- Email: [email protected]
Patterns of Skin Flowmotion in the Lower Limbs of Patients with Chronic Critical Limb Ischaemia (CLI) and Oedema
Eur J Vasc Endovasc Surg 20, 536–544 (2000) doi:10.1053/ejvs.2000.1225, available online at http://www.idealibrary.com on
Patterns of Skin Flowmotion in the Lower Limbs of Patients with Chronic Critical Limb Ischaemia (CLI) and Oedema M. D. Anvar∗1,2, H. Z. Khiabani1,2, A. J. Kroese2 and E. Stranden1 Departments of Vascular 1Diagnosis and Research and 2Vascular Surgery, Aker Hospital, University of Oslo, Norway Objective: to study flowmotion (FM) in lower limbs with critical limb ischaemia (CLI) and oedema and to elucidate FM patterns when skin viability is threatened. Material and methods: fourteen patients with unilateral CLI and oedema and two control groups were included – one consisting of 10 healthy participants and the other nine patients with unilateral CLI without oedema. Laser Doppler was used to evaluate the foot skin microcirculation simultaneously at four different areas, with the limbs in supine and dependent position. FM was expressed using fast Fourier transformation (FFT) as low frequency (LF) and high frequency (HF) waves and their respective FFT-powers. Results: all patients with CLI, both with and without oedema, showed HF waves in both diseased and contralateral limbs. These were absent in healthy controls. There were no regional differences in frequency in the critically ischaemic feet (with and without oedema) and between ischaemic and their contralateral feet. Changing the position of ischaemic limbs from supine to dependency had no significant effect on the frequency, while a significant increase of the median FFT-powers of LF and HF waves at the pulp of the first toe was observed. This manoeuvre resulted in decrease of the median FFT-powers of LF in healthy controls. Conclusions: HF waves are associated with CLI. Ischaemia also appears to influence the FFT-power of each frequency domain. Ischaemic oedema does not seem to affect the pattern of FM in the foot of patients with CLI. Key Words: Chronic critical limb ischaemia; Ischaemic oedema; Flowmotion; Laser-Doppler fluxmetry.
Introduction Oedema is commonly observed in limbs with critical limb ischaemia (CLI) and may be related to the severity of ischaemia.1–6 Electron microscopic studies of skin in patients with CLI and oedema showed that the arterioles were dilated. The same study demonstrated an uneven distribution of erythrocytes in dilated capillaries.2 The vasodilatation of arterioles may occur as a consequence of ischaemia per se and a disturbance in the venoarteriolar reflex (VAR) in the foot of these patients.7–10 The cause of an uneven distribution of erythrocytes in these dilated capillaries is not understood. It might be related to other regulatory mechanisms which try to compensate for the “low flow state” in critically ischaemic tissue, e.g. “flowmotion”. Flowmotion (FM) is a rhythmical, cyclic variation
∗ Please address all correspondence to: M. D. Anvar, Department of Vascular Surgery, Surgical Clinic, Aker Hospital, University of Oslo, 0514 Oslo, Norway. 1078–5884/00/120536+09 $35.00/0 2000 Harcourt Publishers Ltd.
of blood perfusion in a tissue. Three frequency components of FM can be distinguished: (1) low frequency waves (LF: 1–10 cycles per min), (2) high frequency waves (HF: 10–25 cycles per min), (3) pulsations related to cardiac rhythms.11 Its underlying mechanism is called vasomotion, and is due to the changes in smooth muscle cell tone in the wall of small arteries and arterioles.12–15 The precise physiology of vasomotion is not completely understood, but neurogenic, myogenic, humoral, and metabolic factors may all play a role.16 Since oedema probably adds to the severity of CLI, and FM supposedly promotes homeostasis of the microcirculation,17,18 the purpose of the present study was to elucidate the possible consequences of local oedema formation on FM in critically ischaemic skin.
Patients Fourteen patients with unilateral CLI and oedema were included.19 There were nine women and five men with a mean age of 77±9.8 years. The control group
Patterns of Skin Flowmotion in the Lower Limbs
537
Table 1. Summary of patients’ characteristics and vasoactive medication.
4 ch. laser Doppler
Variable
Pressure monitor
CLI with oedema CLI without oedema (n=14) (n=9)
Rest pain Ischaemic skin ulcer Gangrene Signs of infection Ischaemic heart disease Hypertension Prior stroke Pulmonary disease Smoking Antiplatelet drugs Warfarin ∗ ACE-inhibitors -Blocking agents Calcium antagonists
14 6 8 0 9 9 4 1 13 10 4 4 5 5
9 5 4 0 5 4 2 1 7 7 2 1 1 3
∗ Angiotensin Converting Enzyme.
was six women and three men with a mean age of 75±11 with unilateral CLI but without oedema (Table 1). Patients who previously had undergone vascular surgery, those with an amputated contralateral limb, bilateral CLI, clinical signs of congestive heart failure, diabetes mellitus, manifest venous insufficiency or lymph oedema were excluded. Smokers had abstained from smoking since admission to the hospital. Another age-matched control group, 10 healthy voluntary subjects, seven women and three men, without clinical signs of peripheral atherosclerosis and a normal ankle brachial blood pressure index (ABI) were included. The study was accepted by the Regional Ethical Committee of Southern Norway. All patients gave their written consent.
Methods Ankle and brachial pressures index (ABI) was measured with ultrasound Doppler technique.20,21 In the limbs where Doppler signals were not detected, ankle pressure was defined at 15 mmHg.
Laser Doppler fluxmetry Two laser Doppler (LD) fluxmeters (Periflux PF 4001 and 4002, Perimed AB Ja¨rfa¨lla, Sweden),22 with two channels each was used to simultaneously evaluate the microcirculation at four different areas of the skin. The depth of measurement by using PF 4001 and 4002 is approximately 0.5–1 mm, but varies depending on
M
4
Bed
2
3
1 Fig. 1. Schematic illustration of the experimental setup with the laser Doppler fluxmeter probes attached to the limbs: the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2) and at the anterolateral part of ankle (Site 3). A fourth probe was attached to the first toe of the contralateral limb as a reference. “M” is a water filled, op en ended manometer line connected to a pressure monitor for recording of variations in foot position.
blood content of the tissue, degree of oxygenation and skin pigmentation.22 The laser Doppler flux (LDF) values, expressed as perfusion units, (PU) were saved by means of a Perisoft acquisition and analysis software program (Perimed AB) in a computer for offline analysis.
Recording of foot position To record alteration in foot position, a water-filled, open-ended manometer line connected to a pressure monitor (Pressure Monitor, Stranden) was attached to the skin at the dorsum of the foot. This device recorded vertical movements of the foot during the measurements. Changes in pressure indicated variations in the height of the foot.
