Probucol improves symptoms and reduces ... - Semantic Scholar

Rheumatology 1999;38:309–315

Probucol improves symptoms and reduces lipoprotein oxidation susceptibility in patients with Raynaud’s phenomenon C. P. Denton, T. D. Bunce1, M. B. Dorado1, Z. Roberts1, H. Wilson, K. Howell, K. R. Bruckdorfer1 and C. M. Black Academic Unit of Rheumatology and 1Department of Biochemistry and Molecular Biology, Royal Free Hospital School of Medicine, London, UK Abstract Objective. Reactive oxygen species have been implicated in the pathogenesis of inflammatory and vascular disease. We have undertaken a controlled trial to evaluate probucol, a synthetic antioxidant, as a potential therapy for Raynaud’s phenomenon. Methods. The study cohort included patients with systemic sclerosis (SSc; n = 20), primary Raynaud’s phenomenon (n = 15) or ‘autoimmune Raynaud’s’ (n = 5). Patients were allocated to receive either probucol (500 mg daily) or nifedipine (20 mg daily) for 12 weeks. Clinical and biochemical variables at baseline were compared with those at completion of treatment. Evaluation included assessment of Raynaud’s attack frequency and severity by visual analogue scale, measurement of low-density lipoprotein (LDL) oxidation lag time, and plasma concentrations of cholesterol, triglyceride, vitamin E and vitamin C. Results. There was a significant reduction of both the frequency and severity of Raynaud’s attacks in the patients who received probucol, but not in the control group. LDL oxidation lag time, reflecting in vitro susceptibility to oxidation, was also increased by probucol therapy and serum cholesterol levels were significantly reduced. Similar changes were observed in both SSc- and non-SSc-associated Raynaud’s cases. Conclusion. These data suggest that probucol may be useful for the symptomatic treatment of Raynaud’s phenomenon and also reduces LDL oxidation susceptibility. Since oxidized lipoproteins may mediate vascular damage in SSc, the use of probucol could have additional disease-modifying benefits. Based upon the results of this pilot study, further evaluation of this novel form of therapy is warranted. K : Probucol, Raynaud’s phenomenon, Scleroderma, Antioxidant.

Raynaud’s phenomenon manifests as episodic cold- or stress-induced vasospasm affecting the extremities, and can cause major disability in its severe forms [1]. Although most often occurring in otherwise healthy individuals (primary Raynaud’s phenomenon), it is generally more severe when associated with an underlying autoimmune rheumatic disease. Raynaud’s is often a feature of systemic sclerosis (SSc), polymyositis or systemic lupus erythematosus, and when associated with these conditions is termed secondary Raynaud’s phenomenon. A third subtype is recognized in sufferers who have abnormal nailfold capillaries and positive antinuclear antibodies, but no other clinical features of a defined connective tissue disease. Such patients are at

increased risk of developing an autoimmune rheumatic disease and they have been designated ‘autoimmune Raynaud’s’ [2]. The pathophysiology of Raynaud’s phenomenon is complex and its aetiopathogenic relationship to any coexistent connective tissue disease is unclear [3]. Reactive oxygen species (including free radicals) have been implicated in the development of a number of rheumatic diseases [4], and are also potential mediators of the endothelial cell injury which is characteristic of secondary Raynaud’s [5]. One mechanism by which these reactive oxygen species might mediate endothelial cell injury is through their oxidative effect on plasma lipoproteins [6 ]. Oxidized lipoproteins, especially lowdensity lipoproteins (LDLs), may trigger endothelial cell damage in SSc [7], as well as in more prevalent vascular pathologies such as atherosclerosis [8]. It has been shown that they can modulate a number of important endothelial cell properties in vitro, and the pattern of modulation often relates to the extent of LDL oxidation [9].

Submitted 29 May 1998; revised version accepted 30 November 1998. Correspondence to: C. M. Black, Academic Rheumatology Unit, Royal Free Hospital, Pond Street, Hampstead, London NW3 2QG, UK.

