European Journal of Clinical Investigation (1999) 29, 577–581

Endothelial activation and response in patients with handarm vibration syndrome

G. Kennedy, F. Khan, M. McLaren and J. J. F. BelchSection of Vascular Medicine & Biology, University Department of Medicine, Ninewells Hospital & Medical School,Dundee, UK

Abstract Background Hand–arm vibration syndrome (HAVS) is a form of secondary Raynaud’sphenomenon (RP) of occupational origin. In other forms of RP, blood and blood vessel wallinteraction is one factor in the pathophysiology. Cytokines and cell adhesion molecules bothplay an important role in this interaction, and basal vascular tone and vasodilatation areregulated by nitric oxide.

Methods Blood flow responses to acetylcholine (ACh) and sodium nitroprusside (SNP)and levels of soluble intercellular adhesion molecule-1 (sICAM-1) and the inflammatorycytokine interleukin 8 (IL-8) were measured in eight male patients with vibration whitefinger disease, which is part of HAVS, and in eight healthy matched male control subjects.

Results sICAM-1 levels were statistically higher (P ¼ 0·02, Mann–Whitney U-test) andIL-8 levels (P< 0·01, Mann–Whitney) were significantly lower in the patient group. Thepatients with HAVS had significantly reduced vascular responses to SNP (P <0·05, ANOVA).

Conclusions In this study, we reveal differences in vascular responses to SNP that suggestthere may be an impairment of the smooth muscle response to nitric oxide in patients withHAVS. The increase in sICAM-1 that occurs in patients with HAVS suggests that leucocyteadhesion is increased and that adherent neutrophils may contribute to the microvasculardamage seen in this disease. The impeded flow of blood cells through the microcirculationmay result in the low levels of circulating IL-8 due to the cytokine binding to erythrocytes.The possible role of NO activity in HAVS warrants further investigation.

Keywords Endothelium, hand–arm vibration syndrome, intercellular adhesion molecule-1,interleukin 8, Raynaud’s phenomenon, vascular tone.Eur J Clin Invest 1999; 29 (7): 577–581

Introduction

Raynaud’s phenomenon (RP) is a vascular disordercharacterized mainly by episodic digital ischaemia causedby microvascular spasm in response to cold or emotionalstress [1]. This condition can occur alone, as in primaryRaynaud’s disease, or it can be associated with otherconditions, in which case it is termed secondary RP orRaynaud’s syndrome (RS) [2]. Workers who are exposed topolyvinyl chloride or long-term cold temperatures candevelop secondary RP of occupational origin [1]. Anotherform of occupational RS is vibration white finger disease,

which is now know to be a part of the hand–arm vibrationsyndrome (HAVS). Vibration white finger disease occurs inworkers exposed to vibrating machines such as pneumaticdrills, chain saws and buffs. It is estimated that between40% and 90% of all workers using vibratory equipment willhave Raynaud’s syndrome, although the symptoms may bereversible in a quarter of the patients if they change jobsearly in the disease [3]. Recently, there has been anincreasing number of compensation claims from peoplesuffering from vibration white finger disease, and,therefore, it is important to identify and obtain a betterunderstanding of the possible factors involved in thepathophysiology of vibration white finger disease.

Neurogenic mechanisms, blood and blood vessel wallinteractions and an abnormal immunological responsehave been considered to be important factors in thepathophysiology of RP, particularly in RP associated withconnective tissue disorders [4]. The last two potentialcauses of RP involve the cell adhesion molecules

Q 1999 Blackwell Science Ltd

Correspondence to: Dr G. Kennedy, Section of VascularMedicine & Biology, University Department of Medicine,Ninewells Hospital & Medical School, Dundee DD1 9SY, UK.E-mail: G.KIRK@dundee.ac.uk

Received 27 October 1998; accepted 27 February 1999

578 G. Kennedy et al.

(CAMs). These molecules are present on the endothelium,white blood cells and platelets. Intercellular adhesionmolecule-1 (ICAM-1) is basally expressed on endothelialcells and monocytes and can be up-regulated on fibroblastsand lymphocytes [5]. ICAM-1 expression can be inducedby various cytokines, including interleukin 1b and tumournecrosis factor a [6]. Interleukin 8 (IL-8) is a member ofthe chemokine family proinflammatory cytokines. Thiscytokine is chemotactic to neutrophils and can allowunstimulated neutrophils to adhere to endothelial cells byincreasing the expression of the b2-integrins [7].

