Eur J Vasc Endovasc Surg 15, 7-17 (1998)

REVIEW ARTICLE

The Evidence for Exercise-induced in Intermittent Claudication: Should we Encourage Patients to Stop Walking?

P. V. Tisi and C. P. Shearman

Department of Vascular Surgery, Southampton General Hospital, Southampton, U.K.

Objectives: To review clinical and experimental evidence that exercise to the onset of calf pain in patients with intermittent claudication results in an inflammatory response, and to consider whether repeated inflammatory events induced by therapeutic exercise training may lead to progression of . Methods: A literature search was performed to identify studies measuring biochemical markers of exercise-induced ischaemia-reperfusion in patients with intermittent claudication. Current theories of atherogenesis were reviewed and the use of acute-phase as potential markers of vascular disease explored. Results: Exercise to the onset of calf pain results in an inflammatory response with free formation, neutrophil activation and systemic vascular endothelial damage. Acute-phase proteins such as C-reactive and serum amyloid A protein have exciting potential use as stable biochemical markers of disease in claudication. Conclusions: Further studies are needed to determine the effect of long-term exercise training on exercise-induced inflammation in claudication. Early work suggests, in fact, that exercise attenuates this inflammatory response. If this were confirmed then it would support the clinical impression that exercise training is beneficial in terms of symptomatic improvement and cardiovascular health in patients with intermittent claudication.

Key words: Intermittent claudication; Ischaemia-reper)qlsion, Exercise, Acute-phase proteins.

Introduction 30 years, although few of these studies have been well- designed randomised-controlled trials. 6-~° A recent The usual advice given to a patient with stable in- meta-analysis of 21 exercise training programmes termittent claudication is to "stop smoking and keep showed that training for at least 6 months by walking walking".~ In the general population, regular exercise to near-maximum pain tolerance produced significant can decrease the risk of future improvements in pain-free and maximum walking and improve health. 2'3 In view of this, increased phys- distances. 1~ Supervised exercise training has also been ical activity has been promoted as a strategy to improve shown to be more effective than percutaneous trans- the health of the nation. 4 For patients with intermittent luminal angioplasty in terms of walking distance up claudication, exercise undoubtedly has significant to 12 months following randomisation.12 However, benefit in terms of reduction of cardiovascular risk there was no significant difference between the two factors, optimised structural and metabolic per- treatment modalities at 6 years' follow-up. ~3 formance of skeletal muscle, improvement in rheo- In the face of this impressive evidence regarding logical parameters, development of collaterals, the efficacy of exercise training in claudication, why improved walking technique and marked psy- has exercise not been adopted universally as a first- chological benefit.5 Exercise training programs have line treatment? Although locally benign, intermittent been widely reported in the literature over the past claudication is associated with a 10-year mortality of at least 50%, predominantly from cardiovascular disease. 14 This corresponds to 3.8 times the relative risk of compared to a control population. ~5 Evid- * Please address all correspondence to: Mr. P. V. Tisi, Department of Vascular Surgery, E-level, SouthamptonGeneral Hospital, Tremona ence is accumulating that exercise itself may be re- Road, SouthamptonSO16 6YD, U.K. sponsible for adverse inflammatory changes, which

1078-5884/98/010007+ 11 $12.00/0 © 1998 W.B. Saunders CompanyLtd. 8 P.V. Tisi and C. P. Shearman

Claudication Calf pain (Ischaemia-reperfusion) (ischaemia)

/ ff S Neutrophil activation ~ ODFR ATP ", XD

Proteases Purine pathway \ , XO //~ Xanthine Increased vascular endothelial permeability Oxidation LDL Rest (reperfusion) N~ Mieroalbuminuria Atherogenesis Fig. 2. Generation of oxygen-derived free radicals during ischaemia- Fig. 1. Outline of ischaemia-reperfusion injury in intermittent clau- reperfusion in claudication. (ATP=, XD= dication. (ODFR= oxygen-derived free radicals, LDL = low-density , XO = ). iipoprotein). SOD Catalase Superoxide H202 * H20

may be implicated in this excess cardiovascular mor- ,,. Fe 3+ tality. In 1924, Buerger perhaps recognised this prob- lem and suggested that walking for great distances Fenton Glutathione reaction peroxidase should be avoided in claudicants. 16 Instead, he pro- -increase intraeellular Ca 2+ -activates phospholipaseA2 Fe 2+ - _~ posed that passive exercises, such as alternate dis- -liberates arachidonic acid tension and emptying of the peripheral circulation by - synthesis leg elevation and dependency should be used to treat Hydroxyl claudication. Should we therefore encourage our patients to exercise to the onset of calf pain? This article l will explain the putative underlying mechanisms of -lipid peroxidation -oxidation of LDL exercise-induced inflammation in claudication and at- -macrophage activation tempt to rationalise the use of exercise as a treatment -? atherogenesis for intermittent claudication. Fig. 3. The role of oxygen radicals in vascular disease. (SOD= , H202=hydrogen peroxidase, LDL=low- density lipoprotein).

to a fall in tissue adenosine triphosphate levels and accumulation of hypoxanthine (Fig. 2). Biochemical basis of ischaemia-reperfusion injury Physiologically, xanthine dehydrogenase (XD) in vascular endothelial cells converts hypoxanthine to The systemic effects of intermittent claudication may xanthine using nicotine adenine dinucleotide as a co- be subsequent to the phenomenon of ischaemia-re- factor. induces a calcium-induced protease injury (IRI). The underlying biochemical attack on XD forming xanthine oxidase (XO) - the "D- pathways are outlined in Fig. 1. IRI causes greater O conversion". 2° XO requires oxygen as a co-factor tissue damage from restoration of the supply which is provided on reperfusion, generating the following a period of ischaemia than from the original superoxide radical. Superoxide is also a product of ischaemic injury. 17'~s Claudicants experience repeated the "respiratory burst" of activated neutrophils via episodes of calf muscle pain on exercise (ischaemia) oxidation of nicotinamide adenine dinucleotide phos- followed by rest (reperfusion). Transient muscle isch- phate. 21 aemia is difficult to assess clinically. A reduction in Oxygen-derived free radicals (ODFR) are important transcutaneous oxygen tension (TcPO2) in skin over in the pathogenesis of vascular endothelial injury fol- the calf muscle following exercise may, however, reflect lowing ischaemia-reperfusion (Fig. 3). Superoxide in- underlying ischaemic changes. 19 During muscle isch- creases intracellular calcium concentration, which aemia, inadequate oxidative phosphorylation leads activates endothelial phospholipase A2. This liberates