Procedure The laser probes were applied at four different areas by using a probe holder and double adhesive tape (Fig. 1): the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2), at the anterolateral part of ankle (Site 3), the pulp of the first toe of the contralateral foot (Site 4). Measurements were performed after 15 min resting period with the subjects in supine position, in a room with a temperature between 23–25 °C. The LDF values were then continuously recorded Eur J Vasc Endovasc Surg Vol 20, December 2000
538
M. D. Anvar et al.
for 3 min, and the mean value of the last 2.5 min during that period represented the LDF value for the supine position. Subsequently, the CLI limb was passively lowered by flexion in the knee joint with the three laser probes still being attached (Fig. 1). The LDF values were recorded for 3 min and the final 2.5 min represented the LDF values for the dependent limb. The limb was lifted up again to the horizontal position to repeat the same measurements. This procedure was repeated three times for each patient with a total sampling time of more than 18 min for each participant.
site, and between CLI limbs and the contralateral limbs, and between the limbs in the control group. Mann–Whitney and ANOVA tests were used to compare differences in median frequency waves and FFTpower for both LF waves between limbs with CLI, with and without oedema, and healthy control group, both in horizontal and dependent position. The results were presented as medians with ranges in parentheses, and mean±standard deviation (s.d.), with p<0.05 as the level of statistical significance. GraphPad Prism 3.0 was used for analysis and presentation of data (GraphPad Software, Inc., San Diego, CA, U.S.A., www.graphpad.com).
Data analysis The laser Doppler (LD) signal was transformed to a frequency domain by fast Fourier transformation (FFT, Perisoft for Windows ver. 1.19, Perimed AB, Ja¨rfa¨lla, Sweden).22 The software allowed changes in time scale (x-axis) and signal amplification (y-axis) to facilitate the analyses. Various physiological activities which occur in time (“attractors”), such as cardiac activity, respiration, and variations in the microcirculation caused by neuronal or myogenic activity, may be represented in different frequency regions following the FFT. The results of FFT are expressed as frequency (cycles per second) of the attractors and FFT-power (“intensity”, arbitrary units, Fig. 2) of each of the frequency regions.23,24 There is a positive correlation between the derived FFTpower in both LF and HF regions and skin perfusion (LD) at the different measuring sites. A high power indicates that a specific frequency occurs regularly with a LD signal well above the noise level, and acts as a “quality-parameter”.
Leg and foot volume measurements The distal leg and foot volumes were measured by using water displacement volumetry (WDV) as previously described in detail.3 The differences in distal leg and foot volume between the two limbs were calculated.
Statistics Wilcoxon matched-pairs signed-ranks test was used for comparing changes in ABI, ASP, median frequency waves and FFT-power for both LF and HF waves between horizontal and dependent position in each Eur J Vasc Endovasc Surg Vol 20, December 2000
Results There was no statistically significantly difference in mean ankle brachial index (ABI) and ankle systolic pressure (ASP) between the critically ischaemic limbs with (0.19±0.13 with a mean ASP of 28±18 mmHg) and without oedema (0.27±0.14, and 35±20, both p>0.05). In the healthy control group, mean ABI was 0.94±0.28 with a mean ASP of 121±49 mmHg. In seven of 14 limbs with CLI and oedema and three of nine limbs with CLI and without oedema ultrasound Doppler signals could not be detected, and blood pressure at the ankle was defined at 15 mmHg. In patients with CLI with and without oedema, three distinct frequency domains were detected and are denoted respectively as low frequency (LF) waves, high frequency (HF) waves, and cardiac activity related, i.e. heart rate (HR) waves, while high frequency (HF) waves were absent in the healthy control group in all measured sites.
Low frequency (LF) waves With the limbs in horizontal position, there were no significant differences in median frequency of LF waves between the patients with CLI or between patients, and the healthy control group. There were no significant differences in median frequency of LF waves between the limbs within each of these groups either (Fig. 3). Altering the position of the ischaemic limbs from horizontal to dependency did not significantly change the median frequency of LF waves in any of the three groups (Fig. 3). The changes in position resulted, however, in a significant increase of the median FFT-power of LF
Patterns of Skin Flowmotion in the Lower Limbs
539
A
LDF (PU) 120
60
0 Horizontal
Pr.
Dependent
1 min.
B
Ti m
e
(m in
.)
Power
Frequency (Hz) Fig. 2. (A) A sample of recorded laser Doppler flux (LDF) with the limb in horizontal and dependent position in a healthy control participant. In the panel at the top a portion of the curve is expanded to better distinguish fluctuations caused by cardiac activity from flowmotion. (B) A 3D plot of spectrum analysis by fast Fourier transformation (FFT) in the same temporal selection of the curve is shown in the respective time and frequency domains. Pr: manometer line pressure, PU: perfusion unites.
Eur J Vasc Endovasc Surg Vol 20, December 2000
540
M. D. Anvar et al.
Site 3
Healthy controls
Horizontal
CLI without oedema
Dependent
CLI with oedema
10 8
6 4 2 Power
0 0
3
1
2 3 4 Frequency
2
5
1
(cycles/min.)
Site 2
10 8
6 4 2 Power
0 0
Dependent
Dependent
Horizontal
Horizontal
Site 1
1
2 3 4 Frequency
5
(cycles/min.)
10 8
6 4 2 Power
0 0
1
2 3 4 Frequency
5
(cycles/min.)
Fig. 3. Laser Doppler measurements of low frequency (LF) frequency waves with the limb in the horizontal and dependent position in healthy controls, CLI without oedema and CLI with oedema. There were three measuring sites: the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2) and the anterolateral part of the ankle (Site 3). Frequencies are expressed in cycles/min, including FFT-power of each of the frequency regions. Both values are given as medians and range.
[0.12 (0.015–3.3) vs. 0.82 (0.06–5.1), p<0.005] at the pulp of the first toe (Site 1) in the ischaemic limbs while unchanged in the other two sites (Fig. 3). In healthy controls this manoeuvre resulted in a decrease of the median FFT-power of LF in all three sites; site 1 [5.9 (0.8–11.1) vs. 3.7 (0.48–9.0), p<0.03], site 2 [1.8 (1.1–4.4) vs. 0.8 (0.25–3.5), p<0.002], site 3 [1.4 (0.8–2.9) vs. 0.6 (0.24–3.1), p<0.006]. The median FFT-power of LF of ischaemic limbs in site 1, 0.12 (0.015–3.3, Fig. 3), with the limbs in horizontal position was significantly lower than both the contralateral limbs 3.8 (0.34–8.15), p<0.001 and healthy control limbs, 5.9 (0.8–11.1), (p<0.0001) (Fig. 4). Eur J Vasc Endovasc Surg Vol 20, December 2000
High frequency (HF) waves All patients with CLI, (with and without oedema) showed HF waves in both diseased and contralateral limbs, while HF waves were absent in the healthy control group. With the limbs in horizontal position, there were no significant differences in median HF waves between the ischaemic limbs and the contralateral side. Altering the position of the CLI limbs from horizontal to dependency did not significantly change the median frequency of HF waves in the CLI feet (Fig. 5).
Patterns of Skin Flowmotion in the Lower Limbs
541
Contralateral limbs (Site 4) of: Healthy controls CLI patients without oedema
Low frequency
High frequency
CLI patients with oedema
10 8
6 4 2 Power
0 0
1
2 3 4 Frequency
5
1.0 .8
.6 .4
Power
.2
0 0
4
8
12 16 20
Frequency
(cycles/min.)
