Suppressed Cutaneous Endothelial Vascular Control and ...

1017

Suppressed Cutaneous Endothelial Vascular Control and Hemodynamic Changes in Paretic Extremities With Edema in the Extremities of Patients With Hemiplegia Jong-Shyan Wang, PhD, Chih Fang Yang, MD, Mei-Yon Liaw, MD, May-Kuen Wong, MD ABSTRACT. Wang J-S, Yang CF, Liaw M-Y, Wong M-K. Suppressed cutaneous endothelial vascular control and hemodynamic changes in paretic extremities with edema in the extremities of patients with hemiplegia. Arch Phys Med Rehabil 2002;83:1017-23. Objective: To investigate peripheral circulatory function and its underlying mechanisms in the paretic upper extremity after a stroke. Design: Case-control study. Setting: A department of physical medicine and rehabilitation in Taiwan. Participants: A total of 53 hemiplegic patients (28 men, 25 women; mean age ⫾ standard deviation; 58.2⫾3.8y) were studied. Subjects were divided into edema and nonedema groups. The edema group included 29 hemiplegic patients with edematous paretic upper extremities. Twenty-four hemiplegic patients in the nonedema group did not suffer from limb edema in the paretic upper extremity. Interventions: Not applicable. Main Outcome Measures: Cutaneous microvascular perfusion responses to 3 grade levels of iontophoretically applied 1% acetylcholine (ACh), 1% ACh plus 1% NG-monomethylL-arginine (L-NMMA), and 1% sodium nitroprusside (SNP) in the skin of subjects’ forearms were determined by laser Doppler perfusion measurements. Moreover, hemodynamic characteristics in the arterial and venous vessels were measured by impedance plethysmography. Results: Resting arterial inflow and venous capacity, tone, and outflow in paretic extremities did not significantly differ from nonparetic extremities, but the hyperemic arterial inflow was lower in paretic extremities than in nonparetic extremities, and paretic extremities were associated with lower ACh- and ACh plus L-NMMA–induced cutaneous perfusions than nonparetic extremities. ACh-induced cutaneous perfusions also decreased much more significantly in edematous paretic extremities than in nonedematous paretic extremities, and skin vascular responses to SNP do not differ significantly between paretic and nonparetic extremities. Conclusion: Cutaneous microcirculatory function in the paretic upper extremity after stroke may be impaired. The impairment may occur because of decreased endothelium-depen-

From the Departments of Physical Therapy (Wang) and Physical Medicine and Rehabilitation (Yang, Liaw, Chang, Wong), Gung University, Tao-Yuan, Taiwan. Accepted in revised form September 10, 2001. Supported by the National Science Council of Taiwan, ROC (contract no. NSC 90-2314-B-182-055). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Jong-Shyan Wang, PhD, Dept of Physical Therapy, Chang Gung University, 259 Wen-Haw 1st Rd, Kwei-Shan, Tao-Yuan 333, Taiwan, e-mail: [email protected]. 0003-9993/02/8307-6970$35.00/0 doi:10.1053/apmr.2002.33235

dent dilation in skin vasculature. Dysfunction in cutaneous microcirculation tends to be more pronounced in the edematous than in the nonedematous extremities. Key Words: Edema; Endothelium; Hemiplegia; Rehabilitation; Vasodilation. © 2002 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation DEMA, THE PRESENCE OF large quantities of fluid in the extracellular spaces, is likely to develop in extremities E that are in a dependent position or inactive because of weakness, paralysis, or pain.1,2 Hand edema of paretic extremities in patients with hemiplegia may occur as an isolated problem or as part of the shoulder-hand syndrome.3 Earlier investigations3,4 found an incidence of isolated hand edema ranging from 16% to 82.5% after stroke. Although edema of paretic extremities is frequently encountered by medical professionals involved in rehabilitating stroke patients, studies of poststroke hemodynamic changes in persons with paretic edematous extremities have been either controversial or incomplete.2,5-7 Although autonomic vasomotor dysfunction after stroke may open the vascular bed and increase blood flow to various tissues, prolonged inactivity in paretic extremities can also decrease blood flow rather than increase it.2,5-7 Therefore, the mechanisms underlying hemodynamic changes in paretic extremities of hemiplegic patients with or without edema deserve further investigation. Endothelium-derived relaxing factor (EDRF) has been identified as nitric oxide (NO) or a closely related molecule synthesized from the guanidine group of L-arginine.8 Previous studies9,10 showed that inhibiting NO production and bioactivity leads to vascular protein leakage and increases microvascular permeability. The release and bioactivity of EDRF and NO were increased by physical conditioning: conversely, they were decreased by deconditioning.11,12 Accordingly, we hypothesized that muscular flaccidity or disability of the paretic upper extremity after a stroke may be accompanied by impaired peripheral circulatory function, possibly brought on by decreasing EDRF and NO bioactivity or release. Therefore, paretic and nonparetic upper extremities in patients with hemiplegia may respond differently to endothelium-mediated hemodynamic functions. The present study investigated the role of cutaneous endothelial vascular control in paretic extremities after a stroke. We hypothesized that this investigation would elucidate the possible mechanisms of edematous development in paretic extremities after a stroke. Our intent was to clarify peripheral circulatory function and its underlying mechanisms in persons with edematous and nonedematous paretic upper extremities at the flaccid stage after a stroke. To meet this goal, we measured basic hemodynamic characteristics in arterial and venous vessels by impedance plethysmography. To assess cutaneous microcirculatory function, we conducted a laser Doppler perfusion in which we measured the responses of noninvasive cutaneous microcircuArch Phys Med Rehabil Vol 83, July 2002