Eur J Vasc Endovasc Surg Vol 15, January 1998 Exercise-induced Inflammation in Intermittent Claudication 9

arachidonic acid from membrane phospholipids, is essential for neutrophil-mediated injury and re- which is the precursor for pro-inflammatory (thom- cruitment of further neutrophils to the site of in- boxane A2 and leukotriene B4) and anti-inflammatory flammation. (prostacyclin) eicosanoids. Thromboxane A2 (TXA2) Contrary experimental evidence has also suggested causes vasoconstriction, platelet aggregation, pro- that a circulating "neutrophil activator" is released motes neutrophil and increases vascular from ischaemic reperfused tissue, activating systemic permeability by disassembling actin microfilaments neutrophils which then plug the network. In and widening interendothelial tight junctions. 22 Ac- support of this, Paterson used an ex vivo chemotactic tivated neutrophils are probably the main source of model, in which human reperfusion plasma obtained TXA2 in IRI. Leukotriene B~ (LTB4) is chemotactic for from aortic surgery activated neutrophils in neutrophils and upregulates neutrophil-encothelial a rabbit dermabrasion lesion, resulting in increased adhesion in the post-capillary venule, leading to an microvascular permeability,as -8 (IL-8), increase in microvascular permeability. Animal ex- which is synthesised by monocytes in response to perimental work using eicosanoid receptor antagonists interleukin-1 and tumour necrosis factor-a, has been has confirmed that these mediators are certainly im- suggested as the main neutrophil activator. 33 IL-8 acts portant in acute limb ischaemia23 and therefore may specifically on neutrophils promoting chemotaxis, perhaps have a role in claudication. neutrophil-endotheial adhesion, extravascular ac- In vivo, superoxide rapidly dismutates to hydrogen cumulation at the site of injury and release of ODFR peroxide, which may oxidatively damage protein. The and proteases. 3. highly reactive hydroxyl radical is released from hy- Current evidence suggests that IRI has important drogen peroxide in the Fenton reaction, catalysed by effects on endothelial synthesis. Endo- ferrous ions. 24 It attacks polyunsaturated fatty acid thelium derived relaxing factor (EDRF) was isolated side-chains in membrane phospholipids forming the by Furchgott and Zawadzdki35, who demonstrated that was essential for the vasodilator peroxyl radical, which initiates a destructive chain action of acetylcholine. EDRF was subsequently shown reaction of lipid peroxidation. This obviously causes to be identical to nitric oxide (NO). 36 NO is synthesised structural damage to the vascular endothelium,2s but from L-arginine by NO synthase (NOS), which exists perhaps more importantly in claudication may be in several isoforms. These are broadly grouped into implicated in progression of atherosclerosis. constitutive (cNOS) and inducible (iNOS) enzymes. Activated neutrophils play an important role in the cNOS is found in endothelial cells and platelets, and pathogenesis of ischaemia-reperfusion injury and in is continuously expressed in response to shear stress, the extreme situation e.g. acute limb ischaemia, may histamine, thrombin, bradykinin and acetylcholine. initiate significant systemic pathology such as adult The physiological effects of NO are potent vaso- respiratory distress syndrome26'21 and multi-system dilatation and inhibition of both platelet aggregation organ failure. R7'28 Neutrophils exert their effects by and neutrophil adhesion. NO synthesis results in a physical obstruction of (the "no reflow" basal vasodilatory tone which is synergistic with pro- phenomenon), adhesion to the post-capillary venule stacyclin.37 Experimental evidence suggests that in endothelium and release of ODFR, eicosanoids and atherosclerosis there is failure of NO-mediated vaso- proteases. Neutrophil elastase, in particular, causes dilatation: both superoxide generated by IRI and ox- marked damage to the endothelial basement mem- idised LDL can inactivate NO. 38'39 Repetitive low-grade brane, resulting in an increase in microvascular per- IRI, as with intermittent claudication, may result in a meability,a9 The precise mechanism of neutrophil vascular endothelium in which NO-mediated re- activation has not been fully clarified. It has been laxation is ineffective.39 suggested that neutrophils are activated locally by iNOS is synthesised by neutrophils, macrophages, metabolites released from ischaemic reperfused tissue, vascular endothelial and smooth muscle cells (ie. the such as ODFR, TXA2 and LTB,. 30 IRI also leads to whole vascular wall) in response to inflammatory complement activation, releasing C5a which is chemo- mediators such as IL-1 and TNF-<,. iNOS produces tactic for neutrophils. 3~ Activated neutrophils are less significantly greater quantities of NO than cNOS (1000- deformable than inactive cells as a result of increased fold) 38 which may have marked inflammatory effects, cytoplasmic rigidity and projection of pseudopodia, leading to an increase in vascular permeability. 37 In the and are therefore less able to pass through the capillary absence of L-arginine, iNOS may generate superoxide network. 32 This prolonged capillary transit time in- which potentiates the inflammatory response. 3s NO creases neutrophil-endothelial binding via the Mac-1 may also react with superoxide to form the cytotoxic (CD11b/CD18) surface adhesion molecule, which radical. 4°

Eur J Vasc Endovasc Surg Vol 15, January 1998 10 P.V. Tisi and C. P. Shearman

Endothelin may also be involved in the pathogenesis the total antioxidant consumption, as this is in- of IRI. Endothelin-1 (ET-1) is synthesised by vascular dependent of the various metabolic pathways of ODFR endothelial ceils in reponse to ischaemia, shear stress, generation and degradation. Khaira used an enhanced thrombin, angiotensin II, and IL-I: en- chemiluminescent assay to look at exercise-induced dothelin synthesis is inhibited by increased NO syn- changes in antioxidant capacity in 20 claudicants and thesis. *~'~2The net balance between endothelin-induced nine matched controls. 49 Antioxidant levels in clau- vasoconstriction and NO-mediated vasodilatation has dicants (mean 479 ~tmol/1) and controls (438 ~xmol/1) marked effects on the vascular endothelium. En- were similar at rest, while the claudicants showed dothelin itself induces endothelium to synthesise NO, a significant decrease at i min following treadmill suggesting that a fine homeostatic mechanism exists exercise (428 ~tmol/1), returning to baseline at 10 min. in the control of vascular tone. 42 There is evidence that Control subjects showed an insignificant increase in synthesis of endothelin-1 increases during reperfusion antioxidant levels, probably reflecting a protective of ischaemic tissue. Edwards et al. 43 reported a canine mechanism against the deleterious effects of exercise. model of aortic cross-clamping in which ischaemia led to an insignificant increase in ET-1, with a subsequent significant increase in ET-1 on reperfusion. This per- haps should be explored in lesser grades of IRI such Neutrophil activation. Accumulating experimental evi- as in claudication. dence confirms the importance of neutrophils in skel- etal muscle ischaemia-reperfusion. Crinnion, using a rat hindlimb ischaemia model, demonstrated that pre- treatment with a monoclonal antibody to Mac-1 Evidence for ischaemia-reperfusion injury in claudication neutrophil adhesion molecule prevented neutrophil recruitment to the site of injury, attenuated muscle Oxygen-derived free radicals. Demonstrating ODFR syn- necrosis and improved post-ischaemic muscle per- thesis in claudicants is notoriously difficult, as ODFR formance, so This experimental situation is more suited, are highly reactive intermediates with short half-lives. however, to revascularization after a period of acute Many studies have therefore concentrated on measure- limb ischaemia rather than claudication. As discussed ment of stable end products of ODFR-induced lipid above, activated neutrophils are larger and stiffer than peroxidation, such as malondialdehyde. Hickman ex- their inactive counterparts. Neumann reported a study ercised 15 claudicants to maximum calf pain and in which 17 claudicants exercised using repetitive toe demonstrated a significant increase in plasma ma- stands. 32 Immediately following exercise there was an londialdehyde and thrombomodulin (an endothelial increased neutrophil count in femoral-venous blood membrane glycoprotein released with cell damage) (local) compared to femoral-arterial blood (systemic), at 1 rain post-exercise with insignificant changes in an increased proportion of activated neutrophils and healthy controls. 44 The claudicants had significantly decreased neutrophil filtration through 8~tm mi- lower resting red cell superoxide dismutase than con- cropore filters (indicative of the decreased de- trols, suggesting that the free radical scavenging sys- formability of activated neutrophils). After 10min tems are also ineffective. 45 In a further study, lipid these changes were seen in systemic blood. Decreased peroxides were measured in 10 claudicants and 10 neutrophil filterability in systemic venous blood fol- controls,46 Claudicants had slightly higher resting lipid lowing treadmill exercise has also been shown to peroxide levels than control subjects, but showed a correlate with increased serum lysozyme resulting 53% increase following maximal treadmill exercise, from neutrophil degranulation. 5~ This suggests that with no change in controls. In a further study, Stringer activated neutrophils are released locally from the showed an increase in resting lipid peroxides in 50 site of injury into the systemic circulation following patients with arteriographic evidence of peripheral treadmill exercise. Edwards 48 also showed an increase vascular disease, compared to a control population. 47 in systemic neutrophil count following exercise. The A further method of assessing ODFR formation is median neutrophil count was 5.6 x 106/1 at rest (com- to measure individual antioxidant mechanisms, which pared to 2.8 x 106/1 in controls), increasing to 7.1 x 106/ scavenge newly formed ODFR, attenuating the harm- 1 immediately after exercise (p<0.05) and returning to ful effects. Glutathione peroxidase (which detoxifies baseline at 15 min. A more sophisticated method of hydrogen peroxide) and its co-factor selenium are assessing neutrophil activation has recently been de- significantly lower following exercise in claudicants scribed. 52 The cell transit analyser measures neutrophil as compared to control subjects.48 Perhaps a more flow through 8 ~m pores, analysing 1000 cells per appropriate measure of free radical activity is to assess sample. This study showed that claudicants had an