The endothelium plays an important role in maintainingvascular tone by producing substances such asendothelium-derived relaxing factor/nitric oxide (EDRF/NO) and endothelin. Basal vascular tone and vasodilata-tion are regulated by NO [8], and endothelial dysfunctioncan lead to impaired NO release [9] and activity. We havepreviously shown a reduction in vascular responses to theiontophoresis of acetylcholine (ACh) and sodiumnitroprusside (SNP) in patients with primary RP comparedwith responses in control subjects [10].

The aim of the present study was to measure the levels ofsICAM-1 and IL-8 and vascular responses to endothelium-dependent and -independent vasodilators in patients withRP secondary to vibration white finger disease.

Methods

Subjects

Ethical permission was obtained from the local ethicscommittee and all volunteers gave written informedconsent. Eight men with RP secondary to vibration whitefinger disease (median age 59·5 years, range 29–76 years)and eight healthy men (median 59·0 years, range29–71 years) were enrolled into the study. No patient orcontrol subject was taking medication at the time of study.The mean duration of RP was 10·8 years (range2–20 years).

Blood samples

A 10-mL blood sample was taken from the antecubitalfossa using a 19-gauge needle. Light tourniquet pressurewas applied, if required, to assist venepuncture. Thepressure was released for at least 10 s before the bloodwas drawn. Blood was collected into tubes containingclotting beads with no anticoagulant present and placedinto a 378C water bath for 1 h. After this time, the samplewas centrifuged for 15 min at 3500 rpm, the serum wasremoved, aliquoted and stored at ¹708C until assayed. Allblood samples were obtained at the same time of day foreach visit (in the morning) after a standard light breakfastto avoid the effects of circadian variations on sICAM-1

levels as shown by us previously [11]. sICAM-1 and IL-8levels were determined by enzyme immunoassay (R & DSystems, Abington, Oxon, UK).

Assessment of digital cutaneous vascular responses

Experiments were conducted in a constant roomtemperature set at 22–238C. Patients and subjects wereseated comfortably with the arms supported at heart level.Skin perfusion (termed erythrocyte flux) was measuredcontinuously at the dorsal surface of the fingertip using alaser Doppler flowmeter (MBF3/D, Moor Instruments,Axminster, UK). The laser Doppler technique appliedhere measures blood flow in microvessels (capillaries,arterioles, venules) at the dorsal aspect of the fingerwhere vessel diameter is of the order of 17–25 mm in thesubpapillary plexus and approximately 50 mm in the deepdermis. Assessment of endothelium-dependent and-independent vascular responses were made duringiontophoretic application of ACh and SNP, respectively,according to our previously described protocol [10]. Inbrief, the skin of the most affected fingers (usually the indexand middle fingers) was gently cleaned with alcohol and adirect electrode chamber was attached to a site of unbrokenskin, approximately midway between the middle and distalphalangeal joint, using double-sided adhesive tape. Theindifferent electrode was a Velcro strap soaked indeionized, sterile water and placed around the wrist tocomplete the circuit. The leads from the direct electrodechamber and the indifferent electrode were connected to abattery-powered iontophoresis controller (MIC 1, MoorInstruments), which provided a direct current for thedelivery of solutions.

ACh (Sigma Chemical, St Louis, MO, USA) was madeup to a 1% solution in deionized, sterile water and wasiontophoresed for 10 s at an anodal current strength of0·2 mA. Skin erythrocyte flux was measured continuouslyduring the period of iontophoresis, and the subsequentvascular response was measured for 4 min. To increase thedose, ACh was iontophoresed for 20 and 40 s, with theerythrocyte flux being recorded for 4 min between eachdose.