1018

STROKE AND HEMODYNAMIC CHANGES, Wang

lation by transdermal iontophoresis of acetlycholine (ACh) and sodium nitroprusside (SNP) as specific endothelium-dependent and independent vasodilators, respectively.13 Changes in skin microcirculation during NO synthase inhibition by NG-monomethyl-L-arginine (L-NMMA) were also measured. METHODS Participants The protocol was reviewed and approved by an institutional ethics committee to protect the human rights of the participants. All hemiplegic patients were admitted to the rehabilitation ward of Chang Gung Memorial Hospital, and within a month of their stroke onset were evaluated for inclusion in this investigation. A standard medical history was obtained, including details of the stroke and a history of any upper-extremity problems, such as degree of muscle activity and hand edema. Hemiplegic patients with flaccid muscles in the paretic upper extremity were included while patients with recurrent stroke or with peripheral vascular diseases were excluded. For ethical reasons, the subjects were allowed to take regular drugs until 12 hours before the study was conducted. A total of 53 patients (28 men, 25 women; age range, 42–7y; mean age, 58.2⫾3.8y) were studied after they had given informed consent and had the experimental procedures explained to them. Twenty-seven patients had cerebral infarction and 26 had cerebral hemorrhage. The subjects were divided into 2 groups: edema (15 men, 14 women; 58.4⫾3.3y) and nonedema (13 men, 11 women; 58.1⫾3.8y). The edema group included 29 patients with edematous paretic upper extremities, ascertained through the following clinical criteria: (1) dorsal swelling in the hand over the carpal bones; (2) fusiform edema of the metacarpophalangeal and proximal interphalangeal joints; (3) changes in temperature, color, or dryness; and (4) loss of dorsal skin lines and potential change of fingernails.2 The 24 hemiplegic patients comprising the nonedema group did not have limb edema in the paretic upper extremity. Testing of Resting and Hyperemic Arterial Inflow We measured hemodynamic characteristics in the upper extremities of hemiplegic patients by impedance plethysmography.14,a Subjects were permitted 10 minutes of supine rest before testing. The measuring electrodes were placed on the volar side of the forearm, spaced approximately 10cm apart. The arms were raised to position the forearm above the heart. To quantitate arterial inflow at rest, we set the subdiastolic occlusion pressure in the arm at 60mmHg. By using this examination protocol, we conducted 3 occlusion procedures in sequence, each lasting 10 seconds. In subjects with reactive hyperemia caused by a complete arterial occlusion, we determined the arterial inflow over a 4-minute time period. At the beginning of the occlusion phase, cuff pressure was set to a brachial artery systolic pressure of ⫹50mmHg, a level at which a complete occlusion can be expected. After the suprasystolic occlusion phase, the blood flow profile was determined by a sequence of 7 measuring cycles (each with 15s of occlusion and a 5-s refilling phase) in which a subdiastolic cuff pressure was applied. Testing of Venous Capacity and Compliance The subject’s arms were raised at an angle of about 30° from the examination couch. Three different pressures were applied to the occlusion cuff: first 40, then 60, and finally 80mmHg. The venous system was full until the pressure in the veins exceeded the transmural pressure caused by the cuff. The cuff Arch Phys Med Rehabil Vol 83, July 2002