Eur J Vasc Endovasc Surg Vol 15, January 1998 Exercise-induced Inflammation in Intermittent Claudication 11 increase in the median and 90th percentile transit time which is undetectable by qualitative dip-stix55 and has 5 min following exercise, with no significant change been quantified as a urinary albumin concentration of in controls. The change in median transit correlated 15-200 mg/1. 6° Microalbuminuria is usually expressed with an increase in plasma thromboxane, adding sup- as the albumin-creatine ratio (ACR) which allows for port to the idea that activated neutrophils are the main fluctuation in urinary flow rate. Microalbuminuria source of thromboxanes in reperfusion injury. was originally described as an index of early diabetic nephropathy and later shown to be associated with Thromboxane synthesis. Limited studies have looked at pancreatitis,61 the response to surgery,62 hypertension63 the effect of treadmill exercise on eicosanoid synthesis and adult respiratory distress syndrome following in claudicants. Wennmalm found a significant increase aortic surgery. 64 in urinary excretion of thromboxane B2 metabolites The association between microalbuminuria and 20 min following maximal treadmill exercise in clau- claudication has been well documented in several dicants. ~3 In another study the resting plasma throm- studies. Most studies have shown no significant dif- boxane was not significantly different in 11 claudicants ference in the resting ACR between claudicants and (median 32.1pg/ml) compared to seven controls controls,46"52 although others have shown that the base- (25.0 pg/ml). However, following treadmill exercise line ACR is higher in claudicants.65'66 These results only claudicants showed a significant increase which may be explained by the 100 to 1000-fold variation in peaked at 15 min (135.0 pg/ml). 48 Khaira also dem- ACR which would result in an elevated mean but onstrated an increase in TXB2 within 10 min of tread- perhaps not an elevated median ACR. All these studies mill exercise in claudicants, peaking at 60 min. ~2 A have, however, shown a significant increase in ACR similar rise was also seen in controls, although at all in claudicants following maximal treadmill exercise times levels in claudicants were significantly higher with no corresponding change in controls. Hickey than in controls. demonstrated a 153% increase in ACR in 23 claudicants after treadmill exercise (p<0.001), with no significant change in 10 controlsY This exercise-induced increase Vascular endothelial damage. The end-result of isch- in ACR has been shown to correlate with an increase aemia-reperfusion injury in claudication is damage to in serum lipid peroxides,46 total antioxidant con- the target organ - vascular endothelium - reflected by sumption,49 neutrophil activation and degranulation51,52 an increase in systemic vascular permeability. Jellinek and Doppler ankle pressure recovery time. 65 This char- and Detre suggested that increased systemic vascular permeability pre-dates development of atherosclerotic acteristic increase in ACR with exercise may be used lesions. 54 Clinically, an increase in systemic vascular to assess the effect of treatment on claudication. Tsang permeability is difficult to determine. However, there showed that i week of treatment with oxypentifylline, is good reason to correlate renovascular permeability which modulates vascular inflammation, led to a re- with systemic vascular permeability.55 Experimental duction in the post-exercise increase in ACR compared work has shown that in hamsters subjected to lap- to patients taking a placebo. 67 Surgery also appears to arotomy and peritoneal contamination, there is an have an effect on the ACR. Hickey demonstrated that increase in microvascular permeability to fluorescein- bypass surgery had no effect on the resting ACR in labelled dextran in the cheek pouch as well as increased 12 claudicants, but significantly attenuated the post- urinary excretion of high molecular weight dextran. 56 exercise rise in ACR. 51 Similar work by Matsushita et Epidemiological evidence from the prospective Fram- al. 66 confirmed these findings. ingham Study has also shown a three-fold increase in An alternative method of assessing target organ cardiovascular mortality in subjects with proteinuria. 57 damage in atherosclerosis is to measure plasma von Excretion of urinary proteins may therefore reflect the Willebrand factor (vWF). vWF is a specific product of amount of vascular endothelial damage and perhaps damaged vascular endothelial cells and has a role in be indicative of the progression of atherosclerosis.58'59 platelet aggregation and adhesion. 6s Elevated vWF has Urinary albumin excretion is a useful measure of been demonstrated in subjects with known cardio- renovascular permeability. Physiologically, tubular re- vascular risk factors such as diabetes, , sorption mechanisms for albumin are nearly saturated smoking and hypercholesterolaemia.69 A significant and therefore a small increase in vascular permeability increase in vWF has also been shown in subjects will be amplified by the renal concentrating mech- with PAD compared to controls, presumably due to anism leading to a significant increase in urinary repetitive low-grade inflammatory events. 7°'7~ As with albumin excretion - microalbuminuria.46 Micro- urinary albumin, vWF has been shown to increase albuminuria is defined as urinary albumin excretion, following treadmill exercise in claudicants.48