(cycles/min.)
Fig. 4. Laser Doppler measurements of low frequency (LF) and high frequency (HF) frequency waves of the contralateral limbs (Site 4) of the ischaemic limbs, with and without oedema, in the horizontal position. None of the healthy control limbs showed HF-waves. Frequencies are expressed in cycles/min, including FFT-power of each of the frequency regions. Both values are given as medians and range.
By lowering the ischaemic limbs the median FFTpower of HF increased only at the pulp of the first toe in both patients groups 0.08 (0.02–0.38) vs. 0.39 (0.09–1.45), (p<0.001) (Fig. 5). The median FFT-power of HF in the ischaemic limbs were significantly lower than the contralateral limbs [0.06 (0.01–0.12) vs. 0.2 (0.06–1.6), p<0.01)].
Heart rate (HR) waves The median rate of heart waves in patients with CLI and oedema was 72 (58–81), and without oedema 74 (58–84). In the supine position, HR waves were present in all contralateral sites i.e. site 4, in 50% of site 3 but not detected in the analysis of sites 1 and 2 in all CLI limbs both with and without oedema. Lowering the ischaemic foot resulted in appearance of HR waves in all sites. All subjects in the control group showed heart pulsatile flux with the limbs both in supine and dependent position at all measured sites with a median frequency of 68 (48–75) (Fig. 2).
Leg volume There was a mean leg–foot volume difference of 11±7% between limbs with CLI and oedema and contralateral sides measured by WDV. On the other
hand, we did not find any significant differences between the limbs with CLI and without oedema and contralateral sides measured by WDV.
Discussion The LF waves were as frequently observed in ischaemic as in healthy feet. However, HF waves were exclusively seen in the patients with CLI. This finding is in accordance with a previous study and indicates that the prevalence of HF waves is associated with a certain degree of ischaemia.25 This interpretation is supported by the observation that the number of HF waves decreases after successful angioplasty for severe ischaemia.26 Several investigators have reported that LF waves are associated with diameter changes in larger arterioles, and HF waves with diameter changes of terminal arterioles.16,27 Morphological studies by transmission electron microscopy have recently demonstrated dilated terminal skin arterioles in the feet of patients with CLI and oedema.2 However, it is a matter of speculation whether these findings are related to the occurrence of HF waves. Besides, it is not known how HF waves may contribute in compensating for the drastically decreased blood flow in CLI limbs. There were no significant differences in occurrence of LF or HF waves between ischaemic limbs with and without oedema and between the contralateral feet. Thus, it seems that ischaemic oedema formation does Eur J Vasc Endovasc Surg Vol 20, December 2000
542
M. D. Anvar et al.
Site 3 Horizontal
CLI without oedema
Dependent
CLI with oedema
1.0 .8
.6 .4
Power
.2
0 0
3
4
8
2
12 16 20
Frequency
1
(cycles/min.)
Site 1
Dependent
Dependent
Horizontal
Horizontal
Site 2
1.0 .8
.6 .4
Power
.2
0 0
4
8
12 16 20
Frequency
(cycles/min.)
1.0 .8
.6 .4
Power
.2
0 0
4
8
12 16 20
Frequency
(cycles/min.)
Fig. 5. Laser Doppler measurements of high frequency (HF) frequency waves with the limb in the horizontal and dependent position in patients with CLI without oedema and CLI with oedema. There were three measuring sites: the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2) and the anterolateral part of the ankle (Site 3). Frequencies are expressed in cycles/min, including FFT-power of each of the frequency regions. Both values are given as medians and range.
not affect the manifestation of LF and HF waves. The ranges for both LF and HF waves in the present study are comparable with findings of previous studies.11,26 The cause and effect of LF and HF waves are still unknown. Both central and local nervous regulation may play a role.18 The central sympathetic nervous activity promotes homeostasis of blood pressure by constricting the arterioles, a local regulating system maintains blood perfusion of tissues by dilatating the arterioles. However, we found in this study that the distribution of the frequency waves is almost constant in different ischaemic areas of the foot. Furthermore, there was no difference in the frequency of LF and HF waves in the ischaemic limbs and the contralateral sides. These findings may indicate the superimposition of a local metabolic vasodilatory effect of hypoxia on centrally controlled vasoconstriction. Eur J Vasc Endovasc Surg Vol 20, December 2000
There was a positive correlation between the FFTpower and perfusion as measured by LDF, in all frequency domains. The power of LF waves was three times greater in the contralateral and five times greater in the healthy control limbs, compared to the ischaemic limbs. Therefore, it seems that the intensity of flowmotion curves, as indicated by the FFT-power, is related to the severity of ischaemia. A decrease of FFTpower probably varies in parallel with a reduced perfusion of the microcirculation in a certain area. Again, exactly how this effect could contribute to maintain homeostasis in ischaemic tissue is still unclear. Although there were no regional differences in LF or HF waves in the ischaemic feet, either with or without oedema, there were regional differences in the FFT-power for both LF and HF waves in the
Patterns of Skin Flowmotion in the Lower Limbs
ischaemic foot for both horizontal and dependent position. The power was lowest for both frequency waves at site 1 with the ischaemic feet in horizontal position, while the power in ischaemic feet in dependency increased more than 5-fold for both frequency domains at the pulp of the first toe (site 1). In contrast, the power of LF waves in healthy subjects was greatest at site 1 and lowest at the ankle, and lowering the limbs reduced power in all sites (Fig. 2). The effect of dependency and the regional difference in FFT-power of both LF and HF waves could be ascribed the different density of arteriovenous anastomoses (AVA) in the pulp of the toe as compared to site 2 and 3. The pulp of the toe contains the highest number of AVA per skin area,8,28 and hence possesses the greatest potential for varying skin perfusion. This is illustrated by the finding of a more pronounced increase in perfusion and FFT-power in site 1 compared with site 2 and 3, respectively (Figs 3 and 5). Several investigators have previously demonstrated that lowering an ischaemic leg below heart level improves the blood supply of poorly perfused skin regions,7,29,30 while lowering a healthy limb reduces the skin perfusion. There might be several possible physiological mechanisms that are involved in this phenomenon: (1) It has been postulated that paralysed arterioles and collapsed capillaries might be passively distended or even reopened by elevated transmural pressure when the leg is lowered.31 (2) A disturbed VAR may improve the blood perfusion due to lowering the critically ischamic limb, while it reduces blood perfusion in normal limb in dependency.7,10 (3) Subdermal thermoregulatory blood flow might be directed to nutritional capillaries during dependency in limbs with CLI.30,32 Furthermore, blood supply from the proximal leg may be distributed to acral skin regions during dependency.33 Thus, there are indications that redistribution of the microcirculation plays a compensatory role in maintaining optimal perfusion in severely ischaemic tissues. In conclusion, HF waves are typically associated with a state of extreme skin ischaemia. Ischaemia also appears to influence the FFT-power of each frequency domains. It seems that ischaemic oedema per se does not affect the pattern of FM in the foot of patients with CLI. Furthermore, the study showed that alteration in the microcirculation when lowering an ischaemic limb occurs as enhancement of the FM-waves (both occurrence and amplitude), while the frequency domains remain unchanged.