309

© 1999 British Society for Rheumatology

310

C. P. Denton et al.

Our group has previously shown that plasma LDL isolated from SSc patients has an increased susceptibility to oxidation [7]. This is likely to be an indirect consequence of increased LDL oxidation in vivo, either due to increased oxidative stress, or to lower levels of dietary or endogenous antioxidants in these patients. Support for the latter is provided by reports of reduced micronutrient antioxidant levels in SSc patients [7, 10]. These reports of lower levels of antioxidant vitamins are intriguing, especially as many SSc patients appear deliberately to choose a diet rich in nutritional antioxidants such as vitamin E, selenium or vitamin C. Furthermore, a questionnaire administered to patients revealed no reduction in the choice of fresh fruit and vegetables [7]. Probucol is a powerful synthetic antioxidant, developed originally as a drug to lower cholesterol [11]. We hypothesized that probucol would have beneficial clinical effects on Raynaud’s phenomenon through its antioxidant properties and, furthermore, by reducing the propensity of endogenous LDL to oxidation, that it might protect against vascular damage. Here we report the results of a trial comparing probucol with nifedipine, a calcium channel blocker widely used for the treatment of vasospastic symptoms in Raynaud’s.

Methods Patients Raynaud’s phenomenon was defined by triphasic colour changes (pallor–cyanosis–suffusion) affecting the extremities, precipitated by cold or emotion, and occurring at least once on most days. Patients with Raynaud’s phenomenon (n = 40) were enrolled into the study, which was approved by the Royal Free Hospital Ethical Practices Sub-committee. Half of the patients had Raynaud’s phenomenon in association with SSc. Of these, the majority (n = 15) had limited cutaneous SSc; the remainder (n = 5) fulfilled the criteria of the diffuse cutaneous subset [12]. Of the non-SSc patients, 15 had primary Raynaud’s phenomenon, defined by the absence of any clinical features of a connective tissue disease, positive autoimmune serology or significantly abnormal nailfold capillaroscopic appearance [13]. The remaining patients (n = 5) had no clinical evidence of rheumatic disease, but were found to have positive antinuclear antibodies and abnormal nailfold capillaroscopic appearance, and were categorized as ‘autoimmune Raynaud’s’. There were eight current (three with SSc) and eight ex-smokers (four with SSc) in the study group. Patients were randomized to receive either probucol or nifedipine, with half of the SSc cases entering each treatment arm. The clinical and demographic features of each treatment group are summarized in Table 1. Study design and drug formulation This was a randomized parallel group study. Nifedipine is widely prescribed as a treatment for Raynaud’s phenomenon and was chosen for the control arm of the study. It was felt that an active-treatment control arm would facilitate recruitment into the study and provide

T 1. Baseline characteristics of probucol and nifedipine treatment groups Probucol Mean

Nifedipine ..

Mean

..

Demographic Patient number Female Age Smoker Ex-smoker

20 15 43 4 6

– – 13 – –

20 15 47 4 2

– – 14 – –

Clinical SSc-RP Autoimmune RP Primary RP Years RP Years SSc

10 1 9 13.8 5.8

– – – 13.3 4.7

10 4 6 10.3 4.3

– – – 10.1 3.5

RP, Raynaud’s phenomenon.

a useful comparison with current clinical practice. Patients stopped taking maintenance vasodilator drugs, other than those prescribed for treatment of hypertension, 4 weeks prior to commencement of the trial. Following a 2 week baseline assessment period, patients received either probucol or nifedipine for 12 weeks. Nifedipine was formulated as nifedipine retard in 10 mg capsules, and probucol in 250 mg tablets, and patients were instructed to take one tablet twice daily during the treatment period. To minimize weather and seasonal influences on Raynaud’s symptoms, the study took place over 4 months during one winter. Symptom diaries Symptom severity was self-assessed by patients using a visual analogue scale (0–10) for the most severe attack suffered each assessment day, and the number of attacks on that day was also recorded. The primary clinical end points were a change in the mean daily severity or frequency of Raynaud’s attacks between the baseline 2 week period and the final 2 weeks of treatment. Before commencing the study, subjects received careful training in completion of the symptom diary from an experienced investigator (HW ). Although this was not a blinded trial, patients were given their therapy labelled as ‘Study Drug’, which did not state into which arm of the trial they had been randomized, and it was clearly explained that they were equally likely to receive either probucol or nifedipine. Isolation of LDL and measurement of lag times for oxidation Blood (20 ml non-fasting) was drawn into tubes containing 63 m EDTA, as an anticoagulant, and 46 m DTPA, which specifically chelates free transition metal ions that may propagate free radical attack on lipoprotein lipids. Plasma was obtained by centrifugation of whole blood at 3000 g for 11 min. Plasma density was adjusted to 1.3 g/ml and the LDL fraction was isolated by centrifugation underlying a saline solution [7]. The LDL band was removed, concentrated (Amicon