Eur J Vasc Endovasc Surg Vol 15, January 1998 12 P.V. Tisi and C. P. Shearman

Liver Fibrinogen The studies detailed above provide conclusive evid- CRP ence that exercise in claudicants is associated with an SAA inflammatory response with free radical formation, lipid peroxidation, deficient antioxidant mechanisms, Endothelium I~it f~it ~ I~!:,i!!iI neutrophil activation, eicosanoid synthesis and vas- Lipid peroxides Cytokines cular endothelial damage. The significance of these changes is not fully understood. Repeated low-grade OX-LDL inflammatory events, however, may lead to pro- gression of atherosclerosis, 72 perhaps explaining the Immune complexes excess cardiovascular mortality in these patients. Pro- teinuria and the neutrophil count are certainly known Macrophage to be predictive of future cardiovascular events. 57'73 Fig. 4. The role of lipid peroxidation in oxidation of low-density lipoprotein and acute-phase protein synthesis. (OX-LDL= oxidised low-density lipoprotein, CRP=C-reactive protein, SAA=serum amyloid A protein). Exercise-induced inflammation in the pathogenesis of atherosclerosis protein synthesis and atherogenesis. Khaw and Lipid peroxides, generated in claudicants from ex- Woodhouse suggested that repeated exposure to en- ercise-induced IRI, have been suggested as in- vironmental inflammatory stimuli, such as respiratory dependent risk factors for the progression of infection, results in macrophage activation and syn- atherosclerosis. 47 Lipid peroxides are cytotoxic to vas- thesis of acute-phase proteins, such as fibrinogen, cular endothelium, procoagulant and may promote which have atherogenic potential, s3 However, the re- platelet activation. More recently it has been noted sponse to oxidised lipids appears to be a more logical that lipid peroxides are able to oxidise subendothelial hypothesis, although both theories recognise the low-density lipoprotein (LDL) to various grades of importance of acute-phase proteins in atherosclerosis. oxidised LDL (OX-LDL). Oxidation mainly occurs within the arterial intima, as circulating antioxidants Fibrinogen. There is now impressive evidence that fib- such as vitamin E prevent significant intravascular rinogen is an important cardiovascular risk factor. The oxidation of LDL. 74'75 The various oxidised LDLs are Framlingham, Northwick Park and Caerphilly studies cytotoxic to vascular endothelium, thrombogenic, im- demonstrated that an elevated fibrinogen at baseline munogenic and recruit macrophages into the sub- was an important prognostic factor for incident isch- endothelial spacer 6"77Increasing evidence suggests that aemic heart disease. 8446 The Speedwell prospective macrophages play an important role in the de- heart disease study also showed that a raised fib- velopment of atherosclerosis. 72'7s'79Immune complexes, rinogen predicted the development of intermitten clau- formed by antibodies to epitopes of OX-LDL, 75 and dication. 15 Both hospital and population-based studies OX-LDL itself are taken up by "scavenger receptors" in claudicants have shown an elevated fibrinogen in on macrophages, resulting in unregulated ac- claudicants compared to controls, s7 Recent work from cumulation of and lipid debris forming the the Edinburgh Study has also shown an increase characteristic "foam cells" of atherosclerosis, s° Ox- in fibrinogen degradation products (FDP) (urinary idised lipid may also induce macrophages to release fibrinopeptide A and plasma D-dimer) in claudicants cytokines which initiate an hepatic acute-phase re- compared to controls, s8 This study also found an in- sponse 75's~ (Fig. 4). It has been suggested that cytokines dependent association between fibrinogen and a re- themselves are atherogenic, although induction of hep- duction in the Doppler ankle-brachial pressure index. 89 atic synthesis of fibrinogen and other acute-phase Claudicants with a raised fibrinogen are known to proteins may be a more likely explanation. have poor prognosis in terms of local progression of Measurement of acute-phase proteins may therefore disease, 9° graft failure 91 and cardiovascular death. 9a provide an opportunity to assess the intermediate-term The mechanisms of action of fibrinogen are complex. effect of repeated inflammatory episodes as occurs in Undoubtedly, fibrinogen adversely affects micro- claudication. Persisting elevation of acute-phase levels circulatory blood flow via its influence on red cell may also reflect the systemic nature of the disease and platelet activation, activation of coagulation mech- possibly explain the excess cardiovascular mortality anisms and increased plasma viscosity. Recent work in these patients. 8a Alternative theories have been pro- has also suggested that fibrinogen and LDL have posed to explain the association between acute-phase common uptake into atherosclerotic plaque, where

Eur J Vasc Endovasc Surg Vol 15, January 1998 Exercise-induced Inflammation in Intermittent Claudication 13 fibrinogen is converted to fibrin with subsequent fib- in vascular permeability and the resulting acute-phase rinolysis. FDPs may be atherogenic via effects on response. 1°5 smooth muscle cell proliferation, increased micro- The association between CRP and ischaemic heart vascular permeability and neutrophil chemotaxis. 9s'94 disease (IHD) has received more attention, although Some recent work has, however, suggested that an CRP does not appear to correlate with the severity of elevated fibrinogen in atherosclerosis may not be a the presenting IHD but rather with future cardiac consequence of repeated acute-phase responses, but risk. 1°6 The European Concerted Action on may result from genetic polymorphism, producing a and Disabilities Angina Pectoris (ECAT) study showed protein with abnormal atherogenic potential. 95'96 In that a raised CRP within the normal range in patients considering this work and the notorious difficulty in with stable angina was prognostic for myocardial in- interpreting a single fibrinogen result, 97 alternative farction and cardiac death. 1°7 In patients with unstable markers of the acute-phase response may have greater angina, a CRP>3 mg/1 has also been shown to predict value in assessing the inflammatory status in clau- a poor response to medical treatment and re-in- dicants. farction. ~°8 The above studies suggest that CRP is elevated in claudicants, although remaining within the C-reactive protein. C-reactive protein (CRP) has im- normal range, and that a raised CRP may be considered portant potential use as a marker of repeated in- as an adverse cardiovascular risk factor. However, its flammatory events in claudication. CRP is synthesised use may be limited by the fact that a sensitive assay by hepatocytes in response to interleukin-6 derived is required to detect any small changes from normal. from macrophages 98 and, as there is no tissue de- positon, the serum level reflects the synthetic rate. 99 Serum amyloid A protein. Serum amyloid A protein Rather than acting as a passive marker of athero- (SAA) has potential use as a very sensitive marker of sclerosis, CRP has the ability to activate the classical the acute-phase response, with serum levels increasing complement pathway ~°° and may lead to platelet ag- up to 1000 times the physiological value following an gregation and release reactions. In vitro work has also inflammatory response. 1°9 The function of constitutive suggested that CRP complexed to damaged endo- SAA (SAA4) is unknown, although it does not increase thelial cells may selectively bind LDL, which obviously in response to oxidised lipids. 11° Acute-phase SAA has important implications for atherogenesis. ~°1 (SAA1 and SAA 2~/2J3) is structurally an apolipo- There is little published work on the measurement protein and is synthesised by hepatocytes in response of CRP in peripheral arterial disease (PAD). Majewski to cytokines, predominantly interleukin-1. ~1'112 Recent showed that CRP progressively increased according experimental work has also shown that all the im- to the Fontaine staging of PAD, with patients with portant cell-types in atherosclerosis-macrophages - critical ischaemia having significantly greater CRP vascular endothelial cells and smooth muscle cells - levels than claudicants. ~°2 In the 20 claudicants in this express messenger RNA and therefore the ability to study the median CRP was I mg/l: the 90th percentile synthesise SAA. 113'114 of the normal range has been quoted as 3 mg/1. ~°s The The physiological effects of SAA are intrinsically apparently normal values may have been due to the linked with high-density lipoprotein (HDL) meta- relatively insensitive assay used which had a limit of bolism that has important protective effects via the detection of 5 mg/1. In another study of 929 patients "reverse cholesterol transport" mechanism. ~5 SAA ap- admitted to a coronary rehabilitation unit, significantly pears to have both pro-atherogenic and anti-athero- higher CRP levels were found in patients with co- genic functions, suggesting complex control existent PAD, although again these were within the mechanisms are operating. SAA displaces apolipo- normal range. 94 It was suggested that small increments protein A-I from HDL, reducing HDUs lecithin cho- within the normal range represented significant in- lesterol acyl transferase activity, which has the net flammatory events. In support of the environmental effect of increasing cholesterol deposition at sites of inflammatory aetiology of atherosclerosis, a recent endothelial injury, where it may be taken up by LDL. 116 study has shown an increased CRP in those previously SAA bound to HDL reduces the hepatocyte affinity exposed to Helicobacter pylori and chlamydia pneu- for HDL and downregulates hepatocyte expression of monia infection, and that this elevated CRP was as- the HDL receptor, while increasing HDL's affinity for sociated with prevalent intermittent claudication. TM subendothelial macrophages and increasing macro- We have previously shown that the CRP level in 44 phage expression of the HDL receptorY The net result stable claudicants significantly correlates with both is accumulation of the SAA-HDL complex in the sub- plasma fibrinogen and the urinary ACR, providing a endothelium where SAA is then deposited, leaving link between ischaemia-reperfusion induced increase HDL free to scavenge cholesterol debris released from