543
References 1 Khiabani HZ, Anvar MD, Rostad B, Stranden E, Kroese AJ. The distribution of oedema in the lower limb of patients with chronic critical limb ischaemia (CLI). A study with Computed Tomography (CT). VASA 1999; 28: 265–270. 2 Anvar MD, Khiabani HZ, Nesland JM, Kroese AJ, Stranden E. Vascular and stromal features in the skin of the lower limb in patients with chronic critical limb ischaemia (CLI) and oedema. Eur J Vasc Endovasc Surg 2000; 20: 125–131. 3 Khiabani HZ, Anvar MD, Stranden E, Slagsvold CE, Kroese AJ. Oedema in the lower limb of patients with chronic critical ischaemia (CLI). Eur J Vasc Endovasc Surg 1999; 17: 419–423. 4 Khiabani HZ, Anvar MD, Kroese AJ, Stranden E. Transcapillary forces in the development of oedema in the lower limb of patients with chronic critical limb ischemia (CLI). Eur J Vasc Endovasc Surg 2000; 19: 598–604. 5 Anvar MD, Khiabani HZ, Kroese AJ, Stranden E. Alterations in capillary permeability in the lower limb of patients with chronic critical limb ischaemia (CLI) and oedema. VASA 2000; 29: 106–111. 6 Anvar MD, Khiabani HZ, Lande K, Kroese AJ, Stranden E. The concentration of protein-compounds in interstitial tissue of patients with chronic critical limb ischaemia (CLI) and oedema. VASA 2001; 30: in press. 7 Henriksen O, Paaske WP. Local regulation of blood flow in peripheral tissue. Acta Chir Scand 1980; 502: 63–74. 8 Midttun M, Sejrsen P, Paaske WP. Peripheral blood flow rates and microvascular responses to orthostatic pressure changes in claudicants before and after revascularisation. Eur J Vasc Endovasc Surg 1999; 17: 225–229. 9 Ubbink DT, Kitslaar PJ, Tordoir JH et al. The relevance of posturally induced microvasular constriction after revascularisation in patients with chronic leg ishaemia. Eur J Vasc Surg 1992; 6: 525–532. 10 Khiabani HZ, Anvar MD, Kroese AJ, Stranden E. The role of the veno-arteriolar reflex (VAR) in the pathogenesis of peripheral oedema in patients with chronic critical limb ischaemia (CLI). Ann Chir Gyn 2000; 89: 93–98. 11 Hoffmann U, Franzeck UK, Geiger M, Yanar A, Bollinger A. Variability of different patterns of skin oscillatory flux in healthy controls and patients with peripheral arterila occlusive disease. Int J Microcirc 1993; 12: 255–273. 12 Chambers R, Zweifach BW. The topography and fuction of the mesenteric capillary circulation. Am J Anat 1944; 75: 173–205. 13 Nicoll PA, Webb RL. Vascular patterns and active vasomotion as determiners of flow through minutes vessels. Angiology 1955; 6: 291–310. 14 Fagrell B, Fronek A, Intaglietta M. A microscope-television system for studying flow velocity in human skin capillaries. Am J Physiol 1977; 233: 318–321. 15 Mahler F, Muheim MH, Intaglietta M, Bollinger A, Anliker M. Blood pressure fluctuations in human nailfold capillaries. Am J Physiol 1979; 236: 888–893. 16 Meyer JU, Borgstro¨ m P, Intaglietta M. Is vasomotion due to microvascular pacemaker cells? Vasomotion and flow modulation in the microcirculation. Prog Appl Microcirc. Basel, Karger 1989, vol 15, 41–48. 17 Arfors KE, Bergqvist D, Intaglietta M, Westgren B. Measurement of blood flow in the microcirculation. Ups J Med Sci 1975; 80: 27–33. 18 Intaglietta M. Vasomotion as normal microvascular activity and a reaction to impaired homeostasis. Vasomotion and flow modulation in the microcirculation. Prog Appl Microcirc. Basel, Karger 1989, vol 15, 1–9. 19 Second European Consensus Document on Chronic Critical Leg Ischaemia. Circulation 1991; 84, Suppl. 20 Myhre HO, Kroese AJ. Ultrasound in the study of peripheral blood circulation. Acta Chir Scand 1979; Suppl. 488: 15–18. 21 Sørensen TL, Perner L, Hansen L, Schroeder TV. Can Doppler Eur J Vasc Endovasc Surg Vol 20, December 2000
544
22 23 24
25
26
27
M. D. Anvar et al.
ultrasound replace strain gauge in the measurement of ankle blood pressure? Ugeskr Laeger 1992; 154: 3662–3665. Perimed. PeriFlux 4001 Master/4002 Satellite. User’s Manual. Ja¨rfa¨lla, Sweden (www.perimed.se). Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Numerical Recipes in C. The art of scientific computing. 1997 Cambridge University Press, Cambridge. Weidenhagen R, Wichmann A, Koebe HG et al. Analysis of laser Doppler flux motion in man: comparison of autoregressive modelling and fast Fourier transformation. Int J Microcirc 1996; 16: 64–73. Bollinger A, Hoffmann U, Seifert H. Flux motion in peripheral ischemia. In: Messmer K, Hammersen F (eds). Vasomotion and flow modulation in the microcirculation. Prog Appl Microcirc. Basel, Karger 1989, vol 15, 87–92. Hoffmann U, Schneider E, Bollinger A. Flow motion waves with high and low frequency in severe ischaemia before and after percutaneous transluminal angioplasty. Cardiovasc Res 1990; 24: 711–718. Schmid-Scho¨ bein H, Zieges S, Greber R et al. Synergetic interpretation of patterned vasomotor activity in microvascular perfusion: discrete effects of myogenic and neurogenic
Eur J Vasc Endovasc Surg Vol 20, December 2000
28 29 30
31 32 33
vasoconstricion as well as arterial and venous pressure fluctuations. Int J Microcirc 1997; 17: 346–359. Hassan AAK, Tooke JE. Effect of changes in local skin temperature on postural vasoconstriction in man. Clin Science 1988; 74: 201–206. Fagrell B. The skin microsirculation and the pathogenesis of the ischemic necrosis and gangrene. Scand J Clin Lab Invest 1977; 37: 473–476. Ubbink DT, Jacobs MJHM, Slaaf DW, Tangelder GJWJM, Reneman RS. Capillary recruitment and pain relief on leg dependency in patients with severe lower limb ischemia. Circulation 1992; 85: 223–229. Gaskell P, Burton AC. Local postural vasomotor reflexes arising from the limbs vein. Circ Res 1953; 1: 27–39. Bongard O, Fagrell B. Variations in laser Doppler flux and flow motion patterns in the dorsal skin of the human foot. Microvasc Res 1990; 39: 212–222. Scheffler A, Jendryssek J, Rieger H. Redistribution of skin blood flow during leg dependency in peripheral arterial occlusive disease. Clin Physiol 1992; 12: 425–438.
