exercise included increases in heart rate (to as high as 291 beatdmin), systolic blood pressure, and fractionated plasma catecholamine concentrations (8). The animals were sacrificed with intravenous pentobarbital sodium at the end of each experiment and the fresh cadavers were dissected for appropriate tissue samples. Both stifle joints of each dog were exposed by 10-cm midline incisions. Articular fat samples from the infrapatellar fat pad, ligament samples including both cruciates as well as medial collaterals, both menisci, and synovial samples were excised and placed in tared 10-ml scintillation vials. Synovial tissue was taken primarily from the lateral aspect of the suprapatellar pouch. Tissue harvested from the femur included articular cartilage and consisted primarily of the medial and lateral condyles to a depth of approximately 1 cm. Similar amounts of bone tissue were removed from the tibia1 plateaus. Bone tissue was sectioned in a grid-like fashion to facilitate packing in scintillation vials. Both radiocarpal joints were similarly exposed for harvest of synovial tissue, the distal radius, and the intact lunate (intermediate carpal) bone. Free synovium was difficult to obtain from this joint and specimens so labeled contained significant amounts of underlying loose and dense connective tissue. The radius was sectioned as above to a depth of approximately 1 cm. All 18 samples from each dog were individually weighed and several drops of formaldehyde were then added as a preservative. Each sample was counted in a Packard Auto-Gamma Scintillation Spectrometer (Model 5035, Packard Instrument Co., Downers Grove, IL, U.S.A.) for either 1,000 or 2,000 s, depending on the radioactivity of the microspheres. Data obtained from gamma counting were processed on the DEC-10 computer available through the John Locke Computer Center at the University
ARTICULAR BLOOD FLOW IN EXERCISE
of Washington. The counts from the different nuclide labels were separated by simple “stripping” on the basis of the emission spectra of pure samples of each nuclide, yielding the counts and number of beads for each isotope in each tissue sample. These values were then calibrated according to the number of beads in the reference sample of arterial blood, and local blood flow was calculated in microliters per minute per gram of tissue. Because of the small number of beads in many samples, values from articular tissues of the left side were averaged with those from the right to provide tissue flow values for each dog. Although myocardial oxygen consumption and blood flow varied in accord with exercise levels, we could identify no similar relationships with the small number of beads available to quantify flow to articular tissues. The data from all four levels of exercise were therefore combined and compared with data from the same animal at rest. The number of beads in each tissue varied among isotopes and reliability increases at higher bead numbers. Therefore, the mean level of blood flow was calculated after values at each exercise level were weighted according to the number of beads counted. For example, two different isotopes at different exercise levels might yield flow rates of 100 and 50 p,l/min/g in the same specimen. If 400 beads were counted in the first instance and only 100 in the second, the weighted average of these observations would be 90 p,l/min/g [(400 x 100 + 100 x 50)/500]. The effects of exercise were judged by analysis of the geometric mean of exercisehest blood flow ratios. Mean values significantly greater or less than 1 .O, by two-tailed t test, indicated respective increases or decreases in tissue blood flow during exercise.
299
RESULTS Mean weights of the harvested tissues (in grams) were: knee synovium 0.79, femur 9.76, tibia 9.08, fat 1.84, ligaments 0.96, menisci 1.66, wrist synovium 0.34, radius 3.46, and h a t e 3.36. Thus, the osseous samples were significantly heavier than the soft-tissue specimens. The resting blood flow levels i n juxtaarticular bone were also much greater than those found in adjacent soft tissues. The lowest soft-tissue flow was that of the infrapatellar fat pad (9 2 p,l/min/g) whereas the highest was that in the menisci (34 2 6 p,l/min/g). In contrast, bone flow rates ranged from 59 2 13 p,l/min/g in the lunate to 182 2 29 p,l/min/g in the tibial plateau. The combination of small samples and low blood flows means that relatively few radioactive beads were counted in the soft tissue specimens. All values are listed in Table 1 . Blood flow in the canine knee synovium increased over resting flow by a factor of nearly three during exercise (2.66, p < 0.005). Articular fat and ligament blood flow showed similar increases (2.77, p < 0.01, and 2.17, p < 0.005, respectively) over resting levels, and flow to the menisci also increased significantly (1.44, p < 0.05). In contrast, exercise-induced decreases in local blood flow were found in the femoral condyles and tibial plateau (0.71, p < 0.001, and 0.86, p < 0.10, respectively), although the latter change did not reach statistical significance. Wrist synovial blood flow increased more than seven times during exercise (7.25, p < 0.001). Differences between exercise and resting flows for the distal radius and lunate bones of the wrist (exercise/ rest ratios of 0.98 and 1.08, respectively) were not significant. These results are depicted in Fig. 1.
*
*
TABLE 1. Bloodflow to articular tissues at rest and during exercise; values are SEM, weights and bead numbers are per joint and per isotope, EIR is the geometric mean of individual exercisehest ratios, and p is the probability that the true EIR ratio is 1 .O
n
Weight (8)
Resting bead (no.)
Exercise bead (no.)
Resting flow (pVmin/g)
Exercise flow (pVmin/g)
E/R
P
Knee synovium Femoral condyle Tibia1 plateau Articular fat Knee ligament Menisci
10 10 10 7 9 10
0.79 2 0.29 9.76 2 0.36 9.08 2 0.58 1.84 2 0.67 0.96 2 0.08 1.66 2 0.14
8.1 2 1.5 545.9 ? 60.2 732.7 2 140.9 9.2 2 2.7 12.5 ? 3.5 25.4 2 5.5
13.1 2 3.5 218.5 2 43.9 377.6 ? 76.2 22.9 2 10.6 12.2 2 2.2 16.4 ? 4.2
26 2 4 130 2 21 182 ? 29 9 2 2 17 2 3 34 2 6
83 2 24 96 2 16 159 ? 27 30 2 6 41 2 8 46 2 9
2.66 0.71 0.86 2.77 2.19 1.44