was then emptied to allow the venous blood to flow out, and the venous outflow was recorded over 45 seconds. The venous capacity was calculated by the blood volume accumulated in the extremity during the artificial venous occlusion. Venous tone described the dependence of the venous volume on the venous pressure (50mmHg). Meanwhile, venous outflow described the velocity of the change in blood volume during the first 2 seconds after opening the venous occlusions. Finally, flow resistance described the vascular resistance of the proximally located veins against the venous blood flow. Cutaneous Vascular Responsiveness to Agonist We performed laser Doppler perfusion noninvasively to measure how the forearm skin vasculature responded to transdermal iontophoresis of polar drugs.13 After cleaning the subject’s skin with isopropyl alcohol and drying it in air, we affixed a combined probe holder, for iontophoresis and perfusion measurement, on the volar side of forearm by using double-sided adhesive tape. The probe holder had a small chamber, which was closed to the laser Doppler probe,b for depositing the test solutions. A battery-powered, constant-current stimulatorc supplied a direct current for the drug iontophoresis. The active electrode was made of platinum and charged according to the active ions of the drug. A 1% solution of ACh chloride (solvent: deionized water)d and 1% SNP (solvent: deionized water)e were used. The ACh anodal current was used to transfer the cation during iontophoresis, and SNP cathodal current was applied to transfer the anion, as described elsewhere.13 In some experiments, endothelium-derived NO was blocked by coadministration with 1% L-NMMA citratef during ACh-induced cutaneous perfusion. Figure 1 shows the dose-response curve for ACh and SNP by using charges of 2.0mC (0.1mA for 20s), 4mC (0.2mA for 20s), and 8mC (0.2mA for 40s) in which the response was measured for 300 seconds after each charge of iontophoresis. Different dosages of the same substance were applied at the same location, and the test positions for ACh and SNP were separated at 5cm. Statistical Analysis Data were expressed as a mean ⫾ standard error of the mean. The statistical software package StatView® 4,g running on a Macintosh computer, was used to analyze the data. The results were analyzed by 2-factor analysis of variance followed by Tukey multiple comparison. Differences were considered significant at P less than .05. RESULTS Hemodynamic Measurements in Arterial and Venous Vessels In the edema and nonedema groups, and in the combined total group, there were no significant differences in basic hemodynamic variables, such as venous outflow, capacity, tone, and flow resistance, between nonparetic and paretic upper extremities (table 1). Although resting arterial inflow in a paretic extremity did not significantly differ from that in a nonparetic extremity, reactive hyperemia, indicated by a hyperemic arterial inflow, was lower in paretic than in the nonparetic extremity (fig 1A). However, the extent of decreased reactive hyperemia in the paretic extremities of the edema group was similar to that in the nonedema group (fig 1B). Cutaneous Vascular Responsiveness to Agonists From comparable levels of basal cutaneous perfusion, graded iontophoretic administration of ACh, ACh plus

1019


Fig 2. Laser Doppler perfusion of response to increasing concentrations of ACh, ACh plus L-NMMA, and SNP. The dose-response curves were made by using the charges of 2, 4, and 8mC. The response measuring period for each dose was 300 seconds.

Fig 1. Comparison of resting and hyperemic arterial inflow between paretic and nonparetic extremities in total (A) edema and (B) nonedema groups. Although resting arterial inflow in a paretic extremity did not significantly differ from that in a nonparetic extremity, inflow was lower in paretic than in nonparetic extremity (A). However, the extent of decreased reactive hyperemia in the paretic extremities of the edema group was similar to that in the nonedema group (B). Abbreviations: T-NP, nonparetic extremities in total group; T-P, paretic extremities in total group; NE-NP, nonparetic extremities in nonedema group; NE-P, paretic extremities in nonedema group; E-NP, nonparetic extremities in edema group; E-P, paretic extremities in edema group. *P < .05 (paretic vs nonparetic extremities).

L-NMMA, and SNP resulted in successive increases in perfusion (fig 2). Paretic extremities displayed a lower ACh-induced cutaneous perfusion than nonparetic extremities in all subjects (figs 3A, B). The degree of decreased ACh-induced cutaneous perfusion in the paretic extremity of the edema group was larger than in the nonedema group (fig 4B). Paretic extremities

displayed a lower ACh plus L-NMMA–induced cutaneous perfusion than nonparetic extremities in all subjects (figs 3C, D). However, the degree of decreased ACh plus L-NMMA– induced cutaneous perfusion in the paretic extremities of the edema group was similar to that in the nonedema group (figs 4C, D). For SNP-induced cutaneous perfusion, no significant difference existed between nonparetic and paretic extremities in all subjects (figs 3E, F). Furthermore, perfusion in the edema group resembled that in the nonedema group (figs 5A, B). DISCUSSION To our knowledge, the present investigation is the first to show clearly that the paretic upper extremities suffer impaired cutaneous microcirculatory function after cerebrovascular accident (CVA), possibly because of decreased ability to enact endothelium-dependent dilation in the skin vasculature. Moreover, this decrease tended to be more pronounced in the edematous than in the nonedematous extremities. During the flaccid stage after a stroke, the patient is especially susceptible to persistent edema, which can cause an extremity to become painful, disfigured, and disabled.4 Previous investigations have suggested that edematous development in paretic extremities in hemiplegic patients may result from the lack of active movements and/or autonomic vasomotor dysregulation.2,5-7 Possibly, reduced central sympathetic tone relaxes vascular tone and increases blood flow, elevating fil-