Accepted 3 October 2000
Patterns of Skin Flowmotion in the Lower Limbs of Patients with Chronic Critical Limb Ischaemia (CLI) and Oedema M. D. Anvar∗1,2, H. Z. Khiabani1,2, A. J. Kroese2 and E. Stranden1 Departments of Vascular 1Diagnosis and Research and 2Vascular Surgery, Aker Hospital, University of Oslo, Norway Objective: to study flowmotion (FM) in lower limbs with critical limb ischaemia (CLI) and oedema and to elucidate FM patterns when skin viability is threatened. Material and methods: fourteen patients with unilateral CLI and oedema and two control groups were included – one consisting of 10 healthy participants and the other nine patients with unilateral CLI without oedema. Laser Doppler was used to evaluate the foot skin microcirculation simultaneously at four different areas, with the limbs in supine and dependent position. FM was expressed using fast Fourier transformation (FFT) as low frequency (LF) and high frequency (HF) waves and their respective FFT-powers. Results: all patients with CLI, both with and without oedema, showed HF waves in both diseased and contralateral limbs. These were absent in healthy controls. There were no regional differences in frequency in the critically ischaemic feet (with and without oedema) and between ischaemic and their contralateral feet. Changing the position of ischaemic limbs from supine to dependency had no significant effect on the frequency, while a significant increase of the median FFT-powers of LF and HF waves at the pulp of the first toe was observed. This manoeuvre resulted in decrease of the median FFT-powers of LF in healthy controls. Conclusions: HF waves are associated with CLI. Ischaemia also appears to influence the FFT-power of each frequency domain. Ischaemic oedema does not seem to affect the pattern of FM in the foot of patients with CLI. Key Words: Chronic critical limb ischaemia; Ischaemic oedema; Flowmotion; Laser-Doppler fluxmetry.
Introduction Oedema is commonly observed in limbs with critical limb ischaemia (CLI) and may be related to the severity of ischaemia.1–6 Electron microscopic studies of skin in patients with CLI and oedema showed that the arterioles were dilated. The same study demonstrated an uneven distribution of erythrocytes in dilated capillaries.2 The vasodilatation of arterioles may occur as a consequence of ischaemia per se and a disturbance in the venoarteriolar reflex (VAR) in the foot of these patients.7–10 The cause of an uneven distribution of erythrocytes in these dilated capillaries is not understood. It might be related to other regulatory mechanisms which try to compensate for the “low flow state” in critically ischaemic tissue, e.g. “flowmotion”. Flowmotion (FM) is a rhythmical, cyclic variation
∗ Please address all correspondence to: M. D. Anvar, Department of Vascular Surgery, Surgical Clinic, Aker Hospital, University of Oslo, 0514 Oslo, Norway. 1078–5884/00/120536+09 $35.00/0 2000 Harcourt Publishers Ltd.
of blood perfusion in a tissue. Three frequency components of FM can be distinguished: (1) low frequency waves (LF: 1–10 cycles per min), (2) high frequency waves (HF: 10–25 cycles per min), (3) pulsations related to cardiac rhythms.11 Its underlying mechanism is called vasomotion, and is due to the changes in smooth muscle cell tone in the wall of small arteries and arterioles.12–15 The precise physiology of vasomotion is not completely understood, but neurogenic, myogenic, humoral, and metabolic factors may all play a role.16 Since oedema probably adds to the severity of CLI, and FM supposedly promotes homeostasis of the microcirculation,17,18 the purpose of the present study was to elucidate the possible consequences of local oedema formation on FM in critically ischaemic skin.
Patients Fourteen patients with unilateral CLI and oedema were included.19 There were nine women and five men with a mean age of 77±9.8 years. The control group
Patterns of Skin Flowmotion in the Lower Limbs
537
Table 1. Summary of patients’ characteristics and vasoactive medication.
4 ch. laser Doppler
Variable
Pressure monitor
CLI with oedema CLI without oedema (n=14) (n=9)
Rest pain Ischaemic skin ulcer Gangrene Signs of infection Ischaemic heart disease Hypertension Prior stroke Pulmonary disease Smoking Antiplatelet drugs Warfarin ∗ ACE-inhibitors -Blocking agents Calcium antagonists
14 6 8 0 9 9 4 1 13 10 4 4 5 5
9 5 4 0 5 4 2 1 7 7 2 1 1 3
∗ Angiotensin Converting Enzyme.
was six women and three men with a mean age of 75±11 with unilateral CLI but without oedema (Table 1). Patients who previously had undergone vascular surgery, those with an amputated contralateral limb, bilateral CLI, clinical signs of congestive heart failure, diabetes mellitus, manifest venous insufficiency or lymph oedema were excluded. Smokers had abstained from smoking since admission to the hospital. Another age-matched control group, 10 healthy voluntary subjects, seven women and three men, without clinical signs of peripheral atherosclerosis and a normal ankle brachial blood pressure index (ABI) were included. The study was accepted by the Regional Ethical Committee of Southern Norway. All patients gave their written consent.
Methods Ankle and brachial pressures index (ABI) was measured with ultrasound Doppler technique.20,21 In the limbs where Doppler signals were not detected, ankle pressure was defined at 15 mmHg.
Laser Doppler fluxmetry Two laser Doppler (LD) fluxmeters (Periflux PF 4001 and 4002, Perimed AB Ja¨rfa¨lla, Sweden),22 with two channels each was used to simultaneously evaluate the microcirculation at four different areas of the skin. The depth of measurement by using PF 4001 and 4002 is approximately 0.5–1 mm, but varies depending on
M
4
Bed
2
3
1 Fig. 1. Schematic illustration of the experimental setup with the laser Doppler fluxmeter probes attached to the limbs: the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2) and at the anterolateral part of ankle (Site 3). A fourth probe was attached to the first toe of the contralateral limb as a reference. “M” is a water filled, op en ended manometer line connected to a pressure monitor for recording of variations in foot position.
blood content of the tissue, degree of oxygenation and skin pigmentation.22 The laser Doppler flux (LDF) values, expressed as perfusion units, (PU) were saved by means of a Perisoft acquisition and analysis software program (Perimed AB) in a computer for offline analysis.
Recording of foot position To record alteration in foot position, a water-filled, open-ended manometer line connected to a pressure monitor (Pressure Monitor, Stranden) was attached to the skin at the dorsum of the foot. This device recorded vertical movements of the foot during the measurements. Changes in pressure indicated variations in the height of the foot.