Eur J Vasc Endovasc Surg Vol 15, January 1998 14 P.V. Tisi and C. P. Shearman

macrophages. In keeping with its "schizophrenic" shown that exercise training leads to a reduction in function, SAA has been shown to inhibit the neutrophil these biochemical markers of inflammation, m There oxidative burst, inhibit platelet aggregation and release is therefore no clinical evidence to suggest that subjects reactions, m suppress macrophage T-cell interaction 1°° with intermittent claudication should be discouraged and induce chemokinesis of monocytes and neutro- from exercising: further experimental work may in phils. Ha SAA also upregulates expression of the Mac- fact confirm that exercise training decreases the in- 1 neutrophil adhesion molecule] 19 The overall effect flammatory response to walking. This may have bene- of SAA therefore appears to be to direct both in- ficial effects in terms of progression of atherosclerosis. flammatory cells and HDL to the site of endothelial injury where lipid debris can be dealt with effectively. In view of these mechanisms, it therefore seems appealing to measure SAA level as an index of the References inflammatory response in atherosclerotic plaque. To 1 HOUSLE¥ E. Treating claudication in five words. Br Med J 1988; date, no work has previously been published on SAA 296: 1483-1484. in peripheral arterial disease. Limited studies have 2 HOUSLEYE, LENG GC, DONNAN PT, FOWKES FG. Physical activity measured SAA in ischaemic heart disease. Un- and risk of peripheral arterial disease in the general population: Edinburgh Artery Study. J Epidemiol Community Health 1993; surprisingly, SAA is known to rise following myo- 47: 475-480. cardial , due to the marked acute-phase 3 POWELL E, PRATT M. Physical activity and health. Br Med J response. 12° However, Liuzzo showed that in patients 1996; 313: 126-127. 4 SECRETARY OF STATE FOR HEALTH. The Health of the Nation. A with unstable angina, an elevated SAA was predictive strategy for health in England. HMSO: London, 1992; 46-7, 62-4. of further cardiovascular morbidity and mortality. 1°8 5 ERNST E. Exercise: the best therapy for intermittent claudication? Br J Hosp Med 1992; 48: 303-307. 6 LARSEN OA, LASSEN NA. Effect of daily muscular exercise in patients with intermittent claudication. Lancet 1966; ii: 1093- 1095. Conclusion 7 ERICSSON B, HAEGER K, LINDELL SE. Effect of physical training on intermittent claudication. Angiology 1970; 21: 188-192. 8 DAHLLOF AG, BJORNTORP P, HOLM J, SCHERSTEN T. Metabolic Current opinion suggests that patients with in- activity of skeletal muscle in patients with peripheral arterial termittent claudication should be encouraged to take insufficiency. Eur J Clin Invest 1974; 4: 9-15. 9 DAHLLOr AG, HOLM J. Peripheral arterial insufficiency. Effect regular exercise. Clinical trials of exercise training have of physical training on walking tolerance, calf blood flow and shown an undoubted benefit in terms of walking blood resistance. Scand J Rehab Med 1976; 8: 19-26. distance and quality of life. Exercise is also known to 10 HIATT WR; REGENSTEINER J, HARGARTEN ME, WOL~EN EE. Benefit of exercise conditioning for patients with peripheral benefit cardiovascular health in the general population. arterial disease. Circulation 1990; 81: 602-609. However, claudicants have excessive cardiovascular 11 GARDNER AW, POEHLMAN ET. Exercise rehabilitation programs morbidity and mortality rates compared to the general for the treatment of claudication pain. A meta-anatysis. JAMA 1995; 274: 975-980. population. There is overwhelming experimental evid- 12 CREASY TS, McMILLAN PJ, FLETCHER EWL, COLLIN J, MORRIS ence that in claudicants, exercise to the onset of calf PJ. Is percutaneous transluminal angioplasty better than ex- pain followed by rest results in an ischaemia-re- ercise for claudication? Preliminary results from a prospective randomised trial. Eur J Vasc Surg 1990; 4: 135-140. perfusion insult with free radical formation, neutrophil 13 PERKINS JMT, COLLIN J, GREASY TS, FLETCHER EWL, MORRIS activation and systemic vascular endothelial damage. PJ. Exercise training versus angioplasty for stable claudication: Increasing evidence also suggests that repetitive low- long and medium term results of a prospective randomised trial. Eur f Vasc Endovasc Surg 1996; 11: 409-413. grade inflammatory episodes may lead to progression 14 DORMANDY IA, MAHIR M, ASCADY G, BALSANO 1~, DE LEEUW P. of atherosclerosis, which may explain the increased Fate of the patient with chronic leg ischaemia. J Cardiovasc Surg incidence of ischaemic heart disease and in 1989; 30: 50-57. 15 BAINTON D, SWEETNAM P, BAKER I, ELWOOD P. Peripheral vas- these patients. Further studies are need to clarify the cular disease: consequence for survival and association with nature of this inflammatory response in relation to risk factors in the Speedwell prospective heart disease study. exercise, perhaps by measurement of more stable Br Heart J 1994; 72: 128-132. 16 HOLM J. The effect of exercise on intermittent claudication. In: markers of inflammation such as acute-phase proteins. Bell PRF, Jamieson CW, Ruckley CV, eds. Surgical Management This will help determine whether these biochemical of Vascular Diseases. Saunders: Lo~tdon, 1992; 111-118. changes have any clinical significance. The un- 17 HAIMOVlCI H. Muscular, renal and, metabolic complications of acute arterial occlusions: myonephropathic-metabolic syn- answered question is whether exercise training leads drome. Surgery 1979; 85: 461-468. to a cumulative increase in the inflammatory response 18 PARKS DA, GRANGER DN. Contributions of and re- or whether, by increasing the pain-free walking dis- perfusion to mucosal lesion formation. Am J Physiol 1986; 250: G749-G753.. tance, the effect of each individual ischaemia-re- 19 CAILLARD P, MOUREN X, PUJADE B, BLANCHEMAISON P, ELBEZE perfusion event is decreased. We have, in fact, recently Y/ CLOAREC M. Objectifying ~xercise ischaemia in peripheral