Probucol for Raynaud’s phenomenon

Concentrator) and the purity of the LDL fraction was confirmed by SDS–PAGE [14]. Oxidation lag time was defined as the duration of the lag phase before LDL oxidation occurred following addition of 1.6 m Cu2+ to the LDL solution (0.25 g protein/l ). Oxidation was detected by increased absorbency at 234 nm, indicating the formation of conjugated diene intermediates generated in the oxidative attack of polyunsaturated fatty acids from the abstraction of hydrogen [15]. Biochemical variables Blood levels of micronutrient antioxidants were measured using validated assay protocols. Thus, vitamin E [16 ] and vitamin C [17] were quantified at baseline and completion of the study by high-pressure liquid chromatography (HPLC ). Plasma lipid levels (triglyceride and cholesterol ) were also measured by standard enzymatic methods. To confirm compliance, probucol was measured by HPLC, although the results were not made available to investigators until the study was complete. Statistical analysis Pre- and post-treatment values for each variable were compared within groups using Student’s paired t-test, to assess the effect of therapy. Comparisons between groups employed Student’s unpaired t-test. Probability values are given for each test and P < 0.05 was taken as statistically significant.

Results Overall, there was a statistically significant improvement in Raynaud’s symptoms in patients treated with probucol, but not in the control group. Analysis of treatment effect was on an intention-to-treat basis, comparing values at baseline with those after the 12 week treatment period for each variable. These data are given in Table 2, showing mean change from baseline and 95% confidence intervals, together with mean percentage change from basal during the study. There was a mean reduction

311

of 32% from baseline in Raynaud’s severity and of 20% in the frequency of attacks over the treatment period for patients receiving probucol (Fig. 1), but no significant change for the nifedipine treatment group. Improvement in symptoms on probucol was accompanied by a significant increase in lag time for lipoprotein oxidation (mean 164% of pre-treatment time) and was also associated with a modest, though highly significant, reduction in serum cholesterol (Fig. 2). Furthermore, unpaired comparison of mean change for each treatment group confirmed statistically significant differences for oxidation lag time (P = 0.05) and change in cholesterol level (P = 0.01), although comparison of the mean change in Raynaud’s attack frequency or severity did not reach statistical significance (P = 0.25 and P = 0.09, respectively). Levels of vitamins E and C did not change significantly in either treatment arm. Baseline variables were compared between patients with SSc-associated Raynaud’s and those without clinical features of a defined connective tissue disease (primary or autoimmune Raynaud’s). These data are summarized in Table 3. There were some statistically significant differences between these two groups, in keeping with previous observations. Thus, the patients with SSc had significantly greater LDL oxidation lag time and lower levels of vitamin C. There was also a significantly greater triglyceride level in the SSc group, although levels in both groups were within the normal reference range. Mean Raynaud’s attack severity and frequency were somewhat higher in the SSc patients than in those without SSc, but this difference was not statistically significant. When the SSc and non-SSc patients were separated for subgroup analysis of treatment effect, the changes noted above were less marked, although the same trends were observed for Raynaud’s attack severity and frequency ( Table 4). Mean changes were generally greater in the SSc group. Both probucol and nifedipine were well tolerated. Side-effects of nifedipine were consistent with previous reports, and included headache, dizziness and ankle

T 2. Effect of probucol or nifedipine therapy on clinical and biochemical variables 95% CI Mean (..) pre-Rx

Mean (..) post-Rx

Pre-post mean

End point

Treatment

RP severity (0–10) RP frequency (attacks/ day) LDL lag time (s) Cholesterol (m) Triglyceride (m) Vitamin E ( m) Vitamin C ( m)

Nifedipine Probucol Nifedipine Probucol

4.9 6.8 3.3 3.6

(2.3) (2.2) (3.2) (2.2)

5.1 4.6 3.5 2.9

(2.9) (3.0) (3.5) (2.7)

−0.27 2.2 −0.27 0.73

Nifedipine Probucol Nifedipine Probucol Nifedipine Probucol Nifedipine Probucol Nifedipine Probucol

165 160 5.1 5.4 1.4 1.8 16.8 16.9 35 26

(83) (90) (1.2) (1.4) (0.8) (0.8) (5.6) (7.9) (31) (19)

154 262 5 4.8 1.3 1.9 19.3 16.5 38 31

(107) (143) (1.2) (1.3) (0.7) (0.9) (3.6) (7.5) (31) (23)

10.3 −102.2 0.12 0.58 0.07 −0.1 −2.5 0.3 5.1 6.5

*P ∏ 0.05, Student’s paired t-test.