Table 1: Basic Hemodynamic Characteristics of the Upper Extremities in All Groups Total

Arterial inflow (rest, %/min) Venous outflow (2s, %/min) Venous capacity (%) Venous tone (50mmHg, %/mmHg) Flow resistance (mmHg 䡠 % 䡠 min)

Nonedema

Edema

NPE

PE

NPE

PE

NPE

PE

3.99⫾0.33 54.5⫾8.4 5.01⫾0.54 0.028⫾0.001 1.19⫾0.17

3.82⫾0.23 47.7⫾5.4 4.04⫾0.33 0.030⫾0.010 1.27⫾0.20

4.35⫾0.41 53.04⫾13.0 5.46⫾0.98 0.027⫾0.006 1.22⫾0.26

3.83⫾0.0.52 53.9⫾9.1 4.56⫾0.62 0.028⫾0.009 1.15⫾0.28

3.54⫾0.54 55.7⫾13.8 4.40⫾0.59 0.029⫾0.007 1.18⫾0.24

3.81⫾0.44 43.1⫾4.5 3.67⫾0.45 0.031⫾0.010 1.36⫾0.15

NOTE. Values are mean ⫾ standard error of the mean. Abbreviations: NPE, nonparetic extremity; PE, paretic extremity.

Arch Phys Med Rehabil Vol 83, July 2002

1020


Fig 3. Comparison of cutaneous vascular responsiveness to (A, B) ACh, (C, D) ACh plus L-NMMA, and (E, F) SNP between paretic and nonparetic extremities in total group. Paretic extremity had a lower ACh- and ACh plus L-NMMA– induced cutaneous perfusion than nonparetic extremity. However, with SNP-induced cutaneous perfusion, no significant difference existed between nonparetic and paretic extremity. *P < .05 (paretic extremity vs nonparetic extremity).

tration pressure and altering capillary permeability.6,7 In contrast, deconditioning, inactivity, and paralysis may increase peripheral vascular resistance and decrease arterial blood flow.5,15 Studies of poststroke hemodynamic changes of paretic extremities have been controversial or incomplete.2,5-7 According to our present investigation, resting arterial inflow venous capacity, compliance, and outflow did not significantly differ between paretic and nonparetic upper extremities in hemiplegic patients. We found similar results in edematous paretic compared with nonedematous paretic extremities. On the other hand, hyperemic arterial inflow was lower in paretic than nonparetic extremities. Sinoway et al16 suggested that the abnormality in vasodilatory capacity in congestive heart failure may be related to physical deconditioning. Karhunen et al15 indicated that the decreased peripheral vascular resistance and the enhanced blood flow gained during exercise training were attenuated after a detraining period. Based on data from studies on deconditioning in heart failure patients and healthy volunteers, we think it is likely that deconditioning contributed to our Arch Phys Med Rehabil Vol 83, July 2002

finding of diminished reactive hyperemia in the paretic arms. However, the decrease in hyperemic arterial inflow after stroke did not differ significantly between paretic extremities with and without edema in the present study. Mechanisms other than hemodynamic characteristics in venous function or arterial inflow levels may be involved in the edematous development of paretic extremities in hemiplegic patients. The continuous release of NO appears to be an important intrinsic modulator of blood flow in various tissues.8 As widely believed, it is continuously released by the vascular endothelium, and large amounts of NO release can be evoked by administering ACh in humans in vivo.8 Previous studies also found that NO can modulate microvascular permeability.9,10 Inhibiting NO production has been shown to (1) expand the junctions between adjacent capillary endothelial cells,17 (2) increase the formation of release factors of platelet-leukocyte aggregates (possibly in response to platelet-activating factor), with a resultant increase in endothelial cell contraction and vascular protein leakage,10 and (3) permit superoxide to accu-


1021

Fig 4. Comparison of cutaneous vascular responsiveness to (A, B) ACh and (C, D) ACh plus L-NMMA between paretic and nonparetic extremities in edema and nonedema groups. Paretic extremity had a lower ACh- and ACh plus L-NMMA–induced cutaneous perfusion than nonparetic extremity in both groups. The attenuated ACh-induced cutaneous perfusion of paretic extremity meant that the edematous arms’ responses were much greater than those of the nonedematous arms. However, the degree of decreased ACh plus L-NMMA–induced cutaneous perfusion in the paretic extremities of the edema group was similar to that in the nonedema group. *P < .05 (paretic vs nonparetic extremities); ⴙ P