Eur J Vasc Endovasc Surg Vol 15, January 1998 Exercise-induced Inflammation in Intermittent Claudication 15

vascular disease: a study in 120 patients. Angiology 1990; 12: derived free radicals and endothelial damage. Adv Exp Med BioI 469-478. 1994; 36: P565-P570. 20 McCORD JM. Ocygen-derived free radicals in postischemic 45 HICKMAN P, MCCLAREN M, HILL H, McCOLLUM PR, BELCH JJ. tissue injury. N Engl J Med 1985; 312: 159-163. Exercise causes endothelial damage in claudicants but in healthy 21 WINDSOR ACJ, MULLEN PG, FOWLER AA, SUGERMAN HJ. Role controls. Br J Surg 1994; 81: 616. of the neutrophil in adult respiratory distress syndrome. Br J 46 SI~EARMAN CP, GOSLING Pr GWYNN BR, SIMMS MH. Systemic Surg 1993; 80: 10-17. effects associated with intermittent claudication. A model to 22 LELCUK S, ALEXANDER F, VALERI R, SHEPRO D, HECHTMAN H. study biochemical aspects of vascular disease? Eur J Vasc Surg Thromboxane A2 moderates permeability after limb ischaemia. 1988; 2: 401-404. Ann Surg 1985; 202: 642-646. 47 STRINGER MD, GOROG PG, FREEMAN A, KAKKAR MY. Lipid 23 HOMER-VANNIASINKAM S, GOUGH MJ. Role of lipid mediators peroxides and athersclerosis. Br Med J 1989; 298: 281-284. in the pathogenesis of skeletal muscle infarction and oedema 48 EDWARDS AT, BLANN AD, SUAREZ-MENDEZ VJ. Systemic re- during reperfusion after ischaemia. Br J Surg 1994; 81:1500-1503. sponses in patients with intermittent claudication after treadmill 24 SINCLAIR aJ, BARNETT AH, LUNEC J. Free radicals and anti- exercise. Br J Surg 1994; 81: 1738-1741. oxidant systems in health and disease. Br J Hosp Med 1990; 43: 49 KHAIRA HS, MAXWELL SRJ, SHEARMAN CP. Antioxidant con- 344-341. sumption during exercise in intermittent claudication. Br J Surg 25 PARKS DA, BULKLEY GB, GRANGER DN. Role of oxygen free 1995; 82: 1660-1662. radicals in shock, ischemia and organ preservation. Surgery 50 CRINNION IN, HOMER-VANNIASINKAM S, PARKIN SM, GOUGH 1983; 94: 428-438. MJ. The role of neutrophil-endothelial adhesion in skeletal 26 PATERSON IS, KLAUSNER JM, GOLDMAN Get al. Pulmonary muscle reperfusion injury. Br J Surg 1996; 83: 251-254. edema after aneurysm surgery is modified by mannitol. Ann 51 HICKEY NC, GOSLING P, BAAR S, SHEARMAN CP, SIMMS MH. Surg 1989; 210: 796-801. Effect of surgery on the systemic inflammatory response to 27 PATERSON IS, CORSON J, McCOLLUM CN, GOLDMAN MD, intermittent claudication. Br J Surg 1990; 77: 1121-1124. MIDDLETON MD. Activated neutrophils mediate multiple organ 52 KHAIRA HS, NASH GB, BAHRA PS et al. Thromboxane and dysfunction following repair. Br J Surg 1992; neutrophfl changes following intermittent claudication suggest 76: 358. ischaemia-reperfusion injury. Eur J Vase Endovasc Surg 1995; 10: 28 PATERSON IS, SMITH FC% TSANG GMK, HAMER JD, SHEARMAN 31-35. CP. Reperfusion plasma contains a neutrophil activator. Ann 53 WENNMALMA, EDLuND A, SEVASTIKB, FITZGERALDA. Excretion Vasc Surg 1993; 7: 68-75. of thromboxane A2 and prostacyclin metabolites during tread- 29 HICKMAM P, McCOLLUM PT, BELCH JJE Neutrophils may con- mill exercise in patients with intermittent claudication. Clin tribute to the morbidity and mortality of claudicants. Br J Surg Physiol 1988; 8: 243-252. 1994; 81: 790-798. 54 JELLINEK H, DETRE Z. Role of the altered transmural per- 30 KLAUSNERJM, PATERSON IS, MANNICK JA, VALERI CR, SHEPRO meability in the pathomechanism of . Part I. D, HECHTMAN HB. Reperfusion pulmonary edema. JAMA 1989; History of arteriosclerosis theories. Path Res Pract 1986; 181: 261: 1030-1035. 693-712. 31 BENGTSONA, HOLMBERG P~ HEIDEMAN M. The ischaemic leg as 55 WINOCOUR PH. Microalbuminuria. 1992; 304: 1196- a source of complement activation. Br J Surg 1987; 74: 697-700. Br Med J 32 NEUMANN FJ, WAAS W, DIEHM C et al. Activation and decreased 1197. deformability of neutrophils after intermittent claudication. 56 SHEARMANCP I GOSLING PI SIMMS MH. Glomerular permeability Circulation 1990; 82(3): 922-929. to large molecules: a marker of vascular permeability? Br f Surg 33 BAGGIOLINI M, WALZ A, KUNKEL SL. Neutrophil activating 1988; 75: 1273. peptide-1/Interleukin 8, a novel that activates neutro- 57 KANNEL WB, STAMPFER MJ, CASTELLI WE, VERTER J, The pro- phils. J Clin Invest 1989; 84: 1045-1049. gnostic significance of proteinuria: the Framingham study. Am 34 GROSS V, ANDREESEN R, LESER HG, CESKA M. Interleukin-8 and Heart J 1984; 108: 1347-1352. neutrophil activation in acute pancreatitis. Eur ] Clin Invest 58 GOSLING P, SHEARMAN CP. Increased levels of urinary proteins: 1992; 22: 200-203. markers of vascular permeability? Ann Clin Biochem 1988; 25 35 FURCHGOTT RE, ZAWADZKIJV. The obligatory role of endothelial (Suppl.): $150-$151. cells in the relaxation of arterial smooth muscle by acetycholine. 59 GOSLING Pr BEEVERS DG, GOODE GE, HICKEY NC, LITTLER WA, Nature 1980; 288: 373-376. SHEARMAN CP. Lancet 1990; 335: 349-350. 36 PALMER R2~J, FERRIGE AG, MONCADA S. Nitric oxide release 60 WATTS GF, BENNETT JE, ROWE DJ et al. Assessment of im- accounts for the biological activity of endothelium-derived munochemical methods for determining low concentrations of relaxing factor. Nature 1987; 327: 524-526. albumin in urine. Clin Chem 1986; 32: 1544-1548. 37 VALLANCE P, COLLIER J. Biology and clinical relevance of nitric 61 SHEARMAN CPI GOSLING P, WALKER KJ. Is low proteinuria an oxide. Br Med J 1994; 309: 453-459. early predictor of severity of acute pancreatitis? J Clin Pathol 38 DAVIES MG, FULTON GJ, HAGEN PO. Clinical biology of nitric 1989; 42: 1132-1135. oxide. Br J Surg 1995; 82: 1598-1610. 62 GOSLING P, StIEARMAN CPl GWYNN BR, SIMMS MH, BAINBRIDGE 39 WEIGHT SC, BELL PRF, NICHOLSON ML. Renal ischaemia-re- ET. Microproteinuria: response to operation. Br J Med 1988; 296: perfusion injury. Br J Surg 1996; 83: 162-170. 338-339. 40 GRACE P. Ischemia-reperfusion injury. Br J Surg 1994; 81: 637- 63 GOSLING P, BEEVERS DG. Urinary albumin excretion and blood 647. pressure in the general population. Clin Sci 1989; 76: 39-42. 41 VANE JR, ANGGARD EE, BOTTING RM. Regulatory functions of 64 SMITH EC% GOSLING P, SANGHERA K, GREEN MA, PATERSON the vascular endothelium. N Eng J Med 1990; 323: 27-36. IS, SHEARMAN CP. Microproteinuria predicts the severity of 42 HAYNES WG, WEBB DJ. The endothelin family of peptides: local systemic effects of reperfusion injury following infrarenal aortic hormones with diverse roles in health and disease? Clin Sci aneurysm surgery. Ann Vasc Surg 1994; 8: 1-5. 1993; 84: 485-500. 65 HICKEY NC, SHEARMAN CP, GOSLING P, SIMMS MH. Assessment 43 EDWARDS JD, DOVGAN PS, ROWLEY JM, AGRAWAL DK, THORPE of intermittent claudication of exercise-induced micro- PE, ADRIAN TE. Endothelin-1 levels in ischaemia, reperfusion albuminuria. Eur ] Vasc Surg 1990; 4: 603-606. and haemorrhagic shock in the canine infrarenal aortic re- 66 MATSUSHITA M, NISHIKIMI N, SAKURAI % YANO T, NIMURA Y. vascularization model. Eur J Vasc Surg 1994; 8: 729-734. Urinary microalbumin as a marker for intermittent claudication. 44 HiCRMAN P, HAItRIsoN DK, HILL A et aL Exercise in patients with Eur ] Vasc Endovasc Surg 1996; 11: 421-424. intermittent claudication results in the generation of oxygen 67 TSANC GM, SANGI-IERA K, GOSLING Pet al. Pharmacological