SNP was made up to a 1% solution in deionized, sterilewater and was protected from the light. A cathodal currentwas used to iontophorese SNP at similar doses to those forACh (i.e. for 10, 20 and 40 s).

Statistical analysis

Differences in sICAM-1 and IL-8 levels between thepatient and control groups are expressed as medians andinterquartile ranges and were analysed using thenon-parametric Mann–Whitney U-test. Skin erythrocyteflux is expressed in arbitrary units (AUs) and as means6 SEM. The area under the curve (AUC) was calculatedfor a 4-min basal period and for each 4-min period afteriontophoresis of ACh and SNP. The response to each dose

Q 1999 Blackwell Science Ltd, European Journal of Clinical Investigation, 29, 577–581

Vascular responses in vibration white finger 579

Q 1999 Blackwell Science Ltd, European Journal of Clinical Investigation, 29, 577–581

was expressed as the difference between the AUC afteriontophoresis and basal AUC. Differences in the AUC forACh and SNP between patients and control subjects weretested using two-way analysis of variance for repeatedmeasures, followed by t-tests at each dose when asignificant difference was found. The null hypothesis wasrejected at P <0·05. Statistical analyses were performedusing SPSS software (SPSS, IL, USA).

Results

Patients with HAVS had statistically higher sICAM-1 levels(median 284·5 ng mL¹1, range 168·2–531·8 ng mL¹1)than the levels measured from the healthy control group(median 209·7 ng mL¹1, range 99·6–285·5 ng mL¹1,P ¼ 0·02) (Fig. 1). IL-8 levels were statistically lower inthe patient group than in the control group (median13·8 pg mL¹1, range 11·5–25·4 pg mL¹1; and median27·7 pg mL¹1, range 16·1–70·4 pg mL¹1, respectively;P <0·01) (Fig. 1).

The baseline skin erythrocyte flux was not significantlydifferent between HAVS patients and control subjects(23·7 6 5·2 vs. 42·1 6 12·2 AU respectively), in spite of atrend towards lower values in the HAVS group. Figure 2shows that skin vascular responses to ACh were notsignificantly different between control subjects and patientswith HAVS. In contrast, vascular responses to SNP weresignificantly reduced in patients with HAVS (P< 0·05,ANOVA) (Fig. 3). Responses were reduced in the HAVSgroup by factors of 14·3 (P< 0·05, post hoc t-tests), 3·4and 2·4, for the three doses of SNP. There were nosignificant correlations between vascular responses toACh and SNP and levels of sICAM-1 and IL-8.

Discussion

We have shown for the first time that, compared withhealthy subjects, patients with vibration white fingerdisease have elevated sICAM-1 levels, decreased IL-8levels and impaired vascular responses to the

Figure 1 Whisker box plots showing (a) sICAM-1 and (b) IL-8levels in patients with RP secondary to HAVS (open boxes) andhealthy controls (hatched boxes) (median, 5% and 95%percentile points shown) (Mann–Whitney U-test).

Figure 2 Vascular responses to increasing doses of acetylcholinein control subjects (n ¼ 8) and patients with RP secondary toHAVS (n ¼ 8). Skin erythrocyte flux measured in arbitraryperfusion units (PU) and acetylcholine dose in millicoulombs(values are 6 SEM).

Figure 3 Vascular responses to increasing doses of sodiumnitroprusside in control subjects (n ¼ 8) and patients with RPsecondary to HAVS (n ¼ 8). The skin erythrocyte flux wasmeasured in arbitrary perfusion units (PU), and the sodiumnitroprusside dose in millicoulombs (values are 6 SEM).