Upper

t-test (P)

−2.44 0.2 −1.98 0.12

1.9 4.2 1.45 1.34

0.79 0.03* 0.74 0.02*

−66.5 −190.2 −0.07 0.28 −0.16 −0.57 −5.9 −1.5 −9.8 −4.3

87 −14 0.32 0.87 0.3 0.37 0.9 2.2 20 17.2

Lower

0.78 0.03* 0.2 0.001* 0.55 0.66 0.13 0.71 0.48 0.22

312

C. P. Denton et al.

F. 1. Improvement in Raynaud’s attack severity and frequency following probucol treatment. Mean Raynaud’s attack severity and frequency were determined from the self-reported diaries. Baseline values were compared with those recorded during the final 2 weeks of probucol treatment (500 mg/day) using Student’s paired t-test analysis (n = 17).

F. 2. Probucol treatment increases LDL oxidation lag time and reduces serum cholesterol levels. Oxidation lag times for LDL from patients taken before and after 12 weeks of probucol administration are shown. Lag time was determined spectrophotometrically and was significantly greater after treatment with probucol. There was also a modest, but statistically significant, reduction in total serum cholesterol in these patients. Analysis by Student’s paired t-test (n = 17).

swelling. These were severe enough to require cessation of treatment in three cases. Probucol caused headache and nausea in four patients, and two of these individuals stopped treatment before completion of the 12 week study period. There were no major adverse events in either study arm.

Discussion Clinical trials are recognized to be difficult in Raynaud’s phenomenon and SSc [18]. This reflects the heterogen-

eity of these conditions, with the likelihood that different subsets of patients may respond differently to particular therapies, and is compounded by difficulties in objective serial assessment of clinical severity. Probably the most widely used assessment technique for Raynaud’s phenomenon is the self-reported symptom diary. However, its reliability depends on good and consistent patient education, preferably with blinding of the patient and the assessor in interventional studies. Formal blinding was not possible in this study due to the different formulations of the two drugs. It is well recognized that


T 3. Comparison of baseline data for SSc and non-SSc Raynaud’s patients Non-SSc RP

SSc RP

Variable

Mean

..

Mean

..

t-test P

RP severity RP frequency Lag time (s) Vitamin E ( m) Vitamin C ( m) Cholesterol (m) Triglyceride (m)

5.6 2.6 193 15.7 47 4.9 1.3

2.3 1.6 93.2 3.8 25 1.2 0.8

6.1 4.2 128.8 18 25 5.6 1.9

2.7 3.2 63.9 5.3 31.8 1.3 0.7

0.52 0.10 0.03* 0.16 0.04* 0.20 0.03*

Non-SSc RP, primary or ‘autoimmune’ Raynaud’s phenomenon; SSc RP, SSc-associated Raynaud’s phenomenon. *P < 0.05, Student’s unpaired t-test.

there is often a substantial placebo response in Raynaud’s therapies and this was minimized by using a parallel-group controlled-trial design with two active treatment arms. Though statistically significant, the clinical improvement for probucol-treated patients was modest, and it may be advisable to blind future trials formally to reduce potential bias further. Seasonal influences are also a possible problem in Raynaud’s trials, since most patients experience more frequent and severe attacks during the winter. To minimize this effect, our study was undertaken over a single winter period. Although this was a small trial with a relatively short treatment duration, our findings support the hypothesis that systemic administration of a potent synthetic antioxidant could be beneficial in Raynaud’s phenomenon. The efficacy of probucol as an antioxidant is illustrated by the increase in oxidation lag time of lipoprotein fractions from patients after treatment. It is interesting that nifedipine had no demonstrable effects on Raynaud’s symptoms since this agent is widely used for this condition [19]. One explanation could be the low dose used in this study, selected to minimize side-effects which might have increased the withdrawal rate. Also, the mean severity and frequency of Raynaud’s attacks were somewhat greater at baseline for the probucol-