Procedure The laser probes were applied at four different areas by using a probe holder and double adhesive tape (Fig. 1): the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2), at the anterolateral part of ankle (Site 3), the pulp of the first toe of the contralateral foot (Site 4). Measurements were performed after 15 min resting period with the subjects in supine position, in a room with a temperature between 23–25 °C. The LDF values were then continuously recorded Eur J Vasc Endovasc Surg Vol 20, December 2000
538
M. D. Anvar et al.
for 3 min, and the mean value of the last 2.5 min during that period represented the LDF value for the supine position. Subsequently, the CLI limb was passively lowered by flexion in the knee joint with the three laser probes still being attached (Fig. 1). The LDF values were recorded for 3 min and the final 2.5 min represented the LDF values for the dependent limb. The limb was lifted up again to the horizontal position to repeat the same measurements. This procedure was repeated three times for each patient with a total sampling time of more than 18 min for each participant.
site, and between CLI limbs and the contralateral limbs, and between the limbs in the control group. Mann–Whitney and ANOVA tests were used to compare differences in median frequency waves and FFTpower for both LF waves between limbs with CLI, with and without oedema, and healthy control group, both in horizontal and dependent position. The results were presented as medians with ranges in parentheses, and mean±standard deviation (s.d.), with p<0.05 as the level of statistical significance. GraphPad Prism 3.0 was used for analysis and presentation of data (GraphPad Software, Inc., San Diego, CA, U.S.A., www.graphpad.com).
Data analysis The laser Doppler (LD) signal was transformed to a frequency domain by fast Fourier transformation (FFT, Perisoft for Windows ver. 1.19, Perimed AB, Ja¨rfa¨lla, Sweden).22 The software allowed changes in time scale (x-axis) and signal amplification (y-axis) to facilitate the analyses. Various physiological activities which occur in time (“attractors”), such as cardiac activity, respiration, and variations in the microcirculation caused by neuronal or myogenic activity, may be represented in different frequency regions following the FFT. The results of FFT are expressed as frequency (cycles per second) of the attractors and FFT-power (“intensity”, arbitrary units, Fig. 2) of each of the frequency regions.23,24 There is a positive correlation between the derived FFTpower in both LF and HF regions and skin perfusion (LD) at the different measuring sites. A high power indicates that a specific frequency occurs regularly with a LD signal well above the noise level, and acts as a “quality-parameter”.
Leg and foot volume measurements The distal leg and foot volumes were measured by using water displacement volumetry (WDV) as previously described in detail.3 The differences in distal leg and foot volume between the two limbs were calculated.
Statistics Wilcoxon matched-pairs signed-ranks test was used for comparing changes in ABI, ASP, median frequency waves and FFT-power for both LF and HF waves between horizontal and dependent position in each Eur J Vasc Endovasc Surg Vol 20, December 2000
Results There was no statistically significantly difference in mean ankle brachial index (ABI) and ankle systolic pressure (ASP) between the critically ischaemic limbs with (0.19±0.13 with a mean ASP of 28±18 mmHg) and without oedema (0.27±0.14, and 35±20, both p>0.05). In the healthy control group, mean ABI was 0.94±0.28 with a mean ASP of 121±49 mmHg. In seven of 14 limbs with CLI and oedema and three of nine limbs with CLI and without oedema ultrasound Doppler signals could not be detected, and blood pressure at the ankle was defined at 15 mmHg. In patients with CLI with and without oedema, three distinct frequency domains were detected and are denoted respectively as low frequency (LF) waves, high frequency (HF) waves, and cardiac activity related, i.e. heart rate (HR) waves, while high frequency (HF) waves were absent in the healthy control group in all measured sites.
Low frequency (LF) waves With the limbs in horizontal position, there were no significant differences in median frequency of LF waves between the patients with CLI or between patients, and the healthy control group. There were no significant differences in median frequency of LF waves between the limbs within each of these groups either (Fig. 3). Altering the position of the ischaemic limbs from horizontal to dependency did not significantly change the median frequency of LF waves in any of the three groups (Fig. 3). The changes in position resulted, however, in a significant increase of the median FFT-power of LF
Patterns of Skin Flowmotion in the Lower Limbs
539
A
LDF (PU) 120
60
0 Horizontal
Pr.
Dependent
1 min.
B
Ti m
e
(m in
.)
Power
Frequency (Hz) Fig. 2. (A) A sample of recorded laser Doppler flux (LDF) with the limb in horizontal and dependent position in a healthy control participant. In the panel at the top a portion of the curve is expanded to better distinguish fluctuations caused by cardiac activity from flowmotion. (B) A 3D plot of spectrum analysis by fast Fourier transformation (FFT) in the same temporal selection of the curve is shown in the respective time and frequency domains. Pr: manometer line pressure, PU: perfusion unites.
Eur J Vasc Endovasc Surg Vol 20, December 2000
540
M. D. Anvar et al.
Site 3
Healthy controls
Horizontal
CLI without oedema
Dependent
CLI with oedema
10 8
6 4 2 Power
0 0
3
1
2 3 4 Frequency
2
5
1
(cycles/min.)
Site 2
10 8
6 4 2 Power
0 0
Dependent
Dependent
Horizontal
Horizontal
Site 1
1
2 3 4 Frequency
5
(cycles/min.)
10 8
6 4 2 Power
0 0
1
2 3 4 Frequency
5
(cycles/min.)
Fig. 3. Laser Doppler measurements of low frequency (LF) frequency waves with the limb in the horizontal and dependent position in healthy controls, CLI without oedema and CLI with oedema. There were three measuring sites: the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2) and the anterolateral part of the ankle (Site 3). Frequencies are expressed in cycles/min, including FFT-power of each of the frequency regions. Both values are given as medians and range.
[0.12 (0.015–3.3) vs. 0.82 (0.06–5.1), p<0.005] at the pulp of the first toe (Site 1) in the ischaemic limbs while unchanged in the other two sites (Fig. 3). In healthy controls this manoeuvre resulted in a decrease of the median FFT-power of LF in all three sites; site 1 [5.9 (0.8–11.1) vs. 3.7 (0.48–9.0), p<0.03], site 2 [1.8 (1.1–4.4) vs. 0.8 (0.25–3.5), p<0.002], site 3 [1.4 (0.8–2.9) vs. 0.6 (0.24–3.1), p<0.006]. The median FFT-power of LF of ischaemic limbs in site 1, 0.12 (0.015–3.3, Fig. 3), with the limbs in horizontal position was significantly lower than both the contralateral limbs 3.8 (0.34–8.15), p<0.001 and healthy control limbs, 5.9 (0.8–11.1), (p<0.0001) (Fig. 4). Eur J Vasc Endovasc Surg Vol 20, December 2000
High frequency (HF) waves All patients with CLI, (with and without oedema) showed HF waves in both diseased and contralateral limbs, while HF waves were absent in the healthy control group. With the limbs in horizontal position, there were no significant differences in median HF waves between the ischaemic limbs and the contralateral side. Altering the position of the CLI limbs from horizontal to dependency did not significantly change the median frequency of HF waves in the CLI feet (Fig. 5).
Patterns of Skin Flowmotion in the Lower Limbs
541
Contralateral limbs (Site 4) of: Healthy controls CLI patients without oedema
Low frequency
High frequency
CLI patients with oedema
10 8
6 4 2 Power
0 0
1
2 3 4 Frequency
5
1.0 .8
.6 .4
Power
.2
0 0
4
8
12 16 20
Frequency
(cycles/min.)
(cycles/min.)