580 G. Kennedy et al.

nitrovasodilator SNP. Interactions between the blood andendothelium, abnormal immune and inflammatoryresponses and neurogenic mechanisms are probablyinvolved in the pathophysiology of RP. The CAMs areexpressed on a variety of cell types including theendothelium, monocytes, neutrophils, lymphocytes andfibroblasts [12]. Circulating levels of the CAMs are thoughtto be predictive of the extent of the cell to cell interactions[13] and any exacerbation of this mechanism can lead totissue damage [14].

In this preliminary study we are looking at newer aspectsof vascular abnormalities in vibration white finger disease,and these newer observations are supported by ourprevious findings. We have shown that patients withsystemic sclerosis (SSc) have elevated levels of sICAM-1compared with levels found in control subjects [15]. Theelevated levels of sICAM-1 in patients with RS mayindicate increased involvement of the leucocytes and theextracellular matrix in inflammation and damage of theblood vessel wall. We have previously documented aninflammatory element to vibration white finger disease,where we showed an increase in the production of thewhite blood cell proinflammatory product leukotriene B4

and decreased plasma thiol concentrations, indicatingincreased free radical activity, in 25 patients with vibrationwhite finger compared with matched healthy controlsubjects [4]. These findings support two possible effectsof the elevated oxidative stress that is observed in HAVS.ICAM-1 is involved not only in the adherence of the whiteblood cells (WBCs) to the endothelium but also in theadherence of WBCs to other WBCs. These hard adherentWBCs decrease blood flow in the microcirculation [16].ICAM-1 is also up-regulated at the time of reperfusion,the period at which blood flow returns to the occludedvessel. Interactions between ICAM-1 and its ligands (theb2-integrins) results in the adherence of the leucocyte tothe vessel wall, which can then lead to microvascular injury[17]. The extravasated leucocyte can produce cell growthfactors and the generation of cytotoxic factors that canlead to smooth muscle cell damage and death [18]. It hasbeen shown that a high level of plasma sICAM-1 isassociated with a high risk of a future myocardialinfarction [19].

Patients with RS have previously been shown, by thisgroup, to have less deformable (i.e. stiffer) red blood cells(RBCs), which may reduce blood flow [20]. Erythrocytesare capable of binding IL-8, resulting in bound IL-8becoming biologically inactive. This can significantlyreduce the concentration of IL-8 in blood plasma [21].The lower levels of IL-8 found in our patients may be dueto an increased binding of the cytokine to the RBCbecause of the reduced flow of RBC through the micro-circulation, as shown by a 44% reduction in basal skinerythrocyte flux.

The physiological precursor of NO, L-arginine, reduceshuman monocyte adhesion to endothelial cell monolayersassociated with the attenuation of ICAM-1 endothelialexpression levels [22]. It is possible that the elevation insICAM-1 levels in our HAVS patients is related to the

decreased activity of NO. Indeed, vascular responses tothe NO donor, SNP, were significantly reduced in HAVS.The reason for a decreased responsiveness of thevasculature to NO is not clear but may be due to increasedoxidative stress [23], which has been shown in otherdiseases to reduce NO activity [24]. This is supportedby the trend towards a reduction in all blood flowresponses. A possible reason for the lack of statisticalsignificance is that in skin, iontophoresis of AChstimulates both the production of NO and prostanoids[10]. If the fault in vibration white finger disease liesspecifically at the level of NO activity, then the additionalproduction of prostanoids may compensate for thereduced vasodilatory response to ACh. In patients withSSc or Dupuytren’s contracture, the only morphologicalchange found is fibrosis of the skin, and with SSc, fibrosisof the internal organs [25]. Therefore, it could be that theattenuated blood flow response to SNP may reflect a‘stiff ’ basement membrane, an increase in smoothmuscle mass in the vessels, or a stiffening of the tissuearound the vessel resulting in compression. Howeverthese could explain some of the decrease in blood flowresponse in the patients with HAVS after iontopheresis ofSNP, it cannot be the only explanation because thickeningof the tissue would also result in a decrease in blood flowresponse to ACh.