313

treated group. This was not statistically significant, but it is possible that the patients with more symptoms might have responded better to therapy. A third potential explanation is that patients who were refractory to oral vasodilators were more likely to volunteer for our study, although we have no information that this was the case. Evidence supporting a role for oxidant stress in the aetiology and pathogenesis of Raynaud’s phenomenon and SSc is increasing. Tissue hypoperfusion associated with peripheral vasospasm results in a hypoxic environment, probably leading to the generation of free radicals and other reactive oxygen species. This may have a number of consequences pertinent to the development of SSc, and has been proposed as a potential mechanistic link between Raynaud’s phenomenon and SSc [20]. The role of reactive oxygen species in the generation of Raynaud’s symptoms is not directly addressed in our study, but such a link is plausible and our results provide indirect support for this. However, it is important to remember that lipoprotein oxidation in vitro in a Cu2+ oxidizing environment is a somewhat artificial system. It is also likely that some probucol is retained within lipoproteins during isolation and so assay results may not directly reflect in vivo antioxidant activity. It is possible that probucol modulates Raynaud’s symptoms in an entirely different way, e.g. through direct protection from endothelial cell damage. Such a mechanism could also explain its beneficial effects in macrovascular disease, as discussed below. It would be interesting to address this possibility in future studies examining changes in the levels of circulating markers of endothelial cell perturbation during probucol therapy. Markers of endothelial cell perturbation, such as von Willebrand Factor, E-selectin, endothelin-1 and thrombomodulin, have been measured in several crosssectional [21] and longitudinal studies [22, 23] in Raynaud’s phenomenon and SSc, and correlation with clinical severity has been reported. However, interpretation of these markers is complicated by the heterogen-

T 4. Effect of probucol treatment study end points for SSc and non-SSc patients End point

Diagnosis

Mean pre-Rx

..

Pre-post mean

..

t-test (P)

RP severity (0–10) RP frequency (attacks/day) LDL lag time (s) Cholesterol (m) Triglyceride (m) Vitamin E ( m) Vitamin C ( m)

SSc Non-SSc SSc Non-SSc SSc Non-SSc SSc Non-SSc SSc Non-SSc SSc Non-SSc SSc Non-SSc

7.79 5.75 3.43 3.75 127 189 5.95 5 2.13 1.49 18.4 15.5 20.4 54.9

1.78 1.83 1.72 2.76 67 103 1.14 1.42 0.72 0.84 3.8 3.2 28.7 22.9

2.85 1.5 0.85 1 −195 −24 0.94 0.29 −0.26 0.02 0.1 0.5 −0.9 13.9

3.97 3.29 0.89 0.63 144 166 0.56 0.34 1.22 0.54 4.2 2.8 6.7 26.5

0.11 0.12 0.05* 0.11 0.003* 0.34 0.002* 0.02* 0.29 0.45 0.47 0.31 0.36 0.11

*P ∏ 0.05, Student’s paired t-test.