Eur J Vasc Endovasc Surg Vol 15, January 1998 16 P.V. Tisi and C. P. Shearman

reduction of the systemically damaging effects of local isch- adhesiveness, plasma fibrinogen and fibrinolysis. BrJ Surg 1973; aemia. Eur J Vasc Surg 1994; 8: 205-208. 60: 386-389. 68 BLANNAD. von Willebrand factor as a marker of injury to the 92 BANERJEE AK, PEARSON J, GILLILAND EL et aI. A six-year endothelium in inflammatory vascular disease. J Rheumatol prospective study of fibrinogen and other risk factors associated 1993; 20: 1469-1471. with mortality in stable claudicants. Thromb Haemost 1992; 68: 69 BLANNAD, McCOLLUM CN. von Willebrand factor, endothelial 261-263. cell damage and atherosclerosis. Eur J Vasc Surg 1994; 8: 10-15. 93 SMITH EB, K~N GA, GRANT A, STIRK C. Fate of fibrinogen in 70 BLANN AD, McCOLLUM CN. Circulating endothelial cell/leu- human arterial intima. Arteriosclerosis 1990; 10: 263-275. kocyte adhesion molecules in atherosclerosis. Thromb Haemost 94 HEINRICH J, SCHULTEH, SCHONEELDR~ KOHLER E, ASSMANG. 1994; 72: 151-154. Association of variables of coagulation, fibrinolysis and acute- 71 BLANN AD, DOBROTOVA M, KUBISZ P, McCOLLUM CN. von phase with atherosclerosis in coronary and peripheral Willebrand factor, soluble P-selectin, tissue plasminogen ac- and those arteries supplying the brain. Thromb Haemost 1995; tivator and plasminogen activator inhibitor in atherosclerosis. 73: 374-379. Thromb Haemost 1995; 74: 626-630. 95 HAMSTENA, DE FAIRE U, ISELIUSL, BLOMBAOM. Genetic and 72 Ross R. The pathogenesis of atherosclerosis: a perspective for cultural inheritance of plasma fibrinogen concentration. Lancet the 1990s. Nature 1993; 362: 801-809. 1987; 2: 988-990. 73 ERNST EE, HAMMERSCHMIDT DE, BAGGE U, MATRAI A, 96 FOWKES FGR, CONNOR JM, SMITH FB, WOOD J, DONNAN PT, DORMANDY JA. Leukocytes and the risk of ischaemic diseases. LOWE GDO. Fibrinogen genotype and risk of peripheral athero- ]AMA 1987; 257: 2318-2324. sclerosis. Lancet 1992; 339: 693-696. 74 LEAKEDS. Oxidised low density lipoproteins and atherogenesis. 97 MACHIN SJ, MACKIE IJ. Routine measurement of fibrinogen Br Heart J 1993; 69: 476-478. concentration. Br Med J 1993; 307: 882-883. 75 WITZtTMJL. The oxidation hypothesis of atherosclerosis. Lancet 98 CRUIKSHANK AM, FRASER WD, BURNS HJG, VAN DAMME J, 1994; 344: 793-795. SHENKIN A. Response of serum interleukin-6 in patients under- 76 WITZUM JL, STEINBERGD. Role of oxidised low density lipo- going elective surgery of varying severity. Clin Sci 1990; 79: protein in atherogenesis. J Clin Invest 1991; 88: 1785-1792. 161-165. 77 LIAO F, ANDALIBIA, LUSlS AJ, FOGELMANAM. Gentic control 99 PEPYS MB. Rheumatoid arthritis: the role of acute-phase pro- of the inflammatory response induced by oxidised lipids. Am teins. Br J Rheumatol 1993; 32 (Suppl. 3): 1-2. f Cardiol 1995; 75: 65B-66B. 100 KUSHNER I, GEWURZ H, BENSON MD. C-reactive protein and 78 Gown AM, TSUKADA T, Ross R. Human atherosclerosis. II. the acute-phase response. J Lab Clin Med 1981; 97: 739-749. Immunocytochemical analysis of the cellular compositon of 101 PEPYS MB, BALTZ ML. Acute phase proteins with special ref- human atherosclerotic lesions. Am J Path 1986; 125: 191-207. erence to C-reactive protein and related proteins (pentaxins) 79 Ross R. The pathogenesis of atherosclerosis - an update. N and serum amyloid A protein. Adv Immunol 1983; 34: 141-212. 102 MAJEWSKIW, ZIELINSKI A, LACIAK Met aI. C-reactive protein EngI J Med 1986; 314: 488-500. and alpha-l-acid glycoprotein in monitoring of patients with 80 MITCtIINSON MJ, BALL RY. Macrophages and atherogenesis. chronic arterial occlusion of the lower limbs. Eur J Vasc Surg Lancet 1987; 2: 146-148. 1993; 7: 628-623. 81 HEINRICHPC, CASTELLJV, ANDUST. Interleukin-6 and the acute 103 WrLKINSJ, GALLIMOREJR, TENNENTGA et al. Rapid automated phase response. Biochem f 1990; 265: 621-636. enzyme immunoassay of serum amyloid A. Clin Chem 1994; 40: 82 STUARTJ. The acute-phase reaction and haematological stress 1284-1290. syndrome in vascular disease, hzt f Microcirc 1984; Clin Exp 3: 104 MENDALL MA, PATEL P, STRACHAN D, NORTH]FIELD TC. C 115-129. Reactive protein and its relation to cardiovascular risk factors: 83 KHAW KT, WOODMOUSEP. Interrelation of vitamin C, infection, a population based cross sectional study. Br Med J 1996; 312: haemostatic factors and cardiovascular disease. Br J Med 1995; 1061-1065. 310: 1559-1562. 