Fig. 4. Laser Doppler measurements of low frequency (LF) and high frequency (HF) frequency waves of the contralateral limbs (Site 4) of the ischaemic limbs, with and without oedema, in the horizontal position. None of the healthy control limbs showed HF-waves. Frequencies are expressed in cycles/min, including FFT-power of each of the frequency regions. Both values are given as medians and range.
By lowering the ischaemic limbs the median FFTpower of HF increased only at the pulp of the first toe in both patients groups 0.08 (0.02–0.38) vs. 0.39 (0.09–1.45), (p<0.001) (Fig. 5). The median FFT-power of HF in the ischaemic limbs were significantly lower than the contralateral limbs [0.06 (0.01–0.12) vs. 0.2 (0.06–1.6), p<0.01)].
Heart rate (HR) waves The median rate of heart waves in patients with CLI and oedema was 72 (58–81), and without oedema 74 (58–84). In the supine position, HR waves were present in all contralateral sites i.e. site 4, in 50% of site 3 but not detected in the analysis of sites 1 and 2 in all CLI limbs both with and without oedema. Lowering the ischaemic foot resulted in appearance of HR waves in all sites. All subjects in the control group showed heart pulsatile flux with the limbs both in supine and dependent position at all measured sites with a median frequency of 68 (48–75) (Fig. 2).
Leg volume There was a mean leg–foot volume difference of 11±7% between limbs with CLI and oedema and contralateral sides measured by WDV. On the other
hand, we did not find any significant differences between the limbs with CLI and without oedema and contralateral sides measured by WDV.
Discussion The LF waves were as frequently observed in ischaemic as in healthy feet. However, HF waves were exclusively seen in the patients with CLI. This finding is in accordance with a previous study and indicates that the prevalence of HF waves is associated with a certain degree of ischaemia.25 This interpretation is supported by the observation that the number of HF waves decreases after successful angioplasty for severe ischaemia.26 Several investigators have reported that LF waves are associated with diameter changes in larger arterioles, and HF waves with diameter changes of terminal arterioles.16,27 Morphological studies by transmission electron microscopy have recently demonstrated dilated terminal skin arterioles in the feet of patients with CLI and oedema.2 However, it is a matter of speculation whether these findings are related to the occurrence of HF waves. Besides, it is not known how HF waves may contribute in compensating for the drastically decreased blood flow in CLI limbs. There were no significant differences in occurrence of LF or HF waves between ischaemic limbs with and without oedema and between the contralateral feet. Thus, it seems that ischaemic oedema formation does Eur J Vasc Endovasc Surg Vol 20, December 2000
542
M. D. Anvar et al.
Site 3 Horizontal
CLI without oedema
Dependent
CLI with oedema
1.0 .8
.6 .4
Power
.2
0 0
3
4
8
2
12 16 20
Frequency
1
(cycles/min.)
Site 1
Dependent
Dependent
Horizontal
Horizontal
Site 2
1.0 .8
.6 .4
Power
.2
0 0
4
8
12 16 20
Frequency
(cycles/min.)
1.0 .8
.6 .4
Power
.2
0 0
4
8
12 16 20
Frequency
(cycles/min.)
Fig. 5. Laser Doppler measurements of high frequency (HF) frequency waves with the limb in the horizontal and dependent position in patients with CLI without oedema and CLI with oedema. There were three measuring sites: the pulp of the first toe (Site 1), at the level of the second metatarsal body (Site 2) and the anterolateral part of the ankle (Site 3). Frequencies are expressed in cycles/min, including FFT-power of each of the frequency regions. Both values are given as medians and range.
not affect the manifestation of LF and HF waves. The ranges for both LF and HF waves in the present study are comparable with findings of previous studies.11,26 The cause and effect of LF and HF waves are still unknown. Both central and local nervous regulation may play a role.18 The central sympathetic nervous activity promotes homeostasis of blood pressure by constricting the arterioles, a local regulating system maintains blood perfusion of tissues by dilatating the arterioles. However, we found in this study that the distribution of the frequency waves is almost constant in different ischaemic areas of the foot. Furthermore, there was no difference in the frequency of LF and HF waves in the ischaemic limbs and the contralateral sides. These findings may indicate the superimposition of a local metabolic vasodilatory effect of hypoxia on centrally controlled vasoconstriction. Eur J Vasc Endovasc Surg Vol 20, December 2000
There was a positive correlation between the FFTpower and perfusion as measured by LDF, in all frequency domains. The power of LF waves was three times greater in the contralateral and five times greater in the healthy control limbs, compared to the ischaemic limbs. Therefore, it seems that the intensity of flowmotion curves, as indicated by the FFT-power, is related to the severity of ischaemia. A decrease of FFTpower probably varies in parallel with a reduced perfusion of the microcirculation in a certain area. Again, exactly how this effect could contribute to maintain homeostasis in ischaemic tissue is still unclear. Although there were no regional differences in LF or HF waves in the ischaemic feet, either with or without oedema, there were regional differences in the FFT-power for both LF and HF waves in the
Patterns of Skin Flowmotion in the Lower Limbs
ischaemic foot for both horizontal and dependent position. The power was lowest for both frequency waves at site 1 with the ischaemic feet in horizontal position, while the power in ischaemic feet in dependency increased more than 5-fold for both frequency domains at the pulp of the first toe (site 1). In contrast, the power of LF waves in healthy subjects was greatest at site 1 and lowest at the ankle, and lowering the limbs reduced power in all sites (Fig. 2). The effect of dependency and the regional difference in FFT-power of both LF and HF waves could be ascribed the different density of arteriovenous anastomoses (AVA) in the pulp of the toe as compared to site 2 and 3. The pulp of the toe contains the highest number of AVA per skin area,8,28 and hence possesses the greatest potential for varying skin perfusion. This is illustrated by the finding of a more pronounced increase in perfusion and FFT-power in site 1 compared with site 2 and 3, respectively (Figs 3 and 5). Several investigators have previously demonstrated that lowering an ischaemic leg below heart level improves the blood supply of poorly perfused skin regions,7,29,30 while lowering a healthy limb reduces the skin perfusion. There might be several possible physiological mechanisms that are involved in this phenomenon: (1) It has been postulated that paralysed arterioles and collapsed capillaries might be passively distended or even reopened by elevated transmural pressure when the leg is lowered.31 (2) A disturbed VAR may improve the blood perfusion due to lowering the critically ischamic limb, while it reduces blood perfusion in normal limb in dependency.7,10 (3) Subdermal thermoregulatory blood flow might be directed to nutritional capillaries during dependency in limbs with CLI.30,32 Furthermore, blood supply from the proximal leg may be distributed to acral skin regions during dependency.33 Thus, there are indications that redistribution of the microcirculation plays a compensatory role in maintaining optimal perfusion in severely ischaemic tissues. In conclusion, HF waves are typically associated with a state of extreme skin ischaemia. Ischaemia also appears to influence the FFT-power of each frequency domains. It seems that ischaemic oedema per se does not affect the pattern of FM in the foot of patients with CLI. Furthermore, the study showed that alteration in the microcirculation when lowering an ischaemic limb occurs as enhancement of the FM-waves (both occurrence and amplitude), while the frequency domains remain unchanged.