314

C. P. Denton et al.

eous nature of these conditions, by differential involvement of various microvascular beds and by the difficulty of distinguishing between the relative contribution of inflammatory, fibrotic or vasospastic disease processes [24]. Recent studies have suggested that certain tissue antigens are susceptible to fragmentation in an oxidative microenvironment, especially in the presence of heavy metal ions which can catalyse these reactions [25]. New epitopes may be expressed which could trigger the development of hallmark antibodies in SSc. However, the role of these antibodies in disease pathogenesis remains uncertain, since they are not universally present. Another potential role for reactive oxygen species in SSc, and perhaps in Raynaud’s phenomenon, is as a mediator of vascular damage. Interestingly, an increased prevalence of large-vessel atherosclerosis has been reported in limited cutaneous SSc [26 ], a clinical subset often associated with severe vasospastic symptoms. Although free radicals acting on endothelium may be directly injurious, it is also possible that endothelial cell damage occurs indirectly through the toxic effect of oxidized LDLs [6 ]. More recently, effects of free radicals on cell growth and metabolism have also been reported, perhaps through modulation of intracellular signalling pathways [27], providing further plausible mechanisms whereby reactive oxygen species could influence disease pathogenesis. Evidence supporting a role for oxidative stress in SSc has been difficult to obtain in vivo, although elevated levels of oxidized products from prostaglandin metabolism have been reported [28]. Probucol was originally developed to treat hyperlipidaemia, but its lipid-lowering profile appears to be less advantageous than that of other newer drugs such as the statins, since it reduces both high-density lipoprotein (HDL) and LDL cholesterol fractions [11]. A number of side-effects have also been documented, including occasional cardiac arrhythmias [29], although the drug was well tolerated in our study. Probucol is no longer recommended for the treatment of hypercholesterolaemia and its use has declined to such an extent that it is not generally available, despite recent clinical trials suggesting that probucol may be beneficial in patients with coronary artery atheromatous disease [30]. Moreover, a study evaluating probucol and/or antioxidant vitamins showed regression of atheroma in the probucol-treated arms [31], but not in those receiving antioxidant vitamins alone. In conclusion, this pilot study confirms the tolerability and feasibility of using probucol in patients with Raynaud’s and SSc. One explanation for efficacy might be through an antioxidant effect and our study provides indirect evidence for this. It is also possible that probucol could have additional beneficial effects in SSc by reducing the availability of reactive oxygen species, by lowering oxidized LDL levels or perhaps by directly protecting endothelial cells from injury. Although our findings are encouraging, interpretation must be cautious in view of the well-recognized difficulties in conducting clinical trials in Raynaud’s and SSc [18], and

further evaluation of probucol is necessary to define its therapeutic potential in these diseases properly.

Acknowledgements The authors are grateful to Nuala Ryan for help in patient screening and follow-up. Financial support was provided by the Arthritis Research Campaign ( UK ) and the Raynaud’s and Scleroderma Association ( UK ).

References 1. Dowd PM, Goldsmith PC, Bull HA, Burnstock G, Foreman JC, Marshall I. Raynaud’s phenomenon. Lancet 1995;346:283–90. 2. Wollersheim H, Thien T, Hoet MH, Van Venrooy WJ. The diagnostic value of several immunological tests for antinuclear antibody in predicting the development of connective tissue disease in patients presenting with Raynaud’s phenomenon. Eur J Clin Invest 1989; 19:535–41. 3. Kahaleh MB, Mattuci-Cerinic M. Raynaud’s phenomenon and scleroderma: dysregulated neuroendothelial control of vascular tone. Arthritis Rheum 1995;38:1–4. 4. Blake DR, Winyard PG, Maok R. The contribution of hypoxia-reperfusion injury to inflammatory synovitis: the influence of reactive oxygen intermediates on the transcriptional control of inflammation. Ann NY Acad Sci 1994;723:308–17. 5. Blann AD, Illingworth K, Jayson MIV. Mechanisms of endothelial cell damage in systemic sclerosis and Raynaud’s phenomenon. J Rheumatol 1993;20:1325–30. 6. Blake DR, Winyard P, Scott DGI, Brailsford S. Endothelial cytotoxicity in inflammatory vascular diseases—the possible role of oxidised lipoproteins. Ann Rheum Dis 1985;44:176–82. 7. Bruckdorfer KR, Hillary JB, Bunce T, Vancheeswaran R, Black CM. Increased susceptibility to oxidation of lowdensity lipoproteins isolated from patients with systemic sclerosis. Arthritis Rheum 1994;38:1060–7. 8. Steinberg D. Modified forms of low density lipoprotein and atherosclerosis. J Intern Med 1993;233:227–32. 9. Thorne SA, Abbot SE, Winyard PG, Blake DR, Mills PG. Extent of oxidative modification of low density lipoprotein determines the degree of cytotoxicity to human coronary artery cells. Heart 1996;75:11–6. 10. Herrick AL, Rieley F, Schofield D, Hollis S, Braganza JM, Jayson MI. Micronutrient antioxidant status in patients with primary Raynaud’s phenomenon and systemic sclerosis. J Rheumatol 1994;21:1477–83. 11. Salel AF, Zelis R, Sodhi HS, Price J, Mason DT. Probucol: a new cholesterol-lowering drug effective in patients with type II hyperlipoproteinaemia. Clin Pharm Ther 1976;20:690–4. 12. Subcommittee for Scleroderma Criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis Rheum 1980;23:581–90. 13. Blockmans D, Beyens G, Verhaeghe R. Predictive value of nailfold capillaroscopy in the diagnosis of connective tissue diseases. Clin Rheum 1996;15:148–53. 14. Meyer DF, Mayans MO, Groot PHE, Suckling KF, Bruckdorfer KR, Perkins SJ. Time course studies of the structure of polydisperse low density lipoprotein by neu-


15. 16. 17. 18.