However, other possible causes of decreased blood flowdue to mechanical stress caused by vibration are damage tolocal nerve endings resulting in general neuronal loss,especially in those nerves containing the potent vasodilatorcalcitonin gene-related peptide (CGRP) [26].

One factor to take into consideration when usingiontophoresis is the non-specific effect of current on skinerythrocyte flux [27]. At the currents used in this study, wehave shown previously that there is a small electrical effectonly at the 8-mC charge using an anodal current [28] and acathodal current (unpublished observations). When datafor the 8-mC charge are excluded from the statisticalanalysis, the results are the same (i.e. significantly reducedvascular responses to SNP but not to ACh) (data notshown). In a recent study, Morris et al. [29] usediontophoresis of ACh and SNP in patients with diabetesmellitus and stated that their conclusions were the samewhether the non-specific effects of current were accountedfor or not (i.e. reduced skin microvascular vasodilatorresponses in patients with diabetes mellitus comparedwith control subjects).

In this study, we have shown differences in vascularresponses to SNP which suggest that there may be animpairment in the smooth muscle response to NO inpatients with HAVS. The attenuation of IL-8 levels inthe patient group may reflect the impeded flow of blood cellsthrough the microcirculation, which is characteristic inpatients with RP. We have shown an elevation in sICAM-1that demonstrates that WBC adhesion occurs and thatextravasation of leucocytes from the circulation into theextracellular matrix may also contribute to the vasculardamage that is observed in this RP. The possible role ofNO activity in HAVS warrants further investigation.

Q 1999 Blackwell Science Ltd, European Journal of Clinical Investigation, 29, 577–581

Vascular responses in vibration white finger 581

Q 1999 Blackwell Science Ltd, European Journal of Clinical Investigation, 29, 577–581

Acknowledgements

The Raynaud’s and Scleroderma Association and theArthritis and Rheumatism Council (Grant B202) kindlyprovided support for this study. We would like to thank DrMeilien Ho and Dr Douglas Veale for their assistance.

References

1 Belch JJF. Temperature- associated vascular disorders:Raynaud’s phenomenon and erythromelalgia. In: Tooke JE,Lowe GDO, editors. A textbook of vascular medicine. London:Arnold; 1996. p.329–51.

2 Goodfield MJ, Orchard MA, Rowell NR. Whole bloodplatelet aggregation and coagulation factors in patients withsystemic sclerosis. Br J Rheumatol, 1993;55:7–20.

3 Belch JJF. Raynaud’s phenomenon. Curr Opin Rheumatol,1990;2:937–41.

4 Lau C, O’Dowd A, Belch JJF. White cell activation in theRaynaud’s phenomenon of systemic sclerosis and vibrationwhite finger syndrome. Ann Rheum Dis, 1992;51:249–52.

5 Pigott R, Power C. The adhesion molecules facts book. London:Academic Press; 1993.

6 Godin C, Caprani A, Dufaux J, Flaud P. Interactionsbetween neutrophils and endothelial cells. J Cell Sci,1993;106:441–52.

7 Roberts PJ, Pizzey AR, Khwaja A, Carver JE, Mire-Sluis AR.The effects of interleukin-8 on neutrophil fMetLeuPhereceptors, CD11b expression and metabolic activity, incomparison and combination with other cytokines. Br JHaematol, 1993;84:586–94.

8 Vallance P, Collier J, Moncada S. Effects of endothelium-derived nitric oxide on peripheral-arteriolar tone in man.Lancet, 1989;2:99–1000.

9 Lau CS, Khan F, Brown R, Wilson SB, Belch JJF. Digitalblood flow response to whole blood body warming, coolingand rewarming in patients with Raynaud’s phenomenon.Angiology, 1995;46 (1):1–10.

10 Khan F, Litchfield SJ, McLaren M, Veale DJ, Littleford RC,Belch JJF. Oral L-arginine supplementation and cutaneousvascular responses in patients with primary Raynaud’sphenomenon. Arthritis Rheum, 1997;40:352–7.