543
References 1 Khiabani HZ, Anvar MD, Rostad B, Stranden E, Kroese AJ. The distribution of oedema in the lower limb of patients with chronic critical limb ischaemia (CLI). A study with Computed Tomography (CT). VASA 1999; 28: 265–270. 2 Anvar MD, Khiabani HZ, Nesland JM, Kroese AJ, Stranden E. Vascular and stromal features in the skin of the lower limb in patients with chronic critical limb ischaemia (CLI) and oedema. Eur J Vasc Endovasc Surg 2000; 20: 125–131. 3 Khiabani HZ, Anvar MD, Stranden E, Slagsvold CE, Kroese AJ. Oedema in the lower limb of patients with chronic critical ischaemia (CLI). Eur J Vasc Endovasc Surg 1999; 17: 419–423. 4 Khiabani HZ, Anvar MD, Kroese AJ, Stranden E. Transcapillary forces in the development of oedema in the lower limb of patients with chronic critical limb ischemia (CLI). Eur J Vasc Endovasc Surg 2000; 19: 598–604. 5 Anvar MD, Khiabani HZ, Kroese AJ, Stranden E. Alterations in capillary permeability in the lower limb of patients with chronic critical limb ischaemia (CLI) and oedema. VASA 2000; 29: 106–111. 6 Anvar MD, Khiabani HZ, Lande K, Kroese AJ, Stranden E. The concentration of protein-compounds in interstitial tissue of patients with chronic critical limb ischaemia (CLI) and oedema. VASA 2001; 30: in press. 7 Henriksen O, Paaske WP. Local regulation of blood flow in peripheral tissue. Acta Chir Scand 1980; 502: 63–74. 8 Midttun M, Sejrsen P, Paaske WP. Peripheral blood flow rates and microvascular responses to orthostatic pressure changes in claudicants before and after revascularisation. Eur J Vasc Endovasc Surg 1999; 17: 225–229. 9 Ubbink DT, Kitslaar PJ, Tordoir JH et al. The relevance of posturally induced microvasular constriction after revascularisation in patients with chronic leg ishaemia. Eur J Vasc Surg 1992; 6: 525–532. 10 Khiabani HZ, Anvar MD, Kroese AJ, Stranden E. The role of the veno-arteriolar reflex (VAR) in the pathogenesis of peripheral oedema in patients with chronic critical limb ischaemia (CLI). Ann Chir Gyn 2000; 89: 93–98. 11 Hoffmann U, Franzeck UK, Geiger M, Yanar A, Bollinger A. Variability of different patterns of skin oscillatory flux in healthy controls and patients with peripheral arterila occlusive disease. Int J Microcirc 1993; 12: 255–273. 12 Chambers R, Zweifach BW. The topography and fuction of the mesenteric capillary circulation. Am J Anat 1944; 75: 173–205. 13 Nicoll PA, Webb RL. Vascular patterns and active vasomotion as determiners of flow through minutes vessels. Angiology 1955; 6: 291–310. 14 Fagrell B, Fronek A, Intaglietta M. A microscope-television system for studying flow velocity in human skin capillaries. Am J Physiol 1977; 233: 318–321. 15 Mahler F, Muheim MH, Intaglietta M, Bollinger A, Anliker M. Blood pressure fluctuations in human nailfold capillaries. Am J Physiol 1979; 236: 888–893. 16 Meyer JU, Borgstro¨ m P, Intaglietta M. Is vasomotion due to microvascular pacemaker cells? Vasomotion and flow modulation in the microcirculation. Prog Appl Microcirc. Basel, Karger 1989, vol 15, 41–48. 17 Arfors KE, Bergqvist D, Intaglietta M, Westgren B. Measurement of blood flow in the microcirculation. Ups J Med Sci 1975; 80: 27–33. 18 Intaglietta M. Vasomotion as normal microvascular activity and a reaction to impaired homeostasis. Vasomotion and flow modulation in the microcirculation. Prog Appl Microcirc. Basel, Karger 1989, vol 15, 1–9. 19 Second European Consensus Document on Chronic Critical Leg Ischaemia. Circulation 1991; 84, Suppl. 20 Myhre HO, Kroese AJ. Ultrasound in the study of peripheral blood circulation. Acta Chir Scand 1979; Suppl. 488: 15–18. 21 Sørensen TL, Perner L, Hansen L, Schroeder TV. Can Doppler Eur J Vasc Endovasc Surg Vol 20, December 2000
544
22 23 24
25
26
27
M. D. Anvar et al.
ultrasound replace strain gauge in the measurement of ankle blood pressure? Ugeskr Laeger 1992; 154: 3662–3665. Perimed. PeriFlux 4001 Master/4002 Satellite. User’s Manual. Ja¨rfa¨lla, Sweden (www.perimed.se). Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Numerical Recipes in C. The art of scientific computing. 1997 Cambridge University Press, Cambridge. Weidenhagen R, Wichmann A, Koebe HG et al. Analysis of laser Doppler flux motion in man: comparison of autoregressive modelling and fast Fourier transformation. Int J Microcirc 1996; 16: 64–73. Bollinger A, Hoffmann U, Seifert H. Flux motion in peripheral ischemia. In: Messmer K, Hammersen F (eds). Vasomotion and flow modulation in the microcirculation. Prog Appl Microcirc. Basel, Karger 1989, vol 15, 87–92. Hoffmann U, Schneider E, Bollinger A. Flow motion waves with high and low frequency in severe ischaemia before and after percutaneous transluminal angioplasty. Cardiovasc Res 1990; 24: 711–718. Schmid-Scho¨ bein H, Zieges S, Greber R et al. Synergetic interpretation of patterned vasomotor activity in microvascular perfusion: discrete effects of myogenic and neurogenic
Eur J Vasc Endovasc Surg Vol 20, December 2000
28 29 30
31 32 33
vasoconstricion as well as arterial and venous pressure fluctuations. Int J Microcirc 1997; 17: 346–359. Hassan AAK, Tooke JE. Effect of changes in local skin temperature on postural vasoconstriction in man. Clin Science 1988; 74: 201–206. Fagrell B. The skin microsirculation and the pathogenesis of the ischemic necrosis and gangrene. Scand J Clin Lab Invest 1977; 37: 473–476. Ubbink DT, Jacobs MJHM, Slaaf DW, Tangelder GJWJM, Reneman RS. Capillary recruitment and pain relief on leg dependency in patients with severe lower limb ischemia. Circulation 1992; 85: 223–229. Gaskell P, Burton AC. Local postural vasomotor reflexes arising from the limbs vein. Circ Res 1953; 1: 27–39. Bongard O, Fagrell B. Variations in laser Doppler flux and flow motion patterns in the dorsal skin of the human foot. Microvasc Res 1990; 39: 212–222. Scheffler A, Jendryssek J, Rieger H. Redistribution of skin blood flow during leg dependency in peripheral arterial occlusive disease. Clin Physiol 1992; 12: 425–438.
Accepted 3 October 2000