19. 20. 21.

22.

23.

tron and scattering techniques. Biochem J 1995;110: 417–25. Esterbauer H, Strigl G, Puhl H, Rotheneder M. Continuous monitoring of in vitro oxidation of low density lipoproteins. Free Radical Res Commun 1989;6:67–75. Calzada C, Buckdorfer KR, Rice-Evans CA. The influence of antioxidant nutrients on platelet function in healthy volunteers. Atherosclerosis 1997;128:97–105. Jennings PE, Chirico S, Jones AF, Lunec J, Barnett AH. Vitamin C metabolites and microangiopathy in diabetes mellitus. Diabetes Res 1987;6:151–4. White B, Bauer EA, Goldsmith LA, Hochberg MC, Katz LM, Korn JH et al. Guidelines for clinical trials in systemic sclerosis (scleroderma). I. Disease-modifying interventions. The American College of Rheumatology Committee on Design and Outcomes in Clinical Trials in Systemic Sclerosis. Arthritis Rheum 1995;38:351–60. Belch JJ. Raynaud’s phenomenon. Cardiovasc Res 1997;33:25–30. Murrel DF. A radical proposal for the pathogenesis of scleroderma. J Am Acad Dermatol 1993;28:78–85. Herrick AL, Illingworth K, Blann A, Hay CR, Hollis S, Jayson MI. Von Willebrand factor, thrombomodulin, thromboxane, beta-thromboglobulin and markers of fibrinolysis in primary Raynaud’s phenomenon and systemic sclerosis. Ann Rheum Dis 1996;55:122–7. Denton CP, Bickerstaff MCM, Shi-wen X, Carulli MT, Haskard DO, du Bois RM et al. Serial circulating adhesion molecule levels reflect disease severity in systemic sclerosis. Br J Rheumatol 1995;34:1048–54. Gruschwitz MS, Hornstein OP, von Den Dreisch P. Correlation of soluble adhesion molecules in the peripheral

24.

25.

26. 27. 28.

29.

30.

31.

315

blood of scleroderma patients with their in situ expression and with disease activity. Arthritis Rheum 1995;38:184–9. Vancheeswaran R, Magoulas T, Efrat G, Wheeler-Jones C, Olsen I, Penny R et al. Circulating endothelin-1 levels in systemic sclerosis—a marker of fibrosis or vascular dysfunction? J Rheumatol 1994;21:1838–44. Casciola-Rosen L, Wigley F, Rosen A. Scleroderma autoantigens are uniquely fragmented by metal-catalysed oxidation reactions: implications for pathogenesis. J Exp Med 1997;185:71–9. Veale DJ, Collige TA, Belch JJ. Increased prevalence of symptomatic macrovascular disease in systemic sclerosis. Ann Rheum Dis 1995;54:853–955. Winrow VR, Winyard PG, Morris CJ, Blake DR. Free radicals in inflammation: second messengers and mediators of tissue destruction. Br Med Bull 1993;49:506–22. Stein CM, Tanner SB, Awad JA, Roberts LJ, Morrow JD. Evidence of free radical-mediated injury (isoprostane overproduction) in scleroderma. Arthritis Rheum 1996;39:1146–50. Reinoehl J, Frankovich D, Machado C, Kawasaki R, Baga JJ, Pires LA et al. Probucol-associated tachyarrhythmic events and QT prolongation: importance of gender. Am Heart J 1996;131:1184–91. Rodes J, Cote G, Lesperance J, Bourassa MG, Doucet S, Bilodeau L et al. Prevention of restenosis after angioplasty in small coronary arteries with probucol. Circulation 1998;97:429–36. Tardif JC, Cote G, Lesperance J, Bourassa M, Lambert J, Doucet S et al. Probucol and multivitamins in the prevention of restenosis after coronary angioplasty. Multivitamins and Probucol Study Group. N Engl J Med 1997;337:365–72.