105 TIsI PV, CRow AJ, S~EARMAN CP. The use of inflammatory 84 KANNEL WB, WOLF PA, CASTELLI WP, D'AGOSTINO RB. Fib- markers in the assessment of intermittent ciaudication. Eur Surg rinogen and risk of cardiovascular disease. The Framingham Res 1996; 28 (Suppl. 1): 65. Study. JAMA 1987; 258: 1183-1186. 106 MAS~RI A, BIASUCCILM, LIuzzo G. Inflammation in ischaemic 85 MEADET~vV, BROZOVIC M, CHAKRABARTIRR, HAINESAP, IMESON heart disease. Br Med J 1996; 312: 1049-1050. JD. Haemostatic function and ischaemic heart disease: principal 107 T~OMPSON SG, KIENAST J, PYKE SDM, HAVERKATEE VAN DE results of the Northwick Park Heart Study. Lancet 1986; 2: LOO JCW. Haemostatic factors and the risk of myocardial 533-537. infarction or sudden death in patients with angina pectoris. N 86 YARNELL JWG, BAKER Ia, SWEETNAM PM, BAINTON D. Fib- Engl f Med 1995; 332: 635-641. rinogen, viscosity and count are major risk 108 LI•zzo G, BIAsUCCI LM, GALLIMOREJR et al. The prognostic factors for ischaemic heart disease. Circulation 1991; 83: 836-844. value of C-reactive protein and serum amyloid a protein in 87 FOWKES FG. Fibrinogen and peripheral arterial disease. Eur severe unstable angina. N Engl J Med 1994; 331: 417-424. Heart J 1995; 16 (Suppl. A): 36-40. 109 IK/IAURYCPJ. Comparative study of serum amyloid A protein 88 LEE AJ, FOWKES FGR, RATTRAY A, RUMLEY A, LOWE GDO. and C-reactive protein in disease. Clin Sci 1985; 68: 233-238. Haemostatic and theological factors in intermittent claudication: 110 LIAO F, ANDALIBI A, QIAO JH, ALLAYEE H, FOGELMAN AM, the influence of smoking and extent of arterial disease. Br J LUSlS AJ. Genetic evidence for a common pathway mediating Haem 1996; 92: 226-230. , inflammatory gene induction and aortic fatty 89 LowE GD, FOWKES FG, DAWES J, DONNAN PT, LENNIE SE, streak formation in mice. J Clin Invest 1994; 94: 877-884. HOUSLEY E. Blood viscosity, fibrinogen, and activation of co- 111 MALLE E, STEINMETZA, RAYN~S JG. Serum amyloid A (SAA): agulation and leukocytes in peripheral arterial disease and the an acute phase protein and apolipoprotein. Atherosclerosis 1993; normal population in the Edinburgh Artery Study. Circulation 102: 131-146. 1993; 87: 1915-1920. 112 RAYNES JG, EAGLING S, MCADAM KP. Acute-phase protein 90 DORMANDY JA, HOARE E, KHATTAB AH, ARROWSMITH DE, synthesis in human hepatoma cells: differential regulation of DORMANDY TL. Prognostic significance of rheological and bio- serum amyloid A (SAA) and haptoglobulin by interleukin-1 chemical findings in patients with intermittent claudication. Br and interleukin-6. Ctin Exp Immunol 1991; 83: 488-491. Med J 1973; 4: 581-583. 113 URIELI-SHOVALS, MEEK RL, HANSON RH, ERIKSEN N, BENDITT 91 HAMER JD, ASHTON F~ MEYNELL MJ. Factors influencing pro- EP. Human serum amyloid A genes are expressed in monocyte/ gnosis in the surgery of peripheral vascular disease: platelet macrophage cell lines. Am J Pathol 1994; 145: 650-660.

Eur J Vasc Endovasc Surg Vol 15, January 1998 Exercise-induced Inflammation in Intermittent Claudication 17

114 MEEKRL, URIELI-SHOVALS, BENDITTEP. Expression of apolipo- 119 BADOLATOR~ WANG JM, MURPHY WJ et al. Serum amyloid A is protein serum amyloid A mRNA in human atherosclerotic a chemoattractant: induction of migration, adhesion, and tissue lesions and cultured vascular cells: implications for serum infiltration of monocytes and polymorphonuclear leukocytes. J amyloid A function. Proc NatI Acad Sci U.S.A. 1994; 91: 3186- Exp Med 1994; 180: 203-209. 3190. 120 MAURY CPJ, TOTTERMAN KJ, GREF CG, EHNHOLM C. Serum 115 GORDONDJ, RIEKINDBM. High-density lipoprotein - the clinical amyloid A protein, apolipoprotein A-I and apolipoprotein B implications of recent studies. N Engl JMed 1998; 321: 1311-1316. during the course of acute . J Clin Path 116 BLACKBURNWDJ. Validity of acute phase proteins as markers 1988; 41: 1263-1268. of disease activity. J Rheumatol Supp 1994; 42: 9-13. 121 TIsI PV, H~LS~ M, CHULAKADABBAA, GOSLINGP, SHEARMANCP. 117 KISILEVSKYR, SLTBRAFIMANYANL. Serum amyloid A changes Exercise training for intermittent claudication: does it adversely high density lipoprotein's cellular affinity. A clue to serum affect biochemical markers of the exercise-induced in- amyloid A's principal function. Lab Invest 1992; 66: 778-785. flammatory response? Eur J Vasc Endovasc Surg 1997; 14: 344- 118 Xu L, BADOLATOR, MURPHY WJ et al. A novel biologic function 350. of serum amyloid A. Induction of T-lymphocyte migration and adhesion. J ImmunoI 1995; 155: 1184-1190. Accepted 23 July 1997

Eur J Vasc Endovasc Surg Vol 15, January 1998