11 Maple C, Kirk G, McLaren M, Belch JJF. A circadianvariation exists for plasma levels on intercellular adhesionmolecule-1 and E-selectin in healthy volunteers. Clin Sci,1998;94:537–40.

12 Sollberg S, Peltonen J, Uitto J, Jimenez SA. Elevatedexpression of b1 and b2 integrins, intercellular adhesionmolecule-1, and endothelial leukocyte adhesion molecule-1 inthe skin of patients with systemic sclerosis of recent onset.Arthritis Rheum, 1992;35:290–8.

13 Gearing AJH, Newman W. Circulating adhesion molecules indisease. Immunol Today, 1993;14:506–12.

14 Cronstein BN, Weissmann G. The adhesion molecule sininflammation. Arthritis Rheum, 1993;36:14–57.

15 Veale DJ, Kirk G, McLaren M, Belch JJF. Elevation ofcirculating soluble ICAM-1 in response to venous occlusionin patients with systemic sclerosis and their relatives. ArthritisRheum, 1993;36:S271.

16 Boyd AW, Wawryk SO, Burns GF, Fecondo JV. Intercellularadhesion molecule (ICAM-1) has a central role in cell-cellcontact-mediated immune mechanism. Proc Natl Acad SciUSA, 1998;85:3095–9.

17 Welbourn CRB, Golman G, Paterson IS, Valeri CR, Shepro D,Hechtman HB. Pathophysiology of ischaemia reperfusioninjury: central role of the neutrophil. Br J Surg, 1991;78:651–5.

18 Fukuo K, Nakahashi T, Nomura S, et al. Possibleparticipation of Fas-mediated apoptosis in the mechanism ofatherosclerosis. Gerontology, 1997;43:35–42.

19 Ridker PM, Hennekens CH, Roitman-Johnson B, StampferMJ, Allen J. Plasma concentration of soluble intercellularadhesion molecule-1 and risks of future myocardia infarctionin apparently healthy men. Lancet, 1998;351:88–92.

20 Belch JJF, McLaren M, Anderson J, Lowe GD, Sturrock RD,Forbes CD. Increased prostacyclin metabolites and red celldeformability in patients with systemic sclerosis and Raynaud’ssyndrome. Prostaglandins Leukot Med, 1985;17:1–9.

21 Darbonne WC, Rice GC, Mohler MA, et al. Red blood cellsare a sink for interleukin-8, a leukocyte chemotaxin. J ClinInvest, 1991;88:1362–9.

22 Adams MR, Jessup W, Hailstones D, Celermajer DS.L-Arginine reduces human monocyte adhesion to vascularendothelium and endothelial expression of cell adhesionmolecules. Circulation, 1997;95:662–8.

23 Lau CS, Bridges A, Muir A, Scott N, Bancroft A, Belch JJF.Further evidence of increased polymorphonuclear cell activityin patients with Raynaud’s phenomenon. Br J Rheumatol1992 b;31:375–80.

24 Harrison DG, O’Hara Y. Physiologic consequences ofincreased vascular oxidant stresses in hypercholesterolemiaand atherosclerosis: implications for impaired vasomotion.Am J Cardiol, 1995;75:75B–81.

25 Black C. Systemic sclerosis. Topical Rev, 1996;3 (7):1–8.26 Grand Round. Raynaud’s phenomenon. Lancet,

1995;346:283–90.27 Grossman M, Jamieson MJ, Kellog DL, et al. The effect of

iontophoresis on the cutaneous vasculature: evidence forcurrent-induced hyperemia. Microvasc Res, 1995;50:444–52.

28 Khan F, Davidson NC, Littleford RC, Litchfield SJ,Struthers AD, Belch JJF. Cutaneous vascular responses toacetylcholine are mediated by a prostacyclin-dependentmechanism in man. Vasc Med, 1997;2:82–6.

29 Morris SJ, Shore AC, Tooke JE. Responses of the skinmicrocirculation to acetylcholine and nitroprusside in patientswith NIDDM. Diabetologia, 1995;38:1337–44.